Masterclass in Neuroendocrinology 9

Susan Wray
Seth Blackshaw Editors

Developmental
Neuroendocrinology
Masterclass in Neuroendocrinology
Volume 9

Series Editors
John A. Russell
Centre for Discovery Brain Sciences, Edinburgh University Medical School,
CMVM, Edinburgh, UK
William E. Armstrong
Department of Anatomy and Neurobiology, The University of Tennessee Health
Science Center, Memphis, TN, USA
Masterclass in Neuroendocrinology is a co-publication of the INF (International
Neuroendocrine Federation) that aims to illustrate the highest standards and promote
the use of the latest technologies in basic and clinical research, while also providing
inspiration for further exploration into the exciting field of neuroendocrinology. It is
intended for established researchers, trainees and students alike.
Each book
- is edited by leading experts in the field
- is written by a team of internationally respected researchers
- includes assessments of different experimental approaches, both in vivo and in
vitro, and of how the resulting data are interpreted.

Founding Series Co-Editors: William E. Armstrong and John A. Russell

More information about this series at http://www.springer.com/series/15770

Susan Wray • Seth Blackshaw
Editors

Developmental
Neuroendocrinology
Editors
Susan Wray Seth Blackshaw
Cellular and Developmental Neurobiology Department of Neuroscience
NINDS, NIH School of Medicine, Johns Hopkins University
Bethesda, USA Baltimore, MD, USA

ISSN 2662-2068 ISSN 2662-2076 (electronic)

Masterclass in Neuroendocrinology
ISBN 978-3-030-40001-9 ISBN 978-3-030-40002-6 (eBook)
https://doi.org/10.1007/978-3-030-40002-6

# Springer Nature Switzerland AG 2020

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the
material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, expressed or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

Illustration by deblik, based on an image from # Nablys/stock.adobe.com

This Springer imprint is published by the registered company Springer Nature Switzerland AG.
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Series Preface

This series began publication as a joint venture between the International Neuroen-
docrine Federation and Wiley‐Blackwell and now is continuing with Springer-
Nature as publisher for the federation. The broad aim of the series is to provide
established researchers, trainees, and students with authoritative up‐to‐date accounts
of the present state of knowledge and prospects for the future across a range of topics
in the burgeoning field of neuroendocrinology. The series is aimed at a wide
audience as neuroendocrinology integrates neuroscience and endocrinology. We
define neuroendocrinology as the study of the control of endocrine function by the
brain and the actions of hormones on the brain. It encompasses the study of normal
and abnormal function and the developmental origins of disease. It includes the
study of the neural networks in the brain that regulate and form neuroendocrine
systems. It also includes the study of behaviors and mental states that are influenced
or regulated by hormones. It necessarily includes the understanding and study of
peripheral physiological systems that are regulated by neuroendocrine mechanisms.
Clearly, neuroendocrinology embraces many current issues of concern to human
health and well‐being, but research on these issues necessitates reductionist animal
models.
Contemporary research in neuroendocrinology involves the use of a wide range
of techniques and technologies, from the subcellular to systems at the whole-
organism level. A particular aim of the series is to provide expert advice and
discussion about experimental or study protocols in research in neuroendocrinology
and to further advance the field by giving information and advice about novel
techniques, technologies, and interdisciplinary approaches.
To achieve our aims, each book is on a particular theme in neuroendocrinology,
and for each book we have recruited a pair of editors, expert in the field, and they
have engaged an international team of experts to contribute chapters in their individ-
ual areas of expertise. Their mission was to give an update of knowledge and recent
discoveries and to discuss new approaches, “gold‐standard” protocols, translational
possibilities, and future prospects. Authors were asked to write for a wide audience,
to minimize references, and to consider the use of video clips and explanatory text
boxes; each chapter is peer reviewed and has a Glossary and a detailed Index. We
have been guided by an Advisory Editorial Board.

v
vi Series Preface

The MasterClass Series is open‐ended; books in the series published to date are
Neurophysiology of Neuroendocrine Neurons (2014, ed. WE Armstrong & JG
Tasker); Neuroendocrinology of Stress (2015, ed. JA Russell & MJ Shipston);
Molecular Neuroendocrinology: From Genome to Physiology (2016, ed. D Murphy
& H Gainer); Computational Neuroendocrinology (2016, ed. DJ Macgregor & G
Leng); Neuroendocrinology of Appetite (2016; ed. SL Dickson & JG Mercer); The
GnRH Neuron and its Control (2018; ed. AE Herbison & TM Plant); Model Animals
in Neuroendocrinology (2019, ed. M Ludwig and G Levkowitz); and Neurosecre-
tion: Secretory Mechanisms (In Press, ed. J Lemos and G Dayanithi), #8 in the
Series, and will be the first to be published by Springer Nature; books in preparation
include Glial-Neuronal Signaling in Neuroendocrine Systems and Neuroendocrine
Clocks and Calendars.
Feedback and suggestions are welcome.

Advisory Editorial Board

Ferenc A. Antoni, Egis Pharmaceuticals PLC, Budapest

Tracy Bale, University of Pennsylvania
Rainer Landgraf, Max Planck Institute of Psychiatry, Munich
Gareth Leng, University of Edinburgh
Stafford Lightman, University of Bristol
Andrew Loudon, University of Manchester
Tony Plant, University of Pittsburgh

International Neuroendocrine Federation—http://neuroendonow.com/

University of Edinburgh John A. Russell

Edinburgh, UK
University of Tennessee Health Science Center William E. Armstrong
Memphis, TN, USA
Volume Preface

The hypothalamus is the regulatory hub controlling neuroendocrine activity. An

evolutionarily ancient structure, with multiple peptidergic systems contained within
its boundaries, it is here that many innate behaviors and functions essential for survival
are controlled. These include reproduction, homeostasis, feeding, drinking, growth,
circadian rhythms, and resilience to psychological stress. Within the hypothalamus
there are multiple distinct neuronal subtypes, distinguished by their functions and
molecular expression patterns. These neurons are clustered within anatomical nuclei
defined by classic histological approaches and also dispersed across nuclei. These
neuronal subtypes form both intra-hypothalamic circuits and circuits with brainstem
and other limbic structures and may influence daily behavior through their
hypothalamic–pituitary connections. This volume presents the current state of knowl-
edge for three aspects of developmental neuroendocrinology: Part I—the molecular
specification of hypothalamic cells, Part II—developmental modulators and epigenetic
factors that influence hypothalamic development, and Part III—the development of
neuroendocrine circuits. Chapters are presented as succinct summaries of current
progress and future prospects for the respective areas. The research often includes
state-of-the-art techniques for delineating hypothalamic cell lineage and cell fate
specification, for identifying genes altered by epigenetic factors, and for tracing of
functional neural circuits. This book puts together a current and exciting perspective
on developmental neuroendocrinology—covering topics ranging from genes that
specify and mold the embryonic hypothalamus to molecular mechanisms that sculpt
experience-dependent circuit formation in the postnatal period. Each chapter is written
by one or more experts in the field.
Part I focuses on the molecular mechanisms that control specification of hypotha-
lamic and pituitary cells. Cellular and molecular mechanisms that direct the
integrated assembly and differentiation of the neuroendocrine hypothalamus are
addressed in Chap. 1, while Chap. 2 discusses the hypothalamic Shh-Gli code, a
region-specific set of rules governing the relationship between the morphogen Sonic
hedgehog and its downstream effectors, the Gli family of transcription factors.
Developmental patterns of gene expression in the developing hypothalamus that
are conserved across vertebrates, including humans, are discussed in Chap. 3.
Chapters 4 and 5 highlight the gene regulatory program that controls differentiation
of the arcuate nucleus and the transcriptional network guiding specification and

vii
viii Volume Preface

differentiation of the suprachiasmatic nucleus and sleep-regulating neurons of the

dorsolateral hypothalamus. Chapter 6 focuses on transcription factors necessary for
pituitary organogenesis.
The organizing theme for Part II is the role of environmental and epigenetic
factors in regulating development of hypothalamic nucleogenesis and cell specifica-
tion and survival. Chapter 7 discusses the role of GABA-mediated signaling events
in controlling these processes. Chapter 8 addresses the dependence of mammalian
puberty on the interaction of gene sets organized into functionally connected
networks, controlled by transcriptional and epigenetic switches. In Chap. 9, the
focus is on the role of epigenetic factors in the terminal differentiation of GnRH
neurons. Chapter 10 presents an overview of functions of imprinted genes in the
hypothalamus. Chapter 11 completes this section by discussing short- and long-term
oscillations in DNA methylation and histone acetylation in relation to the control of
rhythmic physiology and behavior.
Part III of the book examines development of neuroendocrine neural circuitry.
Research into the neural mechanisms underlying specific behaviors has a long
history that has been repeatedly invigorated with new tools, exciting data, and
ideas. Specific topics discussed in this section include advances in neural circuitry
tracing, opto- and chemogenetics, and the contribution of immune cells to neuronal
maturation. Chapter 12 discusses the development of stress response-regulatory
circuitry in the limbic system. Chapter 13 discusses the formation of neural circuitry
relaying visceral sensory signals from the brainstem to the hypothalamus.
Chapter 14 discusses the role of astrocytes in the development of neuroendocrine
circuits, with an emphasis on regulation of metabolism. Part III closes with an
overview of sexual differentiation and reproductive functions, with Chap. 15
discussing mechanisms of steroid-mediated sexual differentiation, and Chap. 16
discussing the development and modulation of reproductive function by circadian
signals.
Many different structural, developmental, and functional features are found in the
neuroendocrine brain: these involve structurally defined nuclei, cells that are broadly
distributed by tangential migration, cells that show synchronized physiological
rhythms, and a diverse collection of neural circuits fundamental to survival. We
hope the reader is intrigued by the developmental mechanisms explained here that
create the neuroendocrine hypothalamus and that the findings and ideas discussed
will propel readers to look further into the exciting field of developmental neuroen-
docrinology.

Bethesda, MD Susan Wray

Baltimore, MD Seth Blackshaw
Contents

Part I Molecular Specification of Hypothalamic/Pituitary Cells

1 Development of the Neuroendocrine Hypothalamus . . . . . . . . . . . . 3
Marysia Placzek, Travis Fu, and Matthew Towers
2 Sonic hedgehog in Hypothalamus Development . . . . . . . . . . . . . . . . 31
Gonzalo Alvarez-Bolado
3 Development of the Hypothalamus in Xenopus laevis . . . . . . . . . . . 67
Nerea Moreno and Agustín González
4 Gene Regulatory Programs in the Development of Hypothalamic
Arcuate Nucleus Neurons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Jae W. Lee, Christian Huisman, and Seunghee Lee
5 Winding the Clock: Development of Hypothalamic Structures
Controlling Biological Timing and Sleep . . . . . . . . . . . . . . . . . . . . . 105
Dong Won Thomas Kim and Seth Blackshaw
6 Pituitary Development and Organogenesis: Transcription Factors
in Development and Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
Alexandre Z. Daly and Sally A. Camper

Part II Developmental Modulators and Epigenetic Factors

7 Hypothalamic Development: Role of GABA . . . . . . . . . . . . . . . . . . 181
M. Stratton and S. Tobet
8 Epigenetic and Transcriptional Regulation of the Reproductive
Hypothalamus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
Carlos Francisco Aylwin and Alejandro Lomniczi
9 Epigenetic Regulation of the GnRH and Kiss1 Genes:
Developmental Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
Joseph R. Kurian and Ei Terasawa
10 Imprinted Genes and Hypothalamic Function . . . . . . . . . . . . . . . . . 265
Michela Pulix and Antonius Plagge

ix
x Contents

11 Rhythmic Epigenetics in Neuroendocrine and Immune Systems . . . 295

Christopher S. Coyle, Elisabetta Tolla, and Tyler J. Stevenson

Part III Development of Neuroendocrine Circuits

12 Development of Limbic System Stress-Threat Circuitry . . . . . . . . . 317
Newton S. Canteras, Dayu Lin, and Joshua G. Corbin
13 Organization and Postnatal Development of Visceral Sensory
Inputs to the Neuroendocrine Hypothalamus . . . . . . . . . . . . . . . . . 345
Linda Rinaman
14 Astrocytes and Development of Neuroendocrine Circuits . . . . . . . . 367
Lydia L. DonCarlos and Julie A. Chowen
15 Origins of Sex Differentiation of Brain and Behavior . . . . . . . . . . . 393
Margaret M. McCarthy
16 Development and Modulation of Female Reproductive Function
by Circadian Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413
Neta Gotlieb, Jacob Moeller, and Lance J. Kriegsfeld

Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 447
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 459
 Part I
Molecular Specification of Hypothalamic/
Pituitary Cells
Development of the Neuroendocrine
Hypothalamus 1
Marysia Placzek, Travis Fu, and Matthew Towers

Abstract
The neuroendocrine hypothalamus consists of neurosecretory neurons and
specialized glia of the median eminence and posterior pituitary. Neurons and
glia act together to support and regulate reciprocal brain–body communications
that enable the hypothalamus to anticipate and adapt to changing physiological
conditions. Here we summarize recent studies in model organisms that show how
an embryonic program of hypothalamic development unfolds in a manner that
underlies adult function. Fate-mapping studies in chick show that much of the
basal hypothalamus is built from a multipotent Fgf10(+) progenitor population
whose mode of growth in four dimensions suggests a new model for hypotha-
lamic development that we term the ‘Anisotropic growth’ model. Conditional
genetic analyses and pharmacological interventions in mouse, zebrafish, and
chick suggest that a conserved molecular mechanism may mediate anisotropic
growth and establish the basal hypothalamus. A handful of conserved signalling
factors and transcription factors (TFs) direct the progressive development of
Fgf10(+) progenitors to hypothalamic progenitors, hypothalamic neurons, and
then infundibular progenitors of the median eminence and posterior pituitary in a
program where space and time are intrinsically linked. These studies reveal
additionally that Fgf10(+) cell populations are maintained throughout life and
are vital to the long-term building, maintenance, and function of the adult
hypothalamo–pituitary axis. We speculate on whether the embryonic Fgf10(+)
progenitor population harbours a hypothalamic stem cell and discuss the
implications for the ability of the hypothalamus to adapt to changing physiologi-
cal conditions on different time scales through life.

M. Placzek (*) · T. Fu · M. Towers

Bateson Centre and Department of Biomedical Science, University of Sheffield, Sheffield, UK
e-mail: m.placzek@sheffield.ac.uk

# Springer Nature Switzerland AG 2020 3

S. Wray, S. Blackshaw (eds.), Developmental Neuroendocrinology, Masterclass in
Neuroendocrinology 9, https://doi.org/10.1007/978-3-030-40002-6_1
4 M. Placzek et al.

Keywords
Fgf10 · Shh · Anisotropic growth · Infundibulum · Median eminence · Posterior
pituitary

List of Abbreviations

3V 3rd ventricle
AP Anterior Pituitary (aka Adenohypophysis)
ARC Arcuate nucleus
bHLH Basic helix-loop-helix
BMP Bone morphogenetic protein
CRH Corticotropin releasing hormone
DA Dopamine
DMN Dorsomedial nucleus
Fgf Fibroblast growth factor
GHRH Growth hormone releasing hormone
HD Homeodomain
MB Mammillary Body
ME Median Eminence
OXT Oxytocin
PeVN Periventricular nucleus
PP Posterior pituitary (aka neurohypophysis)
PVN Paraventricular nucleus
RDVM Rostral diencephalic ventral midline
Rx Retina and anterior neural fold homeobox
SCN Suprachiasmatic nucleus
Shh Sonic hedgehog
SON Supraoptic nucleus
SST Somatostatin
TMN Tubero-mammillary nucleus
TRH Thyroid releasing hormone
VMN Ventromedial nucleus
VZ Ventricular zone

1.1 Introduction: Neuroendocrine Cells of the Adult

Hypothalamus

The evolutionarily ancient hypothalamus is crucial to life, governing an individual’s

survival and species’ propagation. Its cells enable the body to respond, anticipate, and
adapt to changing physiological conditions. These include physiological fluctuations
1 Development of the Neuroendocrine Hypothalamus 5

Fig.1.1 The adult neuroendocrine hypothalamus. (a) Schematic side view of adult brain. (b) High-
power view of boxed region in (a) to show hypothalamus and approximate position of hypotha-
lamic nuclei relative to optic stalk (OS). (c, d) Transverse views through planes indicated in (b): at
level of anterior hypothalamus (c) or ME (d) to show relative positions of neuroendocrine nuclei
and glia. Different tanycyte subsets are indicated by colour: β2 green: β1 yellow; α2 purple, α1 blue.
(e) High power view of boxed region in (b) to show approximate positions of ME, PP and AP. In the
ME, parvocellular neuronal endfeet intimately contact tanycytes and portal capillaries. In the PP,
magnocellular neuronal endfeet contact pituicytes and capillaries

over different time scales: for instance, short-term changes in stress/energy balance,
or long-term seasonal or lifecourse changes in energy balance and reproduction.
Anatomically and functionally, the hypothalamus is complex. In much of the
central nervous system, such as the cortex, cerebellum, hindbrain or spinal cord,
functionally similar neurons are arranged in columns or layers. By contrast, the
hypothalamus is composed of nuclei, or areas, arranged in a patchwork around the
third ventricle. Each nucleus/area is located in a stereotyped position in the adult
(Fig. 1.1a, b) and harbours functionally diverse neurons. Neuroendocrine neurons
6 M. Placzek et al.

Table 1.1 Characteristics of neuroendocrine neurons and glia

Project
Cell type Location to Function
AVP PVN and SON PP Release AVP
magnocellular
neuron
OXT PVN and SON PP Release OXT
magnocellular
neuron
SST PVN, PeVN ME Regulate somatotropes
parvocellular and ARC
neuron
TRH PVN ME Regulate thyrotropes
parvocellular
neuron
CRH PVN and VMN ME Regulate corticotropes
parvocellular
neuron
GHRH ARC and PVN ME Regulate gonadotropes
parvocellular
neuron
DA PVN, PeVN, ME Regulate prolactin
parvocellular SON and ARC
neuron
AVP SCN PVN, Regulate PVN/SON
parvocellular SON
neuron
Glial Pituicyte PP PP Fenestrated capillary formation
Glial β1 VZ: Lateral ME ARC Stem/progenitor cells, barrier, trafficking
tanycyte functions
Glial VZ: Central ME ME Stem/progenitor cells, barrier, trafficking
β2 tanycyte functions, neurohormone release regulation
Glial VZ: 3V ARC, Stem/progenitor cells, circuit components,
α2 tanycyte VMN metabolite and hormone regulation
Summary of the position, projection and function of the major neuroendocrine neurons and glia

are a subset of hypothalamic neurons, located in multiple hypothalamic nuclei, and

classically fall into two major categories. Magnocellular neurons are located in the
paraventricular (PVN) and supraoptic (SON) nuclei and project to the posterior lobe
of the pituitary (PP; aka posterior neurohypophysis) to release arginine vasopressin
(AVP), and oxytocin (OXT). Parvocellular neurons are located in the PVN, SON,
periventricular (PeVN), arcuate (ARC), and ventromedial (VMN) nuclei and project
to the median eminence (ME) where they release corticotropin-releasing hormone
(CRH), thyrotropin-releasing hormone (TRH), growth hormone-releasing hormone
(GHRH), dopamine (DA), and somatostatin (SST) into the portal vasculature to
control endocrine cells of the anterior pituitary (Swanson 1992; Fig. 1.1b–d;
Table 1.1; List of Abbreviations). Recent studies show that additional neuronal
1 Development of the Neuroendocrine Hypothalamus 7

subsets may be considered as part of the neuroendocrine system. Subsets of OXT(+)

and AVP(+) neurons project internally within the hypothalamus, and mediate
complex behaviours, including trust, aggression, and eating, or may indirectly
regulate classic neuroendocrine behaviour through a network effect. For instance,
subsets of neurons clustered within the suprachiasmatic nucleus (SCN) synthesise
AVP and project to the SON and the PVN (Trudel and Bourque 2010; Gizowski
et al. 2016).
The ME and PP contribute to the unique three-dimensional architecture of the
hypothalamus (Fig. 1.1b, e). The ME and PP are composed of specialized glia cells,
termed tanycytes and pituicytes (Table 1.1), microglia, the endfeet of neurosecretory
neurons and a vast portal capillary network. Together these components provide
blood–brain interfaces that enable communication between the body and the hypo-
thalamus, supporting the release and transport of neurohormones to their endocrine
targets, and providing the route through which the hypothalamus accesses
circulating factors, including metabolites and hormones (Box 1.1). Despite its
complex nature, the overall organization and 3-dimensional morphology of the
hypothalamus, as well as the nature of its resident cell types have been highly
conserved through evolution (Tessmar-Raible et al. 2007). However, understanding
how the hypothalamus develops in embryogenesis has been challenging.

Box 1.1 Brain–Body Interfaces: Composition and Function of the Adult

ME and PP
The ME and PP are composed of numerous cellular types, including glial cells
(ME tanycytes and PP pituicytes), microglia, the endfeet of neurosecretory
neurons, and a network of capillaries. Collectively these form the two-way
circuits between brain and body that underlie hypothalamic functions. Glial
pituicytes regulate the development of the permeable neurovascular interface
in the PP (Anbalagan et al. 2018). Glial tanycytes regulate the proximity and
access of neuronal endfeet to the portal capillaries, and hence the release and
spread of neurotransmitters/neurohormones (reviewed in Clasadonte and
Prevot 2018). At the same time, tanycytes are an active circuit component in
acute homeostatic feedback systems from body to brain: through their
dynamic barrier properties, hormone-regulating properties, transport
functions, and potential to form parts of the neuroendocrine circuitry they
support, modify, and respond to peripheral signals, including metabolites and
hormones, to regulate body–brain communication and short-term/acute-
fluctuating homeostatic mechanisms (reviewed in Rizotti and Lovell-Badge
2017; Lewis and Ebling 2017; Clasadonte and Prevot 2018). Different
tanycyte subsets occupy different regions around the ME and 3V (Fig. 1.1d),
project to particular hypothalamic regions (Table 1.1) and show particular
properties (Table 1.1), including stem/progenitor behaviours (see text).
8 M. Placzek et al.

1.2 Models of Hypothalamic Development

The majority of hypothalamic neuroendocrine neurons and glial originate from

progenitor cells located in ventral parts of the developing forebrain (the exception
being GnRH neurons that migrate from the olfactory placode). Pinpointing the
location of hypothalamic progenitor cells in the early embryo is important. It
provides understanding of the tissues and signals that orchestrate the induction and
specification of distinct hypothalamic cell subsets, potentially an understanding that
is key for future regenerative therapies; it provides insight into how particular
hypothalamic cells develop in concert, important for understanding hypothalamic
circuits and networks; it suggests whether the hypothalamus harbours a stem cell that
is capable of dividing and differentiating over life, able to restore or add cells in the
adult brain; it provides insight into how particular cell subtypes of the hypothalamus
might have evolved. However, understanding the location of progenitor cells, and
how progenitor cells differentiate into hypothalamic neurons and glia has been
challenging: the paucity of markers, the small size and complex morphology of
the hypothalamus, together with the fact that neurons defined on the basis of
neurotransmitter expression are located in multiple nuclei (Table 1.1) have all
hampered its study. Nonetheless, previous studies have suggested two major models
of hypothalamic development: the columnar model (Swanson 1992) and the
(revised) prosomeric model (Puelles et al. 2012, 2013), summarized in Box 1.2.
Each model has moved forward our understanding of this important brain region, but
each has limitations. First, each assumes that adult hypothalamic neuroendocrine
neurons are formed close to sites at which their progenitor cells are specified
(as occurs in many other regions of the CNS: Placzek and Briscoe 2018). Second,
each assumes that hypothalamic progenitor cell domains expand to an equivalent
extent as development proceeds. Third, each has been developed through studies in
rodents, where, for technical reasons, it is difficult to analyse the earliest stages of
hypothalamic development.
Recent studies in the chick (Fu et al. 2017) suggest a new model for hypothalamic
development. In the embryonic chick, a small Fgf10(+) progenitor cell population
(which we refer to as bHyp (basal hypothalamic progenitor) cells) develops in the
early-forming (10 somite stage) ventral diencephalon. Anteriorly, bHyp cells abut
Foxg1(+) telencephalic progenitor cells, suggesting that they are nested within other
diencephalic cell populations (Fig. 1.2a, b). Targeted DiI/DiO fate-mapping studies
show that bHyp cells proliferate and give rise to other progenitor subtypes that
downregulate Fgf10 and extend throughout the basal hypothalamus, from the optic
vesicle to the mammillary pouch (Fig. 1.2c; Fu et al. 2017). Growth from bHyp cells
is anisotropic, and is greater than growth from adjacent diencephalic progenitor
populations. Intriguingly, bHyp cells give rise to two major progenitor subsets,
which grow tangentially and sequentially in opposite directions. First, bHyp cells
generate ‘anterior’ progenitor cells that grow anteriorly (Fig. 1.3 orange), then bHyp
cells generate mammillary progenitors that grow posteriorly (Fig. 1.3 blue).
Throughout the generation of anterior and mammillary progenitors, population
(s) of Fgf10(+) progenitor cells are centrally retained (Fig. 1.3 red). The extensive
1 Development of the Neuroendocrine Hypothalamus 9

Fig. 1.2 Position and growth of chick bHyp cells. (a, b) Schematics of 10-somite stage chick
embryos: ventral (a) and sagittal (b) views. Fgf10(+) bHyp progenitors (red) are nested within
diencephalic progenitors (yellow: note in mouse these express Foxd1), abut Foxg1(+) telencephalic
progenitor cells (green) anteriorly and lie above the prechordal mesendoderm (pm). Arrows indicate
subsequent bidirectional growth (c) Side view, 25 somite embryo. bHyp progenitors give rise to
descendants along the entire basal hypothalamus (red indicates all bHyp-derived cells: note Fgf10 is
maintained only centrally (between arrows). Arrows indicate bidirectional growth

Fig. 1.3 Three-dimensional sequential anisotropic growth from bHyp cells. Schematic sagittal
views of chick embryo (10–40 somites). (a) bHyp cells (red) abut the telencephalon (green) in the
10-somite embryo. (b) By 12 somites, bHyp cells begin to generate anterior progenitors (orange).
(c) By 25 somites, mammillary progenitors are generated (blue): these extend posteriorly from
bHyp cells that are now central (red). (d) Finally, infundibular glial cells are generated and grow
ventrally (arrows). Dotted circle indicates optic stalk (os). Basal hypothalamus shown relative to
underlying tissues: prechordal mesendoderm (a); developing (b) or definitive (c, d) Rathke’s pouch

tangential growth of anterior and mammillary progenitor cells from bHyp cells
means that the adult neurons to which they give rise, including neuroendocrine
neurons, may ultimately lie distant from the sites of their progenitor cell
specification.
10 M. Placzek et al.

Box 1.2 Models of Hypothalamic Development

(A) The columnar model (Swanson 1992) derives from classical studies in the
rat. It suggests that the hypothalamus is a diencephalic-derived structure
whose neurons/nuclei fall into different areas along the rostro–caudal
axis: preoptic/anterior, tuberal, and mammillary, reflecting distinct
diencephalic progenitor subsets arranged in columns along the anterior–
posterior (future rostro–caudal) axis. In this model, neuroendocrine
neurons of the PVN, PeVN, SON and SCN derive from preoptic/anterior
progenitors whereas those of the Arc derive from tuberal progenitors.
Recent single cell RNA-seq studies of wild-type and mutant mice (Kim
et al. 2019) support a version of the columnar model.
(B) The revised prosomeric model (Puelles et al. 2012, 2013), based on gene
expression profiling and genetic inducible fate-mapping studies in mouse
and chick (Sánchez-Arrones et al. 2009; Alvarez-Bolado et al. 2012)

(continued)
1 Development of the Neuroendocrine Hypothalamus 11

Box 1.2 (continued)

suggests that hypothalamic progenitors arise from a common diencephalic–
telencephalic unit termed the secondary prosencephalon. This model
suggests that neurons/nuclei differentiate from progenitors located anterior–
dorsal, or within/posterior–ventral to a diagonal domain of Shh-expressing
cells, and hence are either anterior–dorsal ‘alar progenitors’ or posterior–
ventral ‘basal progenitors’. In this model, neuroendocrine neurons of the
PVN, PeVN, SON and SCN derived from alar progenitors that do not
express Shh whereas neuroendocrine neurons of the Arc and VMN derive
from basal progenitors that expressed Shh (Alvarez-Bolado et al. 2012).
However, in most regions of the CNS, alar progenitors are dorsally located,
so the idea that in the hypothalamus, anterior-most ‘alar’ progenitors
occupy a basal position is controversial.
(C) The anisotropic growth model, described in the text, is based on gene
expression profiling and fate-mapping studies in the early chick embryo
(Fu et al. 2017) and suggests that the basal hypothalamus, at least, is a
diencephalic-derived structure. This model suggests that the basal hypo-
thalamus derives from progenitors that express both Fgf10 and Shh and
abut telencephalic progenitors. In this model, adult neuroendocrine
neurons/nuclei of the PVN/PeVN/SON/SCN may be of mixed progenitor
origin, due to the extensive tangential migration of progenitor cells or
differentiating cells. The figure shows one example, in which alar-derived
neurons migrate ventrally after their specification (yellow arrows:
see text).

Schematics show the location of different domains (ANT, anterior; TUB,

tuberal; MAM, mammillary; alar; basal) and key hypothalamic nuclei. In B
and C, these are shown relative to Shh-expressing cells (shown in grey).

1.3 Mechanisms of Anisotropic Growth

Molecular studies begin to reveal how chick bHyp cells develop and give rise to
anterior and mammillary basal progenitors. These studies, and studies that suggest
the conservation of anisotropic growth mechanisms in other species, have been
described in detail in a recent review (Fu et al. 2019) and here we provide only a
summary.
12 M. Placzek et al.

Fig. 1.4 Successive stages in bHyp cell induction and anterior progenitor growth and differentia-
tion from bHyp cells. Schematic sagittal views of chick embryos (8–17 somites). (a) Induction of
Shh(+) RDVM cells in the 8 somite embryo. (b) Establishment of dorsoventral pattern through a
Shh morphogen gradient: inset shows patterned progenitor domains. (c) Differentiation of Shh(+)
RDVM cells to bHyp cells that co-express Shh, BMP7 and Fgf10 (red area). (d) Resolution of bHyp
into anterior Fgf10/Shh(+) cells (polka dot) and posterior Fgf10/BMP7(+) (red) domains and onset
of growth (depicted by curved arrow). Shh(+) neuroepithelial progenitors (hatched) that
downregulate Fgf10 grow anteriorly from Fgf10/Shh(+) cells. Note although shown in sagittal
view, Shh(+) neuroepithelial progenitors form in an arc around Fgf10(+) cells. (e) Continued
generation and differentiation of anterior progenitors (orange: from Fu et al. 2017)

1.3.1 bHyp Cell Induction

bHyp cells develop from a set of Shh(+) cells, termed rostral diencephalic ventral
midline (RDVM) cells that are induced in the ventral midline of the developing
forebrain (prior to the 10-somite stage in chick). Induction of RDVM cells is
mediated by the signalling ligands Shh and Nodal (reviewed in Placzek and Briscoe
2005), deriving from the underlying prechordal mesendoderm (Fig. 1.4a). At their
anterior limit, RDVM cells abut telencephalic progenitor cells (Fig. 1.4a). Tissue
1 Development of the Neuroendocrine Hypothalamus 13

explant studies in chick suggest that Shh diffuses out of RDVM cells and forms a
spatial morphogen gradient in neighbouring diencephalic neuroepithelial cells
(Fig. 1.4b). This results in the establishment of a dorsoventral pattern. Progenitors
that express the homeodomain transcription factor (TF) Nkx2.2 abut Shh(+) RDVM
cells, and those expressing the homeodomain TF Pax6 are established adjacent to the
Nkx2.2(+) domain (Fig. 1.4b inset). The induction of RDVM cells and their role in
establishing an early dorsoventral pattern appear to be evolutionarily conserved.
Genetic lineage-tracing studies in the mouse (Newman et al. 2018b) that build on
ground-breaking large-scale transcriptomic analyses (Shimogori et al. 2010) indicate
that the hypothalamus is of diencephalic origin. Knockout studies, or analysis of
mutated genes, indicate that the prechordal mesendoderm induces Shh(+) RDVM
cells in mouse, zebrafish, and humans, elegant work in mouse showing that Shh
expression in RDVM cells is driven through a unique enhancer, SBE2 (Shh brain
enhancer 2) (Jeong et al. 2006). Genetic studies in mouse and zebrafish show that
Shh that derives from RDVM cells establishes Nkx2.2(+) and Pax6(+) progenitor
domains. If Shh is selectively deleted in RDVM cells (ShhΔhyp mice; Zhao et al.
2012), the Shh-target gene Gli1 is absent, Nkx2.2 is reduced and Pax6 expands
ventrally (Corman et al. 2018). Cross-repressive TF interactions may then sustain
pattern: in zebrafish a ventral expansion of pax6 is detected after elimination/
reduction of nkx2 genes (Manoli and Driever 2014).
As homeodomain progenitor domains become established adjacent to RDVM
cells the prechordal mesendoderm changes its properties, downregulating Shh and
upregulating BMP7 (Fig. 1.4b). Tissue explant studies in chick suggest that BMP7
from the prechordal tissue has a major influence on overlying RDVM cells, causing
them to upregulate BMP7 and then Fgf10, i.e. to become bHyp cells (Fig. 1.4c;
Manning et al. 2006). Upregulation of BMP7/Fgf10 causes the cells to dramatically
change their behaviour, switching from a quiescent to a proliferative state (Manning
et al. 2006; Fu et al. 2017).

1.3.2 Basal Anterior Progenitor Growth

In the embryonic chick, proliferation is accompanied by a resolution of bHyp cells

into two Fgf10(+) progenitor subtypes: a posterior population that expresses Fgf10
and BMP7 (Fig. 1.4d, red) and an anterior population that expresses Fgf10 and Shh
(Fig. 1.4d, polka-dots). A subset of Fgf10+ Shh+ cells differentiate to progenitor
cells that downregulate Fgf10, but retain/upregulate Shh and are displaced/migrate
anteriorly (Fig. 1.4d, hatched). We term such cells, that derive from bHyp cells,
‘neuroepithelial Shh(+) progenitors’. Neuroepithelial Shh(+) progenitors, in turn,
give rise to wider sets of ‘anterior’ progenitors (Fig. 1.4d orange) an event that
appears to be initiated as neuroepithelial Shh(+) progenitors become displaced/
migrate further anteriorly and upregulate the cyclin-dependent kinase inhibitor,
p57 (Fu et al. 2017). Upregulation of p57 is accompanied by upregulation of
components of the Notch pathway (Ratie et al. 2013, 2014), likely to initiate a
neurogenic differentiation program(s). As p57 and Notch components are
14 M. Placzek et al.

upregulated, Shh is downregulated through an unknown mechanism, so that anterior

progenitors express Shh only transiently. The constant generation and differentiation
of neuroepithelial Shh(+) progenitors into anterior progenitors creates a temporal
dimension that provides the opportunity to build complex arrays of basal hypotha-
lamic neurons. This model predicts that basal hypothalamic anterior progenitor cell
fate and position reflect the amount of time spent as an Fgf10(+) progenitor cell: cells
that spend the shortest time within the Fgf10(+) domain give rise to cells that lie
furthest from it.
Genetic and pharmacological studies suggest that the mechanisms that mediate
the transition from quiescent RDVM to proliferating bHyp cells, then trigger the
resolution of bHyp cells and anterior progenitor growth are evolutionarily
conserved, and present in mouse as well as chick. In both species, dynamic inter-
regulatory interactions between Shh, BMP and Fgf, and the precise balance of these
three signalling ligands in time and space coordinates bHyp induction, resolution
and the onset of proliferation. The transcriptional repressors, Tbx2 and/or Tbx3, play
a critical role in these events (Manning et al. 2006; Trowe et al. 2013: details in Fu
et al. 2019). Wnt signalling, deriving from telencephalic progenitor cells appears to
restrict the bHyp domain (Fig. 1.4b, c) and studies in mouse, zebrafish and chick all
indicate that Wnt signalling levels must be carefully regulated to establish the basal
hypothalamic progenitor territory (Kapsimali et al. 2004; Brinkmeier et al. 2007;
Potok et al. 2008; Manning et al. 2006; Newman et al. 2018b; Kim et al. 2019),
although as yet, little is understood of how Wnt signalling integrates with Shh, BMP
and Fgf signalling to mediate bHyp induction, resolution and proliferation.
Conditional genetic analyses and pharmacological interventions in mouse,
zebrafish and chick indicate that the early anterior basal hypothalamus may be
built from/have a contribution from neuroepithelial Shh(+) progenitors in all of
these species. Feedforward–forward-back regulatory interactions between Shh and
the paired-box transcription factor Rx (or its zebrafish homologue, rx3) appear to
establish a growth loop that provides a dynamic stream of neuroepithelial Shh(+)
progenitors, while maintaining population(s) of Fgf10(+) progenitors (Table 1.2;
Fig. 1.5; Muthu et al. 2016; Orquera et al. 2016; Fu et al. 2017). Key to this growth
loop is the ability of Shh to non-autonomously induce Rx, for Rx to autonomously
induce Shh, and for Shh to autonomously downregulate Rx in order for cells to
realize the anterior progenitor program. These studies begin to suggest that Shh,
deriving from neuroepithelial Shh(+) progenitors, may drive the continued growth
and differentiation of anterior basal hypothalamic progenitors, although this remains
to be formally tested in the mouse. To date, caudal-most Shh(+) RDVM cells and
neuroepithelial Shh(+) progenitors have been lineage traced (Szabó et al. 2009;
Alvarez-Bolado et al. 2012). However, these studies are unlikely to report the fate
of the first-generated neuroepithelial Shh(+) progenitors (i.e. those predicted to give
rise to anterior basal progenitors): the time lag in recombination means that anterior
progenitor cells are likely to have been generated before reporter expression is
initiated.
In summary, Shh has three distinct roles in hypothalamic development, at different
times in development and from different sources. First, it derives from prechordal
1 Development of the Neuroendocrine Hypothalamus 15

Table 1.2 Shh-Rx orchestrate growth and differentiation of the anterior basal hypothalamus
Species Experiment Result References
Zebrafish Rx3 (chokh) Loss of Shh+pea3- progenitors; Tessmar-Raible et al.
mutant or rx3 expansion of basal pea3+ domain; (2007); Dickmeis et al.
morpholino expansion of alar pax6+ domain; loss (2007); Muthu et al.
of anterior and tuberal neurons, (2016)
including pomc, sf1/ff1b, avp, Otp
and TH+ neurons
Mouse Rx2 Loss/reduction of anterior territory Orquera et al. (2016)
conditional (Lhx1, Nkx2.2); loss of anterior and
ablation tuberal neurons, including pomc, Otp,
TH and Sst+ neurons
Chick Timed Reduced Six3+p57+ anterior Fu et al. (2017)
cyclopamine progenitor territory; reduced tract of
(HH9–12) post-optic commissure
Zebrafish Timed Loss of anterior and tuberal neurons Muthu et al. (2016)
cyclopamine including pomc and sf1/ff1b
(30hpf)
Mouse Shh deletion Loss of tuberal and anterior markers Shimogori et al. (2010)
~E10.5 (Lhx1,Lhx6, Lhx9) and neurons
(Nkx2.1-Cre (pomc, sf1)
driver)
Mouse Shh deletion Loss/reduction of anterior markers Zhao et al. (2012);
~9.0 (SBE2- (Six6, Nkx2.1, Nkx2.2); expansion of Corman et al. (2018)
Cre driver) Pax6; loss of tuberal neurons (Hmx3,
Sf1, Pomc, TH, Sst) and reduced
mitotic progenitors
Mouse Shh deletion Loss of anterior territory (Nkx2,1, Carreno et al. (2017)
~E10 (Hesx1- Wnt5a, Tcf4); loss of anterior and
Cre driver) tuberal neurons (GHRH, Pomc, Sst,
Oxt and Avp)
Mouse Shh deletion Reduced Foxd1, Nkx2.1; reduces Szabó et al. (2009)
~E9.0 (Foxb1- anterior/tuberal neurons: Pomc, Hcrt,
Cre driver) reduced mitotic progenitors
Summary of experiments that knockdown, or conditionally knockdown/reduce Rx/rx3 or Shh in
hypothalamic neuroepithelial cells

mesendoderm and acts with Nodal to induce RDVM cells (Fig. 1.4a). Then it derives
from RDVM cells and acts as a morphogen to establish an early dorsoventral pattern
(Fig. 1.4b). Finally (at least in chick and zebrafish: Muthu et al. 2016; Fu et al. 2017),
it derives from Shh(+) neuroepithelial cells to drive growth of the anterior basal
hypothalamus, potentially by regulating cell cycle. Although not yet proved, it
remains possible that Shh(+) neuroepithelial cells continue to act
non-autonomously to pattern adjacent cells. This raises the possibility that progenitor
cells in the basal hypothalamus are continually generated through two routes: from
cells that are responsive to Shh signalling, but do not express Shh; or directly from
Shh(+) neuroepithelial cells, a notion supported through recent work in mouse
(Corman et al. 2018).
16 M. Placzek et al.

Fig. 1.5 Model for cellular

homeostasis and generation
of anterior progenitors from
bHyp cells. (a) Shh induces
Rx non-autonomously; Rx
induces Shh autonomously;
Shh represses Rx
autonomously. (b) Over time,
this loop builds a steady
stream of anterior
neuroepithelial Shh(+)
progenitors (polka dots) while
maintaining Fgf10(+)
progenitors (red). Anterior
neuroepithelial Shh(+)
progenitors upregulate p57
(hatched), promoting anterior
progenitor differentiation
(orange), including the
downregulation of Shh. A
gradient of proliferation and
differentiation is detected in
the developing hypothalamus,
emanating from bHyp
progenitors (Fu et al. 2017)

1.3.3 Mammillary Progenitor Growth

Throughout the generation of anterior progenitors, a pool of Fgf10(+) progenitors is

maintained (Fig. 1.3a, b). The mechanism that selects a steady supply of
neuroepithelial Shh(+) progenitors simultaneously ensures a stable-size pool of
Fgf10(+) progenitors (Fig. 1.5). Thus, genetic and pharmacological interventions
suggest that Shh, deriving from neuroepithelial Shh(+) progenitors, feeds-back to
regulate the size of the Fgf10(+) pool (Fu et al. 2017; Carreno et al. 2017),
potentially acting in concert with Wnt signalling (Brinkmeier et al. 2007; Potok
et al. 2008; Newman et al. 2018b). This concept is not new: in other parts of the
brain, differentiating cells feedback to progenitor cells to maintain their appropriate
numbers and behaviours.
The maintenance of pools of undifferentiated Fgf10(+) progenitor cells is impor-
tant (Fig. 1.3 red area) because after generating anterior progenitors (Fig. 1.3 orange
area), Fgf10(+) progenitors generate mammillary neurogenic progenitors that grow
posteriorly (Fig. 1.3 blue area; Fu et al. 2017). At present, little is understood of
mammillary progenitor growth, although there may be some commonalities with
anterior progenitor growth (described in Fu et al. 2019). Further, it is not clear what
promotes the switch from anterior to mammillary progenitor generation, nor why
mammillary progenitors grow posteriorly. Regardless of the mechanism, the
1 Development of the Neuroendocrine Hypothalamus 17

extensive growth of anterior then mammillary progenitors obscures earlier pattern-

ing. This means that the simple organization of the hypothalamus along the dorso-
ventral axis, established through the early Shh morphogen gradient, is rapidly eroded
through the subsequent extensive growth of anterior and mammillary progenitor
populations. Progenitor cells that lie close to one another at early stages may
ultimately give rise to neurons that lie far apart. This sheds new light on the different
lineage relationships of hypothalamic neuronal and glial subsets (Sect. 1.4).

1.3.4 Infundibular Growth

After generating anterior and mammillary progenitors, Fgf10(+) progenitor cells

generate the infundibulum, a small, but vital structure (Sect. 1.5) that is composed of
glial progenitors that grow ventrally (Fig. 1.3d, arrows; Pearson et al. 2011; Goto
et al. 2015). The sequential anisotropic growth in 3-dimensions from Fgf10(+)
progenitor cells is peculiar and unprecedented within CNS development.
Loss-of-function studies in chick and mouse point to the vital role of Fgf10 in
infundibular development. If the Fgf10(+) progenitor domain(s) is not maintained
normally, or the intensity of Fgf10 expression is altered, then the infundibulum does
not develop (Fu et al. 2017; Carreno et al. 2017). Likewise, in Fgf10-null mice the
infundibulum fails to form properly and infundibular cells undergo apoptosis
(Ohuchi et al. 2000), and in mutant mice where Fgf10 is downregulated, the
infundibular-derived PP is hypoplastic (Mortensen et al. 2015). It is not known
why Fgf10(+) progenitor cells switch from generating neurogenic anterior/mammil-
lary to gliogenic infundibular progenitor cells, but potentially this neurogenic–
gliogenic switch is triggered by the Notch pathway. Elsewhere in the hypothalamus
Ascl1 represses oligodendrocyte differentiation (Marsters et al. 2016), so potentially,
Ascl1 downregulation likewise triggers a switch from neurogenic to infundibular
glial progenitor differentiation. In support of this idea, in Hes1( / ); Hes5(+/ )
mutant embryos, progenitor cells differentiate into neurons at the expense of
pituicytes (derivatives of the infundibulum) (Goto et al. 2015).
Knockout studies in the mouse and analysis of human variants reveal a number of
TFs that, like Hesx1, are required for infundibular formation, including Nkx2.1,
Tbx3, Sox2 and Hes1/Hes5 (reviewed in Burbridge et al. 2016). Many of these are
likely to affect early steps in the development of bHyp progenitors but conditional
knockout studies reveal TFs that act directly downstream of Fgf signalling, including
Rx, Lhx2 and Sox3 (Zhang et al. 2000; Orquera et al. 2016; Miranda-Angulo et al.
2014; Salvatierra et al. 2014; Zhao et al. 2010; Fig. 1.7 red flow-lines).
Growth of an infundibular-like structure can occur in isolated RDVM/bHyp cell
explants (Pearson et al. 2011), suggesting that the instructions for infundibular
growth are intrinsic to the bHyp population. This idea challenges previous models,
which suggest that the infundibulum forms when an underlying tissue, Rathke’s
pouch (see Sect. 5.1) induces Hesx1 (a TF needed for infundibular development: see
Davis et al 2013; Rizzoti 2015) in the overlying ventral diencephalon. One possible
resolution of these findings is that Rathke’s pouch maintains rather than induces
18 M. Placzek et al.

Fig. 1.6 (a) Development of the hypothalamo–pituitary axis. The adenohypophyseal (anterior
pituitary) placode (black hatched) moves ventrally (curved arrow) beneath bHyp cells (red). (b) The
placode invaginates to form a pouch as anterior neuroepithelial Shh(+) progenitors (red hatched)
form. (c) Lhx3 is upregulated, forming the definitive RP as anterior progenitors (orange) form. (d) A
subset of bHyp cells give rise to progenitors that grow ventrally and form the infundibulum, the
precursor of the median eminence (ME) and posterior pituitary (PP). RP gives rise to the anterior
pituitary (AP). Inset shows high power view of these regions. Dotted circle in (c, d) denotes optic
stalk

Hesx1 expression. A second intriguing possibility is that the adenohypophyseal

placode (the precursor to Rathke’s pouch) lies so close to prospective hypothalamic
cells that they interact (Fig. 1.6a hatched; Couly and Le Douarin 1985). Real-time
analyses show cytoneme-like structures that extend between anterior-most midline
cells and prospective hypothalamic cells (Patten et al. 2003), potentially regulating
Wnt signalling between these two populations in the early neural plate (Mattes et al.
2018).
In summary, the bHyp population gives rise, sequentially, to anterior, mammil-
lary and infundibular progenitor cells that grow in different dimensions: anterior,
posterior and ventral, respectively (Fig. 1.3d). Future studies are needed to determine
if bHyp cells are multipotent, contributing to all progenitor subtypes, or are more
limited in their potential, and to establish whether potential correlates with a
particular transcriptional signature. The future interrogation of bHyp cells through
large-scale transcriptomic analyses is likely to reveal intriguing cellular mechanisms
through which the hypothalamus is assembled in four dimensions (space and time)
from this unique population.

1.4 Neuroendocrine Neurogenesis

The transcriptional programs that underpin neuroendocrine neurogenesis have been

described in detail recently (in-depth reviews by Bedont et al. 2015; Burbridge et al.
2016; Alvarez-Bolado 2018), and here we provide only a summary. Gene knockout
studies in the mouse suggest that downstream of Foxd1, the TFs, Six3/Six6 are
required for hypothalamic specification, and are required either transiently, or in a
sustained manner to support two major transcriptional programs for neuroendocrine
neurogenesis: neuroendocrine neurons of the PVN, PeVN and SON differentiate via
1 Development of the Neuroendocrine Hypothalamus 19

Fig. 1.7 Transcriptional programs underlying neuroendocrine neurogenesis and gliogenesis.

Foxd1, Six3 and Six6 are expressed in many hypothalamic neuroendocrine progenitors. Upper
yellow pathway: Many neurons of the PVN, PeVN and SON differentiate via the evolutionarily
conserved zinc finger protein, Fezf2 and the bHLH protein, Olig2, which induce Otp and Sim1.
Parallel activities of Sim1/Arnt2 (Aryl hydrocarbon receptor nuclear translocator 2, dimerization
partner of Sim1) and Otp govern expression of (1) the pou-domain class 3 transcription factor Brain-
2 (Brn2: required for expression of CRH, avp and oxt and (2) Sim2 and hence TRH and Sst
expression. Lower orange pathway: Neurons of the Arc differentiate via an Nkx2.1-dependent
ventral progenitor program that shows inter-regulatory interactions with the TFs Rx, Six3, Six6 and
the signalling ligand Shh (see text). The Arc neurogenic program (orange) is dependent on the
proneural gene Achaete-scute-like 1 (Ascl1: previously known as Mash1). Parallel to and/or
downstream of Ascl1, Neurogenin 3 (Ngn3) and Neurod1 regulate Arc neurogenesis. Middle
grey pathway: The finding that the PVN/PeVN/SON/SCN are reduced in size in Nkx2.1-null
mice, and that ventral otp+ territory and associated TH+/Sst+ neurons, fail to form in rx3 mutant
zebrafish raises the possibility that the PVN/PeVN/SON/SCN are of mixed origin, some neurons
deriving from the ventral progenitor program. The two programs could exert cross-repressive
interactions. Six3 and Six6 are expressed transiently in the PVN and SON, their downregulation
concomitant with Otp, Sim1 and Arnt2 upregulation. Additionally, the transcription factor Lhx2
may act downstream of the Nkx2.1 program as a repressor of Otp and Sim1 and indeed of general
anterior identities. After generating neurons, ventral progenitors give rise to glial pituicytes and
tanycytes (red) the latter, via Rx/Lhx2 interactions

a Shh-responsive Fezf2/Olig2/Otp/Sim1 program (Fig. 1.7 yellow flow lines),

whereas those of the ARC differentiate via a Shh-Nkx2.1/Rx program (Fig. 1.7
orange flow-lines). Initial reports suggested that the PVN/PeVN/SON but not the
Arc can be detected in mice that lack functional Nkx2.1 (Kimura et al. 1996), while
conversely, the PVN/PeVN/SON cannot be detected in mice that lack functional
Sim1. Like neuroendocrine neurons of the PVN, PeVN and SON, Avp(+) neurons of
the SCN are generally accepted to originate from Shh-responsive (rather than
20 M. Placzek et al.

Shh-expressing) progenitors, but via a Six3/Six6/Lhx1/Lhx2 program (Fig. 1.7 grey

flow-lines) that suppresses the Fezf2/Olig2/Otp/Sim1 program [Fig. 1.7; see
in-depth review by Bedont and Blackshaw (2015)].
What are the implications of the anisotropic growth model for neuroendocrine
neurogenesis? As yet, the bHyp chick progenitor domain has not been fate-mapped
to a stage when individual neurons have been generated. However, a prosaic (but not
proved) view is that although many neuroendocrine neurons of the PVN, PeVN,
SON and SCN are likely to derive from Shh-responsive progenitors (potentially
those that expressed Nkx2.2 or Pax6: Fig. 1.4b), some may arise from bHyp cells,
via neuroepithelial Shh(+) progenitors. In this model, some neuroendocrine neurons
of the PVN, PeVN, SON or SCN would share a lineage with those of the Arc
(Figs. 1.4e, 1.5). In support of this idea, the PVN/PeVN/SON/SCN are reduced in
size in the Nkx2.1-null mouse (not remarked on in the original publication: Kimura
et al. 1996). Additionally, conditional inactivation of Rx in mouse leads to a
depletion of both Sst(+) and Avp(+) neurons (Orquera et al. 2016), and loss of
function of rx3 in zebrafish depletes Avp(+) neurons (Dickmeis et al. 2007; Tessmar-
Raible et al. 2007). While these studies do not distinguish between autonomous or
nonautonomous mechanisms, loss of function of Rx/rx3 dramatically prevents/
reduces the growth of basal anterior territories (Muthu et al. 2016; Orquera et al.
2016; Sect. 3.2), including domains that express the homeodomain TF, Otp
(Table 1.2), required for the late development of multiple neuroendocrine neuronal
subsets, including the PVN/PeVN/SON and dopaminergic Arc neurons (reviewed in
Burbridge et al. 2016; Alvarez-Bolado 2018).
Together, these studies raise the possibility that bHyp-derived cells could migrate
extensively into multiple nuclei and contribute to multiple neuroendocrine subsets,
in addition to those of the Arc. The notion that hypothalamic nuclei may be of mixed
progenitor contribution is further supported by lineage-tracing studies that show that
alar progenitors, and/or their differentiating descendants can undergo extensive
tangential migration (Manning et al. 2006; Zhao et al. 2008). Lineage-tracing studies
(a tau-LacZ knockin allele at the Sim1 locus) of mice mutant for Sim1 show that
while Sim1 mutant mice lack a structural PVN and SON, this is not due to the death
of PVN/SON progenitor cells but instead to the fact that they do not migrate
normally (Xu and Fan 2007). Therefore, the adult SON occupies the anterior basal
hypothalamus because differentiating neurons migrate to this location (Box 1.2,
panel C, yellow arrows), and not because its progenitor cells are specified there. In
summary, a prediction is that adult hypothalamic nuclei or regions are populated by
neurons of very distinct origin. Our poor understanding of hypothalamic
neurogenesis highlights the need for state-of-the-art studies that can better resolve
distinct progenitor and neuronal subtypes and their progression over time (Box 1.3).
1 Development of the Neuroendocrine Hypothalamus 21

Box 1.3 Defining Neuroendocrine Neurogenic and Gliogenic Programs

Researchers face many challenges in dissecting the programs that build neu-
roendocrine neurons. First little is known about the heterogeneity of either
progenitor populations or neuronal subsets. Neurohormones such as Oxt, Sst,
TH, and Avp, traditionally used as a proxy for cell fate, are widely expressed
in multiple adult neuronal subtypes/nuclei, and so provide limited insight into
the overall identity and generation of each subtype. This is likely to change
quickly as large-scale omics/single cell RNA-seq studies shed light on the
transcriptional profiles of different subtypes (e.g. Chen et al. 2017; Romanov
et al. 2016; Campbell et al. 2017; Moffitt et al. 2018; Kim et al. 2019), and to
do so in both wild-type and mutant mice (Kim et al. 2019). Second, while
increasing numbers of studies emphasize the importance of progenitor and
neuronal migrations/movements, many more conditional lineage-tracing stud-
ies from defined progenitor populations, linked to multiplex analysis of gene
expression profiles, and real-time analyses of relative cell flows are needed to
understand the full extent of early progenitor cell migrations, or mixing of
progenitors of different origin, and the extent to which boundaries between
hypothalamic regions are fluid or hard-set. Third, it is increasingly clear that
TFs expressed early in development can exert multiple effects at different
times (e.g. Newman et al. 2018a), so that loss of function can reflect a failure of
early progenitor development or a failure of terminal differentiation
(e.g. neurohormone maintenance). Conditional knockout studies and
replacements of coding sequences with a reporter gene are needed to distin-
guish early versus late TF roles and to understand their cross-regulatory
interactions that may maintain distinct domains.

1.5 Maintenance of Fgf10+ Populations for Longer-Term

Building and Lifelong Neurogenesis

The chain of events that unfolds in a sequential manner to build hypothalamic

neurons and the infundibulum and secure a supply of Fgf10(+) progenitor pools
underlies longer-term building steps and lifelong hypothalamic neurogenesis.

1.5.1 Assembly of the Hypothalamo–Pituitary Axis

For many days, Fgf10+ive progenitor cells, and their infundibular daughters, are in
intimate contact with the prospective anterior pituitary (AP, aka adenohypophyseal)
cells. The AP derives from a placode that invaginates to form a rudimentary pouch
that differentiates into Rathke’s pouch, defined through the expression of Lhx3
(a master transcriptional regulator of the endocrine cells of the AP). Sequential
22 M. Placzek et al.

signalling from the ventral-most hypothalamus induces Rathke’s pouch: BMP4

induces the pouch rudiment, then Fgfs activate Lhx3 (Fig. 1.6). Explant recombina-
tion work shows that intimate contact between the infundibulum and the forming
pouch is essential for differentiation of the rudimentary pouch to Rathke’s pouch
(reviewed in Davis et al. 2013; Rizzoti 2015).
New work indicates that Shh plays a role in this sequence. Transient blockade of
Shh signalling, or conditional deletion of Shh from Hesx1(+) cells prevents both
infundibular formation (Sect. 3.3) and the differentiation of Rathke’s pouch from the
rudimentary pouch. Further, instead of forming in intimate contact with the hypo-
thalamus, the rudimentary pouch is separated from the hypothalamus by a loose
meshwork of mesenchymal cells (Fu et al. 2017; Carreno et al. 2017). A prosaic
view is that, in governing the normal movement and growth of anterior hypotha-
lamic progenitor cells (see above), Shh directs the close apposition of the
invaginating rudimentary pouch and the forming infundibulum (Fig. 1.6d). This
close apposition may be important to induce Lhx3. This highlights the importance of
early development to a longer-term building step that is crucial to function: assembly
of the hypothalamo–pituitary neuraxis. The iterative deployment of BMPs, Fgfs and
Shh for different steps in the assembly of the hypothalamo–pituitary neuraxis is
likely to provide a platform through which to interpret human GWAS studies, and
the wide array of mutations associated with pituitary defects.

1.5.2 Assembly of Neurovascular Interfaces in the ME/PP

Cells within the infundibulum, and at its neck, continue to express Fgfs, including
Fgf10 and Fgf3, long after the major neurogenic periods of hypothalamic develop-
ment and play a role in guiding neurosecretory axon terminals and portal capillary
vessels to set up the glial–neuronal–capillary interfaces that underlie the future
functioning of the hypothalamo–pituitary axis and homeostatic balance (Fig. 1.1).
These studies have been reviewed recently (Burbridge et al. 2016) and here we
provide only a summary.
Genetic studies in mouse first indicated that Fgf signalling promotes the guidance
of parvocellular neurosecretory axons (Tsai et al. 2011). Analyses of fgf3 /
zebrafish then revealed that avp(+) and oxt(+) magnocellular cells, and dopaminer-
gic (DA) parvocellular axons fail to project to the ME-like neurohaemal region or
PP, despite neurons initially developing normally (Gutnick et al. 2011; Liu et al.
2013). Expression of a dominant-negative Fgf receptor, in a temporally- or spatially
controlled manner, revealed that Fgf3 directly affects growing axons. Parallel studies
in chick demonstrated that Fgf3/Fgf10 deriving from the infundibulum act as long-
range chemoattractants: low levels of Fgf3/Fgf10 stimulate and orient the growth of
both AVP(+) and DA neurons. Intriguingly, high concentrations of Fgfs stall or repel
growing axons, a mechanism that might help to ensure that neuroendocrine axons
project to, but do not cross the ventral midline: uniquely in this region of the CNS,
axons are non-commissural. Currently, little is known about subsequent mechanisms
that ensure the correct projection of parvocellular axons to the median eminence, and
1 Development of the Neuroendocrine Hypothalamus 23

of magnocellular axons to the posterior lobe. Future studies are needed to examine
this, and to understand how Fgf ligands operate in conjunction with other factors,
including Notch, Shh, BMP and Netrin (reviewed in Burbridge et al. 2016).
Concomitant with guiding axons, FGFs act with other factors to promote local
endothelial vasculogenesis (Liu et al. 2013). These studies point to a role for FGF
signalling during the early modelling of endothelial cells, and the initial formation of
the capillary plexi that characterize the ME and PP (Gutnick and Levkowitz 2012,
Fig.1.1) and suggest that, by attracting axons and stimulating endothelial cells, Fgfs
ensure the coordinated growth of neuroendocrine axons and capillaries to the same
general target. This is then another step in the self-organizing hypothalamic pro-
gram, one that builds the neurovascular interface over time. Elegant studies in
zebrafish show that as oxt(+) magnocellular neurons innervate the PP, they release
oxt from their endfeet. Oxt acts as an angiogenic cue for nearby vascular sprouts,
drawing them towards axon terminals, where they go on to form the hypophyseal
arteries and veins (Gutnick et al. 2011). Thus, parvocellular axons actively partici-
pate in building the vascular networks required for their activity, another proof-of-
principle for the idea that the hypothalamus self-organizes over time.

1.5.3 Pituicyte and Tanycyte Development

The long-term maintenance of Fgf10(+) progenitor populations through embryogen-

esis and into early postnatal life underlies the differentiation of pituicytes and
tanycytes, specialized glial cells that arise in late prenatal/early postnatal life and
that derive from infundibular glia (Fig. 1.7 red flow lines). Pituicytes constitute the
PP: their position in the adult suggests that they derive from Fgf10(+) progenitors at
the distal tip of the infundibulum. As yet, the transcriptional pathways that operate
downstream of Fgf signalling and direct their differentiation remain enigmatic, but
candidates are likely to be obtained through large-scale RNA analyses (Anbalagan
et al. 2018).
Fgf10(+) cells in the proximal infundibulum differentiate first to radial glia and
then to tanycytes—glial cells that are critical to body–brain communication and that
regulate energy homeostasis and reproduction (Fig. 1.1; Table 1.1). Fgf induces and
maintains the TF, Lhx2 (Fig. 1.7, red flow lines; Seth et al. 2006), which first directs
infundibular development and then specifies tanycytes (Salvatierra et al. 2014;
Samms et al. 2015). It is possible that sustained FGF signalling promotes the
transition from infundibular precursors to radial glial cells by maintaining low levels
of Wnt (Wang et al. 2012). An intimate bidirectional loop between Lhx2 and Rx2
operates into postnatal life to induce tanycyte-specific genes such as Gpr50 and
DiO2 (deiodinase 2) in multiciliated tanycytes, and suppress ependymal fates
(Salvatierra et al. 2014). By regulating Lhx2 and DiO2, Fgf signalling is therefore
required throughout life for tanycyte identity and function: DiO2 directly regulates
thyroid hormone (TH) availability and hence appetite and energy expenditure
(Samms et al. 2015, reviewed in Lewis and Ebling 2017).
24 M. Placzek et al.

1.5.4 Tanycytes: Fgf-Responsive Proliferating Progenitors

In addition to their pivotal roles in acute homeostasis (outside of the scope of this
review: see Table 1.1), Fgf10(+) tanycyte subpopulations have been implicated in
longer-term allostatic change through their ability to act as neural stem/progenitor
(NSP) cells.
EdU-labelling studies in postnatal animals first showed that regions around the
ME are hotspots of neurogenesis and gliogenesis, and subsequent studies revealed
that β1, β2 and α2 tanycyte subsets can act as neurogenic and gliogenic NSP cells
(summarized in Rizotti and Lovell-Badge 2017; Lewis and Ebling 2017; Ebling and
Lewis 2018; Clasadonte and Prevot 2018). Lineage-tracing studies showed that
Fgf10(+) tanycytes, although largely quiescent in adulthood, can respond and
adapt to their changing environment by producing new neurons. Thus, Fgf10(+)
tanycytes respond to a high-fat diet by generating neuropeptide Y (NPY)-neurons
that regulate feeding behaviour (Haan et al. 2013). Other lineage-tracing studies
showed that an FGF-responsive α-2 tanycyte subset can self-renew or give rise to
other tanycyte populations, even in unstimulated mice, indicating that localized
proliferation and differentiation to tanycytes maintains the integrity of the ventricular
wall (Robins et al. 2013). Fgf10 and Fgf signal components are maintained in
subsets of tanycytes (including β- and α-2 subsets) throughout life (Haan et al.
2013; Robins et al. 2013) and Fgf signalling can regulate tanycyte proliferation:
infusion of FGF2 into the third ventricle stimulates proliferation of both β-1 and α-2
subtypes (Robins et al. 2013). Tanycyte neurogenesis and gliogenesis can be
stimulated by a range of signals, hormones and metabolites, and ongoing studies
are now seeking to determine whether/how Fgf signalling is involved in supporting
adult neurogenic programs. One study suggests that tanycyte stimulation by FGF1
correlates with the sustained remission of diabetic hyperglycaemia in rodent models
(Scarlett et al. 2016), suggesting that appropriate regulation of Fgf signalling in
tanycytes may be critical to health and disease. Important questions for the future
are: (1) whether adult programs of tanycyte homeostasis and neurogenesis will share
features of embryonic Fgf10(+) progenitor homeostasis and neurogenesis;
(2) whether physiological or epigenetic factors that are known to govern
neurogenesis from adult tanycytes do so via interactions with developmental
signal-TF pathways; and (3) whether neuroendocrine neurons can be generated
throughout life.

1.6 Plasticity Through the Life Course: Is a Hypothalamic Stem

Cell Established in the Embryo?

The presence of neurogenic/gliogenic tanycytes, in close proximity to the extensive

capillary network of the ME, suggests that tanycytes are components of a hypotha-
lamic NSP cell niche. Given their multipotent nature, a likely possibility is that
Fgf10(+) populations that are established in the early embryo form the essential cell
type, and organize the development of an emerging hypothalamic NSP cell niche.
1 Development of the Neuroendocrine Hypothalamus 25

An outstanding question is whether lineage-related progenitors within the niche

share an equivalent state and so represent multipotent stem-like cells, or show
inequivalence and represent adaptive progenitor populations. Currently, we lack
in vivo evidence to distinguish between these possibilities, although ex vivo
neurospherogenic studies demonstrate that some subsets of tanycytes exhibit a
stem-like capacity to self-renew and differentiate (Robins et al. 2013).
The question of whether a multipotent stem cell is established in the early stages
of embryogenesis has implications for how the neuroendocrine hypothalamus might
adapt to changing physiological conditions on different time scales, and for the
robustness of the hypothalamus through life. If a stem cell exists, then by definition,
it has equivalent potential to differentiate throughout life. This implies that a single
cell can respond to multiple stimuli to differentiate into multiple basal neurons over
life. If, on the other hand, cells are not equivalent, i.e. have limited differentiation
potential, this suggests that they become remodelled to anticipate and respond to the
changing environment. How could such a cell become remodelled over time? We
speculate that this could be achieved through a self-organization program that is
driven by hormones that are themselves developmentally regulated by tanycytes.
There is ample evidence that tanycytes can detect and respond to hormones. IGF, for
instance, can stimulate tanycytes and provoke changes in neurogenic programs
(reviewed in Rizotti and Lovell-Badge 2017; Lewis and Ebling 2017), and tanycytes
both regulate (see Sect. 5.3) and respond to TH, through a mechanism that is
intrinsically linked to developmental time (Mohácsik et al. 2016). Together, this
raises the possibility that tanycytes may be key components of an intrinsic hypotha-
lamic clock, responding to developmentally timed steroid signals and supporting
long-term allostatic change.
An important area of future work will be to better understand whether, as
suggested through emerging work, such tanycyte-regulated activity is sensitive to
epigenetic change. This would form a platform to understand how environmental
factors, in particular those that affect steroid signalling, could lead to maladaptive
tanycyte behaviour and homeostatic dysfunction over life. It seems likely that a
continued focus of research on bHyp cells and their derivatives will lead to a far
better understanding of how the hypothalamus can effect both short-term physiolog-
ical regulation of factors such as stress and energy balance, and long-term changes
such as puberty and pregnancy, and to understand sensitive developmental time
points for cells that regulate particular physiological processes.

Acknowledgments Work supported by the Wellcome Trust.

Key References

Alvarez-Bolado et al. (2012)—Lineage-tracing of a subset of Shh(+) neuroepithelial

progenitors shows that they contribute to the tuberal and mammillary
hypothalamus.
26 M. Placzek et al.

Carreno et al. (2017)—Genetic studies in the mouse reveal the integrated develop-
ment of the ventral hypothalamus (including the infundibulum and Shh(+)
neuroepithelial progenitors) and the adenohypophysis.
Chen et al. (2017)—A large scale single cell RNA seq analysis of the adult mouse
hypothalamus confirms that neuroendocrine neurohormones/neurotransmitters
are expressed in heterogeneous cell types.
Corman et al. (2018)—Sophisticated mouse studies, including analyses of condi-
tional knock-out and reporter lines, show that tuberal neurons derive from both
Shh-expressing and Shh-responsive progenitors.
Fu et al. (2017)—Fate-mapping and pharmacological studies in the early chick
provides evidence for the anisotropic growth model of hypothalamic
development.
Haan et al. (2013)—Evidence for de novo adult neurogenesis from Fgf10(+)
tanycytes; together with Fu et al. (2017) and Robins et al. (2013), raises possibil-
ity that an Fgf10(+) population is important for building hypothalamic neurons
throughout life.
Kim et al. (2019)—A new study that sets a benchmark for molecular analysis of the
developing hypothalamus. Single cell RNA seq analyses of the hypothalamus at
different time points in embryogenesis and postnatal life provide evidence for the
early specification of key hypothalamic neuronal subsets. Comparison of wild
type and mutant mice provides evidence that the hypothalamus is a diencephalic-
derived structure.
Manning et al. (2006)—A study in the chick embryo that provided an understanding
of the co-ordination of pattern and proliferation of ventral hypothalamic progeni-
tor cells. Re-interpretation of this paper, in light of the anisotropic growth shown
in Fu et al. (2019), suggests that the cells that express Tbx2/BMP are bHyp cells,
and the Emx2+ cells are mammillary progenitors.
Muthu et al. (2016)—Anisotropic growth of anterior hypothalamus in zebrafish;
absence of growth leads to failure of differentiation of both ‘anterior’ and
‘tuberal’ neuroendocrine neurons.
Newman et al. (2018a)—Sophisticated conditional knock-out and reporter lines
show hypothalamus is of diencephalic origin.
Orquera et al. (2016)—Evidence for conserved mechanism of anterior progenitor
growth and differentiation, via a Shh-Rax pathway.
Robins et al. (2013)—Analysis of adult mouse shows that Fgf10(+) stem/progenitor
cells exist as tanycytes that line the ventricle; with Fu et al. (2019) and Haan et al.
(2013), suggests that Fgf10(+) stem/progenitor cells are retained through life.
Shimogori et al. (2010)—A large-scale screen and in situ analysis identifies many
hypothalamic progenitor domains; conditional knock-out of subsets of Shh(+)
neuroepithelial progenitors shows importance in anterior/tuberal growth and
neurogenesis.
Szabó et al. (2009)—Lineage-tracing of a subset of Shh(+) RDVM cells shows
contribution to tuberal and mammillary hypothalamus.
Tessmar-Raible et al. (2007)—Conservation of neuroendocrine cells and conserva-
tion of key molecules.
1 Development of the Neuroendocrine Hypothalamus 27

Trowe et al. (2013)—Analysis of mouse, analysing control of Shh expression in

RSVM cells. With Manning et al. (2006), suggests a mechanism for resolution of
bHyp cells into Shh(+) and Shh( ) domains.
Xu et al. (2007)—Proof-of-principle that differentiating hypothalamic neurons
undergo tangential migration.
Zhao et al. (2008)—Proof-of-principle that progenitors other than bHyp cells also
undergo tangential migration.

References
Alvarez-Bolado G (2018) Development of neuroendocrine neurons in the mammalian hypothala-
mus. Cell Tissue Res. https://doi.org/10.1007/s00441-018-2859-1
Alvarez-Bolado G, Paul FA, Blaess S (2012) Sonic hedgehog lineage in the mouse hypothalamus:
from progenitor domains to hypothalamic regions. Neural Dev 7:4. https://doi.org/10.1186/
1749-8104-7-4
Anbalagan S, Gordon L, Blechman J, Matsuoka RL, Rajamannar P, Wircer E, Biran J, Reuveny A,
Leshkowitz D, Stainier DYR, Levkowitz G (2018) Pituicyte cues regulate the development of
permeable neuro-vascular interfaces. Dev Cell. https://doi.org/10.1016/j.devcel.2018.10.017
Bedont JL, Blackshaw S (2015) Constructing the suprachiasmatic nucleus: a watchmaker's per-
spective on the central clockworks. Front Syst Neurosci 9:74
Bedont JL, Newman EA, Blackshaw S (2015) Patterning, specification, and differentiation in the
developing hypothalamus. Wiley Interdiscip Rev Dev Biol. https://doi.org/10.1002/wdev.187
Brinkmeier ML, Potok MA, Davis SW, Camper SA (2007) TCF4 deficiency expands ventral
diencephalon signaling and increases induction of pituitary progenitors. Dev Biol 311:396–407
Burbridge S, Stewart I, Placzek M (2016) Development of the neuroendocrine hypothalamus.
Compr Physiol 6(2):623–643
Campbell JN, Macosko EZ, Fenselau H et al (2017) A molecular census of arcuate hypothalamus
and median eminence cell types. Nat Neurosci 20:484–496
Carreno G, Apps JR, Lodge EJ, Panousopoulos L, Haston S, Gonzalez-Meljem JM, Hahn H,
Andoniadou CL, Martinez-Barbera JP (2017) Hypothalamic sonic hedgehog is required for cell
specification and proliferation of LHX3/LHX4 pituitary embryonic precursors. Development
144(18):3289–3302
Chen R, Wu X, Jiang L, Zhang Y (2017) Single-cell RNA-Seq reveals hypothalamic cell diversity.
Cell Rep 18(13):3227–3241
Clasadonte J, Prevot V (2018) The special relationship: glia-neuron interactions in the neuroendo-
crine hypothalamus. Nat Rev Endocrinol 14(1):25–44. https://doi.org/10.1038/nrendo.2017.124
Corman TS, Bergendahl SE, Epstein DJ (2018) Distinct temporal requirements for Sonic hedgehog
signaling in development of the tuberal hypothalamus. Development 145(21). https://doi.org/
10.1242/dev.167379
Couly GF, Le Douarin NM (1985) Mapping of the early neural primordium in quali-chick chimeras.
Dev Bioll 110:422–439
Davis SW, Ellsworth BS, Peréz Millan MI, Gergics P, Schade V, Foyouzi N, Brinkmeier ML,
Mortensen AH, Camper SA (2013) Pituitary gland development and disease: from stem cell to
hormone production. Curr Top Dev Biol 106:1–47
Dickmeis T, Lahiri K, Nica G, Vallone D, Santoriello C, Neumann CJ, Hammerschmidt M, Foulkes
NS (2007) Glucocorticoids play a key role in circadian cell cycle rhythms. PLoS Biol 5:e78
Ebling FJP, Lewis JE (2018) Tanycytes and hypothalamic control of energy balance. Glia 66
(6):1176–1184
Fu T, Towers M, Placzek MA (2017) Fgf10+ progenitors give rise to the chick hypothalamus by
rostral and caudal growth and differentiation. Development 144(18):3278–3288
28 M. Placzek et al.

Fu T, Pearson C, Towers M, Placzek, MA (2019) Development of the basal hypothalamus through

anisotropic growth. J Neuroendocrinol 31(5):e12727
Gizowski C, Zaelzer C, Bourque CW (2016) Clock-driven vasopressin neurotransmission mediates
anticipatory thirst prior to sleep. Nature 537:685–688
Goto M, Hojo M, Ando M, Kita A, Kitagawa M, Ohtsuka T, Kageyama R, Miyamoto S (2015)
Hes1 and Hes5 are required for differentiation of pituicytes and formation of the neurohypoph-
ysis in pituitary development. Brain Res 1625:206–217
Gutnick A, Levkowitz G (2012) The neurohypophysis: fishing for new insights. J Neuroendocrinol
24(6):973–974
Gutnick A, Blechman J, Kaslin J, Herwig L, Belting HG, Affolter M, Bonkowsky JL, Levkowitz G
(2011) The hypothalamic neuropeptide oxytocin is required for formation of the neurovascular
interface of the pituitary. Dev Cell 21:642–654
Haan N, Goodman T, Najdi-Samiei A, Stratford CM, Rice R, El Agha E, Bellusci S, Hajihosseini
MK (2013) Fgf10(+) tanycytes add new neurons to the appetite/energy-balance regulating
centers of the postnatal and adult hypothalamus. J Neurosci 33(14):6170–6180
Jeong Y, El-Jaick K, Roessler E, Muenke M, Epstein DJ (2006) A functional screen for sonic
hedgehog regulatory elements across a 1 Mb interval identifies long-range ventral forebrain
enhancers. Development 133(4):761–772
Kapsimali M, Caneparo L, Houart C, Wilson SW (2004) Inhibition of Wnt/Axin/beta-catenin
pathway activity promotes ventral CNS midline tissue to adopt hypothalamic rather than
floorplate identity. Development 131:5923–5933
Kim DW, Washington PW, Wan ZQ, Lin S, Sun C, Jiang L, Blackshaw S (2019) Single cell
RNA-Seq analysis identifies molecular mechanisms controlling hypothalamic patterning and
differentiation. bioRxiv. https://doi.org/10.1101/657148
Kimura S, Hara Y, Pineau T, Fernandez-Salguero P, Fox CH, Ward JM, Gonzalez FJ (1996) The
T/ebp null mouse: thyroid-specific enhancer-binding protein is essential for the organogenesis
of the thyroid, lung, ventral forebrain, and pituitary. Genes Dev 10:60–69
Lewis JE, Ebling FJ (2017) Tanycytes as regulators of seasonal cycles in neuroendocrine function.
Front Neurol 10:8–79
Liu F, Pogoda HM, Pearson CA, Ohyama K, Lohr H, Hammerschmidt M, Placzek M (2013) Direct
and indirect roles of Fgf3 and Fgf10 in innervation and vascularisation of the vertebrate
hypothalamic neurohypophysis. Development 140:1111–1122
Manning L, Ohyama K, Saeger B, Hatano O, Wilson SA, Logan M, Placzek M (2006) Regional
morphogenesis in the hypothalamus: a BMP-Tbx2 pathway coordinates fate and proliferation
through Shh downregulation. Dev Cell 11:873–885
Manoli M, Driever W (2014) nkx2.1 and nkx2.4 genes function partially redundant during
development of the zebrafish hypothalamus, preoptic region, and pallidum. Front Neuroanat
8:145
Marsters CM, Rosin JM, Thornton HF, Aslanpour S, Klenin N, Wilkinson G, Schuurmans C,
Pittman QJ, Kurrasch DM (2016) Oligodendrocyte development in the embryonic tuberal
hypothalamus and the influence of Ascl1. Neural Dev 11(1):20
Mattes B, Dang Y, Greicius G, Kaufmann LT, Prunsche B, Rosenbauer J, Stegmaier J, Mikut R,
Özbek S, Nienhaus GU, Schug A, Virshup DM, Scholpp S (2018) Wnt/PCP controls spreading
of Wnt/β-catenin signals by cytonemes in vertebrates. Elife 7. https://doi.org/10.7554/eLife.
36953
Miranda-Angulo AL, Byerly MS, Mesa J, Wang H, Blackshaw S (2014) Rax regulates hypotha-
lamic tanycyte differentiation and barrier function in mice. J Comp Neurol 522:876–899
Moffitt JR, Bambah-Mukku D, Eichhorn SW, Vaughn E, Shekhar K, Perez JD, Rubinstein ND,
Hao J, Regev A, Dulac C, Zhuang X (2018) Molecular, spatial, and functional single-cell
profiling of the hypothalamic preoptic region. Science 362(6416). https://doi.org/10.1126/
science.aau5324
Mohácsik P, Füzesi T, Doleschall M, Szilvásy-Szabó A, Vancamp P, Hadadi É, Darras VM,
Fekete C, Gereben B (2016) Increased thyroid hormone activation accompanies the formation
1 Development of the Neuroendocrine Hypothalamus 29

of thyroid hormone-dependent negative feedback in developing chicken hypothalamus. Endo-

crinology 157(3):1211–1221
Mortensen AH, Schade V, Lamonerie T, Camper SA (2015) Deletion of OTX2 in neural ectoderm
delays anterior pituitary development. Hum Mol Genet 24:939–953
Muthu V, Eachus H, Ellis P, Brown S, Placzek M (2016) Rx3 and Shh direct anisotropic growth and
specification in the zebrafish tuberal/anterior hypothalamus. Development (14):2651–2663.
https://doi.org/10.1242/dev.138305
Newman EA, Kim DW, Wan J, Wang J, Qian J, Blackshaw S (2018a) Foxd1 is required for
terminal differentiaiton of anterior hypothalamic neuronal subtypes. Dev Biol 439(2):102–111.
https://doi.org/10.1016/j.ydbio.2018.04.012
Newman EA, Wu D, Taketo MM, Zhang J, Blackshaw S (2018b) Canonical Wnt signaling
regulates patterning, differentiation and nucleogenesis in mouse hypothalamus and prethalamus.
Dev Biol 442(2):236–248
Ohuchi H, Hori Y, Yamasaki M, Harada H, Sekine K, Kato S, Itoh N (2000) FGF10 acts as a major
ligand for FGF receptor 2 IIIb in mouse multi-organ development. Biochem Biophys Res
Commun 277:643–649
Orquera DP, Nasif S, Low MJ, Rubinstein M, de Souza FSJ (2016) Essential function of the
transcription factor Rax in the early patterning of the mammalian hypothalamus. Dev Biol 416
(1):212–224. https://doi.org/10.1016/j.ydbio.2016.05.021
Patten I, Kulesa P, Shen MM, Fraser S, Placzek M (2003) Distinct modes of floor plate induction in
the chick embryo. Development 130:4809–4821
Pearson CA, Ohyama K, Manning L, Aghamohammadzadeh S, Sang H, Placzek M (2011)
FGF-dependent midline-derived progenitor cells in hypothalamic infundibular development.
Development 138:2613–2624
Placzek M, Briscoe J (2005) The floor plate: multiple cells, multiple signals. Nat Rev Neurosci
6:230–240
Placzek M, Briscoe J (2018) Sonic hedgehog in vertebrate neural tube development. Int J Dev Biol
62(1–2–3):225–234. https://doi.org/10.1387/ijdb.170293jb
Potok MA, Cha KB, Hunt A, Brinkmeier ML, Leitges M, Kispert A, Camper SA (2008) WNT
signaling affects gene expression in the ventral diencephalon and pituitary gland growth. Dev
Dyn 237:1006–1020
Puelles L, Martinez-de-la-Torre M, Bardet S, Rubenstein JLR (2012) Hypothalamus. In: Watson C,
Paxinos G, Puelles L (eds) Mouse nervous system. Academic Press, San Diego, CA, pp
221–312
Puelles L, Harrison M, Paxinos G, Watson C (2013) A developmental ontology for the mammalian
brain based on the prosomeric model. Trend Neurosci 36:570–578
Ratie L, Ware M, Barloy-Hubler F, Rome H, Gicquel I, Dubourg C, David V, Dupe V (2013) Novel
genes upregulated when NOTCH signalling is disrupted during hypothalamic development.
Neural Dev 8:25
Ratié L, Ware M, Jagline H, David V, Dupé V (2014) Dynamic expression of Notch-dependent
neurogenic markers in the chick embryonic nervous system. Front Neuroanat 8:158
Rizotti K, Lovell-Badge R (2017) Pivotal role of median eminence tanycytes for hypothalamic
function and neurogenesis. Mol Cell Endocrinol 445:7–13
Rizzoti K (2015) Genetic regulation of murine pituitary development. J Mol Endocrinol 54(2):R55–
R73
Robins SC, Stewart I, McNay DE, Taylor V, Giachino C, Goetz M, Ninkovic J, Briancon N,
Maratos-Flier E, Flier JS, Kokoeva MV, Placzek M (2013) α-Tanycytes of the adult hypotha-
lamic third ventricle include distinct populations of FGF-responsive neural progenitors. Nat
Commun 4:2049. https://doi.org/10.1038/ncomms3049
Romanov RA, Zeisel A, Bakker J et al (2016) Molecular interrogation of hypothalamic organization
reveals distinct dopamine neuronal subtypes. Nat Neurosci 20:176–188
Salvatierra J, Lee DA, Zibetti C, Duran-Moreno M, Yoo S, Newman EA, Wang H, Bedont JL, de
Melo J, Miranda-Angulo AL, Gil-Perotin S, Garcia-Verdugo JM, Blackshaw S (2014) The LIM
30 M. Placzek et al.

homeodomain factor Lhx2 is required for hypothalamic tanycyte specification and differentia-
tion. J Neurosci 34:16809–16820
Samms RJ, Lewis JE, Lory A, Fowler MJ, Cooper S, Warner A, Emmerson P, Adams AC, Luckett
JC, Perkins AC, Wilson D, Barrett P, Tsintzas K, Ebling FJ (2015) Antibody-mediated inhibi-
tion of the FGFR1c isoform induces a catabolic lean state in siberian hamsters. Curr Biol 25
(22):2997–3003
Sánchez-Arrones L, Ferrán JL, Rodríguez-Gallardo L, Puelles L (2009) Incipient forebrain
boundaries traced by differential gene expression and fate mapping in the chick neural plate.
Dev Biol 335:43–65
Scarlett JM, Rojas JM, Matsen ME, Kaiyala KJ, Stefanovski D, Bergman RN, Nguyen HT,
Dorfman MD, Lantier L, Wasserman DH, Mirzadeh Z, Unterman TG, Morton GJ, Schwartz
MW (2016) Central injection of fibroblast growth factor 1 induces sustained remission of
diabetic hyperglycemia in rodents. Nat Med 22(7):800–806
Seth A, Culverwell J, Walkowicz M, Toro S, Rick JM, Neuhauss SC, Varga ZM, Karlstrom RO
(2006) belladonna/(Ihx2) is required for neural patterning and midline axon guidance in the
zebrafish forebrain. Development 133(4):725–735
Shimogori T, Lee DA, Miranda-Angulo A, Yang Y, Wang H, Jiang L, Yoshida AC, Kataoka A,
Mashiko H, Avetisyan M, Qi L, Qian J, Blackshaw S (2010) A genomic atlas of mouse
hypothalamic development. Nat Neurosci 13:767–775
Swanson LW (1992) Brain maps: structure of the rat brain. Elsevier, New York
Szabó NE, Zhao T, Cankaya M, Theil T, Zhou X, Alvarez-Bolado G (2009) Role of neuroepithelial
Sonic hedgehog in hypothalamic patterning. J Neurosci 29(21):6989–7002. https://doi.org/10.
1523/JNEUROSCI.1089-09.2009
Tessmar-Raible K, Raible F, Christodoulou F, Guy K, Rembold M, Hausen H, Arendt D (2007)
Conserved sensory-neurosecretory cell types in annelid and fish forebrain: insights into hypo-
thalamus evolution. Cell 129:1389–1400
Trowe MO, Zhao L, Weiss AC, Christoffels V, Epstein DJ, Kispert A (2013) Inhibition of Sox2-
dependent activation of Shh in the ventral diencephalon by Tbx3 is required for formation of the
neurohypophysis. Development 140:2299–2309
Trudel E, Bourque CW (2010) Central clock excites vasopressin neurons by waking osmosensory
afferents during late sleep. Nat Neurosci 13:467–474
Tsai PS, Brooks LR, Rochester JR, Kavanaugh SI, Chung WC (2011) Fibroblast growth factor
signaling in the developing neuroendocrine hypothalamus. Front Neuroendocrinol 32:95–107
Wang X, Kopinke D, Lin J, McPherson AD, Duncan RN, Otsuna H, Moro E, Hoshijima K,
Grunwald DJ, Argenton F, Chien CB, Murtaugh LC, Dorsky RI (2012) Wnt signaling regulates
postembryonic hypothalamic progenitor differentiation. Dev Cell 23:624–636
Xu C, Fan CM (2007) Allocation of paraventricular and supraoptic neurons requires Sim1 function:
a role for a Sim1 downstream gene PlexinC1. Mol Endocrinol 21(5):1234–1245
Zhang L, Mathers PH, Jamrich M (2000) Function of Rx, but not Pax6, is essential for the formation
of retinal progenitor cells in mice. Genesis 28:135–142
Zhao T, Szabo N, Ma J, Luo L, Zhou X, Alvarez-Bolado G (2008) Genetic mapping of Foxb1-cell
lineage shows migration from caudal diencephalon to telencephalon and lateral hypothalamus.
Eur J Neurosci 28:1941–1955
Zhao Y, Mailloux CM, Hermesz E, Palkóvits M, Westphal H (2010) A role of the LIM-homeobox
gene Lhx2 in the regulation of pituitary development. Dev Biol 337(2):313–323. https://doi.org/
10.1016/j.ydbio.2009.11.002
Zhao L, Zevallos SE, Rizzoti K, Jeong Y, Lovell-Badge R, Epstein DJ (2012) Disruption of SoxB1-
dependent Sonic hedgehog expression in the hypothalamus causes septo-optic dysplasia. Dev
Cell 22:585–596
Sonic hedgehog in Hypothalamus
Development 2
Gonzalo Alvarez-Bolado

Abstract
Sonic hedgehog (SHH) is a secreted signaling protein able to confer ventral cell
fate to the spinal cord on the basis of a concentration gradient that translates into
activation of specific genes through Gli transcription factors. The intricate and
dynamic expression pattern of Shh in the forebrain raises the question of its
function in the developing hypothalamus. Experiments in chick and mouse
show that SHH from the prechordal plate cooperates with Bone Morphogenetic
Proteins (BMPs) to induce a special “hypothalamic floor plate” where Shh
expression will be activated by three transcription factors, Nkx2-1, Six3, and
Sox2/3. Conditional mouse mutants show that this new source of SHH is respon-
sible for hypothalamic growth as well as full differentiation of most hypothalamic
cell types and the formation of the lateral hypothalamus. Experiments on chick
embryos indicate that for the specification of hypothalamic regions Shh
cooperates with Fgf10. For the final differentiation of the basal hypothalamic
regions and hypophysis, repression of Shh by T-box proteins is necessary.
Precisely how SHH unlocks hypothalamic cell fates is not entirely clear. But in
mouse and zebrafish, Shh bestows neuroendocrine cell fates through a complex
relation with Rax/rx3.

Keywords
Floor plate, Gli, Hypothalamus, Lineage, Prechordal plate, Progenitor domain,
Sonic hedgehog

G. Alvarez-Bolado (*)
Department of Neuroanatomy, University of Heidelberg School of Medicine, Heidelberg, Germany
e-mail: [email protected]

# Springer Nature Switzerland AG 2020 31

S. Wray, S. Blackshaw (eds.), Developmental Neuroendocrinology, Masterclass in
Neuroendocrinology 9, https://doi.org/10.1007/978-3-030-40002-6_2
32 G. Alvarez-Bolado

2.1 Introduction

Sonic hedgehog (Shh) encodes a secreted protein essential for the development of the
central nervous system (CNS). SHH has fascinated developmental biologists since
its discovery because of its ability to confer a variety of cell fates on the basis of a
concentration gradient. The most important developmental role of SHH in the CNS
is to confer specific character to the ventral half of the neural tube (Ericson et al.
1995). Progenitor cells in the early neural tube (the stem cells of the CNS) receive
their “ventral character” from SHH secreted by a specific source tissue. In the spinal
cord, the first source of SHH is the axial mesoderm underlying the presumptive
spinal cord, i.e., the notochord. Neuroepithelial cells in the ventral midline of the
presumptive spinal cord are induced by SHH to become floor plate cells and start
expressing Shh and secreting SHH themselves into the neural tube. SHH then forms
a concentration gradient effective over a relatively long distance (20 cell diameters
from the source at least) and imparts ventral character to the progenitor cells that
form the early neural tube according to this gradient. The highest concentration
induces the most ventral structures and progressively lower concentrations of SHH
specify progressively less ventral (more dorsal) cell fates. The complex spatial and
temporal properties of the SHH gradient and the ways in which cells respond to it are
well delineated. The resulting body of knowledge is an achievement of modern
developmental biology (Dessaud et al. 2008; Sagner and Briscoe 2017). Most of this
work was done using the spinal cord as a model. In the forebrain, the dynamic
expression pattern of Shh suggested even a more intricate involvement of this
signaling pathway in the development of the ventral forebrain and of the
hypothalamus.

2.1.1 The Morphogen Protein SHH and Its Signaling Pathway

After translation, SHH cleaves itself into an active N-terminal fragment and is then
covalently attached to two lipid moieties, palmitoyl and cholesterol (which probably
confer SHH the peculiar capacity to signal through relatively long distances). SHH is
then released by source cells through one or maybe several mechanisms. Essential
for SHH to exert its influence on the target cell is a cellular compartment, the primary
cilium (Huangfu et al. 2003), as well as several proteins acting as receptors, second
messengers and effectors of the SHH signal. In vertebrates, the SHH transduction
pathway ends on three effector proteins, the GLI transcription factors (GLI1, 2, and
3). Cells competent to respond to SHH express two transmembrane proteins,
Patched (Ptch1) and Smoothened (Smo), as well as Gli2 and Gli3 (and other specific
genes that regulate the pathway).
While GLI1 is a transcriptional activator, GLI2 and GLI3 can act as activators
(GliA) or repressors (GliR). The GLI DNA-binding domain consists of five zinc
fingers, and the three GLI proteins have similar DNA-binding specificity. The
activator domain is the C-terminal, and the repressor domain of GLI2 and GLI3 is
the N-terminal. In SHH responsive cells, GLI2 and GLI3 are translated and
2 Sonic hedgehog in Hypothalamus Development 33

processed into their repressor forms and repress Shh pathway targets (the primary
repressor is GLI3). Upon binding of SHH to the receptor complex, PTCH1 allows
SMO to change into its active conformation. SMO then activates a signaling cascade
that inhibits processing of GLI proteins into their repressor forms and instead they
are processed into their activator forms (the activator form is mostly GLI2). In
contrast, Gli1 is the first gene expressed upon activation of the pathway and therefore
its expression is a readout of Shh pathway activation. (Note: This is somewhat
divergent in fishes). The components of the SHH transduction pathway are localized
to the cilia, possibly allowing for greater sensitivity of recipient cells to SHH
molecules. The SHH signaling cascade is still being unraveled, and the subject has
been often and expertly reviewed (see for instance Briscoe and Thérond 2013).

2.1.2 Hypothalamic Nomenclatures (Fig. 2.1)

The adult hypothalamus has been classically subdivided into four transverse regions,
from rostral to caudal: preoptic, anterior, tuberal, and mammillary (Fig. 2.1a).
However, developmental gene expression patterns do not entirely conform to the
neuroanatomy of the adult brain as it is traditionally understood (Puelles et al. 2012)
(Fig. 2.1b; from Haddad-Tóvolli et al. 2015). If we take Shh expression (pink in
Fig. 2.1b) as the ultimate ventral marker, the tuberal and mammillary regions are
ventral (i.e., they express Shh during development) and the preoptic and anterior are
dorsal (Fig. 2.1b). To avoid the resulting “clash of nomenclatures” (i.e., adult versus
developmental neuroanatomy), the words “alar” and “basal” (Fig. 2.1b, yellow
versus red, separated by a red line) are sometimes used instead of “dorsal” and
“ventral” when speaking of the developing hypothalamus. The well-known term
“ventral forebrain” becomes “anterior forebrain” and includes alar (dorsal) and basal
(ventral) portions. The most anterior region has received the name of “acroterminal
region” (green in Fig. 2.1b), the basal portion of which (green striped area in
Fig. 2.1b) corresponds to the Rostral Diencephalic Ventral Midline cells (RDVM;
also called hypothalamic floor plate and Medial Progenitor Domain (used through-
out this chapter)).
For this reason, it is preferable to substitute “presumptive hypothalamus” for
“ventral forebrain.” Even more, developmental studies show that the preoptic region
originates in the telencephalon and is not part of the diencephalon (see Sect. 2.7.3).
These questions reflect the current discussion about the developmental and evolu-
tionary origins of the four hypothalamic regions and are not yet settled.
34 G. Alvarez-Bolado

Fig. 2.1 Hypothalamic regional nomenclature and Shh expression. Conventional representation of
the hypothalamus (gray) as ventral region with four regions aligned from rostral to caudal: POA,
preoptic; ANT, anterior; TUB, tuberal; MAM, mammillary. Model of the hypothalamus consider-
ing Shh expression (pink) as basal (ventral) marker. The tuberal and mammillary regions would
then be not “caudal” but rather basal (“ventral”), and the anterior region would form the alar
(“dorsal”) hypothalamus (yellow). In this model, the most anterior portion of the hypothalamus is
called acroterminal region (green); its basal portion (striped green) corresponds to the Medial
Progenitor Domain or hypothalamic floor plate. Additionally, the preoptic region is considered
part of the telencephalon. The red line represents the boundary of Shh expression. ac, anterior
commissure; hp, hypophysis; MB, midbrain; os, optic stalk; POA, preoptic region; PT, pretectum;
PTh, prethalamus; TH, thalamus; ZLI, zona limitans (modified from Haddad-Tóvolli et al. 2015).

Box 2.1: Animal Models and Their Key Contributions

Our current knowledge on the role of Shh in the developing hypothalamus is
based on work on chicken, mouse, and zebrafish. Precise surgical
manipulations can be performed on chick embryos, so that very specific
developmental issues can be asked. Three key contributions made on this
model are those of Manning et al. (2006) on the importance of Shh repression
by T-box2, Dale et al. (1997) on the cooperation between BMP and SHH to
induce the hypothalamic floor plate, and the explanation of how hypothalamic
regions appear through an interplay between Fgf10 and Shh (Fu et al. 2017)

(continued)
2 Sonic hedgehog in Hypothalamus Development 35

Box 2.1 (continued)

(Sects. 2.6.4.4 and 2.7.3). The mouse model offers a tried-and-true way to
learn about gene function: mutating genes and analyzing the resulting
phenotypes on the mammalian brain. The analysis of mutant brains lacking
Shh in the neuroepithelium (Szabó et al. 2009; Shimogori et al. 2010), the
lineage of Shh in the hypothalamus (Alvarez-Bolado et al. 2012) and the
function of Gli genes in hypothalamic development (Haddad-Tóvolli et al.
2015) are among the contributions made on this model. The zebrafish
combines, in principle, the genetic accessibility of the mouse with the practi-
cality and efficiency possible on the chick embryo. The complex relation
between Shh and rx3, essential for the specification and differentiation of
neuroendocrine cell fates has been unraveled in the zebrafish (Muthu et al.
2016). Needless to say, each model has its drawbacks. The chick embryo is not
accessible genetically and the results of experiments cannot be evaluated on
the adult brain. In mice, the questions that can be asked often lack refinement
and obtaining results is a slow process. In fish, development and neuroanat-
omy of the hypothalamus are not as well known as they are in mammals.

2.2 Shh Expression in the Presumptive Hypothalamus

of the Mouse (Fig. 2.2)

At E7.5 (neural fold), Shh is expressed in the prechordal plate, immediately under-
lying the midline of the neural plate (prechordal Shh; (Aoto et al. 2009; Echelard
et al. 1993)) (Fig. 2.2a). This is followed immediately by the onset of Gli1 expres-
sion in the midline of the overlying neural ectoderm, indicating activation of Shh
signaling. Gli1 expression appears in the midline; Gli2 and Gli3, on the contrary, are
expressed widely in the ectoderm, indicating competence to respond to secreted
SHH (Hui et al. 1994).

At E8.5 Shh is expressed in a medial territory overlying the prechordal plate and
extending through the presumptive mammillary and tuberal regions (the two regions
forming the “basal hypothalamus,” Sect. 2.1.2). This territory is the “Medial Pro-
genitor Domain” (Sect. 2.3). This region is characterized not only by Shh expression
like the floor plate in the spinal cord, but for the concomitant expression of Nkx2-1
and Bmp7 genes (while characteristic floor plate marker Foxa2 is expressed here
only transiently) (Dale et al. 1997) (Sect. 2.6.1). Expression of Gli1 and Gli2 overlap
forming a paired lateral band (Fig. 2.2b), while Gli3 is expressed in a peripheral,
non-hypothalamic domain.

At E9.5 Shh expression in the ventral midline extends in the rostral direction but
starts to be downregulated at the tuberal and mammillary level. In this way, a medial
36 G. Alvarez-Bolado

Fig. 2.2 Shh and Gli expression in the developing forebrain. In situ hybridization detection of Shh
(left) and diagrams showing expression domains of the Gli factors and Shh in the presumptive
hypothalamus at the early, middle, and late phases. “lat” and “med,” lateral and medial domains,
respectively. Diagrams (right side) of the corresponding Gli expression patterns with data from
Haddad-Tóvolli et al. (2015) and Hui et al. (1994). The ages indicated (mouse embryonic days, E)
are approximate (since they are based on an estimation of the time of fertilization). Note: Very
similar expression patterns have been found in the mouse, chick, and zebrafish embryos. (a) Shh
expression in the prechordal plate seen on a transverse section of the head folds of an E8.0 embryo.
(b–d) In situ hybridization detection of Shh on wild-type mouse brains. In (c) and (d), the heads
have been sagittally halved to show the neuroepithelium–ventricular zone on the inner side of the
brain; rostral to the left. (b) Shh expression domain in the ventral forebrain at E8.5. (c) Shh
expression is downregulated in the basal hypothalamus at E10.5. (d) Shh expression at E12.5.
Asterisk (), basal hypothalamus; CTX, presumptive cortex; FB, forebrain; HB, hindbrain; MB,
midbrain; os, optic stalk; pm, prechordal mesoderm; PTh, prethalamus; Th, Thalamus; TL, telence-
phalic Shh expression domain (modified from Blaess et al. 2014 and Haddad-Tóvolli et al. 2015)
2 Sonic hedgehog in Hypothalamus Development 37

ventral domain without Shh expression originates, while a lateral band of Shh
expression (the Lateral Progenitor Domain; Sect. 2.3) appears (Fig. 2.2c). This is
the “intriguing” and “dynamic” change in pattern that first attracted attention toward
the role of Shh in the presumptive hypothalamus. The mechanistic explanation for it
was obtained in the chick embryo and revealed one fundamental role of Shh
(Manning et al. 2006) (Sects. 2.6.4.1 and 2.7.2). The lateral band of expression
has been called the “hypothalamic basal plate” (as opposed to the “hypothalamic
floor plate” where Shh was originally expressed). In chick embryos, these cells
migrate “forward” from the diencephalon–mesencephalon boundary (Manning
et al. 2006). A telencephalic domain of Shh expression appears at this stage. This
is the Telencephalic Progenitor Domain and generates neurons for the cortex and
basal ganglia as well as astrocytes for the preoptic region (Sect. 2.3).

At E10.5 (Fig. 2.2c), Gli2 expression is now absent in the presumptive hypothala-
mus. Gli3 and Shh-activation diagnostic marker Gli1 overlap in the medial domain,
indicating an activator function of Gli3 (Gli3A) in the midline at this stage.

At E12.5 the general pattern of Shh in the forebrain does not change; however, the
telencephalic domain becomes a clear separate domain (Fig. 2.2d). Shh expression in
the forebrain peaks at this stage and starts fading, so that after E13.5 it has almost
disappeared. The Lateral Progenitor Domain generates cells also at this stage (Sect.
2.3).

2.3 Shh-Expressing Progenitor Domains in the Hypothalamus

and Their Derivatives

2.3.1 Shh Lineage and Fate-Mapping Strategy

The neuroepithelial cells expressing Shh are in principle stem cells, the cells that
divide to originate neurons and glia for at least part of the hypothalamus. Which
brain cells are born from Shh-expressing neuroepithelium in the ventral forebrain,
i.e., which hypothalamic cells belong to the Shh lineage? (Alvarez-Bolado et al.
2012). In order to trace this cell lineage in mice, a mouse line carrying two different
alleles was generated (Legué and Joyner 2010). The first allele encoded the
recombinase Cre fused to the estrogen receptor (CreER) under the control of
regulatory sequences of Shh (ShhCreER); the corresponding fusion protein remains
in the cytoplasm, not being able to enter the nucleus if not bound to a specific ligand,
the synthetic drug tamoxifen. The second allele, inserted in a permanently active
locus, encodes an easily identifiable and specific gene (e.g., β-galactosidase
a.k.a. LacZ, or enhanced yellow fluorescent protein, EYFP) protected from activa-
tion by a floxed stop cassette. In mouse embryos carrying these two alleles, every
cell producing SHH also produces the protein CreER. Approximately within 6 h of
tamoxifen administration by injection, CreER translocates into the cell nucleus,
where the Cre recombinase remains active for 24 h, enough time to eliminate the
38 G. Alvarez-Bolado

floxed stop cassette so that the cell (and its progeny) will permanently express
β-galactosidase (or EYFP). To label developmental events in the mouse, tamoxifen
is administered as an intraperitoneal injection to pregnant females at the desired
embryonic day; injection at E8.5 (TM8.5) guarantees labeling of any cell expressing
Shh between E9.0 and E10.0 and its progeny (lineage labeling).

2.3.2 The Medial Progenitor Domain (Fig. 2.3a, b)

Cells derived from Shh-expressing progenitors (“Shh-derived cells”) marked

between E8.0 and E9.0 (TM7.5) are found in the notochord and the prechordal
plate and also, in very small numbers, in the midline of the mammillary and caudal
tuberal region at E12.5. In the adult, these cells are found mostly in the mammillary
body, with some in the posterior hypothalamus. Shh-derived cells from progenitors
marked 1 day later (between E9.0 and E10.0, TM8.5; Fig. 2.3a, b) are very
numerous. They occupy the entire mammillary region, the most ventral and medial
portions of the tuberal region including most of the arcuate nucleus and the ventral
and medial parts of the ventromedial nucleus (Fig. 2.3e, g) as well as the infundibu-
lum and many median eminence tanycytes. This territory extends rostrally to the
Retrochiasmatic Area, a small ventral and medial region between optic chiasm and
arcuate nucleus representing the rostral tip of the Shh domain of expression in the
hypothalamic ventral midline (Puelles et al. 2004; Shimamura et al. 1995). In
summary, during the early and middle expression stages, Shh-derived hypothalamic
cells form a caudal domain, which originates in the mammillary region and expands
rostrally and ventrally into the ventral tuberal region. This distribution indicates an
expansion of the Shh-expressing progenitor domain from the mammillary recess to a
large, ventral, and caudal domain including all ventral midline neuroepithelium of
the mammillary and tuberal regions (Fig. 2.3a, b, e, g). These progenitors form the
Medial Progenitor Domain and co-expresses Nkx2-1 (Dale et al. 1997). This domain
is dependent on GLI2 acting as an activator (Gli2A) (Sects. 2.4 and 2.5). The Medial
Progenitor Domain is largely restricted to the Nkx2-1/Six3-positive domain in the
tuberal and mammillary regions at E12.5.

2.3.3 The Lateral Progenitor Domain

Beyond E9.5, the Medial Progenitor Domain does not express Shh anymore, while
two novel expression domains appear (Fig. 2.2c). One of them, the Lateral Progenitor
Domain, forms a paired lateral band spanning the tuberal and mammillary regions;
these bands meet rostrally and ventrally immediately behind the optic fissure
(Fig. 2.2c). This Shh expression domain has been called the hypothalamic basal
plate. Analysis of mouse mutants shows that the formation of this domain is depen-
dent on Shh expressed by the neuroepithelium (Sect. 2.5; Fig. 2.5g–i). From TM9.5 to
TM12.5, the fate-mapped cells are found in the Dbx1 expression domain in the
tuberal hypothalamus, with the dorsal-most fate-mapped cells in the hypothalamus
2 Sonic hedgehog in Hypothalamus Development 39

Fig. 2.3 Lineage of Shh in the mouse hypothalamus. (a) Sagittal section through the E18.5 mouse
hypothalamus labeled for Shh lineage (X-gal reaction, blue) after tamoxifen injection at E8.5 to
identify cells derived from the Medial Progenitor Domain. (b) Summary of A (Medial Progenitor
Domain) in the adult mouse hypothalamus. Insets on the right side: Transverse sections of E12.5
mouse brains showing the distribution of the fate mapped cells at four different rostrocaudal levels.
(c) Right side of hemi-dissected adult brain labeled for Shh-lineage (X-gal reaction, blue), after
tamoxifen injection at E11.5 (Lateral Progenitor Domain). (d) Summary of C (Lateral Progenitor
Domain) in the adult mouse hypothalamus. Insets on the right side: transverse sections of E12.5
40 G. Alvarez-Bolado

abut the Sim1-positive domain localized to the anterior region. In addition, the
Telencephalic Progenitor Domain starts to become obvious rostral to the eye primor-
dium (Sects. 2.3.4 and 2.7.4). At this stage, the predominant source of Shh in the
presumptive hypothalamus is the forebrain neuroepithelium (as opposed to the
prechordal plate, where Shh expression dwindles after E10.0); this agrees with the
loss of the lateral hypothalamus in neural Shh mutants (Szabó et al. 2009) (Sect. 2.5).
GLI2 is probably the predominant Gli activator in this domain, but in Gli2-deficient
mice GLI3 can partially substitute for it (Sect. 2.4). Finally, beyond E12.0 and up to
E14.0 at least (Sect. 2.2), numerous cells are born for the lateral hypothalamus of the
tuberal region and for the ventromedial nucleus (Fig. 2.3f). Although the general
pattern at this stage is similar as in the previous, the total number of cells generated is
smaller, and most of them are astrocytes. In the adult brain, cells originated in the
Lateral Progenitor Domain are found mostly in the tuberal region at all mediolateral
levels except the ventral midline (Fig. 2.3c, d, f, g). The mammillary, supra-
mammillary, and premammillary dorsal and ventral nuclei originate mostly in the
Medial Progenitor Domain (although part of the posterior hypothalamus derives from
the Lateral Progenitor Domain). As is the case with the Medial Progenitor Domain,
the Lateral produces only a few, scattered cells for the anterior region.

2.3.4 The Telencephalic Progenitor Domain

TM12.0–TM14.0 embryos show Shh-derived cells in the preoptic region, which are
mostly astrocytes (identified on the basis of morphology and GFAP expression).
Only very few preoptic neurons are derived from Shh-expressing progenitors; the
large majority of the neurons of the hypothalamic preoptic region does not belong to
the Shh lineage (i.e., are not derived from the Telencephalic Domain of Shh
expression in the mouse) (see Sect. 2.7.4 for a discussion).

Fig. 2.3 (continued) mouse brains showing the distribution of the fate-mapped cells at four
different rostro-caudal levels. The lineage of the telencephalic Shh domain in the hypothalamus
(mostly astrocytes) is labeled in red. (e, f) Diagrams of a transverse section through the adult mouse
tuberal region showing the distribution of fate-mapped cells from the Medial (e) and Lateral (f)
Progenitor Domains. (g) Merged diagrams from (e) and (f) showing clear differential distribution of
early (blue) and late (yellow) fate-mapped cells on the ventral side; at dorsal levels, the respective
patterns of distribution are less clearly demarcated. ac, anterior commissure; ANT, anterior region;
ARH, arcuate nucleus; cc, corpus callosum; DMH, dorsomedial nucleus; LH, lateral hypothalamus;
LM, lateral mammillary nucleus; MAM, mammillary region; ME, median eminence; MM, medial
mammillary nucleus; mth, mammillothalamic tract; ox, optic chiasm; PMD, dorsal premammillary
nucleus; PMV, ventral premammillary nucleus; POA, preoptic region; PVH, paraventricular
nucleus; RCH, retrochiasmatic area; SuM, supramammillary nucleus; VMH, ventromedial nucleus.
In b, d–g, closed circles indicate neurons, open circles indicate astrocytes, and stars indicate
tanycytes (modified from Alvarez-Bolado et al. 2012)
2 Sonic hedgehog in Hypothalamus Development 41

2.3.5 Conclusions

2.3.5.1 Differential Patterns of Neurogenesis Build Different

Hypothalamic Nuclei
Since the area expressing Shh is very extensive, Shh-expressing hypothalamic
progenitors do not show specific timing and rate of neurogenesis. Rather, different
neurogenesis strategies are used which are typical for a given hypothalamic nucleus.
This indicates that there are different Shh-expressing progenitor subpopulations
generating neurons and glia according to specific programs. For instance, the
dorsomedial nucleus and the lateral hypothalamus receive from the progenitors a
slow and steady contribution of scattered cells from TM8.0 through TM12.5. The
ventromedial nucleus, on the contrary, receives in a shorter time large numbers of
neurons, which soon aggregate forming a recognizable entity. Additionally, neurons
derived from progenitors expressing Shh after E10.5 or even after E11.5 (late-born
cells), settle at all mediolateral levels of the tuberal hypothalamus. This indicates
that, as a general rule, hypothalamic neurogenesis proceeds in a more complex way
than three consecutive mediolateral waves generating the periventricular, medial,
and lateral hypothalamic zones (the “three wave” hypothesis (Altman and Bayer
1986)).

2.3.5.2 The Preoptic and Anterior Regions

In the mouse preoptic region, the great majority of Shh-derived cells are astrocytes,
and in the anterior region there are almost no Shh-derived cells at all. This means that
neither the Medial nor the Lateral Progenitor Domains produce cells for the preoptic
and anterior regions. Also, neither the preoptic nor the anterior regions receive
Shh-derived neurons from caudal regions, that is, Shh-derived cells do not cross
the alar–basal boundary. In this sense, the two “rostral” (alar) regions of the adult
hypothalamus derive from the alar portion of the neural tube (which includes the
telencephalon). This notwithstanding, both the preoptic and the anterior regions
depend on Shh signaling (see below Sects. 2.5, 2.7.3, and 2.7.4).
Looking at the early Shh expression pattern (Fig. 2.2) it seems obvious that any
part of the hypothalamus not derived from the Medial or the Lateral Progenitor
Domains has to originate in the telencephalon. Expression of transcription factor
Foxg1 characterizes telencephalic progenitors (Xuan et al. 1995). The scarce data that
we have on Foxg1 lineage in the hypothalamus shows intriguing images of labeled
cells in large parts of the anterior region, including the paraventricular nucleus
(Kawaguchi et al. 2016). Foxg1-null heterozygous mutants show abnormal expres-
sion of Oxt and Avp, two neuropeptide genes specific of the anterior region (Frullanti
et al. 2016). Detailed analyses on chick embryos have uncovered a detailed mecha-
nism for the anterior region in that model (Fu et al. 2017) (see below, Sect. 2.7).
42 G. Alvarez-Bolado

2.4 GLI Factors in the Shh Progenitor Domains

of the Hypothalamus

The secreted protein SHH exerts its influence through transcription factors GLI1,
2, and 3, of which GLI2 and GLI3 can act as activators (GliA) or repressors (GliR).
By comparing the hypothalamic phenotypes of mutant mice deficient in Gli2
(Gli2zfd/+, Gli2 Zinc Finger-Deleted; (Mo et al. 1997)), Gli3 (Gli3Xt-J/+, Extra-
Toes; (Hui and Joyner 1993)), or neural Shh and Gli3 (Foxb1-Cre;ShhloxP/+;
Gli3Xt-J/+; see Sect. 2.5), it is possible to deduce which factors act as GliA or GliR
in the Medial and Lateral Progenitor Domains and if there are differences between
them (Haddad-Tóvolli et al. 2015). We have tried to summarize a complex set of
mutant phenotypes in the diagrams in Fig. 2.4.

2.4.1 The Only Role of Shh in the Alar Regions Is to Suppress Gli3R

The preoptic and anterior regions (alar hypothalamus) show strong expression of
Gli3; despite this fact, the overall specification of these regions is not altered in the
Gli3 mouse mutant brain. It could be that, in the alar portions of the hypothalamus,
Gli2A and Gli3A can substitute for each other. Alternatively, in these regions, the
only role of Shh would be to suppress Gli3R; this would agree with observations that
in double Shh/Gli3 mutant mice the dorsoventral patterning of the telencephalon is
appropriate (Rallu et al. 2002), and supports the hypothesis that both the preoptic and
the anterior regions are telencephalic in origin (Sect. 2.7.4.2). Neither Gli2 nor Gli3
is required for the general patterning of the alar (anterior and preoptic) regions of the
hypothalamus.

2.4.2 Differential Gli Requirements Between Medial and Lateral

Progenitor Domains

In the tuberal and mammillary regions (the basal hypothalamus), Gli2 is required for
the development of the Medial Progenitor Domain (Fig. 2.4a) and its derivatives:
median eminence, pituitary, most of the arcuate nucleus and ventromedial nucleus,
and the mammillary and premammillary nuclei. The results show that the Medial
Progenitor Domain requires nonneural Shh acting through Gli2A (this is in agree-
ment with the phenotype of mice deficient in neural Shh (Shimogori et al. 2010;
Szabó et al. 2009)) (Sect. 2.5). In contrast, the Lateral Progenitor Domain does not
require Gli2 (Fig. 2.4a): the lateral part of the ventromedial and the arcuate nuclei as
well as the dorsomedial nucleus and the lateral hypothalamus. This could mean that
GliA is not required by the Lateral Progenitor Domain. More likely, however, is that
Gli3A is able to compensate in the Gli2zfd/zfd mutant. Gli1 expression (diagnostic of
Shh pathway activation) and Gli3 expression overlap in the Lateral Progenitor
Domain of the Gli2zfd/zfd mutants. In mutants deficient in neurally expressed Shh,
Lateral Progenitor Domain derivatives disappear (Fig. 2.4a) indicating that Shh is
2 Sonic hedgehog in Hypothalamus Development 43

Fig. 2.4 Differential contribution of Gli factors to the Lateral and Medial Progenitor Domains. (a)
Summary diagrams of progenitor domains (neuroepithelium) of the basal hypothalamus in wild
type and mutants as deduced from the phenotypic analysis in the present study. (b) Contribution of
Gli proteins to the specification of the medial and lateral progenitor domains in three successive
stages of development. Colored dots in (a) and (b) represent different kinds of Shh-expressing
progenitors, as indicated on the figure; dotted squares in (b) indicate that, although nonneural Shh
determines the lateral domain through repression of Gli3, a direct influence of Shh through Gli2A
on lateral progenitors before E8.5 is in principle possible (since there are some lateral progenitors in
the double mutant deficient in neural Shh and Gli3). The asterisk () means that loss of GliA2 could
be compensated by Gli3A. MBO, mammillary body (modified from Haddad-Tóvolli et al. 2015)

needed to counter Gli3R. Therefore, the Lateral Progenitor Domain is patterned by

neural Shh blocking Gli3R formation and either promoting Gli2A (in wild-type
animals) or Gli3A (in Gli2-deficient mutants). In agreement, mouse mutants defi-
cient in Shh expressed by the neuroepithelium do not develop a Lateral Expression
Domain (Fig. 2.5g–i) and lack a lateral hypothalamus (Szabó et al. 2009) (Sect. 2.5).
44 G. Alvarez-Bolado

Fig. 2.5 Phenotype of mouse mutants deficient in neural Shh. (a) In situ detection of Shh
expression with a probe detecting exon2, essential for SHH activity; E12.5 wild-type mouse
embryo. (b) Diagram of Shh expression (red); E12.5 Shh cKO-Nkx2-1 mouse embryo. (c) In situ
detection of Shh-exon2; E12.5 Shh cKO-Foxb1 embryo. (d) Diagram of Shh expression (red);
sagittal section of an E12.5 wild type mouse embryo, showing forebrain regions of interest. (e)
Diagram of Shh expression (red); sagittal section of an E12.5 Shh cKO-Nkx2-1 mouse embryo,
showing forebrain regions of interest. (f) Diagram of Shh-exon2 expression (red) on the right half of
the brain of an E12.5 Shh cKO-Foxb1 mouse embryo. (g) In situ detection of Shh-exon2 expression;
E10.5 wild-type mouse embryo. (h) In situ detection of Shh, either intact or truncated (long probe;
see text for details) expression; E10.5 5 Shh cKO-Foxb1 mouse embryo. (i) In situ detection of Shh-
exon2 expression; E10.5 Shh cKO-Foxb1 mouse embryo. (j, k) In situ detection of Lhx1
2 Sonic hedgehog in Hypothalamus Development 45

2.4.3 Suppression of Gli3R by Shh Is Also Essential in the Medial

and Lateral Progenitor Domains

Gli3 is not required for the overall specification of the hypothalamus in wild-type
mice (although Gli3A is specifically required for mammillary proliferation). How-
ever, the ventral forebrain phenotypes of Shh mutants, be it the full Shh null mutant
(Chiang et al. 1996) or the conditional Shh mutants lacking Shh expression in the
presumptive hypothalamus (“neural Shh mutants”) (Shimogori et al. 2010; Szabó
et al. 2009) (Sect. 2.5), are much more severe than the Gli2zfd/zfd mutant phenotype.
This indicates that suppression of Gli3R function by Shh (of neural or prechordal
origin) is required for normal hypothalamic development. In support of this conclu-
sion, the defects in the Medial and Lateral Progenitor Domains found in neural Shh
mutants are largely rescued in double neural Shh –/–; Gli3 –/– mice (Fig. 2.4a).

2.4.4 Orexin (Hcrt)-Expressing Neurons of the Lateral

Hypothalamus Are Particularly Dependent on GliA

The number of Hcrt-expressing neurons in the lateral hypothalamus is reduced in

Gli2zfd/zfd mutants and completely lost in neural Shh mutants and in double neural
Shh –/–; Gli3 –/– mutants. These results uncover a particular dependence of the
progenitors of Hcrt neurons on GliA, maybe because they need GliA later than other
progenitors in the Lateral Progenitor Domain, or for a longer time.

2.5 Role of Shh in Hypothalamic Development: Mouse Mutant

Phenotypes

Box 2.2: From the Fruit Fly Larva to the Mouse Brain: The Pathway
Leading to the Role of Shh in Hypothalamic Specification
The first hedgehog gene (hh) was described in the fruit fly as a gene involved
in larval segmentation (Nüsslein-Volhard and Wieschaus 1980). Three
homologs of hh were later found in chick, zebrafish, and mouse (Echelard

(continued)
ä

Fig. 2.5 (continued) expression; sagittal sections of E18.5 wild type or Shh cKO-Foxb1 mouse
brains. Arrow in b, c, e, f shows remnant Shh expression domains in the conditional mutants.
Arrows in g, h, and i show the lateral Shh expression domain or its theoretical location (black arrow)
and the telencephalic Shh expression domain (white arrow). Asterisk in e, f, and k marks the
abnormal size of the third ventricle in the mutants. ANT, anterior region; CTX, cortex; EmTh,
thalamic eminence; MAM, mammillary region; MBO, mammillary body; PTh, prethalamus; SCH,
suprachiasmatic nucleus; Th, thalamus; TUB, tuberal region (b, d, e with data from Shimogori et al.
2010)
46 G. Alvarez-Bolado

Box 2.2 (continued)

et al. 1993; Krauss et al. 1993; Riddle et al. 1993). One of them, which
received the name of Sonic hedgehog (Shh), was expressed in the floor plate
of the neural tube and in the underlying axial mesoderm (notochord). Thanks
to experiments on cultured embryo explants in the frog, rat, and chick model
species, it was known that these two structures were essential for proper
development of the ventral spinal cord (reviewed by Ruiz i Altaba and Jessell
1993). Ectopic expression experiments in transgenic mice suggested that Shh
was responsible for this role (Echelard et al. 1993), and experiments on
cultured explants of embryonic rat floor plate and chick spinal cord showed
that Shh can specify ventral neural cell types (Roelink et al. 1994). Spectacular
confirmation of the essential role of Shh in cell fate specification in the ventral
neural tube at all rostro-caudal levels was provided by the phenotype of mouse
embryos carrying null mutations of the Shh locus (Chiang et al. 1996).
However, these mutants could not reveal the relative contributions to the
hypothalamic development of Shh expressed by the neural tube versus Shh
expressed by the prechordal plate. This question was addressed by means of
conditional mutants more than a decade later (Szabó et al. 2009; Shimogori
et al. 2010). Shortly after, the same approach was used to show that neural Shh
is required for the appropriate development of hypophysis and eye (Zhao et al.
2012) and to specify glial and ependymal cells in the spinal cord (Yu et al.
2013).

The first Shh null mutant mouse (Chiang et al. 1996) confirmed the ventralizing role
of Shh in the hypothalamus and extended it to the forebrain. However, the phenotype
of this mutant was dominated by extreme holoprosencephaly, so that the entire
hypothalamus together with most of the forebrain were missing. Obviously, Shh
expression in the prechordal plate (“prechordal Shh”) was indispensable for fore-
brain development. Conditional mutants eliminating Shh expression from the neural
tube (“neural Shh”) but not the prechordal plate have been much more informative.
By “subtracting” the neural Shh phenotype we could also infer the role of prechordal
Shh. The first such neural mutant removed Shh expression in the forebrain by using
Foxb1 regulatory sequences as the Cre-driver (Foxb1-Cre  Shh fl/fl or Shh
cKO-Foxb1 (Szabó et al. 2009). The method was used again by using a Nkx2-1 as
Cre-driver (Nkx2-1-Cre  ShhloxP/loxP or Shh cKO-Nkx2-1 (Shimogori et al. 2010).
Further observations on the role of Shh in the ventral forebrain were made using a
Cre line driven by SBE2, an upstream regulatory element of Shh specific for the
hypothalamus (Zhao et al. 2012). Finally, mutants of transcription factor Six3 have
thrown light on the complex role of neural Shh in the splitting of the eye field and
telencephalic vesicle (Geng et al. 2008, 2016). These do not pertain directly to
hypothalamic development (except inasmuch as the retina can be considered a
hypothalamic derivative) but will be included here for the sake of completeness.
2 Sonic hedgehog in Hypothalamus Development 47

2.5.1 Summary of Mouse Mutant Phenotypes

According to the phenotype of mouse mutants deficient in Shh expressed in the

neural tube, neural Shh is:

1. Not necessary to confer basic hypothalamic identity. Prechordal Shh must pro-
vide general hypothalamic specification, although is not enough for full growth
and differentiation
2. Not necessary to specify the four hypothalamic regions and their positions
relative to each other
3. Essential for the hypothalamic regions to differentiate and express their charac-
teristic marker genes
4. Essential for the hypothalamus to reach appropriate size
5. Essential for the establishment of the Lateral Progenitor Domain and therefore for
the development of the lateral hypothalamus (Sect. 2.3). Prechordal Shh can
partially substitute for the Medial Progenitor Domain
6. Not essential for the specification of the neurohypophysis (but required for its
correct development)
7. Essential for the development of the optic disk
8. Not essential for the splitting of the eye field and telencephalic vesicle, although it
has a minor role in this process

2.5.2 The Shh cKO-Foxb1 and Shh cKO-Nkx2-1 Conditional Mutants

2.5.2.1 Comparison of Mutants

The two conditional mutants removing Shh expression from the presumptive hypo-
thalamus were generated using the same classical conditional strategy: crossing a
mouse line carrying a Cre allele under the regulatory sequences of a “driver” gene
expressed in an appropriate region, to a mouse line carrying an allele of Shh where
exon 2, essential for Shh activity, has been floxed (Dassule et al. 2000). These mouse
mutants express a truncated, inactive form of Shh in a specific portion of the
neuroepithelium. The mutant based on Foxb1 regulatory sequences as the Cre-driver
(Szabó et al. 2009) abolishes SHH activity in the entire forebrain (including the
presumptive hypothalamus and the zona limitans of the thalamus) as well as in the
midbrain with one exception: a restricted region with weak full length Shh expres-
sion remains in the telencephalic Shh expression domain (Fig. 2.5a, c, f, i). The
mutant hypothalamus was analyzed at E12.5 and at E18.5.The mutant based on
Nkx2-1 as driver (Shimogori et al. 2010) abolishes neural SHH activity in the
presumptive hypothalamus and also in the telencephalic domain. Two SHH sources,
the zona limitans (dorsal) and the midbrain (caudal) are not affected in this mutant
(Fig. 2.5a, b, e). The mutant hypothalamus was analyzed at E10.5 and E12.5.
By using ISH probes against exon2, expression of active Shh can be detected,
while a “long” probe based on other Shh sequences detects all Shh expression in the
mutant (for instance, compare panels h and i from Fig. 2.5). The results found with
48 G. Alvarez-Bolado

both mutants agree, and some differences (but no contradictions) found are probably
due to the Cre-driver genes used having differential expression and leaving different
Shh expression domains intact (either the telencephalic domain or the midbrain
domain), which in turn could partially rescue the wild-type phenotype. In addition,
some of the conclusions reached may be reflective of the different ages that were
analyzed.

Neural Shh is not required to specify the hypothalamus as a region Transcription

factor Nkx2-1 is specifically expressed in a ventral forebrain region encompassing
the presumptive hypothalamus (Shimamura et al. 1995) and is a major determinant
of hypothalamic fate and the first ventral forebrain marker to be expressed during
development. The hypothalamus without neural Shh expresses Nkx2-1, implying
that expression of Nkx2-1 depends on prechordal Shh (see below, Sect. 2.6). The
hypothalamus without neural Shh does not express (or only in extremely reduced
domains) Dlx2 and Dbx1. In both mutants, the identity of the hypothalamus is
respected (Fig. 2.5d–f), that is, a dorsalizing influence did not overtake the presump-
tive hypothalamic region. This is based on the expression of Foxg1, which does not
expand into diencephalon, and the expression of the transcription factor Rax, key for
tuberal specification (Shimogori et al. 2010). In addition, the hypothalamic regions,
as much as they can be recognized, are found in their appropriate anteroposterior
(or alar–basal, see Sect. 2.1.2) positions.

2.5.2.2 Hypothalamic Regional Phenotypes (See Also Sect. 2.7)

In both the Shh cKO-Foxb1 and the Shh cKO-Nkx2-1 mutants, the mammillary
region is the least affected region. At E18.5, it expresses appropriate markers such as
Sim1, Pitx2, Nkx2-1, Lhx1, Lhx5, and Emx2, but not Foxb1, Otp. It is reduced in size,
but possesses a small axonal bundle in the place of the mammillotegmental tract
(Szabó et al. 2009) (Fig. 2.5j, k). Shimogori et al. (2010) report a reduction of the
premammillary marker Lef1 at E12.5, while Szabó et al. (2009) found that the
premammillary region was present at E18.5 (Fig. 2.5j, k). The tuberal region is
more affected in size and differentiation than the mammillary: characteristic markers
of this region like Nr5a1, Pomc, and Nkx6-2 were lost at E12.5 in the Shh cKO-
Nkx2-1 mutants. At E18.5, Calb expression identified a ventromedial nucleus
reduced in size in the Shh cKO-Foxb1 mutant. The hypothalamic anterior region is
also reduced in these mutants. Only a few scattered cells are left that express
neuropeptide Avp (a marker of the paraventricular nucleus, possibly the most
recognizable nucleus of the hypothalamus). Also, there is attenuated expression of
characteristic transcription factors Otp and Sim1 (which label the paraventricular
nucleus during development). None of the studies analyzes the preoptic region in
depth, although in the Shh cKO-Nkx2-1 it appears extremely reduced in size at
E12.5.
2 Sonic hedgehog in Hypothalamus Development 49

2.5.2.3 General Growth and Dorsoventral Patterning in Neural Shh

Mutants
Mutant analysis at E18.5 showed that an important function of neural Shh is to
stimulate growth and, in this way, coordinate dorsoventral and mediolateral pattern-
ing (Cayuso et al. 2006; Ishibashi and McMahon 2002; Sagner and Briscoe 2017).
Accordingly, the Nkx2-1 territory (presumptive hypothalamus) is reduced in the Shh
cKO-Foxb1 mutant by E12.5, and both neural Shh mutants show a much smaller
hypothalamus. At E18.5, growth has allowed for some recovery of early patterning
defects (Szabó et al. 2009) although in this mutant, remnant telencephalic Shh
expression may be contributing to this partial rescue (Sect. 2.5.2.1). The growth
defect appears as a large gap (asterisk in Fig. 2.5e, f, k) in the mutant forebrain
between thalamus and hypothalamus. Intriguingly, formation of the lateral Shh
expression domain (Lateral Progenitor Domain) appears to depend on neural Shh
and therefore that domain is absent in the mutant (black arrows in Fig. 2.5g–i).
Additionally, a neuroepithelial region termed “intrahypothalamic diagonal” and
expressing Lhx6, Arx, and Nkx2-1 does not develop in the absence of neural Shh
(Shimogori et al. 2010). This suggests that tissue growth stimulated from the lateral
domain has a role in “stitching together” the alar and basal portions of the forebrain.
Finally, the gap between alar and basal forebrain is larger in the Shh cKO-Foxb1
mutant because of the disappearance of the Shh expression domain of the zona
limitans, which is key for growth of the Prethalamus and Thalamic Eminence
(Fig. 2.5d–f).

2.5.2.4 Mediolateral Patterning in Neural Shh Mutants

The Medial Progenitor Domain is probably present in these mutants because it can
be partially rescued by prechordal Shh acting through Gli2 (Sect. 2.4). The Lateral
Progenitor Domain is absent in the neural Shh mutant (Fig. 2.5g–i). This leads to loss
of a large part of the lateral hypothalamus; accordingly, marker genes of the lateral
hypothalamic area are reduced to very few scattered cells (Pmch), or not present at
all (Hcrt) (Szabó et al. 2009).

2.5.3 Hypophysis, Eye, and Holoprosencephaly Phenotypes (See

Also Sects. 2.6.3.2 and 2.7.2)

Analyzing mutant mouse lines carrying mutations that decrease Shh in the
neuroepithelium directly or indirectly, have noticed an important role of Shh in
hypophysis development. Zhao et al. (2012) used an SBE2-Cre line to abolish Shh
expression in the presumptive hypothalamus and observed an ectopic, dysmorphic,
and small infundibulum (primordium of the posterior hypophysis) as well as a
duplicated and not completely differentiated Rathke’s pouch (primordium of the
anterior hypophysis). Failure to form the infundibulum as well as a multiplicity of
Rathke’s pouches have been described in mice mutant for Rax, a transcription factor
activated by Shh and, in turn, essential to maintain Shh expression in the presumptive
hypothalamus (Orquera et al. 2016). In mice carrying null mutations of transcription
factor Tbx3, Shh expression abnormally persists in the tuberal region beyond E9.5.
50 G. Alvarez-Bolado

In these embryos, the infundibulum does not develop and Rathke’s pouch shows
decreased proliferation (Trowe et al. 2013). In conditional mutants of Shh using
Hex1 as a Cre driver, Shh expression in the anterior hypothalamus is lost. These
mutants do not develop an infundibulum or Rathke’s pouch (Carreno et al. 2017).
SBE2-Cre  ShhloxP/loxP mice lack Shh in the presumptive hypothalamus and show
lack of formation of the optic disk (the point of exit of the retinal axons from the eye)
and a hypoplastic optic nerve (Zhao et al. 2012). Early in the development of the
mammalian head, the prechordal plate is essential for splitting of the eye field and the
primary (unpaired) telencephalic vesicle into left and right. Failure of this process
leads to a major phenotype, holoprosencephaly (cyclopia). Some mouse mutants in
transcription factor Six3, expressed in the rostral portion of the ventral telencephalon,
show holoprosencephaly due to deficiency in Shh in that expression domain (Geng
et al. 2008, 2016). However, none of the three neural Shh mutants (Shimogori et al.
2010; Szabó et al. 2009; Zhao et al. 2012) shows holoprosencephaly.

2.6 Interactions of Shh with Other Genes in the Developing

Hypothalamus

2.6.1 Prechordal Shh and Bmp Genes

Experiments on chick embryos have shown that during gastrulation, the prechordal
plate migrates rostrally after detaching from the node. At the same time, the
presumptive hypothalamic floor plate cells migrate rostrally, at times in register
with the prechordal plate, and are influenced by SHH and BMP ligands secreted by
it. If SHH confers “floor plate” identity, the BMPs bestow the hypothalamic floor
plate with “rostral” identity, so that they become “hypothalamic floor plate,” char-
acteristically expressing not only Shh but Nkx2-1 and Bmp7 (Dale et al. 1997; Patten
et al. 2003; Placzek and Briscoe 2005) (summarized in Fig. 2.6a). In experimental
conditions, SHH can by itself rostralize the forebrain floor plate (without the help of
BMP), but only at very high, nonphysiological concentrations (Dale et al. 1997).

2.6.2 Wnt-Based Anteroposterior Patterning of the Neural Plate

Shh secreted by the prechordal plate acts on the overlying neural tube already
patterned along the anteroposterior axis. This early patterning depends on a balance
between the posteriorizing influence of Wnt ligands and the anteriorizing influence
of Wnt antagonists (like Fgf8 from the anterior neural ridge (Shimamura and
Rubenstein 1997)). As a consequence, the forebrain is divided into an anterior
domain expressing transcription factor Six3 and a posterior one expressing Irx3,
meeting each other at the level of the zona limitans (Braun et al. 2003; Kobayashi
et al. 2002). Repression of Wnt1 by Six3 is required to maintain the forebrain
territory and therefore the hypothalamus (Lagutin et al. 2003). Wnt antagonism is
also essential for the ventral forebrain to acquire hypothalamic character (Kapsimali
et al. 2004).The caudalizing Wnt signals (Braun et al. 2003) have to be consistently
2 Sonic hedgehog in Hypothalamus Development 51

Fig. 2.6 Diagrams depicting Shh cooperating and interacting with other genes in the developing
hypothalamus. (a) In most of the neural tube, SHH (orange-colored arrows) from the notochord
(dashed orange-colored line) ventralizes the floor plate (solid orange-colored line), which then
continues secreting SHH; BMP ligands (blue line and arrows) act as dorsalizers. Rostral to the
midbrain–forebrain boundary (dashed black line), the prechordal plate (red dashed line) secretes
SHH plus BMP7, which cooperate to induce the hypothalamic floor plate (red line) (data from Dale
et al. 1997). (b) Shh expression by the hypothalamic floor plate (red) requires Nkx2-1 expression
(pink; at least indirectly, dashed pink arrow) as well as the direct influence (continuous arrows) of
SIX3 (blue) and SOX2/3 (green) non-Shh regulatory sequences. SHH, in turn, maintains Six3 and
Nkx2-1 expression (dashed red arrows) (data from Geng et al. 2008; Sussel et al. 1999; Zhao et al.
2012). (c) The striking change in Shh expression pattern (red) between the middle stage (top left)
and the late stage (bottom right) depends on early secretion of BMP ligands from the prechordal
plate (dashed red line), which activate expression of Bmp genes in the presumptive hypothalamus,
leading to expression of Tbx transcription factors which directly repress Shh expression in a
restricted area (asterisk) (data from Manning et al. 2006; Trowe et al. 2013). (d) Top: The Shh
(red) and Rax (blue) expression domains in the midline of the presumptive hypothalamus by E11.5
overlap in a domain (brown) of the retrochiasmatic area (RCH or ABasM). Bottom: In the zebrafish,
Rax homolog rx3 is activated by Shh (red arrows) in order to confer specific neuroendocrine cell
fates to progenitors; then rx3 is repressed by Shh influence (red lines). The interaction is partially
52 G. Alvarez-Bolado

inhibited in order to achieve and maintain anterior character in the hypothalamic

floor plate (Kapsimali et al. 2004). The early anteroposterior patterning and the need
to antagonize Wnt signaling in order to obtain an anteriorized floor plate (i.e., the
hypothalamic floor plate) become evident in several of the interactions of Shh with
other genes in the ventral forebrain (see below).

2.6.3 Interactions Necessary for Activation of Shh Expression

in the Hypothalamic Floor Plate (Fig. 2.6b)

The DNA regulatory sequence necessary to confer specific hypothalamic Shh

expression does not have binding sites for Gli transcription factors (which act
downstream Shh) (Jeong et al. 2006), indicating that prechordal SHH does not
induce Shh expression by the overlying neuroepithelium directly. Instead, the
expression of a number of other transcription factors in the ventral forebrain
neuroepithelium is necessary to finally activate expression of Shh in the presumptive
hypothalamus at the right time, in the right place, and with the right intensity. As far
as we know, these transcription factors are Nkx2-1 (whose expression is dependent
on prechordal Shh (Dale et al. 1997)), Sox2 and Sox3 (Zhao et al. 2012), and Six3
(Geng et al. 2008).

2.6.3.1 Reciprocal Interactions of Nkx2-1 and Neural Shh

Transcription factor Nkx2-1 is a major determinant of hypothalamic fate and the first
specific hypothalamic marker to be expressed during development. At E9.5, mutant
mice deficient in neural Shh expresses Nkx2-1 similar to wild-type controls. At E11.5,
however, attenuated Nkx2-1 expression and the entire hypothalamic territory is
reduced (Szabó et al. 2009). This indicates that neural Shh is essential to maintain
appropriate levels of Nkx2-1 expression in the presumptive hypothalamus, as well as
for the growth of the region. Mice deficient in Nkx2-1 do not express Shh in the
forebrain at early stages (Sussel et al. 1999). Since the phenotype of Nkx2-1 mutants
(Kimura et al. 1996) is different and milder from that of neural Shh mutants (Sect.
2.5), we assume that Nkx2-1 is not necessary for neural Shh expression, only for its
timely onset. The complexity of the relation between these genes is illustrated by
reports that in Nkx2-1 mutants, the midline territory of expression of Shh in the
presumptive hypothalamic neuroepithelium is expanded; in conditional mutants
abolishing functional Shh expression in the Hesx1 ventral forebrain expression
domain, expression of Nkx2-1 is reduced along the anteroposterior axis (Carreno
et al. 2017); Nkx2-1 expression is maintained by neural Shh in the telencephalic
domain.

Fig. 2.6 (continued) conserved in the mouse (black arrows), where Shh is required to activate Rax
expression to accomplish a similar role, and Rax, in turn, maintains Shh expression at a very specific
point in time; repression of Rax by Shh is also conserved in the mouse (data from Lu et al. 2013;
Muthu et al. 2016; Orquera et al. 2016)
2 Sonic hedgehog in Hypothalamus Development 53

2.6.3.2 Interaction of Neural Shh with Six3

One important determinant of anteroposterior patterning is transcription factor Six3.
Six3 is expressed in the presumptive ventral forebrain, forming a boundary with Irx3
which represents posterior (Braun et al. 2003). Six3 confers “anterior” character to
the Shh-expressing forebrain ventral midline, so that Nkx2-1 is expressed only there.
Six3, however, is not upstream Nkx2-1, it just predisposes the cells to respond to
prechordal Shh (and possibly BMPs)(Kobayashi et al. 2002). Six3 is necessary to
maintain neural Shh expression in the early rostral domain in the ventral forebrain.
For this reason, Six3 is (in certain genetical backgrounds) critical to prevent
holoprosencephaly (Geng et al. 2008; Jeong et al. 2008). Neural Shh interacts with
Foxg1, Six3, Fgf8, Bmp4, and Wnt8b for splitting the eye field and the early,
unpaired telencephalic vesicle (Geng et al. 2016), and subtle alterations in gene
dosage or allelic variations of one or more genes can greatly affect the general
outcome. In agreement with this complexity, the cyclops phenotype appears only in
Six3 mutants of certain backgrounds and, in humans, this phenotype is not fully
penetrant and shows clinical variability (Roessler and Muenke 2010).

2.6.3.3 Sox2 and Sox3

Sox1, Sox2, and Sox3 form the SoxB1 subfamily of transcription factors. The
expression domain of Sox2 and Sox3 overlaps with that of Shh in the early presump-
tive hypothalamic neuroepithelium, and they are necessary for Shh expression to
reach appropriate intensity. They act through direct interaction with the long-range
Shh forebrain enhancer SBE2. This interaction is dose dependent, so that the dose of
SoxB1 factors bound to SBE2 is key, rather than the identity of the factors (Zhao
et al. 2012).

2.6.4 Interactions of Neural Shh with Other Transcription Factors

and Signaling Molecules

After the expression of Shh is activated in the presumptive hypothalamus, it activates

Rax expression (and is maintained by it). Repression of Shh by Tbx2 and Tbx3 in a
restricted domain is necessary for correct development of the infundibulum. Shh
maintains the expression of Bmp7 and/or Pax2 in the developing eye. Finally,
mutual repression of Fgf10 and Shh is emerging as crucial for the specification of
hypothalamic regions.

2.6.4.1 Tbx2 and Tbx3 (Fig. 2.6c)

The interaction of neural Shh with T-box transcription factors Tbx2 and Tbx3 is
particularly complex and interesting, because it helps explain the impressive change
in expression pattern undergone by neural Shh in the tuberal and mammillary
hypothalamus in the late stage (see Chap. 2) and illuminates the development of
the hypophysis. In chick embryo explants (Manning et al. 2006), experimental
blocking of Tbx2 expression by means of chordin (a BMP inhibitor) prevents the
repression of Shh expression; the resulting abnormal presence of Shh causes a
54 G. Alvarez-Bolado

cessation of cell proliferation and prevents the specific marker gene Emx2 to be
expressed in the presumptive mammillary region. In this model, it is not possible to
examine subsequent alterations in hypophysis development. In the mouse, Tbx2 is
not required for hypothalamic development, but Tbx3 shows the same interaction
with Shh as in chicken. Tbx2 and Tbx3 could act redundantly, but Tbx2 expression is
lost in the Tbx3 mutant. In the mouse, however, although Tbx3 is essential to repress
Shh expression in the presumptive tubero-mammillary region, the knockout of Tbx3
causes no obvious changes in differentiation (expression of Emx2 and Fgf10 are
normal) but prevents the formation of the infundibulum (presumptive neurohypoph-
ysis, Trowe et al., 2013). In addition, these studies showed that Tbx2 and 3 can
repress Shh expression in this region directly, through two mechanisms. First, both
TBX2 and TBX3 can bind to SOX2 (or SOX3) (necessary to activate Shh), in this
way making it unavailable to bind to the required Shh long-range enhancer SBE2.
Second, both can also specifically bind to long-range enhancer of Shh SFPE2
repressing Shh expression in this region again.

2.6.4.2 Rax (Fig. 2.6d)

“Retina and anterior neural fold homeobox” (Rax) is a transcription factor essential
for eye development (Mathers et al. 1997) and is also expressed in the presumptive
hypothalamic neuroepithelium. An interaction between this transcription factor and
Shh is essential for the specification of certain hypothalamic neurons in the anterior
and tuberal regions. This has been analyzed in detail in the mouse (Lu et al. 2013;
Orquera et al. 2016) and in the zebrafish (Muthu et al. 2016).
In the zebrafish, rx3 (Rax homolog) is first activated and then repressed by Shh
signaling, important in the patterning of the neuroendocrine system. Rx3 is essential
for the specification of the progenitors of neurons expressing either pomc, ff1b, or
avp. The pomc-expressing neurons are found in the periventricular tuberal nucleus,
functionally analogous to the arcuate nucleus of the tuberal region in mammals; ff1b
is a homolog of mammalian SF1 (Nr5a1), expressed in the ventromedial nucleus of
the tuberal region; avp is expressed in the neurosecretory preoptic area (functionally
analogous to the periventricular nucleus of the anterior region in mammals). Rx3 not
only specifies cell fate for those specific neuronal populations of the anterior/tuberal
region of the zebrafish hypothalamus, but is essential for their survival, proliferation,
and migration. At the same time, rx3 opposes the dorsomedial nucleus and the
neurohypophysis fates in anterior/tuberal progenitors. The effect of rx3 on the
specification of avp progenitors could be mediated by activation of otpb (Otp
homolog), which in zebrafish is essential for the specification of neurons in the
anterior/tuberal regions (Muthu et al. 2016).
In the mouse, Rax expression is activated by neural Shh and, in turn, Rax is
essential to maintain neural Shh expression in the most medial domain of the
presumptive hypothalamus at a very precise time-point (E8.5). Timed-conditional
inactivation experiments show that, after E8.5, Rax is dispensable for hypothalamic
Shh expression in the mouse (Orquera et al. 2016). The same report indicates that
there is an early requirement for Rax in the specification and proliferation of the
presumptive tuberal neuroepithelium. The effects of Rax inactivation on
2 Sonic hedgehog in Hypothalamus Development 55

hypothalamic development could be due to lack of neural Shh (since Rax is needed at
E8.5 to keep Shh activated from then on): reduced proliferation leading to thinning
of the midline forebrain neuroepithelium and/or reduced expression of proneural
genes Ascl1 and Ngn3. Additionally, upon losing neural Shh expression (in the Rax
mutant), the rostral expression boundary of markers like Otx2, Tbx3, and Fgf10
shifts rostrally. Perhaps as a consequence, ectopic Rathke’s pouches (presumptive
anterior hypophysis) are found in these mutants (same as in neural Shh mutants
(Zhao et al. 2012)) and the infundibulum is missing (see Sect. 2.7.2).
About the role of Rax in progenitor specification there is some disagreement. One
report including a very detailed description of the Rax expression domain finds that
this gene is not expressed in the presumptive neuroepithelium of the anterior region
(Lu et al. 2013). As a consequence, Rax does not specify neurons of the anterior
region (i.e., Avp- or Oxt-expressing neurons, for instance). By other reports, Rax is
indeed expressed in the presumptive neuroepithelium of the anterior region (Orquera
et al. 2016; Shimogori et al. 2010). Accordingly, Orquera et al. (2016) find a
reduction in the expression of Six6, Lhx1, Nkx2-2, which are markers of the
suprachiasmatic nucleus. Both analyses of Rax mutants find that Rax is essential
for the specification of some progenitors of the arcuate nucleus. Orquera et al. (2016)
report a reduction in markers found in this nucleus of the tuberal region including
Isl1, Pomc, and Th in neuroendocrine neurons as well as Sst (a marker of some
non-neuroendocrine arcuate neurons; reviewed in (Alvarez-Bolado 2018)).
According to Lu et al. (2013), Rax is essential to confer fate specification to the
progenitors of Pomc-expressing neurons in the arcuate nucleus and of SF1 (Nr5a1)-
expressing neurons in the ventromedial nucleus, both in the tuberal region.
Contrary to what happens in zebrafish, in the mouse, Rax is not involved in the
development of the anterior region (i.e., not involved in the specification of Avp-
expressing neurons of the paraventricular nucleus). This may be due to Rax having
an ancestral developmental mechanism in fishes and a phylogenetically more recent
and more restricted function in mammals. This is consistent with the fact that rx3
acts through activation of optb (Otp homolog), a transcription factor essential for
anterior/tuberal region in the fish. In the mouse, however, Otp is specific of anterior
fates, particularly in the paraventricular and supraoptic nuclei (although it also
regulates the development of Sst-expressing neurons of the arcuate nucleus
(Acampora et al. 1999)). For the sake of completeness, it is worth mentioning that
Rax has an important role in tanycyte differentiation (Miranda-Angulo et al. 2014;
Salvatierra et al. 2014) and a tanycyte phenotype has not been reported in Shh
mutants.

2.6.4.3 Interaction of Neural Shh with Bmp7 and Pax2

in the Development of the Optic Nerve
Neural Shh is required to maintain the expression of Bmp7 and/or Pax2 in the ventral
region of the developing eye so that retinal axons can exit the eye and form the optic
nerve (Zhao et al. 2012).
56 G. Alvarez-Bolado

2.6.4.4 Fgf10
Both Fgf10 and Shh are expressed in the presumptive hypothalamic neuroepithelium
and the territories of expression of the two markers tend to be complementary (rather
than overlapping), so that they seem to repress each other, in a direct or indirect way.
This expression pattern is key to the development of the hypophysis in chick
embryos (Fu et al. 2017) (see Sect. 2.7.2), and, at least in this last model, seems
key to the specification and growth of the anterior region (see Sect. 2.7.3).

2.7 Shh in Hypothalamic Regions and Hypophysis

2.7.1 Neural Shh in the Tuberal Region

Together with the mammillary, the tuberal region forms the basal portion of the
hypothalamus (Sect. 2.1.2) and most of it derives from Shh-expressing
neuroepithelium (Sect. 2.3). The ventromedial, dorsomedial, and arcuate nuclei are
anatomical landmarks of this region. The arcuate nucleus in particular is easy to
recognize by its functionally important neuronal subpopulations expressing Pomc
and Carptp, Npy, and Agrp or Th and Ghrh. The medial portions of the arcuate and
ventromedial nuclei derive from the medial progenitor domain, and the rest derives
from the Lateral Progenitor Domain (Sect. 2.3). In neural Shh mutants, the tuberal
region is reduced in size and mostly undifferentiated (Sect. 2.5). Transcription
factors essential for tuberal development and depending on neural Shh for activation
or maintenance of expression are Nkx2-1 and Rax (Sects. 2.6.3.1 and 2.6.4.2). Mice
missing Nkx2-1 lack differentiation of the tuberal and the mammillary regions,
which appear to be, however, not much smaller than in wild-type brains (Kimura
et al.1996). Of note, the mammillary region of the Nkx2-1 mutant is more affected
than that of neural Shh mutants, while the anterior region is not affected in Nkx2-1
mutants but very much in neural Shh mutants (see below). In Rax mutant mice, the
arcuate nucleus particularly, and also the ventromedial nucleus, are missing specific
cell populations (Lu et al. 2013; Orquera et al. 2016). The Rax homolog rx3 is
essential for the development of the anterior/tuberal region in zebrafish (Muthu et al.
2016).

2.7.2 Neural Shh in the Developing Hypophysis

A series of elegant experiments on chick embryos, supported by analysis of specific

mouse mutant phenotypes, has led to a description of how successive phases of
activation and repression of Shh, tightly regulated in time and space, dynamically
regulate the development of adenohypophysis and neurohypophysis. The role of Shh
in the development of the hypophysis (infundibulum and Rathke’s pouch) can be
tentatively subdivided into three steps, of which the first and second are consecutive
and the second and third approximately simultaneous.
First, Shh is expressed in the entire presumptive tuberal region (i.e., in the
RDVM). In the chick, this expression it is required for the development of a group
2 Sonic hedgehog in Hypothalamus Development 57

of anterior ventral midline cells called “collar cells.” These are characterized by Fgf3
and Sox3 expression and have a key role in the formation of the infundibulum
(Pearson et al. 2011). Next, Shh is repressed by Tbx2 (Tbx3 in mammals) (Manning
et al. 2006; Trowe et al. 2013) in order for the differentiation of the tuberal and
mammillary regions to proceed (Manning et al. 2006), as well as for infundibulum
morphogenesis (Trowe et al. 2013). Therefore, without Shh, no infundibulum
develops. But, if Shh is expressed in the presumptive tuberal–mammillary region
for too long, then no infundibulum develops either. Since Shh expression is
maintained by several factors, in the absence of these, no Shh, therefore no infun-
dibulum: e.g., without Sox3, no Shh, no infundibulum (Zhao et al. 2012); without
Rax, no Shh, no infundibulum (Orquera et al. 2016). Shh is necessary but not
sufficient for hypophysis development, as other factors are also needed. For instance,
as has been observed, in the Nkx2-1 mutant, Shh is expanded, but this does not rescue
the hypophysis phenotype in this mutant (Carreno et al. 2017). Finally, Shh is
required to specify Rathke’s pouch progenitors in the chick (Fu et al. 2017) as
well as in the mouse (Carreno et al. 2017), identified by expression of Lhx3 and
Lhx4. Again, Shh is required but not sufficient to specify them, since BMP and FGF
signaling are also required (Carreno et al. 2017).

2.7.3 Neural Shh in the Anterior Region

The anterior region is in the alar portion of the hypothalamus (Sect. 2.1.2) and its
characteristic nuclei are the paraventricular, the supraoptic, and the suprachiasmatic.
According to mouse mutant phenotypes (Sect. 2.5), the development of the anterior
region is strongly dependent on neural Shh (Sect. 2.3). In zebrafish, transcription
factor rx3 patterns in this region (Muthu et al. 2016). In the mouse, Rax (rx3
homolog) is expressed in the anterior region (Shimogori et al. 2010; Pak et al.
2014) and it could have a role in anterior patterning (Sect. 2.6.4.2). The anterior
region does not derive from Shh-expressing neuroepithelium in the mouse, i.e., it
does not originate in the basal hypothalamus (Alvarez-Bolado et al. 2012). In the
chicken, however, anterior progenitors derive from Fgf10- and Shh-expressing
midline progenitors present in the presumptive tuberal region (Fu et al. 2017). The
most rostral of these progenitors downregulate Fgf10 and move (or proliferate)
forward (or in the alar direction) abandoning the presumptive tuberal region,
downregulate Shh and proliferate intensively, generating the anterior region
(Fu et al. 2017). In this way, the anterior region of the chick would indeed belong
to the Shh lineage.
In both chicken and mouse, the rostral border (alar border) of the Shh-expressing
domain in the presumptive hypothalamus abuts the Foxg1 (BF1) domain that marks
the telencephalon (Fu et al. 2017; Shimamura et al. 1995). Since in the mouse the
anterior region is not Shh-derived, it could be Foxg1-derived, i.e., telencephalic, like
the preoptic region. There is scarce information on the lineage of Foxg1 in the
forebrain, but at least part of the anterior region seems to belong to the Foxg1
lineage, including part of the paraventricular and/or supraoptic nuclei (Kawaguchi
et al. 2016). The Foxg1 mutant mouse shows a phenotype in the magnocellular
58 G. Alvarez-Bolado

neuroendocrine system (Frullanti et al. 2016). Maybe a clue can be found in the fact
that the radial glial fibers that presumably guide the migration of newborn neurons
originate in telencephalon as well as diencephalon and they end in the preoptic and
anterior regions (Tobet et al. 1995).

2.7.4 Neural Shh in the Preoptic Region

In terms of neuronal nuclei and connections, the preoptic region is the most complex
and diverse. Furthermore, several of these nuclei are sexually dimorphic. The
preoptic region is involved in the regulation of neuroendocrine secretion (through
numerous efferents to neuroendocrine nuclei of the anterior and tuberal regions),
cardiovascular responses, fluid homeostasis, sleep, body temperature, and sexual
behavior (Simerly 2004). Regulation of preoptic region patterning and development
is not well understood. The morphological complexity and the lack of neuroendo-
crine nuclei with obvious neuropeptide markers make it difficult to study. These
problems are compounded by a dilemma. The preoptic region was considered a part
of the hypothalamus since it was first defined (Le Gros Clark 1938) and certainly, the
hypothalamus is a singular brain region with specific functions (Thompson and
Swanson 2003). However, the preoptic region is now considered part of the telen-
cephalon (Puelles et al. 2012). The problem of the “clash of nomenclatures” between
development and adult has been commented before (Zhang and Alvarez-Bolado
2016). Unfortunately, when dealing with the hypothalamus, the preoptic region is
considered telencephalic and therefore often left out of the scope of the paper
(Shimogori et al. 2010; Fu et al. 2017), while papers dealing with telencephalic
development do not mention the preoptic region since it is hypothalamic (Xu et al.
2008). The net result being that we do not know much about the patterning of this
functionally important region.

2.7.4.1 Shh Expression in the Developing Preoptic Region of the Mouse

In the embryonic forebrain, the preoptic region is immediately rostral to the optic
recess and ventral to the medial ganglionic eminence (Fig. 2.1). The onset of Shh
expression in the ventral telencephalon takes place at E8.5, simultaneously with
Nkx2-1. Later on, Shh is expressed in the subventricular zone and in the mantle
(in the hypothalamic primordium Shh is expressed only by the neuroepithelium)
(Shimamura et al. 1995; Sussel et al. 1999). Activation of Shh expression in this
region is dependent on Nkx2-1 (Sussel et al. 1999). The role of Shh in the basal
telencephalon probably does not contribute to patterning but to the maintenance of
appropriate gene expression (Machold et al. 2003). For instance, maintenance of
proper levels of Nkx2.1 expression in the progenitors of the medial ganglionic
eminence depends on Shh (Xu et al. 2005). At E13.5, the preoptic region is
characterized by coexpression of Nkx2-1, Nkx2-2, and Shh, and has two different
progenitor domains (one of them co-expresses Nkx6-2, Dbx1, and Lhx2 and the other
does not) (Flames et al. 2007).
2 Sonic hedgehog in Hypothalamus Development 59

2.7.4.2 Shh-Derived Cells in the Preoptic Region

The neuroepithelium of the preoptic region is characterized by Shh expression but,
despite this fact, the only Shh-derived cells found in this region at the end of
development, are astrocytes and some very few scattered neurons (Alvarez-Bolado
et al. 2012). The reason is that the preoptic progenitor domains produce interneurons
for the amygdala (Hirata et al. 2009), the cortex (Gelman et al. 2009), and the lateral
ganglionic eminence (Marin et al. 2000) as well as oligodendrocytes for the optic
nerve and possibly for more caudal hypothalamic regions (Gao and Miller 2006;
Ono et al. 2017) under the control of Shh (Merchán et al. 2007). Where do the
preoptic neurons originate? As for the origin of the neurons of the anterior region
(Sect. 2.7.3), the answer can be given by Foxg1-lineage analysis. The sparse data
seem to indicate that the preoptic region and part of the anterior are of telencephalic
origin (Kawaguchi et al. 2016).

2.7.4.3 Mutant Phenotypes and Experimental Approaches

to the Preoptic Region
The role of neural Shh in the preoptic region is not known. In the Shh cKO-Foxb1
mutant this region was not very affected (Szabó et al. 2009) possibly because
expression in the telencephalic domain was not completely removed. The Nkx2-1
mutant does not maintain the expression of Shh. For this reason, part of its phenotype
could be due to lack of Shh. To judge from the phenotype of the Nkx2-1 mouse
(Kimura et al. 1996), the septal region including the nucleus of the diagonal band,
derives from this region, together with the “preoptic areas.” Sussel et al. (1999)
found that the telencephalic domain of Nkx2-1 gives rise to medial ganglionic
eminence, septum and preoptic region. Phenotypical analyses of Nkx2-1 mutants
(Kimura et al. 1996; Sussel et al. 1999) do not include neuroanatomical descriptions
of the mutant preoptic area. Analysis of the development of the Shh-expressing
ventral midline in chick embryos indicate that the most rostral hypothalamus is the
anterior region; the Shh–Foxg1 boundary is considered the rostral boundary of
the hypothalamus (Shimogori et al. 2010; Fu et al. 2017), in this way excluding
the preoptic region.

2.7.5 Neural Shh in the Mammillary Region

The mammillary region is the most caudal (or the most basal) of the hypothalamic
regions. Although it does not have the neuropeptide-expressing neuroendocrine
nuclei that characterizes the anterior and tuberal regions, the mammillary region is
morphologically very distinctive with the large mammillary body and its prominent
axonal tree forming the mammillothalamic and mammillotegmental branches. In the
mouse, the mammillary region starts receiving Shh-expressing medial progenitor
domain cells at E8.0, peaks from E9.0 to E10.0 and reduced numbers of cells keep
being added through E12.0 (Alvarez-Bolado et al. 2012). In the chicken, this process
is different. Fgf10-expressing progenitors in the tuberal region proliferate beyond
their caudal (or basal) boundary; then they downregulate Fgf10, upregulate specific
60 G. Alvarez-Bolado

mammillary marker Emx2, and resume proliferation to form the mammillary region
(a similar process takes place in the anterior region; Sect. 2.7.3) (Fu et al. 2017).
The mammillary body and premammillary nuclei show differential needs for
neural Shh. In the absence of neural Shh in this region, expression of a specific
marker of the mammillary body, Foxb1, is not maintained. Yet, other specific
markers of this nucleus (Emx2, Lhx1, Nkx2-1, Sim1) are still expressed (although
in very reduced domains). The premammillary area, however, is more affected: it
loses specific marker expression at E12.5 and by E18.5 the entire area seems to have
disappeared. Note: in the Nkx2-1 mutant brain, the mammillary region does not
differentiate at all (Kimura et al. 1996), that is, it is much more affected than in neural
Shh mutants. In chick embryos, repression of Shh by Tbx2 in the tubero-mammillary
region is necessary for Emx2 to be expressed, that is, for the differentiation and
further proliferation of specific mammillary progenitors (Manning et al. 2006).
Analysis of Tbx3 mutants shows that, in the mouse, inhibition of Shh expression
in this region is not necessary for Emx2 to be expressed (Trowe et al. 2013). Gli3A is
specifically required for the proliferation of the mammillary body (Haddad-Tóvolli
et al. 2015).

2.8 Future Goals

In the near future, it would be interesting to see progress in at least these four areas:
the role of the prechordal plate in hypothalamic specification and patterning, the
origin of hypothalamic regions, the relation between the Shh pathway and
neuroepithelial cilia and the genomic networks regulating hypothalamic neuron
specification.

Shh in the Prechordal Plate

The realization that the main function of neural Shh in the developing hypothalamus
is to counteract Gli3R function leads back to the patterning role of Shh secreted by
the prechordal plate, as well as to the patterning of the prechordal plate itself. Much
remains to be done in the molecular genetic control, for instance, are there specific
enhancers to govern Shh expression in the prechordal plate? Which proteins activate
such enhancer(s)? What molecular interactions pattern the prechordal plate (and so
pre-pattern the hypothalamus)?
The phenomenology of prechordal plate formation and migration as well as its
changing influence on the overlying presumptive hypothalamic floor has been
gleaned mostly from studies in chicken embryos. Therefore, an important question
is how similar to each other are the prechordal plates of Birds and Mammals? Rabbit
embryos show a flat gastrula (as opposed to the egg-cylinder gastrula of the mouse)
accessible to manipulation. The mechanistic analysis of this model is in its
beginnings but it is starting to yield unexpected and interesting results (Kremnyov
et al. 2018).
2 Sonic hedgehog in Hypothalamus Development 61

Role of Shh in the Specification of the Four Hypothalamic Regions

How do the hypothalamic regions develop? Studies in the chick have shown how the
regions expand rostrally and caudally from an initial tuberal region. This “aniso-
tropic growth model” (Fu et al. 2017; see above Sect. 2.7.3) has explanatory power
and is supported by careful experiments. Our Shh lineage data suggests that there
could be some intriguing mechanistic differences, however, between birds and
mammals. Again here, the rabbit flat gastrula could be very useful and has the
potential to become a long-range research project. More specifically, we do not
know near enough about the origin of the cells of the preoptic and anterior hypotha-
lamic areas (location of progenitors and migration pathways) and what exactly is the
role of Shh in their specification.

What Is the Role of the Neuroepithelial Cilia in Hypothalamic Development?

The Shh membrane receptor and downstream signaling components of the Shh
pathway are located to the primary cilium (Huangfu et al. 2003). The importance
of the primary cilium of neuroepithelial cells for the patterning function of Shh on
the basal forebrain is only starting to be investigated (Andreu-Cervera et al. 2019)
and has the marks of a fruitful research field for the future.

Shh and the Developmental Genomic Regulatory Networks

of the Hypothalamus
The question about the complexity of the brain (including the hypothalamus) boils
down to a great extent to the question of the regionalization and specification of the
neuroepithelium. A great deal is already known about the regionalization of the
neural plate, including the presumptive hypothalamus by Hh, Fgf, and Bmp ligands.
In the mouse and fish, analysis of mutants guided by microsurgical and explant
experiments in chicken will be needed, as well as expression resources, bioinfor-
matics approaches and mutant analysis. Ultimately, the entire genomic regulatory
network underpinning the regionalization and specification of the neuroepithelium
and cascading from Shh and other factors (Wnt, BMP, Fgf in particular) will have to
be revealed.

Key References

Alvarez-Bolado et al. (2012)—Fate-mapping shows that the Shh-expressing

domains of the presumptive hypothalamus generate the tuberal and mammillary
regions (but not the preoptic and anterior regions).
Chiang et al. (1996)—Inactivation of Shh in transgenic mice confirmed that Shh is
required for the formation of ventral structures (including the entire hypothala-
mus) in the central nervous system.
Dale et al. (1997)—Experiments on chick embryos show that the hypothalamic floor
plate is specified by the prechordal mesoderm.
62 G. Alvarez-Bolado

Geng et al. (2008)—A network of essential regulators controls the partition of the
telencephalon into left and right hemispheres. As a consequence, in certain
genetic backgrounds, neural Shh is required for this process to be successful.
Haddad-Tóvolli et al. (2015)—Analysis of mouse mutant phenotypes shows how
different combinations of Gli transcription factors, acting downstream Shh sig-
nalling, are involved in the specification of the four classical hypothalamic
regions.
Manning L et al. (2006)—The mechanism behind the impressive change in the
pattern of Shh expression in the hypothalamus, and its role, explained.
Szabó et al. (2009)—Thirteen years after the full Shh null mutant (Chiang et al.
1996), the role of neural Shh vs prechordal Shh in hypothalamic development
were dissected by using conditional mouse mutants.
Shimogori et al. (2010)—A second conditional mutant analyzing the role of Shh
from different sources (neural or prechordal) on the developing hypothalamus.
Zhao et al. (2012)—The authors use a novel conditional mutant to show that neural
Shh is essential for the proper development of the eye and hypophysis, a role that
had escaped previous analyses.

References
Acampora D, Postiglione MP, Avantaggiato V, Di Bonito M, Vaccarino FM, Michaud J, Simeone
A (1999) Progressive impairment of developing neuroendocrine cell lineages in the hypothala-
mus of mice lacking the Orthopedia gene. Genes Dev 13:2787–2800
Altman J, Bayer SA (1986) The development of the rat hypothalamus. Adv Anat Embryol Cell Biol
100:1–178
Alvarez-Bolado G (2018) Development of neuroendocrine neurons in the mammalian hypothala-
mus. Cell Tissue Res. https://doi.org/10.1007/s00441-018-2859-1
Alvarez-Bolado G, Paul FA, Blaess S (2012) Sonic hedgehog lineage in the mouse hypothalamus:
from progenitor domains to hypothalamic regions. Neural Dev 7:4. https://doi.org/10.1186/
1749-8104-7-4
Andreu-Cervera A, Anselme I, Karam A, Laclef C, Catala M, Schneider-Maunoury S (2019) The
ciliopathy gene Ftm/Rpgrip1l controls mouse forebrain patterning via region-specific modula-
tion of Hedgehog/Gli signaling. J Neurosci 39(13):2398–2415. https://doi.org/10.1523/J.
NEUROSCI.2199-18
Aoto K, Shikata Y, Imai H, Matsumaru D, Tokunaga T, Shioda S, Yamada G, Motoyama J (2009)
Mouse Shh is required for prechordal plate maintenance during brain and craniofacial morpho-
genesis. Dev Biol 327:106–120. https://doi.org/10.1016/j.ydbio.2008.11.022
Blaess S, Szabó N, Haddad-Tovolli R, Zhou X, Alvarez-Bolado G (2014) Sonic hedgehog
signaling in the development of the mouse hypothalamus. Front Neuroanat 8:156. https://doi.
org/10.3389/fnana.2014.00156
Braun MM, Etheridge A, Bernard A, Robertson CP, Roelink H (2003) Wnt signaling is required at
distinct stages of development for the induction of the posterior forebrain. Development
130:5579–5587
Briscoe J, Thérond PP (2013) The mechanisms of Hedgehog signalling and its roles in development
and disease. Nat Rev Mol Cell Biol 14:416–429. https://doi.org/10.1038/nrm3598
Carreno G, Apps JR, Lodge EJ, Panousopoulos L, Haston S, Gonzalez-Meljem JM, Hahn H,
Andoniadou CL, Martinez-Barbera JP (2017) Hypothalamic sonic hedgehog is required for cell
2 Sonic hedgehog in Hypothalamus Development 63

specification and proliferation of LHX3/LHX4 pituitary embryonic precursors. Development

144:3289–3302. https://doi.org/10.1242/dev.153387
Cayuso J, Ulloa F, Cox B, Briscoe J, Marti E (2006) The Sonic hedgehog pathway independently
controls the patterning, proliferation and survival of neuroepithelial cells by regulating Gli
activity. Development 133:517–528
Chiang C, Litingtung Y, Lee E, Young KE, Corden JL, Westphal H, Beachy PA (1996) Cyclopia
and defective axial patterning in mice lacking Sonic hedgehog gene function. Nature
383:407–413
Dale JK, Vesque C, Lints TJ, Sampath TK, Furley A, Dodd J, Placzek M (1997) Cooperation of
BMP7 and SHH in the induction of forebrain ventral midline cells by prechordal mesoderm.
Cell 90:257–269
Dassule HR, Lewis P, Bei M, Maas R, McMahon AP (2000) Sonic hedgehog regulates growth and
morphogenesis of the tooth. Development 127:4775–4785
Dessaud E, McMahon AP, Briscoe J (2008) Pattern formation in the vertebrate neural tube: a sonic
hedgehog morphogen-regulated transcriptional network. Development 135:2489–2503
Echelard Y, Epstein DJ, St-Jacques B, Shen L, Mohler J, McMahon JA, McMahon AP (1993)
Sonic hedgehog, a member of a family of putative signaling molecules, is implicated in the
regulation of CNS polarity. Cell 75:1417–1430
Ericson J, Muhr J, Placzek M, Lints T, Jessell TM, Edlund T (1995) Sonic hedgehog induces the
differentiation of ventral forebrain neurons: a common signal for ventral patterning within the
neural tube. Cell 81:747–756
Flames N, Pla R, Gelman DM, Rubenstein JL, Puelles L, Marin O (2007) Delineation of multiple
subpallial progenitor domains by the combinatorial expression of transcriptional codes.
J Neurosci 27:9682–9695. https://doi.org/10.1523/JNEUROSCI.2750-07.2007
Frullanti E, Amabile S, Lolli MG, Bartolini A, Livide G, Landucci E, Mari F, Vaccarino FM,
Ariani F, Massimino L, Renieri A, Meloni I (2016) Altered expression of neuropeptides in
FoxG1-null heterozygous mutant mice. Eur J Hum Genet 24:252–257. https://doi.org/10.1038/
ejhg.2015.79
Fu T, Towers M, Placzek MA (2017) Fgf10+ progenitors give rise to the chick hypothalamus by
rostral and caudal growth and differentiation. Development 144:3278–3288. https://doi.org/10.
1242/dev.153379
Gao L, Miller RH (2006) Specification of optic nerve oligodendrocyte precursors by retinal
ganglion cell axons. J Neurosci 26:7619–7628. https://doi.org/10.1523/JNEUROSCI.0855-06.
2006
Gelman DM, Martini FJ, Nóbrega-Pereira S, Pierani A, Kessaris N, Marín O (2009) The embryonic
preoptic area is a novel source of cortical GABAergic interneurons. J Neurosci 29:9380–9389.
https://doi.org/10.1523/JNEUROSCI.0604-09.2009
Geng X, Speirs C, Lagutin O, Inbal A, Liu W, Solnica-Krezel L, Jeong Y, Epstein DJ, Oliver G
(2008) Haploinsufficiency of Six3 fails to activate Sonic hedgehog expression in the ventral
forebrain and causes holoprosencephaly. Dev Cell 15:236–247. https://doi.org/10.1016/j.
devcel.2008.07.003
Geng X, Acosta S, Lagutin O, Gil HJ, Oliver G (2016) Six3 dosage mediates the pathogenesis of
holoprosencephaly. Development 143:4462–4473. https://doi.org/10.1242/dev.132142
Haddad-Tóvolli R, Paul FA, Zhang Y, Zhou X, Theil T, Puelles L, Blaess S, Alvarez-Bolado G
(2015) Differential requirements for Gli2 and Gli3 in the regional specification of the mouse
hypothalamus. Front Neuroanat 9:34. https://doi.org/10.3389/fnana.2015.00034
Hirata T, Li P, Lanuza GM, Cocas LA, Huntsman MM, Corbin JG (2009) Identification of distinct
telencephalic progenitor pools for neuronal diversity in the amygdala. Nat Neurosci
12:141–149. https://doi.org/10.1038/nn.2241
Huangfu D, Liu A, Rakeman AS, Murcia NS, Niswander L, Anderson KV (2003) Hedgehog
signalling in the mouse requires intraflagellar transport proteins. Nature 426:83–87. https://doi.
org/10.1038/nature02061
64 G. Alvarez-Bolado

Hui CC, Joyner AL (1993) A mouse model of greig cephalopolysyndactyly syndrome: the extra-
toesJ mutation contains an intragenic deletion of the Gli3 gene. Nat Genet 3:241–246
Hui CC, Slusarski D, Platt KA, Holmgren R, Joyner AL (1994) Expression of three mouse
homologs of the Drosophila segment polarity gene cubitus interruptus, Gli, Gli-2, and Gli-3,
in ectoderm- and mesoderm-derived tissues suggests multiple roles during postimplantation
development. Dev Biol 162:402–413. https://doi.org/10.1006/dbio.1994.1097
Ishibashi M, McMahon AP (2002) A sonic hedgehog-dependent signaling relay regulates growth of
diencephalic and mesencephalic primordia in the early mouse embryo. Development
129:4807–4819
Jeong Y, El-Jaick K, Roessler E, Muenke M, Epstein DJ (2006) A functional screen for sonic
hedgehog regulatory elements across a 1 Mb interval identifies long-range ventral forebrain
enhancers. Development 133:761–772. https://doi.org/10.1242/dev.02239
Jeong Y, Leskow FC, El-Jaick K, Roessler E, Muenke M, Yocum A, Dubourg C, Li X, Geng X,
Oliver G, Epstein DJ (2008) Regulation of a remote Shh forebrain enhancer by the Six3
homeoprotein. Nat Genet 40:1348–1353. https://doi.org/10.1038/ng.230
Kapsimali M, Caneparo L, Houart C, Wilson SW (2004) Inhibition of Wnt/Axin/beta-catenin
pathway activity promotes ventral CNS midline tissue to adopt hypothalamic rather than
floorplate identity. Development 131:5923–5933
Kawaguchi D, Sahara S, Zembrzycki A, O’Leary DDM (2016) Generation and analysis of an
improved Foxg1-IRES-Cre driver mouse line. Dev Biol 412:139–147. https://doi.org/10.1016/j.
ydbio.2016.02.011
Kimura S, Hara Y, Pineau T, Fernandez-Salguero P, Fox CH, Ward JM, Gonzalez FJ (1996) The
T/ebp null mouse: thyroid-specific enhancer-binding protein is essential for the organogenesis
of the thyroid, lung, ventral forebrain, and pituitary. Genes Dev 10:60–69
Kobayashi D, Kobayashi M, Matsumoto K, Ogura T, Nakafuku M, Shimamura K (2002) Early
subdivisions in the neural plate define distinct competence for inductive signals. Development
129:83–93
Krauss S, Concordet JP, Ingham PW (1993) A functionally conserved homolog of the Drosophila
segment polarity gene hh is expressed in tissues with polarizing activity in zebrafish embryos.
Cell 75:1431–1444. https://doi.org/10.1016/0092-8674(93)90628-4
Kremnyov S, Henningfeld K, Viebahn C, Tsikolia N (2018) Divergent axial morphogenesis and
early Shh expression in vertebrate prospective floor plate. EvoDevo 9:4. https://doi.org/10.1186/
s13227-017-0090-x
Lagutin OV, Zhu CC, Kobayashi D, Topczewski J, Shimamura K, Puelles L, Russell HR,
McKinnon PJ, Solnica-Krezel L, Oliver G (2003) Six3 repression of Wnt signaling in the
anterior neuroectoderm is essential for vertebrate forebrain development. Genes Dev
17:368–379
Le Gros Clark WE (1938) Morphological aspects of the hypothalamus. In: Le Gros Clark WE,
Beattie J, Riddoch G, Dott NM (eds) The hypothalamus. Morphological, functional, clinical and
surgical aspects. Oliver and Boyd, Edinburgh
Legué E, Joyner AL (2010) Genetic fate mapping using site-specific recombinases. In: Wassarman
PM, Soriano PM (eds) Guide to techniques in mouse development, Part B: Mouse molecular
genetics, 2nd edn. Elsevier, Amsterdam
Lu F, Kar D, Gruenig N, Zhang ZW, Cousins N, Rodgers HM, Swindell EC, Jamrich M,
Schuurmans C, Mathers PH, Kurrasch DM (2013) Rax is a selector gene for mediobasal
hypothalamic cell types. J Neurosci 33:259–272. https://doi.org/10.1523/JNEUROSCI.0913-
12.2013
Machold R, Hayashi S, Rutlin M, Muzumdar MD, Nery S, Corbin JG, Gritli-Linde A, Dellovade T,
Porter JA, Rubin LL, Dudek H, McMahon AP, Fishell G (2003) Sonic hedgehog is required for
progenitor cell maintenance in telencephalic stem cell niches. Neuron 39:937–950
Manning L, Ohyama K, Saeger B, Hatano O, Wilson SA, Logan M, Placzek M (2006) Regional
morphogenesis in the hypothalamus: a BMP-Tbx2 pathway coordinates fate and proliferation
through Shh downregulation. Dev Cell 11:873–885
2 Sonic hedgehog in Hypothalamus Development 65

Marin O, Anderson SA, Rubenstein JL (2000) Origin and molecular specification of striatal
interneurons. J Neurosci 20:6063–6076
Mathers PH, Grinberg A, Mahon KA, Jamrich M (1997) The Rx homeobox gene is essential for
vertebrate eye development. Nature 387:603–607. https://doi.org/10.1038/42475
Merchán P, Bribián A, Sánchez-Camacho C, Lezameta M, Bovolenta P, de Castro F (2007) Sonic
hedgehog promotes the migration and proliferation of optic nerve oligodendrocyte precursors.
Mol Cell Neurosci 36:355–368. https://doi.org/10.1016/j.mcn.2007.07.012
Miranda-Angulo AL, Byerly MS, Mesa J, Wang H, Blackshaw S (2014) Rax regulates hypotha-
lamic tanycyte differentiation and barrier function in mice. J Comp Neurol 522:876–899. https://
doi.org/10.1002/cne.23451
Mo R, Freer AM, Zinyk DL, Crackower MA, Michaud J, Heng HH, Chik KW, Shi XM, Tsui LC,
Cheng SH, Joyner AL, Hui C (1997) Specific and redundant functions of Gli2 and Gli3 zinc
finger genes in skeletal patterning and development. Development 124:113–123
Muthu V, Eachus H, Ellis P, Brown S, Placzek M (2016) Rx3 and Shh direct anisotropic growth and
specification in the zebrafish tuberal/anterior hypothalamus. Development 143:2651–2663.
https://doi.org/10.1242/dev.138305
Nüsslein-Volhard C, Wieschaus E (1980) Mutations affecting segment number and polarity in
Drosophila. Nature 287:795–801
Ono K, Yoshii K, Tominaga H, Gotoh H, Nomura T, Takebayashi H, Ikenaka K (2017) Oligoden-
drocyte precursor cells in the mouse optic nerve originate in the preoptic area. Brain Struct Funct
222:2441–2448. https://doi.org/10.1007/s00429-017-1394-2
Orquera DP, Nasif S, Low MJ, Rubinstein M, de Souza FSJ (2016) Essential function of the
transcription factor Rax in the early patterning of the mammalian hypothalamus. Dev Biol
416:212–224. https://doi.org/10.1016/j.ydbio.2016.05.021
Pak T, Yoo S, Miranda-Angulo AL, Wang H, Blackshaw S (2014) Rax-CreERT2 knock-in mice: a
tool for selective and conditional gene deletion in progenitor cells and radial glia of the retina
and hypothalamus. PLoS One 9(4):e90381. https://doi.org/10.1371/journal.pone.0090381
Patten I, Kulesa P, Shen MM, Fraser S, Placzek M (2003) Distinct modes of floor plate induction in
the chick embryo. Development 130:4809–4821. https://doi.org/10.1242/dev.00694
Pearson CA, Ohyama K, Manning L, Aghamohammadzadeh S, Sang H, Placzek M (2011)
FGF-dependent midline-derived progenitor cells in hypothalamic infundibular development.
Development 138:2613–2624. https://doi.org/10.1242/dev.062794
Placzek M, Briscoe J (2005) The floor plate: multiple cells, multiple signals. Nat Rev Neurosci
6:230–240
Puelles L, Martínez S, Martínez-de-la-Torre M, Rubenstein JLR (2004) Gene maps and related
histogenetic domains in the forebrain and midbrain. In: Paxinos G (ed) The rat nervous system,
3rd edn. Academic, San Diego, CA
Puelles L, Martinez-de-la-Torre M, Bardet S, Rubenstein JLR (2012) Hypothalamus. In: Watson C,
Paxinos G, Puelles L (eds) The mouse nervous system. Elsevier-Academic, San Diego
Rallu M, Machold R, Gaiano N, Corbin JG, McMahon AP, Fishell G (2002) Dorsoventral
patterning is established in the telencephalon of mutants lacking both Gli3 and Hedgehog
signaling. Development 129:4963–4974
Riddle RD, Johnson RL, Laufer E, Tabin C (1993) Sonic hedgehog mediates the polarizing activity
of the ZPA. Cell 75:14011416. https://doi.org/10.1016/0092-8674(93)90626-2
Roelink H, Augsburger A, Heemskerk J, Korzh V, Norlin S, Ruiz i Altaba A, Tanabe Y, Placzek M,
Edlund T, Jessell TM, Dodd J (1994) Floor plate and motor neuron induction by vhh-1, a
vertebrate homolog of hedgehog expressed by the notochord. Cell 16:761–775. https://doi.org/
10.1016/0092-8674(94)90514-2
Roessler E, Muenke M (2010) The molecular genetics of holoprosencephaly. Am J Med Genet C
Semin Med Genet 154C:52–61. https://doi.org/10.1002/ajmg.c.30236
Ruiz i Altaba A, Jessell TM (1993) Midline cells and the organization of the vertebrate neuraxis.
Curr Opin Genet Dev. 3:633–640. https://doi.org/10.1016/0959-437X(93)90100-4
66 G. Alvarez-Bolado

Sagner A, Briscoe J (2017) Morphogen interpretation: concentration, time, competence, and

signaling dynamics. Wiley Interdiscip Rev Dev Biol 6. https://doi.org/10.1002/wdev.271
Salvatierra J, Lee DA, Zibetti C, Duran-Moreno M, Yoo S, Newman EA, Wang H, Bedont JL, de
Melo J, Miranda-Angulo AL, Gil-Perotin S, Garcia-Verdugo JM, Blackshaw S (2014) The LIM
homeodomain factor Lhx2 is required for hypothalamic tanycyte specification and differentia-
tion. J Neurosci 34:16809–16820. https://doi.org/10.1523/JNEUROSCI.1711-14.2014
Shimamura K, Rubenstein JL (1997) Inductive interactions direct early regionalization of the mouse
forebrain. Development 124:2709–2718
Shimamura K, Hartigan DJ, Martinez S, Puelles L, Rubenstein JL (1995) Longitudinal organization
of the anterior neural plate and neural tube. Development 121:3923–3933
Shimogori T, Lee DA, Miranda-Angulo A, Yang Y, Wang H, Jiang L, Yoshida AC, Kataoka A,
Mashiko H, Avetisyan M, Qi L, Qian J, Blackshaw S (2010) A genomic atlas of mouse
hypothalamic development. Nat Neurosci 13:767–775. https://doi.org/10.1038/nn.2545
Simerly RB (2004) Anatomical substrates of hypothalamic integration. In: Paxinos G (ed) The rat
nervous system, 3rd edn. Elsevier, Amsterdam
Sussel L, Marin O, Kimura S, Rubenstein JL (1999) Loss of Nkx2.1 homeobox gene function
results in a ventral to dorsal molecular respecification within the basal telencephalon: evidence
for a transformation of the pallidum into the striatum. Development 126:3359–3370
Szabó NE, Zhao T, Cankaya M, Theil T, Zhou X, Alvarez-Bolado G (2009) Role of neuroepithelial
Sonic hedgehog in hypothalamic patterning. J Neurosci 29:6989–7002. https://doi.org/10.1523/
JNEUROSCI.1089-09.2009
Thompson RH, Swanson LW (2003) Structural characterization of a hypothalamic visceromotor
pattern generator network. Brain Res Brain Res Rev 41:153–202
Tobet SA, Paredes RG, Chickering TW, Baum MJ (1995) Telencephalic and diencephalic origin of
radial glial processes in the developing preoptic area/anterior hypothalamus. J Neurobiol
26:75–86. https://doi.org/10.1002/neu.480260107
Trowe MO, Zhao L, Weiss AC, Christoffels V, Epstein DJ, Kispert A (2013) Inhibition of Sox2-
dependent activation of Shh in the ventral diencephalon by Tbx3 is required for formation of the
neurohypophysis. Development 140:2299–2309. https://doi.org/10.1242/dev.094524
Xu Q, Wonders CP, Anderson SA (2005) Sonic hedgehog maintains the identity of cortical
interneuron progenitors in the ventral telencephalon. Development 132:4987–4998. https://
doi.org/10.1242/dev.02090
Xu Q, Tam M, Anderson SA (2008) Fate mapping Nkx2.1-lineage cells in the mouse telencepha-
lon. J Comp Neurol 506:16–29. https://doi.org/10.1002/cne.21529
Xuan S, Baptista CA, Balas G, Tao W, Soares VC, Lai E (1995) Winged helix transcription factor
BF-1 is essential for the development of the cerebral hemispheres. Neuron 14:1141–1152.
https://doi.org/10.1016/0896-6273(95)90262-7
Yu K, McGlynn S, Matise MP (2013) Floor plate-derived sonic hedgehog regulates glial and
ependymal cell fates in the developing spinal cord. Development 140:1594–1604. https://doi.
org/10.1242/dev.090845
Zhang Y, Alvarez-Bolado G (2016) Differential developmental strategies by Sonic hedgehog in
thalamus and hypothalamus. J Chem Neuroanat 75:20–27. https://doi.org/10.1016/j.jchemneu.
2015.11.008
Zhao L, Zevallos SE, Rizzoti K, Jeong Y, Lovell-Badge R, Epstein DJ (2012) Disruption of SoxB1-
dependent Sonic hedgehog expression in the hypothalamus causes septo-optic dysplasia. Dev
Cell 22:585–596. https://doi.org/10.1016/j.devcel.2011.12.023
Development of the Hypothalamus
in Xenopus laevis 3
Nerea Moreno and Agustín González

Abstract
Xenopus laevis (African clawed frog) is a model organism for the study of early
embryonic development of vertebrates. It is especially suited to analyze the
neuroendocrine system because it possesses hormonal systems and mechanisms
of gene regulation comparable to those in rodents and humans. The hypothalamus
is the structure of the forebrain that plays a fundamental role in the neuroendo-
crine and homeostatic regulation of the body, representing a central hub for the
neural networks involved in these functions. As such the hypothalamus is the
nexus between the autonomic and limbic systems, which condition the behavioral
and regulatory responses of the animal. During the development of Xenopus, the
hypothalamus is located rostral to the diencephalon, possesses derivatives of the
alar and basal plates, and includes molecular compartments, comparable to
amniotes (reptiles, birds, and mammals), expressing transcription factors such
as Nkx2.1, Islet 1, and Otp in unique subpopulations.

Keywords
Prosencephalon · Forebrain patterning · Expression patterns · Evolution

Abbreviations

BH Basal hypothalamus
CT Caudal tuberal region
Dien Diencephalon
hp1 Hypothalamic prosomeric domain 1
hp2 Hypothalamic prosomeric domain 2

N. Moreno (*) · A. González

Department of Cell Biology, Faculty of Biology, University Complutense, Madrid, Spain
e-mail: [email protected]

# Springer Nature Switzerland AG 2020 67

S. Wray, S. Blackshaw (eds.), Developmental Neuroendocrinology, Masterclass in
Neuroendocrinology 9, https://doi.org/10.1007/978-3-030-40002-6_3
68 N. Moreno and A. González

Hyp Hypophysis
M Mammillary region
Ma Mammillary area proper
oc Optic chiasm
P1 Prosomere 1
P2 Prosomere 2
P3 Prosomere 3
P3b Basal plate of P3
Pa Paraventricular nucleus
PA Pallium
Pac Caudal paraventricular nucleus
Par Rostral paraventricular nucleus
PO Preoptic region
PT Pretectum
PTh Prethalamus
RM Retromammillary region
RT Rostral tuberal region
Sag Sagittal
SC Suprachiasmatic region
SCc Caudal suprachiasmatic region
SCr Rostral suprachiasmatic region
SPa Subparaventricular area
SPA Subpallium
Th Thalamus
Tub Tuberal region

3.1 Why Is the Xenopus laevis Hypothalamus a Model

of Interest in Developmental Analysis?

From the late 1940s to the 1970s, when modern pregnancy tests were not available,
Xenopus laevis frogs were injected with the urine of possibly pregnant women and a
positive result was indicated when the presence of human chorionic gonadotropin
(hCG) induced the frogs to lay a large number of eggs (Lee-Liu et al. 2017). At the
same time, in many laboratories, this ease with which female frogs could be induced
to lay eggs, which could subsequently be fertilized in vitro, along with the relatively
large size of the eggs and embryos, and their development ex utero, made Xenopus
laevis an excellent model organism to study the early embryonic development of
vertebrates. In addition, due to its metamorphosis phase, this model has provided
insight into the developing neuroendocrine systems, as well as a model to study
endocrine disrupters. In fact, in recent years evidence has accumulated that persistent
organic pollutants, heavy metals, pharmaceutical and personal care products, agri-
cultural chemicals, and aerospace products are acting as endocrine disrupters that
may alter the activity of hypothalamus–pituitary–thyroid axis and subsequently
3 Development of the Hypothalamus in Xenopus laevis 69

interfere with the endocrine systems of both wildlife and humans by affecting
reproductive biology and the thyroid system. In this context, it was also
demonstrated that hormonal control of development during the human perinatal
period is critically important and complex, with multiple hormones that regulate
fetal growth, brain development, and organ maturation in preparation for birth.
Notably, many significant parallels exist between metamorphosis of frogs and the
birth of mammals, and underlying these analogous developmental changes are
conserved neuroendocrine components (Buchholz 2017). Frogs and mammals
have identical hormones that reach their peak at metamorphosis and birth, have
conserved hormone receptors and mechanisms of gene regulation, and share com-
parable roles for hormones in many target organs. Indeed, in both frogs and
mammals, genetic and environmental perturbations of hormonal control can cause
irreversible morphological and physiological changes and can predispose
individuals to diseases of adulthood.
The hypothalamic development in Xenopus has been the focus of thorough
studies in recent years (Domínguez et al. 2013, 2014, 2015). In these studies,
Xenopus was selected mainly due to the convenience of “hormone induced” breed-
ing that makes embryos available for experimentation at all times of the year and that
they are easy to maintain under laboratory conditions. In addition, an accurate
timetable of development exists (Nieuwkoop and Faber 1967) that allows compara-
tive interpretation of developmental processes. Notably, analysis of the amphibian
hypothalamus serves as a model for understanding perinatal events in mammals due
to the analogies in development of these structures.

3.1.1 The Hypothalamus in the Current Prosomeric Model of Brain

Regionalization

In all vertebrates, the hypothalamus plays a key role in survival of the animal, acting
as the bridge between the external (visual, olfactory, and gustatory) and internal
stimuli and endocrine, autonomic and limbic responses. Thus, the hypothalamus is
considered the forebrain territory par excellence for controlling homeostatic pro-
cesses. The hypothalamus has been studied to a great extent, however the interpre-
tation of its regionalization during development has been debated throughout the last
decades and additional studies may shed further light on this fundamental,
anatomical question.
From an anatomical point of view, the strong curvature of the neural tube flexure
at the junction of the midbrain and forebrain forms a sharp cephalic flexure involving
the hypothalamus (see Fig. 3.1). This developmental event has made it hard to
identify the basic units or subdivisions in the mature hypothalamus and understand
the topological relationships between them. Thus, following mere topographical
positions and interpreting “classical” transverse sections, the hypothalamus was
considered a diencephalic region beneath the thalamus and different nomenclatures
have been used for nuclei and hypothalamic regions in different vertebrates. The
70 N. Moreno and A. González

Fig. 3.1 Schematic comparison of the hypothalamic development between mammals and
non-mammals. At the neural plate stage (a), the brain undergoes neurulation forming the neural
tube (b). The hypothalamic anatomy is influenced by the flexure of the neural tube. In mammals the
longitudinal axis bends almost 90
forming a sharp flexure and the rostral tube is moved to a
“ventral” position, whereas in non-mammals this angle is less pronounced. Additionally, the
telencephalic evagination in the case of mammals is huge and proportional over hp1 and hp2,
flattening the hypothalamic territory, whereas in non-mammals, this event pushes hp2, turning the
hypothalamus more “ventral.” In contrast to non-mammals, the hypothalamic evagination in
mammals contributes to the elongation of the hypothalamic territory

different nomenclature established in the past has made it challenging to compare

results within the research community.
During the last 25 years, the “prosomeric model” of forebrain organization has
been reinforced by numerous neuroanatomical studies (Puelles and Rubenstein
2003). This model is based on data from developmental expression patterns of
many different regulatory genes (such as Nkx2, Six3, Dlx, OTx2, Otp, and Isl1)
that allowed investigators to conceptualize and explain the organization of different
brain regions, including the hypothalamus (Puelles et al. 2012; Domínguez et al.
2015) in many different species, including Xenopus. In this model, the forebrain is
organized into transverse segments (prosomeres, see Fig. 3.1) and longitudinal zones
3 Development of the Hypothalamus in Xenopus laevis 71

that are also defined histogenetic domains based on regulatory gene expression
patterns. According to the original prosomeric model and its subsequent revisions,
the hypothalamus is excluded from the diencephalon, which is composed of three
neuromeres (prosomeres P1–P3, Fig. 3.1). The rostral-most forebrain is designated
the secondary prosencephalon that contains the hypothalamus (rostral to
diencephalic P3), the telencephalon impar, and the telencephalic hemispheres
(Puelles and Rubenstein 2003). The interpretation of the parts of the secondary
prosencephalon is fraught with difficulties, mainly derived from the early optic and
hemispheric evaginations and the different degree of development shown across
vertebrates that disturb the primary pattern of this region. However, morphological,
molecular, and hodological (connectivity) results have continued to highlight the
organization of the main parts of the secondary prosencephalon and its subdivisions,
particularly in mice, although comparable results have been obtained in other
amniotes, amphibians, and fishes (reviewed in Domínguez et al. 2015; Ferran
et al. 2015).
Based on gene expression patterns and other molecular markers, interpreted
under the current prosomeric model (Fig. 3.1), such as FoxG1, the preoptic region
(PO, see Fig. 3.2) is part of the subpallium of the telencephalon (SPA, Fig. 3.2),
instead of the hypothalamus, where it was traditionally included (Puelles et al. 2012).
It forms the dorsal boundary of the hypothalamic territory. In the proper hypothala-
mus, a transverse and complete limit, called intrahypothalamic boundary, has been
described marking anteroposterior patterning and defining subdivisions. Thus, ter-
minal (rostral, hp2: hypothalamic prosomeric domain 2, Fig. 3.1 dark blue area) and
peduncular (caudal, hp1: hypothalamic prosomeric domain 1, Fig. 3.1 lighter blue
area) regions were distinct on the basis of the differential expression of transcription
factors in the early neural tube. In addition, a band of Nkx2.2 expressing cells
becomes bent at the cephalic flexure and highlights the continuity of the dorsoventral
patterning across the hypothalamus, revealing the existence of the longitudinal
organization in the forebrain, and indicating alar versus basal territories. As such,
four main domains can be considered in the hypothalamus, because both the hp1 and
hp2 segments contain alar and basal parts (see Fig. 3.1). Furthermore, two longitu-
dinal subdivisions are also formed in the alar and basal hypothalamic regions. The
paraventricular area (Pa), topologically the dorsalmost part of the hypothalamus, is
located adjacent to the preoptic region and mainly expresses the transcription factors
Tbr1, Sim1, and Otp, and produces the magnocellular and parvocellular neurosecre-
tory cell populations of the paraventricular, supraoptic, and periventricular nuclei.
The second domain is the subparaventricular area (SPa), which forms the ventral part
of the alar hypothalamus and is characterized by the expression of Dlx genes and
gives rise to the suprachiasmatic, the anterior hypothalamic nucleus, and the
subparaventricular zone. Sonic hedgehog and the transcription factor Nkx2.1
specifically characterize the basal hypothalamus formed by two longitudinal
subdivisions: tuberal (Tub) and mammillary (M) regions. The tuberal region
produces the anterobasal and posterobasal areas, and the ventromedial, dorsomedial,
and arcuate nuclei. Rostrally to all those areas, in the terminal region a distinct
acroterminal territory has been defined (see Fig. 3.1, yellow dots, Ferran et al. 2015).
72 N. Moreno and A. González

Fig. 3.2 (1) Schematic representation of the timing of the main developmental events during
Xenopus laevis development (based on Nieuwkoop and Faber 1967) pf: post-fertilization. (2) Brain
region changes from embryonic to larval. Photomicrographs of Nissl-stained sagittal (b, h) and
transverse (c–f; i–l) sections through the hypothalamus of an embryonic (b–f) and a larval (h–l)
Xenopus; from anterior to posterior, at the approximate levels indicated on the schemes (a, g). These
photomicrographs illustrate the segregated regions recognized in the hypothalamus using
techniques of the time (Nieuwkoop and Faber 1967). The scale bar in c is valid for b–
f ¼ 100 μm. The scale bar in i is valid for h–l ¼ 200μm

It corresponds to the rostral-most portion of the neural tube and is responsible for the
development of structures such as the lamina terminalis and the optic chiasm in the
alar hypothalamus, whereas in the basal hypothalamus it gives rise to the median
eminence and the associated arcuate nucleus, the infundibulum, and the neurohy-
pophysis (Morales-Delgado et al. 2014; Puelles et al. 2012).

3.1.2 How to Study the Developmental Hypothalamic Region

in Xenopus?

3.1.2.1 Experimental Approaches

Many aspects of developmental and cellular neurobiology have been addressed
using Xenopus as the animal model (Gurdon and Hopwood 2000). Analysis of the
3 Development of the Hypothalamus in Xenopus laevis 73

development of the Xenopus brain, including the hypothalamus, has been the basis
of numerous studies using a variety of experimental techniques and protocols. As
with any laboratory model organism, Xenopus possesses standard protocols for
laboratory breeding and husbandry, techniques to modify gene expression (e.g.,
antisense morpholino oligonucleotides), the generation of transgenic animals, and
genome editing (e.g., CRISPR/Cas9 system; Bhattacharya et al. 2015). In addition,
there are currently genetic and genomic data available from Xenopus to analyze
genes, gene families, and gene networks (see Tandon et al. 2017).

3.1.2.2 Initial Neuroanatomical Studies

Using classic histochemical techniques provided abundant and useful information,
but they are not currently considered sufficiently accurate given that the anatomical
references within a particular region, such as the ventricular sulci, the position of the
ventricles, and even the boundaries between regions, were not clear. Fortunately,
new resources (Pearl et al. 2012) have facilitated experiments that have provided
details of gene regulation and patterning via the identification of morphogens,
transcription factors, neuropeptides, and neurotransmitters, which allow identifica-
tion of developing hypothalamic areas and their limits.

3.1.2.3 High Resolution Analysis of Cells

Immunohistochemistry and/or in situ hybridization show the exact position of
different molecules or genes at different developing stages. These techniques,
together with RTPCR, have identified genes that specify different hypothalamic
regions at very early developmental stages, even before the presence of the
neuropeptides and neurotransmitters that functionally define the region. This new
information, together with classic data, such as 3H-thymidine autoradiography birth-
dating techniques (Tay and Straznicky 1982) are making it possible to delineate the
transition from the neural plate to formation of hypothalamic nuclei. Analyses of
neural circuits using the isolated Xenopus nervous system (Kelley et al. 2017),
together with emerging technologies for whole brain imaging (Tomer et al. 2015),
and behavioral tests will significantly advance our understanding of which nuclei
integrate into the circuits that generate behavioral patterns.

3.2 Hypothalamic Organization During the Development

of Xenopus laevis

Throughout development, brain changes can easily be correlated with the striking
body transformations that take place through embryonic and larval development.
These include changes in the hypothalamus that progressively lead to specific
developmental events related to hormonal control, sexual maturity and behavioral
events which functionally are related to maturation of the hypothalamus.
74 N. Moreno and A. González

3.2.1 Developmental Stages

For Xenopus laevis, Nieuwkoop and Faber (1967) described and illustrated in
precise tables, the different stages of embryonic and larval development mainly
based on external anatomy. This staging is still used today (Fig. 3.2, 1). Xenopus
laevis development encompasses embryos (stages 1–45) and larvae (stages 46–66),
at stage 66 metamorphosis is completed and the tadpole becomes a froglet.
According to these tables, from stage 24 the animal begins to show initial motor
reactions to external stimulation, and by stage 26 such movements are spontaneous.
But it is only from stage 28 that changes in brain morphology related to the
hypothalamus are detected, with the hypophyseal primordium (future pituitary)
penetrating beyond the optic stalk, and the infundibulum rapidly developing by
progressive thinning of the ventral wall. From stage 29/30, the chiasmatic ridge is
rostrally separated from the lamina terminalis, and it is just later, at stage 31, that the
hypophyseal cell plate reaches the caudal border of the chiasmatic ridge, but it will
be connected to the ectoderm until stage 32. All of these events occur at early
embryonic developmental stages. Thus, during late embryonic stages, the hypothal-
amus is already present as an anatomical entity, but its maturity is likely linked to
subsequent changes in the animal as a whole. The next main event that takes place
with respect to behavioral responses is the beginning of independent feeding at stage
45, when larval stages begin. At stage 47, the pars intermedia of the hypophysis
begins to differentiate, and at the same stage, the genital ridges begin to protrude into
the coelomic cavity. At stage 49 the infundibular extension of the neurohypophysis
begins to develop and the first anlage of tuberal part of the adenohypophysis can be
observed. From stage 50, the gonadal rudiments form and the thyroid follicles can be
observed, but it is not until stage 52 that sexual differentiation of the gonads begins,
and at stage 55 the first meiotic divisions are observed.
Thus, in Xenopus, the development of the hypothalamus actually takes place over
a relatively long period, throughout embryonic and larval development (Fig. 3.2, 2).
This makes this system a good model to dissect apart the signals that lead to a
mature, functional hypothalamus. Of note, it is known that from the cell blastomere
of stage 16, hypothalamic dopaminergic cells can already be detected (Huang and
Moody 1992). The development of the anterior hypothalamus occurs mainly around
embryonic stages 20–25, while the posterior part develops later, through the late
embryonic and early larval stages 40–52. In contrast, the preoptic area which is
mainly related to reproductive and parental behaviors, and now defined as part of the
telencephalon, develops almost at the end of development, from stage 58 (Tay and
Straznicky 1982).
3 Development of the Hypothalamus in Xenopus laevis 75

3.2.2 The Main Neuroendocrine Hypothalamus: The Paraventricular

Region

3.2.2.1 Developmental Gene Expression Pattern

Neuroendocrine cells that release their neuropeptides directly into the general
circulation in the neurohypophysis or into the portal system supplying the adenohy-
pophysis, are primarily, but not exclusively, situated in the most dorsal region of the
hypothalamus (Figs. 3.1 and 3.3). In this region, groups of cells in the
paraventricular area (Pa) that secrete neuropeptides have been described
(Domínguez et al. 2015). From the earliest stages of development, this region is
defined molecularly by the expression of the gene orthopedia (Otp, Fig. 3.3a). Otp in
the Xenopus embryonic hypothalamus is involved in the specification of this region.
During development, the Otp-positive territory is distinguished from the adjacent
prethalamic (prosomere p3. Fig. 3.3, PTh) and suprachiasmatic (Fig. 3.3, SC)
territories by the lack of expression of Isl1, Nkx2.1, and Shh (Fig. 3.3b, c, f–h
and r). In addition, the expression of Nkx2.2 in the rostral portion of this
area subdivides it into two regions (Pac and Par; Fig. 3.3i–m, Domínguez et al.
2013).

3.2.2.2 Neuropeptides
Neurochemically different cell populations in the Pa of the hypothalamus in Xenopus
include neurons containing several neuropeptides such as somatostatin (SOM;
Fig. 3.3k), mesotocin (MST; Fig. 3.3j; an oxytocin-like peptide), arginine-vasotocin
(AVT), corticotropin-releasing hormone (CRH), thyrotropin-releasing hormone
(TRH), or orexins (González et al. 1995; Domínguez et al. 2015; López et al.
2017). These cell populations project to the pars nervosa of the pituitary gland
(Tuinhof et al. 1994), and a profuse connection exists between this region and the
tuberal hypothalamus (Tub, Figs. 3.1 and 3.4; Domínguez et al. 2013) that may be
involved in regulating ingestion of food. Numerous opsin-positive neurons have
been found in the previously named “hypothalamic magnocellular preoptic
nucleus,” and most of them co-express MST. Axonal processes from these cells
reach the neural lobe of the pituitary gland, suggesting that in Xenopus,
mesotocinergic hypothalamic neurons, which are directly involved in photorecep-
tion and could be regulating seasonal adaptation changes as well. (Álvarez-Viejo
et al. 2003). Independent of function, the early emergence of cells expressing TRH
or MST and SOM, and the presence of Otp has been correlated (Fig. 3.3j, k;
Domínguez et al. 2013, 2015), suggesting that the transcription factor Otp may
participate in the acquisition of their phenotypes. Thus, in both Xenopus and mouse,
transcription factors are involved in the specification of hypothalamic postmitotic
populations and in the differentiation of independent nuclei within this territory.
76 N. Moreno and A. González

Fig. 3.3 Alar hypothalamic regions identified by molecular markers. Upper panel:
Photomicrographs of sagittal (c, e–h, l, m) and transverse sections (a–d0 , i–k, n–g) through the
developing Xenopus alar hypothalamic region showing the single expressions of Otp (a), xShh
(e, n), and Nkx2.1 (f, o), and the combined expressions of Isl1 and Nkx2.1 (g), Isl1/Otp (b, c, h),
Nkx2.2/Otp (d, i), Nkx2.2/MST (j), Nkx2.2/SOM (k), xDll4/Otp (l), Nkx2.1/Nkx2.2 (m), xShh/TH
(p), and Nkx2.1/TH (q). The arrowheads in d0 and i indicate double labeled cells, and in (a) indicate
Otp ventricular expressing cells. The yellow line in c indicates the level of the section shown in c0 .
Developmental stages are indicated in each photograph. Lower panel: Schematic representation (r)
summarizes the combinatorial code of markers present in the alar hypothalamus and its adjacent
domains in lateral and transverse views (r). MST, mesotocin; TH, tyrosine hydroxylase. Scale bars
in a, b, e, h, j–m ¼ 50 μm; in c, d, g, i, n–q ¼ 100 μm. All photomicrographs are modified from
Domínguez et al. (2013)
3 Development of the Hypothalamus in Xenopus laevis 77

Fig. 3.4 Basal hypothalamic regions identified by molecular markers. Upper panel:
Photomicrographs: transverse (c, e, g, h, i, k–p) and sagittal (a, b, d, f, j, q) sections through the
developing Xenopus basal hypothalamic region showing the single expressions of xShh (a) and the
combined expressions of Isl1 and Nkx2.1 (b, c), Isl1 and Otp (d, e), Nkx2.2 and Otp (f, g), Pax7 and
Nkx2.1 (h–k), Pax7 and Otp (l), TH and Nkx2.1 (m), Shh (n), Isl1 (o), Nkx2.2 (p, q). The
arrowheads in h, k, and m indicate double labeled cells. The red line in q indicates the alar/basal
boundary. Developmental stages (st) are indicated in each photograph. Lower panel: schematic
representation summarizes the combinatorial code of markers present in the basal hypothalamus
and its adjacent domains in lateral and transverse views. TH: tyrosine hydroxylase. Scale bars in a–
i; k–q ¼ 100 μm; in j ¼ 50 μm. All the photomicrographs are modified from Domínguez et al.
(2014)
78 N. Moreno and A. González

3.2.3 The Suprachiasmatic Hypothalamus as the Internal Clock

During Development: The Subparaventricular Hypothalamus

3.2.3.1 Gene Expression Patterns and Connections

The subparaventricular region (Spa; Figs. 3.1 and 3.3) of the alar hypothalamus is
the most ventral part and contains the suprachiasmatic nucleus (SC; Domínguez
et al. 2013). During development, this region is characterized by the expression of
Dlx genes, which control the expression of GABAergic populations, and Isl1
(Fig. 3.3l; Domínguez et al. 2013). The expression of transcription factors Nkx2.1
and Nkx2.2 subdivides the suprachiasmatic nucleus into rostro-caudal subdomains
(Fig. 3.3m; SCr and SCc). In addition, the expression of the morphogen sonic
hedgehog (Shh) in the rostral portion (SCr; Fig. 3.3n) overlaps with the expression
of Nkx2.1 (Fig. 3.3o), suggesting a regulatory role of Shh through Nkx2.1 actions
(Domínguez et al. 2013). The suprachiasmatic region in Xenopus is also
characterized during development by the presence of catecholaminergic cell groups
restricted to the rostral domain (identified by expression of tyrosine hydroxylase,
TH, Fig. 3.3p, q; Moreno et al. 2012; Domínguez et al. 2013). Furthermore, the
suprachiasmatic region has important connections with the tuberal hypothalamus
(Tub, Fig. 3.1) and the mammillary complex (M, Fig. 3.1; Domínguez et al. 2013).

3.2.3.2 Suprachiasmatic Nucleus Functions

Functionally, the SC nucleus is involved in adaptation of an animal to its surround-
ings, a very important mechanism for wildlife survival. The development and
structure of this nucleus has been studied in detail in Xenopus. Specially, a group
of SC neurons inhibits the release of the α-melanophore stimulating hormone
(α-MSH) from the neuroendocrine melanotropic cells of the pituitary gland.
α-MSH is the peptide that causes darkening of the skin during adaptation to a dark
background. In the SC, ventrolateral, dorsomedial, and caudal groups have been
described, based on different cell groups containing neuropeptide Y (NPY). It has
been proposed that in black-adapted animals, NPY-positive neurons in the ventro-
lateral group (known to synaptically inhibit the melanotrope cells in white-adapted
animals) are inhibited by interneurons in the dorsomedial group that contain NPY.
To complicate the circuit further, the NPY-positive neurons in the caudal group have
secretory dynamics similar to dorsomedial NPY neurons, suggesting that they may
also play a role in background adaptation, but probably different from that exerted by
the ventrolateral and dorsomedial group (Kramer et al. 2001).
These SC NPY inhibitory neurons, designated suprachiasmatic melanotrope-
inhibiting neurons (SMINs), possess more and larger synapses on melanotrope
cells in white adapted compared with black-background adapted animals. In addi-
tion, in the latter animals, the melanotropes are larger and produce more
proopiomelanocortin, the precursor of α-MSH. On a white background, presynaptic
SMIN plasticity is reflected by a higher expression of inhibitory NPY and is closely
associated with postsynaptic melanotrope plasticity, namely with a higher expres-
sion of the NPY Y1 receptor. Interestingly, the melanotrope cells in these animals
also show a higher expression of receptors for TRH and urocortin 1, two
3 Development of the Hypothalamus in Xenopus laevis 79

neuropeptides that stimulate α-MSH secretion. Possibly, in white-adapted animals

melanotropes are sensitized to neuropeptide stimulation, so that when the frog
moves to a black background, they can immediately initiate the secretion of
α-MSH to achieve rapid adaptation to the new background condition.

3.2.4 The Basal Hypothalamus: The Tuberal and Mammillary

Territories

3.2.4.1 Tuberal Region Molecular Expression Pattern

In the basal hypothalamus (Fig. 3.4), the tuberal region is the most rostral portion.
From early embryonic stages, it is molecularly characterized by the expression of
Shh and Nkx2.1 close to the ventricles and by Isl1 expression in the adjacent zone
(Fig. 3.4a–c; Domínguez et al. 2014). Furthermore, starting at embryonic stages, the
transcription factor Otp is located exclusively in the rostral tuberal portion (RT;
Fig. 3.4d–e), which constitutes the anterobasal/retrochiasmatic region that gives rise
to the arcuate nucleus. In addition, the caudal domain is characterized by a lack of
Otp and the expression of Nkx2.2 (CT; Fig. 3.4f, g; Domínguez et al. 2014). In the
caudal domain, both Isl1 and Nkx2.2 have been related to the development of the
ventromedial nucleus, including specification of the basal phenotype, specification
of the ventromedial fate, and regulation of reproductive behavior through the
estrogen receptor α (Davis et al. 2004; Kurrasch et al. 2007). In the tuberal hypo-
thalamus, Otp in SOM cell specification has also been described, as in the Pa area
located dorsally. These tuberal SOM cells probably migrate into the arcuate nucleus.
Similarly, TRH cells have been found in the tuberal region of anurans, which
probably originated in Pa of the alar region (Domínguez et al. 2015). Finally,
expression of Dll4 in the more caudal part of the tuberal region in Xenopus is closely
related to the positive GABAergic cell population detected in this location,
suggesting an involvement of Dll4 in the specification of the GABAergic phenotype
in the tuberal region, as in other areas (Domínguez et al. 2014).

3.2.4.2 Mammillary Region Expression Pattern

The boundary between the tuberal and the mammillary territories in amphibians is
mainly defined by the lack of Isl1 expression within the continuous Nkx2.1 positive
tuberal/mammillary region (Fig. 3.4b–e, r; Domínguez et al. 2014). In addition, both
regions can also be distinguished by differential expression of Otp, which is
expressed in the mammillary region and is lacking in the caudal-most tuberal domain
(Fig. 3.4d–g; Domínguez et al. 2014). The mammillary band has been divided into
two different regions: the proper mammillary region (Ma), which occupies a most
rostral position and is Nkx2.1+/Shh-, and the retromammillary region (RM) located
more caudally, where the expression patterns of Shh and Nkx2.1 are inverted being
Nkx2.1-/Shh+ (Fig. 3.4a–c, i, j). This Ma/RM subdivision is demarcated by the
expression of Nkx2.2, Otp, Dll4, and Lhx1 in Ma (Fig. 3.4a–l, p, q). In addition,
some scattered Pax7+ cells are present in the subventricular zone of Ma. The cells
80 N. Moreno and A. González

likely originate from the adjacent Pax7 expressing population of the P3 basal plate
(Fig. 3.4h–l), which indicates that part of the mammillary region arises in the
prethalamic tegmentum. Fate map analysis is needed to confirm this. In Xenopus,
the Lhx7 expression in the alar hypothalamus continues ventrally reaching the
mammillary territory (Domínguez et al. 2013), suggesting the existence of a
tuberomammillary terminal zone in anurans as in mammals (Domínguez et al.
2014). The presence of a catecholaminergic cell population (TH positive, González
et al. 1994; Domínguez et al. 2014) that coexpresses Nkx2.1 and Otp (Fig. 3.4n–q)
also characterizes the mammillary region. As for the specification of this cell
population, there is currently controversy about its origin, but several data support
a contribution of diencephalic areas to the mammillary territory since Pax7
expressing cells likely originated in P3 and subsequently colonized the mammillary
region (Domínguez et al. 2014).

3.3 Concluding Remarks and Perspectives: Xenopus, a Good

Model for Vertebrates or a Distinct Amphibian Condition?

In multiple species analyzed in recent years, it has been described that the hypothal-
amus includes comparable molecular compartments, and each compartment shows a
tendency to a common organization with respect to the profile of molecular expres-
sion, but at the same time, the hypothalamic organization is subject to evolutive/
adaptive changes/modifications during the anamnio-amniotic transition. In this
context, the evolutionary leap of Xenopus, an amphibian (amphibia are the only
group of anamniote tetrapods), to the rest of the tetrapods, implies relevant adaptive
changes for the conquest of the water-independence of the new environment, and
therefore it has some obvious consequences in the organization of the brain. The
main common features of the hypothalamus are its position and regionalization,
since in all models it is topographically located rostral to the diencephalon and it is
divided into alar and basal regions. In addition, each of them is rostro-caudally
subdivided into two different domains according to molecular criteria (reviewed
in Domínguez et al. 2015). Moreover, the chemoarchitectonic profiles of the
territories functionally define them by the presence of multiple postmitotic cell
populations occupying neuroendocrine territories. In contrast, in amniotes a
preoptohypothalamic boundary is evident, which is likely related to the higher
complexity of the paraventricular area that starts to express transcription factors
such as Pax6. In addition, in this alar territory, the restricted expression of Nkx2.1 in
the subparaventricular region observed in Xenopus is gradually more restricted
reaching a total lack in mammals, which could be related to the gradual pallial and
thalamic expansion that take place in amniotes.
3 Development of the Hypothalamus in Xenopus laevis 81

Key References

Bhattacharya et al. (2015)—This study demonstrates that the CRISPR/Cas9 system

in Xenopus can efficiently modify both alleles in the F0 generation.
Buchholz (2017)—The study, describing the roles of thyroid hormones and their
receptors in the tissue transformations comprising development, highlights the
key experimental findings and endocrine, molecular, and genomic resources
available in Xenopus.
Domínguez et al. (2015)—This article summarizes the main data on the expression
patterns of different transcription factors and neuroactive substances in the
developing hypothalamus of Xenopus laevis and the turtle Pseudemys scripta
Ferran et al. (2015)—Review of the main distinct dorsoventral progenitor domains
detected in the mouse developing hypothalamus based on gene expression
markers.
Puelles et al. (2012)—This chapter constitutes one of the most complete revision of
the mouse hypothalamic organization.
Tandon et al. (2017)—This article reports the main advantages in Xenopus to study
multiple aspects of cell and developmental biology and demonstrates its useful-
ness for genetic manipulation approaches.

References
Álvarez-Viejo M, Cernuda-Cernuda R, DeGrip WJ, Álvarez-Lopez C, García-Fernández JM (2003)
Co-localization of mesotocin and opsin immunoreactivity in the hypothalamic preoptic nucleus
of Xenopus laevis. Brain Res 969(1–2):36–43. https://doi.org/10.1016/S0006-8993(03)02273-X
Bhattacharya D, Marfo CA, Li D, Lane M, Khokha MK (2015) CRISPR/Cas9: an inexpensive,
efficient loss of function tool to screen human disease genes in Xenopus. Dev Biol 408
(2):196–204. https://doi.org/10.1016/j.ydbio.2015.11.003
Buchholz DR (2017) Xenopus metamorphosis as a model to study thyroid hormone receptor
function during vertebrate developmental transitions. Mol Cell Endocrinol 459:64–70. https://
doi.org/10.1016/j.mce.2017.03.020
Davis AM, Seney ML, Walker HJ, Tobet SA (2004) Differential colocalization of Islet-1 and
estrogen receptor alpha in the murine preoptic area and hypothalamus during development.
Endocrinology 145(1):360–366. https://doi.org/10.1210/en.2003-0996
Domínguez L, Morona R, González A, Moreno N (2013) Characterization of the hypothalamus of
Xenopus laevis during development. I. The alar regions. J Comp Neurol 521(4):725–759.
https://doi.org/10.1002/cne.23222
Domínguez L, González A, Moreno N (2014) Characterization of the hypothalamus of Xenopus
laevis during development. II. The basal regions. J Comp Neurol 522(5):1102–1131. https://doi.
org/10.1002/cne.23471
Domínguez L, González A, Moreno N (2015) Patterns of hypothalamic regionalization in
amphibians and reptiles: common traits revealed by a genoarchitectonic approach. Front
Neuroanat 9:3. https://doi.org/10.3389/fnana.2015.00003
Ferran JL, Puelles L, Rubenstein JL (2015) Molecular codes defining rostrocaudal domains in the
embryonic mouse hypothalamus. Front Neuroanat 9:46. https://doi.org/10.3389/fnana.2015.
00046
González A, Marin O, Tuinhof R, Smeets WJ (1994) Ontogeny of catecholamine systems in the
central nervous system of anuran amphibians: an immunohistochemical study with antibodies
82 N. Moreno and A. González

against tyrosine hydroxylase and dopamine. J Comp Neurol 346(1):63–79. https://doi.org/10.

1002/cne.903460105
González A, Muñoz A, Muñoz M, Marín O, Smeets WJ (1995) Ontogeny of vasotocinergic and
mesotocinergic systems in the brain of the South African clawed frog Xenopus laevis. J Chem
Neuroanat 9(1):27–40. https://doi.org/10.1016/0891-0618(95)00063-d
Gurdon JB, Hopwood N (2000) The introduction of Xenopus laevis into developmental biology: of
empire, pregnancy testing and ribosomal genes. Int J Dev Biol 44(1):43–50
Huang S, Moody SA (1992) Does lineage determine the dopamine phenotype in the tadpole
hypothalamus?: a quantitative analysis. J Neurosci 12(4):1351–1362
Kelley DB, Elliott TM, Evans BJ, Hall IC, Leininger EC, Rhodes HJ et al (2017) Probing forebrain
to hindbrain circuit functions in Xenopus. Genesis 55(1–2). https://doi.org/10.1002/dvg.22999
Kramer BM, Welting J, Berghs CA, Jenks BG, Roubos EW (2001) Functional organization of the
suprachiasmatic nucleus of Xenopus laevis in relation to background adaptation. J Comp Neurol
432(3):346–355
Kurrasch DM, Cheung CC, Lee FY, Tran PV, Hata K, Ingraham HA (2007) The neonatal
ventromedial hypothalamus transcriptome reveals novel markers with spatially distinct pattern-
ing. J Neurosci 27(50):13624–13634. https://doi.org/10.1523/Jneurosci.2858-07.2007
Lee-Liu D, Mendez-Olivos EE, Munoz R, Larrain J (2017) The African clawed frog Xenopus
laevis: a model organism to study regeneration of the central nervous system. Neurosci Lett
652:82–93. https://doi.org/10.1016/j.neulet.2016.09.054
López JM, Morales L, González A (2017) Spatiotemporal development of the orexinergic
(hypocretinergic) system in the central nervous system of Xenopus laevis. Brain Behav Evol
88(2):127–146. https://doi.org/10.1159/000449278
Morales-Delgado N, Castro-Robles B, Ferran JL, Martínez-de-la-Torre M, Puelles L, Díaz C (2014)
Regionalized differentiation of CRH, TRH, and GHRH peptidergic neurons in the mouse
hypothalamus. Brain Struct Funct 219(3):1083–1111. https://doi.org/10.1007/s00429-013-
0554-2
Moreno N, Domínguez L, Morona R, González A (2012) Subdivisions of the turtle Pseudemys
scripta hypothalamus based on the expression of regulatory genes and neuronal markers.
J Comp Neurol 520(3):453–478. https://doi.org/10.1002/cne.22762
Nieuwkoop PD, Faber J (1967) Normal table of Xenopus laevis (Daudin). North Holland,
Amsterdam
Pearl EJ, Grainger RM, Guille M, Horb ME (2012) Development of Xenopus resource centers: the
National Xenopus Resource and the European Xenopus Resource Center. Genesis 50
(3):155–163. https://doi.org/10.1002/dvg.22013
Puelles L, Rubenstein JL (2003) Forebrain gene expression domains and the evolving prosomeric
model. Trends Neurosci 26(9):469–476. https://doi.org/10.1016/S0166-2236(03)00234-0
Puelles L, Martínez de la Torre M, Bardet S, Rubenstein JLR (2012) Hypothalamus. In: Watson C,
Paxinos G, Puelles L (eds) The mouse nervous system. Academic-Elsevier, San Diego, pp
221–313
Tandon P, Conlon F, Furlow JD, Horb ME (2017) Expanding the genetic toolkit in Xenopus:
approaches and opportunities for human disease modeling. Dev Biol 426(2):325–335. https://
doi.org/10.1016/j.ydbio.2016.04.009
Tay D, Straznicky C (1982) The development of the diencephalon in Xenopus. An autoradiographic
study. Anat Embryol (Berl) 163(4):371–388
Tomer R, Khairy K, Amat F, Keller PJ (2015) Quantitative high-speed imaging of entire developing
embryos with simultaneous multiview light-sheet microscopy. Nat Methods 9(7):755–763.
https://doi.org/10.1038/nmeth.2062
Tuinhof R, González A, Smeets WJ, Scheenen WJ, Roubos EW (1994) Central control of
melanotrope cells of Xenopus laevis. Eur J Morphol 32(2–4):307–310
Gene Regulatory Programs
in the Development of Hypothalamic 4
Arcuate Nucleus Neurons

Jae W. Lee, Christian Huisman, and Seunghee Lee

Abstract
Neurons in the hypothalamic arcuate nucleus (ARC) relay and translate important
cues from the periphery into the central nervous system to maintain critical
homeostatic processes such as energy balance, growth, and reproductive
behaviors. Despite extensive studies dedicated to unraveling the physiological
functions of various neuronal types in the ARC, the gene regulatory networks that
drive their development are still poorly understood. This chapter provides a
concise overview of the transcription factors that control the development of
ARC neurons, followed by discussions of several outstanding challenges in the
field and future directions to address those issues using the recent technological
advances in genome-wide studies.

Keywords
Arcuate nucleus · Gene regulation · Transcription factor · POMC · AgRP · NPY ·
GHRH

J. W. Lee (*)
Department of Biologial Sciences, University at Buffalo, Buffalo, NY, USA
e-mail: [email protected]
C. Huisman
Department of Pediatrics, Oregon Health & Science University, Portland, OR, USA
S. Lee
College of Pharmacy, Research Institute of Pharmaceutical Sciences, Seoul National University,
Seoul, Korea

# Springer Nature Switzerland AG 2020 83

S. Wray, S. Blackshaw (eds.), Developmental Neuroendocrinology, Masterclass in
Neuroendocrinology 9, https://doi.org/10.1007/978-3-030-40002-6_4
84 J. W. Lee et al.

4.1 Introduction

The arcuate nucleus of the hypothalamus (ARC) is located in the mediobasal

hypothalamus. It is adjacent to the third ventricle and the median eminence, and it
contains neuronal cell types that play critical roles in diverse physiological homeo-
static processes, such as energy balance, growth, and reproductive behaviors
(Biebermann et al. 2012; Bluet-Pajot et al. 2001; Hill et al. 2008; Hrabovszky
2014). In particular, five types of arcuate neurons have been extensively
characterized over the past few decades (Fig. 4.1). Arcuate neurons expressing
growth hormone releasing hormone (GHRH) trigger secretion of growth hormone
(GH) in the anterior pituitary gland (Bluet-Pajot et al. 2001). GH induces the hepatic
expression of insulin-like growth factor 1 (IGF1), which controls bone epiphyses,
growth plate development, muscle and adipose tissue development and glucose
homeostasis (Cohen and Rosenfeld 1994). Neurons expressing the orexigenic
neuropeptides “agouti-related peptide (AgRP)” and neuropeptide Y (NPY), herein
termed AgRP-neurons, enhance food intake and reduce energy expenditure, while
neurons expressing the anorexigenic neuropeptides α-melanocyte stimulating hor-
mone (αMSH, cleaved from pro-opiomelanocortin, POMC) and “cocaine and
amphetamine regulated transcript” (CART) inhibit food intake and increase energy
consumption (Biebermann et al. 2012). Neurons expressing Kiss1 activate secretion
from gonadotropin releasing hormone (GnRH)-neurons, thereby playing essential
roles in reproduction (Hill et al. 2008). Tuberoinfundibular dopamine (TIDA)-
neurons in the ARC regulate the secretion of prolactin from the pituitary gland
(Voogt et al. 2001).
Recently, 24 neuronal cell types were cataloged in the adult ARC using the Drop-
seq technology that can simultaneously perform genome-wide analysis of RNA
expression in thousands of individual cells (Campbell et al. 2017) (Box 4.1). This
study also revealed heterogeneity of arcuate neurons that had been previously
considered a homogeneous population of cells, including two types of AgRP-
neurons, three of POMC-neurons, and six of tyrosine hydroxylase (TH)-expressing

Fig. 4.1 Different neurons populate the arcuate nucleus (ARC). The ARC contains as many as
24 neuronal cell types according to the recent scRNA-seq results (Campbell et al. 2017). Only a few
types of neurons have been extensively characterized, including tuberoinfundibular dopaminergic
(TIDA)-, agouti-related peptide (AgRP)-, pro-opiomelanocortin (POMC)-, growth hormone
relasing hormone (GHRH)- and Kiss1-neurons. GH growth hormone, GnRH gonadotropin releas-
ing hormone
4 Gene Regulatory Programs in the Development of Hypothalamic Arcuate. . . 85

neurons (Campbell et al. 2017). Specific roles for each of these ARC neuronal
populations are just beginning to be addressed. For instance, in female mice,
AgRP-neurons that co-express somatostatin (Sst) were shown to specifically mediate
corticotropin-releasing factor (CRF)-driven thermogenesis to cope with environ-
mental challenges (Kuperman et al. 2016). Unfortunately, the gene regulatory
networks that direct the development of arcuate neurons, which are the subject of
this chapter, are poorly understood. Clearly, this information is needed to enable our
understanding of the developmental trajectories through which common progenitors
differentiate to distinct types of arcuate neurons (Biebermann et al. 2012; Bluet-Pajot
et al. 2001; Hill et al. 2008; Hrabovszky 2014) and provide crucial insight into how
and when the seemingly homogeneous populations of neurons segregate to multiple,
functional subpopulations (Campbell et al. 2017).
Development of the hypothalamus involves multiple organizing signals and
transcription factors. In mice, prior to embryonic day 10 (E10), the diencephalic
region is separated from surrounding regions by the influence of organizing signals
such as Wnts (for wingless/integrins), Shh (for sonic hedgehog), Bmps (for bone
morphogenetic proteins, and Fgfs (for fibroblast growth factors). These organizing
signals eventually induce the expression of the homeodomain (HD) transcription
factor (TF) Nkx2.1, which is a key marker of hypothalamic tissue (Martinez-Ferre
and Martinez 2012). During the subsequent neurogenic period (E10–16), neural
stem cells in the ventricular zone give rise to neural precursors (Ifft 1972; Ishii and
Bouret 2012). Because neurons, astrocytes, and oligodendrocytes all derive from the
same progenitor pool, the proneural TFs that characterize the neurogenic period
simultaneously drive neurogenesis and suppress gliogenesis (Rowitch and Kriegstein
2010). Notably, neurogenesis in the hypothalamus proceeds in an outside-in fashion
(Altman and Bayer 1978). Neurons in the most lateral nuclei, such as the lateral
hypothalamus area (LHA), are generated first. This neurogenic period is followed by
birth of cells that will reside in more medially located nuclei such as the dorsomedial
nucleus (DMH) and the ventromedial nucleus (VMH), and finally by neurons in
nuclei along the midline such as the ARC (Byerly and Blackshaw 2009).
In this chapter, we give an overview of the findings related to the gene regulatory
programs in the development of various arcuate neuronal types, which have been
elucidated over the past few decades. We then highlight several outstanding
challenges in the field, together with future studies to address those issues using
the recent technological breakthroughs in genome-wide studies.

4.2 TFs in Progenitors and Fate Specification of ARC Neurons

A number of TFs, including HD and basic-helix-loop-helix (bHLH), are expressed

in spatially and temporally distinct patterns in the developing arcuate neurons and
play crucial roles in their development (Table 4.1; Fig. 4.2). It should be emphasized
that many of these TFs are also involved in the development of other regions of the
brain, although our discussion in this chapter is mostly focused on their develop-
mental function of ARC neurons.
86 J. W. Lee et al.

Table 4.1 Transcription factors in the development of ARC neurons

Transcription factors Type ARC phenotypes in KO models
Nkx2.1 HD No ARC
Rax HD POMC#
Dbx1 HD AgRP#
Isl1 HD POMC/AgRP/GHRH#
Bsx HD AgRP#
Gsx1 HD GHRH#
Otp HD AgRP#
Dlx1/2 HD GHRH#
Ngn3 bHLH POMC#
Nhlh2 bHLH POMC#
Mash1 bHLH GHRH# POMC#
Rbpjk POMC" AgRP"
Ikaros Zinc-finger GHRH#
AgRP agouti-related peptide, Arc arcuate nucleus, bHLH basic-helix-loop-helix, GHRH growth
hormone releasing hormone, HD homeodomain, KO gene knockout, POMC pro-opiomelanocortin

Fig. 4.2 Expression of Otp, NPY, Dlx1, GHRH in the developing arcuate nucleus (ARC). The
expression in transverse brain sections of (left) the transcription factors (TFs) Otp (red) and Dlx1
(green) detected with immuno-fluorescent histochemistry (IHC); and (right) mRNAs for
neuropeptides neuropeptide Y (NPY (a marker of agouti-related peptide (AgRP)-neurons, which
derive from Otp+ cells) and growth hormone releasing hormone (GHRH, a marker of GHRH-
neurons, which derive from Dlx1+ cells) using in situ hybridization (ISH). The ARC region is
denoted, left, as a yellow (dashed) triangle, and right, by autoradiographic sliver grains (black).
E15.5, 17.5: embryonic days of development
4 Gene Regulatory Programs in the Development of Hypothalamic Arcuate. . . 87

4.2.1 Nkx2.1 (aka TTF-1 for Thyroid Transcription Factor-1)

The HD TF Nkx2.1 marks hypothalamic tissue (Martinez-Ferre and Martinez 2012).

By E10.5, Nkx2.1 mRNA is detected in the progenitor zones of the rostroventral
diencephalon and rostroventral telencephalon. Nkx2.1 mRNA and protein are
expressed in both neuroepithelial progenitors and subsets of their postmitotic
neurons (Lazzaro et al. 1991; Marin et al. 2000; Price et al. 1992; Sussel et al.
1999). By E14.5, Nkx2.1 is detectable in the ARC and persists in this region
postnatally (Marin et al. 2000). Accordingly, transgenic mice, in which LacZ
expression is driven by approximately a 4 kb fragment of potential regulatory
DNA from the 50 flanking region of Nkx2.1, show robust expression of LacZ in
the ARC (Pan et al. 2004). In Nkx2.1-deficient mice, the development of the ARC is
absent, while other hypothalamic nuclei are present (Kimura et al. 1996; Marin et al.
2002; Sussel et al. 1999), suggesting that Nkx2.1 is essential for the formation of
ARC during development.

4.2.2 Retina and Anterior Neural Fold Homeobox (Rax, aka Rx)

The HD TF Rax, a key regulator of early retinal development (Zhang et al. 2000),
also plays crucial roles in the formation of the ventral neuroendocrine hypothalamus
(Muranishi et al. 2012). Nkx2.1-expressing neurons, which include many develop-
ing neurons in the ARC, appear to be derived from Rax-expressing cells. Elimination
of Rax throughout ARC progenitors in mice results in loss of ventral hypothalamic
expression of Nkx2.1 and POMC (Lu et al. 2013). These results suggest that Rax
controls the production of POMC and possibly other ARC neuronal cell types, given
the finding that POMC-derived progenitor cells differentiate into multiple ARC
neuronal types (Padilla et al. 2010; Sanz et al. 2015).

4.2.3 Developing Brain Homeobox 1 (Dbx1)

Standard and inducible fate-mapping revealed that progenitors expressing the HD

TF Dbx1 produce some populations of both AgRP- and POMC-neurons in the
developing ARC (Sokolowski et al. 2016). Studies using Nkx2.1-Cre mice
(Xu et al. 2008) to produce Dbx1 hypothalamic-specific conditional knockout
(cKO) mice showed a significant decrease in ARC NPY expression at E13.5 while
no changes in POMC expression were detected (Sokolowski et al. 2015). These
results suggest that, while Dbx1+ progenitors generate both AgRP- and POMC-
neurons in the ARC, Dbx1 only plays a crucial role in the production of AgRP-
neurons (NPY positive). In contrast, Dbx1-independent mechanisms are required for
specification of the POMC-neurons in the ARC.
88 J. W. Lee et al.

4.2.4 Isl1

The HD TF Isl1 is required for the development of POMC-neurons (Nasif et al.

2015). Multiple changes are found in mice lacking Isl1 in Nkx2.1+ expressing cells
(using Nkx2.1-Cre mice). These mice show loss of the orexigenic neuropeptides
AgRP and NPY, the anorexigenic neuropeptide αMSH, and the growth-stimulatory
peptide GHRH, accompanied by severe deficits in both feeding and linear growth
(Lee et al. 2016). Isl1 can directly enhance the expression of AgRP by cooperating
with two key orexigenic TFs, glucocorticoid receptor (GR) and brain-specific
homeobox factor (Bsx) (Lee et al. 2016). Isl1 also activates the expression of
POMC in the ARC by directly binding to the neuronal Pomc enhancers nPE1 and
nPE2 (Lam et al. 2015; Nasif et al. 2015). These data indicate that Isl1 is involved in
the specification of AgRP-neurons, POMC-neurons, and GHRH-neurons within the
ARC (Lee et al. 2016).

4.2.5 Bsx

The HD TF Bsx is expressed in the ARC as early as E15.5 (Cremona et al. 2004). In
the adult mouse brain, Bsx expression is restricted to the hypothalamus with
prominent expression in the ARC and the DMH as well as some cells in the LHA
(Sakkou et al. 2007). In Bsx mutant brains, the expression of AgRP and NPY is
downregulated but not abolished (Sakkou et al. 2007). Although it is unclear if Bsx
is required for the early fate specification of AgRP-neurons, it is possible that Bsx
and Isl1 have essential but redundant roles in the fate specification of AgRP-neurons
as well as their upregulation of AgRP (Lee et al. 2016).

4.2.6 Gsx1 (aka Gsh1)

Mice lacking the HD TF Gsx1 show severe deficits in pituitary development,

accompanied by extreme dwarfism (Li et al. 1996). Interestingly, Gsx1 is expressed
in the regions of the diencephalon, which develops into the thalamus and hypothala-
mus but not in embryonic precursors of the pituitary (Valerius et al. 1995). These
results suggest that the deficits in pituitary development likely reflect the role(s) of
Gsx1 in the hypothalamus. Indeed, Gsx1 directly controls the expression of GHRH
in the ARC (Li et al. 1996).

4.2.7 Orthopedia (Otp)

In the mouse (from E12.5), the HD TF Otp is expressed in neurons that give rise to
the PVN, supraoptic (SON), anterior periventricular (aPV), and ARC nuclei
(Simeone et al. 1994). Otp-null mice die soon after birth and display progressive
impairment of crucial developmental events, including reduced cell proliferation,
4 Gene Regulatory Programs in the Development of Hypothalamic Arcuate. . . 89

abnormal cell migration. In these mice, terminal differentiation of the parvocellular

and magnocellular neurons of the aPV, PVN, SON, and ARC is disrupted
(Acampora et al. 1999). In the ARC of Otp-null brains, the expression of Sst but
not GHRH is abolished (Acampora et al. 1999; Wang and Lufkin 2000). We recently
reported that Opt is highly enriched in AgRP-neurons and plays a critical role in their
development (Lee et al. 2018). Together, these data suggest a role of Otp in the
production of AgRP-neurons and other Sst+ neuronal types in the ARC during
development.

4.2.8 Distal-Less Homeobox-1 (Dlx1) and Dlx2

The HD TFs Dlx1 and Dlx2 are expressed in the region of the developing and adult
ARC (Eisenstat et al. 1999; Saino-Saito et al. 2003). Dlx1/2 plays important and
redundant roles in the fate specification of GABA+ and TH+ neurons in the forebrain
(Andrews et al. 2003; Cobos et al. 2007; Long et al. 2007; Petryniak et al. 2007; Qiu
et al. 1995). The immunochemical analyses with a pan-Dlx antibody detecting
multiple Dlx factors (including Dlx1/2) reveal that Dlx factors are also expressed
in TH+ neurons and non-AgRP-neuronal type GABA+ neurons in the ARC (Yee
et al. 2009). Adult Dlx1-null mice show a two-fold reduction in the number of TH+
neurons in the ARC (Yee et al. 2009), whereas the role of Dlx2 in the hypothalamus
remains unexplored. We recently reported that Dlx1 and Dlx2 are highly expressed
in TH+ neurons in the ARC, including GHRH-neurons and play critical roles in the
production of GHRH-neurons (Lee et al. 2018).

4.2.9 Neurogenin-3 (Ngn3)

Proneural bHLH TF Ngn3 expression begins as early as E9.5 and disappears by

E17.5 (Pelling et al. 2011). Fate mapping studies show that many of the earliest born
POMC-neurons, around E10.5, derive from Ngn3-expressing progenitors, while
later born POMC-neurons develop from non-Ngn3 expressing progenitors (Pelling
et al. 2011). A large fraction of AgRP-neurons also appears to derive from Ngn3-
expressing progenitors (Pelling et al. 2011). Ngn3-null mice show a substantial
reduction in the number of POMC-neurons born between E10.5 and 13.5, although
this number rebounds somewhat by E17.5 (Pelling et al. 2011). Other cell types in
the ARC, such as GHRH-neurons, remained unaffected by deletion of Ngn3. When
Ngn3 is knocked out only in Nkx2.1+ cells using the inducible Nkx2.1-iCre mice
(Tsai et al. 2012), the resulting cKO mice show a reduced number of POMC-neurons
beginning at E10.5 that continues into the adulthood, thereby developing early-onset
obesity and hyperphagia (Anthwal et al. 2013). These results demonstrate critical
roles of Ngn3 in the development of POMC-neurons.
90 J. W. Lee et al.

4.2.10 Nhlh2 (aka NSCL-2)

The bHLH TF Nhlh2 is expressed in the developing nervous system as early as E9.5
(Begley et al. 1992; Gobel et al. 1992). Nhlh2-null mice show hyperphagia and
obesity as well as infertility (Good et al. 1997). These mice show no alteration in
expression of POMC mRNA, although their αMSH levels are reduced, which is due
to a reduction in the expression and activity of prohormone convertases (Pc1/3) that
cleave POMC to adrenocorticotropic hormone (ACTH) and αMSH (Jing et al.
2004). The infertility of Nhlh2-null mice is explained by the finding that migration
of GnRH-neurons to their final hypothalamic sites is significantly reduced in Nhlh1/
Nhlh2-double KO mice (Kruger et al. 2004). Interestingly, POMC-neurons are
reduced in mice when Nhlh2 is deleted specifically in GnRH-expressing neurons
(Schmid et al. 2013), suggesting a critical role for GnRH cells in the development of
POMC-neurons.

4.2.11 Mash1 (aka Ascl1 for Achaete-Scute Complex 1)

The proneural bHLH TF Mash1 is expressed throughout the basal retrochiasmatic

neuroepithelium and loss of Mash1 results in hypoplasia of the ARC (McNay et al.
2006). This is due to a failure of neurogenesis and an increase in apoptosis, defects
that are rescued by ectopic expression of Ngn2 (another proneural bHLH TF) under
the control of the Mash1 promoter (McNay et al. 2006). In addition to its role in
neurogenesis, Mash1 is specifically required for the expression of Gsx1 (McNay
et al. 2006), which drives the expression of GHRH (Li et al. 1996). Although Mash1
is not required for POMC expression, it is required for normal development of
POMC-neurons (McNay et al. 2006). These data demonstrate that Mash1 is both
required for the generation of ventral neuroendocrine neurons as well as for specifi-
cation of these neurons.

4.2.12 Rbpjk

Rbpjk is a key downstream effector TF of Notch signaling (Lasky and Wu 2005).

When the Notch intracellular domain translocates to the nucleus, it binds to Rbpjk
(a transcriptional repressor), converting it into an activator, and inducing the expres-
sion of downstream target genes. Mice lacking Rbpjk in Nkx2.1-expressing cells
show an expansion of the ARC and an increased number of POMC-neurons (Aujla
et al. 2013). Mash1 expression is also increased in these animals, and this could
account for the aberrant POMC expression, as Mash1 is an important factor in the
differentiation of POMC-neurons (McNay et al. 2006). The number of
NPY-expressing AgRP-neurons is also increased (Aujla et al. 2013). Mice with a
constitutively active Notch1 intracellular domain show a reduction in ARC area and
a complete absence of POMC- and AgRP-neurons (Aujla et al. 2013), mirroring the
phenotypes of Mash1-KO mice (McNay et al. 2006).
4 Gene Regulatory Programs in the Development of Hypothalamic Arcuate. . . 91

4.2.13 Ikaros

Ikaros is a zinc finger TF that is critical for lymphohematopoiesis and immune

regulation (Hu et al. 2017). Ikaros is also expressed in GHRH-neurons in the
mouse embryonic ARC. Overexpression of Ikaros enhances GHRH promoter activ-
ity and induces endogenous GHRH gene expression in N3 hypothalamic cells (Ezzat
et al. 2006). Notably, Ikaros-null mice exhibit dwarfism with characteristics of
pituitary GH deficiency, and GHRH immunoreactivity was undetectable in the
hypothalamus, consistent with Ikaros being essential for GHRH expression.

4.3 Additional TFs in Arcuate Neuronal Development

From the public databases such as GenePaint (Visel et al. 2004) and Allen Brain
Atlas (Jones et al. 2009; Lein et al. 2007) as well as the recent expression studies of
Shimogori et al. (2010) (Box 4.2), it is clear that more TFs are expressed in the
developing ARC, which include the TFs Pou2f2, Sox14, Tbx2, Tbx19, Arx, Pbx1,
Insm1, and Lhx5. Future investigation of the role of these TFs in arcuate neuronal
development will allow us to establish which TFs, alone or together, play crucial
roles in regulating the production of the many arcuate neuronal types during
embryogenesis.

4.4 Coordinate Development of Different Arcuate Neurons

It is important to note that what appear to be distinct classes of arcuate neurons are
indeed, functionally interconnected among themselves. For instance, reproduction, a
highly energy-demanding process, is coupled to energy balance, and leptin serves as
a key mediator of this coupling by controlling both food intake via AgRP-/POMC-
neurons and reproduction via Kiss1-neurons (Chehab 2014). In addition, AgRP-
neurons project onto GnRH and Kiss1-neurons while POMC-neurons project onto
Kiss1-neurons, forming a circuit that allows for cross talk between food intake and
reproduction (Chehab 2014). Another example is the well-known correlation
between nutritional status and linear growth. As such, GH/GHRH levels are
inversely correlated with circulating levels of the anorexic hormone leptin, which
in turn targets arcuate neurons that control energy balance (Considine et al. 1996;
Dieguez and Casanueva 1995; Grinspoon et al. 1996). Clearly, the generation of all
of the different ARC neuronal subtypes is critical for homeostasis. Currently,
evidence suggests that production of multiple arcuate neurons during embryogenesis
involves common progenitors. These progenitors then differentiate into multiple
types of neurons via gene regulatory programs that involve distinct fate-specifying
TFs. How these programs are initiated/specified is unclear, but these programs allow
concise regulation of each developmental lineage. In this context, it is interesting to
note that POMC+ precursors generate several arcuate neuronal types, including
POMC-, AgRP-, and Kiss1-neurons (Padilla et al. 2010; Sanz et al. 2015). It is
92 J. W. Lee et al.

also possible that the production of GHRH-neurons is coordinated with the genera-
tion of AgRP- and/or POMC-neurons via common progenitors.

4.5 Challenges to Define Gene Regulatory Programs

and the Genome-Wide Solutions

In addition to identifying and characterizing the TFs that drive the development of
arcuate neuronal subtypes, several challenges remain to be addressed to fully
understand the gene regulatory programs that direct this developmental process. In
particular, three issues are noteworthy.
First, for most TFs that are known to control the development of arcuate neurons
(Table 4.1), information on their critical downstream target genes is largely missing.
Determining the binding sites of these TFs in the different neuronal cell types of the
ARC will help unravel the detailed molecular basis of their action in fate specifica-
tion and development of arcuate neurons. This information should also facilitate our
understanding of how different arcuate neurons coordinately develop during
embryogenesis.
Second, precise assessment of the whole spectrum of actions that each TF plays in
fate specification and development of the different arcuate neuronal types is seri-
ously hampered by the fact that we currently have only a limited number of early fate
markers for each arcuate neuronal population. Thus, we are lacking tools to identify
lineage specification pathways.
Third, directly following is that Cre lines that can delete genes of interest in a
specific population of arcuate neurons during earlier fate specification stages of
development are missing. Such Cre lines will allow us to delineate the developmen-
tal action of each TF in specific populations of arcuate neurons independent of its
action in other cell types. Potential resolutions to these issues, using recent advances
in genome-wide studies, are discussed below.

4.5.1 Connecting Individual TFs in the Gene Regulatory Networks

That Coordinately Integrate Development of Different
Arcuate Neurons

The first step to reconstruct the gene regulatory programs for arcuate neuronal
development is to identify TFs that play critical roles in arcuate neuronal develop-
ment (Table 4.1). The next step is to identify the genome-wide target genes of each
of these TFs in the developing ARC. This will also facilitate our effort to decipher
how these TFs function, alone and/or in combination, to coordinate the production of
different populations of arcuate neurons. A strategy used in our recent study provides
a template for such work. We found that the HD TFs Otp and Dlx1 (and its homolog
Dlx2) play essential roles in the development of AgRP- and GHRH-neurons,
respectively (Lee et al. 2018). In addition to a loss of GHRH-neurons, we found
that mice with Nkx2.1-Cre-directed cKO for the adjoining Dlx1 and Dlx2 genes
4 Gene Regulatory Programs in the Development of Hypothalamic Arcuate. . . 93

Fig. 4.3 The Dl1/2-Otp regulatory axis in the development of GHRH- and AgRP-neurons. (a)
Summary of deficits in the development of growth hormone releasing hormone (GHRH)- and
agouti-related peptide (AgRP)-neurons in Otp/ and Dlx1/2-conditional knockout (cKO) mice.
While Otp/ mice lose AgRP-neurons, Dlx1/2-cKO mice lose GHRH-neurons but gain AgRP-
neurons. (b) CHIP-seq schematic, including cross-linking of tissues, preparation of lysed chromatin
and sonification, immunoprecipitation of chromatin fragments bound by transcription factor
(TF) using antibody against the TF, and reverse cross-linking and isolation of DNA, followed by
library constructiuon, sequencing and analyses. (c) Our CHIP-seq experiments using anti-Dlx1
antibody as well as the ensuing functional assays revealed that Dlx1 suppresses the expression of
Otp by directly binding the silencer region located downstream of the Otp gene in developing
GHRH-neurons (Lee et al. 2018). Direct target genes of Otp in developing AgRP-neurons and of
Dlx1/2 in developing GHRH-neurons remain to be investigated.

show a significant increase in AgRP-neurons (Fig. 4.3a). Consistent with these

results, these mice exhibit not only dwarfism, likely reflecting the loss of GHRH-
neurons, but also many postnatal phenotypes that appear to reflect the increased
number of AgRP-neurons, including reduced levels of energy expenditure and
brown fat thermogenesis (Lee et al. 2018). Our subsequent genome-wide
experiments revealed a surprising link between Otp and Dlx1/2.
First, we performed chromatin immunoprecipitation (CHIP) of E16.5 hypothala-
mus using an anti-Dlx1 antibody followed by sequencing, a technology named
CHIP-seq that enables the genome-wide mapping of binding sites for TFs with
extremely high resolution (Barski and Zhao 2009) (Fig. 4.3b). Notably, a functional
Dlx1-binding site in the downstream region of the Otp gene was identified. Occupa-
tion of this site by Dlx1 resulted in suppression of Otp expression (Fig. 4.3c). We
genetically showed that upregulation of Otp increased the number of AgRP-neurons
in Dlx1/2-cKO mice (Lee et al. 2018). Given the finding that Dlx1/2 is expressed in
developing GHRH-neurons, we concluded that GHRH- and AgRP-neurons share
94 J. W. Lee et al.

Fig. 4.4 De novo motif analysis and partner transcription factors (TFs). (a) Unbiased search of
CHIP-seq peaks for motifs that recruit the target TF reveals multiple modes of DNA binding by the
TF. TFs often bind their target sites indirectly through other TFs (e.g., X in this diagram) or in
synergy with another TF (e.g., Y in this diagram). (b) The homeodomain (HD) TF Isl1 is known to
pair with other TFs in a cell type-specific manner during development, and this combinatorial mode
of actions allows one TF to control multiple different fates

common progenitors and that developing GHRH-neurons are still plastic with the
potential to differentiate to AgRP-neurons. Thus, suppression of Otp expression by
Dlx1/2 forces the progenitors to take the lineage to GHRH-neurons rather than the
erroneous pathway to AgRP-neurons. These results demonstrate that the Dlx1/2-Otp
gene regulatory axis directs a balanced production of GHRH- and AgRP-neurons
during embryogenesis, contributing to the postnatal homeostatic functions of hypo-
thalamic arcuate neurons that control energy balance and growth.
In addition to directly binding their cognate binding motifs, TFs often bind DNA
either indirectly through protein–protein interactions with other DNA-bound TFs
(e.g., X in Fig. 4.3a) or synergistically through adjacent motifs for other TFs (e.g., Y
in Fig. 4.4a). Once the genome-wide occupancy of a TF is determined using the
CHIP-seq technology, de novo motif analysis of CHIP-seq peaks (Bailey et al. 2009;
Machanick and Bailey 2011) will allow us to predict its different modes of DNA
binding as well as partner TFs. Hence, de novo motif analyses, coupled with the
ensuing validation experiments, also provide crucial insights into the identity of
partner TFs that function together with the original TF (e.g., X and Y in Fig. 4.4a),
contributing to our understanding of how distinct fate-determining TFs are
integrated into the gene regulatory programs for arcuate neuronal development. It
is possible that some of the TF in Table 4.1 may be engaged in such protein–protein
interactions, resulting in an additional layer of cross talk among key TFs. Such cross
talk may direct the development of neuronal heterogeneity now known to exist in the
ARC. In this regard, it is interesting to note that Isl1 directs fate determination of
“specific” cell types by forming cell type-specific complexes with other TFs (Cho
et al. 2014; Habener et al. 2005; Lee and Pfaff 2001; Mazzoni et al. 2013; Peng et al.
2005) (Fig. 4.4b). As Isl1 is required for the generation of several distinct types of
arcuate neurons (Lee et al. 2016), Isl1 may partner with other TFs in specifying cell-
fate of each arcuate neuronal subtype.
4 Gene Regulatory Programs in the Development of Hypothalamic Arcuate. . . 95

4.5.2 The Need for Early Fate Markers of Arcuate Neurons

The lack of early fate markers has prohibited us from deciphering the full spectrum
of activities that each TF exerts in the production of arcuate neuronal populations.
Most fate markers that are currently being used in the field, such as GHRH for
GHRH-neurons and AgRP for AgRP-neurons, represent late fate markers that
identify already committed or differentiated neurons. In addition, markers currently
used as early fate markers, label multiple neuronal types. For example, POMC marks
early ARC cells but lacks specificity given POMC-progenitors produce POMC-
neurons, AgRP-neurons as well as Kiss1-neurons (Padilla et al. 2010; Sanz et al.
2015). Similarly, we recently found that NPY+ progenitors likely give rise to AgRP-
and GHRH-neurons (Lee et al. 2018). The multiple lineages from POMC+ and
NPY+ precursors highlight the need for additional early fate markers, which will
enable us to understand the exact roles of each TF in specifying different arcuate
neuronal cell types. Future investigation into the ontogeny of arcuate neurons, such
as the relationship among POMC+ and NPY+ precursors and their progenies,
coupled with our understanding of the gene regulatory programs for arcuate neuronal
development, will reveal how different types of arcuate neurons are produced in the
right ratio during development to perform highly interconnected functions at post-
natal stages.
The recent advances in single cell RNA-seq (scRNA-seq) technology (Fig. 4.5)
will greatly facilitate the efforts to identify additional early fate markers of develop-
ing arcuate neurons, as this technology provides the transcriptome of individual
developing arcuate neurons. Using this technology in the hypothalamus at various
embryonic stages, we can identify additional early fate markers for each cell type. In
addition, developmental trajectories of each arcuate neuronal population and when
common progenitors begin to diverge to different arcuate neuronal lineages during
embryogenesis may be unraveled. Single cell RNA-seq can also be performed for
control and KO/cKO models for each TF, and their comparative analyses will

Fig. 4.5 Schematics for scRNA-seq. Dissociated cells are highly diluted so that only a single cell
mixed with reagents is encapsulated into nanoliter-sized GEMs (gel bead in emulsion) together with
a bead containing a barcoded oligonucleotide. GEMs are subjected to reverse transcription, and the
resulting cDNA is isolated and constructed into a library, followed by sequencing and analyses
96 J. W. Lee et al.

identify target genes of each TF. Integration of these results with the above described
CHIP-seq results will identify target genes that are directly controlled by the TF,
ultimately allowing us to fully reconstruct the gene regulatory programs that direct
the development of individual arcuate neurons.

4.5.3 The Need for More Specific Cre Lines

Cre lines that can delete genes of interests only in a specific population of developing
arcuate neurons will allow us to investigate the developmental action of each TF in a
specific arcuate neuronal type independent of its action in other arcuate and
non-arcuate neuronal types. Notably, while Nkx2.1-Cre transgenic line has been
widely used for studying roles of genes of interests in the developing ARC, it deletes
genes in most neuronal types in the developing ARC (Xu et al. 2008). Moreover,
Nkx2.1 is also expressed in the thyroid, lung, and other parts of the brain (Lazzaro
et al. 1991). This specificity issue also extends to other Cre lines that delete genes of
interests in developing arcuate neurons, such as Npy-Cre (Milstein et al. 2015) and
Pomc-Cre (Balthasar et al. 2004). The promoters of early fate markers to be identified
from the aforementioned single cell RNA-seq approaches may allow us to create
more specific Cre lines. However, many of these early markers may still be expressed
in more than a single cell type. To help resolve this issue, an additional layer of
specificity can be achieved using the split-Cre system, in which co-expression of
catalytically inactive N-terminal and C-terminal fragments of Cre reactivates Cre
activity (Casanova et al. 2003; Hirrlinger et al. 2009) (Fig. 4.6). In this system, for
instance, two different promoters from early fate marker genes that are expressed in
arcuate cell types A and B (promoter 1) and B, C, and D (promoter 2), respectively,
will drive the expression of reconstituted active Cre in the overlapping cell type B
(Fig. 4.5).
Demonstrating a postnatal homeostatic phenotype that correlates with a develop-
mental deficit in an arcuate neuronal subtype is an excellent way to validate cell
specification. Notably, different arcuate neurons are connected to each other as well

Fig. 4.6 Schematics for the split-Cre system. To obtain higher specificity, expression of the
catalytically inactive N-terminal and C-terminal fragments of Cre is driven by two less-specific
promoters, respectively. Only in cell types in which both promoters are active, N-terminal and
C-terminal fragments of Cre are coexpressed restoring Cre activity
4 Gene Regulatory Programs in the Development of Hypothalamic Arcuate. . . 97

as with neurons in other brain regions, forming complex neural circuitries that
mediate homeostatic processes. Therefore, less-specific Cre lines, such as the
Nkx2.1-Cre transgenic line (Xu et al. 2008), pose a challenge for deciphering such
complex circuitry. For instance, mice with conditional deletions of Isl1 or Dlx1/2 in
developing arcuate neurons display developmental deficits in arcuate neurons that
control energy homeostasis and linear growth. Postnatally, these mice show dwarf-
ism as well as reduced energy expenditure and late-onset obesity (Lee et al. 2016,
2018). However, the specificity issues of Nkx2.1-Cre line make it difficult to
conclude that the postnatal phenotypes observed with Isl1- or Dlx1/2-cKO mice
are mainly due to the developmental deficits in arcuate neurons (Lee et al. 2016,
2018). For instance, Isl1 may also play a role in the development of the thyroid gland
(Westerlund et al. 2008). Thus, potential hypothyroidism resulting from the lack of
Isl1 may be the major culprit or have also contributed to the metabolic phenotypes
observed in the Isl1-cKO mice (Lee et al. 2016). These experiments highlight the
necessity for the development of more specific Cre lines that will enable us to
successfully link developmental deficits to adult ARC-related phenotypes.

4.6 Future Perspectives

The technological breakthroughs, such as CHIP-seq and scRNA-seq, combined with

the publicly available databases including GenePaint (Visel et al. 2004) and Allen
Brain Atlas (Jones et al. 2009; Lein et al. 2007) as well as the pioneering expression
studies of Dr. Blackshaw’s group (Shimogori et al. 2010) provide an unprecedented
opportunity to fully reconstruct the gene regulatory programs that direct the balanced
development of diverse neuronal types and subtypes in the developing ARC. Future
studies are needed to identify transcriptional coregulators that enable or disable the
action of TFs in the development of arcuate neurons (Mannervik et al. 1999). In
particular, roles of epigenetic chromatin regulators in arcuate neuronal development
is an interesting subject, given their broad effect on brain development and autism
spectrum disorders (De Rubeis et al. 2014; Iossifov et al. 2014; Michaelson et al.
2012; O’Roak et al. 2012). Moreover, some patients with inactivated chromatin
regulators show phenotypes that may involve deficit in arcuate neuronal develop-
ment, such as dwarfism (Lintas and Persico 2018; Niikawa et al. 1981). With the
arrival of more specific Cre lines, we may be able to connect prominent metabolic
features of those patients to deficits in the development of specific types of arcuate
neurons. For the next few years, we will begin to witness the emergence of fully
decoded gene regulatory programs that direct the development of arcuate neurons.

Acknowledgments This work was supported by grants from NIH/NIDDK (R01 DK064678 and
R01 DK103661 to J.W.L.), the Korea Health Technology R&D Project through the Korea Health
Industry Development Institute, funded by the Ministry of Health & Welfare, Republic of Korea
(HI17C0447 to S.L.), and National Research Foundation of Korea (NRF-2012M3A9D1054705 to
S.L.).
98 J. W. Lee et al.

Key References

Campbell et al. (2017)—This article reports different types of arcuate nucleus

neurons in the adult for the first time using scRNA-seq approach.
Lee et al. (2016)—This article reports crucial roles of the TF Isl1 in the development
of multiple arcuate nucleus neuronal types.
Lee et al. (2018)—This article shows how different TFs involved in the production
of GHRH- and AgRP-neurons are integrated into a regulatory axis that prevents
developing GHRH-neurons from erroneously taking AgRP-neuronal fate during
embryogenesis.
Nasif et al. (2015)—This article describes the developmental role of the TF Isl1 in
specifying POMC-neurons.
Shimogori et al. (2010)—Using microarray analysis at different developmental time
points, and developmental in situ hybridization, for genes expressed over the
course of hypothalamic neurogenesis, this article identifies markers that stably
label each major hypothalamic nucleus over the entire course of neurogenesis and
constructs a molecular atlas of the developing hypothalamus.

Box 4.1 Identification of the Adult Arcuate Nucleus Neuronal Types

This box contains interesting information on the discovery of hypothalamic
arcuate nucleus neurons in the adult. Despite extensive studies on the physio-
logical roles of many arcuate nucleus neurons, their exact composition has not
been well-defined. Using the recent advances in scRNA-seq technology, the
group in Boston cataloged arcuate nucleus neuronal types in an unbiased way
for the first time [7]. This paradigm-shifting study identified many transcrip-
tionally distinct cell populations, including 24 arcuate nucleus neurons. One
unexpected finding from this study was the presence of multiple subtypes for
some neurons, such as three subtypes of POMC-neurons, two subtypes of
AgRP-neurons and six subtypes of TH-neurons (GHRH- and five other
TH-subtypes). This study also serves as an invaluable resource for genes
specifically expressed in each adult arcuate neuronal type.

Box 4.2 A Molecular Atlas of the Developing Hypothalamus

Understanding of the developmental processes for various hypothalamus
neurons has been seriously hampered by the lack of information on genes
specifically expressed in each developing cell type, including upstream sig-
naling molecules and downstream gene regulators such as transcription factors
and coregulators. Dr. Seth Blackshaw’s group at Johns Hopkins University
performed microarray analysis at 12 different developmental time points, and
then in situ hybridization for 1045 genes that were dynamically expressed
during hypothalamic neurogenesis [66]. This paradigm-shifting study

(continued)
4 Gene Regulatory Programs in the Development of Hypothalamic Arcuate. . . 99

Box 4.2 (continued)

identified markers that stably labeled each major hypothalamic nucleus during
the course of neurogenesis, which will serve as a fundamental platform to
study the molecular pathways that mediate hypothalamic development.

References
Acampora D, Postiglione MP, Avantaggiato V, Di Bonito M, Vaccarino FM, Michaud J, Simeone
A (1999) Progressive impairment of developing neuroendocrine cell lineages in the hypothala-
mus of mice lacking the Orthopedia gene. Genes Dev 13:2787–2800
Altman J, Bayer SA (1978) Development of the diencephalon in the rat. II. Correlation of the
embryonic development of the hypothalamus with the time of origin of its neurons. J Comp
Neurol 182:973–993
Andrews GL, Yun K, Rubenstein JL, Mastick GS (2003) Dlx transcription factors regulate
differentiation of dopaminergic neurons of the ventral thalamus. Mol Cell Neurosci 23:107–120
Anthwal N, Pelling M, Claxton S, Mellitzer G, Collin C, Kessaris N, Richardson WD, Gradwohl G,
Ang SL (2013) Conditional deletion of neurogenin-3 using Nkx2.1iCre results in a mouse
model for the central control of feeding, activity and obesity. Dis Model Mech 6:1133–1145
Aujla PK, Naratadam GT, Xu L, Raetzman LT (2013) Notch/Rbpjkappa signaling regulates
progenitor maintenance and differentiation of hypothalamic arcuate neurons. Development
140:3511–3521
Bailey TL, Boden M, Buske FA, Frith M, Grant CE, Clementi L, Ren J, Li WW, Noble WS (2009)
MEME SUITE: tools for motif discovery and searching. Nucleic Acids Res 37:W202–W208
Balthasar N, Coppari R, McMinn J, Liu SM, Lee CE, Tang V, Kenny CD, McGovern RA, Chua SC
Jr, Elmquist JK et al (2004) Leptin receptor signaling in POMC neurons is required for normal
body weight homeostasis. Neuron 42:983–991
Barski A, Zhao K (2009) Genomic location analysis by ChIP-Seq. J Cell Biochem 107:11–18
Begley CG, Lipkowitz S, Gobel V, Mahon KA, Bertness V, Green AR, Gough NM, Kirsch IR
(1992) Molecular characterization of NSCL, a gene encoding a helix-loop-helix protein
expressed in the developing nervous system. Proc Natl Acad Sci USA 89:38–42
Biebermann H, Kuhnen P, Kleinau G, Krude H (2012) The neuroendocrine circuitry controlled by
POMC, MSH, and AGRP. Handb Exp Pharmacol 209:47–75
Bluet-Pajot MT, Tolle V, Zizzari P, Robert C, Hammond C, Mitchell V, Beauvillain JC, Viollet C,
Epelbaum J, Kordon C (2001) Growth hormone secretagogues and hypothalamic networks.
Endocrine 14:1–8
Byerly MS, Blackshaw S (2009) Vertebrate retina and hypothalamus development. Wiley
Interdiscip Rev Syst Biol Med 1:380–389
Campbell JN, Macosko EZ, Fenselau H, Pers TH, Lyubetskaya A, Tenen D, Goldman M,
Verstegen AM, Resch JM, McCarroll SA et al (2017) A molecular census of arcuate hypothala-
mus and median eminence cell types. Nat Neurosci 20:484–496
Casanova E, Lemberger T, Fehsenfeld S, Mantamadiotis T, Schutz G (2003) Alpha complementa-
tion in the Cre recombinase enzyme. Genesis 37:25–29
Chehab FF (2014) 20 years of leptin: leptin and reproduction: past milestones, present undertakings,
and future endeavors. J Endocrinol 223:T37–T48
Cho HH, Cargnin F, Kim Y, Lee B, Kwon RJ, Nam H, Shen R, Barnes AP, Lee JW, Lee S et al
(2014) Isl1 directly controls a cholinergic neuronal identity in the developing forebrain and
spinal cord by forming cell type-specific complexes. PLoS Genet 10:e1004280
100 J. W. Lee et al.

Cobos I, Borello U, Rubenstein JL (2007) Dlx transcription factors promote migration through
repression of axon and dendrite growth. Neuron 54:873–888
Cohen P, Rosenfeld RG (1994) Physiologic and clinical relevance of the insulin-like growth factor
binding proteins. Curr Opin Pediatr 6:462–467
Considine RV, Sinha MK, Heiman ML, Kriauciunas A, Stephens TW, Nyce MR, Ohannesian JP,
Marco CC, McKee LJ, Bauer TL et al (1996) Serum immunoreactive-leptin concentrations in
normal-weight and obese humans. N Engl J Med 334:292–295
Cremona M, Colombo E, Andreazzoli M, Cossu G, Broccoli V (2004) Bsx, an evolutionary
conserved brain specific homeoboX gene expressed in the septum, epiphysis, mammillary
bodies and arcuate nucleus. Gene Expr Patterns 4:47–51
De Rubeis S, He X, Goldberg AP, Poultney CS, Samocha K, Cicek AE, Kou Y, Liu L, Fromer M,
Walker S et al (2014) Synaptic, transcriptional and chromatin genes disrupted in autism. Nature
515:209–215
Dieguez C, Casanueva FF (1995) Influence of metabolic substrates and obesity on growth hormone
secretion. Trends Endocrinol Metab 6:55–59
Eisenstat DD, Liu JK, Mione M, Zhong W, Yu G, Anderson SA, Ghattas I, Puelles L, Rubenstein
JL (1999) DLX-1, DLX-2, and DLX-5 expression define distinct stages of basal forebrain
differentiation. J Comp Neurol 414:217–237
Ezzat S, Mader R, Fischer S, Yu S, Ackerley C, Asa SL (2006) An essential role for the
hematopoietic transcription factor Ikaros in hypothalamic-pituitary-mediated somatic growth.
Proc Natl Acad Sci USA 103:2214–2219
Gobel V, Lipkowitz S, Kozak CA, Kirsch IR (1992) NSCL-2: a basic domain helix-loop-helix gene
expressed in early neurogenesis. Cell Growth Differ 3:143–148
Good DJ, Porter FD, Mahon KA, Parlow AF, Westphal H, Kirsch IR (1997) Hypogonadism and
obesity in mice with a targeted deletion of the Nhlh2 gene. Nat Genet 15:397–401
Grinspoon S, Gulick T, Askari H, Landt M, Lee K, Anderson E, Ma Z, Vignati L, Bowsher R,
Herzog D et al (1996) Serum leptin levels in women with anorexia nervosa. J Clin Endocrinol
Metab 81:3861–3863
Habener JF, Kemp DM, Thomas MK (2005) Minireview: transcriptional regulation in pancreatic
development. Endocrinology 146:1025–1034
Hill JW, Elmquist JK, Elias CF (2008) Hypothalamic pathways linking energy balance and
reproduction. Am J Phys Endocrinol Metab 294:E827–E832
Hirrlinger J, Scheller A, Hirrlinger PG, Kellert B, Tang W, Wehr MC, Goebbels S, Reichenbach A,
Sprengel R, Rossner MJ et al (2009) Split-cre complementation indicates coincident activity of
different genes in vivo. PLoS One 4:e4286
Hrabovszky E (2014) Neuroanatomy of the human hypothalamic kisspeptin system. Neuroendocri-
nology 99:33–48
Hu Y, Yoshida T, Georgopoulos K (2017) Transcriptional circuits in B cell transformation. Curr
Opin Hematol 24:345–352
Ifft JD (1972) An autoradiographic study of the time of final division of neurons in rat hypothalamic
nuclei. J Comp Neurol 144:193–204
Iossifov I, O’Roak BJ, Sanders SJ, Ronemus M, Krumm N, Levy D, Stessman HA, Witherspoon
KT, Vives L, Patterson KE et al (2014) The contribution of de novo coding mutations to autism
spectrum disorder. Nature 515:216–221
Ishii Y, Bouret SG (2012) Embryonic birthdate of hypothalamic leptin-activated neurons in mice.
Endocrinology 153:3657–3667
Jing E, Nillni EA, Sanchez VC, Stuart RC, Good DJ (2004) Deletion of the Nhlh2 transcription
factor decreases the levels of the anorexigenic peptides alpha melanocyte-stimulating hormone
and thyrotropin-releasing hormone and implicates prohormone convertases I and II in obesity.
Endocrinology 145:1503–1513
Jones AR, Overly CC, Sunkin SM (2009) The Allen Brain Atlas: 5 years and beyond. Nat Rev
Neurosci 10:821–828
4 Gene Regulatory Programs in the Development of Hypothalamic Arcuate. . . 101

Kimura S, Hara Y, Pineau T, Fernandez-Salguero P, Fox CH, Ward JM, Gonzalez FJ (1996) The
T/ebp null mouse: thyroid-specific enhancer-binding protein is essential for the organogenesis
of the thyroid, lung, ventral forebrain, and pituitary. Genes Dev 10:60–69
Kruger M, Ruschke K, Braun T (2004) NSCL-1 and NSCL-2 synergistically determine the fate of
GnRH-1 neurons and control necdin gene expression. EMBO J 23:4353–4364
Kuperman Y, Weiss M, Dine J, Staikin K, Golani O, Ramot A, Nahum T, Kuhne C, Shemesh Y,
Wurst W et al (2016) CRFR1 in AgRP neurons modulates sympathetic nervous system activity
to adapt to cold stress and fasting. Cell Metab 23:1185–1199
Lam DD, de Souza FS, Nasif S, Yamashita M, Lopez-Leal R, Otero-Corchon V, Meece K,
Sampath H, Mercer AJ, Wardlaw SL et al (2015) Partially redundant enhancers cooperatively
maintain mammalian pomc expression above a critical functional threshold. PLoS Genet 11:
e1004935
Lasky JL, Wu H (2005) Notch signaling, brain development, and human disease. Pediatr Res
57:104r–109r
Lazzaro D, Price M, de Felice M, Di Lauro R (1991) The transcription factor TTF-1 is expressed at
the onset of thyroid and lung morphogenesis and in restricted regions of the foetal brain.
Development 113:1093–1104
Lee SK, Pfaff SL (2001) Transcriptional networks regulating neuronal identity in the developing
spinal cord. Nat Neurosci 4(Suppl):1183–1191
Lee B, Lee S, Lee SK, Lee JW (2016) The LIM-homeobox transcription factor Isl1 plays crucial
roles in the development of multiple arcuate nucleus neurons. Development 143:3763–3773
Lee B, Kim J, An T, Kim S, Patel EM, Raber J, Lee S-K, Lee S, Lee JW (2018) Dlx1/2 and Otp
coordinate the production of hypothalamic GHRH- and AgRP-neurons. Nat Commun 9:2026
Lein ES, Hawrylycz MJ, Ao N, Ayres M, Bensinger A, Bernard A, Boe AF, Boguski MS,
Brockway KS, Byrnes EJ et al (2007) Genome-wide atlas of gene expression in the adult
mouse brain. Nature 445:168–176
Li H, Zeitler PS, Valerius MT, Small K, Potter SS (1996) Gsh-1, an orphan Hox gene, is required
for normal pituitary development. EMBO J 15:714–724
Lintas C, Persico AM (2018) Unraveling molecular pathways shared by Kabuki and Kabuki-like
syndromes. Clin Genet 94(3–4):283-295
Long JE, Garel S, Alvarez-Dolado M, Yoshikawa K, Osumi N, Alvarez-Buylla A, Rubenstein JL
(2007) Dlx-dependent and -independent regulation of olfactory bulb interneuron differentiation.
J Neurosci Off J Soc Neurosci 27:3230–3243
Lu F, Kar D, Gruenig N, Zhang ZW, Cousins N, Rodgers HM, Swindell EC, Jamrich M,
Schuurmans C, Mathers PH et al (2013) Rax is a selector gene for mediobasal hypothalamic
cell types. J Neurosci Off J Soc Neurosci 33:259–272
Machanick P, Bailey TL (2011) MEME-ChIP: motif analysis of large DNA datasets. Bioinformat-
ics 27:1696–1697
Mannervik M, Nibu Y, Zhang H, Levine M (1999) Transcriptional coregulators in development.
Science 284:606–609
Marin O, Anderson SA, Rubenstein JL (2000) Origin and molecular specification of striatal
interneurons. J Neurosci Off J Soc Neurosci 20:6063–6076
Marin O, Baker J, Puelles L, Rubenstein JL (2002) Patterning of the basal telencephalon and
hypothalamus is essential for guidance of cortical projections. Development (Cambridge,
England) 129:761–773
Martinez-Ferre A, Martinez S (2012) Molecular regionalization of the diencephalon. Front
Neurosci 6:73
Mazzoni EO, Mahony S, Closser M, Morrison CA, Nedelec S, Williams DJ, An D, Gifford DK,
Wichterle H (2013) Synergistic binding of transcription factors to cell-specific enhancers
programs motor neuron identity. Nat Neurosci 16:1219–1227
McNay DE, Pelling M, Claxton S, Guillemot F, Ang SL (2006) Mash1 is required for generic and
subtype differentiation of hypothalamic neuroendocrine cells. Mol Endocrinol 20:1623–1632
102 J. W. Lee et al.

Michaelson JJ, Shi Y, Gujral M, Zheng H, Malhotra D, Jin X, Jian M, Liu G, Greer D, Bhandari A
et al (2012) Whole-genome sequencing in autism identifies hot spots for de novo germline
mutation. Cell 151:1431–1442
Milstein AD, Bloss EB, Apostolides PF, Vaidya SP, Dilly GA, Zemelman BV, Magee JC (2015)
Inhibitory gating of input comparison in the CA1 microcircuit. Neuron 87:1274–1289
Muranishi Y, Terada K, Furukawa T (2012) An essential role for Rax in retina and neuroendocrine
system development. Develop Growth Differ 54:341–348
Nasif S, de Souza FS, Gonzalez LE, Yamashita M, Orquera DP, Low MJ, Rubinstein M (2015) Islet
1 specifies the identity of hypothalamic melanocortin neurons and is critical for normal food
intake and adiposity in adulthood. Proc Natl Acad Sci USA 112:E1861–E1870
Niikawa N, Matsuura N, Fukushima Y, Ohsawa T, Kajii T (1981) Kabuki make-up syndrome: a
syndrome of mental retardation, unusual facies, large and protruding ears, and postnatal growth
deficiency. J Pediatr 99:565–569
O’Roak BJ, Vives L, Girirajan S, Karakoc E, Krumm N, Coe BP, Levy R, Ko A, Lee C, Smith JD
et al (2012) Sporadic autism exomes reveal a highly interconnected protein network of de novo
mutations. Nature 485:246–250
Padilla SL, Carmody JS, Zeltser LM (2010) Pomc-expressing progenitors give rise to antagonistic
neuronal populations in hypothalamic feeding circuits. Nat Med 16:403–405
Pan Q, Li C, Xiao J, Kimura S, Rubenstein J, Puelles L, Minoo P (2004) In vivo characterization of
the Nkx2.1 promoter/enhancer elements in transgenic mice. Gene 331:73–82
Pelling M, Anthwal N, McNay D, Gradwohl G, Leiter AB, Guillemot F, Ang SL (2011) Differential
requirements for neurogenin 3 in the development of POMC and NPY neurons in the hypothal-
amus. Dev Biol 349:406–416
Peng SY, Wang WP, Meng J, Li T, Zhang H, Li YM, Chen P, Ma KT, Zhou CY (2005) ISL1
physically interacts with BETA2 to promote insulin gene transcriptional synergy in non-beta
cells. Biochim Biophys Acta 1731:154–159
Petryniak MA, Potter GB, Rowitch DH, Rubenstein JL (2007) Dlx1 and Dlx2 control neuronal
versus oligodendroglial cell fate acquisition in the developing forebrain. Neuron 55:417–433
Price M, Lazzaro D, Pohl T, Mattei MG, Ruther U, Olivo JC, Duboule D, Di Lauro R (1992)
Regional expression of the homeobox gene Nkx-2.2 in the developing mammalian forebrain.
Neuron 8:241–255
Qiu M, Bulfone A, Martinez S, Meneses JJ, Shimamura K, Pedersen RA, Rubenstein JL (1995)
Null mutation of Dlx-2 results in abnormal morphogenesis of proximal first and second
branchial arch derivatives and abnormal differentiation in the forebrain. Genes Dev
9:2523–2538
Rowitch DH, Kriegstein AR (2010) Developmental genetics of vertebrate glial-cell specification.
Nature 468:214–222
Saino-Saito S, Berlin R, Baker H (2003) Dlx-1 and Dlx-2 expression in the adult mouse brain:
relationship to dopaminergic phenotypic regulation. J Comp Neurol 461:18–30
Sakkou M, Wiedmer P, Anlag K, Hamm A, Seuntjens E, Ettwiller L, Tschop MH, Treier M (2007)
A role for brain-specific homeobox factor Bsx in the control of hyperphagia and locomotory
behavior. Cell Metab 5:450–463
Sanz E, Quintana A, Deem JD, Steiner RA, Palmiter RD, McKnight GS (2015) Fertility-regulating
Kiss1 neurons arise from hypothalamic POMC-expressing progenitors. J Neurosci Off J Soc
Neurosci 35:5549–5556
Schmid T, Gunther S, Mendler L, Braun T (2013) Loss of NSCL-2 in gonadotropin releasing
hormone neurons leads to reduction of pro-opiomelanocortin neurons in specific hypothalamic
nuclei and causes visceral obesity. J Neurosci Off J Soc Neurosci 33:10459–10470
Shimogori T, Lee DA, Miranda-Angulo A, Yang Y, Wang H, Jiang L, Yoshida AC, Kataoka A,
Mashiko H, Avetisyan M et al (2010) A genomic atlas of mouse hypothalamic development.
Nat Neurosci 13:767–775
4 Gene Regulatory Programs in the Development of Hypothalamic Arcuate. . . 103

Simeone A, D’Apice MR, Nigro V, Casanova J, Graziani F, Acampora D, Avantaggiato V (1994)

Orthopedia, a novel homeobox-containing gene expressed in the developing CNS of both
mouse and Drosophila. Neuron 13:83–101
Sokolowski K, Esumi S, Hirata T, Kamal Y, Tran T, Lam A, Oboti L, Brighthaupt SC, Zaghlula M,
Martinez J et al (2015) Specification of select hypothalamic circuits and innate behaviors by the
embryonic patterning gene dbx1. Neuron 86:403–416
Sokolowski K, Tran T, Esumi S, Kamal Y, Oboti L, Lischinsky J, Goodrich M, Lam A, Carter M,
Nakagawa Y et al (2016) Molecular and behavioral profiling of Dbx1-derived neurons in the
arcuate, lateral and ventromedial hypothalamic nuclei. Neural Dev 11:12
Sussel L, Marin O, Kimura S, Rubenstein JL (1999) Loss of Nkx2.1 homeobox gene function
results in a ventral to dorsal molecular respecification within the basal telencephalon: evidence
for a transformation of the pallidum into the striatum. Development 126:3359–3370
Tsai HH, Li H, Fuentealba LC, Molofsky AV, Taveira-Marques R, Zhuang H, Tenney A, Murnen
AT, Fancy SP, Merkle F et al (2012) Regional astrocyte allocation regulates CNS
synaptogenesis and repair. Science (New York, NY) 337:358–362
Valerius MT, Li H, Stock JL, Weinstein M, Kaur S, Singh G, Potter SS (1995) Gsh-1: a novel
murine homeobox gene expressed in the central nervous system. Dev Dyn 203:337–351
Visel A, Thaller C, Eichele G (2004) GenePaint.org: an atlas of gene expression patterns in the
mouse embryo. Nucleic Acids Res 32:D552–D556
Voogt JL, Lee Y, Yang S, Arbogast L (2001) Regulation of prolactin secretion during pregnancy
and lactation. Prog Brain Res 133:173–185
Wang W, Lufkin T (2000) The murine Otp homeobox gene plays an essential role in the specifica-
tion of neuronal cell lineages in the developing hypothalamus. Dev Biol 227:432–449
Westerlund J, Andersson L, Carlsson T, Zoppoli P, Fagman H, Nilsson M (2008) Expression of
Islet1 in thyroid development related to budding, migration, and fusion of primordia. Dev Dyn
237:3820–3829
Xu Q, Tam M, Anderson SA (2008) Fate mapping Nkx2.1-lineage cells in the mouse telencepha-
lon. J Comp Neurol 506:16–29
Yee CL, Wang Y, Anderson S, Ekker M, Rubenstein JL (2009) Arcuate nucleus expression of
NKX2.1 and DLX and lineages expressing these transcription factors in neuropeptide Y(+),
proopiomelanocortin(+), and tyrosine hydroxylase(+) neurons in neonatal and adult mice.
J Comp Neurol 517:37–50
Zhang L, Mathers PH, Jamrich M (2000) Function of Rx, but not Pax6, is essential for the formation
of retinal progenitor cells in mice. Genesis 28:135–142
Winding the Clock: Development
of Hypothalamic Structures Controlling 5
Biological Timing and Sleep

Dong Won Thomas Kim and Seth Blackshaw

Abstract
The daily cycle of sleep and wake governs all aspects of behavior and is under tight
homeostatic regulation. The core circadian oscillator in turn restricts sleep to
specific intervals of the solar day. The neurons that regulate both sleep and
circadian timing reside largely in the hypothalamus, and are linked in a complex,
reciprocally connected, and still incompletely characterized network. While our
knowledge of the neural circuitry controlling circadian rhythms and sleep has
advanced considerably, our understanding of the mechanisms by which they are
assembled during embryonic and postnatal development lag far behind. In this
chapter, we review advances in the understanding of hypothalamic circuitry
controlling circadian rhythms and sleep and our current knowledge of the gene
regulatory networks that guide their development. In addition, we discuss the

D. W. T. Kim
Solomon H. Snyder Department of Neuroscience, Johns Hopkins University School of Medicine,
Baltimore, MD, USA
S. Blackshaw (*)
Solomon H. Snyder Department of Neuroscience, Johns Hopkins University School of Medicine,
Baltimore, MD, USA
Department of Ophthalmology, Johns Hopkins University School of Medicine, Baltimore, MD,
USA
Department of Neurology, Johns Hopkins University School of Medicine, Baltimore, MD, USA
Center for Human Systems Biology, Johns Hopkins University School of Medicine, Baltimore,
MD, USA
Institute for Cell Engineering, Johns Hopkins University School of Medicine, Baltimore, MD, USA
e-mail: [email protected]

# Springer Nature Switzerland AG 2020 105

S. Wray, S. Blackshaw (eds.), Developmental Neuroendocrinology, Masterclass in
Neuroendocrinology 9, https://doi.org/10.1007/978-3-030-40002-6_5
106 D. W. T. Kim and S. Blackshaw

potential for future studies of hypothalamic development to provide new insights

into the mechanisms by which animals regulate the timing and duration of sleep.

Keywords
Circadian · Development · Hypocretin · Sleep · Suprachiasmatic · Timing ·
Transcription factor · Zona incerta

5.1 Fast, Slow, and Timely Hypothalamic Processes

Controlling Sleep

5.1.1 Introduction

The hypothalamus has long been known to be a central and critical regulator of
sleep. Focal lesions to the anterior hypothalamus and preoptic area often result in
insomnia, while lesions to the posterior hypothalamus result in lethargy and
hypersomnia (Scammell et al. 2017; Weber and Dan 2016). Several distinct neural
processes interact to regulate sleep. First of these is the neural circuitry that directly
regulates sleep/wake transitions. Sleep and wakefulness are mutually exclusive
states, and transitions between these states occur rapidly, with “half asleep” being
a state that lasts only seconds. Reciprocal transitions between wake and sleep can
occur frequently, particularly in species such as mice and human infants. An
analogous process in turn mediates rapid, biphasic, and bidirectional transitions
between non-rapid eye movement (NREM) and rapid eye movement (REM) sleep
phases. The neural circuitry that controls sleep and wake involves three major
components. The first and most extensively studied directly regulates the transition
between wake, NREM, and REM sleep. The second and third, which are much less
well understood, regulate homeostatic sleep pressure, and circadian changes in sleep
pressure, respectively. Sensory signals and changes in internal states can in turn
modulate sleep/wake transitions. An overview of hypothalamic cell types that
control sleep, wakefulness, and biological timing is shown in Fig. 5.1.

5.1.2 Neural Circuitry Directly Regulating Sleep/Wake Transitions

Recent studies, particularly since the advent of optogenetics, have identified several
hypothalamic neuronal subtypes whose activity is sleep regulated, and which rapidly
induce either wakefulness or sleep following activation or inhibition (Fig. 5.2a).
Primary wake-promoting neurons act by sending widespread, and typically excit-
atory, projections throughout the forebrain. In the hypothalamus, these include the
histaminergic neurons of the tuberomammillary nucleus (TMN), whose wake-
promoting signals are blunted by conventional antihistamines; orexinergic neurons
of the lateral hypothalamus (LH), which degenerate in narcoleptic patients; and other
populations such as glutamatergic neurons of the supramammilary nucleus (Pedersen
5 Winding the Clock: Development of Hypothalamic Structures Controlling. . . 107

Fig. 5.1 Locations of hypothalamic sleep and wake-promoting neurons. (a) Sleep-promoting
neurons of the preoptic area. (b) Sleep and wake-promoting neurons of the dorsolateral hypothala-
mus. (c) Neuronal subtypes of the suprachiasmatic nucleus regulating biological timing. (d) Wake-
promoting neurons of the posteroventral hypothalamus. Diagrams of transverse sections through
hyopothalamus, centered on third ventricle (black). Abbreviations: Avp arginine vasopressin, Cck
cholecystokinin, Crh corticotropin-releasing hormone, DMH dorsomedial hypothalamic nucleus,
GABA gamma-amino butyric acid, Gal galanin, Grp gastrin-related peptide, Hdc histamine decar-
boxylase, LH lateral hypothalamus, Pmch pro-melanin concentrating hormone, SCN
suprachiasmatic nucleus, Tac1 tachykinin 1, TMN tuberomammillary nucleus, Vip vasoactive
intestinal peptide, VLPO ventrolateral preoptic area, ZIv zona incerta, ventral region

et al. 2017). Alternatively, wake-promoting neurons can be inhibitory and often

function by directly inhibiting the activity of sleep-promoting neurons. Examples
of such inhibitory wake-promoting hypothalamic neurons include subpopulations of
GABAergic neurons in the LH (Venner et al. 2016) and in the preoptic area (POA)
(Chung et al. 2017) (Fig. 5.2a).
108 D. W. T. Kim and S. Blackshaw

Fig. 5.2 Connections and actions of hypothalamic sleep-regulating neurons. (a) Schematic of
hypothalamic neurons directly regulating sleep and wake. (b) Pre- and postsynaptic connections of
sleep pressure-regulating Lhx6-expressing neurons of the zona incerta. (c) Delayed effects on sleep
of chemogenetic-dependent activation and inhibition of Cre-dependent DREADD constructs deliv-
ered Lhx6-expressing neurons of the zona incerta by bilateral stereotaxic injection [adapted from
Liu et al. (2017)]. (a, b) Sagittal brain sections; (c) coronal brain section. Abbreviations: CNO
clozapine-N-oxide, DB ventral diagonal band, DMH dorsomedial hypothalamic nucleus, DREADD
designer receptor exclusively activated by designer drugs, DRV dorsal raphe nucleus, ventral
region, LH lateral hypothalamus, LS lateral septum, POA preoptic area, PPTG pedunculopontine
tegmental nucleus, REM rapid eye movement, RtTg rostral tegmental nucleus, SNPC substantia
nigra, pars compacta, TMN tuberomammillary nucleus, vlPAG ventrolateral periaqueductal grey,
VTA ventral tegmental area, ZIv zona incerta, ventral region, ZT zeitgeber time

Conversely, primary sleep-promoting neurons identified thus far are invariably

GABAergic, and send direct, inhibitory inputs to wake-promoting neurons. Examples
include sleep-active GABAergic POA neurons, which directly suppress the activity of
histaminergic and orexinergic neurons (Chung et al. 2017; Saito et al. 2013). Other
sleep-promoting hypothalamic neurons, such as MCH-positive LH neurons, do so by
directly inhibiting wake-promoting neurons in the midbrain and brainstem (Ferreira
et al. 2017). Sleep-promoting hypothalamic neurons, in turn, generally directly
promote either NREM or REM sleep, with GABAergic POA neurons selectively
promoting NREM and galaninergic DMH neurons promoting REM (Chen et al. 2018;
Chung et al. 2017). Wake-promoting neurons send reciprocal projections to sleep-
promoting neurons, and in turn inhibit these cells, usually by activation of Gi-coupled
GPCRs. This biphasic nature of sleep and wakefulness is thus encoded in the core
neural circuitry that controls the initiation and maintenance of sleep.
5 Winding the Clock: Development of Hypothalamic Structures Controlling. . . 109

5.1.3 Neural Circuitry Regulating Homeostatic Sleep Pressure

A second, and much less well understood, component of sleep-regulatory circuitry

are the neural mechanisms that regulate homeostatic sleep pressure (Fig. 5.1).
Homeostatic sleep pressure accumulates continuously during time spent awake, is
measured separately for NREM and REM sleep phases, and is gradually dissipated
during sleep (Borbely 1982). Sleep deprivation, once relieved, leads to a compensa-
tory rebound in sleep time (Brunner et al. 1990). The molecular signals of sleep
pressure are still mostly obscure but are known to include byproducts of normal
cellular metabolism such as adenosine, which activates GPCRs that are blocked by
caffeine (Elmenhorst et al. 2007, 2017). The neural circuitry that encodes sleep
pressure, and relays these signals to the neural circuitry that directly regulates sleep/
wake transition, is even less well characterized. In contrast to the fast-acting core
circuitry that directly regulates sleep/wake transitions, it is both slow acting and
highly stable in the face of sensory input and changes in internal state. Although
nNOS-positive cortical interneurons are activated by elevated sleep pressure
(Dittrich et al. 2015; Morairty et al. 2013), little is still known about how sleep
pressure is detected by hypothalamic sleep-regulatory circuitry.
Insight into how such circuitry may be organized comes from the recent identifi-
cation of Lhx6-positive GABAergic neurons of the hypothalamic zona incerta (ZI)
as candidate regulators of homeostatic sleep pressure (Liu et al. 2017). These
neurons are activated in response to accumulating sleep pressure, and are both
directly presynaptic to multiple populations of wake-promoting neurons and exten-
sively interconnected with another (Fig. 5.2b). Direct activation and inhibition of
these cells produce robust but slow-onset changes in sleep/wake patterns (Fig. 5.2c),
likely reflecting the effect of inputs from these, which counteract and modulate direct
inhibitory inputs onto directly wake-promoting neurons. Other extensively
interconnected populations of GABAergic neurons are found in both the hypothala-
mus and other brain regions and may also play a similar role in signaling homeostatic
sleep pressure.

5.1.4 Neural Circuitry Regulating Circadian Sleep Pressure

A third component is controlled by the core circadian clock located in the

suprachiasmatic nucleus (SCN) of the hypothalamus and modulates sleep pressure
depending on time of day (Fig. 5.1) (Daan et al. 1984). It signals elevated sleep
pressure during the dark phase for diurnal species and during the light phase for
nocturnal species. This master clock regulates not only the sleep/wake cycle but also
controls virtually all aspects of mammalian physiology (Reddy and O’Neill 2009).
Much like the circuitry regulating homeostatic sleep pressure, changes in the activity
of this component of the sleep-regulatory circuit occur slowly, though cyclically.
Moreover, the core circadian oscillator makes extensive use of negative feedback at
both the cellular level and in local circuitry to encode these slow but robust changes
in activity levels.
110 D. W. T. Kim and S. Blackshaw

Fig. 5.3 Suprachiasmatic nucleus (SCN): circadian rhythms and neuronal differentiation. (a)
Overview of light-dependent regulation of the core circadian oscillator in the SCN. Light is relayed
to the SCN via intrinsically photosensitive retinal ganglion cells to regulate rhythmic gene expres-
sion. SCN neurons. (b) Transcriptional regulatory networks (left) controlling SCN differentiation,
(right) and controlling expression of clock-regulating neuropeptides (Avp arginine vasopressin, Grp
gastrin releasing peptide, Vip vasoactive intestinal peptide)

Like all cells in the adult, a biochemical circadian oscillator is active in neurons of
the SCN (Fig. 5.3). This takes the form of a negative feedback loop, with a
heterodimeric transcriptional complex consisting of the transcription factors Bmal1
and Clock, which drive the expression of Per and Cry genes, which in turn feedback
to inhibit their own transcription. This core oscillator is reinforced and stabilized by
linked transcriptional negative feedback loops, involving the nuclear hormone
receptor superfamily members Rora, Rorb, and Nr1d1 (Partch et al. 2014). In
sharp contrast to virtually all other tissues, however, the cellular clocks of SCN
neurons undergo spontaneous synchronization when grown in dispersed cell culture
and remain robustly synchronized when maintained ex vivo in intact organ culture
(Mohawk and Takahashi 2011). Moreover, while cellular clocks in other tissues are
readily reset and synchronized by a range of external signals—including
5 Winding the Clock: Development of Hypothalamic Structures Controlling. . . 111

temperature, metabolites, and growth factors—cells in the SCN robustly retain their
phase and synchrony under a broad range of treatments. These unique features arise
at least in part as a result of the organization of local neural circuitry in the SCN. Like
the ZI, the SCN consists of interconnected GABAergic neurons, and this extensive
collateral inhibition buffers the SCN against perturbation. Inhibition of neuronal
excitability in the SCN reverses this stability, allowing cellular oscillators to be
readily reset by light and temperature signals that would normally produce no
change in phase (Buhr et al. 2010; Yamaguchi et al. 2013). SCN neurons also
release a diverse array of neuropeptides—including Vip, Avp, Grp, Penk, and
Vgf—which likewise contribute to phase synchrony and stabilization. While the
molecular mechanisms and, to a lesser extent, the local circuitry that controls
circadian timekeeping in the SCN, are now fairly well understood, the neural
circuitry that relays these signals from the SCN to other brain regions is not.

5.1.5 Sensory and Internal Signals Regulating Sleep/Wake

Transitions

Finally, a range of sensory and internal signals can modulate sleep, usually by
directly promoting wakefulness without altering either homeostatic or circadian
sleep pressure. Sensory signals include loud noises, temperature extremes, and
pain, with the most complex of these in its effects being light. Light directly inhibits
(in the case of diurnal mammals) or promotes (in the case of nocturnal mammals) the
onset of sleep. Moreover, while circadian sleep pressure is highly robust in the face
of most sensory and internal signals, it can be reset by light signals relayed directly
from intrinsically photosensitive retinal ganglion cells (ipRGCs) (Hattar et al. 2002).
These ipRGCs directly activate neurons in the ventral SCN, which in turn leads to a
gradual and progressive resetting of cellular SCN clocks. This phase shift occurs at
the rate of approximately 1 h per day, with misalignment of circadian and clock time
leading to the phenomenon of jet lag (Sack et al. 2007). Some examples of internal
states that acutely modulate sleep (though not sleep pressure) include sympathetic
and behavior arousal, such as hunger and thirst, or psychotropic drugs that directly
target arousal circuitry (Saper et al. 2010). Despite the strong and rapid effects of
these signals on sleep, surprisingly little is known about the circuitry mediating
clock-dependent regulation of sleep. In addition, virtually nothing is known about
the mechanisms by which changes in sleep state in turn induce sleep-dependent
changes in internal states and sensory processing.

5.1.6 Ontogeny of Sleep–Wake Cycles and Behavioral Rhythmicity

As any parent of a newborn knows, light and circadian-dependent regulation of

sleep/wake cycle develops gradually during the postnatal period. Using standard
EEG recordings, wake, NREM, and REM sleep can be clearly distinguished in
premature human infants at 6 months gestational age (Foreman et al. 2008), and in
112 D. W. T. Kim and S. Blackshaw

rodents between postnatal day 12 and 20 (Daszuta and Gambarelli 1985; Frank and
Heller 1997; Nelson et al. 2013). However, analysis of other sleep-regulated
behaviors, such as muscle tone measured by electromyogram (EMG), suggest that
sleep/wake transitions may be evident in rodents as early as postnatal day 2 (Gall
et al. 2008). Light-dependent regulation of sleep–wake cycles is evident by weaning
in rodents (Nelson et al. 2013; Sack et al. 2007) and begin to emerge by 10–12 weeks
postnatal in humans (Davis et al. 2004). Circadian control of circulating cortisol and
core body temperature also becomes evident around this age (Joseph et al. 2015),
and this is likely the point at which sleep/wake cycles also come under circadian
regulation (Blumberg et al. 2014; Galland et al. 2012). This process occurs long after
molecular rhythms are first detected in the SCN. These are first observed by E15,
coincident with expression of the first SCN-enriched neuropeptides, and become
strong and well consolidated by postnatal day 2 (Carmona-Alcocer et al. 2018),
implying a substantial delay in the development of efferent neural circuitry relaying
SCN output. There is a gradual shift in sleep homeostasis through life, with progres-
sively less time spent in sleep, with major reductions observed first in REM, then in
deep NREM (Dzierzewski et al. 2018; Miles and Dement 1980). Sleep/wake phases
are relatively phase-advanced in childhood, phase-delayed in adolescence and early
adulthood, and phase-advanced and less well consolidated in the elderly. Both
Mendelian and multifactorial genetic variation control the extent to which sleep/
wake cycles are phase-advanced or phase-delayed relative to the solar day, giving
rise to what are known as chronotypes (Roenneberg et al. 2003). This chronotypic
variation in sleep/wake timing peaks in early adulthood and declines slowly through-
out the lifespan (Fischer et al. 2017).

5.2 Development of Sleep-Regulatory Circuitry

5.2.1 Introduction

Like all CNS structures, the developing hypothalamus is initially patterned by

gradients of growth and guidance factors, which then induce expression of transcrip-
tion factors in a concentration-dependent manner. These broadly expressed tran-
scription factors then act in a combinatorial fashion to specify individual subregions
prior to the onset of hypothalamic neurogenesis. Once neurogenesis begins, many of
these same transcription factors have additional roles in controlling the specification
and differentiation of individual neuronal subtypes that regulate biological timing
and sleep, among other physiological behaviors. Finally, a largely uncharacterized
set of molecular cues control the formation of axonal projections and synaptic
connections that underpin these circuits. We will review our current knowledge of
how these processes guide the formation of the sleep-regulating cells for which this
is best understood: the neurons of the SCN and the dorsolateral hypothalamus.
5 Winding the Clock: Development of Hypothalamic Structures Controlling. . . 113

5.2.2 Early Hypothalamic Patterning and Role of Extrinsic Factors

The POA derives from Foxg1-positive telencephalic neuroepithelium, while the

hypothalamus proper arises from Foxd1-positive diencephalic neuroepithelium.
The extrinsic factors discussed here arise from different sources and have substan-
tially different effects on the patterning of these structures.

5.2.2.1 Shh
Shh (sonic hedgehog) has been studied more extensively than any other extrinsic
factor in the context of hypothalamic development. Shh expression in the notochord,
and its anterior derivative the prechordal plate, leads to induction of Shh expression
in the ventral floor plate all along the neuraxis, including in hypothalamic
neuroepithelium. Loss of function of Shh at this stage blocks development of all
hypothalamic structures (Blaess et al. 2014). Soon thereafter, Shh expression is
induced in a more dorsally located region known at the basal plate domain, and
floorplate expression is suppressed. Selective loss of function of Shh in the basal
plate domain prior to the onset of neurogenesis leads to the loss of all anterior and
tuberal hypothalamic markers, along with severe defects in neural progenitor prolif-
eration. This prevents development of all sleep-associated neurons, with the excep-
tion of supramammilary nucleus neurons (Shimogori et al. 2010).

5.2.2.2 Canonical Wnt Signaling

Canonical Wnt signaling also plays a critical role in hypothalamic patterning and the
development of sleep-regulatory neurons. Selective loss of function of beta-catenin—
which is essential for mediating canonical Wnt signaling—in hypothalamic
neuroepithelium leads to a reduction in the number of multiple sleep-regulatory
neurons, including GABAergic neural precursors in the SCN and LH, as well as
orexinergic neural precursors (Newman et al. 2018b). Overexpression of constitu-
tively active beta-catenin blocks specification of all hypothalamic GABAergic
neurons but increases generation of LH orexinergic neurons (Newman et al.
2018b). Milder phenotypes are seen in mice and zebra fish showing altered expres-
sion of Lef1, a major transcriptional effector for canonical Wnt signaling (Lee et al.
2006; Xie et al. 2017). Loss of function of Lef1 selectively disrupts development of
LH MCH neurons, while the gain of function increases their number.

5.2.2.3 Other Extrinsic Factors

Other extrinsic factors are known to regulate hypothalamic development. In chick, a
temporal gradient of FGF signaling plays a critical role in anteroposterior patterning
and proliferation of early hypothalamic progenitors (Fu et al. 2017). At later stages
of chick development, BMP signaling specifies anterioventral identity and drives
proliferation of ventral hypothalamic progenitors (Manning et al. 2006; Ohyama
et al. 2008). Finally, Notch/Delta signaling has been shown to regulate proliferation
and neurogenesis in both anterior and tuberal hypothalamic progenitors (Aujla et al.
2013; Biehl et al. 2018). Although it is likely that all of these pathways play an
114 D. W. T. Kim and S. Blackshaw

Fig. 5.4 Molecular organization of the hypothalamus and prethalamus in the embryonic (day E12)
mouse brain. (a) Parasagittal schematic (adapted from Shimogori et al. (2010)). (b) Selected
molecular markers for the indicated regions. Abbreviations: from top to bottom, left group: ZLI
zona limitans intrathalamica, PreThal prethalamus, EmThal thalamic eminence, PVN hypothalamic
paraventricular nucleus, vAH ventral anterior hypothalamus, ID intrahypothalamic diagonal, TT
tuberomammillary terminal, VMH ventromedial hypothalamus, ArcN arcuate nucleus, PMN
premammillary nucleus, MM mammillary nucleus, SMM supramammillary nucleus, Shh sonic
hedgehog

important role in development of sleep-regulatory neurons in mammals, their spe-

cific function has not been investigated.

5.2.3 A Temporal Cascade of Intrinsic Factors Controlling SCN

Development

Despite its small size (~20,000 neurons total), the physiological importance and
compact, easily identifiable structure of the SCN has resulted in the transcription
factors that control its development being better characterized than any other struc-
ture in the hypothalamus. Nearly a decade of work has made considerable progress
in identifying a hierarchical network of transcription factors that controls specifica-
tion, differentiation, and adult function of the SCN (Fig. 5.4).

5.2.3.1 Dual Regulation of SCN Identity by Lhx2

The LIM homeodomain transcription factor Lhx2 is expressed during early stages of
development in both optic neuroepithelium and adjacent neuroepithelium that give
rise to the anterioventral hypothalamus (Porter et al. 1997; Shimogori et al. 2010).
Loss of function of Lhx2 blocks expression of genes specific to both the eye field and
the SCN. Strikingly, this also leads to an ectopic upregulation of genes specific to the
adjacent neuroendocrine paraventricular nucleus (PvN) and thalamic eminence
5 Winding the Clock: Development of Hypothalamic Structures Controlling. . . 115

(EmThal) (Roy et al. 2013). In the neuroretina, the repressive function of Lhx2
persists even after neurogenesis has begun (de Melo et al. 2016; Roy et al. 2013).

5.2.3.2 Promotion of SCN Identity and Nucleogenesis by Six3, Six6,

and Vax1
The homeodomain transcription factors Six3 is broadly expressed in anterior hypo-
thalamic neuroepithelium, while Six6 and Vax1 show a more restricted expression
that is confined to anterioventral hypothalamus and eye field. Lhx2 is necessary to
maintain expression of Six6 and Vax1, though not Six3, in this region (Roy et al.
2013). Loss of function of Six3 in hypothalamic neuroepithelium leads to a loss of all
SCN-specific molecular markers, along with the absence of a histologically defined
SCN (VanDunk et al. 2011). Six6 loss of function results in milder and more variable
defects, with some persistence of SCN-specific markers, and moderate to severe
defects in SCN nucleogenesis (Clark et al. 2013). Vax1 defects are milder yet and
exhibit only moderate defects in expression of SCN-specific genes (Hanne et al.
2016). Although both Six3 and Six6 expression persist in adult SCN (Clark et al.
2013; VanDunk et al. 2011), loss of Six3 expression in SCN neural precursors
produced no obvious developmental phenotype.

5.2.3.3 Delayed Action of Foxd1 in Maintenance of SCN Differentiation

The Forkhead domain transcription factor Foxd1 is broadly, transiently, and selec-
tively expressed in hypothalamic neuroepithelial cells, although expression is
maintained longest in anterior hypothalamus (Salvatierra et al. 2014; Shimogori
et al. 2010). The loss of function of Foxd1 results in a severe, although puzzlingly
delayed, defect in SCN differentiation. Foxd1 mutants initially show normal expres-
sion of early SCN markers, although they display modest reductions in expression of
both Six3 and Vax1 expression (Newman et al. 2018a). However, beginning at E16
in mouse, once neurogenesis is completed, progressive defects in expression of
SCN-specific genes are observed. By P0, virtually all SCN-specific markers are
lost, along with all histological evidence for a distinct SCN (Newman et al. 2018a).
The mechanism mediating these effects is unclear, and the long delay between
termination of Foxd1 expression and phenotype onset is striking. It is possible that
loss of function of Foxd1 triggers epigenetic changes that selectively inactivate
cis-regulatory elements that maintain expression of SCN-specific genes or other
transcription factors required for SCN differentiation at later developmental ages.

5.2.3.4 Lhx1 Controls SCN Terminal Differentiation

The first selective marker of postmitotic SCN neural precursors is the LIM
homeodomain factor Lhx1, which is detected beginning at E11 in mouse. Lhx1
expression in SCN is lost in Lhx2- and Six3-mutants and, while initially normal in
Foxd1-mutants, is lost by P0 (Table 5.1). Selective loss of function of Lhx1 in
anterior hypothalamic neuroepithelium does not disrupt either initial SCN specifica-
tion or nucleogenesis. Instead, mutant animals lose expression of genes that are
selectively expressed in mature SCN neurons, including virtually all SCN-enriched
neuropeptides, including Vip, Avp, and Grp (Bedont et al. 2014). Selective loss of
116 D. W. T. Kim and S. Blackshaw

Table 5.1 Behavioral defects in mouse mutants that disrupt development of hypothalamic neurons
controlling sleep and circadian timing
Mutant line Cellular phenotype Behavioral phenotype References
Six3-Cre: Broadly decreased Loss of SCN-enriched Bedont et al.
Lhx1lox/lox expression of neuropeptides. Disruption of (2014, 2017)
SCN-enriched activity, sleep, and body
neuropeptides. temperature rhythms. Loss of
resistance to temperature-
dependent resetting of SCN clock.
Loss of light-dependent regulation
of sleep.
Rora-Cre: Decreased expression Moderate defects in circadian Hatori et al.
Lhx1lox/lox of a subset of activity rhythms. Enhanced light- (2014)
SCN-enriched dependent resetting of circadian
neuropeptides. phase.
Zfhx3sci/sci Broadly reduced Defective circadian activity Balzani et al.
expression of rhythms. Altered sleep and interval (2016) and
SCN-enriched timing. Parsons et al.
neuropeptides. (2015)
Ubc- Not determined. Disrupted circadian activity Wilcox et al.
CreER; rhythms. (2017)
Zfhx3lox/lox
Foxd1- Loss of hypothalamic Reduced sleep time, increased Liu et al. (2017)
Cre; Lhx6+ neurons. wake time.
Lhx6lox/lox
Lhx9 / Reduction in Hcrt+ Reduced locomotor activity and Dalal et al.
neuron number. increased wake time. (2013)

function in postmitotic SCN neural precursors results in a milder but broadly similar
phenotype (Hatori et al. 2014). Lhx1 expression is maintained in adult SCN
(VanDunk et al. 2011) and is regulated by both the circadian clock and by light
(Hatori et al. 2014).
Changes in gene expression in Lhx1-deficient mice appear to reflect both the
direct and indirect effects of Lhx1. Lhx1 directly regulates Vip expression (Hatori
et al. 2014) and Vip shows the strongest effects of all SCN-enriched neuropeptides
on regulation of cellular and behavioral rhythms (Maywood et al. 2011). Loss of
function of Vip leads to defects in circadian rhythmicity, and loss of expression of a
large fraction of Lhx1-dependent genes, including neuropeptides such as Avp and
Penk, implying that Lhx1-dependent regulation of Vip expression accounts for these
effects (Bedont et al. 2017). However, expression of other SCN-enriched
neuropeptides such as Grp, Nms, and Prok2 is unaffected by Vip loss of function,
implying that Lhx1 may act as a master regulator of SCN neuropeptide expression
(Table 5.1).

5.2.3.5 Zfhx3 Regulates Expression of SCN-Enriched Neuropeptides

The zinc finger homeodomain factor Zfhx3 is expressed in both neural progenitors
and postmitotic precursor cells of the anterioventral hypothalamus (Parsons et al.
5 Winding the Clock: Development of Hypothalamic Structures Controlling. . . 117

2015). Zfhx3 hypomorphic mutants resemble Lhx1 mutants in showing lower

expression levels of SCN-enriched neuropeptides—including Vip, Grp, Nms, and
Prok2 (though not Avp)—along with a limited subset of other genes. Both mutants
show normal SCN specification and nucleogenesis are maintained. Like Lhx1, Zfhx3
directly regulates transcription of SCN-enriched neuropeptides (Parsons et al. 2015).
However, the total number of SCN neurons expressing these neuropeptides is
unchanged, implying that Zfhx3 is not necessary for the expression of these genes
but rather regulates their expression levels. Also like Lhx1, Zfhx3 is expressed in
mature SCN neurons and is required in adult to maintain normal expression of many
of these same SCN-enriched genes (Wilcox et al. 2017) (Table 5.1).

5.2.3.6 Other Candidate Regulators of SCN Differentiation and Function

Gene expression analysis has identified other transcription factors expressed in
anterioventral hypothalamus and/or SCN neural precursors that are strong
candidates for controlling SCN development. These include homeodomain tran-
scription factors such as Dlx family members and Arx, which are also necessary for
normal development of telencephalic interneurons (Shimogori et al. 2010). Two
Zfhx3 paralogues—Zfhx2 and Zfhx4—are also expressed in SCN and associated
structures. Like Zfhx3, they show reduced expression in Foxd1 mutant mice
(Newman et al. 2018a) and may play a similar role in control of SCN differentiation
and/or function.

5.2.4 Intrinsic Factors Controlling Specification of Wake

and Sleep-Promoting Neurons in the Posterolateral
Hypothalamus

Much less is known about transcription factors that control development of sleep-
regulating neurons in the posterior hypothalamus, with work on this topic having
mostly focused on the function of the LIM homeodomain factors Lhx6 and Lhx9.
Lhx6 and Lhx9 are expressed in adjacent, nonoverlapping regions of the posterolat-
eral hypothalamus from E11 into adulthood in mouse (Shimogori et al. 2010). Lhx6
expression is ultimately restricted to GABAergic neurons in the ventral ZI, DMH,
and LH (Liu et al. 2017), while Lhx9 is restricted to excitatory orexinergic neurons
(Dalal et al. 2013). Although loss of function of Lhx6 disrupts the function of sleep-
promoting ZI neurons, exactly how this gene regulates the specification and/or
function of these cells is still unclear. Lhx9 expression is detected in orexinergic
cells long before orexin itself is expressed (Shimogori et al. 2010), and in zebra fish,
Lhx9 directly regulates orexin expression (Liu et al. 2017). In mice, loss of function
of Lhx9 leads to a reduction in both the levels of orexin expression and a 30%
reduction in the number of orexigenic neurons (Dalal et al. 2013) (Table 5.1). Lhx9-
deficient zebra fish shows a complete loss of orexin expression (Liu et al. 2017), In
contrast, overexpression of Lhx9 in both species increases the number of orexinergic
neurons, but in mice did so only in posterior hypothalamus (Liu et al. 2015).
Moreover, viral overexpression of Lhx9 in adult Lhx9-deficient mice does not rescue
118 D. W. T. Kim and S. Blackshaw

defects in hypocretin expression (Liu et al. 2015). Gene expression profiling analysis
of orexinergic neurons has identified other transcription factors that are selectively
expressed in orexinergic neurons, although their role in differentiation and function
of these cells has not been investigated.

5.2.5 Development of Other Hypothalamic Cell Types Regulating

Sleep and Wake

Little is known about the transcriptional regulatory network controlling development

of other sleep-regulatory circuitry, although multiple candidate genes have been
identified. These include the homeodomain factors Hmx2/3, which are expressed in
both the POA and posterior hypothalamic neural progenitors that give rise to both
histaminergic and orexinergic neurons (Liu et al. 2015). Dlx family genes are also
expressed in developing GABAergic neurons throughout the hypothalamus and
POA and may play important roles in specification and differentiation of multiple
wake and sleep-promoting neuronal subtypes.

5.2.6 Control of Nucleogenesis, Axonal Targeting,

and Synaptogenesis

Virtually nothing is known about the molecular mechanisms that control

nucleogenesis and circuit formation, with most of what is known being negative
data. Despite the relative ease in visualizing the retinohypothalamic tract, no genes
are known to selectively regulate innervation of the SCN by ipRGCs. Despite its
prominent role in SCN development, retinohypothalamic tract formation occurs
normally in Lhx1 mice (Bedont et al. 2014). Likewise, nothing is known about the
mechanisms guiding the extensive reciprocal synaptic connectivity of the SCN or ZI,
which are likely to be critical for their circadian and/or sleep-regulating function.
Further studies on this topic are urgently needed.

5.3 Insights into Hypothalamic Circuitry and Physiology

Gained from Studies of Hypothalamic Development

5.3.1 Lhx1 and Control of SCN Clock Function

Mice mutant for Lhx1 have provided important insights into the molecular
mechanisms regulating circadian synchrony and behavioral output of the SCN.
SCN neurons express a diverse collection of neuropeptides, including Vip, Avp,
Grp, Prok2, Nms, Penk, and Vgf (Abrahamson and Moore 2001). Expression of
many of these is regulated by both circadian time and light, and they have been
proposed to both regulate the synchrony of cellular oscillators within the SCN and
mediate output from the SCN to other brain regions, allowing them to keep time with
5 Winding the Clock: Development of Hypothalamic Structures Controlling. . . 119

the master clock (Aton and Herzog 2005). Genetic and pharmacological analysis
have identified a hierarchy of function among these neuropeptides, at least with
respect to synchronization of cellular oscillators, with Vip being dominant over Avp,
and Avp over Grp (Maywood et al. 2011). The sheer number of candidate
neuropeptides, however, has not made it feasible to investigate the role of
neuropeptides in central clock function more generally, as opposed to investigating
the function of individual neuropeptides.
Disruption of Lhx1 function in hypothalamic neuroepithelial cells leads to a near-
complete loss of expression of all SCN-enriched neuropeptides without grossly
affecting either regional organization or GABAergic neurotransmission in the
SCN (Bedont et al. 2014). Cellular rhythms within the SCN remain locally
synchronized in explants from Lhx1-deficient mice, but desynchrony is observed
among different regions within the SCN, and this becomes more pronounced over
time (Bedont et al. 2014). This is reflected in the behavioral phenotype of mutant
mice, which though grossly normal, show either phase splitting or lack of circadian
rhythmicity in locomotor and body temperature rhythms (Bedont et al. 2014).
Interestingly, the loss of neuropeptide expression in Lhx1-deficient mice renders
the core behavioral oscillator highly susceptible to perturbation and resetting. SCN
explants from Lhx1-deficient mice lose their resistance to resetting by temperature
pulses, essentially phenocopying samples in which neuronal excitability is
suppressed using tetrodotoxin (Bedont et al. 2017). Moreover, induction of fever
by rhythmic injections of lipopolysaccharide is sufficient to reset both core body
temperature and locomotor rhythms of these mutant mice (Bedont et al. 2017). These
same animals are also rendered susceptible to phase shifts induced by
intracerebroventricular canulation of SCN-enriched neuropeptides that do not nor-
mally influence behavioral rhythms (Bedont et al. 2014). Mice showing later loss of
function of Lhx1 show milder phenotypes that also reflect reduced coupling of
cellular oscillators in the SCN (Hatori et al. 2014). Taken together, these data
show that Lhx1-regulated neuropeptide gene expression plays a central role in the
stabilization of cellular oscillators within the SCN, buffering them against resetting
by internal signals and also in mediating efferent signaling from the SCN to other
brain regions.

5.3.2 Lhx1 and Light-Dependent Regulation of Sleep and Circadian

Rhythms

Lhx1-deficient mice also show severe defects in light-dependent regulation of both

behavioral rhythms and sleep. Although projections from ipRGCs to the SCN
remain intact following early deletion of Lhx1 (Bedont et al. 2014), these animals
do not show increases in c-fos expression in the SCN in response to 30 min light
pulses (Bedont et al. 2017). Moreover, while activity rhythms of Lhx1-deficient mice
synchronize with 12 h light-dark cycles, light periods shorter than 6 h are not
effective at doing so. This stands in sharp contrast to wild-type mice, which robustly
synchronize and entrain to skeleton photoperiods where light phases are as short as
120 D. W. T. Kim and S. Blackshaw

1 h (Jud et al. 2005). This implies that SCN-enriched neuropeptides are essential for
relaying light signals from ipRGCs to cellular clocks in the SCN. These phenotypes
are not observed when Lhx1 is selectively deleted in SCN neural precursors, with
these mutants showing enhanced resetting of circadian activity rhythms by light
pulse (Hatori et al. 2014).
Early deletion of Lhx1 leads to an unexpected loss of acute light-dependent
regulation of sleep (Bedont et al. 2017). This phenotype is distinct from light-
dependent regulation of the core circadian clock, which is completely disrupted in
these mutants, and identifies an unexpected role for the SCN in direct regulation of
sleep/wake transitions. These mutants do not show any changes in sleep homeosta-
sis. The neural circuitry downstream of the SCN that controls acute light-dependent
regulation of sleep remains unclear, however.

5.3.3 Zfhx3 and Control of Circadian Rhythms, Interval Timing,

and Sleep

Like Lhx1, Zfhx3 has an essential role in control of circadian locomotor rhythms and
sleep. Dominant ENU-induced missense mutations in Zfhx3 show shortened bio-
chemical and circadian activity periods, while these were lengthened by Zfhx3
knockdown (Parsons et al. 2015). In contrast, selective knockout of Zhfx3 in adult
animals results in either shortened rhythms or else led to complete arrhythmicity.
Animals that retained rhythmicity rapidly entrained to phase-advanced (but not
phase-delayed) light cycles, implying that coupling among cellular SCN oscillators
was attenuated (Parsons et al. 2015).
Missense mutations in Zfhx3 mutants also showed evidence for shortened interval
timing, when trained to respond with a 10s delay to a conditioned stimulus (Balzani
et al. 2016), suggesting that they may show a more general defect in biological
timing. Zfhx3 missense mutants also show a reduction in rebound NREM sleep
following 6 h. of sleep deprivation but do not show any significant differences in
sleep homeostasis under baseline conditions (Balzani et al. 2016). Light-dependent
regulation of sleep is also intact in these mutants (Balzani et al. 2016). Since these
data were obtained using constitutive mutants, and Zfhx3 is expressed broadly in the
forebrain, it is unclear whether defects in interval timing and sleep homeostasis are a
direct result of loss of Zfhx3 in the SCN or other hypothalamic structures.

5.3.4 Lhx6 and Detection of Sleep Pressure

Lhx6-expressing neurons in the ventral ZI are directly activated by sleep pressure

and send inhibitory GABAergic projections to wake-promoting neurons in both the
LH and midbrain (Liu et al. 2017). Slice recordings show that optogenetic activation
of Lhx6 ZI neurons directly inhibits wake-promoting hypocretin-expressing and
GABAergic cells in the LH (Liu et al. 2017). In addition, Lhx6-positive ZI neurons
receive presynaptic inputs from multiple sleep/wake-regulating neurons.
5 Winding the Clock: Development of Hypothalamic Structures Controlling. . . 121

Conditional deletion of Lhx6 from the developing diencephalon leads to decreases in

both NREM and REM sleep. Furthermore, selective activation and inhibition of
Lhx6-positive ZI neurons bidirectionally regulate sleep time in adult animals, in part
through hypocretin-dependent mechanisms (Liu et al. 2017). There is a striking
contrast between the very rapid inhibition of wake-promoting cells seen using
ex vivo optogenetic analysis in slice preparations and the much slower changes in
sleep/wake patterns seen following in vivo chemogenetic modulation of the activity
of ZI Lhx6-positive neurons. This likely reflects the fact that Lhx6-positive ZI
neurons are in turn functionally interconnected through inhibitory collateral
projections, resulting in slow global shifts in activity levels even following direct
stimulation. This pattern of connectivity may underpin the role of these neurons in
sensing and signaling sleep pressure to neuronal subtypes that directly regulate
sleep–wake transitions.

5.3.5 Lhx9 and Orexinergic Neurons

Lhx9-deficient mice show reduced locomotor activity and increased rest time,
although sleep was not directly investigated (Dalal et al. 2013). Unfortunately, in
zebra fish, the effects of gain and loss of function of Lhx9 on sleep behavior have
also not been investigated (Liu et al. 2015)—but since Lhx9 is crucial for Hcrt-
expressing neuronal differentiation, we expect there is a defect in sleep behavior.

5.4 General Conclusions and Future Directions

5.4.1 The Urgent Need for Unambiguous Identification

of Sleep-Regulating Neurons

Aside from the transcriptional regulatory network controlling specification of the

SCN, little is known about the molecular mechanisms controlling development of
the central circadian pacemaker and the sleep/wake cycle. Since congenital defects
in these neural circuits are of considerable medical importance, this is somewhat
surprising, and needs to be prioritized in future. A key problem here is the lack of
unambiguous molecular markers for sleep-regulating neuronal subtypes. Most phys-
iological studies rely on the use of fast neurotransmitters or neuropeptides to identify
individual cell types linked to sleep regulation. These are generally fairly poor
markers of cell identity, except in the rare cases where the cell that expresses
genes show limited and highly restricted expression (e.g., Hcrt, MCH, histaminergic
neurons). ScRNA-Seq studies, which have barely scratched the surface in
characterizing hypothalamic cell diversity, have already identified hundreds of
distinct GABAergic and glutamatergic neuronal subtypes, and considerable diver-
sity among neurons expressing individual neuropeptides or calcium binding proteins
(Campbell et al. 2017; Chen et al. 2017; Romanov et al. 2017).
122 D. W. T. Kim and S. Blackshaw

Knowledge of a cell’s transcription factor expression profile and, in particular, the

expression profile of that cell immediately following terminal differentiation, can
greatly assist in accurate classification of major cell types, avoiding potentially
confounding effects of both aging and the effects of experience and
environmental-driven variation in gene expression that accumulate over the lifespan
(Hrvatin et al. 2018; Martinez-Jimenez et al. 2017) Developmental studies can thus
be of great help in helping identify a unique molecular address for each main subtype
of sleep-regulating cells—an essential step for unambiguous analysis of their
function.

5.4.2 Comparative Studies: What Can Be Learned

from Nonmammalian Vertebrates?

It is unclear how evolutionarily conserved this circuitry is likely to be. The location
of the central circadian clock in nonmammalian vertebrates is still unclear, with the
location and function of SCN being ambiguous even in birds. Some conservation of
key sleep-regulatory neurons (e.g., Hcrt, Lhx6, Pmch) is observed in zebra fish (Liu
et al. 2015; Wolf and Ryu 2013; Xie et al. 2017). In some cases, these have also been
shown to be functionally homologous (Liu et al. 2015). Studies of nonmammalian
vertebrates will help resolve this question—in particular, a systematic and unbiased
analysis of gene expression in developing hypothalamus like that performed in
developing mouse hypothalamus (Shimogori et al. 2010).

5.4.3 Human Studies and ES/iPS-Based Approaches

to Understanding Sleep

Recent years have seen rapid improvement in protocols for directed differentiation of
human ES and iPS cells into a broad range of neuronal subtypes, including a number
of mediobasal hypothalamic cell types (Merkle et al. 2015; Ogawa et al. 2018; Wang
et al. 2015; Wataya et al. 2008; Yao et al. 2017). The more recent development of
hypothalamic organoid preparations (Qian et al. 2016, 2018) raises the possibility of
potentially analyzing the formation and function of sleep-regulating structures such
as the SCN or ZI in vitro.
Before this can come about, however, we will need to know much more about the
basic molecular mechanisms controlling hypothalamic patterning and neurogenesis.
In particular, we lack a clear understanding of general principles guiding organiza-
tion and function of even the SCN. For instance, we have no idea how individual
transcription factors that have been found to be necessary to regulate rhythmicity
actually do so. What are their genomic targets? Is expression of these factors
sufficient for initiation of rhythmicity and synchrony of cellular clocks, as well as
necessary? The ability to perform gain and loss of function analysis in organoid
preparations (Eldred et al. 2018; Mariani et al. 2015), in combination with
techniques for globally profiling transcription factor target sites such as ATAC-
5 Winding the Clock: Development of Hypothalamic Structures Controlling. . . 123

Seq-based footprinting, offers the possibility of identifying the gene regulatory

networks that control the formation of key components of human sleep-regulatory
circuitry. Ultimately, this may allow us to build a sleeping brain in a dish.

Key References

Bedont et al. (2014)—Identified Lhx1 as a master regulation of neuropeptide

expression in neurons of the suprachiasmatic nucleus, and demonstrated the
necessity of these neuropeptides in stabilizing the master circadian clock.
Carmona-Alcocer et al. (2018)—Demonstrated that synchronized activity rhythms
first emerge in the mouse suprachiasmatic nucleus by embryonic day 15, and
show mature organization by postnatal day 2.
Dalal et al. (2013)—Identified Lhx9 as a critical regulator of development of
hypocretin neurons.
Liu et al. (2017)—Identified Lhx6-positive neurons of the zona incerta as responsive
to sleep pressure, and demonstrated that they are necessary and sufficient for
promotion of sleep.
Parsons et al. (2015)—Identified Zfhx3 as a master regulator of the development of
the suprachiasmatic nucleus.

References
Abrahamson EE, Moore RY (2001) Suprachiasmatic nucleus in the mouse: retinal innervation,
intrinsic organization and efferent projections. Brain Res 916:172–191
Aton SJ, Herzog ED (2005) Come together, right. . .now: synchronization of rhythms in a mamma-
lian circadian clock. Neuron 48:531–534
Aujla PK, Naratadam GT, Xu L, Raetzman LT (2013) Notch/Rbpjkappa signaling regulates
progenitor maintenance and differentiation of hypothalamic arcuate neurons. Development
140:3511–3521
Balzani E, Lassi G, Maggi S, Sethi S, Parsons MJ, Simon M, Nolan PM, Tucci V (2016) The Zfhx3-
mediated axis regulates sleep and interval timing in mice. Cell Rep 16:615–621
Bedont JL, LeGates TA, Slat EA, Byerly MS, Wang H, Hu J, Rupp AC, Qian J, Wong GW, Herzog
ED et al (2014) Lhx1 controls terminal differentiation and circadian function of the
suprachiasmatic nucleus. Cell Rep 7:609–622
Bedont JL, LeGates TA, Buhr E, Bathini A, Ling JP, Bell B, Wu MN, Wong PC, Van Gelder RN,
Mongrain V et al (2017) An LHX1-regulated transcriptional network controls sleep/wake
coupling and thermal resistance of the central circadian clockworks. Curr Biol 27:128–136
Biehl MJ, Kaylan KB, Thompson RJ, Gonzalez RV, Weis KE, Underhill GH, Raetzman LT (2018)
Cellular fate decisions in the developing female anteroventral periventricular nucleus are
regulated by canonical Notch signaling. Dev Biol 442:87–100
Blaess S, Szabo N, Haddad-Tovolli R, Zhou X, Alvarez-Bolado G (2014) Sonic hedgehog
signaling in the development of the mouse hypothalamus. Front Neuroanat 8:156
Blumberg MS, Gall AJ, Todd WD (2014) The development of sleep-wake rhythms and the search
for elemental circuits in the infant brain. Behav Neurosci 128:250–263
Borbely AA (1982) A two process model of sleep regulation. Hum Neurobiol 1:195–204
124 D. W. T. Kim and S. Blackshaw

Brunner DP, Dijk DJ, Tobler I, Borbely AA (1990) Effect of partial sleep deprivation on sleep
stages and EEG power spectra: evidence for non-REM and REM sleep homeostasis.
Electroencephalogr Clin Neurophysiol 75:492–499
Buhr ED, Yoo SH, Takahashi JS (2010) Temperature as a universal resetting cue for mammalian
circadian oscillators. Science 330:379–385
Campbell JN, Macosko EZ, Fenselau H, Pers TH, Lyubetskaya A, Tenen D, Goldman M,
Verstegen AM, Resch JM, McCarroll SA et al (2017) A molecular census of arcuate hypothala-
mus and median eminence cell types. Nat Neurosci 20:484–496
Carmona-Alcocer V, Abel JH, Sun TC, Petzold LR, Doyle FJ 3rd, Simms CL, Herzog ED (2018)
Ontogeny of circadian rhythms and synchrony in the suprachiasmatic nucleus. J Neurosci
38:1326–1334
Chen R, Wu X, Jiang L, Zhang Y (2017) Single-cell RNA-Seq reveals hypothalamic cell diversity.
Cell Rep 18:3227–3241
Chen KS, Xu M, Zhang Z, Chang WC, Gaj T, Schaffer DV, Dan Y (2018) A hypothalamic switch
for REM and non-REM sleep. Neuron 97(1168–1176):e1164
Chung S, Weber F, Zhong P, Tan CL, Nguyen TN, Beier KT, Hormann N, Chang WC, Zhang Z,
Do JP et al (2017) Identification of preoptic sleep neurons using retrograde labelling and gene
profiling. Nature 545:477–481
Clark DD, Gorman MR, Hatori M, Meadows JD, Panda S, Mellon PL (2013) Aberrant development
of the suprachiasmatic nucleus and circadian rhythms in mice lacking the homeodomain protein
Six6. J Biol Rhythm 28:15–25
Daan S, Beersma DG, Borbely AA (1984) Timing of human sleep: recovery process gated by a
circadian pacemaker. Am J Phys 246:R161–R183
Dalal J, Roh JH, Maloney SE, Akuffo A, Shah S, Yuan H, Wamsley B, Jones WB, de Guzman
Strong C, Gray PA et al (2013) Translational profiling of hypocretin neurons identifies candidate
molecules for sleep regulation. Genes Dev 27:565–578
Daszuta A, Gambarelli F (1985) Early postnatal development of EEG and sleep-waking cycle in
two inbred mouse strains. Brain Res 354:39–47
Davis KF, Parker KP, Montgomery GL (2004) Sleep in infants and young children: part one:
normal sleep. J Pediatr Health Care 18:65–71
de Melo J, Zibetti C, Clark BS, Hwang W, Miranda-Angulo AL, Qian J, Blackshaw S (2016) Lhx2
is an essential factor for retinal gliogenesis and Notch signaling. J Neurosci 36:2391–2405
Dittrich L, Morairty SR, Warrier DR, Kilduff TS (2015) Homeostatic sleep pressure is the primary
factor for activation of cortical nNOS/NK1 neurons. Neuropsychopharmacology 40:632–639
Dzierzewski JM, Dautovich N, Ravyts S (2018) Sleep and cognition in older adults. Sleep Med Clin
13:93–106
Eldred KC, Hadyniak SE, Hussey KA, Brenerman B, Zhang PW, Chamling X, Sluch VM, Welsbie
DS, Hattar S, Taylor J et al (2018) Thyroid hormone signaling specifies cone subtypes in human
retinal organoids. Science 362:eaau6348
Elmenhorst D, Meyer PT, Winz OH, Matusch A, Ermert J, Coenen HH, Basheer R, Haas HL,
Zilles K, Bauer A (2007) Sleep deprivation increases A1 adenosine receptor binding in the
human brain: a positron emission tomography study. J Neurosci 27:2410–2415
Elmenhorst D, Elmenhorst EM, Hennecke E, Kroll T, Matusch A, Aeschbach D, Bauer A (2017)
Recovery sleep after extended wakefulness restores elevated A1 adenosine receptor availability
in the human brain. Proc Natl Acad Sci USA 114:4243–4248
Ferreira JGP, Bittencourt JC, Adamantidis A (2017) Melanin-concentrating hormone and sleep.
Curr Opin Neurobiol 44:152–158
Fischer D, Lombardi DA, Marucci-Wellman H, Roenneberg T (2017) Chronotypes in the US—
influence of age and sex. PLoS One 12:e0178782
Foreman SW, Thomas KA, Blackburn ST (2008) Individual and gender differences matter in
preterm infant state development. J Obstet Gynecol Neonatal Nurs 37:657–665
Frank MG, Heller HC (1997) Development of diurnal organization of EEG slow-wave activity and
slow-wave sleep in the rat. Am J Phys 273:R472–R478
5 Winding the Clock: Development of Hypothalamic Structures Controlling. . . 125

Fu T, Towers M, Placzek MA (2017) Fgf10(+) progenitors give rise to the chick hypothalamus by
rostral and caudal growth and differentiation. Development 144:3278–3288
Gall AJ, Todd WD, Ray B, Coleman CM, Blumberg MS (2008) The development of day-night
differences in sleep and wakefulness in Norway rats and the effect of bilateral enucleation. J Biol
Rhythm 23:232–241
Galland BC, Taylor BJ, Elder DE, Herbison P (2012) Normal sleep patterns in infants and children:
a systematic review of observational studies. Sleep Med Rev 16:213–222
Hanne HM, Meadows JD, Trang C, Hereford B, Bharti K, Gorman MR, Mellon PL (2016) Deletion
of SIX3 or VAX1 in the SCN impairs circadian rhythms and fertility. In Endocrine Society’s
98th Annual Meeting and Expo, April 1–4, 2016, Boston
Hatori M, Gill S, Mure LS, Goulding M, O’Leary DD, Panda S (2014) Lhx1 maintains synchrony
among circadian oscillator neurons of the SCN. Elife 3:e03357
Hattar S, Liao HW, Takao M, Berson DM, Yau KW (2002) Melanopsin-containing retinal ganglion
cells: architecture, projections, and intrinsic photosensitivity. Science 295:1065–1070
Hrvatin S, Hochbaum DR, Nagy MA, Cicconet M, Robertson K, Cheadle L, Zilionis R, Ratner A,
Borges-Monroy R, Klein AM et al (2018) Single-cell analysis of experience-dependent
transcriptomic states in the mouse visual cortex. Nat Neurosci 21:120–129
Joseph D, Chong NW, Shanks ME, Rosato E, Taub NA, Petersen SA, Symonds ME, Whitehouse
WP, Wailoo M (2015) Getting rhythm: how do babies do it? Arch Dis Child Fetal Neonatal Ed
100:F50–F54
Jud C, Schmutz I, Hampp G, Oster H, Albrecht U (2005) A guideline for analyzing circadian wheel-
running behavior in rodents under different lighting conditions. Biol Proced Online 7:101–116
Lee JE, Wu SF, Goering LM, Dorsky RI (2006) Canonical Wnt signaling through Lef1 is required
for hypothalamic neurogenesis. Development 133:4451–4461
Liu J, Merkle FT, Gandhi AV, Gagnon JA, Woods IG, Chiu CN, Shimogori T, Schier AF, Prober
DA (2015) Evolutionarily conserved regulation of hypocretin neuron specification by Lhx9.
Development 142:1113–1124
Liu K, Kim J, Kim DW, Zhang YS, Bao H, Denaxa M, Lim SA, Kim E, Liu C, Wickersham IR et al
(2017) Lhx6-positive GABA-releasing neurons of the zona incerta promote sleep. Nature
548:582–587
Manning L, Ohyama K, Saeger B, Hatano O, Wilson SA, Logan M, Placzek M (2006) Regional
morphogenesis in the hypothalamus: a BMP-Tbx2 pathway coordinates fate and proliferation
through Shh downregulation. Dev Cell 11:873–885
Mariani J, Coppola G, Zhang P, Abyzov A, Provini L, Tomasini L, Amenduni M, Szekely A,
Palejev D, Wilson M et al (2015) FOXG1-dependent dysregulation of GABA/glutamate neuron
differentiation in autism spectrum disorders. Cell 162:375–390
Martinez-Jimenez CP, Eling N, Chen HC, Vallejos CA, Kolodziejczyk AA, Connor F, Stojic L,
Rayner TF, Stubbington MJT, Teichmann SA et al (2017) Aging increases cell-to-cell tran-
scriptional variability upon immune stimulation. Science 355:1433–1436
Maywood ES, Chesham JE, O’Brien JA, Hastings MH (2011) A diversity of paracrine signals
sustains molecular circadian cycling in suprachiasmatic nucleus circuits. Proc Natl Acad Sci
USA 108:14306–14311
Merkle FT, Maroof A, Wataya T, Sasai Y, Studer L, Eggan K, Schier AF (2015) Generation of
neuropeptidergic hypothalamic neurons from human pluripotent stem cells. Development
142:633–643
Miles LE, Dement WC (1980) Sleep and aging. Sleep 3:1–220
Mohawk JA, Takahashi JS (2011) Cell autonomy and synchrony of suprachiasmatic nucleus
circadian oscillators. Trends Neurosci 34(7):349–358
Morairty SR, Dittrich L, Pasumarthi RK, Valladao D, Heiss JE, Gerashchenko D, Kilduff TS (2013)
A role for cortical nNOS/NK1 neurons in coupling homeostatic sleep drive to EEG slow wave
activity. Proc Natl Acad Sci USA 110:20272–20277
Nelson AB, Faraguna U, Zoltan JT, Tononi G, Cirelli C (2013) Sleep patterns and homeostatic
mechanisms in adolescent mice. Brain Sci 3:318–343
126 D. W. T. Kim and S. Blackshaw

Newman EA, Kim DW, Wan J, Wang J, Qian J, Blackshaw S (2018a) Foxd1 is required for
terminal differentiation of anterior hypothalamic neuronal subtypes. Dev Biol 439:102–111
Newman EA, Wu D, Taketo MM, Zhang J, Blackshaw S (2018b) Canonical Wnt signaling
regulates patterning, differentiation and nucleogenesis in mouse hypothalamus and prethalamus.
Dev Biol 442:236–248
Ogawa K, Suga H, Ozone C, Sakakibara M, Yamada T, Kano M, Mitsumoto K, Kasai T, Kodani Y,
Nagasaki H et al (2018) Vasopressin-secreting neurons derived from human embryonic stem
cells through specific induction of dorsal hypothalamic progenitors. Sci Rep 8:3615
Ohyama K, Das R, Placzek M (2008) Temporal progression of hypothalamic patterning by a dual
action of BMP. Development 135:3325–3331
Parsons MJ, Brancaccio M, Sethi S, Maywood ES, Satija R, Edwards JK, Jagannath A, Couch Y,
Finelli MJ, Smyllie NJ et al (2015) The regulatory factor ZFHX3 modifies circadian function in
SCN via an AT motif-driven Axis. Cell 162:607–621
Partch CL, Green CB, Takahashi JS (2014) Molecular architecture of the mammalian circadian
clock. Trends Cell Biol 24:90–99
Pedersen NP, Ferrari L, Venner A, Wang JL, Abbott SBG, Vujovic N, Arrigoni E, Saper CB, Fuller
PM (2017) Supramammillary glutamate neurons are a key node of the arousal system. Nat
Commun 8:1405
Porter FD, Drago J, Xu Y, Cheema SS, Wassif C, Huang SP, Lee E, Grinberg A, Massalas JS,
Bodine D et al (1997) Lhx2, a LIM homeobox gene, is required for eye, forebrain, and definitive
erythrocyte development. Development 124:2935–2944
Qian X, Nguyen HN, Song MM, Hadiono C, Ogden SC, Hammack C, Yao B, Hamersky GR,
Jacob F, Zhong C et al (2016) Brain-region-specific organoids using mini-bioreactors for
modeling ZIKV exposure. Cell 165:1238–1254
Qian X, Jacob F, Song MM, Nguyen HN, Song H, Ming GL (2018) Generation of human brain
region-specific organoids using a miniaturized spinning bioreactor. Nat Protoc 13:565–580
Reddy AB, O’Neill JS (2009) Healthy clocks, healthy body, healthy mind. Trends Cell Biol
20:36–44
Roenneberg T, Wirz-Justice A, Merrow M (2003) Life between clocks: daily temporal patterns of
human chronotypes. J Biol Rhythm 18:80–90
Romanov RA, Zeisel A, Bakker J, Girach F, Hellysaz A, Tomer R, Alpar A, Mulder J, Clotman F,
Keimpema E et al (2017) Molecular interrogation of hypothalamic organization reveals distinct
dopamine neuronal subtypes. Nat Neurosci 20:176–188
Roy A, de Melo J, Chaturvedi D, Thein T, Cabrera-Socorro A, Houart C, Meyer G, Blackshaw S,
Tole S (2013) LHX2 is necessary for the maintenance of optic identity and for the progression of
optic morphogenesis. J Neurosci 33:6877–6884
Sack RL, Auckley D, Auger RR, Carskadon MA, Wright KP Jr, Vitiello MV, Zhdanova IV,
American Academy of Sleep, M (2007) Circadian rhythm sleep disorders: part I, basic
principles, shift work and jet lag disorders. An American Academy of Sleep Medicine review.
Sleep 30:1460–1483
Saito YC, Tsujino N, Hasegawa E, Akashi K, Abe M, Mieda M, Sakimura K, Sakurai T (2013)
GABAergic neurons in the preoptic area send direct inhibitory projections to orexin neurons.
Front Neural Circuits 7:192
Salvatierra J, Lee DA, Zibetti C, Duran-Moreno M, Yoo S, Newman EA, Wang H, Bedont JL, de
Melo J, Miranda-Angulo AL et al (2014) The LIM homeodomain factor Lhx2 is required for
hypothalamic Tanycyte specification and differentiation. J Neurosci 34:16809–16820
Saper CB, Fuller PM, Pedersen NP, Lu J, Scammell TE (2010) Sleep state switching. Neuron
68:1023–1042
Scammell TE, Arrigoni E, Lipton JO (2017) Neural circuitry of wakefulness and sleep. Neuron
93:747–765
Shimogori T, Lee DA, Miranda-Angulo A, Yang Y, Wang H, Jiang L, Yoshida AC, Kataoka A,
Mashiko H, Avetisyan M et al (2010) A genomic atlas of mouse hypothalamic development.
Nat Neurosci 13:767–775
5 Winding the Clock: Development of Hypothalamic Structures Controlling. . . 127

VanDunk C, Hunter LA, Gray PA (2011) Development, maturation, and necessity of transcription
factors in the mouse suprachiasmatic nucleus. J Neurosci 31:6457–6467
Venner A, Anaclet C, Broadhurst RY, Saper CB, Fuller PM (2016) A novel population of wake-
promoting GABAergic neurons in the ventral lateral hypothalamus. Curr Biol 26:2137–2143
Wang L, Meece K, Williams DJ, Lo KA, Zimmer M, Heinrich G, Martin Carli J, Leduc CA, Sun L,
Zeltser LM et al (2015) Differentiation of hypothalamic-like neurons from human pluripotent
stem cells. J Clin Invest 125:796–808
Wataya T, Ando S, Muguruma K, Ikeda H, Watanabe K, Eiraku M, Kawada M, Takahashi J,
Hashimoto N, Sasai Y (2008) Minimization of exogenous signals in ES cell culture induces
rostral hypothalamic differentiation. Proc Natl Acad Sci USA 105:11796–11801
Weber F, Dan Y (2016) Circuit-based interrogation of sleep control. Nature 538:51–59
Wilcox AG, Vizor L, Parsons MJ, Banks G, Nolan PM (2017) Inducible knockout of mouse Zfhx3
emphasizes its key role in setting the pace and amplitude of the adult circadian clock. J Biol
Rhythm 32:433–443
Wolf A, Ryu S (2013) Specification of posterior hypothalamic neurons requires coordinated
activities of Fezf2, Otp, Sim1a and Foxb1.2. Development 140:1762–1773
Xie Y, Kaufmann D, Moulton MJ, Panahi S, Gaynes JA, Watters HN, Zhou D, Xue HH, Fung CM,
Levine EM et al (2017) Lef1-dependent hypothalamic neurogenesis inhibits anxiety. PLoS Biol
15:e2002257
Yamaguchi Y, Suzuki T, Mizoro Y, Kori H, Okada K, Chen Y, Fustin JM, Yamazaki F,
Mizuguchi N, Zhang J et al (2013) Mice genetically deficient in vasopressin V1a and V1b
receptors are resistant to jet lag. Science 342:85–90
Yao L, Liu Y, Qiu Z, Kumar S, Curran JE, Blangero J, Chen Y, Lehman DM (2017) Molecular
profiling of human induced pluripotent stem cell-derived hypothalamic neurones provides
developmental insights into genetic loci for body weight regulation. J Neuroendocrinol 29(2)
Pituitary Development and Organogenesis:
Transcription Factors in Development 6
and Disease

Alexandre Z. Daly and Sally A. Camper

Abstract
Many transcription factors that regulate pituitary gland development and specifi-
cation of hormone-producing cell types have been discovered. Here we focus on
13 transcription factors that play key roles in these processes, many of which are
mutated in patients with congenital pituitary insufficiency disorders. We present
the methods by which each of these transcription factors were identified. Each of
the genes we discuss has been well studied in genetically engineered mice,
facilitating an understanding of the impact of each single gene defect on the
expression of downstream targets. Finally, we highlight technical advances that
promise to yield future insights into epigenetic regulation of cell fate.

Keywords
Cell specification · Stem cell · Hypopituitarism

6.1 Introduction

The pituitary gland is located at the base of the brain, and it is a small organ,
approximately the size of a pea in humans. It is often referred to as the master
gland because it is a key regulator of multiple organ systems, controlling growth,
fertility, stress response, and homeostasis. The pituitary gland develops between
5 and 9 weeks of gestation in humans and embryonic day 10–19 in mice (Xue et al.
2013). An invagination of oral ectoderm, called Rathke’s pouch, forms the anterior
and intermediate lobes of the pituitary gland, and evagination of the neural ectoderm
forms the posterior lobe of the pituitary gland and pituitary stalk. The posterior lobe
contains the axon terminals for oxytocin and vasopressin and pituicytes. The anterior

A. Z. Daly · S. A. Camper (*)

Department of Human Genetics, University of Michigan, Ann Arbor, MI, USA
e-mail: [email protected]

# Springer Nature Switzerland AG 2020 129

S. Wray, S. Blackshaw (eds.), Developmental Neuroendocrinology, Masterclass in
Neuroendocrinology 9, https://doi.org/10.1007/978-3-030-40002-6_6
130 A. Z. Daly and S. A. Camper

pineal
gland
cerebellum
cortex

midbrain
thalamus
olfactory
blub hypothalamus medulla
pons
pituitary
gland

supraoptic
nucleus
hypothalamus
paraventricular
nucleus
optic
chiasm

posterior lobe
oxytocin,
pituitary vasopressin
stalk

pituitary gland anterior lobe

GH, TSH, PRL
ACTH intermediate lobe
LH, FSH MSH,
-endorphin

Fig. 6.1 Anatomy of the pituitary gland and base of the hypothalamus. The pituitary gland
contains anterior (ant) and intermediate (int) lobes derived from the oral ectoderm that forms
Rathke’s pouch. The posterior lobe (post) and pituitary stalk are derived from neural ectoderm of
the ventral diencephalon, and they contain the axon terminals of vasopressin and oxytocin neurons
that project from both the supraoptic nucleus and the paraventricular nucleus of the hypothalamus.
Hypothalamic peptides regulate the release of anterior pituitary hormones into the hypophyseal
portal system [adapted from Rappaport and Amselem (2001)]

lobe is composed of cells specialized in the secretion of polypeptide hormones

(Fig. 6.1). Hypothalamic neurons secrete factors into the hypophyseal portal system
that regulate pituitary function. In the past three decades, a great deal of progress has
been made in understanding the mechanisms that drive the differentiation of
hormone-producing cells of the anterior pituitary gland during development. Multi-
ple transcription factors and signaling pathways are involved, and defects in the
genes that encode these factors and pathways are an important contributor to
congenital pituitary hormone deficiency (Castinetti et al. 2016; Fang et al. 2016;
Kelberman et al. 2009; Rizzoti 2015). The focus of this review will be on the
6 Pituitary Development and Organogenesis: Transcription Factors in. . . 131

pituitary transcription factors that drive the differentiation of pituitary hormone-

producing cells with an emphasis on the history of discovery for those that are most
relevant to human pituitary insufficiency. Other chapters in this volume review the
development of the hypothalamus and its regulation of pituitary function (See the
reviews by Placzek and Towers on the development of the hypothalamus, the review
from Alvarez-Bolado on SHH and GLI in the hypothalamus and Lee’s review on the
arcuate nucleus.)

6.1.1 The Hypophyseal Portal System, Pituitary Vasculature,

and Accessory Cells of the Pituitary Gland

The hypophyseal portal system refers to the blood vessels that provide connections
between the hypothalamus and the anterior pituitary gland. This vasculature
develops e14.5–e18.5 in mice and appears complete by 12 weeks of gestation in
humans (Xue et al. 2013). The capillaries are highly fenestrated, facilitating rapid
molecular exchange between the hypothalamus and pituitary lobes. Oxygen levels
and blood flow are regulated and have an effect on pituitary hormone production and
secretion (Lafont et al. 2010; Schaeffer et al. 2010).
The mature pituitary gland has six distinct hormone-producing cell types, as well
as folliculo-stellate cells and pituicytes (Fig. 6.2). The hormone-producing cell types
in the anterior lobe include corticotropes that produce adrenocorticotropin (ACTH)
from the pro-hormone pro-opiomelanocortin (POMC), somatotropes that produce
GH, lactotropes that produce prolactin, gonadotropes that produce the heterodimeric
glycoprotein hormones luteinizing hormone (LH) and follicle-stimulating hormone
(FSH) and thyrotropes that produce the heterodimeric glycoprotein hormone thy-
roid-stimulating hormone (TSH). The intermediate lobe, which remains distinct in
rodents, contains melanotropes that produce melanocyte-stimulating hormone from
POMC. Folliculo-stellate (FS) cells are non-endocrine cells that were named because
of the star-like morphology of cytoplasmic processes (Rinehart and Farquhar 1953).
FS cells appear between postnatal days 10 and 20 in the rat, throughout the anterior
lobe parenchyma, directly adjacent to the hormone-producing cells (Soji et al. 1997).
FS cells serve a support function for hormone-producing cells through the release of
cytokines and growth factors. FS cells are excitatory and are involved in the
coordinated, pulsatile release of hormones from the hormone-producing cells,
which themselves form homotypic networks (Fauquier et al. 2001; Mollard et al.
2012). Pituicytes are glial-like cells located within the posterior lobe. The larger
parenchymatous pituicytes regulate hormone output by completely enveloping the
neurosecretory processes of the oxytocin and vasopressin axons when hormone
secretion is low, and they recede as hormone secretion increases. They also regulate
hormone output by secretion of taurine (Rosso and Mienville 2009). The smaller
fibrous pituicytes do not appear to be involved in this regulatory function and are less
well understood (Pow et al. 1989).
The vasculature is important for pituitary gland function because it transports
molecules from pituitary target organs that feedback at the level of the pituitary gland
132 A. Z. Daly and S. A. Camper

Corticotrope
Melanotrope
TPit Pax7

Somatotrope

Prop1 Pou1f1

Lactotrope

SOX2 Thyrotrope
stem cell

SF1

Gonadotrope
pituitary population
Fraction of total

Proliferating
Precursors
Quiescent
Precursors
Differentiated
Cells
E9.5 E11.5 E13.5 E17.5

Fig. 6.2 Critical transcription factors control the development of specialized hormone-producing
cells. The major hormone-producing cell types of the anterior and intermediate lobes of the pituitary
gland are derived from SOX2-expressing progenitor cells. Combinations of transcription factors
drive specific cell fates and antagonize differentiation into alternate fates. Each cell type has
distinctive secretory granules and shape (Matsumoto and Ishii 1992). Proliferating precursors
leave the cell cycle and differentiate into hormone-producing cells during gestation

and hypothalamus, maintaining homeostasis. This multi-organ feedback regulation is

called an axis. The best known of these axes are the hypothalamic–pituitary–gonadal
axis, hypothalamic–pituitary–adrenal axis, and the hypothalamic–pituitary–thyroid
axis, although axes involving bone, liver, and other organs also exist (Fekete and
Lechan 2014; Imam et al. 2009; Jin and Yang 2014; Keller-Wood 2015; Simon et al.
2002).

6.1.2 Cell Signaling During Pituitary Gland Development

Some of the major signaling pathways that have been implicated in pituitary devel-
opment are SHH, FGF, BMP, Wnt, Notch, Hippo, EGF, and retinoic acid (Cohen
et al. 1999; Djakoure et al. 1996; Lodge et al. 2019; Zhu et al. 2007). Hypothalamic
SHH is necessary for inducing expression of the pituitary transcription factors LHX3
and LHX4, which drive progenitor proliferation and prevent cell death (Carreno et al.
6 Pituitary Development and Organogenesis: Transcription Factors in. . . 133

2017). FGF8 and FGFR1 mutations cause pituitary hormone deficiency in humans
(Fang et al. 2016), and mouse studies show reduced cell proliferation and enhanced
cell death in the pituitary primordia of Fgf8, Fgf10, and Fgfr2 mutants (Ohuchi et al.
2000). BMP and the antagonist noggin affect pituitary growth and shape during
development, and altered BMP signaling is associated with pituitary adenomas
(Davis and Camper 2007; Ericson et al. 1998; Giacomini et al. 2006; Treier et al.
2001). Similarly, WNT signaling regulates pituitary organogenesis and acts in a
paracrine manner to stimulate excess cell proliferation in adenomas (Chambers
et al. 2013). The ligand, beta-catenin, is a cofactor for several key pituitary transcrip-
tion factors including PITX2, TCF7L2 (TCF4), LEF1, NR5A1, and PROP1. Notch
signaling regulates the timing of pituitary progenitor cell cycle exit and differentiation
(Cheung et al. 2018a; Raetzman et al. 2004; Zhu et al. 2006). YAP and TAZ are
transcriptional regulators downstream of the Hippo signaling pathway, which
suppresses their function by phosphorylation. This pathway plays an essential role
in the regulation of pituitary progenitor expansion and normal organ size (Lodge et al.
2019). There is a great deal of cross talk between these pathways, in that perturbation
of one signaling pathway typically affects the other signaling pathways. EGF and
retinoic acid signaling are recognized as important signaling pathways but their
precise roles during organogenesis have not been established (Rhodes et al. 1993;
Roh et al. 2001).

6.1.3 Hypopituitarism

The major pathologies associated with the pituitary are adenomas, which can be
associated with either over- or underproduction of hormones, and congenital or
acquired hypopituitarism (Higham et al. 2016; Melmed 2011). The main emphasis
of this review will be on congenital hypopituitarism, which had an estimated
prevalence of 45.5 per 100,000 people (Regal et al. 2001). This is a genetically
heterogeneous condition that can present with a single pituitary hormone deficiency
or with multiple deficiencies. About 50% of cases that originally present with
isolated growth hormone deficiency (IGHD) progress to combined pituitary hor-
mone deficiency (CPHD). CPHD is defined as a reduction of at least two pituitary
hormones, and usually, GH is one of them. It has a prevalence of 1 in 8000
individuals, and more than 30 genes have been implicated as causal factors for
CPHD (Fang et al. 2016).

6.1.4 Overview

Here we review the transcription factors that are critical for developing the specialized
cells of the anterior and intermediate lobes of the pituitary gland, emphasizing the
history of scientific discovery surrounding each factor, from its discovery through to
its most recent characterization. This reveals the continuing evolution and diversity of
effective technical approaches. These include identification of cis-acting sequences
134 A. Z. Daly and S. A. Camper

important for regulation of hormone gene expression and identification of the trans-
acting factors that bind those sites, positional cloning of mutations in mice and human
patients, and exploration of epigenetic regulation of chromatin accessibility. We also
highlight areas for future discovery that may take advantage of high-throughput
sequencing and high-efficiency, targeted germline disruption tools.

6.2 SOX2 (LCC, YSB)

6.2.1 Gene Discovery

Sox2 was first discovered in a whole embryo cDNA screen in search of genes similar
to the male, sex-determining gene Sry, which binds DNA through a motif or box
characteristic of high mobility group proteins (Gubbay et al. 1990). Sox2 is
expressed in progenitor cell populations in many tissues, and its expression is
negatively correlated with differentiation (Avilion et al. 2003; Li et al. 1998;
Zappone et al. 2000). SOX2 is one of the factors that, along with OCT3/4,
c-MYC, and KLF4, were demonstrated by Yamanaka and colleagues to be sufficient
to induce pluripotency in differentiated cells from mice and humans (Takahashi et al.
2007; Takahashi and Yamanaka 2006).

6.2.2 Knockout Mouse Phenotype, Downstream Targets,

and Interacting Factors

Sox2 is expressed in the developing pituitary and hypothalamus, and it has a role in
pituitary development (Takahashi et al. 2007; Takahashi and Yamanaka 2006).
Embryos homozygous for Sox2 loss of function alleles die around implantation,
and heterozygous mice have genetic background-dependent reduction in body size
and male infertility (Avilion et al. 2003). The developing pituitaries of heterozygous
mutants are bifurcated, similar to the dysmorphology observed in Wnt5a mutants,
and both males and females exhibit reduced differentiation into somatotropes and
gonadotropes (Fig. 6.3) (Cha et al. 2004; Rizzoti and Lovell-Badge 2005). About
two-thirds of the heterozygotes survive to adulthood, and in these survivors, the
reduction in pituitary size is modest, and GH and LH content is only significantly
reduced in males.
Martinez-Barbera and colleagues generated a conditional deletion of Sox2 using
the Hesx1-cre strain, deleting Sox2 in the pituitary gland and areas of the brain
between e10.5 and e12.5 (Rizzoti and Lovell-Badge 2005). This resulted in severe
pituitary hypoplasia and greatly reduced the expressions of Pou1f1, Gh, and Tshb.
Gonadotropin deficiency appeared to be secondary to the reduction in GnRH
neurons in these mice. Cell proliferation is significantly decreased in the condition-
ally deleted Sox2 pituitary glands, and it appears that the progenitors have not
migrated from the progenitor zone of Rathke’s pouch to colonize the anterior lobe.
This suggests that Sox2 has an important role in pituitary progenitor proliferation and
transition to differentiation.
6 Pituitary Development and Organogenesis: Transcription Factors in. . . 135

Fig. 6.3 The effect of transcription factor mutations on pituitary development has been revealed
using genetically engineered and spontaneous mutant mice. Deletion of Pitx1 has modest effects
including slight reduction of TSH and LH, slight increase in POMC and no change in GH and
αGSU, as observed by in situ hybridization in P0 animals (Szeto et al. 1999). In situ hybridization in
136 A. Z. Daly and S. A. Camper

Robinson and colleagues discovered that Sox2-expressing cells in the adult mouse
pituitary were still mitotically active and that, given appropriate culture conditions,
these SOX2-positive cells can self-renew and differentiate in vitro into all of the
hormone-producing cells of the anterior lobe of the pituitary gland (Fauquier et al.
2008). Both Lovell-Badge and Martinez-Barbera and colleagues used lineage tracing
to demonstrate that SOX2-positive cells give rise to all of the different hormone-
producing cells in the pituitary gland in vivo (Andoniadou et al. 2013; Rizzoti et al.
2013). SOX2-positive cells contribute to tissue homeostasis, and they are involved in
responding to end-organ ablation, such as adrenalectomy and gonadectomy. Mice
with a constitutively active allele of β-catenin exhibit clusters of pituitary cells that
overexpress β-catenin, express Sox2, and induce transformation of neighboring cells
that mimic craniopharyngioma, a childhood tumor (Gaston-Massuet et al. 2011).
Taken together these studies established that SOX2-expressing cells represent
progenitors of the hormone-producing cells in the developing pituitary.

6.2.3 Patient Mutations

Most patients with SOX2 mutations have serious eye abnormalities that can include
anophthalmia, microphthalmia, and/or optic nerve hypoplasia, and hypopituitarism
characterized by pituitary gland hypoplasia and hypogonadotropic hypogonadism

Fig. 6.3 (continued) newborn mice homozygous for a hypomorphic allele of Pitx2, Pitx2neo/neo,
revealed normal Pomc expression, reduction in Gh and Tshb expression, and no detectable Lhb or
Fshb transcripts (panels bordered in purple) (Suh et al. 2002). Pitx2
/
embryos had arrested the
development of Rathke’s pouch and reduced expression of αGSU at e13.5 mice (Lin et al. 1999).
Trace amounts of Pomc transcripts were present, but all other hormones were not detectable (not
shown). In situ hybridization in the Lhx3 knockout mice revealed a near-total loss of Pomc, Tshb,
Gh, and Cga (αGSU) transcripts, and immunohistochemistry failed to detect significant amounts of
LH. In situ hybridization of Lhx4 knockout pituitaries revealed a significant reduction in Tshb, Lhb,
Gh, and Cga transcripts at e18.5 (Sheng et al. 1997), and immunohistochemistry showed largely
similar POMC expression in e14.5 mice (Sheng et al. 1997). Immunohistochemistry of Sox2
heterozygous null mice revealed reduced staining for GH and LH at e18.5 (Kelberman et al.
2006). There is no detectable Pou1f1 staining in Hesx cre/+; Sox2flox/flox mice at e14.5 or later in
gestation (not shown) (Jayakody et al. 2012). The staining for αGSU and POMC was somewhat
normal at e15.5, considering the hypoplasia of the organ. Only Pou1f1 independent
TSH-expressing cells were detected. Immunohistochemistry of Prop1
/
pituitaries revealed
POMC expression but little or no detectable caudomedial staining for TSH, LH, or GH at e18.5
(Nasonkin et al. 2004). αGSU staining was modestly reduced at e14.5 in Prop1df/df mice, but
NR5A1 cells were normal at birth (Raetzman et al. 2002). In situ hybridization analysis of e17.5
Hesx1R160C/R160C mutants revealed approximately normal levels of Pomc, Tshb, Lhb, Gh, and Cga
expression, although the pituitaries were dysmorphic (Sajedi et al. 2008). These missense mutant
mice have similar pituitary features to mice homozygous for a null allele, but the forebrain is less
affected. Analysis of newborn SF1-mutant mice revealed undetectable LH, diminished Cga
transcripts and normal immunostaining for ACTH, TSH, and GH (Ingraham et al. 1994).
Pou1f1dw/dw mutant newborns have normal POMC immunostaining, undetectable TSH and GH,
and increased LH (Mortensen et al. 2011). Cga appeared reduced. Immunohistochemical analysis
of Tpit knockouts revealed undetectable POMC, no effect on caudomedial staining of GH, TSH,
LH, or αGSU in the anterior lobe, but ectopic expression of TSH, LH, and αGSU in the intermediate
lobe (Pulichino et al. 2003b).
6 Pituitary Development and Organogenesis: Transcription Factors in. . . 137

(Gaston-Massuet et al. 2011; Gubbay et al. 1990; Kelberman et al. 2006; Ragge et al.
2005). While all of the patients with SOX2 mutations exhibited reduced gonadotrope
function, the degree of GH deficiency varied. The response of patients to GnRH
stimulation was also variable.

6.3 PITX Factors (PITX1: Ptx1, Bft, P-OTX; PITX2: Brx, Munc30,
Otlx2, Rieg)

6.3.1 Gene Discovery

The PITX transcription factors, PITX1 and PITX2, are expressed early in pituitary
gland development and persist through adulthood. Their names derive from their
pituitary expression and paired-like homeobox DNA-binding domains, and they
have important functions in the hind limb, heart, eye, and other tissues (Gage et al.
1999; Kitamura et al. 1999; Lin et al. 1999; Lu et al. 1999; Szeto et al. 1999). These
factors were found in very different ways, and they have overlapping functions
within the pituitary gland (Charles et al. 2005).
Pitx1 was discovered by searching for factors that bind the regulatory elements of
the pro-opiomelanocortin gene (Pomc). POMC is a pro-hormone that is expressed in
both corticotropes of the anterior lobe and melanotropes or the intermediate lobe.
POMC undergoes differential cleavage to produce adrenocorticotropin (ACTH) in
corticotropes and melanocyte-stimulating hormone (MSH) and β-endorphin in
melanotropes. In vitro analysis revealed that the major regulatory region of the Pomc
promoter-proximal region was between 166 and 480 bp upstream of the transcription
start site (Therrien and Drouin 1991). Drouin and colleagues found that deleting a small
section from the middle of this element nearly ablated Pomc expression (Lamonerie
et al. 1996). They screened a cDNA expression library from AtT-20 cells, an
immortalized corticotrope-like cell, for factors that bind this element, and identified
Pitx1, which has high homology with the mouse Otx genes. Pitx1 was identified
independently by screening the same cDNA expression library for factors that interact
with POU1F1 (Szeto et al. 1996), a pituitary transcription factor described below.
PITX2 was discovered as one cause of Rieger syndrome, which was first
described in the 1800s as a genetic disorder characterized by eye malformations,
dental hypoplasia, craniofacial dysmorphism, and umbilical stump abnormalities.
The causal gene was mapped to a 50-kb region on chromosome 4 (Datson et al.
1996). Genomic DNA from CpG islands in this region was used to screen human
craniofacial and fetal brain cDNA libraries, and a contig of RIEG cDNAs was
obtained (Semina et al. 1996). RIEG was most closely related to Pitx1, and it was
later renamed PITX2. Murray and colleagues reported mutations in six families with
Rieger syndrome. Pitx2 was discovered independently by screening an adult mouse
pituitary cDNA library for homeobox genes (Gage and Camper 1997). It was
predicted to be a regulator of anterior structure formation in mice based on its
expression pattern and its location on mouse chromosome 3 within a region of
synteny homology to the region of human 4q25 where Rieger syndrome had been
mapped.
138 A. Z. Daly and S. A. Camper

6.3.2 Gene Expression

The expression patterns of Pitx1 and Pitx2 differ, but there are some regions of
striking overlap. Pitx1 is expressed in the pituitary, intestine, tongue, oral epithelium,
the mandible, salivary glands, duodenum, nasal epithelium, and the hind limb
(Lamonerie et al. 1996; Lanctot et al. 1999; Szeto et al. 1996). Pitx1 is expressed
in the pituitary primordia by embryonic day 9.5 (e9.5) (Treier et al. 1998). Pitx2 is
expressed in the eye, the dental lamina, limb mesenchyme, umbilical vessels, and in
Rathke’s pouch as early as e8.5 (Semina et al. 1996). The early expression of both
PITX factors suggested their importance for pituitary development.

6.3.3 Downstream Targets

PITX1 binds the promoters of several genes that encode pituitary hormones, includ-
ing Cga (chorionic gonadotropin alpha, the common subunit of the heterodimeric
hormones LH, FSH, and TSH), Gh, and Prl, and induces their transcription (Szeto
et al. 1996). PITX1 was also implicated in the transcription of Gnrhr, Pou1f1, Tshb,
Nr5a1, and Lhx3 using cell transfection assays (Tremblay et al. 1998). However,
immunostaining suggests that gonadotropes and thyrotropes have the most abundant
expression of PITX1 and PITX2 in the adult pituitary gland (Charles et al. 2006).
PITX2 activity is enhanced by β-catenin, and the complex regulates Cyclin D2
expression (Kioussi et al. 2002).

6.3.4 Knockout Mice

Mice homozygous for Pitx1 deletion had disrupted hind limb formation and a very
mild pituitary phenotype with a slight reduction of gonadotropes and thyrotropes
(Szeto et al. 1999). All of the major transcription factors were largely unaffected, and
Pomc expression was normal.
By comparison, Pitx2 had a much stronger effect on the developing pituitary.
Four independent Pitx2 knockout mouse lines were published in 1999 (Gage et al.
1999; Kitamura et al. 1999; Lin et al. 1999; Lu et al. 1999). Camper and colleagues
developed several alleles, which included a hypomorphic or reduced function allele,
Pitx2neo, a null allele, Pitx2
, and an inducible null allele (Suh et al. 2002). Various
combinations of these alleles were used to assess dosage-sensitive effects of Pitx2
loss of function, i.e., an allelic series. Pitx2
/
and Pitx2neo/neo embryos die in utero,
at e13.5 and e18.5, respectively. Both exhibit right lung isomerism, and dosage-
sensitive effects on eye, pituitary gland, and heart development. The ventral body
wall of Pitx2
/
embryos fails to close, resulting in the externalization of the heart,
liver, and abdominal organs. Pitx2 heterozygotes accurately model the ocular
features of Rieger syndrome (Chen and Gage 2016).
The initial induction of Rathke’s pouch occurs at e10.5 in Pitx2
/
embryos, but
there are extensive cell death and reduced cell proliferation, resulting in a very small
pituitary primordium at e12.5 (Charles et al. 2005; Gage et al. 1999). Hesx1, Prop1,
6 Pituitary Development and Organogenesis: Transcription Factors in. . . 139

and Lhx4 are not expressed in these mutants, but Lhx3 is activated. Little or no
hormone cell specification is detected (Lin et al. 1999). In contrast to the small
pituitaries in Pitx2
/
and Pitx2neo/
embryos, there was no obvious reduction in the
size of Pitx2neo/neo pituitaries (Suh et al. 2002). Somatotropes and thyrotropes were
reduced, and gonadotropes were ablated. Pitx2neo/neo pituitaries had reduced expres-
sion of the transcription factors Gata2, Egr1, Nr5a1, and Pou1f1.
The overlapping expression patterns of Pitx1 and Pitx2 suggested the possibility
of compensatory function. Indeed, double heterozygotes (Pitx1+/
, Pitx2+/
and
Pitx1+/
, Pitx2+/neo) have very poor viability, and hypoplastic pituitaries (Charles
et al. 2005). Lhx3 expression was not detectable in rare double mutants, suggesting
that Pitx genes are required for Lhx3 expression.

6.3.5 Human Mutations

Heterozygous PITX1 loss of function mutations have been reported in two families
with lower extremity abnormalities (Alvarado et al. 2011; Gurnett et al. 2008;
Klopocki et al. 2012). Genomic rearrangements that result in ectopic expression of
PITX1 in the developing forelimbs cause Liebenberg syndrome, a homeotic trans-
formation of the arms to a leg morphology (Kragesteen et al. 2019; Spielmann et al.
2012).
Rieger syndrome is a rare, autosomal dominant disorder affecting the develop-
ment of the eyes, teeth, and abdominal wall. It can be caused by mutations in PITX2
or FOXC1. The majority of patients with Rieger syndrome have normal height,
although there are some anecdotal reports of Rieger syndrome patients with growth
insufficiency. For example, a three-generation family with Rieger syndrome and
either short stature or growth hormone deficiency was reported (Sadeghi-Nejad and
Senior 1974). The proband responded to GH replacement therapy. The nature of the
mutation in this family and other historical cases is unknown. Many Rieger syn-
drome patients have been screened for PITX2 mutations, but no systematic screening
of CPHD patients without eye defects has been reported.

6.4 LIM Homeodomain Proteins: LHX2 (ap, apterous, Lh-2,

LH2A), LHX3 (P-LIM, Lim3), and LHX4 (Gsh-4)

6.4.1 Gene Discovery

LHX proteins contain both LIM-domains, composed of two contiguous zinc fingers
and a Homeobox-domain. Three LHX factors involved in pituitary development,
Lhx2, Lhx3, and Lhx4, and they were discovered independently. Lhx2 was discov-
ered in a screen of cell lines for VDJ recombinase activity (Xu et al. 1993). Lhx3 was
discovered in a screen for Lhx factors in a cDNA library generated from Xenopus
gastrula (called Xlim3 in Xenopus) (Taira et al. 1992). Lhx4 (previously known as
Gsh-4) was discovered in a screen for homeobox genes in mouse genomic DNA
(Singh et al. 1991). Lhx2 is expressed in developing B cells, telencephalon, retina,
140 A. Z. Daly and S. A. Camper

ventral diencephalon, posterior pituitary gland, and infundibulum (Porter et al.

1997). Lhx3 is expressed in the pituitary and pineal glands, the retina, spinal cord,
and hindbrain. Lhx4 is expressed in the developing pituitary, neural tube, hindbrain,
and spinal cord (Li et al. 1994; Sheng et al. 1997).

6.4.2 Knockout Mice

Mouse knockouts of Lhx2, Lhx3, and Lhx4 revealed the importance of each gene for
pituitary development (Ellsworth et al. 2008; Li et al. 1994; Raetzman et al. 2002;
Sheng et al. 1996, 1997; Zhao et al. 2010).
Lhx2 is expressed in the developing ventral diencephalon from e9.5 to e12.5 in
regions that become the pituitary infundibulum and posterior lobe (Zhao et al. 2010).
Embryos homozygous for Lhx2 null alleles have pituitary dysmorphology including
bifurcation of Rathke’s pouch, excess proliferation in the infundibular area, and
failure of the neural ectoderm to evaginate. Despite this, all hormone-producing cell
types of the anterior lobe are detectable, consistent with the expression of critical
signaling molecules such as BMP and FGFs by the abnormal infundibular structure.
The mice are not viable and also have anophthalmia, anemia, and malformations of
the cortex (Porter et al. 1997). A cohort of 59 patients with pituitary deficiency and
eye abnormalities were negative for LHX2 mutations (Perez et al. 2012).
Lhx3 is expressed in the pituitary placode at e9.5 and persists through adulthood
(Sheng et al. 1996). Lhx3+/
mice are phenotypically normal and fertile, but mice
homozygous for the null allele were either stillborn or died within 24 h after birth.
Rathke’s pouch formed in Lhx3
/
embryos but failed to expand. The thin, under-
developed pouch fails to separate from the oral ectoderm. These mutant pouches
exhibited reduced cell proliferation and enhanced cell death, and there is evidence of
abnormal dorsal–ventral polarity (Ellsworth et al. 2008). The cyclin-dependent
kinase inhibitor Cdkn1a (p21) is expressed more dorsally than normal, and
Cdkn1c (p57) is absent in the caudal region of the pouch (Gergics et al. 2015).
Hesx1 and Isl1 expression were initiated at e10.5 but failed to be sustained at e12.5.
Induction of lineage-specific transcription factor gene expression—Tbx19, Neurod1,
Nr5a1, and Pou1f1—were very poor. Lhx3
/
embryos contained a few
differentiated corticotropes but no other differentiated hormone-producing cells
were detected.
Lhx4 is expressed in the pituitary primordium between e9.5 and e12.5, and it is
restricted to the prospective anterior lobe at e12.5 (Sheng et al. 1997). Expression
diminishes at e15.5, but it is detectable in adult pituitary anterior and intermediate
lobes. Lhx4
/
newborns appear normal but die shortly after birth due to failure of
the lungs to inflate (Li et al. 1994). Dexamethasone treatment, used to accelerate
lung development, improved but did not completely rescue survival. Pituitary
insufficiency can result in poor lung development in certain genetic backgrounds
(Nasonkin et al. 2004). Lhx4
/
embryos have anterior pituitary lobe hypoplasia
(Sheng et al. 1997). Enhanced cell death is evident at e12.5, and Lhx3 expression is
significantly delayed (Gergics et al. 2015; Raetzman et al. 2002). Isl1 expression
6 Pituitary Development and Organogenesis: Transcription Factors in. . . 141

does not become ventralized at e11.5 and it is not detected at e12.5. Pitx2 expression
is also not maintained. Expression of pituitary hormone genes is reduced but
detectable.
The functions of Lhx3 and Lhx4 overlap (Sheng et al. 1997). Both Lhx3+/
,
Lhx4
/
and Lhx3
/
, Lhx4+/
embryos can form Rathke’s pouch, but it fails to
expand and is more severely affected than in either single mutant. Lhx3
/
, Lhx4
/

double mutants have a very severe pituitary phenotype (Sheng et al. 1997). The
double mutants form a pouch rudiment, and invagination of the oral ectoderm
occurs, but the rudiment fails to separate from the oral ectoderm and remains at
the pharyngeal cavity beneath the palate.

6.4.3 Human Mutations

Following characterization of the Lhx3
/
mice, patients with Combined Pituitary

Human Deficiency (CPHD) were screened for LHX3 mutations. Loss of function
mutations were found in several unrelated patients (Bechtold-Dalla Pozza et al. 2012;
Bhangoo et al. 2006; Bonfig et al. 2011; Gregory et al. 2015; Jullien et al. 2019;
Kristrom et al. 2009; Netchine et al. 2000; Pfaeffle et al. 2007; Rajab et al. 2008;
Ramzan et al. 2017; Sobrier et al. 2012). Generally, the mutations were recessive and
associated with multiple pituitary hormone deficiencies. Limited neck rotation,
hyperplastic anterior pituitary gland, sensorineural hearing impairment, respiratory
problems, and skeletal abnormalities were found in some cases. Heterogeneity in the
presentation of pituitary size is observed in CPHD patients with mutations in other
genes including PROP1 (Riepe et al. 2001).
Mutations in LHX4 have been identified in multiple unrelated patients with
CPHD. One is a homozygous lethal missense variant (p.T126M) (Gregory et al.
2015), but the rest are heterozygous loss of function mutations with incomplete
penetrance (Dateki et al. 2010; Filges et al. 2012; Pfaeffle et al. 2008; Rochette et al.
2015; Tajima et al. 2007, 2010). This indicates that the dosage sensitivity to LHX4
loss of function is more extreme in humans and mice.
In conclusion, LHX3 and LHX4 have essential and overlapping roles in anterior
pituitary development in mouse and man, while LHX2 has a role in posterior
pituitary lobe development.

6.5 HESX1 (Rpx)

6.5.1 Gene Discovery and Gene Expression

Hesx1 was discovered by a screen for Homeobox factors in an Embryonic Stem cell
cDNA library, and independently, in a screen for homeobox factors expressed
during gastrulation, at e7.5 (Hermesz et al. 1996; Thomas et al. 1995; Thomas and
Rathjen 1992). Hesx1 is expressed sequentially in the anterior visceral endoderm,
anterior definitive endoderm, and the cephalic neural plate. While officially known
142 A. Z. Daly and S. A. Camper

as Hesx1, it was also named Rathke’s Pouch homeobox (Rpx1) because of its robust
expression in the developing pituitary primoridia at e9.5. Its expression in the
developing pituitary gland is extinguished by PROP1 at e14.5 (Dasen et al. 2001;
Gage et al. 1996).

6.5.2 Knockout Mouse Phenotype, Downstream Targets, Upstream

Regulators

Robinson and colleagues generated a Hesx1 deletion in mice and demonstrated that
93% of homozygous mutants died before weaning (Dattani et al. 1998). The
phenotype was variable and included developmental abnormalities of the forebrain,
eye, and pituitary gland. The pituitary glands were missing (5%), misplaced, dys-
morphic, and/or small (Dasen et al. 2001). Hesx1 deficiency permits excessive FGF
signaling, resulting in multiple invaginations of oral ectoderm instead of a single
Rathke’s pouch (Dasen et al. 2001). Most mice heterozygous for the deletion were
normal, but 1% displayed mild abnormalities of the same organs affected in
homozygous mice.
Rosenfeld and colleagues demonstrated that HESX1 acts as a repressor of Pou1f1
expression by binding the Pou1f1 early enhancer (Dasen et al. 2001; Olson et al.
2006). HESX1 binds the co-repressor transducin-like enhancer of split 1 (TLE1)
through its engrailed homology domain and also interacts with a corepressor, histone
deacetylase complex containing Nuclear Receptor Co-Repressor (NCoR). Forced
expression of HESX1 and TLE1 during pituitary development is sufficient to
suppress POU1F1-dependent cell lineages. Both HESX1 and PROP1 are paired-
type homeodomain transcription factors that can compete for the same DNA binding
sites. In the presence of beta-catenin, PROP1 binds the early enhancer of Pou1f1 and
initiates Pou1f1 gene expression.
Both Lhx3 and Pitx2 are necessary upstream regulators of Hesx1 in the develop-
ing pituitary gland (Gage et al. 1999; Sheng et al. 1996). A 100-bp element that
directs spatial expression of Hesx1 in Rathke’s pouch lies ~3.3 kb 30 of the gene and
is bound by PITX2 and GATA2 (Chou et al. 2006). The promoter proximal

570 bp, exons 1 and 2 and intron 1 of Hesx1 are sufficient for transgene reporter
expression in the anterior visceral endoderm, anterior neural ectoderm, and anterior
neural plate. A negative element that suppresses Hesx1 expression in the hypothala-
mus, and responds to inductive signals from Rathke’s pouch, has been mapped
568
to
532 bp upstream of the gene.

6.5.3 Patient Mutations

The phenotypes of Hesx1 knockout mice suggested that HESX1 mutations might
cause pituitary and anterior structure malformations in humans. A screen of
61 patients with holoprosencephaly, Septo-Optic Dysplasia (SOD), or pituitary
insufficiency uncovered two siblings with agenesis of the corpus callosum and
6 Pituitary Development and Organogenesis: Transcription Factors in. . . 143

panhypopituitarism that were homozygous for a missense mutation in HESX1, p.

R53C, which abrogates DNA binding (Dattani et al. 1998). This disease association
was confirmed by screening a larger cohort of 228 patients with congenital pituitary
defects for mutations in HESX1 (Thomas et al. 2001). Three heterozygous missense
mutations were identified, p.S170L, p.C509T, and p.A541G, which were associated
with incomplete penetrance and variable phenotypic presentation from SOD to
hypopituitarism, or Isolated Growth Hormone Deficiency (IGHD).

6.5.4 Summary

Hesx1 was discovered in two independent cDNA screens for homeobox transcription
factors. LHX1, LHX3, and LMX1B activate its expression in early head development,
and PITX2 and GATA2 activate its expression in Rathke’s pouch. Pituitary expres-
sion is extinguished by PROP1, which also releases HESX1-mediated repression of
Pou1f1, activating the development of somatotropes, lactotropes, and thyrotropes.
Homozygous mutations in mice and heterozygous mutations in humans cause variable
phenotypes that include reduced forebrain, anophthalmia, and pituitary dysfunction.
The cause of variable presentation is not known.

6.6 PROP1

6.6.1 Gene Discovery

Prophet of Pit1 or PROP1 is a pituitary-specific transcription factor that was

identified by positionally cloning a spontaneous dwarf mouse mutant, Ames
dwarf, df (Sornson et al. 1996). These mice were first described phenotypically
58 years ago by Drs. Schaible and Gowen at Iowa State University in Ames, Iowa, as
they studied the descendants of an irradiation experiment (Schaible and Gowen
1961). Homozygous mutant mice have anterior pituitary hypoplasia, growth insuffi-
ciency, hypothyroidism, and infertility, which are attributed to lack of growth
hormone, prolactin, and thyroid-stimulating hormone (Bartke 1965; Cheng et al.
1983; Gage et al. 1995). Hesx1 expression was not properly repressed in the df/df
mutant pituitaries, placing the df locus downstream of Hesx1 (Gage et al. 1995). No
expression of Pou1f1 could be detected in the pituitaries of developing df/df mutants
by in situ hybridization, indicating that Prop1 is required for activation of Pou1f1
(Andersen et al. 1995; Gage et al. 1996). However, small clusters of cells expressing
Pou1f1 and GH, PRL or TSH could be detected in adult df/df mutant pituitaries by
immunohistochemistry (Gage et al. 1995). These committed somatotropes did not
express detectable levels of Ghrhr and were insensitive to increased stimulation by
growth hormone-releasing hormone (Gage et al. 1995).
Genetic mapping placed the df locus on mouse chromosome 11 in a region of
synteny homology with human chromosome 5q (Buckwalter et al. 1991). Using
genetically directed representational difference analysis, the critical region was
144 A. Z. Daly and S. A. Camper

narrowed to a 400–600 kb window (Sornson et al. 1996). Genomic clones from this
region were probed with cDNA from the developing pituitary, resulting in the
identification of Prop1, which encodes a 233 amino acid paired homeodomain-
containing protein. The Ames dwarf mice had a missense mutation in the
homeodomain, p.S83P, that abrogates DNA binding. A null allele of Prop1 was
genetically engineered by inserting a LacZ/Neo cassette into the first exon of Prop1
(Nasonkin et al. 2004). These Prop1
/
mice essentially phenocopy the Prop1df/df
mice when compared to the same genetic background. Prop1 expression is highly
restricted to the pituitary, beginning at embryonic day 10.5, reaching a peak at e12.5,
declining through development, and remaining detectable in the postnatal period
(Perez Millan et al. 2016; Sornson et al. 1996).

6.6.2 Downstream Targets

PROP1 binds β-catenin and competes with the HESX1-corepressor complex for
binding at the early enhancer of Pou1f1 to induce Pou1f1 expression (Dasen et al.
2001; Olson et al. 2006; Sornson et al. 1996). Deletion of β-catenin before the
expression of Pou1f1 blocks Pou1f1 expression, while deletion of β-catenin after the
onset of Pou1f1 expression had no effect, suggesting a narrow critical window of
β-catenin expression in lineage determination. This β-catenin phenotype occurs
independently of LEF1, TCF3, or TCF4 action. PROP1 interacts directly with the
corepressors TLE, Reptin, and HDAC1 to bind and repress the expression of Hesx1.
Lineage tracing reveals that all hormone-producing cell types of the anterior and
intermediate lobes of the pituitary descend from Prop1-expressing cells (Davis et al.
2016). This places Prop1 just downstream of Sox2 expression in pituitary
progenitors (Perez Millan et al. 2016) (Fig. 6.2). Prop1df/df mice have a vasculariza-
tion defect, reduced expression of cyclin D1 and cyclin D2 in pituitary progenitors,
and a failure of proliferating cells to migrate away from the stem cell niche into the
parenchyma, which appears to be a failed epithelial to mesenchymal-like transition
(Ward et al. 2005). Many mutant cells undergo apoptosis. Notch signaling is
normally active in the transitional zone between the stem cell niche and the
differentiating cells, but Prop1 mutants do not express Notch2 (Himes and Raetzman
2009; Raetzman et al. 2004). Colony forming assays and RNA-Seq were used to
assess the effect of Prop1 deficiency on stem cells. Mutants had a reduced stem cell
pool, abnormal colony morphology, and changes in gene expression consistent with
a failed epithelial-to-mesenchymal transition. To identify additional downstream
targets of Prop1 that might underlie the features of mutant pituitaries, Pérez Millán
and colleagues developed a biotin-tagged allele of PROP1 in an immortalized
pituitary cell line, facilitating direct identification of PROP1-binding sites in chro-
matin (Perez Millan et al. 2016). Binding sites were identified in or near Notch2,
Gli2, and Zeb2. Zeb2 activation appeared to be an essential step in the process of
engaging progenitors to undergo EMT.
6 Pituitary Development and Organogenesis: Transcription Factors in. . . 145

6.6.3 Upstream Regulators

Multiple species alignment of PROP1 genomic sequences from a variety of

mammals revealed three areas of high conservation: upstream, downstream, and
within intron 1 of Prop1 (Ward et al. 2007). In the context of the Cga promoter, the
intronic enhancer is sufficient to drive expression in the stem cell niche, located in
the dorsal aspect of the pituitary gland in transgenic mice. The Cga promoter alone
drives expression in the more ventral aspect of the organ, consistent with the location
of differentiating gonadotropes and thyrotropes. This enhancer sequence is bound by
RBpJK, indicating that Notch signaling feeds back to upregulate Prop1 expression
(Zhu et al. 2006).

6.6.4 Patient Mutations

Quickly after the discovery of mouse Prop1 by positional cloning, the human
ortholog was identified and recessive loss of function mutations was found in four
unrelated families with CPHD (Wu et al. 1998). All three mutations either reduced or
completely eliminated PROP1’s ability to bind DNA, convincingly implicating
these lesions as the cause of CPHD. Mutations in PROP1 are the most common
known cause of CPHD (Cogan et al. 1998; Deladoey et al. 1999). The most frequent
mutations are frameshifts causing premature termination; many patients are homo-
zygous for c.[301-302delAG] or c.[150delA]. Portuguese (Lemos et al. 2006),
Lithuanian (Navardauskaite et al. 2014), and Czech (Lebl et al. 2005) cohorts have
a high incidence of the 2-bp deletion, and the 1-bp deletion is common in Croatia
(Dusatkova et al. 2016). A thorough investigation of the haplotype revealed that the
2-bp deletion founder mutation arose approximately 101 generations ago in central
eastern Europe, and it either reoccurred or was transferred to the Iberian peninsula
approximately 23 generations ago (Dusatkova et al. 2016). The Iberian haplotype
and mutation were subsequently transferred to Latin America. Combining all popu-
lation groups analyzed so far, PROP1 mutations are estimated to cause ~11% of
familial and sporadic cases of CPHD (Fang et al. 2016). All mutations discovered to
date are recessive, and the phenotype includes deficiency of GH, TSH,
gonadotropins, and progressive loss of ACTH. We speculate that the progressive
hormone loss may result from depletion of the pituitary stem cell pool.

6.7 POU1F1 (PIT-1, GHF-1)

6.7.1 Gene Discovery

Karin and Rosenfeld independently discovered the first pituitary-specific transcrip-

tion factor, demonstrated that it regulated expression of the Gh and Prl genes, and
named it GHF1 and PIT1, respectively (Bodner et al. 1988; Ingraham et al. 1988). It
was later renamed POU1F1. Rosenfeld’s group identified cis-acting sequences in the
146 A. Z. Daly and S. A. Camper

Gh and Prl genes that were necessary for their expression in pituitary cell cultures
and used them as probes to screen a pituitary cDNA expression library (Ingraham
et al. 1988; Nelson et al. 1986). They identified the Pou1f1 cDNA, detected
transcripts only in the pituitary gland, and showed that Pou1f1 expression vectors
were sufficient to activate Gh and Prl reporter genes in transfected, heterologous
cells. Karin’s group took the approach of biochemically purifying the protein that
bound the cis-acting sequences in the Gh promoter and obtaining the amino acid
sequence of a peptide from the binding protein (Bodner et al. 1988; Bodner and
Karin 1987). Probes complementary to the peptide sequence were used to screen
pituitary cDNA libraries and GHF1 was identified. Loss of GHF1 resulted in the loss
of Gh expression (McCormick et al. 1988). Subsequent structure–function analyses
revealed the presence of the nuclear localization signal and the roles of the
homeodomain and POU-specific domain in specificity of DNA binding (Theill
et al. 1989).

6.7.2 Knockout Mice

George D. Snell described the phenotype of an autosomal recessive, spontaneous

dwarf mouse mutant in 1929, and it was named dw (Snell 1929). The growth
insufficiency of the mutant mice is evident 14 days postnatally, and adult mutants
also exhibit hypothyroidism and hypogonadism. Tissue transplantation experiments
demonstrated that the multi-organ defects were rescued by pituitary tissue, and
mutant pituitaries did not improve function when implanted in the sella of normal
mice, pinpointing the pituitary as the root cause of the hormone deficiencies (Carsner
and Reynolds 1960; Smith and MacDowell 1930).
Finally, 61 years after the dw Snell dwarf mutation was described phenotypically,
lesions in Pou1f1 were implicated as causal (Camper et al. 1990; Li et al. 1990).
Pou1f1 was tightly linked genetically to the dw locus, and the gene was rearranged in
an allelic variant, dwJ (Eicher and Beamer 1980). DNA sequencing revealed a
G ! T missense mutation, resulting in a p.W251C change of a highly conserved
residue in the homeodomain that abrogates DNA binding. These findings extended
the role of Pou1f1 from direct activation of Gh and Prl gene expression to include
stimulating the development and specification of cells that express Gh, Prl, and
Tshb. These three specialized cell types became known as the Pou1f1 lineage.

6.7.3 Patient Mutations

Two years after linking the mouse dwarfism phenotype to lesions in Pou1f1, Khono
and colleagues identified the first POU1F1 mutation found in a patient with TSH,
GH, and prolactin deficiencies (Tatsumi et al. 1992). This c.C638T nonsense
mutation resulted in a truncation, p.R172X, N-terminal to the homeodomain of
POU1F1. Numerous POU1F1 mutations have been described in patients with
CPHD (reviewed in Fang et al. 2016). The majority are recessive loss of function
6 Pituitary Development and Organogenesis: Transcription Factors in. . . 147

mutations that present with deficiency of GH, PRL, and TSH. However, there are a
few notable exceptions. The p.R271W mutation is dominant negative and incom-
pletely penetrant, possibly due to variability in mono-allelic expression of the mutant
versus normal allele (Ohta et al. 1992; Okamoto et al. 1994; Radovick et al. 1992).
The p.R271W mutation interferes with binding to C/EBPα, and it permits recruit-
ment to centromeric heterochromatin (Enwright et al. 2003). The mutation also
interferes with tethering of POU1F1 dimers to the nuclear matrix via interactions
with β-catenin, SATB1, and matrin-3 (Skowronska-Krawczyk et al. 2014). A p.
K216E mutation in the homeodomain is also a dominant cause of CPHD (Cohen
et al. 1999). This mutation acts by interfering with the ability of POU1F1 to
autoactivate its late enhancer in response to retinoic acid stimulation (see below).
A p.P76L mutation in the transactivation domain causes dominant IGHD in humans
and recessive growth insufficiency in mice (Sobrier et al. 2016). TSH and PRL were
not measured in the mouse model. The amino acid substitution increases DNA
binding at the human GH1 locus control region and increases binding to several
transcription factors, including PITX1, LHX3a, and ELK1. The p.P76L change
decreases transactivation of Gh, but it has no effect on the Prl promoter. Two
other proline substitutions in the transactivation domain have been reported in
children with GH, PRL, and TSH deficiency: p.P14L and p.P24L (Fofanova et al.
1998; Ohta et al. 1992). Detailed functional studies were not carried out. The p.P14L
variant appears dominant with incomplete penetrance, and no inheritance data were
presented for the p.P24L allele. In sum, mutations in POU1F1 can cause CPHD or
IGHD and be recessive or dominant, and dominant mutations are informative about
the normal mechanism of action.

6.7.4 Upstream Regulators

Pou1f1 expression is selectively high in the anterior lobe of the pituitary. The
cis-acting sequences necessary for Pou1f1 expression were mapped extensively in
transgenic mice (Rhodes et al. 1993). A 390-bp enhancer located ~10.4–10.2 kb
upstream of the Pou1f1 transcription start site is sufficient for Pou1f1 expression
postnatally in the context of the minimal 327 bp Pou1f1 promoter (DiMattia et al.
1997). This distal or “late enhancer” is bound by POU1F1 in an autoregulatory loop
necessary for maintaining expression. Another enhancer element located between
~8 kb upstream of Pou1f1 is bound by PROP1 and β-catenin, inducing initial
expression of Pou1f1, but this proximal or “early enhancer” is not sufficient for
the sustained expression of Pou1f1 into adulthood (DiMattia et al. 1997; Olson et al.
2006; Sornson et al. 1996).

6.7.5 Downstream Targets

The role of POU1F1 in establishing thyrotropes and Tshb expression was discovered
later (Charles et al. 2006; Dasen et al. 1999; Gordon et al. 1997). Ridgway and
148 A. Z. Daly and S. A. Camper

colleagues mapped cis-acting sequences necessary for Tshb expression in thyrotro-

pic tumor cells and identified binding sites for POU1F1 and GATA2 within
133 to

88 of the Tshb transcription start site. Neither transcription factor had a major
impact on Tshb transcription individually, but together they synergistically activated
Tshb. Using dominant-negative and ectopic gene expression transgenic approaches,
Rosenfeld and colleagues discovered that POU1F1 and GATA2 interact to promote
thyrotrope development, while GATA2 alone promotes gonadotrope development.
Specifically, they showed that broadly expressing Gata2 throughout the developing
pituitary blocks Pou1f1 expression, reduces thyrotrope and somatotrope differentia-
tion, and increases gonadotrope specification. Conversely, expressing Pou1f1
broadly throughout the developing pituitary drives high levels of thyrotrope and
somatotrope differentiation, and blocks gonadotrope specification. Camper and
Ridgway collaborated to show that a pituitary-specific knockout of GATA2 dimin-
ished thyrotrope and gonadotrope differentiation. At birth, the mice have very few
thyrotropes, but they recover as the mice age, likely through sensitive feedback loops
that maintain thyroid hormone production at the proper level. These mice also
became fertile. Marked upregulation of Gata3 was observed, which may compensate
for loss of Gata2. A pituitary-specific Gata2, Gata3 double knockout would be
necessary to assess whether loss of both GATA factors is sufficient to block
gonadotrope and thyrotrope development. Nevertheless, these studies established
an important role for both Pou1f1 and Gata2 in pituitary development.
POU1F1-mediated regulation of gene expression and cell fate requires interaction
with chromatin-modifying factors. POU1F1 binds to the Gh promoter by e13.5–
e14.5, but Gh transcription is not activated until a complex containing the histone
lysine demethylase, LSD1 is recruited (Wang et al. 2007). A pituitary-specific
deletion of Lsd1 has little effect on Pou1f1 expression, but there is little or no
expression of Gh, Tshb, or Prl. Differentiation appears to be suppressed because
Notch signaling fails to be silenced. LSD1 binds and activates the Gh promoter
through interaction with an MLL1 coactivator complex. A region of the Gh promoter
located at
161 to
146 is necessary for silencing Gh expression in lactotropes
postnatally. ZEB1, a homeodomain transcription factor with seven Zn fingers, binds
this element with a constellation of cofactors including LSD1 and corepressors.
Estrogen, an activator of Prl expression, enhances the recruitment of LSD1- and
ZEB1-mediated repression of Gh in lactotropes during the postnatal period in mice.
Recently, other POU1F1 binding partners were identified, shedding light on its
mechanism of action (Skowronska-Krawczyk et al. 2014). Using immunoprecipita-
tion and mass spectrometry, POU1F1 was discovered to interact with matrin-3,
β-catenin, and SATB1. The dominant-negative POU1F1 p.R271W mutation
interferes with POU1F1’s interaction with β-catenin and SATB1 which, in turn,
results in the failure to interact with matrin-3 and reduced expression of POU1F1
target genes. Gene expression is rescued by fusing POU1F1 (p.R271W) with matrin-
3. This shows convincingly that the multi-protein complex must be tethered to the
nuclear matrix for optimal activity.
6 Pituitary Development and Organogenesis: Transcription Factors in. . . 149

6.8 NR5A1 (Steroidogenic Factor 1, SF1, Ftz-f1, Adrenal

4-Binding Protein)

6.8.1 Gene Discovery

NR5A1 is considered to be the signature transcription factor for driving pituitary

gonadotrope cell fate. It was initially discovered in the search for transcription
factors regulating steroidogenic enzyme biosynthesis in the adrenal cortex. Parker
and Morohashi independently made the observation that multiple cis-regulatory
elements driving expression of steroidogenic enzyme genes and cytochrome P450
genes in the adrenal cortex had highly similar AGGTC motifs, suggesting that a
single factor might regulate the expression of these proteins (Morohashi et al. 1992;
Rice et al. 1991). In support of this, the factor(s) that bound all six of these elements
were the same molecular weight and had identical chromatographic properties. They
took a biochemical purification approach to identify the transcription factor binding
these elements. Y1 adrenocortical tumor cells did not produce sufficient protein for
purification and characterization. To circumvent this limitation, they took advantage
of the evolutionary conservation of regulatory networks among mammals. The
promoters of these three genes have high sequence similarity in cows and mice
(Honda et al. 1990), and the bovine factor that binds the mouse regulatory elements
is the same size as the mouse factor (Honda et al. 1993; Lala et al. 1992). They
collected bovine adrenal cortex tissues, purified the binding protein, and named it
steroidogenic factor 1 or SF1. This 53-kD protein could bind all six regulatory
elements. Because the AGGTC motif is similar to the binding sites of nuclear
hormone receptors, they predicted that SF1 would have a DNA-binding domain
similar to the nuclear hormone receptor family. They screened a Y1 cDNA library
with a probe designed to hybridize to the consensus sequence encoding a nuclear
hormone receptor DNA binding domain. One of the cDNAs fit the expected
expression profile of SF1, namely expression in the adrenal glands, the corpus
luteum, and the testis. They proved that they had cloned the correct cDNA by
expressing the protein in bacteria and demonstrating that it had the same binding
affinity as that of SF1 purified from the bovine adrenal gland. The SF1 cDNA
sequence encoded an orphan nuclear hormone receptor related to the Drosophila
segmentation gene, fushi tarazu or ftz, which also specifies neuronal identity in flies.
The gene was renamed as NR5A1 according to the nomenclature for orphan nuclear
receptors.

6.8.2 Gene Expression

In mice, Nr5a1 transcripts were detected by in situ hybridization in the developing

hypothalamus and the urogenital ridge, which is the precursor to the adrenal glands
and gonads (Ikeda et al. 1994). Nr5a1 sustains high expression in the adrenal glands
through adulthood. Conversely, it is expressed highly in the developing bipotential
gonad, and it is maintained in testes, particularly in Sertoli cells. Meanwhile, its
150 A. Z. Daly and S. A. Camper

expression in ovaries is reduced when the ovary first becomes morphologically

distinct from the testis. Nr5a1 expression in the bipotential gonad suggested a func-
tion in sexual development. Nr5a1 was expressed in the ventromedial hypothalamus
(VMH), which is involved in feeding, fear, thermoregulation, and sexual activ-
ity (Shima et al., 2005). These studies indicated that Nr5a1 was expressed early in
organ development and could, therefore, be a driver of cell fate in mammals.

6.8.3 Downstream Targets

Mellon and colleagues identified a DNA element that regulates gonadotrope expres-
sion of Cga, which encodes the alpha-subunit of the heterodimeric glycoprotein
hormones, luteinizing hormone (LH), follicle-stimulating hormone (FSH), and
thyroid-stimulating hormone (TSH) (Horn et al. 1992). Mellon showed that this
element (which they called GSE for Gonadotrope-Specific Element) was bound by a
54-kD protein, which they labeled GSEB-1 (GSE-Binding factor 1). They
demonstrated that GSEB-1 is NR5A1 by showing that NR5A1 binds specifically
to the GSE in vitro and that NR5A1-specific antibodies antagonize GSEB-1 function
(Barnhart and Mellon 1994). This was the first evidence that NR5A1 had a role in the
pituitary gland.
Mellon discovered that NR5A1 regulates the transcription of Lhb. This
represented the second known direct target of NR5A1 in the pituitary gland.
NR5A1 binds a GSE in the promoter region of Lhb and increases its expression
(Halvorson et al. 1996; Horn et al. 1992). Nilson and colleagues showed that 776 bp
of the bovine LHB proximal promoter region was sufficient to drive gonadotrope-
specific expression in the transgenic mice (Keri et al. 1994). This sequence contained
the GSE, was responsive to GnRH regulation but was not androgen or estrogen
responsive. A mutation in the GSE element reduced transgene expression ten-fold
(Keri and Nilson 1996). This suggested that NR5A1 stimulates LHB expression by
binding the promoter-proximal GSE in cell culture and in vivo. Morohashi and
colleagues found two regulatory elements of Nr5a1, showing that NR5A1 binds one
of them to stimulate its own expression in adrenal cells (Honda et al. 1993), while
Clay and colleagues demonstrated that NR5A1 directly regulates expression of
Gnrhr via a GSE in the promoter-proximal region (Duval et al. 1997).

6.8.4 Interacting Factors

NR5A1 interacts with other transcription factors to promote gene expression. The
Drouin lab demonstrated that PITX1 and NR5A1 interact directly and synergize to
activate Lhb promoter but not Cga gene expression (Tremblay et al. 1998, 1999).
The PITX1 binding site is not required for this activity. NR5A1 also interacts with
the zinc-finger transcription factor EGR1 to regulate the expression of Lhb (Lee et al.
1996; Tremblay and Drouin 1999). Both NR5A1 and EGR1 binding sites are
required for the synergism. Milbrandt and Charnay independently demonstrated
6 Pituitary Development and Organogenesis: Transcription Factors in. . . 151

that EGR1 is necessary for Lhb expression in mice (Lee et al. 1996; Topilko et al.
1998). There are no exact obvious NR5A1 binding sites in the Fshb promoter, but
Mellon and colleagues found two sites bound by NR5A1 (Jacobs et al. 2003).
NR5A1 interacts with Nuclear Factor Y (NFYA) on the Fshb promoter to synergis-
tically activate Fshb transcription.

6.8.5 Knockout Mouse Phenotype

Nr5a1 knockout mice were generated by homologous recombination in embryonic

stem (ES) cells (Luo et al. 1994). Homozygous mutant mice lack adrenal glands and
gonads, revealing that NR5A1 is not only critical for expression of steroidogenic
enzymes in these tissues, but it has a critical role early in adrenal and gonadal
organogenesis. Moreover, the pituitaries of Nr5a1-null mice lack detectable
transcripts for Lhb, Fshb, and the gonadotropin-releasing hormone receptor, Gnrhr
(Ingraham et al. 1994). Expression of other pituitary hormones was normal. Nr5a1
transcripts were detected at e13.5–e14.5, prior to the expression of other
gonadotrope markers, suggesting that NR5A1 could be involved in gonadotrope
differentiation, but it was still unclear whether Nr5a1 was involved in differentiation
or maintenance.
Nr5a1 mutants had structural abnormalities in the VMH, implicating NR5A1 in
three levels of the reproductive axis: hypothalamus–pituitary–gonad (Ikeda et al.
1995). However, the GnRH neurons were normal. A 24-h treatment regime of
GnRH injections can rescue expressions of Lhb and Fshb in hypogonadal mutants,
Gnrhhpg/hpg, and Nr5a1 knockout mutants, suggesting that the gonadotrope
progenitors are present in Nr5a1 mutants, and that NR5A1 is not strictly required
for gonadotrope fate (Ikeda et al. 1995). GnRH induces expression of many
transcription factors in the pituitary gland, including Egr1, Atf3, c-Jun, and
TCF/LEF, and increased expression of these factors may compensate for NR5A1
deficiency (Kaiser et al. 2000; Melamed et al. 2006; Salisbury et al. 2009; Tremblay
and Drouin 1999).
To distinguish direct effects of NR5A1 on the pituitary gland from indirect effects
resulting from altered hypothalamic and/or gonadal inputs to the pituitary gland,
researchers developed a floxed allele of Nr5a1 and induced deletion in the pituitary
with a Cga-cre transgene (Cushman et al. 2000; Zhao et al. 2001). This ablated
expression of Fshb, Lhb, and Gnrhr. These deficiencies caused failure to develop
mature reproductive organs and infertility. Injection of these pituitary-specific
knockout mice with pregnant mare serum gonadotropin (PMSG), a hormone pro-
duced by the equine placenta that has luteinizing- and follicle-stimulating activity,
was sufficient to induce pituitary expression of Lhb and Fshb. This demonstrated
that Nr5a1 deficiency impacts pituitary development and function directly, but it can
be overcome by supraphysiological hormonal stimulation.
152 A. Z. Daly and S. A. Camper

6.8.6 Patient Mutations

The first patient with an NR5A1 mutation was reported in 1999 (Achermann et al.
1999). The patient presented with primary adrenal failure, complete XY sex reversal,
and streak-like gonads. Her pituitary gland responded to GnRH stimulation. She was
heterozygous for a 2-bp mutation that causes a p.G35A missense in the first zinc
finger of NR5A1 that abrogates DNA binding. Her phenotype is consistent with
NR5A1’s role in adrenal function and stimulation of LH and FSH expression. The
sex reversal was surprising because mice heterozygous for a null allele of Nr5a1 are
phenotypically normal. Loss-of-function mutations have been associated with other
cases of 46XY sex reversal (Achermann et al. 2002; Correa et al. 2004; Lin et al.
2007; Mallet et al. 2004), 46XX sex reversal (Hasegawa et al. 2004; Swartz et al.
2017b), undervirilization (Swartz et al. 2017a), adrenocortical insufficiency (Biason-
Lauber and Schoenle 2000; Guran et al. 2016), premature ovarian failure (Lourenco
et al. 2009), and spermatogenic failure (Bashamboo et al. 2010). Most mutations
were dominant, some were associated with incomplete penetrance, and some were
recessive (Baetens et al. 2017; Bashamboo et al. 2016; Igarashi et al. 2017; Sekido
and Lovell-Badge 2008).

6.8.7 Upstream Regulators

Mice homozygous for a hypomorphic Pitx2neo allele have no Nr5a1 expression and,
consequently, no Lhb, Fshb, or Gnrhr expression (Suh et al. 2002). This suggests
that Pitx2 is upstream of Nr5a1 in the transcriptional hierarchy regulating pituitary
development. Later, PITX2 was shown to bind a pituitary gonadotrope-specific
enhancer within intron 6 of the Nr5a1 gene (Shima et al. 2008). Several groups
used transgenic mice to localize elements necessary for Nr5a1 expression in various
tissues (Karpova et al. 2005; Shima et al. 2005; Stallings et al. 2002). The action of
Nr5a1 in promoting gonadotrope fate is antagonized by TBX19 (Pulichino et al.
2003b).

6.9 POMC Transcription in Corticotropes (TBX19, NEUROD1)

6.9.1 Factor Discovery

Pro-opiomelanocortin (POMC) is a prohormone that is cleaved in anterior pituitary

corticotropes to become adrenocorticotropic hormone (ACTH) and in melanotropes
of the intermediate lobe into melanocyte-stimulating hormone (MSH) and beta-
endorphin. Drouin used the AtT-20 mouse corticotrope tumor cell line to identify
cis-acting elements that regulate Pomc expression. One element contained an E-box
motif, which is typically bound by helix-loop-helix (HLH) transcription factors
(Therrien and Drouin 1993). The beta HLH factor NeuroD1 is expressed in
corticotrope cells and binds this element, acting together with PITX1 to activate
6 Pituitary Development and Organogenesis: Transcription Factors in. . . 153

Pomc expression (Poulin et al. 1997). The detection of Neurod1 expression in the
developing pituitary gland from e12 to e16 is consistent with a role in initiating
Pomc expression. Neurod1 is transiently expressed in corticotropes, suggesting that
there are other factors involved in sustained Pomc transcription (Poulin et al. 2000).
PITX1 was identified in an effort to identify a trans-acting factor that bound
another cis-acting sequence within the promoter-proximal region of Pomc
(Lamonerie et al. 1996). This element was sufficient to confer reporter gene expres-
sion in AtT-20 cells but not heterologous cells. Drouin and colleagues used this
sequence to screen a cDNA expression library for DNA binding proteins, which led
to the discovery of a novel homeodomain transcription factor PTX, now known as
Pitx1. Pitx1 is expressed in the pituitary primordium, and is detected in all pituitary
cell types, but it is enriched in adult gonadotropes and thyrotropes (Charles et al.
2005; Lanctot et al. 1999). Mice homozygous for Pitx1 knockout die at birth and
have severe defects in mandible and hind limb development, but there is little effect
on pituitary development due to functional overlap with the related Pitx2 gene
(Charles et al. 2005; Lanctot et al. 1999; Szeto et al. 1999).
A cis-acting sequence adjacent to the PITX1 binding site is required for Pomc
expression, and the sequence motif, TCACACCA, is a T box transcription factor-
binding site (Lamolet et al. 2001). PCR amplification of Tbox cDNAs in AtT-20 cells
identified a novel factor, TPIT (officially known as TBX19), which is expressed early
in pituitary development and is enriched in corticotropes and melanotropes. Ectopic
expression of Tpit under the control of the Cga promoter in mice is sufficient to drive
Pomc expression, but not Neurod1 expression. Patients with isolated ACTH defi-
ciency were screened and two unrelated individuals were identified with mutations in
TPIT, confirming the critical role of this gene in POMC expression. Homozygous
deletion of Tpit results in almost complete loss of POMC in both the anterior and the
intermediate lobes, and the intermediate lobe is hypoplastic (Pulichino et al. 2003b).
Neurod1 expression is not obviously affected, suggesting that the commitment to
corticotrope and melanotrope fate has been initiated. In the absence of TPIT, the
progenitors in the intermediate lobe express Nr5a1 and the alpha and beta subunits of
TSH, FSH, and LH, suggesting that TPIT has a repressive effect on gonadotrope and
thyrotrope cell fates. TPIT interacts directly with NR5A1 to inhibit its transcriptional
activity. It is noteworthy that Pou1f1, Gh and Prl are not ectopically expressed in the
intermediate lobes of Tpit mutants. Overexpressing Tpit under the control of the Cga
promoter had little effect on Tshb expression, but it is sufficient to suppress expres-
sion of Nr5a1, Lhb, and Fshb, consistent with the idea that TPIT drives corticotrope
and melanotrope fate while suppressing gonadotrope fate.

6.9.2 Patient Mutations

TPIT mutations are the most common, known cause of congenital, isolated ACTH
deficiency (Couture et al. 2012; Metherell et al. 2004; Pulichino et al. 2003a;
Vallette-Kasic et al. 2005). Patients with homozygous loss-of-function mutations
typically present with undetectable plasma ACTH and corticosterone, and hypoplas-
tic adrenal glands. Affected babies may have cholestatic jaundice and potentially
154 A. Z. Daly and S. A. Camper

fatal hypoglycemia. Tpit null mice have more yellow pigment than wild-type mice,
due to the lack of MSH, but no obvious pigment defects are present in human
patients.

6.9.3 Summary

Pomc transcription and the corticotrope and melanotrope cell fates are promoted by
several transcription factors including TPIT, NeuroD1, PITX1, and PITX2. TPIT is a
T box transcription factor that binds promoter-proximal region of POMC and
induces its expression in concert with PITX factors. It is selectively expressed in
corticotropes and melanotropes, activating the corticotrope and melanotrope lineage
while repressing the gonadotrope lineage by antagonizing NR5A1 function in a
DNA binding-independent manner. Loss of TPIT results in neonatal-onset isolated
ACTH deficiency, but not juvenile onset.

6.10 PAX7

6.10.1 Gene Discovery

Pax7
/
mice were generated more than 20 years ago and revealed the importance of
Pax7 for muscle and brain development (Jostes et al. 1990; Kassar-Duchossoy et al.
2005; Mansouri and Gruss 1998; Oustanina et al. 2004; Relaix et al. 2005, 2006;
Seale et al. 2000). The homozygous mutants die 3 weeks after birth, but the pituitary
gland was not characterized (Mansouri et al. 1996). Karlstrom and colleagues
showed that the SHH-responsive gene Pax7 is expressed in the melanotropes of
the developing zebra fish pituitary gland (Guner et al. 2008). Spatial and temporal
gene expression profiling revealed that Pax7 is expressed in progenitors that give
rise to intermediate lobe melanotropes in mice, and it was hypothesized to be a
critical factor for intermediate lobe cell fate through direct regulation of Notch
signaling (Budry et al. 2012; Goldberg et al. 2011; Hosoyama et al. 2010).

6.10.2 Downstream Targets, Pioneering Activity

Drouin and colleagues showed that Pax7
/
mice lack intermediate lobe
melanotropes, and the deficiency of Pax7 permits ectopic differentiation of
progenitors into corticotropes (Budry et al. 2012). Moreover, ectopic expression of
Pax7 in the immortalized, corticotrope-like AtT-20 cells is sufficient to convert them
to a melanotrope identity associated with both reduced expression of corticotrope-
enriched transcripts such as the NeuroD1, glucocorticoid receptor, CRH receptor,
and vasopressin receptor 1b and activation of melanotrope-enriched transcripts such
as the dopamine receptor Drd2 and prohormone convertase 2, Pcsk2, which cleaves
Pomc to produce MSH. Thus, Pax7 is a selector of intermediate lobe melanotrope
cell identity.
6 Pituitary Development and Organogenesis: Transcription Factors in. . . 155

To understand the underlying mechanism of PAX7 action in reprogramming

AtT-20 cells to the melanotrope fate, binding sites for TPIT, PAX7, and
H3K4me1 were identified by chromatin immunoprecipitation (CHIP-Seq) and
open chromatin was identified using FAIRE-Seq (Formaldehyde-Assisted Isolation
of Regulatory Elements). Methylation of histone H3 at lysine 4 (H3K4me1) is
associated with enhancers. Drouin and colleagues demonstrated that PAX7 can
bind heterochromatin and act as pioneer factors to open the chromatin, yielding a
FAIRE-Seq signal and permitting the binding of additional transcription factors such
as TPIT (Mayran et al. 2018).
To further characterize PAX7 action, the corticotrope and melanotrope
transcriptomes and epigenomes were compared. In so doing, these authors con-
firmed much of what they had shown previously, that PAX7 is necessary and
sufficient for establishing a melanotrope transcriptome and epigenome. These
authors made the interesting discovery that the majority of PAX7 action on the
epigenome occurs at regions far from gene promoters, as the promoters of most
genes in corticotropes and melanotropes are similar. They went on to show that the
areas of the genome that PAX7 opens and activates represents a minor, but sizable
fraction of PAX7 binding sites. Interestingly, they revealed a large fraction of sites
containing a PAX7-binding motifs that are impervious to PAX7 binding. These sites
were characterized by CTCF binding, and constitutive heterochromatin. While
constitutive heterochromatin was resistant to PAX7 remodeling, facultative hetero-
chromatin was readily bound by PAX7. Using inducible nuclear localization of
PAX7 in AtT20
PAX7+ cells, these authors showed that PAX7 binds to open
DNA very quickly (within 30 min), while it takes much longer (nearly 24 h) for
PAX7 to bind areas of closed DNA. Unsurprisingly, it takes even longer for PAX7 to
remodel bound, heterochromatin sites, requiring more than 3 days to fully open these
areas of DNA. Finally, areas of PAX7 binding and opening were subject to reduction
in CpG methylation, a very stable form of epigenetic memory, and most sites opened
by PAX7 pioneering activity remain open long after PAX7 has been removed from
the nucleus. With this information, the authors put forth the model that involves
PAX7 binding to both euchromatin and facultative heterochromatin, remodeling of
the areas of heterochromatin into areas of euchromatin, and allowing factors like
TBX19 and STAT3 to bind and drive expression. This clearly defines the function of
a pioneering transcription factor, and leaves us with the question of whether this
mechanism of pioneering action represents a broader pattern of pioneer action, or is
specific to PAX7.

6.10.3 Patient Mutations

PAX7 translocations resulting in fusion with the forkhead in rhabdomyosarcoma

gene, FKHR, can cause pediatric alveolar rhabdomyosarcoma (Barr et al. 1996).
Loss-of-function mutations would be expected to be lethal in the homozygous state,
and none have been reported in humans (Mansouri et al. 1996).
156 A. Z. Daly and S. A. Camper

6.10.4 Summary

PAX7 is the first validated pioneer transcription factor acting within the pituitary.
Important for both muscle and brain development, it was not implicated in pituitary
function until 2010. Melano/corticotrope precursor cells can become either cell type
in development, and the presence of PAX7 railroads the development to a
melanotrope fate whereas the absence of PAX7 results in a corticotrope fate.
PAX7 does this by making melanotrope-specific TPIT-binding sites available to
TPIT by binding to heterochromatin and converting it to euchromatin. The Drouin
lab has shown a very thorough mechanism by which PAX7 does this, revealing the
amount of time this chromatin-changing process takes, and showing its stability
through repeated DNA replication cycles. However, the exact complex required for
this change in the epigenetic landscape is unclear. Furthermore, all of the work done
to date has focused on the “opening” of DNA that is specific to melanotropes, while
it is yet unclear whether PAX7 “closes” corticotrope-specific DNA.

6.11 Current Model for the Transcription Factors

of the Pituitary Gland and Open Questions

A complex cascade of transcription factors is responsible for the development and

maintenance of the pituitary gland. PITX and LHX proteins are some of the earliest
factors in promoting pituitary formation. Pitx1 and Pitx2 compensate for one another
and are responsible for turning on Lhx3. LHX3 and LHX4 also have overlapping
functions, and together, they are necessary for the pouch rudiment to expand. During
these early stages, the majority of cells are undifferentiated, proliferating progenitor
cells, defined by their expression of SOX2. Then, Lhx3 activates Prop1 throughout
Rathke’s pouch from e10.5 to e12.5, and Prop1 quickly becomes restricted to the
progenitors that are undergoing an epithelial to mesenchymal-like transition, involv-
ing activation of Zeb2, and have silenced Sox2 expression. PROP1 represses Hesx1
expression and induces Pou1f1, the major lineage determining transcription factor
for somatotropes, lactotropes, and thyrotropes. POU1F1 interacts with epigenetic
regulators to activate and repress expression of target genes necessary for the
sub-specialization into three different hormone-producing cell fates. LSD1, ZEB1,
estrogen receptor, and GATA2 are some of the factors involved in
sub-specialization. During the last few days of embryogenesis, GATA2, NR5A1,
and EGR1 drive gonadotrope fate, and TPIT and NEUROD1 drive corticotrope fate.
PAX7 and TPIT act together to drive melanotrope fate.
This attractive model represents decades of sophisticated, pioneering work, and
explains many basic steps in pituitary development. There are, however, questions
that remain unanswered. How is the switch from proliferation to differentiation
regulated? What additional transcription factors and signaling pathways are involved
in cell specification?
SOX2 is firmly established as the marker for cycling progenitor cells in the
developing mouse pituitary, and it is also expressed in the quiescent stem cell reserve
6 Pituitary Development and Organogenesis: Transcription Factors in. . . 157

in adult pituitary. The steps involved in switching cells from their SOX2 expressing,
cycling state to their specific lineage is unclear. A thorough analysis of cell cycle
regulators from e11.5 to e16.5 helped clarify this, as it revealed three distinct
populations (Bilodeau et al. 2009). There is a cycling precursor population (defined
as Ki67-positive, and presumably SOX2-positive), a non-cycling precursor popula-
tion (defined as p57-positive and p27-negative), and a non-cycling differentiated
population (p57-negative and p27-positive). Drouin and colleagues showed that the
deletion of the major transcription factor TPIT resulted in an unusual population of
non-cycling, undifferentiated progenitors that were p57+ and p27+. This suggests
that p57 marks cell cycle exit, and major factors (PROP1, TPIT, NR5A1) silence
p57, permitting differentiation, and locking cells into their fate. Unlike the temporal
phasing of neuronal differentiation in the brain and retina, nearly all cells in the
pituitary leave the cell cycle at approximately the same time between e11.5 and
e13.5 (Davis et al. 2011). Two major questions remain: what triggers the transition
from the proliferating to non-proliferating state, and how are the major transcription
factors exclusively turned on in different progenitors? Notch, WNT, and YAP/TAZ
signaling play important roles in regulating progenitor proliferation (Cheung et al.
2018a; Gaston-Massuet et al. 2011; Lodge et al. 2019) The second question
concerning how each factor is turned on is almost certainly specific to the factor,
as common regulatory networks across cell lineages would fail to explain the highly
unique transcription network in the different cell types.
The prevailing model of cell-type differentiation in the pituitary involves an exit
from the cell cycle with a major transcription factor that inexorably pushes it down a
single path to induce a specific and exclusive fate. There is evidence to suggest,
however, that the differentiation is not as linear as once thought, and the fate may be
more plastic than the model suggests. Early analyses of the pituitary led Romeis to
propose a “one cell one hormone” model (von Möllendorff and Bargmann 1940).
However, a single gonadotrope can express and secrete both LH and FSH (Phifer
et al. 1973; Robyn et al. 1973) and GH and TSH are co-expressed in hypothyroid
mice (Horvath et al. 1990). Cells that express both GH and PRL may be a transient
population that represents a precursor of both somatotropes and lactotropes (Borrelli
et al. 1989; Frawley et al. 1985). Or, they may represent a transition from a
GH-expressing cell to PRL-expressing through the action of POU1F1, LSD1, and
ZEB1 (Wang et al. 2007). Boehm and colleagues discovered that at e17.5, nearly all
cells expressing FSH also express TSH, and that many of these cells express NR5A1
suggesting a close relationship between gonadotropes and thyrotropes (Wen et al.
2010). These studies highlight how closely related the various cell types are,
challenging the idea that major transcription factors direct an inflexible cell fate.
158 A. Z. Daly and S. A. Camper

6.12 Technical Developments and Future Directions

Four major technological innovations are now available to gain insight into cell fate
decisions during development: innovative DNA library preparation methods cou-
pled with high-throughput sequencing, single-cell sequencing, CRISPR, and
organoids.

6.12.1 High-Throughput Sequencing

Advances in sequencing technology have made genome-wide analyses of gene

expression feasible. Sequencing mRNA from genetically marked pituitary cells
has been invaluable to the pituitary field. Subpopulations of pituitary cells have
been purified from transgenic mice with fluorescently labeled cells and fluorescence-
activated cell sorting (FACS). This approach revealed heterogeneity of gonadotrope
gene expression in different sexes at different times (Qiao et al. 2016) and the
similarity between corticotropes and melanotropes, with only ~500 differentially
expressed genes (Mayran et al. 2018). This method also revealed novel regulators of
cell identity for somatotropes and lactotropes (Peel et al. 2018). There are
limitations, however, to applying this technology to small numbers of cells collected
from developing pituitaries, as they may contain different factors that drive cell fate
from those that are employed to maintain it in adult animals.
Identification of genome-wide regions of open chromatin and transcription factor
binding sites in specialized cells is valuable for understanding cell fate decisions.
DNase hypersensitivity mapping has been largely replaced by Assay for
Transposase-Accessible Chromatin with sequencing (ATAC-seq) for the identifica-
tion of putative regulatory elements (Buenrostro et al. 2013). These powerful
methods have successfully been used in the pituitary, finding regulatory elements
in melanotropes and corticotropes (Budry et al. 2012), implicating PAX7 as a
pioneering transcription factor, and further characterizing the speed at which
PAX7 euchromatinizes DNA and the longevity of the subsequent epigenetic
marks (Mayran et al. 2018). Similarly, they have shown the chromatin accessibility
at STAT3 and Glucocorticoid receptor binding sites in corticotropes (Langlais et al.
2012). Software programs are available to identify footprints of transcription factors
in ATAC-seq data and DNA-binding motifs that underlie transcription factor bind-
ing (Baek and Sung 2016). Genome-wide binding sites for PROP1 (Perez Millan
et al. 2016) were identified using Chromatin Immunoprecipitation with sequencing
(CHIP-seq) (Barski et al. 2007). POU1F1 binding at active enhancers with Matrin3
have been reported also (Skowronska-Krawczyk et al. 2014).
CHIP-seq for epigenetic marks on histones can identify putative poised enhancers
(histone H3 lysine 4 methylation, H3K4me1 alone) and active enhancers (H3K4me1
coupled with histone H3 lysine 27 acetylation, H3K27ac) (Creyghton et al. 2010).
H3K27me3 marking is a little more nuanced, as the presence of this mark across the
gene body is correlated with repression, whereas enrichment at the transcription start
site or the promoter may be a sign of bivalency or even active transcription (Young
et al. 2011).
6 Pituitary Development and Organogenesis: Transcription Factors in. . . 159

Chromatin Conformation Capture (3C) enables researchers to determine which

regulatory elements and promoters are interacting. This is important because
enhancer elements may not regulate the closest gene, and they may lie at a substan-
tial distance from the gene they regulate. Liebhaber and colleagues initially used
traditional techniques of DNase I hypersensitivity mapping and transgenic animal
models to discover and characterize the Locus Control Region (LCR) for the human
growth hormone gene cluster (Jones et al. 1995). There are several different
enhancers within the LCR that selectively regulate the expression of GH1 and the
surrounding GH-related genes in the pituitary and the placenta, respectively. The 3C
method was used to discover the interaction of different enhancers within the LCR
with the promoters of GH1 and the neighboring related genes (Ho et al. 2008, 2013;
Tsai et al. 2016). This method was also used to show that the early and distal
enhancers of Pou1f1 interact with the Pou1f1 promoter sequentially, and this shift
requires POU1F1, establishing an autoregulatory network, that results in the enrich-
ment of H3K27ac at the Pou1f1 promoter (Ho et al. 2015). Further application of this
method will link enhancers found using CHIP-seq to the genes they regulate. A
remaining challenge for the field is scaling the technology to small numbers of cells
available during pituitary development and the tendency for individual enhancers to
be redundant (Sarro et al. 2018).

6.12.2 Single Cell Sequencing

RNA sequencing of single cells (scRNA-seq) is now possible (Navin et al. 2011).
Various methods can be used to capture single cells including fluidics, manual
capture, and FACS (Shapiro et al. 2013), and there are different approaches for
obtaining the RNA sequences from individual cells, including bar coding RNA
library preparations from single cells followed by next-generation sequencing
(Zheng et al. 2017; Ziegenhain et al. 2017). scRNA-seq analysis algorithms allow
researchers to map out a shared differentiation pathway in an asynchronous popula-
tion of cells (Trapnell et al. 2014). Temporal changes in gene expression that occur
as differentiation proceeds can be inferred, revealing candidate genes that may drive
these differentiation pathways. The scRNA-seq methods are particularly valuable for
analysis of rare cell types or situations in which the relative contribution of individ-
ual cell types to the whole is shifting. Currently, very few publications containing
single-cell sequencing data from pituitary exist (Cheung et al. 2018b; Ruf-Zamojski
et al. 2018). But the work that has been done already shows that this approach is
powerful and can reveal previously unappreciated subpopulations of cells and novel
markers of specialized cell types.

6.12.3 CRISPR

Genetically engineered mice have been invaluable for understanding the role of
individual genes in pituitary development (Camper et al. 1995). Transgenic mice
160 A. Z. Daly and S. A. Camper

were used to identify the regulatory elements of the Gh gene (Jones et al. 1995) and
Nr5a1 (Karpova et al. 2005; Stallings et al. 2002) and to tease apart the differentia-
tion of the somatolactotrope (Borrelli et al. 1989). Homologous recombination in
embryonic stem cells was used to characterize the function of Hesx1 (Dasen et al.
2001) and numerous other pituitary genes, as described earlier. Despite the impor-
tance of these technologies, there are significant shortcomings, including expense
and inefficiency, and position effects associated with random integration of
transgenes. CRISPR-Cas9 (clustered regularly interspersed short palindromic
repeats-CRISPR-associated protein 9) technology has reduced the time and expense
needed to make genetic modifications in mammalian genomes (Cong et al. 2013).
Suddenly, knockouts, knockins, conditional alleles, and even more complex tissue-
specific CRISPR mice were developed. This technology will have a high impact on
the field as it is more broadly implemented.

6.12.4 Organoids

Organoids offer the opportunity to study developmental processes ex vivo, which

has helped in understanding the differentiation the gut (Sato et al. 2009), retina
(Nakano et al. 2012), kidneys (Xia et al. 2013), and pituitary gland (Ozone et al.
2016; Suga et al. 2011). Pituitary Sox2-expressing and verapamil-sensitive cells can
be cultured as pituispheres and differentiate into each of the anterior pituitary
hormone-producing cell types (Chen et al. 2005; Fauquier et al. 2008). Subse-
quently, Sasai and colleagues demonstrated an incredible ability of mouse and
human embryonic stem cells to self-organize and differentiate into functional pitui-
tary hormone-producing cells that could rescue hormone deficiency, giving promise
for future cell-based therapies for hypopituitarism (Ozone et al. 2016; Suga et al.
2011).

6.13 Future Directions

These powerful technological advances will help us clarify many questions that
remain in understanding pituitary development, namely understanding the switch
from proliferation to differentiation and the factors responsible for the differentiation
of the cell types. Single-cell sequencing will reveal more nuanced populations of
cells that will clarify the exact pathway from a Sox2-expressing progenitor cell,
through a non-cycling intermediate precursor, to a cell type that has some combina-
tion of major transcription factors and hormones, to finally, a differentiated cell type
that expresses high levels of hormones. Knowing exactly which factor is in which
cell will rigorously generate a model for differentiation that will be less prone to
revision. Highly enriched motifs and footprints in ATAC-seq data will implicate
transcription factors that are critical to this process, shortening the long list of factors
that will be generated by single-cell RNA-seq. ATAC-seq will also (in concert with
CHIP-seq for histone marks) find the regulatory elements for these major factors,
6 Pituitary Development and Organogenesis: Transcription Factors in. . . 161

revealing transcriptional networks that are active at different timepoints during

pituitary development. Using CRISPR, each factor’s necessity and function will be
directly assessed in vivo. Finally, once a larger suite of factors has been implicated,
whose expression profile is clear, and whose function has been characterized, we
will be able to generate highly accurate organoids, knowing which ingredients are
necessary and sufficient for creating a pituitary gland.
One such area that requires further investigations are the thyrotropes. These
represent ~5% of the total number of cells in the pituitary gland (Kulig et al.
1998), making them difficult to study. It is clear from the Pou1f1-dwarfed mice
that POU1F1 is required for the adult population of thyrotropes (Fig. 6.3), and it
interacts with GATA2 to drive thyrotrope fate (Dasen et al. 1999). There could be
additional factors that are necessary to drive thyrotrope differentiation (Charles et al.
2006). The regulatory elements of Tshb are not known. POU1F1, GATA2, and
MED220 all synergize on the Tshb promoter to activate its expression (Gordon et al.
1997, 2006), but these sequences are insufficient for expression in transgenic mice
(Camper et al. 1995). Using single-cell sequencing, ATAC-seq and CHIP-seq for
POU1F1 and histone marks, the factors responsible for thyrotropes, and the
cis-regulatory elements to which they bind will become clear.

Acknowledgments We thank Karine Rizzoti (Crick, London) for images of Sox2 mutant embryos,
Hironori Bando for critical reading of the manuscript, and the National Institutes of Health for
funding (T32GM007544 and T32HG000040 to AZD and HD034283 and HD097096 to SAC).

Key References

Sox2 deficiency impairs the differentiation of multiple pituitary cell types in humans
and mice (Kelberman et al. 2006). Sox2 expressing pituitary cells have the
potential to self-renew and generate all other hormone producing cells of the
pituitary gland (Fauquier et al. 2008). Sox2 expressing progenitors are recruited to
generate new hormone producing cells in adult animals challenged to increase
hormone production (Andoniadou et al. 2013).
Pitx1 was cloned as a transcription factor that binds a critical element of the Pomc
gene (Lamonerie et al. 1996). Pitx2 was cloned as a cause of Rieger syndrome
and in a search for homeobox genes expressed in the pituitary gland (Gage and
Camper 1997; Semina et al. 1996). Pitx1 deletion has a modest effect on pituitary
gonadotropes and thyrotropes (Szeto et al. 1999). Pitx2 has multiple roles in
pituitary development including early induction of Rathke’s pouch, differentia-
tion of gonadotropes and the Pou1f1 lineage, and overlapping function with Pitx1
(Charles et al. 2005; Gage et al. 1999).
Lhx3 knockout affects early pituitary development and causes apoptosis of the
pituitary primordium (Ellsworth et al. 2008; Sheng et al. 1996). Lhx4 knockout
affects early pituitary development and causes apoptosis in the pituitary
162 A. Z. Daly and S. A. Camper

primordium (Raetzman et al. 2002; Sheng et al. 1997). Lhx2 knockout affects
pituitary posterior lobe development (Zhao et al. 2010).
Hesx1 was identified as a transcription factor expressed in early development and
then concentrated in Rathke’s pouch (Hermesz et al. 1996). Hesx1 deficient mice
have multiple anomalies including pituitary hypoplasia, and severely affected
individuals have septo-optic dysplasia (Dattani et al. 1998). HESX1 suppresses
Prop1 expression through interactions with co-repressors, and in the presence of
beta-catenin, Prop1 activites Pou1f1 expression (Olson et al. 2006).
PROP1 is a pituitary-specific transcription factor that is necessary for activation of
Pou1f1 and for differentiation of thyrotropes, somatotropes and lactotropes
(Sornson et al. 1996). Prop1 is necessary for developing a normal pool of
pituitary stem cells, for driving an epithelial to mesenchymal-like transition
process through the activation of Zeb2 and suppression of cell adhesion (Perez
Millan et al. 2016). All hormone producing cells of the anterior and intermediate
lobes of the pituitary gland are derived from a Prop1 expressing progenitor
(Davis et al. 2016).
POU1F1 was cloned based on its ability to bind critical elements of the GH and PRL
genes (Bodner et al. 1988; Ingraham et al. 1988). Pou1f1 is mutated in mice that
lack somatotropes, lactotropes and thyrotropes (Camper et al. 1990; Li et al.
1990). POU1F1 interacts with GATA2 to drive thyrotrope fate, and Gata2 may
have overlapping function with Gata3 (Charles et al. 2006; Dasen et al. 1999).
Somatotropes and lactotropes are very similar cell types derived from POU1F1
expressing progenitors. Epigentic regulation involving LSD1 and ZEB1 affects
the specialization of progenitors to somatotropes and lactotropes (Wang et al.
2007). POU1F1 acts at the nuclear matrix to regulate gene expression
(Skowronska-Krawczyk et al. 2014).
NR5A1 is essential for the development of adrenal and multiple organs of the
reproductive axis including the pituitary gland gonadotropes (Ingraham et al.
1994; Zhao et al. 2010). NR5A1 interacts with GATA2 to promote gonadotrope
fate (Charles et al. 2006; Dasen et al. 1999).
TBX19 is critical for melanotrope and corticotrope fate and it antagonizes the action
of NR5A1 in promoting gonadotrope fate (Lamolet et al. 2001; Pulichino et al.
2003). Homozygous deletion of Tpit results in almost complete loss of POMC in
both the anterior and intermediate lobes, and the intermediate lobe is hypoplastic
(Pulichino et al. 2003).
Pax7
/
mice lack intermediate lobe melanotropes, and the deficiency of Pax7
permits ectopic differentiation of progenitors into corticotropes (Budry et al.
2012). PAX7 is a pioneering transcription factor that opens chromatin permitting
access of transcription factors such as TPIT.
Single cell sequencing is a powerful technique for determining gene expression in
individual pituitary cell types without the need for genetic labeling or cell sorting
(Cheung et al. 2018b).
Pituitary organoids hold promise for studying pituitary development ex vivo
(Ozone et al. 2016; Suga et al. 2011).
6 Pituitary Development and Organogenesis: Transcription Factors in. . . 163

References
Achermann JC, Ito M, Ito M, Hindmarsh PC, Jameson JL (1999) A mutation in the gene encoding
steroidogenic factor-1 causes XY sex reversal and adrenal failure in humans. Nat Genet
22:125–126
Achermann JC, Ozisik G, Ito M, Orun UA, Harmanci K, Gurakan B, Jameson JL (2002) Gonadal
determination and adrenal development are regulated by the orphan nuclear receptor steroido-
genic factor-1, in a dose-dependent manner. J Clin Endocrinol Metab 87:1829–1833
Alvarado DM, McCall K, Aferol H, Silva MJ, Garbow JR, Spees WM, Patel T, Siegel M, Dobbs
MB, Gurnett CA (2011) Pitx1 haploinsufficiency causes clubfoot in humans and a clubfoot-like
phenotype in mice. Hum Mol Genet 20:3943–3952
Andersen B, Pearse RV 2nd, Jenne K, Sornson M, Lin SC, Bartke A, Rosenfeld MG (1995) The
Ames dwarf gene is required for Pit-1 gene activation. Dev Biol 172:495–503
Andoniadou CL, Matsushima D, Mousavy Gharavy SN, Signore M, Mackintosh AI, Schaeffer M,
Gaston-Massuet C, Mollard P, Jacques TS, Le Tissier P et al (2013) Sox2(+) stem/progenitor
cells in the adult mouse pituitary support organ homeostasis and have tumor-inducing potential.
Cell Stem Cell 13:433–445
Avilion AA, Nicolis SK, Pevny LH, Perez L, Vivian N, Lovell-Badge R (2003) Multipotent cell
lineages in early mouse development depend on SOX2 function. Genes Dev 17:126–140
Baek S, Sung MH (2016) Genome-scale analysis of cell-specific regulatory codes using nuclear
enzymes. Methods Mol Biol 1418:225–240
Baetens D, Stoop H, Peelman F, Todeschini AL, Rosseel T, Coppieters F, Veitia RA, Looijenga
LH, De Baere E, Cools M (2017) NR5A1 is a novel disease gene for 46,XX testicular and
ovotesticular disorders of sex development. Genet Med 19:367–376
Barnhart KM, Mellon PL (1994) The orphan nuclear receptor, steroidogenic factor-1, regulates the
glycoprotein hormone alpha-subunit gene in pituitary gonadotropes. Mol Endocrinol 8:878–885
Barr FG, Nauta LE, Davis RJ, Schafer BW, Nycum LM, Biegel JA (1996) In vivo amplification of
the PAX3-FKHR and PAX7-FKHR fusion genes in alveolar rhabdomyosarcoma. Hum Mol
Genet 5:15–21
Barski A, Cuddapah S, Cui K, Roh TY, Schones DE, Wang Z, Wei G, Chepelev I, Zhao K (2007)
High-resolution profiling of histone methylations in the human genome. Cell 129:823–837
Bartke A (1965) The response of two types of dwarf mice to growth hormone, thyrotropin, and
thyroxine. Gen Comp Endocrinol 5:418–426
Bashamboo A, Ferraz-de-Souza B, Lourenco D, Lin L, Sebire NJ, Montjean D, Bignon-Topalovic J,
Mandelbaum J, Siffroi JP, Christin-Maitre S et al (2010) Human male infertility associated with
mutations in NR5A1 encoding steroidogenic factor 1. Am J Hum Genet 87:505–512
Bashamboo A, Donohoue PA, Vilain E, Rojo S, Calvel P, Seneviratne SN, Buonocore F,
Barseghyan H, Bingham N, Rosenfeld JA et al (2016) A recurrent p.Arg92Trp variant in
steroidogenic factor-1 (NR5A1) can act as a molecular switch in human sex development.
Hum Mol Genet 25:5286
Bechtold-Dalla Pozza S, Hiedl S, Roeb J, Lohse P, Malik RE, Park S, Duran-Prado M, Rhodes SJ
(2012) A recessive mutation resulting in a disabling amino acid substitution (T194R) in the
LHX3 homeodomain causes combined pituitary hormone deficiency. Horm Res Paediatr
77:41–51
Bhangoo AP, Hunter CS, Savage JJ, Anhalt H, Pavlakis S, Walvoord EC, Ten S, Rhodes SJ (2006)
Clinical case seminar: a novel LHX3 mutation presenting as combined pituitary hormonal
deficiency. J Clin Endocrinol Metab 91:747–753
Biason-Lauber A, Schoenle EJ (2000) Apparently normal ovarian differentiation in a prepubertal
girl with transcriptionally inactive steroidogenic factor 1 (NR5A1/SF-1) and adrenocortical
insufficiency. Am J Hum Genet 67:1563–1568
Bilodeau S, Roussel-Gervais A, Drouin J (2009) Distinct developmental roles of cell cycle
inhibitors p57Kip2 and p27Kip1 distinguish pituitary progenitor cell cycle exit from cell
cycle reentry of differentiated cells. Mol Cell Biol 29:1895–1908
164 A. Z. Daly and S. A. Camper

Bodner M, Karin M (1987) A pituitary-specific trans-acting factor can stimulate transcription from
the growth hormone promoter in extracts of nonexpressing cells. Cell 50:267–275
Bodner M, Castrillo JL, Theill LE, Deerinck T, Ellisman M, Karin M (1988) The pituitary-specific
transcription factor GHF-1 is a homeobox-containing protein. Cell 55:505–518
Bonfig W, Krude H, Schmidt H (2011) A novel mutation of LHX3 is associated with combined
pituitary hormone deficiency including ACTH deficiency, sensorineural hearing loss, and short
neck-a case report and review of the literature. Eur J Pediatr 170:1017–1021
Borrelli E, Heyman RA, Arias C, Sawchenko PE, Evans RM (1989) Transgenic mice with inducible
dwarfism. Nature 339:538–541
Buckwalter MS, Katz RW, Camper SA (1991) Localization of the panhypopituitary dwarf mutation
(df) on mouse chromosome 11 in an intersubspecific backcross. Genomics 10:515–526
Budry L, Balsalobre A, Gauthier Y, Khetchoumian K, L’Honore A, Vallette S, Brue T, Figarella-
Branger D, Meij B, Drouin J (2012) The selector gene Pax7 dictates alternate pituitary cell fates
through its pioneer action on chromatin remodeling. Genes Dev 26:2299–2310
Buenrostro JD, Giresi PG, Zaba LC, Chang HY, Greenleaf WJ (2013) Transposition of native
chromatin for fast and sensitive epigenomic profiling of open chromatin, DNA-binding proteins
and nucleosome position. Nat Methods 10:1213–1218
Camper SA, Saunders TL, Katz RW, Reeves RH (1990) The Pit-1 transcription factor gene is a
candidate for the murine Snell dwarf mutation. Genomics 8:586–590
Camper SA, Saunders TL, Kendall SK, Keri RA, Seasholtz AF, Gordon DF, Birkmeier TS, Keegan
CE, Karolyi IJ, Roller ML et al (1995) Implementing transgenic and embryonic stem cell
technology to study gene expression, cell-cell interactions and gene function. Biol Reprod
52:246–257
Carreno G, Apps JR, Lodge EJ, Panousopoulos L, Haston S, Gonzalez-Meljem JM, Hahn H,
Andoniadou CL, Martinez-Barbera JP (2017) Hypothalamic sonic hedgehog is required for cell
specification and proliferation of LHX3/LHX4 pituitary embryonic precursors. Development
144:3289–3302
Carsner RL, Reynolds EG (1960) Primary site of gene action in anterior pituitary dwarf mice.
Science 131:829
Castinetti F, Reynaud R, Saveanu A, Jullien N, Quentien MH, Rochette C, Barlier A, Enjalbert A,
Brue T (2016) Mechanisms in endocrinology: an update in the genetic aetiologies of combined
pituitary hormone deficiency. Eur J Endocrinol 174(6):R239–R247
Cha KB, Douglas KR, Potok MA, Liang H, Jones SN, Camper SA (2004) WNT5A signaling affects
pituitary gland shape. Mech Dev 121:183–194
Chambers TJ, Giles A, Brabant G, Davis JR (2013) Wnt signalling in pituitary development and
tumorigenesis. Endocr Relat Cancer 20:R101–R111
Charles MA, Suh H, Hjalt TA, Drouin J, Camper SA, Gage PJ (2005) PITX genes are required for
cell survival and Lhx3 activation. Mol Endocrinol 19:1893–1903
Charles MA, Saunders TL, Wood WM, Owens K, Parlow AF, Camper SA, Ridgway EC, Gordon
DF (2006) Pituitary-specific Gata2 knockout: effects on gonadotrope and thyrotrope function.
Mol Endocrinol 20:1366–1377
Chen L, Gage PJ (2016) Heterozygous Pitx2 null mice accurately recapitulate the ocular features of
Axenfeld-Rieger syndrome and congenital glaucoma. Invest Ophthalmol Vis Sci 57:5023–5030
Chen J, Hersmus N, Van Duppen V, Caesens P, Denef C, Vankelecom H (2005) The adult pituitary
contains a cell population displaying stem/progenitor cell and early embryonic characteristics.
Endocrinology 146:3985–3998
Cheng TC, Beamer WG, Phillips JA 3rd, Bartke A, Mallonee RL, Dowling C (1983) Etiology of
growth hormone deficiency in little, Ames, and Snell dwarf mice. Endocrinology
113:1669–1678
Cheung L, Le Tissier P, Goldsmith SG, Treier M, Lovell-Badge R, Rizzoti K (2018a) NOTCH
activity differentially affects alternative cell fate acquisition and maintenance. Elife 7:e33318
6 Pituitary Development and Organogenesis: Transcription Factors in. . . 165

Cheung LYM, George AS, McGee SR, Daly AZ, Brinkmeier ML, Ellsworth BS, Camper SA
(2018b) Single-cell RNA sequencing reveals novel markers of male pituitary stem cells and
hormone-producing cell types. Endocrinology 159:3910–3924
Chou SJ, Hermesz E, Hatta T, Feltner D, El-Hodiri HM, Jamrich M, Mahon K (2006) Conserved
regulatory elements establish the dynamic expression of Rpx/HesxI in early vertebrate develop-
ment. Dev Biol 292:533–545
Cogan JD, Wu W, Phillips JA 3rd, Arnhold IJ, Agapito A, Fofanova OV, Osorio MG, Bircan I,
Moreno A, Mendonca BB (1998) The PROP1 2-base pair deletion is a common cause of
combined pituitary hormone deficiency. J Clin Endocrinol Metab 83:3346–3349
Cohen LE, Zanger K, Brue T, Wondisford FE, Radovick S (1999) Defective retinoic acid regulation
of the Pit-1 gene enhancer: a novel mechanism of combined pituitary hormone deficiency. Mol
Endocrinol 13:476–484
Cong L, Ran FA, Cox D, Lin S, Barretto R, Habib N, Hsu PD, Wu X, Jiang W, Marraffini LA et al
(2013) Multiplex genome engineering using CRISPR/Cas systems. Science 339:819–823
Correa RV, Domenice S, Bingham NC, Billerbeck AE, Rainey WE, Parker KL, Mendonca BB
(2004) A microdeletion in the ligand binding domain of human steroidogenic factor 1 causes
XY sex reversal without adrenal insufficiency. J Clin Endocrinol Metab 89:1767–1772
Couture C, Saveanu A, Barlier A, Carel JC, Fassnacht M, Fluck CE, Houang M, Maes M, Phan-
Hug F, Enjalbert A et al (2012) Phenotypic homogeneity and genotypic variability in a large
series of congenital isolated ACTH-deficiency patients with TPIT gene mutations. J Clin
Endocrinol Metab 97:E486–E495
Creyghton MP, Cheng AW, Welstead GG, Kooistra T, Carey BW, Steine EJ, Hanna J, Lodato MA,
Frampton GM, Sharp PA et al (2010) Histone H3K27ac separates active from poised enhancers
and predicts developmental state. Proc Natl Acad Sci USA 107:21931–21936
Cushman LJ, Burrows HL, Seasholtz AF, Lewandoski M, Muzyczka N, Camper SA (2000)
Cre-mediated recombination in the pituitary gland. Genesis 28:167–174
Dasen JS, O’Connell SM, Flynn SE, Treier M, Gleiberman AS, Szeto DP, Hooshmand F, Aggarwal
AK, Rosenfeld MG (1999) Reciprocal interactions of Pit1 and GATA2 mediate signaling
gradient-induced determination of pituitary cell types. Cell 97:587–598
Dasen JS, Martinez Barbera JP, Herman TS, Connell SO, Olson L, Ju B, Tollkuhn J, Baek SH, Rose
DW, Rosenfeld MG (2001) Temporal regulation of a paired-like homeodomain repressor/TLE
corepressor complex and a related activator is required for pituitary organogenesis. Genes Dev
15:3193–3207
Dateki S, Fukami M, Uematsu A, Kaji M, Iso M, Ono M, Mizota M, Yokoya S, Motomura K,
Kinoshita E et al (2010) Mutation and gene copy number analyses of six pituitary transcription
factor genes in 71 patients with combined pituitary hormone deficiency: identification of a single
patient with LHX4 deletion. J Clin Endocrinol Metab 95:4043–4047
Datson NA, Semina E, van Staalduinen AA, Dauwerse HG, Meershoek EJ, Heus JJ, Frants RR, den
Dunnen JT, Murray JC, van Ommen GJ (1996) Closing in on the Rieger syndrome gene on
4q25: mapping translocation breakpoints within a 50-kb region. Am J Hum Genet 59:1297–
1305
Dattani MT, Martinez-Barbera JP, Thomas PQ, Brickman JM, Gupta R, Martensson IL,
Toresson H, Fox M, Wales JK, Hindmarsh PC et al (1998) Mutations in the homeobox gene
HESX1/Hesx1 associated with septo-optic dysplasia in human and mouse. Nat Genet
19:125–133
Davis SW, Camper SA (2007) Noggin regulates Bmp4 activity during pituitary induction. Dev Biol
305:145–160
Davis SW, Mortensen AH, Camper SA (2011) Birthdating studies reshape models for pituitary
gland cell specification. Dev Biol 352:215–227
Davis SW, Keisler JL, Perez-Millan MI, Schade V, Camper SA (2016) All hormone-producing cell
types of the pituitary intermediate and anterior lobes derive from Prop1-expressing progenitors.
Endocrinology 157:1385–1396
Deladoey J, Fluck C, Buyukgebiz A, Kuhlmann BV, Eble A, Hindmarsh PC, Wu W, Mullis PE
(1999) “Hot spot” in the PROP1 gene responsible for combined pituitary hormone deficiency.
J Clin Endocrinol Metab 84:1645–1650
166 A. Z. Daly and S. A. Camper

DiMattia GE, Rhodes SJ, Krones A, Carriere C, O’Connell S, Kalla K, Arias C, Sawchenko P,
Rosenfeld MG (1997) The Pit-1 gene is regulated by distinct early and late pituitary-specific
enhancers. Dev Biol 182:180–190
Djakoure C, Guibourdenche J, Porquet D, Pagesy P, Peillon F, Li JY, Evain-Brion D (1996)
Vitamin A and retinoic acid stimulate within minutes cAMP release and growth hormone
secretion in human pituitary cells. J Clin Endocrinol Metab 81:3123–3126
Dusatkova P, Pfaffle R, Brown MR, Akulevich N, Arnhold IJ, Kalina MA, Kot K, Krzisnik C,
Lemos MC, Malikova J et al (2016) Genesis of two most prevalent PROP1 gene variants
causing combined pituitary hormone deficiency in 21 populations. Eur J Hum Genet
24:415–420
Duval DL, Nelson SE, Clay CM (1997) A binding site for steroidogenic factor-1 is part of a
complex enhancer that mediates expression of the murine gonadotropin-releasing hormone
receptor gene. Biol Reprod 56:160–168
Eicher EM, Beamer WG (1980) New mouse dw allele: genetic location and effects on lifespan and
growth hormone levels. J Hered 71:187–190
Ellsworth BS, Butts DL, Camper SA (2008) Mechanisms underlying pituitary hypoplasia and failed
cell specification in Lhx3-deficient mice. Dev Biol 313:118–129
Enwright JF 3rd, Kawecki-Crook MA, Voss TC, Schaufele F, Day RN (2003) A PIT-1
homeodomain mutant blocks the intranuclear recruitment of the CCAAT/enhancer binding
protein alpha required for prolactin gene transcription. Mol Endocrinol 17:209–222
Ericson J, Norlin S, Jessell TM, Edlund T (1998) Integrated FGF and BMP signaling controls the
progression of progenitor cell differentiation and the emergence of pattern in the embryonic
anterior pituitary. Development 125:1005–1015
Fang Q, George AS, Brinkmeier ML, Mortensen AH, Gergics P, Cheung LY, Daly AZ, Ajmal A,
Perez Millan MI, Ozel AB et al (2016) Genetics of combined pituitary hormone deficiency:
roadmap into the genome era. Endocr Rev 37:636–675
Fauquier T, Guerineau NC, McKinney RA, Bauer K, Mollard P (2001) Folliculostellate cell
network: a route for long-distance communication in the anterior pituitary. Proc Natl Acad
Sci USA 98:8891–8896
Fauquier T, Rizzoti K, Dattani M, Lovell-Badge R, Robinson IC (2008) SOX2-expressing progen-
itor cells generate all of the major cell types in the adult mouse pituitary gland. Proc Natl Acad
Sci USA 105:2907–2912
Fekete C, Lechan RM (2014) Central regulation of hypothalamic-pituitary-thyroid axis under
physiological and pathophysiological conditions. Endocr Rev 35:159–194
Filges I, Bischof-Renner A, Rothlisberger B, Potthoff C, Glanzmann R, Gunthard J, Schneider J,
Huber AR, Zumsteg U, Miny P et al (2012) Panhypopituitarism presenting as life-threatening
heart failure caused by an inherited microdeletion in 1q25 including LHX4. Pediatrics 129:
e529–e534
Fofanova OV, Takamura N, Kinoshita E, Yoshimoto M, Tsuji Y, Peterkova VA, Evgrafov OV,
Dedov II, Goncharov NP, Yamashita S (1998) Rarity of PIT1 involvement in children from
Russia with combined pituitary hormone deficiency. Am J Med Genet 77:360–365
Frawley LS, Boockfor FR, Hoeffler JP (1985) Identification by plaque assays of a pituitary cell type
that secretes both growth hormone and prolactin. Endocrinology 116:734–737
Gage PJ, Camper SA (1997) Pituitary homeobox 2, a novel member of the bicoid-related family of
homeobox genes, is a potential regulator of anterior structure formation. Hum Mol Genet
6:457–464
Gage PJ, Lossie AC, Scarlett LM, Lloyd RV, Camper SA (1995) Ames dwarf mice exhibit
somatotrope commitment but lack growth hormone-releasing factor response. Endocrinology
136:1161–1167
Gage PJ, Brinkmeier ML, Scarlett LM, Knapp LT, Camper SA, Mahon KA (1996) The Ames dwarf
gene, df, is required early in pituitary ontogeny for the extinction of Rpx transcription and
initiation of lineage-specific cell proliferation. Mol Endocrinol 10:1570–1581
6 Pituitary Development and Organogenesis: Transcription Factors in. . . 167

Gage PJ, Suh H, Camper SA (1999) Dosage requirement of Pitx2 for development of multiple
organs. Development 126:4643–4651
Gaston-Massuet C, Andoniadou CL, Signore M, Jayakody SA, Charolidi N, Kyeyune R, Vernay B,
Jacques TS, Taketo MM, Le Tissier P et al (2011) Increased wingless (Wnt) signaling in
pituitary progenitor/stem cells gives rise to pituitary tumors in mice and humans. Proc Natl
Acad Sci USA 108:11482–11487
Gergics P, Brinkmeier ML, Camper SA (2015) Lhx4 deficiency: increased cyclin-dependent kinase
inhibitor expression and pituitary hypoplasia. Mol Endocrinol 29:597–612
Giacomini D, Paez-Pereda M, Theodoropoulou M, Gerez J, Nagashima AC, Chervin A, Berner S,
Labeur M, Refojo D, Renner U et al (2006) Bone morphogenetic protein-4 control of pituitary
pathophysiology. Front Horm Res 35:22–31
Goldberg LB, Aujla PK, Raetzman LT (2011) Persistent expression of activated Notch inhibits
corticotrope and melanotrope differentiation and results in dysfunction of the HPA axis. Dev
Biol 358:23–32
Gordon DF, Lewis SR, Haugen BR, James RA, McDermott MT, Wood WM, Ridgway EC (1997)
Pit-1 and GATA-2 interact and functionally cooperate to activate the thyrotropin beta-subunit
promoter. J Biol Chem 272:24339–24347
Gordon DF, Tucker EA, Tundwal K, Hall H, Wood WM, Ridgway EC (2006) MED220/thyroid
receptor-associated protein 220 functions as a transcriptional coactivator with Pit-1 and GATA-
2 on the thyrotropin-beta promoter in thyrotropes. Mol Endocrinol 20:1073–1089
Gregory LC, Humayun KN, Turton JP, McCabe MJ, Rhodes SJ, Dattani MT (2015) Novel lethal
form of congenital hypopituitarism associated with the first recessive LHX4 mutation. J Clin
Endocrinol Metab 100:2158–2164
Gubbay J, Collignon J, Koopman P, Capel B, Economou A, Munsterberg A, Vivian N,
Goodfellow P, Lovell-Badge R (1990) A gene mapping to the sex-determining region of the
mouse Y chromosome is a member of a novel family of embryonically expressed genes. Nature
346:245–250
Guner B, Ozacar AT, Thomas JE, Karlstrom RO (2008) Graded hedgehog and fibroblast growth
factor signaling independently regulate pituitary cell fates and help establish the pars distalis and
pars intermedia of the zebrafish adenohypophysis. Endocrinology 149:4435–4451
Guran T, Buonocore F, Saka N, Ozbek MN, Aycan Z, Bereket A, Bas F, Darcan S, Bideci A, Guven
A et al (2016) Rare causes of primary adrenal insufficiency: genetic and clinical characterization
of a large Nationwide cohort. J Clin Endocrinol Metab 101:284–292
Gurnett CA, Alaee F, Kruse LM, Desruisseau DM, Hecht JT, Wise CA, Bowcock AM, Dobbs MB
(2008) Asymmetric lower-limb malformations in individuals with homeobox PITX1 gene
mutation. Am J Hum Genet 83:616–622
Halvorson LM, Kaiser UB, Chin WW (1996) Stimulation of luteinizing hormone beta gene
promoter activity by the orphan nuclear receptor, steroidogenic factor-1. J Biol Chem
271:6645–6650
Hasegawa T, Fukami M, Sato N, Katsumata N, Sasaki G, Fukutani K, Morohashi K, Ogata T
(2004) Testicular dysgenesis without adrenal insufficiency in a 46,XY patient with a heterozy-
gous inactive mutation of steroidogenic factor-1. J Clin Endocrinol Metab 89:5930–5935
Hermesz E, Mackem S, Mahon KA (1996) Rpx: a novel anterior-restricted homeobox gene
progressively activated in the prechordal plate, anterior neural plate and Rathke’s pouch of
the mouse embryo. Development 122:41–52
Higham CE, Johannsson G, Shalet SM (2016) Hypopituitarism. Lancet 388:2403–2415
Himes AD, Raetzman LT (2009) Premature differentiation and aberrant movement of pituitary cells
lacking both Hes1 and Prop1. Dev Biol 325:151–161
Ho Y, Tadevosyan A, Liebhaber SA, Cooke NE (2008) The juxtaposition of a promoter with a
locus control region transcriptional domain activates gene expression. EMBO Rep 9:891–898
Ho Y, Shewchuk BM, Liebhaber SA, Cooke NE (2013) Distinct chromatin configurations regulate
the initiation and the maintenance of hGH gene expression. Mol Cell Biol 33:1723–1734
168 A. Z. Daly and S. A. Camper

Ho Y, Cooke NE, Liebhaber SA (2015) An autoregulatory pathway establishes the definitive

chromatin conformation at the pit-1 locus. Mol Cell Biol 35:1523–1532
Honda S, Morohashi K, Omura T (1990) Novel cAMP regulatory elements in the promoter region
of bovine P-450(11 beta) gene. J Biochem 108:1042–1049
Honda S, Morohashi K, Nomura M, Takeya H, Kitajima M, Omura T (1993) Ad4BP regulating
steroidogenic P-450 gene is a member of steroid hormone receptor superfamily. J Biol Chem
268:7494–7502
Horn F, Windle JJ, Barnhart KM, Mellon PL (1992) Tissue-specific gene expression in the
pituitary: the glycoprotein hormone alpha-subunit gene is regulated by a gonadotrope-specific
protein. Mol Cell Biol 12:2143–2153
Horvath E, Lloyd RV, Kovacs K (1990) Propylthiouracyl-induced hypothyroidism results in
reversible transdifferentiation of somatotrophs into thyroidectomy cells. A morphologic study
of the rat pituitary including immunoelectron microscopy. Lab Investig 63:511–520
Hosoyama T, Nishijo K, Garcia MM, Schaffer BS, Ohshima-Hosoyama S, Prajapati SI, Davis MD,
Grant WF, Scheithauer BW, Marks DL et al (2010) A postnatal Pax7 progenitor gives rise to
pituitary adenomas. Genes Cancer 1:388–402
Igarashi M, Takasawa K, Hakoda A, Kanno J, Takada S, Miyado M, Baba T, Morohashi KI,
Tajima T, Hata K et al (2017) Identical NR5A1 missense mutations in two unrelated 46,XX
individuals with testicular tissues. Hum Mutat 38:39–42
Ikeda Y, Shen WH, Ingraham HA, Parker KL (1994) Developmental expression of mouse steroido-
genic factor-1, an essential regulator of the steroid hydroxylases. Mol Endocrinol 8:654–662
Ikeda Y, Luo X, Abbud R, Nilson JH, Parker KL (1995) The nuclear receptor steroidogenic factor
1 is essential for the formation of the ventromedial hypothalamic nucleus. Mol Endocrinol
9:478–486
Imam A, Iqbal J, Blair HC, Davies TF, Huang CL, Zallone A, Zaidi M, Sun L (2009) Role of the
pituitary-bone axis in skeletal pathophysiology. Curr Opin Endocrinol Diabetes Obes
16:423–429
Ingraham HA, Chen RP, Mangalam HJ, Elsholtz HP, Flynn SE, Lin CR, Simmons DM, Swanson L,
Rosenfeld MG (1988) A tissue-specific transcription factor containing a homeodomain specifies
a pituitary phenotype. Cell 55:519–529
Ingraham HA, Lala DS, Ikeda Y, Luo X, Shen WH, Nachtigal MW, Abbud R, Nilson JH, Parker
KL (1994) The nuclear receptor steroidogenic factor 1 acts at multiple levels of the reproductive
axis. Genes Dev 8:2302–2312
Jacobs SB, Coss D, McGillivray SM, Mellon PL (2003) Nuclear factor Y and steroidogenic factor
1 physically and functionally interact to contribute to cell-specific expression of the mouse
follicle-stimulating hormone-beta gene. Mol Endocrinol 17:1470–1483
Jayakody SA, Andoniadou CL, Gaston-Massuet C, Signore M, Cariboni A, Bouloux PM, Le
Tissier P, Pevny LH, Dattani MT, Martinez-Barbera JP (2012) SOX2 regulates the
hypothalamic-pituitary axis at multiple levels. J Clin Invest 122:3635–3646
Jin JM, Yang WX (2014) Molecular regulation of hypothalamus-pituitary-gonads axis in males.
Gene 551:15–25
Jones BK, Monks BR, Liebhaber SA, Cooke NE (1995) The human growth hormone gene is
regulated by a multicomponent locus control region. Mol Cell Biol 15:7010–7021
Jostes B, Walther C, Gruss P (1990) The murine paired box gene, Pax7, is expressed specifically
during the development of the nervous and muscular system. Mech Dev 33:27–37
Jullien N, Romanet P, Philippon M, Quentien MH, Beck-Peccoz P, Bergada I, Odent S, Reynaud R,
Barlier A, Saveanu A et al (2019) Heterozygous LHX3 mutations may lead to a mild phenotype
of combined pituitary hormone deficiency. Eur J Hum Genet 27:216–225
Kaiser UB, Halvorson LM, Chen MT (2000) Sp1, steroidogenic factor 1 (SF-1), and early growth
response protein 1 (egr-1) binding sites form a tripartite gonadotropin-releasing hormone
response element in the rat luteinizing hormone-beta gene promoter: an integral role for SF-1.
Mol Endocrinol 14:1235–1245
6 Pituitary Development and Organogenesis: Transcription Factors in. . . 169

Karpova T, Presley J, Manimaran RR, Scherrer SP, Tejada L, Peterson KR, Heckert LL (2005) A
FTZ-F1-containing yeast artificial chromosome recapitulates expression of steroidogenic factor
1 in vivo. Mol Endocrinol 19:2549–2563
Kassar-Duchossoy L, Giacone E, Gayraud-Morel B, Jory A, Gomes D, Tajbakhsh S (2005) Pax3/
Pax7 mark a novel population of primitive myogenic cells during development. Genes Dev
19:1426–1431
Kelberman D, Rizzoti K, Avilion A, Bitner-Glindzicz M, Cianfarani S, Collins J, Chong WK, Kirk
JM, Achermann JC, Ross R et al (2006) Mutations within Sox2/SOX2 are associated with
abnormalities in the hypothalamo-pituitary-gonadal axis in mice and humans. J Clin Invest
116:2442–2455
Kelberman D, Rizzoti K, Lovell-Badge R, Robinson IC, Dattani MT (2009) Genetic regulation of
pituitary gland development in human and mouse. Endocr Rev 30:790–829
Keller-Wood M (2015) Hypothalamic-pituitary—adrenal axis-feedback control. Compr Physiol
5:1161–1182
Keri RA, Nilson JH (1996) A steroidogenic factor-1 binding site is required for activity of the
luteinizing hormone beta subunit promoter in gonadotropes of transgenic mice. J Biol Chem
271:10782–10785
Keri RA, Wolfe MW, Saunders TL, Anderson I, Kendall SK, Wagner T, Yeung J, Gorski J, Nett
TM, Camper SA et al (1994) The proximal promoter of the bovine luteinizing hormone beta-
subunit gene confers gonadotrope-specific expression and regulation by gonadotropin-releasing
hormone, testosterone, and 17 beta-estradiol in transgenic mice. Mol Endocrinol 8:1807–1816
Kioussi C, Briata P, Baek SH, Rose DW, Hamblet NS, Herman T, Ohgi KA, Lin C, Gleiberman A,
Wang J et al (2002) Identification of a Wnt/Dvl/beta-catenin!Pitx2 pathway mediating cell-
type-specific proliferation during development. Cell 111:673–685
Kitamura K, Miura H, Miyagawa-Tomita S, Yanazawa M, Katoh-Fukui Y, Suzuki R, Ohuchi H,
Suehiro A, Motegi Y, Nakahara Y et al (1999) Mouse Pitx2 deficiency leads to anomalies of the
ventral body wall, heart, extra- and periocular mesoderm and right pulmonary isomerism.
Development 126:5749–5758
Klopocki E, Kahler C, Foulds N, Shah H, Joseph B, Vogel H, Luttgen S, Bald R, Besoke R, Held K
et al (2012) Deletions in PITX1 cause a spectrum of lower-limb malformations including mirror-
image polydactyly. Eur J Hum Genet 20:705–708
Kragesteen BK, Brancati F, Digilio MC, Mundlos S, Spielmann M (2019) H2AFY promoter
deletion causes PITX1 endoactivation and Liebenberg syndrome. J Med Genet 56:246–251
Kristrom B, Zdunek AM, Rydh A, Jonsson H, Sehlin P, Escher SA (2009) A novel mutation in the
LIM homeobox 3 gene is responsible for combined pituitary hormone deficiency, hearing
impairment, and vertebral malformations. J Clin Endocrinol Metab 94:1154–1161
Kulig E, Camper SA, Kuecker S, Jin L, Lloyd RV (1998) Remodeling of hyperplastic pituitaries in
hypothyroid us-subunit knockout mice after thyroxine and 1713-estradiol treatment: role of
apoptosis. Endocr Pathol 9:261–274
Lafont C, Desarmenien MG, Cassou M, Molino F, Lecoq J, Hodson D, Lacampagne A,
Mennessier G, El Yandouzi T, Carmignac D et al (2010) Cellular in vivo imaging reveals
coordinated regulation of pituitary microcirculation and GH cell network function. Proc Natl
Acad Sci USA 107:4465–4470
Lala DS, Rice DA, Parker KL (1992) Steroidogenic factor I, a key regulator of steroidogenic
enzyme expression, is the mouse homolog of fushi tarazu-factor I. Mol Endocrinol 6:1249–1258
Lamolet B, Pulichino AM, Lamonerie T, Gauthier Y, Brue T, Enjalbert A, Drouin J (2001) A
pituitary cell-restricted T box factor, Tpit, activates POMC transcription in cooperation with Pitx
homeoproteins. Cell 104:849–859
Lamonerie T, Tremblay JJ, Lanctot C, Therrien M, Gauthier Y, Drouin J (1996) Ptx1, a bicoid-
related homeo box transcription factor involved in transcription of the pro-opiomelanocortin
gene. Genes Dev 10:1284–1295
Lanctot C, Gauthier Y, Drouin J (1999) Pituitary homeobox 1 (Ptx1) is differentially expressed
during pituitary development. Endocrinology 140:1416–1422
170 A. Z. Daly and S. A. Camper

Langlais D, Couture C, Balsalobre A, Drouin J (2012) The Stat3/GR interaction code: predictive
value of direct/indirect DNA recruitment for transcription outcome. Mol Cell 47:38–49
Lebl J, Vosahlo J, Pfaeffle RW, Stobbe H, Cerna J, Novotna D, Zapletalova J, Kalvachova B,
Hana V, Weiss V et al (2005) Auxological and endocrine phenotype in a population-based
cohort of patients with PROP1 gene defects. Eur J Endocrinol 153:389–396
Lee SL, Sadovsky Y, Swirnoff AH, Polish JA, Goda P, Gavrilina G, Milbrandt J (1996) Luteinizing
hormone deficiency and female infertility in mice lacking the transcription factor NGFI-A
(Egr-1). Science 273:1219–1221
Lemos MC, Gomes L, Bastos M, Leite V, Limbert E, Carvalho D, Bacelar C, Monteiro M,
Fonseca F, Agapito A et al (2006) PROP1 gene analysis in Portuguese patients with combined
pituitary hormone deficiency. Clin Endocrinol 65:479–485
Li S, Crenshaw EB 3rd, Rawson EJ, Simmons DM, Swanson LW, Rosenfeld MG (1990) Dwarf
locus mutants lacking three pituitary cell types result from mutations in the POU-domain gene
pit-1. Nature 347:528–533
Li H, Witte DP, Branford WW, Aronow BJ, Weinstein M, Kaur S, Wert S, Singh G, Schreiner CM,
Whitsett JA et al (1994) Gsh-4 encodes a LIM-type homeodomain, is expressed in the develop-
ing central nervous system and is required for early postnatal survival. EMBO J 13:2876–2885
Li M, Pevny L, Lovell-Badge R, Smith A (1998) Generation of purified neural precursors from
embryonic stem cells by lineage selection. Curr Biol 8:971–974
Lin CR, Kioussi C, O’Connell S, Briata P, Szeto D, Liu F, Izpisua-Belmonte JC, Rosenfeld MG
(1999) Pitx2 regulates lung asymmetry, cardiac positioning and pituitary and tooth morphogen-
esis. Nature 401:279–282
Lin L, Philibert P, Ferraz-de-Souza B, Kelberman D, Homfray T, Albanese A, Molini V, Sebire NJ,
Einaudi S, Conway GS et al (2007) Heterozygous missense mutations in steroidogenic factor
1 (SF1/Ad4BP, NR5A1) are associated with 46,XY disorders of sex development with normal
adrenal function. J Clin Endocrinol Metab 92:991–999
Lodge EJ, Santambrogio A, Russell JP, Xekouki P, Jacques TS, Johnson RL, Thavaraj S, Bornstein
SR, Andoniadou CL (2019) Homeostatic and tumourigenic activity of SOX2+ pituitary stem
cells is controlled by the LATS/YAP/TAZ cascade. Elife 8:e43996
Lourenco D, Brauner R, Lin L, De Perdigo A, Weryha G, Muresan M, Boudjenah R, Guerra-
Junior G, Maciel-Guerra AT, Achermann JC et al (2009) Mutations in NR5A1 associated with
ovarian insufficiency. N Engl J Med 360:1200–1210
Lu MF, Pressman C, Dyer R, Johnson RL, Martin JF (1999) Function of Rieger syndrome gene in
left-right asymmetry and craniofacial development. Nature 401:276–278
Luo X, Ikeda Y, Parker KL (1994) A cell-specific nuclear receptor is essential for adrenal and
gonadal development and sexual differentiation. Cell 77:481–490
Mallet D, Bretones P, Michel-Calemard L, Dijoud F, David M, Morel Y (2004) Gonadal dysgenesis
without adrenal insufficiency in a 46, XY patient heterozygous for the nonsense C16X mutation:
a case of SF1 haploinsufficiency. J Clin Endocrinol Metab 89:4829–4832
Mansouri A, Gruss P (1998) Pax3 and Pax7 are expressed in commissural neurons and restrict
ventral neuronal identity in the spinal cord. Mech Dev 78:171–178
Mansouri A, Stoykova A, Torres M, Gruss P (1996) Dysgenesis of cephalic neural crest derivatives
in Pax7-/- mutant mice. Development 122:831–838
Matsumoto A, Ishii S (1992) Atlas of endocrine organs: vertebrates and invertebrates. Springer,
Berlin
Mayran A, Khetchoumian K, Hariri F, Pastinen T, Gauthier Y, Balsalobre A, Drouin J (2018)
Pioneer factor Pax7 deploys a stable enhancer repertoire for specification of cell fate. Nat Genet
50:259–269
McCormick A, Wu D, Castrillo JL, Dana S, Strobl J, Thompson EB, Karin M (1988) Extinction of
growth hormone expression in somatic cell hybrids involves repression of the specific trans-
activator GHF-1. Cell 55:379–389
6 Pituitary Development and Organogenesis: Transcription Factors in. . . 171

Melamed P, Zhu Y, Tan SH, Xie M, Koh M (2006) Gonadotropin-releasing hormone activation of
c-jun, but not early growth response factor-1, stimulates transcription of a luteinizing hormone
beta-subunit gene. Endocrinology 147:3598–3605
Melmed S (2011) Pathogenesis of pituitary tumors. Nat Rev Endocrinol 7:257–266
Metherell LA, Savage MO, Dattani M, Walker J, Clayton PE, Farooqi IS, Clark AJ (2004) TPIT
mutations are associated with early-onset, but not late-onset isolated ACTH deficiency. Eur J
Endocrinol 151:463–465
Mollard P, Hodson DJ, Lafont C, Rizzoti K, Drouin J (2012) A tridimensional view of pituitary
development and function. Trends Endocrinol Metab 23:261–269
Morohashi K, Honda S, Inomata Y, Handa H, Omura T (1992) A common trans-acting factor,
Ad4-binding protein, to the promoters of steroidogenic P-450s. J Biol Chem 267:17913–17919
Mortensen AH, MacDonald JW, Ghosh D, Camper SA (2011) Candidate genes for
panhypopituitarism identified by gene expression profiling. Physiol Genomics 43:1105–1116
Nakano T, Ando S, Takata N, Kawada M, Muguruma K, Sekiguchi K, Saito K, Yonemura S,
Eiraku M, Sasai Y (2012) Self-formation of optic cups and storable stratified neural retina from
human ESCs. Cell Stem Cell 10:771–785
Nasonkin IO, Ward RD, Raetzman LT, Seasholtz AF, Saunders TL, Gillespie PJ, Camper SA
(2004) Pituitary hypoplasia and respiratory distress syndrome in Prop1 knockout mice. Hum
Mol Genet 13:2727–2735
Navardauskaite R, Dusatkova P, Obermannova B, Pfaeffle RW, Blum WF, Adukauskiene D,
Smetanina N, Cinek O, Verkauskiene R, Lebl J (2014) High prevalence of PROP1 defects in
Lithuania: phenotypic findings in an ethnically homogenous cohort of patients with multiple
pituitary hormone deficiency. J Clin Endocrinol Metab 99:299–306
Navin N, Kendall J, Troge J, Andrews P, Rodgers L, McIndoo J, Cook K, Stepansky A, Levy D,
Esposito D et al (2011) Tumour evolution inferred by single-cell sequencing. Nature 472:90–94
Nelson C, Crenshaw EB 3rd, Franco R, Lira SA, Albert VR, Evans RM, Rosenfeld MG (1986)
Discrete cis-active genomic sequences dictate the pituitary cell type-specific expression of rat
prolactin and growth hormone genes. Nature 322:557–562
Netchine I, Sobrier ML, Krude H, Schnabel D, Maghnie M, Marcos E, Duriez B, Cacheux V,
Moers A, Goossens M et al (2000) Mutations in LHX3 result in a new syndrome revealed by
combined pituitary hormone deficiency. Nat Genet 25:182–186
Ohta K, Nobukuni Y, Mitsubuchi H, Fujimoto S, Matsuo N, Inagaki H, Endo F, Matsuda I (1992)
Mutations in the Pit-1 gene in children with combined pituitary hormone deficiency. Biochem
Biophys Res Commun 189:851–855
Ohuchi H, Hori Y, Yamasaki M, Harada H, Sekine K, Kato S, Itoh N (2000) FGF10 acts as a major
ligand for FGF receptor 2 IIIb in mouse multi-organ development. Biochem Biophys Res
Commun 277:643–649
Okamoto N, Wada Y, Ida S, Koga R, Ozono K, Chiyo H, Hayashi A, Tatsumi K (1994) Monoallelic
expression of normal mRNA in the PIT1 mutation heterozygotes with normal phenotype and
biallelic expression in the abnormal phenotype. Hum Mol Genet 3:1565–1568
Olson LE, Tollkuhn J, Scafoglio C, Krones A, Zhang J, Ohgi KA, Wu W, Taketo MM, Kemler R,
Grosschedl R et al (2006) Homeodomain-mediated beta-catenin-dependent switching events
dictate cell-lineage determination. Cell 125:593–605
Oustanina S, Hause G, Braun T (2004) Pax7 directs postnatal renewal and propagation of myogenic
satellite cells but not their specification. EMBO J 23:3430–3439
Ozone C, Suga H, Eiraku M, Kadoshima T, Yonemura S, Takata N, Oiso Y, Tsuji T, Sasai Y (2016)
Functional anterior pituitary generated in self-organizing culture of human embryonic stem
cells. Nat Commun 7:10351
Peel MT, Ho Y, Liebhaber SA (2018) Transcriptome analyses of female Somatotropes and
Lactotropes reveal novel regulators of cell identity in the pituitary. Endocrinology
159:3965–3980
Perez Millan MI, Brinkmeier ML, Mortensen AH, Camper SA (2016) PROP1 triggers epithelial-
mesenchymal transition-like process in pituitary stem cells. Elife 5:e14470
172 A. Z. Daly and S. A. Camper

Perez C, Dastot-Le Moal F, Collot N, Legendre M, Abadie I, Bertrand AM, Amselem S, Sobrier
ML (2012) Screening of LHX2 in patients presenting growth retardation with posterior pituitary
and ocular abnormalities. Eur J Endocrinol 167:85–91
Pfaeffle RW, Savage JJ, Hunter CS, Palme C, Ahlmann M, Kumar P, Bellone J, Schoenau E,
Korsch E, Bramswig JH et al (2007) Four novel mutations of the LHX3 gene cause combined
pituitary hormone deficiencies with or without limited neck rotation. J Clin Endocrinol Metab
92:1909–1919
Pfaeffle RW, Hunter CS, Savage JJ, Duran-Prado M, Mullen RD, Neeb ZP, Eiholzer U, Hesse V,
Haddad NG, Stobbe HM et al (2008) Three novel missense mutations within the LHX4 gene are
associated with variable pituitary hormone deficiencies. J Clin Endocrinol Metab 93:1062–1071
Phifer RF, Midgley AR, Spicer SS (1973) Immunohistologic and histologic evidence that follicle-
stimulating hormone and luteinizing hormone are present in the same cell type in the human pars
distalis. J Clin Endocrinol Metab 36:125–141
Porter FD, Drago J, Xu Y, Cheema SS, Wassif C, Huang SP, Lee E, Grinberg A, Massalas JS,
Bodine D et al (1997) Lhx2, a LIM homeobox gene, is required for eye, forebrain, and definitive
erythrocyte development. Development 124:2935–2944
Poulin G, Turgeon B, Drouin J (1997) NeuroD1/beta2 contributes to cell-specific transcription of
the proopiomelanocortin gene. Mol Cell Biol 17:6673–6682
Poulin G, Lebel M, Chamberland M, Paradis FW, Drouin J (2000) Specific protein-protein
interaction between basic helix-loop-helix transcription factors and homeoproteins of the Pitx
family. Mol Cell Biol 20:4826–4837
Pow DV, Perry VH, Morris JF, Gordon S (1989) Microglia in the neurohypophysis associate with
and endocytose terminal portions of neurosecretory neurons. Neuroscience 33:567–578
Pulichino AM, Vallette-Kasic S, Couture C, Gauthier Y, Brue T, David M, Malpuech G, Deal C,
Van Vliet G, De Vroede M et al (2003a) Human and mouse TPIT gene mutations cause early
onset pituitary ACTH deficiency. Genes Dev 17:711–716
Pulichino AM, Vallette-Kasic S, Tsai JP, Couture C, Gauthier Y, Drouin J (2003b) Tpit determines
alternate fates during pituitary cell differentiation. Genes Dev 17:738–747
Qiao S, Nordstrom K, Muijs L, Gasparoni G, Tierling S, Krause E, Walter J, Boehm U (2016)
Molecular plasticity of male and female murine gonadotropes revealed by mRNA sequencing.
Endocrinology 157:1082–1093
Radovick S, Nations M, Du Y, Berg LA, Weintraub BD, Wondisford FE (1992) A mutation in the
POU-homeodomain of Pit-1 responsible for combined pituitary hormone deficiency. Science
257:1115–1118
Raetzman LT, Ward R, Camper SA (2002) Lhx4 and Prop1 are required for cell survival and
expansion of the pituitary primordia. Development 129:4229–4239
Raetzman LT, Ross SA, Cook S, Dunwoodie SL, Camper SA, Thomas PQ (2004) Developmental
regulation of Notch signaling genes in the embryonic pituitary: Prop1 deficiency affects Notch2
expression. Dev Biol 265:329–340
Ragge NK, Lorenz B, Schneider A, Bushby K, de Sanctis L, de Sanctis U, Salt A, Collin JR, Vivian
AJ, Free SL et al (2005) SOX2 anophthalmia syndrome. Am J Med Genet A 135:1–7.
discussion 8
Rajab A, Kelberman D, de Castro SC, Biebermann H, Shaikh H, Pearce K, Hall CM, Shaikh G,
Gerrelli D, Grueters A et al (2008) Novel mutations in LHX3 are associated with hypopituita-
rism and sensorineural hearing loss. Hum Mol Genet 17:2150–2159
Ramzan K, Bin-Abbas B, Al-Jomaa L, Allam R, Al-Owain M, Imtiaz F (2017) Two novel LHX3
mutations in patients with combined pituitary hormone deficiency including cervical rigidity
and sensorineural hearing loss. BMC Endocr Disord 17:17
Rappaport R, Amselem S (2001) Hypothalamic-pituitary development: genetic and clinical aspects.
Karger, Basel
Regal M, Paramo C, Sierra SM, Garcia-Mayor RV (2001) Prevalence and incidence of hypopitu-
itarism in an adult Caucasian population in northwestern Spain. Clin Endocrinol 55:735–740
6 Pituitary Development and Organogenesis: Transcription Factors in. . . 173

Relaix F, Rocancourt D, Mansouri A, Buckingham M (2005) A Pax3/Pax7-dependent population

of skeletal muscle progenitor cells. Nature 435:948–953
Relaix F, Montarras D, Zaffran S, Gayraud-Morel B, Rocancourt D, Tajbakhsh S, Mansouri A,
Cumano A, Buckingham M (2006) Pax3 and Pax7 have distinct and overlapping functions in
adult muscle progenitor cells. J Cell Biol 172:91–102
Rhodes SJ, Chen R, DiMattia GE, Scully KM, Kalla KA, Lin SC, Yu VC, Rosenfeld MG (1993) A
tissue-specific enhancer confers Pit-1-dependent morphogen inducibility and autoregulation on
the pit-1 gene. Genes Dev 7:913–932
Rice DA, Mouw AR, Bogerd AM, Parker KL (1991) A shared promoter element regulates the
expression of three steroidogenic enzymes. Mol Endocrinol 5:1552–1561
Riepe FG, Partsch CJ, Blankenstein O, Monig H, Pfaffle RW, Sippell WG (2001) Longitudinal
imaging reveals pituitary enlargement preceding hypoplasia in two brothers with combined
pituitary hormone deficiency attributable to PROP1 mutation. J Clin Endocrinol Metab
86:4353–4357
Rinehart JF, Farquhar MG (1953) Electron microscopic studies of the anterior pituitary gland.
J Histochem Cytochem 1:93–113
Rizzoti K (2015) Genetic regulation of murine pituitary development. J Mol Endocrinol 54:R55–
R73
Rizzoti K, Lovell-Badge R (2005) Early development of the pituitary gland: induction and shaping
of Rathke’s pouch. Rev Endocr Metab Disord 6:161–172
Rizzoti K, Akiyama H, Lovell-Badge R (2013) Mobilized adult pituitary stem cells contribute to
endocrine regeneration in response to physiological demand. Cell Stem Cell 13:419–432
Robyn C, Leleux P, Vanhaelst L, Golstein J, Herlant M, Pasteels JL (1973) Immunohistochemical
study of the human pituitary with anti-luteinizing hormone, anti-follicle stimulating hormone
and anti-thyrotrophin sera. Acta Endocrinol 72:625–642
Rochette C, Jullien N, Saveanu A, Caldagues E, Bergada I, Braslavsky D, Pfeifer M, Reynaud R,
Herman JP, Barlier A et al (2015) Identifying the deleterious effect of rare LHX4 allelic variants,
a challenging issue. PLoS One 10:e0126648
Roh M, Paterson AJ, Asa SL, Chin E, Kudlow JE (2001) Stage-sensitive blockade of pituitary
somatomammotrope development by targeted expression of a dominant negative epidermal
growth factor receptor in transgenic mice. Mol Endocrinol 15:600–613
Rosso L, Mienville JM (2009) Pituicyte modulation of neurohormone output. Glia 57:235–243
Ruf-Zamojski F, Fribourg M, Ge Y, Nair V, Pincas H, Zaslavsky E, Nudelman G, Tuminello SJ,
Watanabe H, Turgeon JL et al (2018) Regulatory architecture of the LbetaT2 Gonadotrope cell
underlying the response to gonadotropin-releasing hormone. Front Endocrinol 9:34
Sadeghi-Nejad A, Senior B (1974) Autosomal dominant transmission of isolated growth hormone
deficiency in iris-dental dysplasia (Rieger’s syndrome). J Pediatr 85:644–648
Sajedi E, Gaston-Massuet C, Signore M, Andoniadou CL, Kelberman D, Castro S, Etchevers HC,
Gerrelli D, Dattani MT, Martinez-Barbera JP (2008) Analysis of mouse models carrying the
I26T and R160C substitutions in the transcriptional repressor HESX1 as models for septo-optic
dysplasia and hypopituitarism. Dis Model Mech 1:241–254
Salisbury TB, Binder AK, Grammer JC, Nilson JH (2009) GnRH-regulated expression of Jun and
JUN target genes in gonadotropes requires a functional interaction between TCF/LEF family
members and beta-catenin. Mol Endocrinol 23:402–411
Sarro R, Kocher AA, Emera D, Uebbing S, Dutrow EV, Weatherbee SD, Nottoli T, Noonan JP
(2018) Disrupting the three-dimensional regulatory topology of the Pitx1 locus results in overtly
normal development. Development 145:dev158550
Sato T, Vries RG, Snippert HJ, van de Wetering M, Barker N, Stange DE, van Es JH, Abo A,
Kujala P, Peters PJ et al (2009) Single Lgr5 stem cells build crypt-villus structures in vitro
without a mesenchymal niche. Nature 459:262–265
Schaeffer M, Hodson DJ, Lafont C, Mollard P (2010) Functional importance of blood flow
dynamics and partial oxygen pressure in the anterior pituitary. Eur J Neurosci 32:2087–2095
Schaible R, Gowen JW (1961) A new dwarf mouse. Genetics 46:896
174 A. Z. Daly and S. A. Camper

Seale P, Sabourin LA, Girgis-Gabardo A, Mansouri A, Gruss P, Rudnicki MA (2000) Pax7 is

required for the specification of myogenic satellite cells. Cell 102:777–786
Sekido R, Lovell-Badge R (2008) Sex determination involves synergistic action of SRY and SF1 on
a specific Sox9 enhancer. Nature 453:930–934
Semina EV, Reiter R, Leysens NJ, Alward WL, Small KW, Datson NA, Siegel-Bartelt J, Bierke-
Nelson D, Bitoun P, Zabel BU et al (1996) Cloning and characterization of a novel bicoid-
related homeobox transcription factor gene, RIEG, involved in Rieger syndrome. Nat Genet
14:392–399
Shapiro E, Biezuner T, Linnarsson S (2013) Single-cell sequencing-based technologies will
revolutionize whole-organism science. Nat Rev Genet 14:618–630
Sheng HZ, Zhadanov AB, Mosinger B Jr, Fujii T, Bertuzzi S, Grinberg A, Lee EJ, Huang SP,
Mahon KA, Westphal H (1996) Specification of pituitary cell lineages by the LIM homeobox
gene Lhx3. Science 272:1004–1007
Sheng HZ, Moriyama K, Yamashita T, Li H, Potter SS, Mahon KA, Westphal H (1997) Multistep
control of pituitary organogenesis. Science 278:1809–1812
Shima Y, Zubair M, Ishihara S, Shinohara Y, Oka S, Kimura S, Okamoto S, Minokoshi Y, Suita S,
Morohashi K (2005) Ventromedial hypothalamic nucleus-specific enhancer of Ad4BP/SF-1
gene. Mol Endocrinol 19:2812–2823
Shima Y, Zubair M, Komatsu T, Oka S, Yokoyama C, Tachibana T, Hjalt TA, Drouin J, Morohashi
K (2008) Pituitary homeobox 2 regulates adrenal4 binding protein/steroidogenic factor-1 gene
transcription in the pituitary gonadotrope through interaction with the intronic enhancer. Mol
Endocrinol 22:1633–1646
Simon FR, Fortune J, Iwahashi M, Sutherland E (2002) Sexual dimorphic expression of ADH in rat
liver: importance of the hypothalamic-pituitary-liver axis. Am J Physiol Gastrointest Liver
Physiol 283:G646–G655
Singh G, Kaur S, Stock JL, Jenkins NA, Gilbert DJ, Copeland NG, Potter SS (1991) Identification
of 10 murine homeobox genes. Proc Natl Acad Sci USA 88:10706–10710
Skowronska-Krawczyk D, Ma Q, Schwartz M, Scully K, Li W, Liu Z, Taylor H, Tollkuhn J, Ohgi
KA, Notani D et al (2014) Required enhancer-matrin-3 network interactions for a homeodomain
transcription program. Nature 514:257–261
Smith PE, MacDowell EC (1930) An hereditary anterior-pituitary deficiency in the mouse. Anat
Rec 46:249–257
Snell GD (1929) Dwarf, a new Mendelian recessive character of the house mouse. Proc Natl Acad
Sci USA 15:733–734
Sobrier ML, Brachet C, Vie-Luton MP, Perez C, Copin B, Legendre M, Heinrichs C, Amselem S
(2012) Symptomatic heterozygotes and prenatal diagnoses in a nonconsanguineous family with
syndromic combined pituitary hormone deficiency resulting from two novel LHX3 mutations. J
Clin Endocrinol Metab 97:E503–E509
Sobrier ML, Tsai YC, Perez C, Leheup B, Bouceba T, Duquesnoy P, Copin B, Sizova D, Penzo A,
Stanger BZ et al (2016) Functional characterization of a human POU1F1 mutation associated
with isolated growth hormone deficiency: a novel etiology for IGHD. Hum Mol Genet
25:472–483
Soji T, Mabuchi Y, Kurono C, Herbert DC (1997) Folliculo-stellate cells and intercellular commu-
nication within the rat anterior pituitary gland. Microsc Res Tech 39:138–149
Sornson MW, Wu W, Dasen JS, Flynn SE, Norman DJ, O’Connell SM, Gukovsky I, Carriere C,
Ryan AK, Miller AP et al (1996) Pituitary lineage determination by the prophet of Pit-1
homeodomain factor defective in Ames dwarfism. Nature 384:327–333
Spielmann M, Brancati F, Krawitz PM, Robinson PN, Ibrahim DM, Franke M, Hecht J, Lohan S,
Dathe K, Nardone AM et al (2012) Homeotic arm-to-leg transformation associated with
genomic rearrangements at the PITX1 locus. Am J Hum Genet 91:629–635
Stallings NR, Hanley NA, Majdic G, Zhao L, Bakke M, Parker KL (2002) Development of a
transgenic green fluorescent protein lineage marker for steroidogenic factor 1. Endocr Res
28:497–504
6 Pituitary Development and Organogenesis: Transcription Factors in. . . 175

Suga H, Kadoshima T, Minaguchi M, Ohgushi M, Soen M, Nakano T, Takata N, Wataya T,

Muguruma K, Miyoshi H et al (2011) Self-formation of functional adenohypophysis in three-
dimensional culture. Nature 480:57–62
Suh H, Gage PJ, Drouin J, Camper SA (2002) Pitx2 is required at multiple stages of pituitary
organogenesis: pituitary primordium formation and cell specification. Development
129:329–337
Swartz JM, Ciarlo R, Guo MH, Abrha A, Diamond DA, Chan YM, Hirschhorn JN (2017a) Two
unrelated undervirilized 46,XY males with inherited NR5A1 variants identified by whole-
exome sequencing. Horm Res Paediatr 87:264–270
Swartz JM, Ciarlo R, Guo MH, Abrha A, Weaver B, Diamond DA, Chan YM, Hirschhorn JN
(2017b) A 46,XX ovotesticular disorder of sex development likely caused by a steroidogenic
factor-1 (NR5A1) variant. Horm Res Paediatr 87:191–195
Szeto DP, Ryan AK, O’Connell SM, Rosenfeld MG (1996) P-OTX: a PIT-1-interacting
homeodomain factor expressed during anterior pituitary gland development. Proc Natl Acad
Sci USA 93:7706–7710
Szeto DP, Rodriguez-Esteban C, Ryan AK, O’Connell SM, Liu F, Kioussi C, Gleiberman AS,
Izpisua-Belmonte JC, Rosenfeld MG (1999) Role of the bicoid-related homeodomain factor
Pitx1 in specifying hindlimb morphogenesis and pituitary development. Genes Dev 13:484–494
Taira M, Jamrich M, Good PJ, Dawid IB (1992) The LIM domain-containing homeo box gene
Xlim-1 is expressed specifically in the organizer region of Xenopus gastrula embryos. Genes
Dev 6:356–366
Tajima T, Hattori T, Nakajima T, Okuhara K, Tsubaki J, Fujieda K (2007) A novel missense
mutation (P366T) of the LHX4 gene causes severe combined pituitary hormone deficiency with
pituitary hypoplasia, ectopic posterior lobe and a poorly developed sella turcica. Endocr J
54:637–641
Tajima T, Yorifuji T, Ishizu K, Fujieda K (2010) A novel mutation (V101A) of the LHX4 gene in a
Japanese patient with combined pituitary hormone deficiency. Exp Clin Endocrinol Diabetes
118:405–409
Takahashi K, Yamanaka S (2006) Induction of pluripotent stem cells from mouse embryonic and
adult fibroblast cultures by defined factors. Cell 126:663–676
Takahashi K, Tanabe K, Ohnuki M, Narita M, Ichisaka T, Tomoda K, Yamanaka S (2007)
Induction of pluripotent stem cells from adult human fibroblasts by defined factors. Cell
131:861–872
Tatsumi K, Miyai K, Notomi T, Kaibe K, Amino N, Mizuno Y, Kohno H (1992) Cretinism with
combined hormone deficiency caused by a mutation in the PIT1 gene. Nat Genet 1:56–58
Theill LE, Castrillo JL, Wu D, Karin M (1989) Dissection of functional domains of the pituitary-
specific transcription factor GHF-1. Nature 342:945–948
Therrien M, Drouin J (1991) Pituitary pro-opiomelanocortin gene expression requires synergistic
interactions of several regulatory elements. Mol Cell Biol 11:3492–3503
Therrien M, Drouin J (1993) Cell-specific helix-loop-helix factor required for pituitary expression
of the pro-opiomelanocortin gene. Mol Cell Biol 13:2342–2353
Thomas PQ, Rathjen PD (1992) HES-1, a novel homeobox gene expressed by murine embryonic
stem cells, identifies a new class of homeobox genes. Nucleic Acids Res 20:5840
Thomas PQ, Johnson BV, Rathjen J, Rathjen PD (1995) Sequence, genomic organization, and
expression of the novel homeobox gene Hesx1. J Biol Chem 270:3869–3875
Thomas PQ, Dattani MT, Brickman JM, McNay D, Warne G, Zacharin M, Cameron F, Hurst J,
Woods K, Dunger D et al (2001) Heterozygous HESX1 mutations associated with isolated
congenital pituitary hypoplasia and septo-optic dysplasia. Hum Mol Genet 10:39–45
Topilko P, Schneider-Maunoury S, Levi G, Trembleau A, Gourdji D, Driancourt MA, Rao CV,
Charnay P (1998) Multiple pituitary and ovarian defects in Krox-24 (NGFI-A, Egr-1)-targeted
mice. Mol Endocrinol 12:107–122
176 A. Z. Daly and S. A. Camper

Trapnell C, Cacchiarelli D, Grimsby J, Pokharel P, Li S, Morse M, Lennon NJ, Livak KJ,

Mikkelsen TS, Rinn JL (2014) The dynamics and regulators of cell fate decisions are revealed
by pseudotemporal ordering of single cells. Nat Biotechnol 32:381–386
Treier M, Gleiberman AS, O’Connell SM, Szeto DP, McMahon JA, McMahon AP, Rosenfeld MG
(1998) Multistep signaling requirements for pituitary organogenesis in vivo. Genes Dev
12:1691–1704
Treier M, O’Connell S, Gleiberman A, Price J, Szeto DP, Burgess R, Chuang PT, McMahon AP,
Rosenfeld MG (2001) Hedgehog signaling is required for pituitary gland development. Devel-
opment 128:377–386
Tremblay JJ, Drouin J (1999) Egr-1 is a downstream effector of GnRH and synergizes by direct
interaction with Ptx1 and SF-1 to enhance luteinizing hormone beta gene transcription. Mol Cell
Biol 19:2567–2576
Tremblay JJ, Lanctot C, Drouin J (1998) The pan-pituitary activator of transcription, Ptx1 (pituitary
homeobox 1), acts in synergy with SF-1 and Pit1 and is an upstream regulator of the
Lim-homeodomain gene Lim3/Lhx3. Mol Endocrinol 12:428–441
Tremblay JJ, Marcil A, Gauthier Y, Drouin J (1999) Ptx1 regulates SF-1 activity by an interaction
that mimics the role of the ligand-binding domain. EMBO J 18:3431–3441
Tsai YC, Cooke NE, Liebhaber SA (2016) Long-range looping of a locus control region drives
tissue-specific chromatin packing within a multigene cluster. Nucleic Acids Res 44:4651–4664
Vallette-Kasic S, Brue T, Pulichino AM, Gueydan M, Barlier A, David M, Nicolino M,
Malpuech G, Dechelotte P, Deal C et al (2005) Congenital isolated adrenocorticotropin defi-
ciency: an underestimated cause of neonatal death, explained by TPIT gene mutations. J Clin
Endocrinol Metab 90:1323–1331
von Möllendorff W, Bargmann W (1940) Handbuch der mikroskopischen Anatomie des
Menschen: Hypofyse. In: Blutgefäss- und Lymphgefässapparat; Innersekretorische Drüsen.
Springer, Berlin
Wang J, Scully K, Zhu X, Cai L, Zhang J, Prefontaine GG, Krones A, Ohgi KA, Zhu P, Garcia-
Bassets I et al (2007) Opposing LSD1 complexes function in developmental gene activation and
repression programmes. Nature 446:882–887
Ward RD, Raetzman LT, Suh H, Stone BM, Nasonkin IO, Camper SA (2005) Role of PROP1 in
pituitary gland growth. Mol Endocrinol 19:698–710
Ward RD, Davis SW, Cho M, Esposito C, Lyons RH, Cheng JF, Rubin EM, Rhodes SJ, Raetzman
LT, Smith TP et al (2007) Comparative genomics reveals functional transcriptional control
sequences in the Prop1 gene. Mamm Genome 18:521–537
Wen S, Ai W, Alim Z, Boehm U (2010) Embryonic gonadotropin-releasing hormone signaling is
necessary for maturation of the male reproductive axis. Proc Natl Acad Sci USA
107:16372–16377
Wu W, Cogan JD, Pfaffle RW, Dasen JS, Frisch H, O’Connell SM, Flynn SE, Brown MR, Mullis
PE, Parks JS et al (1998) Mutations in PROP1 cause familial combined pituitary hormone
deficiency. Nat Genet 18:147–149
Xia Y, Nivet E, Sancho-Martinez I, Gallegos T, Suzuki K, Okamura D, Wu MZ, Dubova I, Esteban
CR, Montserrat N et al (2013) Directed differentiation of human pluripotent cells to ureteric bud
kidney progenitor-like cells. Nat Cell Biol 15:1507–1515
Xu Y, Baldassare M, Fisher P, Rathbun G, Oltz EM, Yancopoulos GD, Jessell TM, Alt FW (1993)
LH-2: a LIM/homeodomain gene expressed in developing lymphocytes and neural cells. Proc
Natl Acad Sci USA 90:227–231
Xue L, Cai JY, Ma J, Huang Z, Guo MX, Fu LZ, Shi YB, Li WX (2013) Global expression profiling
reveals genetic programs underlying the developmental divergence between mouse and human
embryogenesis. BMC Genomics 14:568
Young MD, Willson TA, Wakefield MJ, Trounson E, Hilton DJ, Blewitt ME, Oshlack A, Majewski
IJ (2011) ChIP-seq analysis reveals distinct H3K27me3 profiles that correlate with transcrip-
tional activity. Nucleic Acids Res 39:7415–7427
6 Pituitary Development and Organogenesis: Transcription Factors in. . . 177

Zappone MV, Galli R, Catena R, Meani N, De Biasi S, Mattei E, Tiveron C, Vescovi AL, Lovell-
Badge R, Ottolenghi S et al (2000) Sox2 regulatory sequences direct expression of a (beta)-geo
transgene to telencephalic neural stem cells and precursors of the mouse embryo, revealing
regionalization of gene expression in CNS stem cells. Development 127:2367–2382
Zhao L, Bakke M, Krimkevich Y, Cushman LJ, Parlow AF, Camper SA, Parker KL (2001)
Steroidogenic factor 1 (SF1) is essential for pituitary gonadotrope function. Development
128:147–154
Zhao Y, Mailloux CM, Hermesz E, Palkovits M, Westphal H (2010) A role of the LIM-homeobox
gene Lhx2 in the regulation of pituitary development. Dev Biol 337:313–323
Zheng GX, Terry JM, Belgrader P, Ryvkin P, Bent ZW, Wilson R, Ziraldo SB, Wheeler TD,
McDermott GP, Zhu J et al (2017) Massively parallel digital transcriptional profiling of single
cells. Nat Commun 8:14049
Zhu X, Zhang J, Tollkuhn J, Ohsawa R, Bresnick EH, Guillemot F, Kageyama R, Rosenfeld MG
(2006) Sustained Notch signaling in progenitors is required for sequential emergence of distinct
cell lineages during organogenesis. Genes Dev 20:2739–2753
Zhu X, Gleiberman AS, Rosenfeld MG (2007) Molecular physiology of pituitary development:
signaling and transcriptional networks. Physiol Rev 87:933–963
Ziegenhain C, Vieth B, Parekh S, Reinius B, Guillaumet-Adkins A, Smets M, Leonhardt H,
Heyn H, Hellmann I, Enard W (2017) Comparative analysis of single-cell RNA sequencing
methods. Mol Cell 65(631–643):e634
 Part II
Developmental Modulators and Epigenetic
Factors
Hypothalamic Development: Role of GABA
7
M. Stratton and S. Tobet

Abstract
A number of cellular developmental processes occur in an orchestrated fashion to
mediate the transition from a simple neural tube to the complexity of the adult
brain. A combination of internal programming and external cues provides the
molecular specificity needed to direct cells to the correct location, initiate correct
gene expression, and connect to the appropriate circuits. The heterogeneous cells
of the hypothalamus affect all aspects of physiology through the regulation of the
autonomic nervous system, motivated behavior, and endocrine balance. Small
disruptions (genetic or environmental) to hypothalamic development can impact
adult physiology and contribute to pathology. GABAergic signaling has drawn
the attention of developmental neuroendocrinologists as a potential effector
system of hypothalamic development. This chapter discusses the role of GABA
in directing the development of the hypothalamus, specifically related to the
migration of neurons that release gonadotropin-releasing hormone (GnRH), and
in the formation of the ventromedial (VMH) and paraventricular (PVN) nuclei of
the hypothalamus (Fig. 7.1).

Keywords
Neuron migration · Hypothalamic development · Gamma-aminobutyric acid
(GABA) · Gonadotropin-releasing hormone (GnRH) · Paraventricular nucleus of
the hypothalamus (PVN) · Ventromedial nucleus of the hypothalamus (VMH) ·
Developmental patterning

M. Stratton (*)
Department of Physiology and Cell Biology, Ohio State University, Columbus, OH, USA
e-mail: [email protected]
S. Tobet (*)
Department of Biomedical Sciences, School of Biomedical Engineering, Colorado State University,
Fort Collins, CO, USA
e-mail: [email protected]

# Springer Nature Switzerland AG 2020 181

S. Wray, S. Blackshaw (eds.), Developmental Neuroendocrinology, Masterclass in
Neuroendocrinology 9, https://doi.org/10.1007/978-3-030-40002-6_7
182 M. Stratton and S. Tobet

Abbreviations

AOB Accessory olfactory bulb

GABA Gamma aminobutyric acid
GAD Glutamic acid decarboxylase
GnRH Gonadotropin releasing hormone
OB Olfactory bulb
OE Olfactory epithelium
PVN Paraventricular nucleus of the hypothalamus
VMH Ventromedial nucleus of the hypothalamus
VNO Vomeronasal organ

7.1 Hypothalamic Organization

The hypothalamus is a master regulator of physiological homeostasis. Cells in the

hypothalamus survey information from the periphery through either systemic circu-
lation (blood/cerebrospinal fluid/lymph) or peripheral sensory inputs and integrate
information from other brain regions to elicit coordinated responses through neuro-
endocrine and autonomic outlets. The complexity of this master regulator is astound-
ing given the sheer diversity of inputs and outputs. The oldest parcellations of cell
grouping in the brain were based on a simple dye-based method known as Nissl
staining, which visualizes neuronal cell bodies according to the accumulation of
rough endoplasmic reticulum. For more than 100 years, people have subdivided the
hypothalamus into different numbers of subdivisions. In general, there are about a
dozen nuclear groups and several named areas delineated based on Nissl stains and
functional circuits in adults. From another perspective (Puelles and Rubenstein
2015), you might consider 33 molecularly discrete hypothalamic progenitor areas
that produce as many as 150 derived “nuclei” (distinct cell populations).
Early chapters in this series deal with the molecular specification of hypothalamic
cells (e.g., transcription factors) and some deal with expressed genes that directly
signal developmental processes (e.g., sonic hedgehog). The chapter presented here
discusses a secreted factor that is the product of an enzymatic reaction (e.g., the
decarboxylation of glutamate) that is a well-known neurotransmitter in the central
nervous system; gamma-aminobutyric acid (GABA) (Fig. 7.1).
Beginning with studies in the telencephalon and extending to studies of the
diencephalon, evidence suggests that GABA release in early development is capable
of impacting the cytoarchitectural landscape of the brain and for this chapter, in
particular, the hypothalamus. The effects of GABA on cell positioning and cell
phenotype within the hypothalamus during early development (and interactions with
perinatal stress and sex) are antecedent to hypothalamic regulation of adult homeo-
stasis. This chapter focuses on the formation of the ventromedial (VMH) and
paraventricular (PVN) hypothalamic nuclei as well as neurons that secrete
7 Hypothalamic Development: Role of GABA 183

Fig. 7.1 Introduction to the role of GABA in hypothalamic development and potential
implications for adult physiology and pathophysiology. There is a striking absence of GABA and
GABA-producing enzymes inside select developing regions of the hypothalamus (e.g., VMH–
GAD67 depicted and PVN–GABA depicted) which is contrasted by enrichment of GABA
receptors within these regions (GABABR1 depicted). Subsequent text and figures summarize
work done to determine the role of GABA in directing the development of these regions and for
the GnRH system as well as potential consequences of altered GABA signaling during hypotha-
lamic development. The digital image in the background was taken from archival material in the
Tobet laboratory and shows immunoreactive GAD67 in adult mouse brain; dashed circles highlight
the VMH and PVN nuclei. Inset VMH images were adapted from Davis et al. (2002) and inset PVN
images were adapted from McClellan et al. (2010)

gonadotropin-releasing hormone (GnRH), which are scattered through the basal

forebrain including in the hypothalamus (Fig. 7.2). Functions associated with these
hypothalamic regions/cells include reproduction, feeding, drinking, stress responses,
cardiac output, blood pressure, growth signaling, and arousal.
184 M. Stratton and S. Tobet

Fig. 7.2 Sagittal schematic diagram of an embryonic head illustrates hotspots for GABA regula-
tion of neuron migration in the developing hypothalamus. GABA has been shown to act generally
as a “brake” for hypothalamic cells as they exit nasal cavity and enter the brain (GnRH neurons), in
the developing ventromedial nucleus of the hypothalamus (VMH) and in the developing
paraventricular nucleus of the hypothalamus (PVN). GnRH neurons mostly track fibers of the
vomeronasal nerve from the vomeronasal organ (VNO) across the cortical plate (CP) toward the
accessory olfactory bulb (AOB). When a subset of fibers bifurcates and turn toward the basal
hypothalamus, GnRH neurons start making directional choices that are likely influenced by GABA
from neighboring GnRH neurons and from other neurons in the region. The other major source of
fibers is from main olfactory neuroepithelial cells (OE) that travel to the main olfactory bulb (OB).
For the PVN and VMH, cells originate medially and enter the region of the nuclei before they are
impacted by GABA at the boundaries of the nuclei as they coalesce

7.1.1 Patterning Processes

From a mechanistic perspective, there are two major organizational schemes in brain
development that individual cells must navigate; joining layers (e.g., cerebral cortex)
or groups (e.g., diencephalic nuclei). GnRH neurons indicate a less common third
case that does not do either (in most species) as these cells are not distributed in
either layer or confined to nuclear group. Due to the inability to categorize cells
solely based on distance from a proliferative zone, developmental studies of the
7 Hypothalamic Development: Role of GABA 185

hypothalamus may be intrinsically more complex than in areas like the cortex. As in
any tissue, the processes of proliferation, growth, migration, differentiation, and
death mediate the development of the hypothalamus. Process extension (axon and
dendrite) and circuit integration are additional crucial steps in brain development.
During development, each of these steps impacts the other and altered regulation of
one process can lead to profound effects in downstream processes. For instance, the
microenvironment and epigenetic program of the progenitor pool directs gene
expression which influences migration. The microenvironment through which a
cell migrates, further influences its epigenetic and gene expression program, which
influences its growth, differentiation, process extension, and circuit integration.
Regardless of whether you categorize functional entities of brain development
into larger units like prosomeres or split things into smaller units based on molecular
genetic characteristics (early chapters, Puelles and Rubenstein 2015), the business
end of development is operated by the molecules that provide signals from one cell
to another (Fig. 7.3). They can be recognition, adhesion or receptive entities on a cell
surface, or secreted, diffusive molecules that signal attraction or repulsion. The
current chapter addresses the secreted molecule GABA as a contributor to cell
movement and ultimately anatomical position. For secreted molecules to provide
patterning information they must have a way to address the issue of spatial distribu-
tion relative to nuclear boundaries that characterize the hypothalamus. Whereas a
radially migrating cell in the cerebral cortex can get 2-dimensional data to signal stop
or go relative to a layer, radially migrating cells in the hypothalamus must sense
curvilinear boundaries.
In regions outside of the hypothalamus, GABA had been suggested to elicit
multiple effects in neurodevelopment. These include effects on proliferation, migra-
tion, DNA synthesis, differentiation, growth, and morphology (reviewed in Wu and
Sun 2015). These developmental effects of GABA are thought to be mediated by
non-synaptic release/secretion of GABA and typically occur prior to the emergence
of structural synapses. Likely mechanisms for GABA release through non-synaptic
mechanisms include GABA transporters or through non-vesicular channel-mediated
release.

7.1.2 The Birth and Organization of Hypothalamic Neurons

For hypothalamic cells, the first cells to leave the proliferative regions are usually the
ones that migrate the farthest. This is in contrast to structures such as the cerebral
cortex, where cells that populate the farthest layers, are the last ones to leave the
proliferative zone. Derived from the ventral and inferior lobes of the diencephalic
neuroepithelium, Altman and Bayer (1986) describe three waves of neurogenesis in
development of the rat hypothalamus. The first wave is between embryonic day (E)
13 and E15 (E0 ¼ first day post timed breeding; for mouse E11–E13) and these
neurons form lateral hypothalamic structures. The second wave is between E15 and
E17 with these neurons becoming part of the medial hypothalamus. The final wave
occurs between E17 and E19. These neurons make up the periventricular
186 M. Stratton and S. Tobet

Fig. 7.3 Schematic diagram illustrates a cellular perspective of the developmental boundaries of
the PVN and VMH that are defined by the presence/absence of GABA synthesizing cells. The cell
on the left in (a) suggests a cell found outside of the boundary that synthesizes and secretes GABA
following glutamic acid decarboxylase (GAD) acting on glutamate precursor. The cell on the right
in (a) suggests a cell found inside of the boundary that receives GABA through GABAA (forming a
chloride channel) or GABAB (G-protein coupled) receptors. Transcription factor programs initiated
by classic hypothalamic differentiation genes dictate cells that secrete and/or respond to GABA
signaling. Digital images in (b) depict the developmental boundaries of the VMH. Images demon-
strate GAD67 (left) and GABABR1 (right) immunoreactivities in the developing VMH, and were
adapted from Davis et al. (2002)

hypothalamic populations with notable exception of the PVN, for which most
neurons complete terminal mitosis prior to E15.
The field of birthdate labeling with tritiated thymidine to follow cells was
pioneered by investigators such as Sidman, Rakic, Altman, and Bayer. After 1988
(Miller and Nowakowski 1988), the field was gradually commandeered by studies
using bromodeoxyuridine (BrdU) because results could be generated more quickly
7 Hypothalamic Development: Role of GABA 187

by immunohistochemistry instead of waiting for the development of radioauto-

graphic images. Interestingly, the technique of choice is shifting again, this time to
5-ethynyl-20 -deoxyuridine (EdU) with the publication of a simpler method that
makes it even easier to process tissue (Salic and Mitchison 2008). Tritiated thymi-
dine remains the gold standard by virtue of the least modification to the nucleic acid
being introduced to the DNA for tracing purposes.
Neurons that synthesize an essential peptide for the regulation of reproduction,
gonadotropin-releasing hormone (GnRH), traverse significant distances in almost all
vertebrates that have been so characterized. The realization that the GnRH neurons
originate outside of the CNS in the olfactory periphery was a major neuroendocrine
“aha” moment for 1989 (Schwanzel-Fukuda and Pfaff 1989; Wray et al. 1989). Both
publications indicated that GnRH neurons were born in a narrow temporal window
in mice between embryonic days 10 and 13 based on tritiated thymidine autoradi-
ography. This time frame fits well, albeit a little early, with other hypothalamic
neurons that do not migrate nearly as far. GnRH neurons originating in the nasal
compartment migrate over the nasal septum, cross the cribriform plate, and enter the
basal forebrain where they must turn toward the hypothalamus. While their fibers are
directed to two major sites (organum vasculosum of the lamina terminalis and
median eminence), the cell bodies scatter across a much larger expanse. The signals
that direct this unusual migration include soluble factors that can work by diffusing
across a distance as well as factors that regulate GnRH neuron adhesion to guiding
fibers (reviewed Wierman et al. 2011), and it is likely that none of the selective
signals drive more than 30–40% of the GnRH neuronal population. This redundancy
of signaling is probably a major factor in ensuring reproductive success leading to
the preservation of species.
Medial-to-lateral migration in the hypothalamus, like in the cerebral cortex, is
aided by the presence of radial glia, which arises at the stage of the neural tube and
stretches from the ventricle to the pia surface. This creates a series of railroad tracks
like fibers that immature neurons can crawl along during the course of migration
(Rakic 1972). While the rules and roles of radial glia are well studied in the cortex,
fewer studies exist in the diencephalon. Nonetheless, there is ample evidence of
radial glial fibers throughout the developing diencephalon in a number of species
(Tobet et al. 1995).
Combining birthdate labeling with radial glia labeling in the same experiment
allowed for the observation that radial glia may provide paths for cells to cross
putative boundaries into diencephalic regions (Tobet et al. 1995) as well as paths
across forming cell groups such as the VMH (Dellovade et al. 2001). GABA
regulation of cell adhesion to guidance fibers (radial glial or neuronal) may be one
mechanism of migratory control. Importantly for this chapter, there are patterns of
GABAergic patterning elements present throughout hypothalamic cell migratory
periods that may influence cell movements (Bless et al. 2000; Davis et al. 2004).
Several hypothalamic cell groups are defined by neurochemical phenotypes of
cells in regions (e.g., GABA, see Fig. 7.1) before they become recognizable using
Nissl stains. This is readily apparent in mice for the VMH (Tobet et al. 1999) and the
PVN (McClellan et al. 2010). Neither VMH or PVN is discernible as a cell group
188 M. Stratton and S. Tobet

using Nissl stains in mice until E17. However, many of the neurons that make up
these nuclear groups are present in groups earlier and express immunoreactive
peptides before E17. For example, immunoreactive corticotropin-releasing hormone
(CRH) and arginine vasopressin (AVP) can be found in the PVN by E15. It is not yet
known what event occurs between E15 and E17 to allow for Nissl stain detection,
although possible mechanisms include the proliferation of glia and/or tighter cell
packing mediated by migration. Since there is a GABAergic surround to both VMH
and PVN (see below), it is possible that GABA contributes to the fine-tuning of cell
positions that results in nuclear appearance.

7.2 Reconstructing GABA and GABA Receptor Distributions

Reveals Unique Patterns

The first requirement for connecting a signaling factor with hypothalamic develop-
ment is the demonstration of presence. Within the developing hypothalamus, multi-
ple immature cell groupings, including those destined to form PVN and VMH, are
enriched with GABA receptors while being bordered by areas with many cells
containing immunoreactive GABA or expressing GABA synthesizing enzymes
(Figs. 7.1 and 7.2). This striking pattern, first observed in the 1990s, provided the
first suggestions of an important role for GABA in directing hypothalamic
development.

7.2.1 GABAergic Elements

GABA Sources: The patterning of GABA and its synthesizing enzymes can be
visualized efficiently through immunohistochemistry. Notably, GABA is found in
many cells throughout the hypothalamus where its synthetic enzymes glutamic acid
decarboxylases 1 and 2 (GAD aka GAD67 and GAD65) are found in cell bodies that
are excluded from locations within the developing VMH and PVN during develop-
ment (Tobet et al. 1999; McClellan et al. 2008; e.g., Fig. 7.1) and in adulthood (by in
situ hybridization for GAD; Okamura et al. 1990). Thus, the embryonic PVN and
VMH are ringed by regions of cells that contain immunoreactive GABA and GAD
while there is an absence of GABA and GAD within cells of the developing nuclear
groups. This pattern of expression also holds for GABA transporters GAT1 and
GAT4 (Jursky and Nelson 1996).
This void of GABAergic components in developing nuclei is apparent as early as
E12–13 in mice when the various developing nuclei are first detectable by
immunohistochemistry (but not yet Nissl). As GABAergic axons grow into these
hypothalamic nuclei, the pattern is lost (earlier for PVN than VMH) and the entire
hypothalamus becomes highly immunoreactive for GABAergic protein elements
soon after birth (GAD mRNA is still restricted to cells outside of the VMH and PVN;
Okamura et al. 1990). It is unclear if all of the sources of GABA are neuronal or if
some might be glial in nature. Thoughts of GAD as strictly neuronal might be only
7 Hypothalamic Development: Role of GABA 189

partially true as there is evidence of glial expression of GAD enzymes. Glia can also
synthesize GABA through the putrescine degradation pathway and is dependent on
the activity of monoamine oxidase B (Yoon and Lee 2014). More recent work also
points toward endothelial cells associated with developing blood vessels as being
sources of GABA in other brain regions (Li et al. 2018).
The case is less clear for patterning of GABAergic elements along the develop-
mental path of GnRH neurons. Consistent with this notion, a subset of GnRH
neurons (~30%) contain GABA during development (Tobet et al. 1996a) while
exposure to GABA varies across the long path of migration from the nasal compart-
ment to destinations in the hypothalamus and other brain regions. It is likely that
GABA from within and outside the population is one of the soluble secreted factors
that contribute to their migration.

7.2.2 The A’s, B’s, and C’s of GABA Receptors

GABA binds to three classes of cell surface receptors, GABAA, GABAA-rho

(previously GABAC), and GABAB receptors. Given the paucity of literature
investigating GABAA-rho in the development of the hypothalamus, the GABAA-
rho receptor, a ligand-gated Cl channel composed of 5 rho subunits predominantly
studied in the retina, will not be discussed further in this chapter. Notably, GABAA
and GABAB receptor subunits are enriched in the developing PVN (McClellan et al.
2010) and VMH (Dellovade et al. 2001; Davis et al. 2002) at least as early as E15.
GABAA receptors have been found in GnRH neurons from the earliest ages exam-
ined (E12.5; Temple and Wray 2005) and virtually all GnRH neurons contain
GABAA receptors, but with heterogeneous complements of subunit composition
(Simonian et al. 2000).
GABAA receptors are a family of ligand-gated Cl channels. Five subunits (e.g.,
2 α subunits, 2 β subunits, and 1 γ subunit) comprise a functional GABAA channel.
There are at least 21 subunits that can be used to build the channel (Lujan et al.
2005). Depending on the intracellular versus extracellular concentration gradient of
Cl, the activation of GABAA receptors can exert a depolarizing or hyperpolarizing
influence on membrane potential. In the embryo, Cl concentrations are higher
inside the cell than outside, and thus GABAA receptor activation is excitatory in
the embryo. The Cl gradient usually reverses at some point in development with the
expression of the K+/Cl cotransporter KCC2, which shunts Cl ions to the outside
of the cell. There have been several reports of this embryonic Cl gradient being
present in specialized locations within the adult brain, for example, the progenitor
pool of the adult hippocampus and some magnocellular neurons in the PVN, which
coincide with a lack of KCC2 expression.
GABAB receptors, among the first G protein-coupled receptors to be identified,
were cloned after searching for molecules that could explain responses to GABA
that were insensitive to chloride and bicuculline (GABAA antagonist). GABAB
receptors have been extensively studied and excellent reviews are available (e.g.,
Frangaj and Fan 2018). The functional G protein-coupled receptor is a heterodimer
190 M. Stratton and S. Tobet

of R1 and R2 subunits, for which multiple splice variants exist. GABAB receptor
activation has been shown to affect the adenylate cyclase system, decrease Ca2+
conductance, and increase K+ conductance. Originally thought to be mediated by
pertussis toxin-sensitive Giα and Goα G proteins, some pertussis toxin insensitive
baclofen (agonist) effects have been found in magnocellular neurons of PVN and
supraoptic nucleus (SON), in presynaptic receptors and whole-brain
co-immunoprecipitation mass spectrometry studies point toward Gβ1/β2, Gγ2, and
GαΖ as being predominant effector molecules (Schwenk et al. 2016).

7.3 Organizing Principles

The role of homeotic genes in the formation of the hypothalamus is complicated and
all the more so by a lack of understanding downstream molecular consequences.
When molecular genetics have been used to disrupt the expression of any of a
number of transcription factors, hypothalamic patterning has been dramatically
altered (see preceding chapters in this series). Whether the mechanisms are through
proliferation, migration, and terminal differentiation is often an open question.
Determining the impact on the molecules that mediate cell–cell communication
and induction of cellular processes is still less understood and infrequently
attempted. GABAergic signaling falls in this latter category.
Relative to understanding a role for GABA, a major homeodomain protein of
interest is the vertebrate homolog of the fly gene distal-less (DLX 1 and 2). DLX
protein exerts a direct regulatory impact on the transcription of glutamic acid
decarboxylases (GAD)65 and 67 (Le et al. 2017) and establishes the distribution
of GABA sources in multiple brain regions including the developing hypothalamus.
When a number of different transcription factors have been disrupted, a common
phenotype is aberrant Dlx expression and, presumably GABA. One homeobox-
containing gene, Orthopedia (Otp) is expressed in neurons that come to reside in
the PVN, SON, anterior periventricular, and arcuate nuclei. In mice in which Otp
was replaced by a lacZ reporter gene, cells migrated abnormally and were apparently
displaced by cells expressing Dlx1, in a manner suggesting that Dlx1-positive cells
replaced PVN-committed cells (positive for lacZ) that migrated abnormally
(Acampora et al. 1999). Interestingly, these types of knockout mice usually, when
they can be raised to adulthood, display altered physiology and behavior associated
with the hypothalamus including, response to dehydration, feeding, sexual function,
and stress responses. A key question is what downstream molecular changes follow
the transcription factor action(s) to alter cell positions? Are they secreted protein-
coding genes like sonic hedgehog or wnt’s or might they be GAD 65 or 67 leading to
changes in GABA that drive the movement of other cells?
The transcription factor HES1, a Notch effecter gene, is also expressed in the
developing pituitary and hypothalamus. In mice lacking HES1, distribution of GAD
67 immunoreactivity was shifted from surrounding the PVN to within the PVN. In
this case, AVP neurons were found to be located outside of the PVN boundaries in
the HES1 null (Aujla et al. 2011). The SON also showed ectopic GAD expression
7 Hypothalamic Development: Role of GABA 191

Fig. 7.4 Schematic diagram illustrates genetic manipulations that alter GABAergic cell
populations relative to boundaries in the developing ventromedial nucleus of the hypothalamus
(VMH) by different mechanisms. Loss of SF-1 function most likely causes cellular rearrangements
that result in GABAergic cells moving into the location of the VMH (Dellovade et al. 2000; Davis
et al. 2004). Loss of Rax function (knockout—KO) most likely causes cells in the location of the
VMH to change fate (transfate) to express glutamic acid decarboxylases (GAD) and make GABA
(Lu et al. 2013). In both cases, the phenotypes of cells in the VMH are altered and importantly,
GABA synthesizing cells become inappropriately located within the nucleus. Digital images from
sections taken from wild-type (WT) and SF1-KO mice were adapted from Dellovade et al. (2000).
The image of a section from a Rax KO mouse was adapted from one provided by Dr. D. Kurrasch
related to Lu et al. (2013)

in the HES1 null, which corresponded with the dispersion of typical AVP expressing
neurons in the SON. Based on the hypothesis that an inappropriately located cell is
exposed to alternate signaling milieus that have further functional consequences,
axons from these misplaced AVP neurons were unable to appropriately target the
posterior pituitary.
The gene Rax codes for a homeodomain protein that was found to be critical to
the formation of the VMH based on changes in the fates or molecular identities of
multiple cells (Lu et al. 2013). Mice lacking functional Rax also exhibited a change
in the pattern of GAD immunoreactivity. Cells containing GAD become localized
inside the VMH in contrast to its normal pattern of expression in cells surrounding
the nucleus. The authors pose two possible models to explain the Rax phenotype.
First, that mutant VMH cells “transfate” into a hypothalamic non-VMH identity, or
second that hypothalamic non-VMH cells (e.g., GAD containing) could migrate into
the mutant VMH region (Fig. 7.4). However, the former change in fate hypothesis
was considered more likely, based on the lack of change in the distribution of cells
expressing another transcription factor, Nkx2.1. Nonetheless, the new area of GAD
immunoreactivity within the VMH in the Rax null excludes several VMH cell types
that typically reside there.
When the transcription factor steroidogenic factor 1 (SF-1) was disrupted, the
positions of cells in and around the VMH change, including those expressing
192 M. Stratton and S. Tobet

Nkx2.1. Similar to Rax, cells that express GAD and normally surround the VMH
become localized within regions closer to the third ventricle and the distribution of
other markers of cell phenotype, including estrogen receptor-α, neuropeptide Y, and
galanin all become localized more medially (Dellovade et al. 2000). When the
amount of immunoreactive GAD was quantified across hypothalamic tissue
sections, the total amount of GAD did not appear to change. This suggests that
GAD expressing cells inappropriately invade the developing VMH in the SF-1 null,
rather than GAD expression being turned on inappropriately in typically VMH
resident cells. Also, cells that would normally express SF-1 and reside in the
VMH (assessed by SF-1-GFP transgenic reporter) become located more laterally
in the SF-1 null (Davis et al. 2004). The question in the SF-1 findings, like those
noted above, is identifying the mechanism behind the change. Transcription factors
are great markers for cell phenotypes and frequently code for genes that direct cell
differentiation, but to impact migration they must act by influencing genes that affect
cell movements. In this case, if GABA is an active agent influencing movement, and
if GABA acts as a stop signal for normal movement of SF-1 positive cells (more
discussion below), then when SF-1 is disrupted, GABA must no longer provide that
stop signal (Davis et al. 2004).

7.4 Finding Useful Model Animals

As alluded to above, a major tool for determining the impact of gene expression on
hypothalamic nuclear formation has been derived from animal models with selective
gene disruptions. It is particularly useful when the expression of a key regulatory
gene is somewhat selective for a given cell group. This is true for Sf-1 being
relatively restricted to the VMH and to some extent true for Sim-1, which is
somewhat restricted to the PVN (but does extend laterally and ventrally more so
than is usually discussed). Many of the other transcription factors discussed above
(e.g., Dlx) are expressed in many regions, but with patterns that still allow localized
interpretations when there are genetic disruptions. Unfortunately for GnRH neurons,
the only reliable marker is GnRH itself. Most of the genetic disruptions of transcrip-
tion factors that have been characterized as impacting GnRH neuron development,
do so indirectly, acting on elements other than the neurons themselves. However, an
advantage in this system is that most GnRH neurons found in immunohistochemical
studies can now be linked to one site of origin. Thus, genetic disruptions that
influence GnRH neuron movement can be characterized by changes in the distribu-
tion of GnRH immunoreactive cells at different developmental stages. The same
cannot be said of most other neurons in the brain. For example, neurons containing
CRH can arise from several different progenitor populations.
Much of the work done to investigate GABA’s role in directing neuron migration
in the hypothalamus has used a mouse line generated with YFP expressed under the
control of a modified neuronal Thy-1 promoter. In efforts to find a good reporter to
visualize synaptogenesis in motor neurons and retinal ganglion cells, 25 fluorescent
reporter lines were generated. Sixteen of these lines included YFP. Though the same
genetic element was used to drive fluorescent reporter expression, the lines have
7 Hypothalamic Development: Role of GABA 193

remarkably divergent expression patterns. One of these lines showed subpopulation-

specific fluorescence in the hypothalamus and has been used as a tool to query
mechanisms of hypothalamic development (reviewed in Tobet et al. 2003). For
studies of GnRH migration, transgenic fluorescent mice (Suter et al. 2000) again
played an important role (Bless et al. 2005), although significant work was done
before GnRH promoter driving GFP mice were available.

7.5 Choosing Ex Vivo Approaches

A key question posed in Sect. 7.3 is whether the molecular changes following
transcription factor actions involve secreted protein-coding genes like sonic hedge-
hog or wnt’s or might they be acting through GAD 65/67 leading to changes in
GABA that drive the movement of other cells? This is not a question easily answered
in vivo. To get a better look at the movement of cells in real-time and in response to
specific factors that influence motion, in vitro methods are needed. There are two
usual methods to evaluate influences on motion. In one case, dispersed cell cultures
have been used. This is perhaps most appropriate when there is a relatively homog-
enous population to work with, such as immortalized cell lines that represent GnRH
neurons (e.g., NLT, GT-1). Transwell migration assays and wound-healing migra-
tion assays were both able to demonstrate roles for secretory products impacting
NLT cells in vitro (reviewed Wierman et al. 2011). The problem for GnRH neurons
in vitro and for many other regions in the hypothalamus is the cellular diversity and
the dependence of cell behaviors on the surrounding cellular environment. For
GABA, the transwell migration assays have been done with cerebral cortical cells
(e.g., Behar et al. 1998), but were also complemented by later studies using
organotypic slice cultures (e.g., Behar et al. 2000).
Organotypic cultures have an advantage when cell diversity and local environ-
ment is important for understanding influences. It is important to note that in the
hypothalamus, the plane of sectioning is critical as the tissue slice should contain the
source of the migrating cells, the destination, and any cues used by the cell to direct
migration. These are different planes of sectioning for radial hypothalamic migration
(coronal) versus migration of GnRH neurons, which originate in the nasal compart-
ment (sagittal). Tissue explants offer relatively intact systems to test hypotheses of
cellular behavior but suffer limitations of inclusion of all necessary cues, imaging
hurdles for depth, as well as tissue health deep in tissue that has been removed from
nutrient and oxygen vascular perfusion. Two approaches have been used for creating
in vitro models of GnRH neuron migration. The first was an explant model (Fueshko
and Wray 1994) and then a head slice model (Tobet et al. 1996b). Both models have
been used to test the role of GABA; first in explants and then in head slices (see
Wierman et al. 2011). Live views of GABA influences on GnRH neurons were
discerned in slices (Bless et al. 2005) and then in explants (Casoni et al. 2012). The
slice model maintained the arrangements of axonal fiber guides as they existed
in vivo, while the explant model allows free extension of those axons away from
their nasal origins. Each technique has its place, and all have been successfully used
to shed light on many aspects of neurodevelopment.
194 M. Stratton and S. Tobet

7.6 Manipulating GABA Signaling Alters Hypothalamic Cell

Behavior

Complementary Pharmacology and Molecular Genetics Results from

organotypic brain slice experiments suggest that GABA acts via GABAA and
GABAB receptors to restrict cell movements in the hypothalamus. This is apparent
as receptor agonists slow cell movement speeds while receptor antagonists increase
cell movement speeds, indicating that basal migration may be dampened by endoge-
nous GABA gradients emanating from developmental boundaries. This appears true
for GnRH neurons, as well as for cells in the VMH and PVN that are discussed
separately below. The availability of selective models of genetic disruption makes it
possible to consider complementary studies of in vivo function relative to secreted
factors (gene product, e.g., WNTs or BDNF, or enzymes that produce ligands of
interest, e.g., GAD for GABA) and the receptors for any secreted factor. Gene
disruptions are not simple in interpretation. Multi-subunit structures like the
GABAA receptor present the challenge of choosing which of 21 subunits to disrupt
(not all of which impact function and there are no more than 5 subunits per receptor).
Since there are several routes to GABA production, it is also unclear what GABA-
producing enzymes would be most beneficial to disrupt to assay GABA effects in
particular aspects of development in particular regions. Nonetheless, the complemen-
tary data of pharmacology and genetics helps draw more generalizable conclusions.

GnRH Neurons: Pharmacology First tested via in utero exposure then

organotypic head slices exposed to the GABAA receptor agonist muscimol or
antagonist bicuculline, GABA was shown to regulate the transition of migrating
GnRH neurons exiting the nasal compartment and entering the brain at the cribri-
form plate (Fig. 7.5). Muscimol decreased the number of GnRH neurons in the

Fig. 7.5 Schematic diagram illustrates GABAA receptor activation decreasing GnRH neuron exit
from the nasal compartment. (a) Montage shows immunochemically labeled GnRH neurons in
migration in a mouse sagittal brain section at E15 (adapted from Tobet et al. 2001). (b) GnRH
neurons originate in the vomeronasal organ (VNO) and migrate toward the brain along olfactory
axons, crossing the cribriform plate. (c) When slice cultures are treated with a GABAA receptor
agonist, fewer GnRH neurons cross the cribriform plate. Additional molecular complexity is present
in this system: GABA-producing GnRH neurons are shown in red and GABAA receptor containing
cells marked with red tags. BF Basal forebrain, OE olfactory epithelia, AOB accessory olfactory
bulb, OB olfactory bulb
7 Hypothalamic Development: Role of GABA 195

olfactory bulb and increased their number in the nasal compartment. Meanwhile,
bicuculline treatment resulted in GnRH neurons more scattered in the brain and
olfactory bulb (Bless et al. 2000). When fluorescence video microscopy was used to
measure acute effects of GABA signaling on GnRH cell movement, GABAA
receptor blockade resulted in increased migration and decreased turning (Bless
et al. 2005). These results were confirmed in explants where picrotoxin, a GABAA
receptor antagonist increased total distance traveled for GnRH neurons while
muscimol again restricted movement (Casoni et al. 2012). Interestingly, the work
addressing GABA’s role in GnRH cell movements was preceded by work showing
that GABA regulates GnRH electrophysiology in explant cultures started from
mouse embryos at E11.5 (Kusano et al. 1995) when GnRH neuron birth and
migration is just beginning. Interpretations of GABA effects on GnRH migration
are complicated by the fact that ~30% of migrating GnRH neurons are themselves
GABA positive and thus are not simply receivers of GABA cues (see genetic
manipulation below).

GnRH Neurons: Genetics Given the complications of multiple sources of GABA

along the migratory path of GnRH neurons, a gain-of-function approach was used in
which the synthetic enzyme for GABA—glutamic acid decarboxylase—was driven
by the GnRH promoter. This approach was attractive as a minority subset of GnRH
neurons already express GAD. Transgenic mice with GAD67 selectively
overexpressed in GnRH neurons increased the number impacted directly by
GABA, but still only influenced a percentage of the total population (reviewed in
Wierman et al. 2011). In contrast, GAD67 null mice displayed increased numbers of
GnRH neurons out of the nasal placode at E14.5 and E17.5, although the changes
were not balanced along the migratory route, making assessment of cell migration
difficult to evaluate (Lee et al. 2008). More recent data using GAD67 and GAD65
knockout mice showed differential effects of knocking out the two forms of GAD
and although the results were slightly different from the earlier study, the bottom line
was a confirmation for a role of GABA in GnRH neuronal development (Vastagh
et al. 2015). The role of GABAA receptors was probed using mice in which the
γ2 subunit was disrupted (Simonian et al. 2000), but this had no apparent effect on
the distribution of GnRH neurons. There was a preliminary examination of mice
with GABAA, β3 subunit knockdown using cohorts of mice used for studies of VMH
(Dellovade et al. 2001); however, the results were inconclusive.

VMH: Pharmacology Manipulation of GABAA receptors affected cell movements

and the distribution of identified cells in the region of the VMH when examined
ex vivo (Dellovade et al. 2001). When treated with the antagonist bicuculline, the
number of migrating cells increased (Dellovade et al. 2001) and the speed of
movement increased (McClellan et al. 2008). Since bicuculline treatment also
decreased the movement of neurons along the orientation of radial glial fibers in
the central VMH, a possible mechanism of GABA action may be the regulation cell
associations with radial glial fibers. Interestingly, at the extreme medial and lateral
regions of the VMH, where GABA concentrations should be highest, cell
movements are typically perpendicular to the radial fibers. When treated with
196 M. Stratton and S. Tobet

bicuculline, movement patterns in these areas switched to being more parallel to the
orientation of radial fibers. This data, in agreement with work done in the cortex
(Behar et al. 1998), indicates that the functional result of GABA signaling can
change based on location, GABA concentration, or developmental stage. GABA
signaling through the GABAB receptor also has a strong influence on VMH devel-
opment. Activation of GABAB receptors with baclofen decreased cell movements in
the VMH (Davis et al. 2002) while inhibition with saclofen increased cell movement
speed (McClellan et al. 2008).

VMH: Genetics Many GABAA receptor subunits are expressed in the developing
hypothalamus, some of which are expressed in subregions. As the β3 subunit is
highly expressed in the VMH and its disruption was known to reduce GABAA
receptor binding by 50% across the brain, β3 subunit-disrupted mice were a logical
place to start dissecting the role of GABAA receptors from a genetic perspective.
Mice lacking GABAA β3 subunits had larger VMH areas and cells containing ERα
were more widely dispersed than in controls (Dellovade et al. 2001). In mice lacking
functional GABAB receptors, the size of the VMH also was larger and there was
increased dorsal-to-ventral spread of cells containing ERα in the ventrolateral
quadrant (McClellan et al. 2008). Both of these findings are consistent with the
hypothesis that GABA produced at the lateral edges of the VMH provides a slowing
or stop signal (Fig. 7.4).

PVN: Pharmacology The influence of GABAA receptor signaling on cell

movements in this region has not been robustly characterized, however, when treated
in utero with bicuculline, the PVN and lateral hypothalamic areas contain signifi-
cantly fewer cells that express ERα (McClellan et al. 2010), pointing to potential
roles in proliferation, survival, or gene expression regulation. More of the work
investigating GABA influences in the development of the PVN has centered on
actions through the GABAB receptor. When the GABAB receptor was chemically
inhibited in utero, a lateral shift in the placement of cells containing ERα was noted
in neonates, which was preserved through adulthood. This corresponded with
increased cell movement speeds in tissue slices taken from the embryonic brain
when the GABAB receptor was inhibited with antagonists, either saclofen or CGP
55845, suggesting that GABA acts to slow or stop migrating cells at predefined
boundaries of the immature PVN (Stratton et al. 2014).

PVN: Genetics It was not surprising that the disruption of the β3 subunit of the
GABAA receptor did not impact the PVN as β3 subunit expression in the PVN was
found to be quite low (McClellan et al. 2010). On the other hand, in mice lacking
functional GABAB receptors, the cytoarchitecture of the PVN was altered. Specifi-
cally, relative to littermate controls, cells containing nNOS were less tightly packed,
cells containing ERα are found more laterally, and there was decreased expression of
BDNF in the PVN (McClellan et al. 2010) while CRH expression was increased
(Stratton et al. 2011). There was an unexpected finding of reduced blood vessel
density in these mice (Frahm et al. 2012) that provided a new target and new
7 Hypothalamic Development: Role of GABA 197

mechanisms hypothalamic differentiation and disruption based on glucocorticoid

actions.

GABA Receptor Synergy? As alluded to above, GABA may influence cell posi-
tioning by affecting the guidance of cells by neuronal (as per GnRH; Bless et al.
2000) or glial (as per VMH; Dellovade et al. 2001) fibers. Manipulation of GABAA
receptor signaling altered the orientation of cell movements (Dellovade et al. 2001),
whereas GABAB receptor signaling decreased the rate of cell movement (Davis et al.
2002). Since the two receptors may be differentially sensitive to GABA levels, the
interaction of the two mechanisms may be spatially regulated by the locations of
different concentrations of GABA (e.g., high at the nuclear boundary). It is still
unknown whether GABAA and GABAB receptors act synergistically, antagonisti-
cally, or independently to influence movements in the developing hypothalamus.

7.7 Important Discoveries Along the Way Functional

Significance

Taken together, the breadth of studies on the developing hypothalamus indicate

significant effects of GABA on cell placement and/or gene/protein expression in the
developing hypothalamus. In this section, we consider the potential functional
consequences of such early actions. When using genetic approaches, it is difficult
to link early critical period GABA impacts to adult function because GABA is a
potent neurotransmitter throughout the life span. Given that the hypothalamus
impacts many aspects of organismal homeostasis and how individuals respond to
their environment, there are many potential functions to test. While it would be best
if long-term functional consequences of GABA signaling were able to be interpreted
from in vivo experiments where the time and location of manipulation were tightly
controlled, that data does not yet exist. We will discuss below the consequences of
manipulating GABA signaling globally but for discreet developmentally confined
periods using chemicals and consequences of manipulating GABA signaling
throughout the life span using genetics. A handful of data sets allow evaluation of
the impact of GABA signaling where genetic manipulation was accomplished
selectively in GnRH cell populations throughout the life span.
By virtue of GnRH being a hypothalamic peptide that directly regulates pituitary
function, it is easy to consider what to examine for GnRH neuron failures. Never-
theless, GnRH neurons can still be difficult to evaluate for functional impact if the
endpoint is a reproductive failure. Assuming that GABA manipulations affect all of
the GnRH neurons that might be sensitive to GABA effects on movement that might
only be 30% of the GnRH neuronal population (even though virtually all GnRH
neurons contain a form of GABA receptor). Is 30% enough to impact fertility? One
study indicated that fertility in males is preserved with only 12% of the GnRH
neuronal population, while in females between 12 and 34% are necessary (Herbison
et al. 2008). Treatments that influence GnRH neurons in development also influence
reproductive functions in adults. Corresponding with developmental effects noted
198 M. Stratton and S. Tobet

above, female mice that overexpressed GAD67 in a subset of GnRH neurons

displayed altered estrus cyclicity and rates of pregnancy, although it did not disrupt
the onset of sexual maturation. GABA may also play a role in normal reproductive
function in the adult through other receptors since the examination of GABABR1
subunit null mice demonstrated again a normal onset of sexual maturation, but
abnormal estrus cyclicity and impaired fertility (reviewed in Wierman et al. 2011).
Functionally, exposure to a GABAB receptor antagonist affected stress-related
physiology and behaviors associated with the PVN. Adult mice who were embryon-
ically exposed to the GABAB receptor antagonist, CGP55845 behaved differently in
elevated plus maze, open field, tail suspension, and sucrose preference tests. These
same animals displayed altered HPA regulation as indicated by altered post-restraint
corticosterone levels and FOS activation in the PVN. Many of these same behavioral
changes are present in the GABAB R1 knockout mice, which displayed increased
anxiety-like behavior, reduced depression-like behaviors, and were also hyperactive
(reviewed in Tiao and Bettler 2007).
It is also important to highlight the interaction between sex of the animal and
susceptibility of the hypothalamus to GABA manipulation. Specifically, in the
GABAB receptor manipulation studies, there are quite dramatic sex-selective effects.
In the receptor knockout studies, CRH expression in the PVN was only affected
in females, while lateral spread of ERα-containing cells was more predominant in
females. With pharmacological manipulation of the GABAB receptor, cells in
ex vivo slices from females preferentially showed altered cell movement speeds,
and increased lateral spreads for in vivo cell placement. Finally, females exposed in
utero to GABAB receptor inhibition displayed increased anxiety-like behaviors as
adults while littermate males displayed a hyperactivity phenotype.

7.8 Perspectives

GABA is intimately tied to hypothalamic development, whether it be in the richness

of GABAergic cells within particular areas (e.g., preoptic area) or nuclei (e.g.,
arcuate) or in the more striking pattern of cells surrounding particular nuclei (e.g.,
VMH and PVN). In the PVN and VMH, there is evidence of GABAergic role(s) in
affecting cell placements or differentiation. GnRH neurons are different by nature
with a unique origin outside the CNS yet also falls into the category of being subject
to GABAergic influences on cell movements/migration. Understanding molecular
mechanisms by which GABA receptor activation impact these changes has been
more challenging, largely due to the heterogeneity of the hypothalamus. While some
insight can be gained from the investigation of mechanisms underlying GABA
effects in the cortex, these have not all been tested in the hypothalamus. Further-
more, directly demonstrating cause/effect relationships of physiological and behav-
ioral consequences to altered GABA signaling during hypothalamic development is
extremely complex and circumstantial evidence has not been conclusive. Even when
one can tie a particular cellular phenotype to a particular molecule in a particular
location using genetic targeting, it is difficult to tie that cellular phenotype to a
7 Hypothalamic Development: Role of GABA 199

particular functional endpoint at a later time point in the animal’s life. Although it
may be easier to tie a phenotype to a particular time point with chemical targeting, if
the drug is given systemically, it is still difficult to prove the functional phenotype is
brought about via chemical activity at that particular location of interest.
Despite these limitations, there is sufficient evidence to suspect phenotypes/
disorders that are related to hypothalamic function (and other brain regions) in
humans subjected to in utero exposure to drugs that manipulate GABA signaling.
Exposures to barbiturates, benzodiazepines, and other drugs (e.g., valproate) that
impact GABAergic systems can occur when the mother is being treated for epilepsy,
anxiety, or depression disorders. This is particularly true for pregnancies prior to the
mid-1980s when evidence for potential effects on the fetus began to mount. While
these drugs are at times, still used during pregnancy, it is after an informed risk–
benefit analysis. Most studies assessing effects in the fetus are limited in time of
follow-up and typically do not extend into the post-pubescent period (or even middle
age) when subtle alterations in hypothalamic function may manifest. In one rare
study that followed the offspring of mothers treated with barbiturates, potential
effects were seen in sexual preference and gender identity (Dessens et al. 1999).
What other phenotypes might be seen? For cells in hypothalamic cell groups such
as the VMH, PVN, and ARC that are all part of the classic feeding/obesity circuitry,
it is easy to consider that as a functional endpoint. The PVN also regulates systemic
glucocorticoids and autonomic tone in the cardiovascular system. Thus, develop-
mental GABA signaling in the hypothalamus represents a potential mechanism that
might link the highly comorbid conditions of depression, cardiovascular disease
(CVD), and obesity (Tobet et al. 2013). Comorbidity may point toward interactions
in adult physiology, for which obvious mechanisms exist between obesity and CVD,
and mechanisms linking depression to CVD or obesity are being developed
(Fig. 7.6). Nonetheless, a common developmental root to all three may lie in
common hypothalamic or even GABAergic etiology. Strikingly, Prader–Willi Syn-
drome (PWS) includes in its spectrum of symptoms, depression, obesity, CVD, and
reproductive phenotypes that could be due to effects in the GnRH system, and
additional behavioral phenotypes potentially rooted in the VMH. Dysregulation of
multiple hypothalamic circuits/functions is well documented in PWS patients.
Though the exact molecular roots of PWS are complex and still under investigation,
the spectrum of symptoms may well represent the potential phenotypes, albeit
severe, to be expected in humans subjected to conditions that impact the function
of GABAergic systems in the hypothalamus during early development.

Box 7.1 Multiple Roads Lead to GABA

Historically, decarboxylation of glutamic acid was thought to be the predomi-
nant if not only source of physiologically relevant gamma-aminobutyric acid
(GABA). Decarboxylase enzymes GAD1 and GAD2 efficiently remove ter-
minal COOH from glutamic acid (C5H10NO4) to form GABA (C4H9NO2)
(Roberts and Frankel 1951). More recent work has demonstrated that a number

(continued)
200 M. Stratton and S. Tobet

Box 7.1 (continued)

of enzymes have the capacity to convert pyrroline (C4H7N) and other four
carbon amine containing molecules to GABA (e.g. Laschet et al. 1992). These
precursor molecules are typically sourced from amino acid metabolism.
Mechanisms that utilize non-glutamic acid precursors are critical for GABA
production in several contexts including in glia (Laschet et al. 1992) and in
dopamine neurons of the mid-brain (Kim et al. 2015). GABA can also be
converted to succinic semialdehyde (C4H6O3) and subsequently succinate
(C4H6O4) for use in the tricarboxylic acid (TCA) cycle (e.g. Bessman et al.
1953). The TCA cycle generates electromotive force in the mitochondria for
ATP production. Precursors of glutamic acid include glutamine (C5H10N2O3)
and α-ketoglutarate (C5H6O5) (Peng et al. 1991), another TCA cycle interme-
diate. Thus, GABA production and the myriad important functions of GABA
can be intimately linked with two ubiquitous metabolic processes, the TCA
cycle and amino acid metabolism (Fig. 7.7).

Fig. 7.6 Schematic diagram illustrates potential comorbid consequences of altered GABAergic
signaling. Hypothalamic dysfunction may provide a shared anatomical substrate for clinical
comorbidity. There are high rates of comorbidity among depression, cardiovascular disease,
obesity, and reproductive dysfunction. Altered hypothalamic development through genetic or
environmental insult could predispose individuals to one or more of the above conditions. Once
present, pathophysiology in one system is likely to disrupt physiology of the other systems and
cause additional disease. Brain circuits that regulate mood, cardiovascular function, metabolism,
and sexual function are all connected within the hypothalamus
7 Hypothalamic Development: Role of GABA 201

Fig. 7.7 The “GABA Shunt” uses α-Ketoglutarate, an intermediate of the TCA cycle, to produce
glutamic acid, which is decarboxylated by GAD enzymes to GABA. The 4-carbon chain of GABA
can re-enter the TCA cycle following conversion to succinic semialdehyde and ultimately succinate.
GABA is also produced in putrescine degradation or again from glutamic acid but derived from
glutamine. Many of these reactions are also reversible (e.g. glutamic acid can be converted to
α-ketoglutarate)

Acknowledgments We would like to thank Dr. Kristy McClellan for helpful discussion, Mr. Luke
Schwerdtfeger for expert assistance in preparing figures, and Dr. Deborah Kurrasch for providing an
original digital image for use in Fig. 7.4. MS was supported by 1K01AG056848.

Key References

Altman and Bayer (1986). This is a classic description of the birth of cells in the
diencephalon.
Frahm et al. (2012). First paper to indicate a role for regulated development of
hypothalamic vasculature.
Herbison et al. (2008). This paper is notable for being the first to directly address the
redundancy in the GnRH population needed for fertility.
Miller and Nowakowski (1988). This paper introduced the BrdU method to the
neuroscience community and kicked off a plethora of studies of cell proliferation
in the nervous system due to the relative ease compared to the use of tritiated
thymidine.
Okamura et al. (1990). This paper was the first to note the special relationship
between cell bodies for GABA synthesis based on the localization of glutamic
acid decarboxylases and how they lie outside of major hypothalamic cell groups.
202 M. Stratton and S. Tobet

Puelles and Rubenstein (2015). This paper updates history and approaches to
parcellation of the hypothalamus on the basis of large anatomical assumptions
to smaller units informed by molecular data.
Rakic (1972). This paper defined the classic view of radial glia providing migratory
guidance for neuroblasts in cerebral cortex.
Salic and Mitchison (2008). This paper provides an impetus to switch from BrdU to
EdU as a marker for DNA synthesis and thereby (indirectly) proliferation.

References
Acampora D, Postiglione MP, Avantaggiato V, Di Bonito M, Vaccarino FM, Michaud J, Simeone
A (1999) Progressive impairment of developing neuroendocrine cell lineages in the hypothala-
mus of mice lacking the Orthopedia gene. Genes Dev 13(21):2787–2800
Altman J, Bayer SA (1986) The development of the rat hypothalamus. Advances in anatomy,
embryology and cell biology. Springer, Berlin. Monograph
Aujla PK, Bora A, Monahan P, Sweedler JV, Raetzman LT (2011) The Notch effector gene Hes1
regulates migration of hypothalamic neurons, neuropeptide content and axon targeting to the
pituitary. Dev Biol 353(1):61–71
Behar TN, Schaffner AE, Scott CA, O’Connell C, Barker JL (1998) Differential response of cortical
plate and ventricular zone cells to GABA as a migration stimulus. J Neurosci 18(16):6378–6387
Behar TN, Schaffner AE, Scott CA, Greene CL, Barker JL (2000) GABA receptor antagonists
modulate postmitotic cell migration in slice cultures of embryonic rat cortex. Cereb Cortex
10(9):899–909
Bessman SP, Rosen J, Layne EC (1953) Gamma-aminobutyric acid-glutamic acid transamination in
brain. J Biol Chem 201(1):385–391
Bless EP, Westaway WA, Schwarting GA, Tobet SA (2000) Effects of gamma-aminobutyric acid
(A) receptor manipulation on migrating gonadotropin-releasing hormone neurons through the
entire migratory route in vivo and in vitro. Endocrinology 141(3):1254–1262
Bless EP, Walker HJ, Yu KW, Knoll JG, Moenter SM, Schwarting GA, Tobet SA (2005) Live view
of gonadotropin-releasing hormone containing neuron migration. Endocrinology
146(1):463–468
Casoni F, Hutchins BI, Donohue D, Fornaro M, Condie BG, Wray S (2012) SDF and GABA
interact to regulate axophilic migration of GnRH neurons. J Cell Sci 125(Pt 21):5015–5025
Davis AM, Henion TR, Tobet SA (2002) Gamma-aminobutyric acid B receptors and the develop-
ment of the ventromedial nucleus of the hypothalamus. J Comp Neurol 449(3):270–280
Davis AM, Seney ML, Stallings NR, Zhao L, Parker KL, Tobet SA (2004) Loss of steroidogenic
factor 1 alters cellular topography in the mouse ventromedial nucleus of the hypothalamus.
J Neurobiol 60(4):424–436
Dellovade TL, Young M, Ross EP, Henderson R, Caron K, Parker K, Tobet SA (2000) Disruption
of the gene encoding SF-1 alters the distribution of hypothalamic neuronal phenotypes. J Comp
Neurol 423(4):579–589
Dellovade TL, Davis AM, Ferguson C, Sieghart W, Homanics GE, Tobet SA (2001) GABA
influences the development of the ventromedial nucleus of the hypothalamus. J Neurobiol
49(4):264–276
Dessens AB, Cohen-Kettenis PT, Mellenbergh GJ, vd Poll N, Koppe JG, Boer K (1999) Prenatal
exposure to anticonvulsants and psychosexual development. Arch Sex Behav 28(1):31–44
Frahm KA, Schow MJ, Tobet SA (2012) The vasculature within the paraventricular nucleus of the
hypothalamus in mice varies as a function of development, subnuclear location, and GABA
signaling. Horm Metab Res 44(8):619–624
7 Hypothalamic Development: Role of GABA 203

Frangaj A, Fan QR (2018) Structural biology of GABAB receptor. Neuropharmacology

136(Pt A):68–79. https://doi.org/10.1016/j.neuropharm.2017.10.011
Fueshko S, Wray S (1994) LHRH cells migrate on peripherin fibers in embryonic olfactory explant
cultures: an in vitro model for neurophilic neuronal migration. Dev Biol 166(1):331–348
Herbison AE, Porteous R, Pape JR, Mora JM, Hurst PR (2008) Gonadotropin-releasing hormone
neuron requirements for puberty, ovulation, and fertility. Endocrinology 149(2):597–604
Jursky F, Nelson N (1996) Developmental expression of GABA transporters GAT1 and GAT4
suggests involvement in brain maturation. J Neurochem 67(2):857–867
Kim K, Ganesan S, Luo SX, Wu Y, Park E, Huang EJ, Chen L, Ding JB (2015) Aldehyde
dehydrogenase 1a1 mediates a GABA synthesis pathway in midbrain dopaminergic neurons.
Science 350(6256):102–106
Kusano K, Fueshko S, Gainer H, Wray S (1995) Electrical and synaptic properties of embryonic
luteinizing hormone-releasing hormone neurons in explant cultures. Proc Natl Acad Sci U S A
92(9):3918–3922
Laschet J, Grisar T, Bureau M, Guillaume D (1992) Characteristics of putrescine uptake and
subsequent GABA formation in primary cultured astrocytes from normal C57BL/6J and
epileptic DBA/2J mouse brain cortices. Neuroscience 48:151–157
Le TN, Zhou QP, Cobos I, Zhang S, Zagozewski J, Japoni S, Vriend J, Parkinson T, Du G,
Rubenstein JL, Eisenstat DD (2017) GABAergic interneuron differentiation in the basal fore-
brain is mediated through direct regulation of glutamic acid decarboxylase isoforms by Dlx
homeobox transcription factors. J Neurosci 37(36):8816–8829
Lee JM, Tiong J, Maddox DM, Condie BG, Wray S (2008) Temporal migration of gonadotrophin-
releasing hormone-1 neurones is modified in GAD67 knockout mice. J Neuroendocrinol
20(1):93–103
Li S, Kumar TP, Joshee S, Kirschstein T, Subburaju S, Khalili JS, Kloepper J, Du C, Elkhal A,
Szabó G, Jain RK, Köhling R, Vasudevan A (2018) Endothelial cell-derived GABA signaling
modulates neuronal migration and postnatal behavior. Cell Res 28(2):221–248
Lu F, Kar D, Gruenig N, Zhang ZW, Cousins N, Rodgers HM, Swindell EC, Jamrich M,
Schuurmans C, Mathers PH, Kurrasch DM (2013) Rax is a selector gene for mediobasal
hypothalamic cell types. J Neurosci 33(1):259–272
Luján R, Shigemoto R, López-Bendito G (2005) Glutamate and GABA receptor signalling in the
developing brain. Neuroscience 130(3):567–580
McClellan KM, Calver AR, Tobet SA (2008) GABAB receptors role in cell migration and
positioning within the ventromedial nucleus of the hypothalamus. Neuroscience
151(4):1119–1131
McClellan KM, Stratton MS, Tobet SA (2010) Roles for gamma-aminobutyric acid in the develop-
ment of the paraventricular nucleus of the hypothalamus. J Comp Neurol 518(14):2710–2728
Miller MW, Nowakowski RS (1988) Use of bromodeoxyuridine-immunohistochemistry to exam-
ine the proliferation, migration and time of origin of cells in the central nervous system. Brain
Res 457(1):44–52
Okamura H, Abitbol M, Julien JF, Dumas S, Bérod A, Geffard M, Kitahama K, Bobillier P,
Mallet J, Wiklund L (1990) Neurons containing messenger RNA encoding glutamate decarbox-
ylase in rat hypothalamus demonstrated by in situ hybridization, with special emphasis on cell
groups in medial preoptic area, anterior hypothalamic area and dorsomedial hypothalamic
nucleus. Neuroscience 39(3):675–699
Peng L, Schousboe A, Hertz L (1991) Utilization of alpha-ketoglutarate as a precursor for
transmitter glutamate in cultured cerebellar granule cells. Neurochem Res 16:29
Puelles L, Rubenstein JL (2015) A new scenario of hypothalamic organization: rationale of new
hypotheses introduced in the updated prosomeric model. Front Neuroanat 9:27. https://doi.org/
10.3389/fnana.2015.00027
Rakic P (1972) Mode of cell migration to the superficial layers of fetal monkey neocortex. J Comp
Neurol 145(1):61–83
204 M. Stratton and S. Tobet

Roberts E, Frankel S (1951) Glutamic acid decarboxylase in brain. J Biol Chem 188:789–795
Salic A, Mitchison TJ (2008) A chemical method for fast and sensitive detection of DNA synthesis
in vivo. Proc Natl Acad Sci U S A 105(7):2415–2420
Schwanzel-Fukuda M, Pfaff DW (1989) Origin of luteinizing hormone-releasing hormone neurons.
Nature 338(6211):161–164
Schwenk J, Pérez-Garci E, Schneider A, Kollewe A, Gauthier-Kemper A, Fritzius T, Raveh A,
Dinamarca MC, Hanuschkin A, Bildl W, Klingauf J, Gassmann M, Schulte U, Bettler B, Fakler
B (2016) Modular composition and dynamics of native GABAB receptors identified by high-
resolution proteomics. Nat Neurosci 19(2):233–242
Simonian SX, Skynner MJ, Sieghart W, Essrich C, Luscher B, Herbison AE (2000) Role of the
GABA(A) receptor gamma2 subunit in the development of gonadotropin-releasing hormone
neurons in vivo. Eur J Neurosci 12(10):3488–3496
Stratton MS, Searcy BT, Tobet SA (2011) GABA regulates corticotropin releasing hormone levels
in the paraventricular nucleus of the hypothalamus in newborn mice. Physiol Behav
104(2):327–333
Stratton MS, Staros M, Budefeld T, Searcy BT, Nash C, Eitel C, Carbone D, Handa RJ, Majdic G,
Tobet SA (2014) Embryonic GABA(B) receptor blockade alters cell migration, adult hypotha-
lamic structure, and anxiety- and depression-like behaviors sex specifically in mice. PLoS One
9(8):e106015
Suter KJ, Wuarin JP, Smith BN, Dudek FE, Moenter SM (2000) Whole-cell recordings from
preoptic/hypothalamic slices reveal burst firing in gonadotropin-releasing hormone neurons
identified with green fluorescent protein in transgenic mice. Endocrinology 141(10):3731–3736
Temple JL, Wray S (2005) Developmental changes in GABA receptor subunit composition within
the gonadotrophin-releasing hormone-1 neuronal system. J Neuroendocrinol 17(9):591–599
Tiao J, Bettler B (2007) Characteristics of GABAB receptor mutant mice. In: The GABA receptors,
3rd edn. Springer, Berlin
Tobet SA, Paredes RG, Chickering TW, Baum MJ (1995) Telencephalic and diencephalic origin of
radial glial processes in the developing preoptic area/anterior hypothalamus. J Neurobiol
26(1):75–86
Tobet SA, Chickering TW, King JC, Stopa EG, Kim K, Kuo-Leblank V, Schwarting GA (1996a)
Expression of gamma-aminobutyric acid and gonadotropin-releasing hormone during neuronal
migration through the olfactory system. Endocrinology 137(12):5415–5420
Tobet SA, Hanna IK, Schwarting GA (1996b) Migration of neurons containing gonadotropin
releasing hormone (GnRH) in slices from embryonic nasal compartment and forebrain. Brain
Res Dev Brain Res 97(2):287–292
Tobet SA, Henderson RG, Whiting PJ, Sieghart W (1999) Special relationship of gamma-
aminobutyric acid to the ventromedial nucleus of the hypothalamus during embryonic develop-
ment. J Comp Neurol 405(1):88–98
Tobet SA, Bless EP, Schwarting GA (2001) Developmental aspect of the gonadotropin-releasing
hormone system. Mol Cell Endocrinol 185(1–2):173–184
Tobet SA, Walker HJ, Seney ML, Yu KW (2003) Viewing cell movements in the developing
neuroendocrine brain. Integr Comp Biol 43(6):794–801
Tobet SA, Handa RJ, Goldstein JM (2013) Sex-dependent pathophysiology as predictors of
comorbidity of major depressive disorder and cardiovascular disease. Pflugers Arch
465(5):585–594
Vastagh C, Schwirtlich M, Kwakowsky A, Erdélyi F, Margolis FL, Yanagawa Y, Katarova Z,
Szabó G (2015) The spatiotemporal segregation of GAD forms defines distinct GABA signaling
functions in the developing mouse olfactory system and provides novel insights into the origin
and migration of GnRH neurons. Dev Neurobiol 75(3):249–270
Wierman ME, Kiseljak-Vassiliades K, Tobet S (2011) Gonadotropin-releasing hormone (GnRH)
neuron migration: initiation, maintenance and cessation as critical steps to ensure normal
reproductive function. Front Neuroendocrinol 32(1):43–52
7 Hypothalamic Development: Role of GABA 205

Wray S, Nieburgs A, Elkabes S (1989) Spatiotemporal cell expression of luteinizing hormone-

releasing hormone in the prenatal mouse: evidence for an embryonic origin in the olfactory
placode. Brain Res Dev Brain Res 46(2):309–318
Wu C, Sun D (2015) GABA receptors in brain development, function, and injury. Metab Brain Dis
30(2):367–379
Yoon BE, Lee CJ (2014) GABA as a rising gliotransmitter. Front Neural Circuits 8:141
Epigenetic and Transcriptional Regulation
of the Reproductive Hypothalamus 8
Carlos Francisco Aylwin and Alejandro Lomniczi

Abstract
During the last few years, new genes required for the advent of puberty have been
identified thanks to the use of new deep sequencing associated technologies.
Evidence has emerged suggesting that these genes, organized into functional
networks, are regulated by a mechanism of epigenetic control. While the
Polycomb group (PcG) of transcriptional repressors is in charge of the repressive
arm of this mechanism, the Trithorax Group (TrxG) of epigenetic activators
counteract the PcG-mediated silencing during the infantile to juvenile stage
where increased Kiss1 expression is needed for activation of gonadotropin-
releasing hormone secretion from the hypothalamus. Here we will discuss the
recent advances and tools used for large-scale genomic and epigenomic studies of
the developing hypothalamus.

Keywords
Epigenetic regulators · Female puberty gene networks · GnRH neurons · KNDy
neurons · Reproductive development · Transcriptional activators · Transcriptional
repressors

8.1 Introduction

Puberty begins when pulsatile gonadotropin-releasing hormone (GnRH) release

from neurosecretory neurons of the hypothalamus increases in a diurnal manner
for a prolonged period of time, inducing the pituitary gland to increase its pulsatile
release of luteinizing hormone (LH). Changes in trans-synaptic and glial inputs to

C. F. Aylwin · A. Lomniczi (*)

Division of Neuroscience, Oregon National Primate Research Center, Oregon Health and Science
University, Beaverton, OR, USA
e-mail: [email protected]

# Springer Nature Switzerland AG 2020 207

S. Wray, S. Blackshaw (eds.), Developmental Neuroendocrinology, Masterclass in
Neuroendocrinology 9, https://doi.org/10.1007/978-3-030-40002-6_8
208 C. F. Aylwin and A. Lomniczi

GnRH neurons are responsible for the increased GnRH secretion. These changes
involve a reduction in inhibitory trans-synaptic inputs combined with increased
trans-synaptic and glial excitatory inputs to the GnRH neuronal network (Ojeda
and Skinner 2006).
The time of puberty has a strong genetic component, but very recently epigenetic
mechanisms have been implicated as part of the regulatory control of the develop-
mental process by which GnRH release is kept in check during infancy, and also
during the increase in GnRH secretion that brings about the pubertal process.
Agreeing with this emergent concept, a central target of epigenetic control is the
transcriptional machinery of neurons in charge of stimulating GnRH release. In this
chapter, we review key findings as well as research methodologies used to support
the existence of an epigenetically controlled gene regulatory network residing at the
core of the neuroendocrine process that controls pubertal timing.

8.2 Neuroendocrine Regulatory Mechanisms Involved

in the Hormonal Changes of Puberty

An increase in pulsatile LH release is the first endocrine indication of the initiation of

puberty (Boyar et al. 1972). In peripubertal girls, before any somatic manifestation
of pubertal development is evident, the amplitude of LH pulses detected in the
systemic circulation increases shortly after the initiation of sleep in the absence of
changes in circulating ovarian steroid levels (Boyar et al. 1972). Increased LH
pulsatility is also observed in rodents and occurs in the afternoon hours at the end
of the juvenile period, at postnatal days (PND) 28–30 (Urbanski and Ojeda 1985,
1987). As a consequence of this, gonadal output of sex steroids rises, leading to the
development of secondary sexual characteristics. In females, another event needs to
occur for puberty to be accomplished; the first preovulatory surge of gonadotropins,
which occurs upon maturation of the central mechanism of estrogen positive feed-
back. It is clear that the pubertal alteration in LH output is determined by an increase
in pulsatile GnRH release from the hypothalamus [reviewed in Ojeda and Terasawa
(2002)]. GnRH neurons are able to synthesize and release GnRH long before
puberty, thus the ultimate responsibility for the initiation of puberty rests on the
neuronal and glial networks that control GnRH neuron excitability (Ojeda and
Terasawa 2002).
It is considered that the primary trans-synaptic mechanism underlying pulsatile
GnRH release to be the synchronized action of a subclass of arcuate nucleus (ARC)
neurons termed KNDy neurons (Lehman et al. 2010; Navarro et al. 2011) because
they produce Kisspeptin, NKB (neurokinin B), and Dynorphin (Navarro et al. 2011;
Wakabayashi et al. 2010). NKB released from KNDy neurons acts on other KNDy
neurons via tachykinin receptors to stimulate kisspeptin release (Navarro et al. 2011;
Wakabayashi et al. 2010). NKB and kisspeptin are released in a rhythmic fashion
that appears to be determined by an inhibitory, periodic effect of dynorphin on NKB
release (Meijs-Roelofs et al. 1985; Navarro et al. 2011). Direct evidence for a role of
KNDy neurons in the genesis of pulsatile LH release was recently provided (Beale
et al. 2014). A second population of kisspeptin neurons, which do not contribute to
8 Epigenetic and Transcriptional Regulation of the Reproductive Hypothalamus 209

the control of pulsatile GnRH release, is located in the anteroventral periventricular

nucleus (AVPV) of rodents [reviewed in Pinilla et al. (2012)]. These neurons are
involved in the preovulatory surge of gonadotropins (Pinilla et al. 2012) and they do
not seem to be involved in the initiation of female puberty, because the gonadotropin
surge occurs only after the pubertal process is well underway.
During the prepubertal period (or infancy), the secretory activity of GnRH
neurons is under predominant trans-synaptic inhibitory control. At puberty, this
inhibition is lifted while there is a concomitant increase of excitatory inputs to the
GnRH network (Ojeda 1991; Ojeda and Terasawa 2002). This concept is strongly
supported by the evidence that activation of kisspeptin neurons, which provide a
substantial portion of the stimulatory inputs controlling GnRH neurons [reviewed in
Pinilla et al. (2012)], is essential for puberty to occur (de Roux et al. 2003; Seminara
et al. 2003).
Recent evidence suggests that this trans-synaptic regulatory mechanism is con-
trolled by a molecular switch that regulates the timing of puberty by epigenetically
coordinating the transcriptional activity of neurons implicated in stimulating GnRH
release (Lomniczi et al. 2013a, 2015; Toro et al. 2018).

8.2.1 Gene Networks Involved in the Neuroendocrine Control

of Puberty

With the advent of new high-throughput genomic technologies, coupled with a novel
and developing systems biology tools, great advances have been made in determin-
ing new pathways involved in the neuroendocrine control of puberty. This novel
approach is based on the concept that a diversity of genes that affect the time of
puberty are functionally organized into networks capable of generating a coordinated
output of biological signals. About 10 years ago, using expression data from female
rats and monkeys, we described the first gene network operating in the peripubertal
hypothalamus. This gene network is organized into a structure containing “central”
nodes that reside at the core of the network, and a number of peripherally located
subordinate genes that are transcriptionally controlled by the central nodes. The
central nodes of this network were earlier identified as being implicated in tumor
suppression/tumor formation (Roth et al. 2007), now referred to as Tumor-Related
Genes (TRGs). The TRG network has five central nodes or “hubs”: CDP/CUTL1/
CUX1, MAF, p53, YY1, and USF2. These hubs are strongly connected to each other
and also to additional upper-echelon genes (OCT2, TTF1, and EAP1) previously
found to be involved in the transcriptional regulation of the pubertal process (Heger
et al. 2007; Mastronardi et al. 2006; Ojeda et al. 1999), and TRGs themselves
(Lomniczi et al. 2013b).
We (Heger et al. 2007; Lomniczi et al. 2012; Mueller et al. 2011, 2012) and others
(Cukier et al. 2013) have experimentally verified the role of EAP1 as a central
transcriptional regulator of subordinate genes like KISS1 and GnRH. Interestingly,
KISS1 was formerly known as a Metastasis Suppressor or Metastin (Steeg et al.
2003), while EAP1 was described as part of a transcriptional repressive complex that
regulates apoptosis in breast cancer (Yeung et al. 2011). Most recently, we found a
210 C. F. Aylwin and A. Lomniczi

series of transcriptional regulators with epigenetic capabilities to be involved in the

transcriptional control of the Kiss1 gene. We validated the function of EED, a
member of the Polycomb family of transcriptional repressors (Lomniczi et al.
2013a) as well as MLL1 and MLL3 (Toro et al. 2018), two members of the Trithorax
complex of transcriptional activators. These two families, also involved in cancer
biology, are in charge of epigenetically controlling the expression pattern of Kiss1
and Tac3 during the infantile–pubertal transition (Lomniczi et al. 2013a; Toro et al.
2018). Finally, SynCAM1 (also known as a tumor suppressor of lung cancer, TSLC1)
plays a central role in the developmental control of glia to GnRH neuron adhesive
communications. This adhesion molecule is part of the mechanism by which astro-
cytic erbB4 receptors facilitate GnRH release and modulate female reproductive
function (Sandau et al. 2011a, b).
In addition to the TRG network, there is a posttranscriptional repressor structure
that may contribute to pubertal development. LIN28b encodes an RNA-binding
protein that prevents the maturation of let7 miRNAs (Hagan et al. 2009; Heo et al.
2009; Lehrbach et al. 2009), a family of microRNAs with tumor suppressor activity
(Chang et al. 2009). Genome-wide association studies found a single nucleotide
polymorphism near the human LIN28B gene associated with earlier puberty and
shorter stature in girls (Ong et al. 2009; Perry et al. 2009; Sulem et al. 2009). In the
hypothalamus of male and female rats, the expression of Lin28b decreases during
pubertal development, while two well-known LIN28b targets, let7a and let7b,
increase (Sangiao-Alvarellos et al. 2013). As the repressive effect of LIN28b on
let7a and let7b maturation diminishes, degradation of let7 target genes increases due
to increased availability of these miRNAs. It is predicted that the Lin28b-let7
posttranscriptional regulatory pathway should predominantly control inhibitory
genes in the pubertal process. Still today, there are no formal targets for this mode
of action in the hypothalamic control of pubertal development.

8.3 Genome-Wide Technologies

With the advent of relatively inexpensive whole-genome sequencing technologies

(see Box 8.1), we can today interrogate the mammalian hypothalamus for RNA
expression changes using RNA-seq (Wang et al. 2009), Global Run-On sequencing
(GRO-seq) (Garcia-Martinez et al. 2004), or Precision nuclear Run-On sequencing
(PRO-seq) (Kwak et al. 2013). Using the same sequencing technologies, we can
study gene regulation by Chromatin Immunoprecipitation (CHIP) followed by deep
sequencing (CHIP-seq), Chromatin conformation analysis like 3C, 4C, Hi-C, and
Chia-Pet (Barutcu et al. 2016) or chromatin accessibility assays to determine
nucleosome-free areas of the genome like Assay for Transposase-Accessible Chro-
matin sequencing (ATAC-seq) (Buenrostro et al. 2015), Formaldehyde-Assisted
Isolation of Regulatory Elements sequencing (FAIRE-seq) (Giresi et al. 2007) or
DNase I hypersensitive sites sequencing (DNase-seq) (Crawford et al. 2006). In this
section, we highlight the methods used to study whole-genome changes during
pubertal development.
8 Epigenetic and Transcriptional Regulation of the Reproductive Hypothalamus 211

Box 8.1 Next-Generation Sequencing Technologies

Next-generation sequencing (NGS) allows for massively parallel sequenc-

ing reactions that enable us to analyze up to hundreds of millions of sequenc-
ing products at the same time, supporting a broad range of applications such as
genome (DNA)- and transcriptome (mRNA)-based methods. Methods differ
in the sample isolation and library preparation steps and can be processed by
Illumina platforms. DNA technologies can be used to detect direct genome-
wide protein to DNA interactions (CHIP-seq), 3D DNA conformation to
identify promoter/enhancer interactions (3C, 4C, etc.) or regions of open
chromatin that allows for transcription factor binding (ATAC-seq, FAIRE-
seq, etc.). RNA technologies are mainly used to identify new transcripts,
splicing variants or to quantify mRNA synthesis globally (RNA-seq) or
nascent transcripts associated with endogenous RNA polymerase activity
(GRO and PRO-seq). Data analysis requires bioinformatics tools in order to
filter output data by passing quality controls, whole-genome alignment, quan-
titation, and statistical analysis. Experimental validation by qPCR is required
for accurate results.
212 C. F. Aylwin and A. Lomniczi

8.3.1 Phenotypic Characterization and Tissue Dissection

The first step into analyzing RNA expression or chromatin modifications throughout
critical developmental periods is to characterize the phenotype of interest, link it to a
specific day of pre- or postnatal development and to a tissue or brain region. In the rat
model, the Hypothalamus–Pituitary–Gonadal (HPG) axis remains quiescent until
postnatal day 28 (PND28) when the first endocrine manifestations of increased
GnRH pulsatility are detected as LH pulse frequency increases during the evening
hours (Urbanski and Ojeda 1985). After a few days of increased diurnal LH
pulsatility, and around PND32–34, the first preovulatory surge of gonadotropins
occurs, followed by the first ovulation. In general terms, pubertal development can
be divided into three stages: juvenile period (Juv ¼ PND21), late juvenile
(LJ ¼ PND28), and late proestrus (LP ¼ PND32–34) (Fig. 8.1).
In the rhesus macaque, we define four different stages of reproductive develop-
ment: neonatal, infantile, juvenile, and pubertal. As in humans, during the first
months to a year of life, high levels of LH pulsatility are detected in a period called
“mini-puberty” This is followed by the infantile period in which there is a marked
quiescence of the gonadotropin axis. During the early pubertal period, there is a

Fig. 8.1 Plasma LH concentrations during postnatal development in humans, rhesus macaques,
and rats. In humans and monkeys, during the early postnatal period, high LH pulsatility is detected
in a period called “mini-puberty.” This is followed by the infantile period where a “repressed”
central drive generates low LH levels. During the juvenile stage, 8–9 years of age in humans and
approximately 2 years of age for the rhesus macaques, there is a reawakening of the GnRH network
evidenced by increased LH pulsatility during the night hours. In the rodent, this same phenomenon
is first detected in the evening of postnatal days 28–29. As puberty approaches in females, ovarian
estrogen secretion increases, reaching a sustained magnitude capable of activating the positive
feedback mechanism that results in the preovulatory surge of gonadotropins. Days (d), months (m),
and years (yrs)
8 Epigenetic and Transcriptional Regulation of the Reproductive Hypothalamus 213

reawakening of the GnRH pulse generator and therefore increased LH pulsatility

generating a clear a.m./p.m. difference in LH pulse frequency. Finally, during
puberty, the rhesus macaque female is capable of sustaining ovulatory cycles with
the presence of a dominant ovulatory follicle and a sustained luteal phase followed
by menses, as is typical of primates (Fig. 8.1).
The rodent model is ideal to detect genome changes during peripubertal develop-
ment because of the anatomical nature of the GnRH network. The AVPV Kiss1–
GnRH neurons involved in the preovulatory surge of gonadotropins are distinctively
separated from the ARC KNDy neurons to GnRH terminals implicated in GnRH
pulsatility. ARC and AVPV tissue dissection can be achieved simply by rapid
microdissection using surgical Iris Scissors under binocular magnifying glasses, or
by micro-punching the area of interest from freshly prepared 400 μm vibratome
sections. The advantage of direct microdissection is that the tissue is more rapidly
recovered and flash frozen in liquid nitrogen, while sectioning and micro-punching a
more accurate region is recovered but with longer processing times (Fig. 8.2a, b).

8.3.2 RNA-Seq: mRNA Isolation and Library Preparation

A typical medial basal hypothalamic fragment weighs between 3 and 5 mg and

provides in average 5 μg of total RNA. Total RNA can be extracted using TRIzol
reagent (Invitrogen cat# 15596026) or any of the ion exchange-based methods like
the RNeasy mini kit (Qiagen cat# 74104). High-quality total RNA samples must
have an RNA Integrity Number (RIN) greater or equal to 8 and a ratio 28S/18S
rRNA of 2:1 (Fig. 8.3a). For mRNA library preparation, total RNA must be depleted

Fig. 8.2 Schematic representation of the hypothalamic region of the rat. (a) Ventral view of the rat
brain. Major regions (olfactory bulb, neocortex, optic chiasma, hypothalamus) are labeled. The
hypothalamic region, shaded in orange, is used for micro-punching or coronal slicing. (b) Coronal
view of the hypothalamic region depicting the arcuate nucleus (Arc), median eminence (ME),
ventromedial hypothalamus (VMH), and dorsomedial hypothalamus (DMD)
 214

Fig. 8.3 RNA-Seq library preparation process from rat MBH tissue. (a) Agilent Bioanalyzer electrophoretic trace of a high-quality total RNA sample. Shaded
green area shows the 18S and 28S rRNA bands. (b) Electrophoretic trace after rRNA depletion using poly-A magnetic beads. The green shaded area highlights
the absence of 18S and 28S signals depicted by dotted lines. The blue line corresponds to the purified mRNA 1000–5000 nt. (c) Electrophoretic trace after
chemical fragmentation of mRNA. Fragmented mRNA can be observed by a shift of the enriched area toward the lower molecular weight, 100–800 nt (orange
line) as opposed to intact mRNA (dotted blue line). (d) Electrophoretic trace of a finalized sequencing library after amplification using primers P5 and P7
followed by magnetic beads size selection to discard both adapter and primer dimers. An enrichment in 400–600 bp long amplicons is detected
C. F. Aylwin and A. Lomniczi
8 Epigenetic and Transcriptional Regulation of the Reproductive Hypothalamus 215

of ribosomal RNA using either active ribosome binding technology like Ribo-Zero
rRNA removal kit (Illumina cat# MRZH116) or by further isolating the poly-A
mRNA from the total RNA fraction by means of poly-(A) paramagnetic beads (Bioo
Scientific Cat# NOVA-51979). These magnetic beads are coupled to oligo
(dT) probes capable of isolating poly-A mRNA starting from 10 ng to 10 μg of
total RNA (See Box 8.2). Poly-A beads are more price effective than any of the
ribosomal depletion methods, with the caveat of generating libraries that only
represent poly-A containing mRNA. The first step is to test the best condition for
mRNA isolation. One efficient way to do this is to use equal volumes of poly-(A)
beads with increasing amounts of total RNA. Efficient removal of ribosomal RNA
can be identified by the absence of 18S and 28S fractions commonly seen in total
RNA samples in Bioanalyzer RNA nano- or pico chip runs (Agilent cat# 5067-1511
and 5067-1514) (Fig. 8.3b).

Box 8.2 RNA-Seq Library Preparation Pipeline

RNA transcripts can be isolated from total RNA by two different methods:
(a) Ribosomal depletion, a method where all ribosomal RNAs are captured and
discarded from the sample, allows for isolation of both mRNA and lncRNA
and (b) Poly-A-enrichment, specifically captures and purifies poly-A-
containing mRNAs using poly-T-containing magnetic beads. After isolation,
samples are subjected to fragmentation and cDNA synthesis. Adapter ligation
allows for the attachment of Illumina compatible adapter sequences and
barcodes for downstream sequencing and sample identification. Before
sequencing, library quantification by qPCR is necessary for accuracy.
216 C. F. Aylwin and A. Lomniczi

Several Illumina compatible kits for RNA-seq library preparation are available on
the market. In general, the basic steps to prepare a next-generation sequencing
(NGS) library include (a) fragmentation of the RNA sample; (b) converting RNA
into double-stranded DNA; (c) attaching oligonucleotide adapters to each end of the
target fragments; and (d) quantitating the library by qPCR. The first step of this
process is RNA fragmentation and can be achieved by (a) physical fragmentation
like acoustic or hydrodynamic shearing; (b) enzymatic shearing like RNase III
treatment; and (c) chemical fragmentation by incubating the mRNA sample at
95  C with divalent cations. The optimal size fragmentation for Illumina compatible
libraries ranges from 100 to 400 nt, this can also be detected by Bioanalyzer RNA
nano- or pico chip (Fig. 8.3c). After fragmentation, first strand synthesis is
performed with random primers, if it is desired to retain the strand orientation of
the transcripts, dUTP is added into the second strand synthesis, which prevents
further amplification with the PCR polymerase that is unable to recognize uracil in
the template strand (NEXTflex Rapid Directional RNA-seq kit, Bioo Scientific cat#
NOVA-5138-07). Bead cleanup using Agencourt AMPure XP Magnetic Beads is
performed in order to remove free oligonucleotides and dNTPs from the reaction
mix. This is followed by 30 end adenylation and adapter ligation. The Illumina
compatible adapters carry the barcodes used to identify the sample after sequencing
as well as the primer recognition sites used for library amplification as well as
sequencing on the Illumina flow cell. After adapter ligation, limited cycle PCR
amplification of the library is performed in order to reach the minimum amount of
DNA needed for sequencing. Size selection is performed on the library with AMPure
XP Magnetic beads, this step is needed to remove the primers from the mix that
could interfere with the sequencing reaction (Fig. 8.3d). The library is quantitated by
qPCR using the Kapa Library Quantitation Kit (Roche cat# 07960140001). Six to
eight samples with different barcodes are pooled into a single tube at a concentration
of 3 nM and send for sequencing in an Illumina HiSeq 3000 or 4000 sequencer that
delivers around 300 million reads per lane, providing about 50 million reads per
sample when using pools of six samples per sequencing lane.

8.3.3 RNA-Seq: Genome Alignment and Data Analysis

The Illumina platform delivers raw reads in fasqc formatted files. The first step is to
run quality controls on the sequencing procedure, followed by alignment of the
sequencing reads to a reference genome and quantitation of reads per gene. All
the computational tools needed to analyze NGS datasets are freely available at the
Bioconductor webpage (http://www.bioconductor.org/), and are coded on R lan-
guage (https://www.r-project.org/). Another platform widely used for high-
throughput data analysis is Galaxy (https://usegalaxy.org/), which is especially
helpful for users with less experience in programming on R. Finally, DNASTAR
(https://www.dnastar.com/) is a commercial package that fulfills all the steps from
quality control, data alignment, transcript quantitation as well as data visualization
and gene ontology analysis.
8 Epigenetic and Transcriptional Regulation of the Reproductive Hypothalamus 217

FastQC (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/) is a stan-

dard tool used for quality control of the raw reads that include GC content, adaptor
identification, duplicated reads, and overrepresented k-mers. These parameters help
identify sequencing errors and PCR artifacts. As sequencing quality decreases
toward the 30 end of the reads, the Trimmomatic (Bolger et al. 2014) tool is run to
trim the ends of the sequence and discard low-quality reads and trim adaptor
sequences. Alignment of the remaining reads to a reference genome or transcriptome
is performed by Bowtie (Langmead et al. 2009) or STAR (Dobin et al. 2013)
aligners. When transcriptome annotation is comprehensive, Bowtie performs
non-gapped alignment in a very fast and efficient manner. In the case of less studied
genomes, genome alignment helps to identify novel genes or transcripts, which is
performed with gapped or spliced mappers capable of reading the splice junctions
correctly. The most important parameters to consider when aligning transcripts are
the strandness of the library, the number of mismatches to accept and the length of
the sequenced fragments.
To estimate gene expression, featureCounts (Liao et al. 2014) quantitates the
aggregate raw counts of the previously exonic mapped reads. This measure is
affected by factors like transcript length, total number of reads, and library size. In
order to compare expression levels among samples, normalization methods are used
to correct for the aforementioned biases. RPKM (Mortazavi et al. 2008) (reads per
kilobase of exon per million reads), a within sample normalization method, will
eliminate the feature length and library size effects. This normalization method
performs well only when transcript distribution is relatively homogeneous. When
highly and differentially expressed features skew the count distribution, TPM
(Li et al. 2008) (transcripts per million reads) and TMM (Robinson and Smyth
2007) (trimmed mean of M-values) normalization methods are preferred because
they disregard highly variable and highly expressed features. Since RNA-seq quan-
titation is based on read counts that are probabilistically assigned to different
transcripts, discrete probabilistic distributions, like Poisson or the negative binomial
distribution are used to compute differential expression (Anders and Huber 2010;
Robinson et al. 2010 #7592). Several R packages are available to perform multiple
comparison analyses including edgeR (Robinson et al. 2010), DESeq (Anders and
Huber 2010), and limma-voom (Law et al. 2014). We recommend that RNA-seq
experiments be run in tri- or quadruplicates in order for the differential expression
methods to generate reproducible results in the ample spectrum of mRNA expression
levels in the brain.

8.3.4 CHIP-Seq: DNA Isolation, Fragmentation, and Library

Preparation

CHIP-seq analysis allows for the determination of transcription factor binding or

positioning of histone PTMs modification throughout the genome. We have suc-
cessfully performed several CHIP-qPCR and CHIP-seq studies from rat MBH
(Lomniczi et al. 2013a, 2015; Toro et al. 2018). The basic principle of this technique
218 C. F. Aylwin and A. Lomniczi

is immobilization of the protein–DNA interactions by crosslinking the tissue with

4% paraformaldehyde (PF) followed by chromatin extraction and fragmentation by
sonication. Finally, a fraction of the chromatin is immunoprecipitated with the
antibody of choice using protein-A or -G beads. The immune-isolated DNA
fragments are purified and used for Illumina NGS library preparation followed by
deep sequencing.
A single 5 mg MBH produces about 2–3 μg of DNA, enough material to perform
about 10–12 independent CHIP assays. First, the frozen tissue is crosslinked with
4% PF for 10 min at room temperature. This is followed by sonication in SDS lysis
buffer. Sonication needs to generate fragments of DNA with an average size of
150–500 bp. In our laboratory, we generate these fragments with about 45 s of
sonication time using a microtip sonicator. It is extremely important that the investi-
gator sets up the fragmentation protocol by testing different sonication times and
checking for DNA fragmentation in agarose gel (Fig. 8.4a) or Bioanalyzer DNA
nano chip (Agilent # 5067-4626). Fragmented chromatin is diluted, aliquoted, and
stored at 80  C. In order to determine the linear range of detection, it is necessary to
establish the best chromatin-to-antibody ratio. For that, we perform CHIP reactions
with a fixed amount of antibody and increasing volumes of chromatin followed by
qPCR of promoter regions corresponding to high and low expressed genes
(Fig. 8.4b). The selection of antibodies is crucial, several commercial antibodies
have been validated for CHIP, especially for histone modifications (see: https://
www.diagenode.com/en/categories/antibodies and https://genome.ucsc.edu/encode/
antibodies.html). If antibodies have not been validated, it is safe to assume that an
antibody that works for immunocytochemistry could work in CHIP assays. Negative
controls are usually species-specific IgGs or antibodies directed against proteins that
are not present in the sample. We have successfully used antibodies anti-Green
Fluorescent Protein or anti-Beta Galactosidase as negative controls for CHIPs from
rat hypothalamus. The antibody–protein immune complexes are isolated using
protein A or G agarose beads or paramagnetic Dynabeads (ThermoFisher cat#
10001D and 10007D). The use of paramagnetic beads has particularly improved
this technique because they have extremely low background DNA and protein
binding. DNA is recovered from the paramagnetic beads using a high pH SDS
buffer at high temperatures and purified using commercially available columns
(Zymo research cat# D5205). Qubit (ThermoFisher cat# Q33226) fluorometric
quantitation is used to determine the DNA concentration that is usually in the
range 0.2–2 ng/μl.
Several Illumina compatible kits are available for library preparation from
sub-nanogram samples. NEBNext Ultra Library Prep Kit (New England cat# NEB
E7370) or DNA SMART CHIP-Seq Kit (Takara cat# 634865) are well suited to
generate Illumina compatible libraries starting with around 1 ng of total DNA.
Library preparation is similar to the description in Sect. 8.3.3. Every biological
sample will be sequenced two times, one will correspond to the immunoprecipitated
chromatin while the other corresponds to the “Input DNA” or non-precipitated
chromatin that serves as total loading in the qPCR and as peak detection after
NGS. It is highly recommended to perform double size selection with AMPure XP
 8
Epigenetic and Transcriptional Regulation of the Reproductive Hypothalamus

Fig. 8.4 CHIP-Seq library preparation process from rat MBH tissue. (a) In order to determine the sonication time to generate optimum chromatin
219

fragmentation, the tissue is treated with increasing number of pulses followed by de-crosslinking, DNA purification, and gel electrophoresis. Increasing the
 220

Fig. 8.4 (continued) number of pulses reduces the average size of the fragments. Large fragments (>500 bp) reduce the accuracy of the CHIP procedure. The
optimum fragment size is between 150 and 500 bp. (b) Linearity of chromatin capture is determined by using a constant amount of Antibody (2–5 ug) and
increasing amounts of chromatin. After the CHIP assay is completed, cycle threshold (ct) values amplification of a specific loci by qPCR must decrease linearly
according to the amount of chromatin used (blue circles corresponding to CHIPed H3K27me3). The green shaded area shows the linear range of chromatin/
antibody interaction. The red and yellow areas depict saturated chromatin/antibody conditions where the qPCR is not capable to detect changes in DNA capture.
The orange circles correspond to an anti beta-Galactosidase antibody (βGAL) used as negative control showing no efficient chromatin/antibody interactions. (c)
Agilent Bioanalyzer electrophoretic trace of an amplified chromatin library generates a signal corresponding to the barcoded library (green shaded area). The low
molecular weight peak corresponds to adapter dimers (red shaded area), while the high molecular weight peaks are considered artifacts due to overamplification
of the library (yellow shaded area). (d) Electrophoretic trace of an amplified chromatin library after double size selection to discard both low- and high-molecular
peaks. First, and to cleanup large molecular weight fragments, a high magnetic bead:library ratio (0.8–1) is used to bind large DNA fragments while the library
and the small fragments remain unbound. The soluble unbound DNA is transferred into a new tube and a low magnetic bead:library ratio (0.2) is used, the
desired library is captured by the beads, while the small fragments remain unbound. After several washes, the bound library fragments are eluted
C. F. Aylwin and A. Lomniczi
8 Epigenetic and Transcriptional Regulation of the Reproductive Hypothalamus 221

Magnetic beads before PCR amplification of the library, to remove free unligated
adapters and high molecular weight DNA that has been selected during immunopre-
cipitation (Fig. 8.4c, d). To correct for technical variability during CHIP/Library
preparation, it is suggested to add a spike-in control to each CHIP sample (Egan et al.
2016). We usually add 1% of chromatin from Drosophila and antibodies to immu-
noprecipitate histone H2Av, only present in this species. Since this step will pull
down the same amount of Drosophila histone 2Av between samples, it allows for
normalization of the CHIP-seq values from the mammalian genome to the total
number of reads obtained from the Drosophila genome.

8.3.5 CHIP-Seq: Genome Alignment and Data Analysis

Platforms used for CHIP-seq analysis are the same as for RNA-seq. It is estimated
that a reading depth of about 20 million reads per sample is adequate to detect
transcription factor binding in the mammalian genome. Proteins with more binding
sites as well as histone marks require up to 60 million reads. After assessment of the
reads quality metrics, a mapping tool like Bowtie (Langmead et al. 2009) or MAQ
(Li and Durbin 2009) is used to align the short reads to the genome. The analysis is
followed by peak calling algorithms like MACS (Zhang et al. 2008) that predict the
regions of the genome where the immunoprecipitated protein is bound. Peak callers
will provide a p-value and FDR as enrichment metrics for each peak. For enzyme or
transcription factor binding, peaks are readily detected, but in the case of histone
PTMs where marks are more broadly positioned, it is more accurate to define
“chromatin domains.” Quantitative analysis is based on the total read count per
region of interest. DBCHIP (Liang and Keles 2012) and MAnorm (Shao et al. 2012)
compute read density over peak regions and give a precise statistical assessment of
differentially affected regions across biological conditions. Peaks or domains
identified by CHIP-seq are coded as BED or GFF files that can be uploaded to a
genome browser for further identification of promoter and enhancer regions as well
as gene transcription start sites.

8.4 Important Discoveries Made with NGS Technologies

The first evidence of epigenetic mechanisms involved in pubertal development was

published by our laboratory (Lomniczi et al. 2013a, 2015). In these papers, we
looked for transcriptional repressors highly expressed in the MBH during the
juvenile period and lower expression as the animal reaches puberty. First, using
the rat model, we demonstrated that the Polycomb Group (PcG) of transcriptional
repressors prevents the initiation of puberty by repressing ARC Kiss1 expression
(Lomniczi et al. 2013a). During the prepubertal period, two central members of the
PcG, Eed, and Cbx7, are highly expressed within ARC kisspeptin neurons, and their
protein products are coupled to the Kiss1 promoter (Lomniczi et al. 2013a). As the
animal reaches puberty, DNA methylation of Eed and Cbx7 promoter regions
222 C. F. Aylwin and A. Lomniczi

increases, lowering their expression, inducing the loss of PcG components from the
Kiss1 promoter. This is accompanied by increased levels of the epigenetic marks
associated with gene activation like H3K9ac, H3K14ac, and H3K4me3, ultimately
inducing increased Kiss1 mRNA expression in the ARC (Lomniczi et al. 2013a). In a
second paper, and using rhesus macaques as animal model, we provided evidence
that members of the zinc finger (ZNF) family of transcriptional repressors play a role
in keeping the GnRH pulse generator in check during prepubertal development
(Lomniczi et al. 2015). Using MBH of both ovary-intact females and agonadal
male rhesus monkeys during the juvenile–pubertal transition we demonstrated that
while KISS1 and TAC3 mRNA expression increases, GATAD1 expression decreases.
GATAD1 recruits the histone eraser KDM1A, inducing a decrease in the activating
H3K4me3/2 marks at the Kiss1 and TAC3 genes 50 -regulatory regions (Lomniczi
et al. 2015; Maisonpierre et al. 1991; Shi et al. 2004). Stereotaxic delivery of
lentiviral particles overexpressing GATAD1 in the ARC of immature rats, delays
puberty, and disrupts estrous cyclicity (Lomniczi et al. 2015). The decrease in
GATAD1 and KDM1A association to the KISS1 and TAC3 promoters and the
concomitant increase in H3K4me2 abundance observed in the monkey MBH at
the time of puberty, strongly support the notion that during juvenile development
there is a mechanism of epigenetic repression, that is lifted during the reawakening
of the GnRH pulse generator (Lomniczi et al. 2015).
The finding that the chromatin landscape at the Kiss1 promoter switches from a
primarily repressed state, brought about by the PcG and GATAD1/KDM1A com-
plex, to an activating state characterized by trimethylation and acetylation of H3 at
lysines 4 and 9, respectively, during the peri-pubertal period indicates the action of
an activating transcriptional complex occurring simultaneously to the loss of inhibi-
tion. This idea was studied by RNA-seq analysis of the rat MBH during peri-pubertal
development. In this study, we found that members of the Trithorax Group (TrxG) of
transcriptional activators, well known for antagonizing PcG activity (Shilatifard
2012; Simon and Kingston 2009), increase their expression in the MBH as puberty
progresses (Toro et al. 2018). MLL1 and MLL3, two members of the TrxG complex,
act on the Kiss1 promoter and enhancer regions, respectively, to provide trans-
activational activity at the time when the inhibitory influence of the PcG complex is
declining (Toro et al. 2018). To confirm TrxG function, siRNA-mediated knock-
down of Mll1 in the ARC, inhibited Kiss1 expression and delayed puberty (Toro
et al. 2018).
Since the TrxG member MLL3 has a preference for binding at enhancer sites, and
active enhancers display high levels of H3K4me1 and H3K27ac (Kim and
Shiekhattar 2015), we performed CHIP-seq analysis of these two histone
modifications in the MBH of prepubertal rats to pinpoint potential enhancers of
the ARC Kiss1 gene. In a 12-kb segment of the Kiss1 gene 50 upstream region, we
detected two areas showing high levels of H3K4me1 and H3K27ac. CHIP-qPCR
analysis of Site 1 and Site 2 in the rat genome revealed that only Site 1 (about
3000 bp from TSS) displayed increased association of p300/CBP and H3K27ac
during prepubertal development, a hallmark of active enhancers (Kim et al. 2010;
Kim and Shiekhattar 2015). In agreement with this concept, MLL3 recruitment to
8 Epigenetic and Transcriptional Regulation of the Reproductive Hypothalamus 223

Site 1 (but not Site 2) increased during pubertal development, while the
EED-mediated H3K27m3 mark weans down (Toro et al. 2018). In order to deter-
mine the function of this new Kiss1 enhancer region, we used a CRISPR-CAS9-
KRAB epigenetic remodeling approach that inhibits the action of Mll3 by site-
specific deposition of the repressive H3K9m3 histone mark. Animals stereotaxically
injected with an AAV9 virus carrying this construct into the ARC experienced a
downregulation of Kiss1 expression, delayed puberty and disrupted estrous cyclicity
(Toro et al. 2018). In conclusion, the TrxG of epigenetic activators is capable of
controlling Kiss1 expression in the ARC, by implementing histone PTM changes at
promoter and enhancer regions around the time of puberty.

Box 8.3 The Epigenetic Yin-Yang of Puberty

During the last few years, we learned that there is a basic mechanism of
inhibition–activation that operates at a transcriptional level of ARC kisspeptin
neurons. This mechanism, of epigenetic nature, ultimately controls pulsatile
GnRH secretion. In 2013, we demonstrated that Kiss1 expression in the ARC
of prepubertal female rats is subjected to an epigenetic repressive state exerted
by the polycomb group of transcriptional repressors. At puberty, the associa-
tion of the polycomb proteins to the Kiss1 promoter declines, coinciding with
increased Kiss1 expression and changes in the chromatin structure of the Kiss1
promoter. Increased levels of two histone marks associated with gene activa-
tion are detected, while a gradual loss was found in repressive histone
modifications catalyzed by the polycomb group complex. In 2017, we deter-
mined that the increase in activating histone marks at puberty is mediated by
the action of two members of the Trithorax Group of epigenetic activators.
Using deep-sequencing techniques, we demonstrated that the Trithorax binds
to the Kiss1 promoter and enhancer regions during the pubertal period to
facilitate open chromatin states, ultimately enhancing Kiss1 expression and
hence, increased GnRH release.

8.5 Limitations and Caveats of NGS Technologies

NGS technologies, used by researchers and clinicians, have provided invaluable

information into genome variation, stability, as well as gene regulation. With the
widespread use of this technology several limitations have arisen. First, for efficient
library preparation and sequencing, DNA samples need to be above the ng range,
limiting the size of the starting material. In our hands, we start with 200–300 ng of
chromatin from rat MBH (about 1/10th of the total chromatin obtained from a single
MBH), and at the end of the CHIP protocol we end up with 500 pg to 5 ng total DNA
available for library preparation depending on the antibody used for immunoprecip-
itation. This approach, while not ideal because it generates data from a complex
mixture of different cell types, is still today the only method available to overcome
224 C. F. Aylwin and A. Lomniczi

the critical mass needed for library preparation and sequencing. While single-cell
mRNA sequencing technologies have been developed (Eberwine et al. 2014),
allowing for picogram amounts of starting material to be used, heavy amplification
is needed for sample and library preparation, resulting in uneven coverage, high
noise, and inaccurate quantitation. Needless to say, CHIP-seq technologies have
improved tremendously, but efficiency and sequence quality get compromised when
starting material lies below 104 cells (Gilfillan et al. 2012).
The second limitation is related to the massive amount of information generated
and the wide range of platforms and algorithms that can be used to analyze and
visualize the data. Although the ENCODE Consortium (https://www.encodeproject.
org/) has set the standard for the analysis of most of these technologies, researchers
have to rely on external bioinformatics cores, recruit informaticians in their
laboratories or even learn coding and statistics for high-throughput data analysis.
This can be especially challenging for small laboratories that do not have easy access
to core facilities and other informatics resources.

8.6 Gene Network Visualization

With the widespread use of NGS technologies, scientists struggle to find ways to
identify the relevant genes or genome signatures from long gene lists. For instance,
genes are scored by their differential expression or chromatin signatures between
different biological states. Using an arbitrary expression threshold (we usually use
1.5-fold expression change), gene lists are screened for the top most variable genes.
Searching for enrichment of predefined functionally related genes can be easily done
with stand-alone software like Ingenuity Pathway Analysis (IPA) from Qiagen
(https://www.qiagenbioinformatics.com/products/ingenuity-pathway-analysis/),
web-based program like DAVID Bioinformatic Database (https://david.ncifcrf.gov/
home.jsp) (Dennis et al. 2003) and GeneMANIA (https://genemania.org/) (Mostafavi
et al. 2008) or open-source options like several plugins for Cytoscape (http://www.
cytoscape.org/) (Cline et al. 2007).

8.6.1 Enrichment Analysis

Enrichment analysis is a statistically powerful method to analyze and interpret large

datasets using a priori knowledge. Most of the enrichment algorithms use Gene
Ontology (GO) annotations to identify gene sets that participate or belong to a
certain category (e.g., participation in a signaling pathway, biological process,
protein domain, chromosomal location, specific disease, and genetic perturbation).
DAVID uses fuzzy gene clusters to identify enriched gene sets, classifies them into
partially overlapped groups and display results in a tabular form (Dennis et al. 2003).
For fast and easy visualization, the plugins BiNGO (Maere et al. 2005) and ClueGO
(Bindea et al. 2009) for Cytoscape display the hierarchical structure of a network of
enriched GO terms. For annotated gene function like co-expression, co-localization,
8 Epigenetic and Transcriptional Regulation of the Reproductive Hypothalamus 225

Fig. 8.5 Gene ontology (GO) analysis of the genes with SNPs associated with age at menarche.
DAVID (https://david.ncifcrf.gov/home.jsp) GO categories: Biological Process and Molecular
Function of the 339 genes with SNPs associated with the age at menarche show a high enrichment
in, e.g., gene transcription, chemical synapses, steroid signaling, and others

genetic interactions, physical interactions, and pathway classification, web-based

GeneMANIA, as well as the plugin, for Cytoscape is a great tool for network
visualization for validated gene interactions.
As an example, we used DAVID to analyze the GO function of the genes located
nearest to the 389 SNPs reported as associated with age at menarche (Day et al.
2017). Surprisingly, this gene set is significantly enriched in functions related to
transcriptional regulation like Negative/Positive regulation of transcription, chroma-
tin binding, and DNA binding in addition to steroid hormone signaling and chemical
synaptic transmission (Fig. 8.5). To determine if this group of genes is co-regulated
in any way to the three families of epigenetic modifiers we described in our
laboratory, we ran GeneMANIA on the same data set and added the Polycomb
group of epigenetic writers (EED, EZH1, and EZH2), the epigenetic reader/eraser
226 C. F. Aylwin and A. Lomniczi

Fig. 8.6 Gene network representation of the genes with SNPs associated with age at menarche and
their GeneMANIA connections with the epigenetic machinery. The round nodes of the network are
the 339 genes associated with age at menarche, red diamonds are genes of the Polycomb Group
(PcG) of transcriptional repressors, orange diamonds represent Gatad1 and Kdm1a and green
squares are the Trithorax Group (TrxG) of transcriptional activators. The epigenetic machinery
(depicted on the first layer of the network) is connected to 229 out of 339 of the genes associated
with age at menarche (second layer of nodes). The third layer depicts the 110 genes that are not
directly connected to the epigenetic machinery. Yellow nodes are genes involved in neurotransmis-
sion, blue nodes are transcriptional repressors and cyan nodes are transcriptional activators. Gray
lines, or edges, are connections established by the GeneMANIA database representing: Genetic
interactions, Pathways, Physical interactions, Predicted, and Shared protein domains

complex (GATAD1/LSD1) and the recently described Trithorax members (MLL1

and MLL3) (Fig. 8.6). Interestingly, most of the human dataset (229 out of
339 genes) seems to be co-regulated or controlled by at least one of the epigenetic
complexes mentioned before, raising the tantalizing possibility that the majority of
8 Epigenetic and Transcriptional Regulation of the Reproductive Hypothalamus 227

the SNP’s described by Day et al. (2017) as associated with the age at menarche, are
in some way altering the epigenetic milieu of their downstream target genes.

8.6.2 Gene Co-expression Analysis

Another way to infer gene function is by co-expression network analysis, a method

that creates networks of genes whose expression across samples is highly correlated
and thus, potentially co-regulated. Gene co-expression networks only specify which
genes are active simultaneously, often indicating that they are active in the same
biological process (Amar et al. 2013; Xue et al. 2013; Pierson et al. 2015). This
approach serves in the identification of transcription factors or methylation patterns
that can affect the expression of expression modules (Glass et al. 2013). The first step
in the analysis is to express the individual relationship between genes as a correlation
measure between each pair of genes. This correlation measure describes the similar-
ity between expression patterns of each gene across all samples (Steuer et al. 2002;
Margolin et al. 2006). Typically, a Pearson’s or Spearman’s correlation, as well as a
Bayesian approach, can be used to create a co-expression network (Friedman et al.
2000). In the second step of the analysis, the associations between genes are used to
construct gene networks where the nodes represent each gene and the edges repre-
sent the strength of the co-expression correlation. The final step is to use the k-means
clustering and hierarchical clustering to identify modules of genes or “clusters” to
group genes with similar patterns of expression (D’Haeseleer 2005). These modules
can finally be studied for functional enrichment to identify overrepresented func-
tional categories using the enrichment analysis tools mentioned above.
An example of this approach is our study of the function of GATAD1 on pubertal
development. To this aim, we analyzed the network configuration and connectivity
of 147 selected transcripts important for pubertal development. We used a
compressive-sensing approach that deduces gene networks by identifying robust
co-expression links between genes (Lomniczi et al. 2015). We showed that the top
20 genes gaining cumulative connectivity are sparsely connected to themselves,
while the top 20 losing connectivity is strongly connected to themselves in the MBH
at PND14 (Fig. 8.7a). Later on, during the juvenile-to-puberty transition (at PND28)
connectivity shifts and the 20 genes gaining connectivity become more
interconnected. When overexpressing GATAD1 in the ARC, the pattern of connec-
tivity resembles more that of an immature (PND14) than of a pubertal (PND28)
animal (Lomniczi et al. 2015) (Fig. 8.7a). Gene Ontology (GO) enrichment analysis
showed that for the epigenetic repressors Kdm1a, Kdm5b, Ehmt1, Pcgf2, and Rnf2
their loss of connectivity during pubertal maturation is reversed in the presence of
GATAD1 (Fig. 8.7b). Another set of epigenetic regulators, Kdm1b, Ezh1, and
Hdac1, that gain connectivity during normal puberty, fail to do so when
overexpressing GATAD1, keeping gene connectivity in an “undeveloped” or infan-
tile stage (Fig. 8.7b). It is known that GATAD1 recruits lysine demethylase 1A
(KDM1A/LSD1) to promoter regions to induce the loss of the activating histone
3 trimethylated at lysine 4 (H3K4me3). These analyses assisted us in the
228 C. F. Aylwin and A. Lomniczi

Fig. 8.7 Effect of GATAD1 overexpression on gene co-expression networks in the hypothalamus
of rhesus macaques. (a) Inferred co-expression network between the top 20 genes losing edges
(green) and the top 20 genes gaining edges/connections (red) during prepubertal development,
based on the sum of changes in connectivity between PND14–PND21 and PND21–PND28.
Network at PND 28 with the top 20 genes losing edges (yellow) and gaining edges (blue) under
8 Epigenetic and Transcriptional Regulation of the Reproductive Hypothalamus 229

demonstration that during the infantile period, when Kiss1 expression is low, the
GATAD1/LSD1 complex is associated to the Kiss1 promoter keeping low levels of
H3K4me3. As puberty approaches and Gatad1 expression decreases, the GATAD1/
LSD1 complex is evicted from the promoter, H3K4me2/3 rises, and Kiss1 expres-
sion increases (Lomniczi et al. 2015).

8.6.3 Limitations and Caveats of Network Visualization

One of the biggest limitations is related to the animal species of choice. Most of the
interactomics databases is limited to human and/or mouse datasets. During the last
few years, efforts have been made to widen this repertoire to include rats, Drosoph-
ila, C. elegans, and others. With that in mind, researchers many times opt to use
human or mouse databases when investigating rat genomic interactions due to the
higher characterization of gene function and interactions in these species. Another
issue that comes out frequently is that not all genes are yet classified on the Gene
Ontology database, producing bias toward a subclass of previously well-studied
genes present in the dataset.
Network analysis and visualization are powerful tools for whole-genome studies,
but it is crucial for investigators to take all the precautions for the correct statistical
analysis and subsequent validation of the identified pathways or gene interactions.

8.7 Perspectives

High-throughput sequencing technologies have transformed basic biology as well as

clinical research. As sequencing costs keep falling and informatics tools become
more easily available, these technologies will be embraced by a higher number of
laboratories involved in neuroendocrine research. With this in mind, it is key that
data become openly available for the discovery of common pathways across differ-
ent models.
We just started uncovering the first epigenetic complexes involved in normal
pubertal progression, but the technologies discussed in this chapter will be crucial in
ä

Fig. 8.7 (continued) GATAD1 overexpression as compared with the network inferred at PND28.
In this network, genes that are among the top 20 genes gaining edges during normal prepubertal
development are shown as triangles, while genes that are among the top 20 losers of connectivity
during development are shown as squares. For all networks, positive correlation edges are
represented by solid black lines, negative correlation edges are shown as dashed lines. (b) Changes
in the connectivity of genes belonging to selected transcription-related functional categories that
lose (Gene Ontology category GO:0045892: negative regulation of transcription, DNA templated)
or gain (Protein Information Resource, http://pir.georgetown.edu/; Keyword “transcription regula-
tion”) connectivity (green and red bars, respectively) from PND14 to PND28. For each gene, the
alteration in connectivity resulting from GATAD1 overexpression as compared with PND28 is
shown as blue bars. Modified from Lomniczi et al. (2015)
230 C. F. Aylwin and A. Lomniczi

studying alterations of reproductive development induced by external factors like

metabolism, endocrine disruptors, or changes in circadian rhythms.

Acknowledgments We thank every past member of our laboratory who has contributed to the
understanding of the genomic events involved in the maturation of the reproductive hypothalamus.
The authors are supported by the National Institute of Health (1R01HD084542 and
P51OD011092).

Key References

Boyar et al. (1972). This study was the first demonstrating that LH pulsatility
increases before sleep in pubertal patients.
Lomniczi et al. (2013). This study describes how the Polycomb family of epigenetic
repressors control the epigenetic landscape at the Kiss1 promoter region, keeping
ARC Kiss1 transcription low during the infantile period.
Lomniczi et al. (2015). This study describes how the GATAD1/LSD1 controls the
epigenetic landscape at the Kiss1 promoter region in rhesus macaques, keeping
ARC Kiss1 transcription low during the infantile period.
Roth et al. (2007). This study was the first proposing a tumor-related gene network
that controls mammalian pubertal development.
Toro et al. (2018). This study describes how the Trithorax family of epigenetic
activators control the epigenetic landscape at the Kiss1 promoter/enhancer
regions during the pubertal activation on Kiss1 transcription in the ARC.

References
Amar D, Safer H, Shamir R (2013) Dissection of regulatory networks that are altered in disease via
differential co-expression. PLoS Comput Biol 9:E1002955
Anders S, Huber W (2010) Differential expression analysis for sequence count data. Genome Biol
11:R106
Barutcu AR, Fritz AJ, Zaidi SK, Van Wijnen AJ, Lian JB, Stein JL, Nickerson JA, Imbalzano AN,
Stein GS (2016) C-Ing the genome: a compendium of chromosome conformation capture
methods to study higher-order chromatin organization. J Cell Physiol 231:31–35
Beale KE, Kinsey-Jones JS, Gardiner JV, Harrison EK, Thompson EL, Hu MH, Sleeth ML, Sam
AH, Greenwood HC, McGavigan AK, Dhillo WS, Mora JM, Li XF, Franks S, Bloom SR,
O'Byrne KT, Murphy KG (2014) The physiological role of arcuate kisspeptin neurons in the
control of reproductive function in female rats. Endocrinology 155(3):1091–1098
Bindea G, Mlecnik B, Hackl H, Charoentong P, Tosolini M, Kirilovsky A, Fridman WH, Pages F,
Trajanoski Z, Galon J (2009) Cluego: a cytoscape plug-in to decipher functionally grouped gene
ontology and pathway annotation networks. Bioinformatics 25:1091–1093
Bolger AM, Lohse M, Usadel B (2014) Trimmomatic: a flexible trimmer for illumina sequence
data. Bioinformatics 30:2114–2120
Boyar R, Finkelstein J, Roffwarg H, Kapen S, Weitzman E, Hellman L (1972) Synchronization of
augmented luteinizing hormone secretion with sleep during puberty. N Engl J Med 287:582–
8 Epigenetic and Transcriptional Regulation of the Reproductive Hypothalamus 231

586. [This study was the first demonstrating that LH pulsatility increases before sleep in pubertal
patients]
Buenrostro JD, Wu B, Chang HY, Greenleaf WJ (2015) Atac-seq: a method for assaying chromatin
accessibility genome-wide. Curr Protoc Mol Biol 109:21.29.1–21.29.9
Chang TC, Zeitels LR, Hwang HW, Chivukula RR, Wentzel EA, Dews M, Jung J, Gao P, Dang
CV, Beer MA, Thomas-Tikhonenko A, Mendell JT (2009) Lin-28B transactivation is necessary
for Myc-mediated let-7 repression and proliferation. Proc Natl Acad Sci U S A 106(9):3384–
3389
Cline MS, Smoot M, Cerami E, Kuchinsky A, Landys N, Workman C, Christmas R, Avila-
Campilo I, Creech M, Gross B, Hanspers K, Isserlin R, Kelley R, Killcoyne S, Lotia S,
Maere S, Morris J, Ono K, Pavlovic V, Pico AR, Vailaya A, Wang PL, Adler A, Conklin
BR, Hood L, Kuiper M, Sander C, Schmulevich I, Schwikowski B, Warner GJ, Ideker T, Bader
GD (2007) Integration of biological networks and gene expression data using cytoscape. Nat
Protoc 2:2366–2382
Crawford GE, Holt IE, Whittle J, Webb BD, Tai D, Davis S, Margulies EH, Chen Y, Bernat JA,
Ginsburg D, Zhou D, Luo S, Vasicek TJ, Daly MJ, Wolfsberg TG, Collins FS (2006) Genome-
wide mapping of dnase hypersensitive sites using massively parallel signature sequencing
(Mpss). Genome Res 16:123–131
Cukier P, Wright H, Rulfs T, Silveira LF, Teles MG, Mendonca BB, Arnhold IJ, Heger S, Latronico
AC, Ojeda SR, Brito VN (2013) Molecular and gene network analysis of thyroid transcription
factor 1 (TTF1) and enhanced at puberty (EAP1) genes in patients with GnRH-dependent
pubertal disorders. Horm Res Paediatr 80(4):257–266
D’haeseleer P (2005) How does gene expression clustering work? Nat Biotechnol 23:1499–1501
Day FR, Thompson DJ, Helgason H, Chasman DI, Finucane H, Sulem P, Ruth KS, Whalen S,
Sarkar AK, Albrecht E, Altmaier E, Amini M, Barbieri CM, Boutin T, Campbell A,
Demerath E, Giri A, He C, Hottenga JJ, Karlsson R, Kolcic I, Loh PR, Lunetta KL,
Mangino M, Marco B, Mcmahon G, Medland SE, Nolte IM, Noordam R, Nutile T,
Paternoster L, Perjakova N, Porcu E, Rose LM, Schraut KE, Segre AV, Smith AV, Stolk L,
Teumer A, Andrulis IL, Bandinelli S, Beckmann MW, Benitez J, Bergmann S, Bochud M,
Boerwinkle E, Bojesen SE, Bolla MK, Brand JS, Brauch H, Brenner H, Broer L, Bruning T,
Buring JE, Campbell H, Catamo E, Chanock S, Chenevix-Trench G, Corre T, Couch FJ,
Cousminer DL, Cox A, Crisponi L, Czene K, Davey Smith G, De Geus E, De Mutsert R, De
Vivo I, Dennis J, Devilee P, Dos-Santos-Silva I, Dunning AM, Eriksson JG, Fasching PA,
Fernandez-Rhodes L, Ferrucci L, Flesch-Janys D, Franke L, Gabrielson M, Gandin I, Giles GG,
Grallert H, Gudbjartsson DF, Guenel P, Hall P, Hallberg E, Hamann U, Harris TB, Hartman CA,
Heiss G, Hooning MJ, Hopper JL, Hu F, Hunter DJ, Ikram MA, Im HK, Jarvelin MR, Joshi PK,
Karasik D, Kellis M et al (2017) Genomic analyses identify hundreds of variants associated with
age at menarche and support a role for puberty timing in cancer risk. Nat Genet 49:834–841
de Roux N, Genin E, Carel J-C, Matsuda F, Chaussain J-L, Milgrom E (2003) Hypogonadotropic
hypogonadism due to loss of function of the KiSS1-derived peptide receptor GPR54. Proc Natl
Acad Sci U S A 100:10972–10976
Dennis G Jr, Sherman BT, Hosack DA, Yang J, Gao W, Lane HC, Lempicki RA (2003) David:
database for annotation, visualization, and integrated discovery. Genome Biol 4:P3
Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, Batut P, Chaisson M, Gingeras TR
(2013) Star: ultrafast universal RNA-seq aligner. Bioinformatics 29:15–21
Eberwine J, Sul JY, Bartfai T, Kim J (2014) The promise of single-cell sequencing. Nat Methods
11:25–27
Egan B, Yuan CC, Craske ML, Labhart P, Guler GD, Arnott D, Maile TM, Busby J, Henry C, Kelly
TK, Tindell CA, Jhunjhunwala S, Zhao F, Hatton C, Bryant BM, Classon M, Trojer P (2016) An
alternative approach To ChIP-seq normalization enables detection of genome-wide changes in
histone h3 lysine 27 trimethylation upon EZH2 inhibition. PLoS One 11:E0166438
Friedman N, Linial M, Nachman I, Pe'er D (2000) Using Bayesian networks to analyze expression
data. J Comput Biol 7:601–620
Garcia-Martinez J, Aranda A, Perez-Ortin JE (2004) Genomic run-on evaluates transcription rates
for all yeast genes and identifies gene regulatory mechanisms. Mol Cell 15:303–313
232 C. F. Aylwin and A. Lomniczi

Gilfillan GD, Hughes T, Sheng Y, Hjorthaug HS, Straub T, Gervin K, Harris JR, Undlien DE, Lyle
R (2012) Limitations and possibilities of low cell number ChIP-seq. BMC Genomics 13:645
Giresi PG, Kim J, Mcdaniell RM, Iyer VR, Lieb JD (2007) Faire (formaldehyde-assisted isolation
of regulatory elements) isolates active regulatory elements from human chromatin. Genome Res
17:877–885
Glass K, Huttenhower C, Quackenbush J, Yuan GC (2013) Passing messages between biological
networks to refine predicted interactions. PLoS One 8:E64832
Hagan JP, Piskounova E, Gregory RI (2009) Lin 28 recruits the TUTase Zcchc11 to inhibit let-7
maturation in mouse embryonic stem cells. Nat Struct Mol Biol 16(10):1021–1025
Heger S, Mastronardi C, Dissen GA, Lomniczi A, Cabrera R, Roth CL, Jung H, Galimi F,
Sippell W, Ojeda SR (2007) Enhanced at puberty 1 (EAP1) is a new transcriptional regulator
of the female neuroendocrine reproductive axis. J Clin Invest 117(8):2145–2154
Heo I, Joo C, Kim YK, Ha M, Yoon MJ, Cho J, Yeom KH, Han J, Kim VN (2009) TUT4 in concert
with Lin28 suppresses microRNA biogenesis through pre-microRNA uridylation. Cell 138
(4):696–708
Kim TK, Shiekhattar R (2015) Architectural and functional commonalities between enhancers and
promoters. Cell 162(5):948–959
Kim TK, Hemberg M, Gray JM, Costa AM, Bear DM, Wu J, Harmin DA, Laptewicz M, Barbara-
Haley K, Kuersten S, Markenscoff-Papadimitriou E, Kuhl D, Bito H, Worley PF, Kreiman G,
Greenberg ME (2010) Widespread transcription at neuronal activity-regulated enhancers.
Nature 465(7295):182–187
Kwak H, Fuda NJ, Core LJ, Lis JT (2013) Precise maps of RNA polymerase reveal how promoters
direct initiation and pausing. Science 339:950–953
Langmead B, Trapnell C, Pop M, Salzberg SL (2009) Ultrafast and memory-efficient alignment of
short DNA sequences to the human genome. Genome Biol 10:R25
Law CW, Chen Y, Shi W, Smyth GK (2014) Voom: precision weights unlock linear model analysis
tools for RNA-seq read counts. Genome Biol 15:R29
Lehman MN, Coolen LM, Goodman RL (2010) Minireview: kisspeptin/neurokinin B/dynorphin
(KNDy) cells of the arcuate nucleus: a central node in the control of gonadotropin-releasing
hormone secretion. Endocrinology 151(8):3479–3489
Lehrbach NJ, Armisen J, Lightfoot HL, Murfitt KJ, Bugaut A, Balasubramanian S, Miska EA
(2009) LIN-28 and the poly(U) polymerase PUP-2 regulate let-7 microRNA processing in
Caenorhabditis elegans. Nat Struct Mol Biol 16(10):1016–1020
Li H, Durbin R (2009) Fast and accurate short read alignment with Burrows-Wheeler transform.
Bioinformatics 25:1754–1760
Li H, Ruan J, Durbin R (2008) Mapping short DNA sequencing reads and calling variants using
mapping quality scores. Genome Res 18:1851–1858
Liang K, Keles S (2012) Detecting differential binding of transcription factors with ChIP-seq.
Bioinformatics 28:121–122
Liao Y, Smyth GK, Shi W (2014) Featurecounts: an efficient general purpose program for assigning
sequence reads to genomic features. Bioinformatics 30:923–930
Lomniczi A, Garcia-Rudaz C, Ramakrishnan R, Wilmont B, Khouangsathiene S, Ferguson B,
Dissen GA, Ojeda SR (2012) A single nucleotide polymorphism in the EAP1 gene is associated
with amenorrhea/oligomenorrhea in nonhuman primates. Endocrinology 153(1):339–349
Lomniczi A, Loche A, Castellano JM, Ronnekleiv OK, Bosh M, Kaidar G, Knoll JG, Wright H,
Pfeifer GP, Ojeda SR (2013a) Epigenetic control of female puberty. Nat Neurosci 16:281–289.
[This study describes how the Polycomb family of epigenetic repressors control the epigenetic
landscape at the Kiss1 promoter region, keeping ARC Kiss1 transcription low during the
infantile period]
Lomniczi A, Wright H, Castellano JM, Sonmez K, Ojeda SR (2013b) A system biology approach to
identify regulatory pathways underlying the neuroendocrine control of female puberty in rats
and nonhuman primates. Horm Behav 64(2):175–186
8 Epigenetic and Transcriptional Regulation of the Reproductive Hypothalamus 233

Lomniczi A, Wright H, Castellano JM, Matagne V, Toro CA, Ramaswamy S, Plant TM, Ojeda SR
(2015) Epigenetic regulation of puberty via Zinc finger protein-mediated transcriptional repres-
sion. Nat Commun 6:10195. [This study describes how the GATAD1/LSD1 controls the
epigenetic landscape at the Kiss1 promoter region in rhesus macaques, keeping ARC Kiss1
transcription low during the infantile period]
Maere S, Heymans K, Kuiper M (2005) Bingo: a cytoscape plugin to assess overrepresentation of
gene ontology categories in biological networks. Bioinformatics 21:3448–3449
Maisonpierre PC, Le Beau MM, Espinosa R III, Ip NY, Belluscio L, de la Monte SM, Squinto S,
Furth ME, Yancopoulos GD (1991) Human and rat brain-derived neurotrophic factor and
neurotrophin-3: gene structures, distributions, and chromosomal localizations. Genomics
10:558–568
Margolin AA, Nemenman I, Basso K, Wiggins C, Stolovitzky G, Dalla Favera R, Califano A
(2006) Aracne: an algorithm for the reconstruction of gene regulatory networks in a mammalian
cellular context. BMC Bioinformatics 7(Suppl 1):S7
Mastronardi C, Smiley GG, Raber J, Kusakabe T, Kawaguchi A, Matagne V, Dietzel A, Heger S,
Mungenast AE, Cabrera R, Kimura S, Ojeda SR (2006) Deletion of the Ttf1 gene in
differentiated neurons disrupts female reproduction without impairing basal ganglia function.
J Neurosci 26:13167–13179
Meijs-Roelofs HMA, Kramer P, Cappellen WAV, Woutersen PJA (1985) Changes in serum
concentration and ovarian content of 5a-androstane-3a-17b-diol in the female rat approaching
first ovulation. Biol Reprod 32:301–308
Mortazavi A, Williams BA, McCue K, Schaeffer L, Wold B (2008) Mapping and quantifying
mammalian transcriptomes by RNA-Seq. Nat Methods 5:621–628
Mostafavi S, Ray D, Warde-Farley D, Grouios C, Morris Q (2008) Genemania: a real-time
multiple association network integration algorithm for predicting gene function. Genome Biol
9(Suppl 1):S4
Mueller JK, Dietzel A, Lomniczi A, Loche A, Tefs K, Kiess W, Danne T, Ojeda SR, Heger S (2011)
Transcriptional regulation of the human KiSS1 gene. Mol Cell Endocrinol 342(1–2):8–19
Mueller JK, Koch I, Lomniczi A, Loche A, Rulfs T, Castellano JM, Kiess W, Ojeda S, Heger S
(2012) Transcription of the human EAP1 gene is regulated by upstream components of a
puberty-controlling tumor suppressor gene network. Mol Cell Endocrinol 351(2):184–198
Navarro VM, Gottsch ML, Wu M, Garcia-Galiano D, Hobbs SJ, Bosch MA, Pinilla L, Clifton DK,
Dearth A, Ronnekleiv OK, Braun RE, Palmiter RD, Tena-Sempere M, Alreja M, Steiner RA
(2011) Regulation of NKB pathways and their roles in the control of Kiss1 neurons in the
arcuate nucleus of the male mouse. Endocrinology 152(11):4265–4275
Ojeda SR (1991) The mystery of mammalian puberty: how much more do we know? Perspect Biol
Med 34:365–383
Ojeda SR, Hill J, Hill DF, Costa ME, Tapia V, Cornea A, Ma YJ (1999) The Oct-2 POU-domain
gene in the neuroendocrine brain: a transcriptional regulator of mammalian puberty. Endocri-
nology 140:3774–3789
Ojeda SR, Terasawa E (2002) Neuroendocrine regulation of puberty. In: Pfaff D et al (eds)
Hormones, brain and behavior, vol 4. Elsevier, New York, pp 589–659
Ojeda SR, Skinner MK (2006) Puberty in the rat. In: Neill JD (ed) The physiology of reproduction,
3rd edn. Academic Press/Elsevier, San Diego, pp 2061–2126
Ong KK, Elks CE, Li S, Zhao JH, Luan J, Andersen LB, Bingham SA, Brage S, Smith GD,
Ekelund U, Gillson CJ, Glaser B, Golding J, Hardy R, Khaw KT, Kuh D, Luben R, Marcus M,
McGeehin MA, Ness AR, Northstone K, Ring SM, Rubin C, Sims MA, Song K, Strachan DP,
Vollenweider P, Waeber G, Waterworth DM, Wong A, Deloukas P, Barroso I, Mooser V, Loos
RJ, Wareham NJ (2009) Genetic variation in LIN28B is associated with the timing of puberty.
Nat Genet 41(6):729–733
Perry JR, Stolk L, Franceschini N, Lunetta KL, Zhai G, McArdle PF, Smith AV, Aspelund T,
Bandinelli S, Boerwinkle E, Cherkas L, Eiriksdottir G, Estrada K, Ferrucci L, Folsom AR,
Garcia M, Gudnason V, Hofman A, Karasik D, Kiel DP, Launer LJ, van MJ, Nalls MA,
234 C. F. Aylwin and A. Lomniczi

Rivadeneira F, Shuldiner AR, Singleton A, Soranzo N, Tanaka T, Visser JA, Weedon MN,
Wilson SG, Zhuang V, Streeten EA, Harris TB, Murray A, Spector TD, Demerath EW,
Uitterlinden AG, Murabito JM (2009) Meta-analysis of genome-wide association data identifies
two loci influencing age at menarche. Nat Genet 41(6):648–650
Pierson E, GTEx Consortium, Koller D, Battle A, Mostafavi S, Ardlie KG, Getz G, Wright FA,
Kellis M, Volpi S, Dermitzakis ET (2015) Sharing and specificity of co-expression networks
across 35 human tissues. PLoS Comput Biol 11:E1004220
Pinilla L, Aguilar E, Dieguez C, Millar RP, Tena-Sempere M (2012) Kisspeptins and reproduction:
physiological roles and regulatory mechanisms. Physiol Rev 92(3):1235–1316
Robinson MD, Smyth GK (2007) Moderated statistical tests for assessing differences in tag
abundance. Bioinformatics 23:2881–2887
Robinson MD, Mccarthy DJ, Smyth GK (2010) Edger: a bioconductor package for differential
expression analysis of digital gene expression data. Bioinformatics 26:139–140
Roth CL, Mastronardi C, Lomniczi A, Wright H, Cabrera R, Mungenast AE, Heger S, Jung H,
Dubay C, Ojeda SR (2007) Expression of a tumor-related gene network increases in the
mammalian hypothalamus at the time of female puberty. Endocrinology 148(11):5147–5161.
[This study was the first proposing a tumor-related gene network that controls mammalian
pubertal development]
Sandau US, Mungenast AE, Alderman Z, Sardi SP, Fogel AI, Taylor B, Parent AS, Biederer T,
Corfas G, Ojeda SR (2011a) SynCAM1, a synaptic adhesion molecule, is expressed in
astrocytes and contributes to erbB4 receptor-mediated control of female sexual development.
Endocrinology 152(6):2364–2376
Sandau US, Mungenast AE, McCarthy J, Biederer T, Corfas G, Ojeda SR (2011b) The synaptic cell
adhesion molecule, SynCAM1, mediates astrocyte-to-astrocyte and astrocyte-to-GnRH neuron
adhesiveness in the mouse hypothalamus. Endocrinology 152(6):2353–2363
Sangiao-Alvarellos S, Manfredi-Lozano M, Ruiz-Pino F, Navarro V, Sanchez-Garrido MA,
Leon S, Dieguez C, Cordido F, Matagne V, Dissen GA, Ojeda SR, Pinilla L, Tena-Sempere
M (2013) Changes in hypothalamic expression of the Lin28/let-7 system and related
microRNAs during postnatal maturation and after experimental manipulations of puberty.
Endocrinology 154:942–955
Seminara SB, Messager S, Chatzidaki EE, Thresher RR, Acierno JS Jr, Shagoury JK, Bo-Abbas Y,
Kuohung W, Schwinof KM, Hendrick AG, Zahn D, Dixon J, Kaiser UB, Slaugenhaupt SA,
Gusella JF, O'Rahilly S, Carlton MB, Crowley WF Jr, Aparicio SA, Colledge WH (2003) The
GPR54 gene as a regulator of puberty. N Engl J Med 349:1614–1627
Shao Z, Zhang Y, Yuan GC, Orkin SH, Waxman DJ (2012) Manorm: a robust model for
quantitative comparison of ChIP-seq data sets. Genome Biol 13:R16
Shi Y, Lan F, Matson C, Mulligan P, Whetstine JR, Cole PA, Casero RA, Shi Y (2004) Histone
demethylation mediated by the nuclear amine oxidase homolog LSD1. Cell 119(7):941–953
Shilatifard A (2012) The COMPASS family of histone H3K4 methylases: mechanisms of regula-
tion in development and disease pathogenesis. Annu Rev Biochem 81:65–95
Simon JA, Kingston RE (2009) Mechanisms of polycomb gene silencing: knowns and unknowns.
Nat Rev Mol Cell Biol 10(10):697–708
Steeg PS, Ouatas T, Halverson D, Palmieri D, Salerno M (2003) Metastasis suppressor genes: basic
biology and potential clinical use. Clin Breast Cancer 4:51–62
Steuer R, Kurths J, Daub CO, Weise J, Selbig J (2002) The mutual information: detecting and
evaluating dependencies between variables. Bioinformatics 18(Suppl 2):S231–S240
Sulem P, Gudbjartsson DF, Rafnar T, Holm H, Olafsdottir EJ, Olafsdottir GH, Jonsson T,
Alexandersen P, Feenstra B, Boyd HA, Aben KK, Verbeek AL, Roeleveld N, Jonasdottir A,
Styrkarsdottir U, Steinthorsdottir V, Karason A, Stacey SN, Gudmundsson J, Jakobsdottir M,
Thorleifsson G, Hardarson G, Gulcher J, Kong A, Kiemeney LA, Melbye M, Christiansen C,
Tryggvadottir L, Thorsteinsdottir U, Stefansson K (2009) Genome-wide association study
identifies sequence variants on 6q21 associated with age at menarche. Nat Genet 41(6):734–738
8 Epigenetic and Transcriptional Regulation of the Reproductive Hypothalamus 235

Toro CA, Wright H, Aylwin CF, Ojeda SR, Lomniczi A (2018) Trithorax dependent changes in
chromatin landscape at enhancer and promoter regions drive female puberty. Nat Commun 9
(1):57. [This study describes how the Trithorax family of epigenetic activators control the
epigenetic landscape at the Kiss1 promoter/enhancer regions during the pubertal actiation on
Kiss1 transcription in the ARC]
Urbanski HF, Ojeda SR (1985) The juvenile-peripubertal transition period in the female rat:
establishment of a diurnal pattern of pulsatile luteinizing hormone secretion. Endocrinology
117:644–649
Urbanski HF, Ojeda SR (1987) Gonadal-independent activation of enhanced afternoon luteinizing
hormone release during pubertal development in the female rat. Endocrinology 121:907–913
Wakabayashi Y, Nakada T, Murata K, Ohkura S, Mogi K, Navarro VM, Clifton DK, Mori Y,
Tsukamura H, Maeda K, Steiner RA, Okamura H (2010) Neurokinin B and dynorphin A in
kisspeptin neurons of the arcuate nucleus participate in generation of periodic oscillation of
neural activity driving pulsatile gonadotropin-releasing hormone secretion in the goat. J
Neurosci 30(8):3124–3132
Wang Z, Gerstein M, Snyder M (2009) RNA-seq: a revolutionary tool for transcriptomics. Nat Rev
Genet 10:57–63
Xue Z, Huang K, Cai C, Cai L, Jiang CY, Feng Y, Liu Z, Zeng Q, Cheng L, Sun YE, Liu JY,
Horvath S, Fan G (2013) Genetic programs in human and mouse early embryos revealed by
single-cell RNA sequencing. Nature 500:593–597
Yeung KT, Das S, Zhang J, Lomniczi A, Ojeda S, Xu CF, Neubert TA, Samuels HH (2011) A novel
transcription complex that selectively modulates apoptosis of breast cancer cells through
regulation of FASTKD2. Mol Cell Biol 31(11):2287–2298
Zhang Y, Liu T, Meyer CA, Eeckhoute J, Johnson DS, Bernstein BE, Nusbaum C, Myers RM,
Brown M, Li W, Liu XS (2008) Model-based analysis of ChIP-seq (Macs). Genome Biol 9:
R137
Epigenetic Regulation of the GnRH
and Kiss1 Genes: Developmental 9
Perspectives

Joseph R. Kurian and Ei Terasawa

Abstract
Gonadotropin-releasing hormone (GnRH) and kisspeptin neurons are indispens-
able for reproductive function. This chapter focuses on the epigenetic regulation
of the GnRH and kisspeptin (Kiss1) genes in the context of neuronal develop-
ment, puberty onset, and maintenance of adult reproductive function. Proper
function of GnRH and kisspeptin neurons in the hypothalamus requires coordi-
nated embryonic and postnatal maturation. Recent studies indicate that diverse
epigenetic phenomena including the formation of chromatin loops, activation of
bivalent domains, maintenance of “stable” histone modifications, and DNA
methylation, active demethylation, and hydroxymethylation are all involved in
epigenetic regulation of these neurons. This chapter focuses on how these epige-
netic regulation components interact with each other and which enzymes or
binding factors are involved in specific stages of neuronal development or
aging. By emphasizing similar epigenetic mechanisms operating within these
two peptidergic neuronal populations, we highlight likely targets for future
investigations of environmental influences over puberty timing and progression.

Keywords
Gonadotropin-releasing hormone (GnRH) · Kisspeptin · Puberty · Bivalent
chromatin domain · DNA hydroxymethylation · Tet enzymes

J. R. Kurian
Departments of Obstetrics/Gynecology and Internal Medicine, Southern Illinois University School
of Medicine, Springfield, IL, USA
St. Johns Hospital, Carol Jo Vecchie Women and Children’s Center, Springfield, IL, USA
E. Terasawa (*)
Department of Pediatrics, University of Wisconsin-Madison, Madison, WI, USA
Wisconsin National Primate Research Center, University of Wisconsin-Madison, Madison, WI,
USA
e-mail: [email protected]

# Springer Nature Switzerland AG 2020 237

S. Wray, S. Blackshaw (eds.), Developmental Neuroendocrinology, Masterclass in
Neuroendocrinology 9, https://doi.org/10.1007/978-3-030-40002-6_9
238 J. R. Kurian and E. Terasawa

9.1 GnRH Neurons

Gonadotropin-releasing hormone (GnRH) neurons are critical relays in the axis

controlling reproductive function. These neurons reside in the preoptic area and basal
hypothalamus and project to the median eminence, where they release the decapeptide
GnRH into the portal circulation. GnRH subsequently stimulates anterior pituitary
release of the gonadotropins luteinizing hormone (LH) and follicle-stimulating hor-
mone (FSH), which promote gametogenesis and gonadal steroidogenesis in both sexes
as well as ovulation in females. In order to properly stimulate gonadotropin release,
GnRH must be supplied to the portal circulation in sufficient, episodic pulses. In fact,
the onset of puberty and maintenance of reproductive function depends on the elevated
and finely coordinated pulsatility of GnRH release. While upstream mechanisms fine-
tune GnRH neuronal activity, GnRH neurons also have an intrinsic capacity to release
GnRH peptide in a pulsatile manner (Terasawa 2001).
GnRH-expressing neurons reside in a complex milieu of other neurons and glia;
this presents a challenge for studying the cellular and molecular mechanisms of
neuronal differentiation and function. Fortunately, their unique ontogeny provides
an opportunity to isolate GnRH-expressing neurons for in vitro studies. In the rhesus
monkey, these neurons differentiate from progenitor cells in the nasal placode
between embryonic days (E) 32–34 (Ronnekleiv and Resko 1990; Quanbeck et al.
1997). GnRH neurons subsequently begin migrating into the brain at about E36 and
settle in the hypothalamus by E55 (Terasawa et al. 2001). Isolating the nasal placode
after E34 but prior to migration provides a neuronal population enriched in GnRH
neurons. We found that placode tissue isolated on E36 developed typical patterns of
mature activity (e.g., GnRH peptide release; Terasawa et al. 1993, 1999; Kurian et al.
2010a) after about 2–2.5 weeks in vitro. Wray and colleagues, who developed a
similar murine in vitro culture model (Fueshko and Wray 1994), have also reported a
period of gradual maturation after isolation from the nasal placode (Constantin et al.
2009). In addition, they report that development of GnRH peptide release patterns is
paralleled by increasing GnRH gene expression and peptide biosynthesis (Maurer
and Wray 1997; Moore and Wray 2000). A question arises: what mechanism triggers
increasing levels of gene expression during maturation of GnRH-expressing
neurons?

9.2 DNA Methylation in GnRH Neurons

The genetic control of GnRH gene expression is well characterized and depends on
several cis-regulatory sequences in the 50 region of the gene (Box 9.1). The rat gene
contains a neuron-specific enhancer region between 1863 and 1571 (Kepa et al.
1992, 1996b; Clark and Mellon 1995; Whyte et al. 1995). This spans a major region
of homology between rat (1786 to 1559) and human (2766 to 2539) genes.
Interestingly, however, this portion of the human gene does not appear to activate
gene expression. In fact, based on serial truncations of the 50 human GnRH gene in
luciferase assay constructs, it appears that this region represses gene expression
9 Epigenetic Regulation of the GnRH and Kiss1 Genes: Developmental Perspectives 239

(Kepa et al. 1996a). Importantly, this area has sequence similarity to a 50 portion of
the rhesus monkey GnRH gene. We noticed that this distal 50 region of the rhesus
monkey gene contains a 243-bp segment (2126 to 1863) that has 60% GC
content and a CpG (cytosine–guanine) dinucleotide observed to expected ratio of
0.65 (14 CpG sites). These characteristics define the region as a CpG island (CGI;
Gardiner-Garden and Frommer 1987). CGIs, when associated with gene promoters,
are related to the epigenetic regulation (DNA methylation) of gene expression
(Deaton and Bird 2011).

Box 9.1 Genetic and Epigenetic Regulation of the GnRH Gene During
Development
Understanding transcriptional regulation of the GnRH gene is important, as
deficiency of this process leads to abnormal timing of puberty and infertility.
The mammalian form of GnRH (GnRH1) encoded by the GnRH gene,
consists of a ten amino acid-peptide (pyro-Glu-His-Trp-Ser-Tyr-Gly-Leu-Arg-
Pro-Gly-NH2), cleaved from the 92-amino acid GnRH precursor protein
(prepro-GnRH peptide). The GnRH gene contains four exons and three
introns. The 50 GnRH gene-regulatory region comprises a proximal promoter
and three enhancer regions. Transcription factors bind to conserved enhancer
regions and the proximal promoter region.
Genetic Regulation: Since the 1990s several groups reported using in vitro
GnRH cell lines and in vivo transgenic mouse approaches to identify various
transcription factors that bind the 50 GnRH gene-regulatory region, and activate
or repress GnRH gene transcription. Among the activators are the homeodomain
transcription factors OCT1, OTX2, Six3/6, MSX1 and VAX1, neuron-specific
elements, GATA4 (Radovick et al. 1994; Clark and Mellon 1995; Lawson et al.
1998; Kelley et al. 2000; Kim et al. 2007; Larder and Mellon. 2009; Gan et al.
2012; Xie et al. 2015; Larder et al. 2013; Hoffmann et al. 2018), an estrogen
response element (human/primates only, Radovick et al. 1991b, 1994), and a
kisspeptin response element (Novaira et al. 2012, 2016). Repressors are C/EBPβ
(Belsham and Mellon 2000; Messina et al. 2016), MSX1/2 (Givens et al. 2005,
Rave-Harel et al. 2005), NKX2.1/TTF-1 (Provenzano et al. 2010), and ZEB1
(Messina et al. 2016). See further discussion in Hoffmann and Mellon (2018).
Epigenetic Regulation: More recently, GnRH gene expression was also
found to be altered by environmental factors through mechanisms such as
DNA methylation and histone modifications without changes in nucleotide
sequences. Recent findings are summarized in this chapter.

DNA methylation is the covalent addition of a methyl (–CH3) group to

nucleotides. Mammalian DNA-methyltransferase (DNMT) enzymes catalyze this
reaction. Mammalian DNA methylation occurs primarily at the 50 carbon of
cytosines in CpG dinucleotides and to a lesser extent at CpH (H¼A, T, C)
dinucleotides (Ramsahoye et al. 2000; Guo et al. 2014). There are three primary
240 J. R. Kurian and E. Terasawa

DNMT enzymes (DNMT1, 3a, and 3b), each critical to development as

demonstrated by embryonic or early postnatal lethality in monogenic null mouse
models (Li et al. 1992; Okano et al. 1999). DNMT1 is responsible for faithful
maintenance of DNA methylation after replication through cell division (Bestor
and Ingram 1983; Bestor et al. 1988; Hermann et al. 2004). DNMTs 3a and 3b are
de novo methyltransferases responsible for newly acquired methylation, such as
during the initial establishment of methylation patterns during early embryonic
development (Okano et al. 1998, 1999) and throughout neuronal maturation and
aging (Guo et al. 2014). Once established, DNA methylation can have several
impacts on gene transcription. Methylated DNA can directly alter transcription
factor binding to cis-regulatory sequences or attract methyl-binding proteins,
which in turn block transcription factor binding. In addition, methyl-binding proteins
(e.g., MeCP2, Mbd2, and Kaiso) interact with histone-modifying factors to alter
chromatin structure.
Until our recent studies (Kurian et al. 2010a), there were no reports on the
epigenetic aspects of GnRH neuron maturation or function. The CGI in the 50 region
of the rhesus monkey GnRH gene suggested to us that DNA methylation had some
role in neuronal function (Fig. 9.1). We hypothesized that increasing peptide release
during in vitro maturation of GnRH neurons resulted from increased GnRH tran-
scription, and changing DNA methylation patterns across the rhesus monkey GnRH
gene, particularly within the 50 CGI. As suspected, we found that GnRH mRNA
levels were low at day 0 but rose dramatically by day 20 of in vitro culture. This
increase was paralleled by a dramatic decrease in CpG methylation status in the 50
CGI. We also found that a comparable phenomenon occurs across puberty. Similar
to the report by Plant and colleagues which showed that GnRH mRNA levels
increase between juvenile and pubertal stages in the medial basal hypothalamus
(MBH) of orchidectomized male rhesus monkeys (El Majdoubi et al. 2000), we
observed that GnRH mRNA levels increase across puberty in the male monkey
MBH. Furthermore, methylation status of the 50 CGI of the GnRH gene was lower in
adults compared to prepubertal male rhesus monkeys (Kurian et al. 2011; Kurian and
Terasawa 2013). This suggests that the developmental rise in GnRH gene expression
is, at least partly, the result of DNA demethylation across a CGI in the GnRH gene.
When evaluating these findings, it is important to consider the distinct pattern of
GnRH neuronal activity across development. In primates, GnRH release is elevated,
as indicated by peripheral luteinizing hormone levels, during a brief perinatal period
(this is commonly known as “mini-puberty”), but then decreases during juvenile
development before gradually increasing again through puberty. Our findings sug-
gest that each period of elevated GnRH release is linked to sequential waves of
demethylation of the GnRH gene CGI. Specifically, demethylation observed during
in vitro maturation mirrors that seen during embryonic development, and leads to
elevated GnRH expression during the postnatal period. In contrast, measurements in
MBH tissue reflect developmental changes across pubertal maturation. The decrease
in DNA methylation seen during puberty is indicative of two potential scenarios.
One possibility is that the process of DNA demethylation initiated in embryonic
development might stall during the juvenile period, and subsequently resume at
9 Epigenetic Regulation of the GnRH and Kiss1 Genes: Developmental Perspectives 241

Fig. 9.1 Embryonic development of primate GnRH neurons is marked by a dramatic shift in Gnrh
gene expression and active demethylation of a 50 CpG island. (a) Total RNA was extracted from
nasal placode cultures containing GnRH neurons at 0, 14, and 20 days (div). Gnrh mRNA levels,
measured by quantitative PCR, started to increase after 14 div, reaching the highest level at 20 div.
(P < 0.05 versus 0 div; #P ¼ 0.05 versus 14 div). (b) A schematic representation of the rhesus
monkey GnRH gene depicting the location of the 50 CpG island (CGI) and nucleotide sequence of
this region. This region is the only classifiable CGI within 2500 bases upstream of the Gnrh
transcription start site. CpG sites are indicated with superscript numbers corresponding to the CpG
sites in the bottom graph. (c) Nasal placode regions, containing GnRH neurons, were dissected from
two rhesus monkey embryos at E36 and E37, plated and then harvested on 0, 14, or 20 div. DNA
extracted from pooled samples (four cultures) at each time point was bisulfite sequenced. Percent
changes in methylation at each CpG site on 0 div (black bars), 14 div (gray bars), and 20 div (white
bars) are shown. CpG methylation status was significantly higher at 0 div compared with 20 div at
sites 4, 5, 6, 8, 9, 11, 12, and 14 (P  0.01). CpG methylation status was significantly higher at 0 div
compared with 14 div at site 5 (#P < 0.05) and at 0 div compared with 14 div but not 20 div at site
10 (##P < 0.05). Modified from Kurian et al. Endocrinology 151: 5359–5368 (2010a)
242 J. R. Kurian and E. Terasawa

puberty onset. Alternatively, the CGI may become re-methylated during juvenile
development, and again demethylated at puberty onset. The latter scenario would
suggest that lower CpG methylation status must be maintained to enable elevated
GnRH gene expression. Consequently, a mechanism responsible for DNA demeth-
ylation and perhaps another mechanism responsible for maintaining hypomethylated
DNA might be necessary for the transition to puberty and maintenance of reproduc-
tive function. A comparison of CpG methylation status between perinatal and early
pubertal MBH tissue will be necessary to differentiate between these two scenarios.
Nonetheless, our current findings suggest that DNA demethylation is an important
aspect of GnRH neuronal development and function.
DNA demethylation is achieved via two general mechanisms, passive or active.
Passive demethylation occurs through cell divisions and interruption of maintenance
methyltransferase activity. Current pharmacological approaches for DNA demethyl-
ation target this mechanism. For example, 5-azacytidine (5-aza) is metabolized and
subsequently incorporated into DNA, where it then acts as a substrate that covalently
traps DNMTs after methyl transfer to the 5-aza nitrogen (Schermelleh et al. 2005;
Svedruzic 2008). Given the postmitotic/nondividing state of GnRH neurons in our
in vitro cultures and across the pubertal transition measured in MBH samples, the
process of DNA demethylation we discovered must be active. The mechanisms
responsible for active DNA demethylation are not well characterized, and until
recently skepticism has remained over the existence of this process in mammalian
systems. Several mechanisms are now proposed, and intense efforts to validate or
better characterize these mechanisms continue. For a thorough background, we
suggest a recent review (Wu and Zhang 2010) that outlines several promising
pathways discovered during the past decade. Presently, a well-accepted mechanism
is the sequential enzymatic process beginning with the oxidation of
5-methylcytosine (5mC) to 5-hydroxymethylcytosine (5hmC), which is carried out
by anyone of three ten-eleven-translocation enzymes (Tet 1–3). 5hmC is a stable
epigenetic modification, though under certain circumstances it is recognized and
excised by thymine DNA glycosylases. Base excision repair subsequently completes
the transition. Importantly, brain tissue, and particularly the hypothalamus, shows
the highest reported levels of 5-hydroxymethylcytosine (5hmC) (Branco et al. 2012).
Emerging physiological evidence also supports a role for Tet enzymes in neuro-
endocrine development, and particularly the control of reproductive function. For
example, Tet1 knockout mice exhibits deficits in fecundity. This effect is primarily
a female-specific abnormality, as typical litter sizes result from crossing wild-type
females with Tet1 knockout males, whereas wild-type males mated to Tet1 knockout
females produce significantly fewer offspring per litter (Dawlaty et al. 2011). A
subsequent report suggests this defect in fecundity might be the consequence of
abnormal progression of female germ cell development through the second meiotic
division prior to ovulation (Yamaguchi et al. 2012). Interestingly, stimulation of the
second meiotic division and ovulation requires a GnRH-driven LH surge
(Mehlmann 2005). Consequently, given the apparent active DNA demethylation
associated with GnRH neuronal development, the effect of Tet depletion on abnor-
mal reproductive function might be the consequence of altered Tet-mediated
9 Epigenetic Regulation of the GnRH and Kiss1 Genes: Developmental Perspectives 243

differentiation of the neuroendocrine hypothalamus. While hypothalamic expression

patterns of Tet enzymes are not yet reported, we recently found that basal hypotha-
lamic TET2 expression increases across puberty in female rhesus monkeys (Kurian
and Terasawa; unpublished observation). That the deficit in fecundity was even more
pronounced in Tet1/Tet2 double knockout mice (Dawlaty et al. 2013) gives further
credence to the suggestion that Tet enzymes act in the hypothalamus to promote
differentiation of neurons controlling stage-specific reproductive functions.

9.3 Histone Modifications in GnRH Neurons

Histones are an integral component of nucleosomes, the primary units for genome
organization or compaction. Consequently, through their interaction with DNA,
these proteins have a critical role in determining gene expression patterns. There
are four primary classes of histones, 1 through 4. Histones 2 through 4 are
components of the core octamer, which DNA circumnavigates in about 146 base
pairs to form a single nucleosome. Histone 1 is a scaffold protein, which tightly
packages nucleosomes when present. Several variants of this histone are distinguish-
able by sensitivity to hormones (Banks et al. 2001). Histone 2 also has several
variants including A, AX, and B. Histone 2AX is particularly intriguing in the
context of neuronal maturation (Lee et al. 2010) and function based on its association
with activity-dependent DNA double-strand breaks in neurons (Suberbielle et al.
2013). Histones 3 and 4 complete the nucleosome octamer with histone 3 the most
heavily investigated in the context of neuroendocrine function. Histone 3 has two
variants, H3.3A and H3.3B. These variants are integral to DNA replication-
independent histone switching, which would be presumed an important mechanism
in the regulation of post-mitotic cell activity. To our knowledge, nothing is reported
regarding the relationship between histone switching and hypothalamic neuronal
maturation or function.
Histone proteins package DNA largely due to their predominant positive charge
attracting negatively charged DNA. Posttranslational modifications (PTMs) alter the
strength of that attraction, and recruit or repel transcriptional machinery and histone-
modifying enzymes. Consequently, these PTMs alter gene accessibility and rates of
transcription. Several known modifications include acetylation, phosphorylation,
methylation, ubiquitination, sumoylation, and GlcNAcylation. To date,
measurements of histone PTMs in neuroendocrine systems have focused on acety-
lation and methylation. Acetylation leaves a more negative charge on histones and
consequently promotes transcription. Methylation is neutral, and depending on the
location and degree (mono, di, or tri-methylation), it can either promote or inhibit
transcription.
Mellon and colleagues were the first to report a pattern of permissive histone
modifications enabling elevated or mature Gnrh gene transcription (Iyer et al. 2011).
Their studies capitalized on the distinct stages of development between two GnRH
neuronal cell lines. GN11 cells, originally isolated from a tumor in the mouse nasal
placode (Radovick et al. 1991a), express Gnrh at very low levels, whereas GT1 cells,
244 J. R. Kurian and E. Terasawa

which were isolated from an analogous tumor in the mouse hypothalamus (Mellon
et al. 1990), are characterized by mature activity patterns including elevated Gnrh
expression. In essence, comparisons between these two cell lines are similar to our
evaluations of embryonic nasal placode derived neurons from days 0 and 20 in vitro
described above. They found that the GnRH gene promoter and enhancer regions in
immature GN cells were more heavily associated with a repressive histone modifi-
cation: histone 3 (H3) lysine 9 (K9) di-methylation (me2). On the other hand, the
same genomic regions in GT1 cells were associated with the permissive H3K9
acetylation and H3K4me3 PTMs. For comparison, they also evaluated these histone
PTM patterns in a nonneuronal (NIH3T3) cell line. The repressive PTMs were high
in NIH3T3 cells, intermediate in GN cells and low in GT1 cells. The presence of
permissive PTMs was low and similar between NIH3T3 and GN cells but signifi-
cantly higher in mature GT1 cells.
The intermediate repressive chromatin state in GN cells is indicative of a bivalent
promoter, where both repressive and permissive histone PTMs maintain genes in a
repressed albeit primed position. These chromatin domains were first described in
embryonic stem (ES) cells and associated with developmentally regulated transcrip-
tion factors (Azuara et al. 2006; Bernstein et al. 2006; Pan et al. 2007). Several recent
reports, taken together, indicate that establishment of bivalent domains through
polycomb repressive complex 2 (PRC2) and COMPASS families is critical for
neural lineage differentiation from ES cells (Yu et al. 1995; Yagi et al. 1998; Glaser
et al. 2006; Pasini et al. 2007; Shen et al. 2008). The PRC2 component, Ezh2, is
responsible for H3K27 methylation, which is subsequently bound by the PRC1
complex to maintain tri-methylation at H3K27. COMPASS complexes methylate
H3K4. Together, these complexes are proposed to establish a bivalent promoter with
the heavily repressive H3K27me3 mark in close proximity to H3K4me3. Interest-
ingly, this bivalent domain also appears related to the juvenile repression and
subsequent activation of Kiss1 (kisspeptin) gene expression during the pubertal
transition in female rats (Lomniczi et al. 2013). As a major stimulant of GnRH
release, this epigenetic regulation of Kiss1 expression (discussed in detail below)
may be instrumental in pubertal development. Together, this also suggests that
bivalent domain activation may be a consistent mechanism across reproductive
neuroendocrine systems.

9.4 Tet Enzyme Activity in GnRH Neurons and Reproductive

Function

A recent discovery points to Tet enzymes as critical mediators of activation at

bivalent promoters during neuronal differentiation. Specifically, while Ezh2 (the
H3K27 methyltransferase component of PRC2) is critical for progression of neuro-
nal precursor cells toward a neuronal fate, Tet2 or Tet3 appear to complete the
process of differentiation (Hahn et al. 2013). In fact, these studies also found that
accumulation of intragenic 5-hydroxymethylcytosine is associated with loss of
H3K27me3 near regions with the most significant gene activation during neuronal
9 Epigenetic Regulation of the GnRH and Kiss1 Genes: Developmental Perspectives 245

differentiation. In addition, Tet2 was recently reported to promote H3K4me3

through association with the Set1/COMPASS complex (Deplus et al. 2013). Based
on this background, our recent studies have focused on Tet enzyme activity in the
context of GnRH neuron development and function. Initially we utilized the devel-
opmentally distinct GnRH neuronal cell lines. We discovered that Tet2 expression is
low in immature GN11 cells, but substantially higher in mature GT1–7 cells.
Subsequently we found that overexpression of Tet2 in immature cells led to signifi-
cant elevation of Gnrh gene expression. This increase was potentiated when cultured
cells were supplemented with glutamine, which enhances Tet2 activity by increasing
intracellular concentrations of the cofactor alpha-ketoglutarate (Yang et al. 2014;
Carey et al. 2015). Using chromatin immunoprecipitation, we evaluated Tet2 bind-
ing as well as H3K4me3 and H3K27me3 abundance across the mouse GnRH gene 50
region (Kurian 2016; Fig. 9.2). Chromatin conditions were evaluated in four
conditions: (1) GN11, (2) GT1–7, (3) GN11 cells overexpressing Tet2, and
(4) GT1–7 cells with Tet2 eliminated by CRISPR/Cas9-mediated gene disruption.
Tet2 binding was low in GN11 cells across the entire 50 region, but significantly
elevated in GT1–7 cells at the promoter and neuron-specific enhancer region. The
activating H3K4me3 modification largely mirrored Tet2 binding patterns, with
significantly higher abundance in GT1–7 compared to GN11 cells. Overexpression
of Tet2 in GN11 cells significantly increased Tet2 binding to the Gnrh promoter and,
to a lesser extent, the neuron-specific enhancer; H3K4me3 levels were also signifi-
cantly higher at the promoter compared to non-transfected GN11 cells. Disruption of
Tet2 in GT1–7 cells decreased Tet2 binding across the entire region; remarkably,
H3K4me3 levels significantly dropped at the neuron-specific enhancer after Tet2
disruption in GT1–7 cells. H3K27 levels were largely consistent between all
conditions at each site evaluated. Altogether, these results show that Tet2 influences
bivalent promoter-specific histone modification patterns near the Gnrh gene and
contributes to the elevation and maintenance of elevated Gnrh transcription
(Fig. 9.3). Our subsequent studies evaluating the impact of Tet2 activity on puberty
progression and reproductive function support this conclusion, as described below.
Using Cre-lox technology, we recently generated mice with Tet2 loss of function
in GnRH neurons (gTKO animals). We expected that this would disrupt GnRH
neuron epigenetic maturation and consequently the acquisition of reproductive
competence (i.e., puberty onset and progression). Surprisingly, we found no evi-
dence of pubertal disruption though, remarkably, gTKO males exhibited an
age-dependent decrease of circulating LH and impairment in fecundity not seen in
wild type animals (Kurian et al. 2014). Considered together with the loss of
H3K4me3 abundance at the Gnrh promoter after Tet2 disruption in a cell line
derived from mature GnRH neurons, these physiological findings suggest that
epigenetic patterns must be actively maintained, as opposed to established during
development.
In fact, Messina and colleagues (2016), focusing on the micro-RNAs (miR-200
and miR-155), have demonstrated the reliance on active maintenance of Gnrh gene
expression. Using Cre-lox technology in mice, they specifically disrupted the
miRNA processing enzyme Dicer in GnRH neurons. Compared to control animals,
246 J. R. Kurian and E. Terasawa

Fig. 9.2 Tet2 binds the

mouse Gnrh gene and
promotes accumulation and
maintenance of an activating
histone modification. (a)
Schematic diagram of the
mouse GnRH gene showing
relative locations of
3 enhancer regions (E1–3) and
the gene promoter
(P) [adapted from Iyer et al.
(2010)]. (b) Graphs represent
relative abundance of Tet2,
H3K4me3 and H3K27me3
detected at each GnRH gene
region using Chromatin
Immunoprecipitation from
immature GnRH neurons
(GN11), differentiated GnRH
neurons (GT1–7), GN11 cells
transfected with Tet2 (GN11
(+) tet2) and GT1–7 cells with
Tet2 knockdown (GT1–7 ()
tet2). Tet2 binds the promoter
and neuron-specific enhancer
(E1) in GT1–7 and GN11 (+)
tet2 cells. H3K4me3 is
significantly more abundant at
the promoter and E1 region in
GT1–7 and GN11 (+) tet2
compared to GN11 cells.
H3K27me3 abundance was
similar across all cell lines at
all regions evaluated.
Comparisons were made
within each chromatin-
associated factor. Levels
assigned different letters are
significantly different
(P < 0.05) from each other by
Tukey post hoc after
Two-Way ANOVA across all
cell lines and genomic
regions. From Kurian
et al. Endocrinology 157:
3588–3603 (2016)
9 Epigenetic Regulation of the GnRH and Kiss1 Genes: Developmental Perspectives 247

Fig. 9.3 Tet2 influences accumulation and maintenance of the activating H3K4 trimethylation in
immature and differentiated GnRH neurons, respectively. (a) Immature GnRH neuronal cell lines
(GN11) express Gnrh at very low levels and less Tet2 than differentiated GnRH neuronal cell lines
(GT1–7). Increasing Tet2 expression in GN11 cells by transient transfection leads to an elevation of
Gnrh mRNA and H3K4me3 at the Gnrh gene neuron-specific enhancer and promoter. The histone
modifications are suspected to be generated by the Tet2 interacting histone-modifying enzyme (MLL:
mixed lineage leukemia) or complex (SET1: a histone methyltransferase complex) (b) Knockdown of
Tet2 in differentiated GnRH neuronal cultures (GT1–7 cells) leads to a loss of Tet2 binding near the
Gnrh gene promoter, loss of H3K4me3 abundance in the same region and reduction in Gnrh mRNA
levels. Modified from Kurian et al. Endocrinology 157: 3588–3603 (2016)

these mice exhibited a gradual decline in GnRH neurons, from typical numbers and
distribution at P0 to no detectable Gnrh expression in adult animals. These animals
also developed hypogonadotropic hypogonadism and had significantly lower
circulating LH and FSH. While mRNA and miRNA expression patterns have not
been profiled in mice lacking Dicer in GnRH neurons, it is clear that miR-200 and
miR-155, which target the GnRH gene repressors Zeb-1 and Cebpp, respectively, are
critical for GnRH neuron maintenance in rodents (Messina et al. 2016). Moreover,
the work by Messina et al. (2016) supports the concept that a set of microRNAs
regulates Gnrh gene expression during sexual development. Whether additional
miRNAs are involved in rodents, whether a similar mechanism can be seen in
primates, and whether the microRNA-dependent epigenetic regulation of GnRH
release is involved in the pathophysiology of disease states, such as idiopathic
hypogoanadotropic hypogonadism in human patients remain open questions.
248 J. R. Kurian and E. Terasawa

9.5 Kisspeptin Neurons

Kisspeptin peptides were originally coined “metastins” after they were discovered in
the mid-1990s as novel ligands that suppressed metastasis of breast cancer and
melanoma (Lee et al. 1996). Since that discovery, the biological roles of kisspeptin
have broadened, most notably for their participation in reproductive maturation
(puberty) and cyclic ovulation. In 2003, two independent reports revealed a novel
signaling mechanism crucial to reproductive function. Specifically, mutations in the
human gene encoding G-protein coupled receptor 54 (GPR54 or KISS1R) were
linked to the absence of puberty or severe disruption of pubertal progression
(Seminara et al. 2003; de Roux et al. 2003). More recently, a familial mutation in
KISS1 (the gene encoding kisspeptins) was linked to the impairment of pubertal
progress (Topaloglu et al. 2012). Mice lacking either Gpr54 or Kiss1 exhibits
delayed puberty and hypogonadotropic hypogonadism (Funes et al. 2003;
d’Anglemont de Tassigny et al. 2007; Lapatto et al. 2007), whereas exogenous
kisspeptin exposures potently stimulate gonadotropin release in prepubertal rodents,
sheep, and primates (Gottsch et al. 2004; Messager et al. 2005; Navarro et al.
2005a, b; Shahab et al. 2005; Dhillo et al. 2007; Guerriero et al. 2012). These
kisspeptin stimulations of gonadotropin release occur through activation of GPR54
(KISS1R) located on GnRH neurons (Hrabovszky et al. 2010; Kirilov et al. 2013;
Yeo et al. 2014), suggesting GnRH neurons are capable of adult levels of activity
prior to puberty onset, and that developmental activation of kisspeptin systems may
play a role in the timing of puberty. Consequently, based on the proposed environ-
mental influence over timing of puberty, significant efforts are currently directed
toward characterizing the epigenetic development and regulation of kisspeptin
neurons.
Kisspeptin cell bodies are concentrated in the arcuate nucleus (ARC) and the
anteroventral periventricular nucleus (AVPV)/periventricular nucleus (PEN)
(Gottsch et al. 2004; Smith et al. 2005; Clarkson and Herbison 2006; Kauffman
et al. 2007). In rodents, the presence of kisspeptin cell bodies in the medial amygdala
is also described (Kim et al. 2011). While KISS1 expression in the ARC is consistent
across species (Smith et al. 2005, 2010; Ramaswamy et al. 2008; Shahab et al. 2005;
Smith 2008; Alcin et al. 2013), there are some species differences in the distribution
pattern of kisspeptin neurons in the anterior population. In rodents, the AVPV
population is very clear but in primates, including humans, a similar but less
conspicuous population is present (Rometo et al. 2007; Hrabovszky et al. 2012;
Smith et al. 2010; Watanabe et al. 2014; Vargas Trujillo et al. 2017). In sheep and
goats, in contrast, kisspeptin neurons are diffusely distributed in the preoptic area
(POA), although within the POA a surge activated population is restricted in the
AVPV area (Franceschini et al. 2006; Smith et al. 2009; Hoffman et al. 2011;
Matsuda et al. 2015). A clear sexual dimorphism of the AVPV population, i.e., a
larger number of kisspeptin neurons in female, exists in mice and rats (Clarkson and
Herbison 2006; Kauffman et al. 2007; Ohkura et al. 2009, Semaan and Kauffman
2010), and this reflects the significance of AVPV kisspeptin neurons in the preovu-
latory surge in rodents. However, there is little sex difference in the number of the
9 Epigenetic Regulation of the GnRH and Kiss1 Genes: Developmental Perspectives 249

anterior kisspeptin neurons in sheep and monkeys and anterior kisspeptin neurons
respond to the estrogen challenge in castrated male monkeys, but not in the castrated
ram (Cheng et al. 2010; Watanabe et al. 2014). Therefore, despite the well-established
concept that negative feedback effects of estradiol are mediated through estrogen
receptor alpha (ERα) located in kisspeptin neurons in the ARC, whereas the positive
feedback effects of estradiol are mediated through ERα in kisspeptin neurons in the
rodent AVPV (Smith 2013; Herbison 2016), the role of kisspeptin neurons in
regulation of GnRH release in primates and ruminants requires further investigation.
In rodents, kisspeptin neuronal systems develop in characteristic regional and
sex-specific patterns. Kiss1 is detectable in the rat and mouse ARC by E13–14
(Desroziers et al. 2012a; Kumar et al. 2014) and persists stably throughout postnatal
and pubertal development (Poling and Kauffman 2013; Semaan and Kauffman
2015). In contrast, AVPV Kiss1 expression becomes apparent at P7–10 (Semaan
et al. 2010; Desroziers et al. 2012b; Clarkson et al. 2009) with only a small amount
of kisspeptin peptide detectable by P15 (Clarkson et al. 2009) but levels substantially
elevate prior to the onset of puberty (Clarkson et al. 2009; Semaan and Kauffman
2015). In addition, the appearance of Kiss1 AVPV briefly lags behind in males and is
significantly lower in pubertal and adult males compared to females (Kauffman et al.
2007; Semaan et al. 2010; Clarkson and Herbison 2006; Homma et al. 2009). The
sex- and region-specific differences in kisspeptin expression are critical to
sex-specific reproductive function and have therefore been a focus of early
investigations of rodent kisspeptin neuron epigenetic regulation. The origin and
prenatal development of prenatal nonhuman primate kisspeptin neurons have not
been studied. Nevertheless, the kisspeptin neurons during the neonatal mini-puberty
are active at levels consistent with adult neurons, and subsequently enter a dormant
state during the prepubertal period until the onset of puberty (Ramaswamy et al.
2013; Shahab et al. 2018).

9.6 Epigenetic Aspects of Sexual Differentiation of Kisspeptin

Neurons

Sex differences in the brain are largely created during a prenatal or perinatal “critical
period” by a sex-specific steroid environment (Phoenix et al. 1959). During this
critical period, a male-specific acute surge of testicular testosterone secretion
coordinates the masculinization/defeminization of the brain and associated physiol-
ogy and behaviors (Simerly 1998, 2002). Importantly, prior to acute sex steroid
exposure, the brain exhibits the potential to develop both male- and female-specific
neural circuitry. While in rodents testosterone aromatized into estradiol is responsi-
ble for this organizational action (Arnold and Gorski 1984; McCarthy 1994), in
ruminants and primates, androgens themselves are responsible (Wallen 2005;
Puttabyatappa and Padmanabhan 2017). As evidence, male-like brain development
can be initiated by perinatal estradiol treatment of female rats and mice, whereas
androgen treatment of early to mid-embryonic sheep and monkey embryos directs
male-like brain development (Phoenix et al. 1959; Arai and Gorski 1968; see Wallen
250 J. R. Kurian and E. Terasawa

2009). Conversely, castration of newborn male rats and mice leads to a female-like
brain in adulthood (Döhler et al. 1984). The critical period in rodents extends
through approximately P10, after which sex steroid manipulations no longer alter
sexually dimorphic brain structures (Gorski 1985). Some brain traits, however,
appear to be influenced by steroid exposures during the pubertal stage (Schulz and
Sisk 2006; Schulz et al. 2009).
Importantly, in addition to sex steroids, sexually dimorphic experiences during
the perinatal critical period also appear to masculinize the brain. For example, male
rodents experience a higher degree of maternal grooming than female littermates
during this period. Subjecting female rats to a sex steroid surge or simulated maternal
grooming in the first week of life is reported to masculinize the ERα gene CpG
methylation pattern and gene expression profile in the MBH (Kurian et al. 2010b).
That study was the first to report an equivalent impact of steroids and environmental
influence over sexual differentiation of epigenetic patterns in the brain. This
suggested that epigenetic mechanisms likely had a significant role in the sexual
differentiation of the brain in general. In fact, several “critical periods” for sex steroid
exposures were soon after related to epigenetic factors, including bed nucleus of the
stria terminalus (BNST) size and vasopressin fiber projections (Murray et al. 2011),
male sexual behavior in rats and olfactory behavior in mice (Matsuda et al. 2011),
and steroid receptor and DNA methyltransferase 3a expression (Schwarz et al. 2010;
Kolodkin and Auger 2011).
To date, only one study has investigated whether sexual differences of the two
kisspeptin systems are driven by epigenetic factors. Specifically, Semaan,
Kauffman, and colleagues evaluated whether histone deacetylation and DNA meth-
ylation contributed to sexually dimorphic regulation of AVPV Kiss1 gene expres-
sion and cell number (Semaan et al. 2012). They pharmacologically blocked histone
deacetylation during the postnatal critical period with valproic acid, and evaluated
male and female AVPV Kiss1 expression in adulthood. While the authors observed
the expected disruption of sex differences in the BNST similar to that reported by
Murray et al. (2011), Kiss1 gene expression in the AVPV was elevated in both sexes
compared to same sex controls and there was no sex-specific effect per se. Whether,
the elevation of Kiss1 expression in males had a functional impact (i.e., feminization
of estradiol-induced LH surge potential) was not evaluated and remains an important
question. Additionally, potential off-target effects of valproic acid are a significant
concern in interpreting these results, as valproic acid is also a voltage-gated sodium
channel blocker and elevates GABA levels in the brain, which could indirectly alter
GnRH neuronal activity.
In the same study, Semaan et al. (2012) further evaluated DNA methylation over
70 unique sites of the Kiss1 gene in microdissected AVPV and ARC tissues. CpG
methylation was significantly higher in females, primarily at the putative Kiss1
promoter in the AVPV but not in the ARC. This finding was unexpected, as
methylation is typically associated with reduced gene expression, though AVPV
Kiss1 expression is substantially higher in females than males. Consequently, it is
currently unclear how methylation alters Kiss 1 gene expression, though it is likely
involved in chromatin conformation, which is critical to hormone-dependent Kiss
9 Epigenetic Regulation of the GnRH and Kiss1 Genes: Developmental Perspectives 251

1 expression in the AVPV (discussed in the following section). Whether the perinatal
hormone surge drives epigenetic changes in the Kiss1 gene in the AVPV also
remains an open critical question.

9.7 Epigenetic Aspects of Estrogen Signaling on Kiss1 Gene

Expression

An essential aspect of female reproductive function is the ability to generate LH

surges in response to estradiol. LH surges are critical to gamete maturation and
ovulation and require ERα signaling in AVPV kisspeptin neurons (d’Anglemont de
Tassigny and Colledge 2010; Dubois et al. 2015). As stated above, acute estradiol
(E2) exposure in adult females induces suppression of Kiss1 gene expression in the
ARC, whereas it activates Kiss1 expression in the AVPV (Smith et al. 2005). This
phenomenon was recently utilized to interrogate the epigenetic effects of E2 action
on Kiss1 gene expression in both regions (Fig. 9.4).
First, using only adult female mice, Tomikawa and colleagues reported that E2
exposure increased ERα binding at the Kiss1 promoter in the AVPV, but not in the
ARC (Tomikawa et al. 2012). In addition, total histone 3 (H3) acetylation at the
Kiss1 promoter increased in the AVPV and simultaneously dropped in the ARC
during E2 exposure. Interestingly, H3 acetylation in only the ARC also changed in a
50 region upstream of the Kiss1 promoter (increased acetylation) and a 30 intergenic
region (decreased acetylation) located downstream of the last exon of the Kiss1 gene.
These later differences in H3 acetylation are probably the consequence of brain
region specific chromatin loop conformations. Specifically, using chromatin confor-
mation capture assays, Tomikawa and colleagues exquisitely found that in the ARC,
the Kiss1 gene is held in a tight loop, with interactions between a 50 region and a
proximal 30 region centered at the promoter. In contrast, no chromatin loop is
apparent in the AVPV; however, during E2 exposure, a loop forms between the
promoter and a distal intergenic 30 region, later shown to act as an enhancer in the
AVPV. That same distal 30 region also associates with the promoter in the ARC
during E2 exposure. Whereas this region is necessary for enhancing E2-driven
AVPV Kiss1 expression, it is also necessary for E2 inhibition of ARC Kiss1
expression. We consequently refer to this region as a “transcriptional regulatory
region.” Importantly, while histone acetylation is typically associated with increased
gene expression, the ARC 50 region exhibits increased H3 acetylation even while
Kiss1 expression drops significantly. Examination of the DNA loop structures
generated by E2 exposure in the ARC (Fig. 9.4) sheds light on this apparent
discordance. It is possible that the tight looping of DNA results in histone displace-
ment without any loss of histone acetylation, thereby shifting the regional
associations between the Kiss1 gene and histone core. In essence, this would mean
no net gain of histone acetylation, but rather that a regional shift in H3 acetylation
patterns. To clarify the relationship between histone acetylation and Kiss 1 gene
expression in the ARC and AVPV, closer examination of the histone
252 J. R. Kurian and E. Terasawa

Fig. 9.4 Estradiol (E2) exposure has brain region-specific influence over Kiss1 gene expression,
chromatin conformation and histone 3 (H3) acetylation (Tomikawa et al. 2012). (a) A schematic
representation of the mouse Kiss1 gene depicts several critical regions as defined by color and shapes
including a 50 region (black), a coding region (green) that includes a promoter (lime green box) 3 exons
(numbered boxes) and a CpG island (red box, within the third exon), a 30 region (blue) and a
transcriptional modifier region (red). Small gray shapes indicate sites utilized for analysis by
Tomikawa et al. (2012). The table indicates the direction of H3 acetylation abundance detected in
each gene region in the ARC and AVPV during E2 exposure. H3 acetylation significantly increased
("), decreased (#), or remained the same ($); N.D. not determined. (b) Representations of DNA
looping at the Kiss1 gene in the ARC and AVPV. Yellow circles represent H3 acetylation abundance.
Location of yellow circles enables visualization of Kiss1 regional interactions with the acetylated
histone. Prior to E2 exposure, the Kiss1 gene structure is similar in the ACR and AVPV, though a
second loop is apparent in only the ARC. E2 exposure causes formation of a complex DNA structure
and decrease in Kiss1 expression in the ARC. On the other hand, the AVPV 50 loop is lost in favor of a
long range 30 loop and increased Kiss1 expression during E2 exposure. Created, based on Tomikawa
et al., Proc Natl Acad Sci USA 109:E1294-1301 (2012), Uenoyama et al., Neuroendocrinology
103:640–649 (2016), and Kurian et al., Endocrinology 157: 3588–3603 (2016)

acetyltransferase and histone deacetylase activities in kisspeptin neurons during E2

exposure is necessary.
Based on these intriguing discoveries, critical questions arise. First, how do these
different populations of kisspeptin neurons evolve to create distinct orientations of
the kisspeptin gene? Are there sex-specific differences in Kiss1 gene conformation?
9 Epigenetic Regulation of the GnRH and Kiss1 Genes: Developmental Perspectives 253

Also, how might DNA methylation status affect chromatin conformation?

Tomikawa and colleagues found no substantial differences in CpG methylation
status in the basal or hormone exposed states, but they only assessed six CpG sites
near the transcription start site. Importantly, the most pronounced CpG island (CGI)
of the Kiss1 gene is located in the third exon. There are presently no reports of
differences in methylation status across this region between ARC and AVPV tissue.
This could be critical, as the Kiss1 gene contains dense CCCTC-binding factor
(CTCF) sites surrounding the last exon as well as between the promoter and first
exon. As CTCF binding and consequential modification of chromatin loop structure
is profoundly influenced by DNA methylation status (Ito et al. 2013; Ong and Corces
2014), it is tempting to speculate that DNA methylation could factor heavily into
brain region-specific Kiss1 gene conformations. Nevertheless, at this time, there are
no reports of CTCF binding within the Kiss1 gene or differences in methylation
status across the third exon CGI.

9.8 Epigenetic Regulation of the Kiss1 Gene and Kisspeptin

Cells in the Context of Puberty

Kisspeptin signaling is essential for puberty onset and progression. Because puberty
onset is heavily influenced by environmental conditions including climate, stress,
body weight, and environmental estrogens, interest in the epigenetic regulation of
kisspeptin cell development and puberty-related Kiss1 gene expression has grown
substantially. As mentioned earlier, similar to the findings of Mellon and colleagues
comparing Gnrh gene promoter structure of immature and mature cell lines,
Lomniczi et al. (2013) reported that the female rat pubertal transition is accompanied
by increased prevalence of activating histone PTMs at the Kiss1 promoter. Specifi-
cally, prior to puberty onset, the kisspeptin promoter was associated with a bivalent
domain (i.e., H3K27me3 and H3K4me3) and occupied by a component of a
polycomb repressive complex, EED. The transition to puberty from prepubertal
period was accompanied by decreased EED occupancy of the kisspeptin promoter,
a gradual loss of the repressive H3K27me3 PTM, and increased levels of the
permissive H3K9,14 acetylation and H3K4me3 PTMs. The authors suggest this
process is the consequence of increased DNA methylation of the EED promoter,
leading to lower EED expression and consequential decreased occupancy of the
kisspeptin promoter by the repressive PRC2 (EED, SUZ12, Ezh2) complex. How-
ever, these conclusions were largely based on observations of dramatically delayed
puberty in female mice when DNA methylation was inhibited by peripheral admin-
istration of the pharmacological DNA methyltransferase inhibitor 5-azacytidine
(5-aza). While the findings are consistent with the expected results, this 5-aza
activity may not be specific to inhibition of postmitotic neuronal DNA methylation.
Because the well-characterized mechanism of 5-aza requires nucleoside
incorporation into DNA (Schermelleh et al. 2005; Stresemann and Lyko 2008;
Svedruzic 2008) the mechanism by which this compound inhibits DNMT in
postmitotic cells remains unclear. In addition, the significant reduction in growth
254 J. R. Kurian and E. Terasawa

rate and elevated levels of circulating corticosterone that occurs following initiation
of drug treatment (Lomniczi et al. 2013) is indicative of toxicity that likely
contributes to delays in sexual maturation. Because of these concerns, more direct
approaches (e.g., cell-specific genetic or enzyme expression manipulations) are
necessary to clarify the relationship between DNA methylation, chromatin
modifications, and puberty onset. Nonetheless, these studies and a recent report
that isolated another vital repressor gene (GATAD1) through an extensive gene
array analysis in orchiectomized male monkey hypothalamus (Lomniczi et al. 2015),
further establish the concept that structural modification of bivalent promoters in the
neuroendocrine hypothalamus is an integral step toward puberty onset and repro-
ductive function in primates and rodents.
Lomniczi and colleagues’ focused approach toward measuring histone modifica-
tion status during development at one gene has tremendous value for clarifying the
temporal progression of bivalent promoter transactivation. Importantly, they report
that permissive histone PTMs (H3K4me3 and H3K9,14 acetylation) accumulate
near the kisspeptin promoter during the transition from juvenile to early pubertal
stages. This preceded the loss of the repressive H3K27me3 PTM, suggesting that the
recruitment of activating complexes is imperative for transactivation and increased
gene expression. Based on the temporal relationship between H3K4me3 accumula-
tion and loss of H3K27me3, as well as the association between Tet2 and H3K4me3,
we recently evaluated the impact of kisspeptin neuron-specific ablation of Tet2
(kTKO animals) on puberty onset and progression in mice (Kurian et al. 2014).
Remarkably, Tet2 disruption delayed puberty onset and progression in both male
and female kTKO animals. Whether this impact on puberty is the consequence of
Tet2-mediated DNA demethylation, altered DNA hydroxymethylation patterns,
H3K4me3 accumulation or another mechanism remain open and important
questions. More recent evidence indicates that histone-modifying Tet2 interactors
(i.e., MLL1 and MLL3) are also critical mediators of increased rodent kiss1 expres-
sion and pubertal progression; suggesting factors that drive Tet2/MLL interactions
could represent signals that are critical for initiating puberty (Toro, et al. 2018).
Which genomic regions in kisspeptin neurons, in addition to Kiss1, are altered
following Tet2 disruption also remain unknown.
The similar genomic targets as integral positions of epigenetic regulation in
kisspeptin cells are predicated on the recent discoveries of additional peptides that
are co-expressed in ARC kisspeptin neurons, namely neurokinin B (Tac2) and
dynorphin. Neurons that express all three peptides are known as KNDy neurons
(Goodman et al. 2014). In fact, Tac2 expression in the ARC in mice is an early
marker of puberty onset (Gill et al. 2012a), and is likely regulated by a bivalent
promoter. As evidence, recent preliminary studies (Gill et al. 2012b) found that
female mice heterozygous for a loss-of-function mutation in Lsd1, a histone lysine
demethylase of di- and mono-methylated H3K4, exhibit precocious vaginal opening
and first ovulation (3–4 days prior to wild-type littermates), with early elevations of
circulating gonadotropins and hypothalamic expression of Tac2 (Gill et al. 2012b).
Importantly, these findings parallel those of Lomniczi et al. (2013), in that lower
expression of an epigenetic repressive enzyme is related to the elevated
9 Epigenetic Regulation of the GnRH and Kiss1 Genes: Developmental Perspectives 255

hypothalamic expression of a puberty-related gene. However, it is not known

whether Lsd1 expression or activity changes across development. In addition,
Lsd1 also stimulates hormone (ligand associated androgen receptor) mediated
gene activation through demethylation of H3K9 (Metzger et al. 2005). Conse-
quently, while preliminary evidence suggests the intriguing possibility that Lsd1
directly alters hypothalamic development (including kisspeptin neurons) during the
pubertal transition, models for cell or region-specific genetic manipulations will be
necessary to verify this interpretation. These models will be critical for determining
the primary activity of Lsd1 (i.e., H3K4 or H3K9 demethylation) as it relates to
reproductive maturation. Nonetheless, it is increasingly clear that Tac2 is an addi-
tional genomic loci within kisspeptin neurons that is the target of repressive and
activating epigenetic complexes related to the transition to puberty (Toro et al.
2018).
The concept that KNDy neurons in the arcuate nucleus are critical to the regula-
tion of GnRH release has been proposed, based on the finding that kisspeptin,
neurokinin B and dynorphin are all co-localized in the same neuron in rodents and
ruminants (Goodman et al. 2014). In primates, including humans, however, the rate
of co-localization of these 3 peptides is not as high as reported in other mammalian
species (Ramaswamy et al. 2010; Hrabovszky 2014). Nevertheless, studies in this
lab indicate that kisspeptin, neurokinin B and dynorphin have considerable
interactions in the median eminence of the hypothalamus as each independent
entry (Garcia et al. 2017, 2018). Moreover, the degree of interactions between
these 3 peptidergic neurons varies during development, and different hormonal
and physiological conditions (Terasawa et al. 2018). Whether and how epigenetic
mechanisms are involved in the interactions between 3 peptidergic neurons remain
as a great challenge.

9.9 Conclusion

Taken together, the findings detailed in this chapter highlights exciting connections
between epigenetic regulations in different cells of the reproductive neuroendocrine
hypothalamus. Nevertheless, our understanding of epigenetic regulation of GnRH
and kisspeptin neurons and reproductive neuroendocrine function, in general, is still
in its infancy. In fact, the findings described in this chapter are all based on changes
in known epigenetic markers in a given physiological condition and the question of
whether actual modifications of environmental conditions result in similar epigenetic
modifications in GnRH and kisspeptin neurons, along with altered physiology, has
not been examined to date. For example, it is known that exposure of prepubertal
animals to stress results in delayed puberty, whereas exposure of prepubertal animals
to a high-calorie diet induces precocious puberty. Do these environmental changes
cause consistent opposite epigenetic modifications to GnRH? The field is
progressing rapidly. We are inching closer to the ultimate promise of neuroendocrine
epigenetic research: to define how the environment influences our physiology.
256 J. R. Kurian and E. Terasawa

Acknowledgments The authors express their sincere appreciation for many current and past
postdoctoral research fellows, graduate students, and research specialists including Kim L. Keen,
who has contributed to research in the Terasawa laboratory as well as Somaja Louis and Jennifer Li,
who contributed to research in the Kurian laboratory. This work was supported by the NIH grants,
HD011355, HD015344, and HD077447 for ET, ES020878 for JRK, and OD01106 and RR0061 for
the WNPRC.

Key References

Clark and Mellon (1995). This is the first paper identifying that the homeodomain
transcription factor Oct 1 regulates the GnRH gene transcription.
Iyer et al. (2011). This is the first paper demonstrating chromatin modifications of
GnRH gene expression during neuronal differentiation.
Kepa et al. (1996a). This is one of the first works to analyze the structure of GnRH
gene promoter.
Kurian et al. (2010a). This is the first paper to demonstrate that epigenetic changes
occur during GnRH neuronal maturation.
Lomniczi et al. (2015). This is a seminal work demonstrating that Zinc finger protein
is involved in prepubertal GnRH suppression in non-human primates.
Messina et al. (2016). This paper identified that microRNAs regulate GnRH gene
expression, which correlates with the timing of puberty.

References
Alcin E, Sahu A, Ramaswamy S, Hutz ED, Keen KL, Terasawa E, Bethea CL, Plant TM (2013)
Ovarian regulation of kisspeptin neurones in the arcuate nucleus of the rhesus monkey (Macaca
mulatta). J Neuroendocrinol 5:488–496
Arai Y, Gorski RA (1968) Critical exposure time for androgenization of the developing hypothala-
mus in the female rat. Endocrinology 82:1010–1014
Arnold AP, Gorski RA (1984) Gonadal steroid induction of structural sex differences in the central
nervous system. Annu Rev Neurosci 7:413–442
Azuara V, Perry P, Sauer S, Spivakov M, Jørgensen HF, John RM, Gouti M, Casanova M,
Warnes G, Merkenschlager M, Fisher AG (2006) Chromatin signatures of pluripotent cell
lines. Nat Cell Biol 8:532–538
Banks GC, Deterding LJ, Tomer KB, Archer TK (2001) Hormone-mediated dephosphorylation of
specific histone H1 isoforms. J Biol Chem 276:36467–36473
Belsham DD, Mellon PL (2000) Transcription factors Oct-1 and C/EBPbeta (CCAAT/enhancer-
binding protein-beta) are involved in the glutamate/nitric oxide/cyclic-guanosine 5'-
monophosphate-mediated repression of mediated repression of gonadotropin-releasing hor-
mone gene expression. Mol Endocrinol 14:212–228
Bernstein BE, Mikkelsen TS, Xie X, Kamal M, Huebert DJ, Cuff J, Fry B, Meissner A, Wernig M,
Plath K, Jaenisch R, Wagschal A, Feil R, Schreiber SL, Lander ES (2006) A bivalent chromatin
structure marks key developmental genes in embryonic stem cells. Cell 125:315–326
Bestor TH, Ingram VM (1983) Two DNA methyltransferases from murine erythroleukemia cells:
purification, sequence specificity, and mode of interaction with DNA. Proc Natl Acad Sci U S A
80:5559–5563
9 Epigenetic Regulation of the GnRH and Kiss1 Genes: Developmental Perspectives 257

Bestor T, Laudano A, Mattaliano R, Ingram V (1988) Cloning and sequencing of a cDNA encoding
DNA methyltransferase of mouse cells. The carboxyl-terminal domain of the mammalian
enzymes is related to bacterial restriction methyltransferases. J Mol Biol 203:971–983
Branco MR, Ficz G, Reik W (2012) Uncovering the role of 5-hydroxymethylcytosine in the
epigenome. Nat Rev Genet 13:7–13
Carey BW, Finley LW, Cross JR, Allis CD, Thompson CB (2015) Intracellular α-ketoglutarate
maintains the pluripotency of embryonic stem cells. Nature 518(7539):413–416
Cheng G, Coolen LM, Padmanabhan V, Goodman RL, Lehman MN (2010) The kisspeptin/
neurokinin B/dynorphin (KNDy) cell population of the arcuate nucleus: sex differences and
effects of prenatal testosterone in sheep. Endocrinology 151:301–311
Clark ME, Mellon PL (1995) The POU homeodomain transcription factor Oct-1 is essential for
activity of the gonadotropin-releasing hormone neuron-specific enhancer. Mol Cell Biol
15:6169–6177
Clarkson J, Herbison AE (2006) Postnatal development of kisspeptin neurons in mouse hypothala-
mus; sexual dimorphism and projections to gonadotropin-releasing hormone neurons. Endocri-
nology 147:5817–5825
Clarkson J, Boon WC, Simpson ER, Herbison AE (2009) Postnatal development of an estradiol-
kisspeptin positive feedback mechanism implicated in puberty onset. Endocrinology
150:3214–3220
Constantin S, Caraty A, Wray S, Duittoz AH (2009) Development of gonadotropin-releasing
hormone-1 secretion in mouse nasal explants. Endocrinology 150:3221–3227
d’Anglemont de Tassigny X, Colledge WH (2010) The role of kisspeptin signaling in reproduction.
Physiology 25:207–217
d’Anglemont de Tassigny X, Fagg LA, Dixon JP, Day K, Leitch HG, Hendrick AG, Zahn D,
Franceschini I, Caraty A, Carlton MB, Aparicio SA, Colledge WH (2007) Hypogonadotropic
hypogonadism in mice lacking a functional Kiss1 gene. Proc Natl Acad Sci U S A
104:10714–10719
Dawlaty MM, Ganz K, Powell BE, Hu YC, Markoulaki S, Cheng AW, Gao Q, Kim J, Choi SW,
Page DC, Jaenisch R (2011) Tet1 is dispensable for maintaining pluripotency and its loss is
compatible with embryonic and postnatal development. Cell Stem Cell 9:166–175
Dawlaty MM, Breiling A, Le T, Raddatz G, Barrasa MI, Cheng AW, Gao Q, Powell BE, Li Z,
Xu M, Faull KF, Lyko F, Jaenisch R (2013) Combined deficiency of Tet1 and Tet2 causes
epigenetic abnormalities but is compatible with postnatal development. Dev Cell 24:310–323
de Roux N, Genin E, Carel JC, Matsuda F, Chaussain JL, Milgrom E (2003) Hypogonadotropic
hypogonadism due to loss of function of the KiSS1-derived peptide receptor GPR54. Proc Natl
Acad Sci U S A 100:10972–10976
Deaton AM, Bird A (2011) CpG islands and the regulation of transcription. Genes Dev
25:1010–1022
Deplus R, Delatte B, Schwinn MK, Defrance M, Méndez J, Murphy N, Dawson MA, Volkmar M,
Putmans P, Calonne E, Shih AH, Levine RL, Bernard O, Mercher T, Solary E, Urh M, Daniels
DL, Fuks F (2013) TET2 and TET3 regulate GlcNAcylation and H3K4 methylation through
OGT and SET1/COMPASS. EMBO J 32:645–655
Desroziers E, Droguerre M, Bentsen AH, Robert V, Mikkelsen JD, Caraty A, Tillet Y, Duittoz A,
Franceschini I (2012a) Embryonic development of kisspeptin neurones in rat.
J Neuroendocrinol 24:1284–1295
Desroziers E, Mikkelsen JD, Duittoz A, Franceschini I (2012b) Kisspeptin-immunoreactivity
changes in a sex- and hypothalamic-region-specific manner across rat postnatal development.
J Neuroendocrinol 24:1154–1165
Dhillo WS, Chaudhri OB, Thompson EL, Murphy KG, Patterson M, Ramachandran R, Nijher GK,
Amber V, Kokkinos A, Donaldson M, Ghatei MA, Bloom SR (2007) Kisspeptin-54 stimulates
gonadotropin release most potently during the preovulatory phase of the menstrual cycle in
women. J Clin Endocrinol Metab 92:3958–3966
Döhler KD, Hancke JL, Srivastava SS, Hofmann C, Shryne JE, Gorski RA (1984) Participation of
estrogens in female sexual differentiation of the brain; neuroanatomical, neuroendocrine and
behavioral evidence. Prog Brain Res 61:99–117
258 J. R. Kurian and E. Terasawa

Dubois SL, Acosta-Martínez M, DeJoseph MR, Wolfe A, Radovick S, Boehm U, Urban JH, Levine
JE (2015) Positive, but not negative feedback actions of estradiol in adult female mice require
estrogen receptor α in kisspeptin neurons. Endocrinology 156:1111–1120
El Majdoubi M, Sahu A, Ramaswamy S, Plant TM (2000) Neuropeptide Y: a hypothalamic brake
restraining the onset of puberty in primates. Proc Natl Acad Sci U S A 97:6179–6184
Franceschini I, Lomet D, Cateau M, Delsol G, Tillet Y, Caraty A (2006) Kisspeptin immunoreac-
tive cells of the ovine preoptic area and arcuate nucleus co-express estrogen receptor alpha.
Neurosci Lett 401:225–230
Fueshko S, Wray S (1994) LHRH cells migrate on peripherin fibers in embryonic olfactory explant
cultures: an in vitro model for neurophilic neuronal migration. Dev Biol 166:331–348
Funes S, Hedrick JA, Vassileva G, Markowitz L, Abbondanzo S, Golovko A, Yang S, Monsma FJ,
Gustafson EL (2003) The KiSS-1 receptor GPR54 is essential for the development of the murine
reproductive system. Biochem Biophys Res Commun 312:1357–1363
Gan L, Ni PY, Ge Y, Xiao YF, Sun CY, Deng L, Zhang W, Wu SS, Liu Y, Jiang W, Xin HB (2012)
Histone deacetylases regulate gonadotropin-releasing hormone I gene expression via
modulating Otx2-driven transcriptional activity. PLoS One 7:e39770
Garcia JP, Guerriero KA, Keen KL, Kenealy BP, Seminara SB, Terasawa E (2017) Kisspeptin and
neurokinin B signaling network underlies the pubertal increase in GnRH release in female
rhesus monkeys. Endocrinology 158:3269–3280
Garcia JP, Keen KL, Kenealy BP, Seminara SB, Terasawa E (2018) Role of kisspeptin and
neurokinin B signaling in male rhesus monkey puberty. Endocrinology 159:3048–3060
Gardiner-Garden M, Frommer M (1987) CpG islands in vertebrate genomes. J Mol Biol
196:261–282
Gill JC, Navarro VM, Kwong C, Noel SD, Martin C, Xu S, Clifton DK, Carroll RS, Steiner RA,
Kaiser UB (2012a) Increased neurokinin B (Tac2) expression in the mouse arcuate nucleus is an
early marker of pubertal onset with differential sensitivity to sex steroid-negative feedback than
Kiss1. Endocrinology 153:4883–4893
Gill J, Kwong C, Clark E, Carroll RS, Shi YG, Kaiser UB (2012b) A role for the histone
demethylase LSD1 in controlling the timing of pubertal onset. In: Abstracts for Endocrine
Society Annual Meeting, Huston, TX, No. OR12–3
Givens ML, Rave-Harel N, Goonewardena VD, Kurotani R, Berdy SE, Swan CH, Rubenstein JL,
Robert B, Mellon PL (2005) Developmental regulation of gonadotropin-releasing hormone
gene expression by the MSX and DLX homeodomain protein families. J Biol Chem
280:19156–19165
Glaser S, Schaft J, Lubitz S, Vintersten K, van der Hoeven F, Tufteland KR, Aasland R,
Anastassiadis K, Ang SL, Stewart AF (2006) Multiple epigenetic maintenance factors
implicated by the loss of Mll2 in mouse development. Development 133:1423–1432
Goodman RL, Coolen LM, Lehman MN (2014) A role for neurokinin B in pulsatile GnRH
secretion in the ewe. Neuroendocrinology 99:18–32
Gorski RA (1985) Sexual dimorphisms of the brain. J Anim Sci 61(Suppl 3):38–61
Gottsch ML, Cunningham MJ, Smith JT, Popa SM, Acohido BV, Crowley WF, Seminara S, Clifton
DK, Steiner RA (2004) A role for kisspeptins in the regulation of gonadotropin secretion in the
mouse. Endocrinology 145:4073–4077
Guerriero KA, Keen KL, Millar RP, Terasawa E (2012) Developmental changes in GnRH release in
response to kisspeptin agonist and antagonist in female Rhesus monkeys (Macaca mulatta):
implication for the mechanism of puberty. Endocrinology 153:825–836
Guo JU, Su Y, Shin JH, Shin J, Li H, Xie B, Zhong C, Hu S, Le T, Fan G, Zhu H, Chang Q, Gao Y,
Ming GL, Song H (2014) Distribution, recognition and regulation of non-CpG methylation in
the adult mammalian brain. Nat Neurosci 17:215–222
Hahn MA, Qiu R, Wu X, Li AX, Zhang H, Wang J, Jui J, Jin SG, Jiang Y, Pfeifer GP, Lu Q (2013)
Dynamics of 5-hydroxymethylcytosine and chromatin marks in mammalian neurogenesis. Cell
Rep 3:291–300
Herbison AE (2016) Control of puberty onset and fertility by gonadotropin-releasing hormone
neurons. Nat Rev Endocrinol 12:452–466
9 Epigenetic Regulation of the GnRH and Kiss1 Genes: Developmental Perspectives 259

Hermann A, Goyal R, Jeltsch A (2004) The Dnmt1 DNA-(cytosine-C5)-methyltransferase

methylates DNA processively with high preference for hemimethylated target sites. J Biol
Chem 279:48350–48359
Hoffmann HM, Mellon PL (2018) Regulation of GnRH gene expression. In: Herbison AE, Plant
TM (eds) Master class in neuroendocrinology series.The GnRH neuron and its control. Wiley,
Hoboken, NJ, pp 95–119
Hoffmann GE, Le WW, Franceschini I, Caraty A, Advis JP (2011) Expression of fos and in vivo
median eminence release of LHRH identifies an active role for preoptic area kisspeptin neurons
in synchronized surges of LH and LHRH in the ewe. Endocrinology 152:214–222
Hoffmann HM, Gong P, Tamrazian A, Mellon PL (2018) Transcriptional interaction between cFOS
and the homeodomain-binding transcription factor VAX1 on the GnRH promoter controls
Gnrh1 expression levels in a GnRH neuron maturation specific manner. Mol Cell Endocrinol
461:143–154
Homma T, Sakakibara M, Yamada S, Kinoshita M, Iwata K, Tomikawa J, Kanazawa T, Matsui H,
Takatsu Y, Ohtaki T, Matsumoto H, Uenoyama Y, Maeda K, Tsukamura H (2009) Significance
of neonatal testicular sex steroids to defeminize anteroventral periventricular kisspeptin neurons
and the GnRH/LH surge system in male rats. Biol Reprod 81:1216–1225
Hrabovszky E (2014) Neuroanatomy of the human hypothalamic kisspeptin system. Neuroendocri-
nology 99:33–48
Hrabovszky E, Ciofi P, Vida B, Horvath MC, Keller E, Caraty A, Bloom SR, Ghatei MA, Dhillo
WS, Liposits Z, Kallo I (2010) The kisspeptin system of the human hypothalamus: sexual
dimorphism and relationship with gonadotropin-releasing hormone and neurokinin B neurons.
Eur J Neurosci 31:1984–1998
Hrabovszky E, Sipos MT, Molnár CS, Ciofi P, Borsay BÁ, Gergely P, Herczeg L, Bloom SR,
Ghatei MA, Dhillo WS, Liposits Z (2012) Low degree of overlap between kisspeptin,
neurokinin B, and dynorphin immunoreactivities in the infundibular nucleus of young male
human subjects challenges the KNDy neuron concept. Endocrinology 153:4978–4989
Ito Y, Nativio R, Murrell A (2013) Induced DNA demethylation can reshape chromatin topology at
the IGF2-H19 locus. Nucleic Acids Res 41:5290–5302
Iyer AK, Miller NL, Yip K, Tran BH, Mellon PL (2010) Enhancers of GnRH transcription
embedded in an upstream gene use homeodomain proteins to specify hypothalamic expression.
Mol Endocrinol 24:1949–1964
Iyer AK, Brayman MJ, Mellon PL (2011) Dynamic chromatin modifications control GnRH gene
expression during neuronal differentiation and protein kinase C signal transduction. Mol
Endocrinol 25:460–473
Kauffman AS, Gottsch ML, Roa J (2007) Sexual differentiation of Kiss1 gene expression in the
brain of the rat. Endocrinology 148:1774–1783
Kelley CG, Lavorgna G, Clark ME, Boncinelli E, Mellon PL (2000) The Otx2 homeoprotein
regulates expression from the gonadotropin-releasing hormone proximal promoter. Mol
Endocrinol 14:1246–1256
Kepa JK, Wang C, Neeley CI, Raynolds MV, Gordon DF, Wood WM, Wierman ME (1992)
Structure of the rat gonadotropin releasing hormone (rGnRH) gene promoter and functional
analysis in hypothalamic cells. Nucleic Acids Res 20:1393–1399
Kepa JK, Spaulding AJ, Jacobsen BM, Fang Z, Xiong X, Radovick S, Wierman ME (1996a)
Structure of the distal human gonadotropin releasing hormone (hGnRH) gene promoter and
functional analysis in GT1-7 neuronal cells. Nucleic Acids Res 24:3614–3620
Kepa JK, Jacobsen BM, Boen EA, Prendergast P, Edwards DP, Takimoto G, Wierman ME (1996b)
Direct binding of progesterone receptor to nonconsensus DNA sequences represses rat GnRH.
Mol Cell Endocrinol 117:3927–3910
Kim HH, Wolfe A, Cohen RN, Eames SC, Johnson AL, Wieland CN, Radovick S (2007) In vivo
identification of a 107-base pair promoter element mediating neuron-specific expression of
mouse gonadotropin-releasing hormone. Mol Endocrinol 21:457–471
Kim J, Semaan SJ, Clifton DK, Steiner RA, Dhamija S, Kauffman AS (2011) Regulation of Kiss1
expression by sex steroids in the amygdala of the rat and mouse. Endocrinology 152:2020–2030
260 J. R. Kurian and E. Terasawa

Kirilov M, Clarkson J, Liu X, Roa J, Campos P, Porteous R, Schütz G, Herbison AE (2013)

Dependence of fertility on kisspeptin-Gpr54 signaling at the GnRH neuron. Nat Commun
4:2492
Kolodkin MH, Auger AP (2011) Sex difference in the expression of DNA methyltransferase 3a in
the rat amygdala during development. J Neuroendocrinol 23:577–583
Kumar D, Freese M, Drexler D, Hermans-Borgmeyer I, Marquardt A, Boehm U (2014) Murine
arcuate nucleus kisspeptin neurons communicate with GnRH neurons in utero. J Neurosci
34:3756–3766
Kurian JR (2016) Epigenetic regulation of the GnRH and Kiss1 genes. In: Binder E (ed) Epigenetics
and neuroendocrinology, vol 1. Springer Nature, Switzerland, pp 243–264
Kurian JR, Terasawa E (2013) Epigenetic control of gonadotropin releasing hormone neurons.
Front Endocrinol (Lausanne) 4:61
Kurian JR, Keen KL, Terasawa E (2010a) Epigenetic changes coincide with in vitro primate GnRH
neuronal maturation. Endocrinology 151:5359–5368
Kurian JR, Olesen KM, Auger AP (2010b) Sex differences in epigenetic regulation of the estrogen
receptor-alpha promoter within the developing preoptic area. Endocrinology 151:2297–2305
Kurian JR, Kapke JA, Terasawa E (2011) Epigenetic contribution to GnRH neuronal development
in male pubertal development. In: Abstracts for the Society for Neuroscience Annual Meeting,
Washington, DC, No. 500.03
Kurian JR, Terasawa E, Levine JE (2014) Ten eleven translocase 2 (Tet2) drives GnRH gene
expression and enables the typical progression through puberty. In: Abstracts for the Endocrine
Society Annual Meeting, Chicago, IL, No.OR03-3
Lapatto R, Pallais JC, Zhang D, Chan YM, Mahan A, Cerrato F, Le WW, Hoffman GE, Seminara
SB (2007) Kiss1/ mice exhibit more variable hypogonadism than Gpr54/ mice. Endo-
crinology 148:4927–4936
Larder R, Mellon PL (2009) Otx2 induction of the gonadotropin-releasing hormone promoter is
modulated by direct interactions with Grg co-repressors. J Biol Chem 284:16966–16978
Larder R, Kimura I, Meadows J, Clark DD, Mayo S, Mellon PL (2013) Gene dosage of Otx2 is
important for fertility in male mice. Mol Cell Endocrinol 377:16–22
Lawson MA, Buhain AR, Jovenal JC, Mellon PL (1998) Multiple factors interacting at the GATA
sites of the gonadotropin-releasing hormone neuron-specific enhancer regulate gene expression.
Mol Endocrinol 12:364–377
Lee JH, Miele ME, Hicks DJ, Phillips KK, Trent JM, Weissman BE, Welch DR (1996) KiSS-1, a
novel human malignant melanoma metastasis-suppressor gene. J Natl Cancer Inst
88:1731–1737
Lee SY, Lau AT, Jeong CH, Shim JH, Kim HG, Kim J, Bode AM, Dong Z (2010) Histone XH2AX
is required for Xenopus anterior neural development: critical role of threonine 16 phosphoryla-
tion. J Biol Chem 285:29525–29534
Li E, Bestor TH, Jaenisch R (1992) Targeted mutation of the DNA methyltransferase gene results in
embryonic lethality. Cell 69:915–926
Lomniczi A, Loche A, Castellano JM, Ronnekleiv OK, Bosch M, Kaidar G, Knoll JG, Wright H,
Pfeifer GP, Ojeda SR (2013) Epigenetic control of female puberty. Nat Neurosci 16:281–289
Lomniczi A, Wright H, Castellano JM, Matagne V, Toro CA, Ramaswamy S, Plant TM, Ojeda SR
(2015) Epigenetic regulation of puberty via zinc finger protein-mediated transcriptional repres-
sion. Nat Commun 6:10195
Matsuda KI, Mori H, Nugent BM, Pfaff DW, McCarthy MM, Kawata M (2011) Histone
deacetylation during brain development is essential for permanent masculinization of sexual
behavior. Endocrinology 152:2760–2767
Matsuda F, Nakatsukasa K, Suetomi Y, Naniwa Y, Ito D, Inoue N, Wakabayashi Y, Okamura H,
Maeda KI, Uenoyama Y, Tsukamura H, Ohkura S (2015) The luteinising hormone surge-
generating system is functional in male goats as in females: involvement of kisspeptin neurones
in the medial preoptic area. J Neuroendocrinol 27:57–65
Maurer JA, Wray S (1997) Luteinizing hormone-releasing hormone (LHRH) neurons maintained in
hypothalamic slice explant cultures exhibit a rapid LHRH mRNA turnover rate. J Neurosci
17:9481–9491
9 Epigenetic Regulation of the GnRH and Kiss1 Genes: Developmental Perspectives 261

McCarthy MM (1994) Molecular aspects of sexual differentiation of the rodent brain.

Psychoneuroendocrinology 19:415–427
Mehlmann LM (2005) Stops and starts in mammalian oocytes: recent advances in understanding the
regulation of meiotic arrest and oocyte maturation. Reproduction 30:791–799
Mellon PL, Windle JJ, Goldsmith PC, Padula CA, Roberts JL, Weiner RI (1990) Immortalization of
hypothalamic GnRH neurons by genetically targeted tumorigenesis. Neuron 5:1–10
Messager S, Chatzidaki EE, Ma D, Hendrick AG, Zahn D, Dixon J, Thresher RR, Malinge I,
Lomet D, Carlton MB, Colledge WH, Caraty A, Aparicio SA (2005) Kisspeptin directly
stimulates gonadotropin-releasing hormone release via G protein-coupled receptor 54. Proc
Natl Acad Sci U S A 102:1761–1766
Messina A, Langlet F, Chachlaki K, Roa J, Rasika S, Jouy N, Gallet S, Gaytan F, Parkash J, Tena-
Sempere M, Giacobini P, Prevot V (2016) A microRNA switch regulates the rise in hypotha-
lamic GnRH production before puberty. Nat Neurosci 19:835–844
Metzger E, Wissmann M, Yin N, Müller JM, Schneider R, Peters AH, Günther T, Buettner R,
Schüle R (2005) LSD1 demethylates repressive histone marks to promote androgen-receptor-
dependent transcription. Nature 437:436–439
Moore JP Jr, Wray S (2000) Luteinizing hormone-releasing hormone biosynthesis and secretion in
embryonic LHRH neurons. Endocrinology 141:4486–4495
Murray EK, Varnum MM, Fernandez JL, de Vries GJ, Forger NG (2011) Effects of neonatal
treatment with valproic acid on vasopressin immunoreactivity and olfactory behaviour in mice.
J Neuroendocrinol 10:906–914
Navarro VM, Castellano JM, Fernández-Fernández R, Tovar S, Roa J, Mayen A, Barreiro ML,
Casanueva FF, Aguilar E, Dieguez C, Pinilla L, Tena-Sempere M (2005a) Effects of KiSS-1
peptide, the natural ligand of GPR54, on follicle-stimulating hormone secretion in the rat.
Endocrinology 146:1689–1697
Navarro VM, Castellano JM, Fernández-Fernández R, Tovar S, Roa J, Mayen A, Nogueiras R,
Vazquez MJ, Barreiro ML, Magni P, Aguilar E, Dieguez C, Pinilla L, Tena-Sempere M (2005b)
Characterization of the potent luteinizing hormone-releasing activity of KiSS-1 peptide, the
natural ligand of GPR54. Endocrinology 146:156–163
Novaira HJ, Fadoju D, Diaczok D, Radovick S (2012) Genetic mechanisms mediating kisspeptin
regulation of GnRH gene expression. J Neurosci 32:17391–17400
Novaira HJ, Sonko ML, Radovick S (2016) Kisspeptin induces dynamic chromatin modifications to
control GnRH gene expression. Mol Neurobiol 53:3315–3325
Ohkura S, Takase K, Matsuyama S, Mogi K, Ichimaru T, Wakabayashi Y, Uenoyama Y, Mori Y,
Steiner RA, Tsukamura H, Maeda KI, Okamura H (2009) Gonadotrophin-releasing hormone
pulse generator activity in the hypothalamus of the goat. J Neuroendocrinol 21:813–821
Okano M, Xie S, Li E (1998) Cloning and characterization of a family of novel mammalian DNA
(cytosine-5) methyltransferases. Nat Genet 19:219–220
Okano M, Bell DW, Haber DA, Li E (1999) DNA methyltransferases Dnmt3a and Dnmt3b are
essential for de novo methylation and mammalian development. Cell 99:247–257
Ong CT, Corces VG (2014) CTCF: an architectural protein bridging genome topology and function.
Nat Rev Genet 15:234–246
Pan G, Tian S, Nie J, Yang C, Ruotti V, Wei H, Jonsdottir GA, Stewart R, Thomson JA (2007)
Whole-genome analysis of histone H3 lysine 4 and lysine 27 methylation in human embryonic
stem cells. Cell Stem Cell 1:299–312
Pasini D, Bracken AP, Hansen JB, Capillo M, Helin K (2007) The polycomb group protein Suz12 is
required for embryonic stem cell differentiation. Mol Cell Biol 27:3769–3779
Phoenix CH, Goy RW, Gerall AA, Young WC (1959) Organizing action of prenatally administered
testosterone propionate on the tissues mediating mating behavior in the female guinea pig.
Endocrinology 65:369–382
Poling MC, Kauffman AS (2013) Organizational and activational effects of sex steroids on
kisspeptin neuron development. Front Neuroendocrinol 34:3–17
262 J. R. Kurian and E. Terasawa

Provenzano C, Pascucci B, Lupari E, Civitareale D (2010) Large scale analysis of transcription

factor TTF-1/NKX2.1 target genes in GnRH secreting cell line GT1-7. Mol Cell Endocrinol
323:215–223
Puttabyatappa M, Padmanabhan V (2017) Prenatal testosterone programming of insulin resistance
in the female sheep. Adv Exp Med Biol 1043:575–596
Quanbeck C, Sherwood NM, Millar RP, Terasawa E (1997) Two populations of luteinizing
hormone-releasing hormone neurons in the forebrain of the rhesus macaque during embryonic
development. J Comp Neurol 380:293–309
Radovick S, Wray S, Lee E, Nicols DK, Nakayama Y, Weintraub BD, Westphal H, Cutler GB Jr,
Wondisford FE (1991a) Migratory arrest of gonadotropin-releasing hormone neurons in trans-
genic mice. Proc Natl Acad Sci U S A 88:3402–3406
Radovick S, Ticknor CM, Nakayama Y, Notides AC, Rahman A, Weintraub BD, Cutler GB Jr,
Wondisford FE (1991b) Evidence for direct estrogen regulation of the human gonadotropin-
releasing hormone gene. J Clin Invest 88:1649–1655
Radovick S, Wray S, Muglia L, Westphal H, Olsen B, Smith E, Patriquin E, Wondisford FE (1994)
Steroid hormone regulation and tissue-specific expression of the human GnRH gene in cell
culture and transgenic animals. Horm Behav 28:520–529
Ramaswamy S, Guerriero KA, Gibbs RB, Plant TM (2008) Structural interactions between
kisspeptin and GnRH neurons in the mediobasal hypothalamus of the male rhesus monkey
(Macaca mulatta) as revealed by double immunofluorescence and confocal microscopy. Endo-
crinology 149:4387–4395
Ramaswamy S, Seminara SB, Ali B, Ciofi P, Amin NA, Plant TM (2010) Neurokinin B stimulates
GnRH release in the male monkey (Macaca mulatta) and is colocalized with kisspeptin in the
arcuate nucleus. Endocrinology 151:4494–4503
Ramaswamy S, Dwarki K, Ali B, Gibbs RB, Plant TM (2013) The decline in pulsatile GnRH
release, as reflected by circulating LH concentrations, during the infant-juvenile transition in the
agonadal male rhesus monkey (Macaca mulatta) is associated with a reduction in kisspeptin
content of KNDy neurons of the arcuate nucleus in the hypothalamus. Endocrinology
154:1845–1853
Ramsahoye BH, Biniszkiewicz D, Lyko F, Clark V, Bird AP, Jaenisch R (2000) Non-CpG
methylation is prevalent in embryonic stem cells and may be mediated by DNA
methyltransferase 3a. Proc Natl Acad Sci U S A 97:5237–5242
Rave-Harel N, Miller NL, Givens ML, Mellon PL (2005) The Groucho-related gene family
regulates the gonadotropin-releasing hormone gene through interaction with the homeodomain
proteins MSX1 and OCT1. J Biol Chem 280:30975–30983
Rometo AM, Krajewski SJ, Voytko ML, Rance NE (2007) Hypertrophy and increased kisspeptin
gene expression in the hypothalamic infundibular nucleus of postmenopausal women and
ovariectomized monkeys. J Clin Endocrinol Metab 92:2744–2750
Ronnekleiv OK, Resko JA (1990) Ontogeny of gonadotropin-releasing hormone-containing
neurons in early fetal development of rhesus macaques. Endocrinology 126:498–511
Schermelleh L, Spada F, Easwaran HP, Zolghadr K, Margot JB, Cardoso MC, Leonhardt H (2005)
Trapped in action: direct visualization of DNA methyltransferase activity in living cells. Nat
Methods 2:751–756
Schulz KM, Sisk CL (2006) Pubertal hormones, the adolescent brain, and the maturation of social
behaviors: lessons from the Syrian hamster. Mol Cell Endocrinol 254–255:120–126
Schulz KM, Molenda-Figueira HA, Sisk CL (2009) Back to the future: the organizational-
activational hypothesis adapted to puberty and adolescence. Horm Behav 55:597–604
Schwarz JM, Nugent BM, McCarthy MM (2010) Developmental and hormone-induced epigenetic
changes to estrogen and progesterone receptor genes in brain are dynamic across the life span.
Endocrinology 151:4871–4881
Semaan SJ, Kauffman AS (2010) Sexual differentiation and development of forebrain reproductive
circuits. Curr Opin Neurobiol 20:424–431
9 Epigenetic Regulation of the GnRH and Kiss1 Genes: Developmental Perspectives 263

Semaan SJ, Kauffman AS (2015) Daily successive changes in reproductive gene expression and
neuronal activation in the brains of pubertal female mice. Mol Cell Endocrinol 401:84–97
Semaan SJ, Murray EK, Poling MC, Dhamija S, Forger NG, Kauffman AS (2010) BAX-dependent
and BAX-independent regulation of Kiss1 neuron development in mice. Endocrinology
151:5807–5817
Semaan SJ, Dhamija S, Kim J, Ku EC, Kauffman AS (2012) Assessment of epigenetic
contributions to sexually-dimorphic Kiss1 expression in the anteroventral periventricular
nucleus of mice. Endocrinology 153:1875–1886
Seminara SB, Messager S, Chatzidaki EE, Thresher RR, Acierno JS Jr, Shagoury JK, Bo-Abbas Y,
Kuohung W, Schwinof KM, Hendrick AG, Zahn D, Dixon J, Kaiser UB, Slaugenhaupt SA,
Gusella JF, O’Rahilly S, Carlton MB, Crowley WF Jr, Aparicio SA, Colledge WH (2003) The
GPR54 gene as a regulator of puberty. N Engl J Med 349:1614–1627
Shahab M, Mastronardi C, Seminara SB, Crowley WF, Ojeda SR, Plant TM (2005) Increased
hypothalamic GPR54 signaling: a potential mechanism for initiation of puberty in primates.
Proc Natl Acad Sci U S A 102:2129–2134
Shahab M, Lippincott M, Chan YM, Davies A, Merino PM, Plummer L, Mericq V, Seminara S
(2018) Discordance in the dependence on kisspeptin signaling in mini puberty vs adolescent
puberty: human genetic evidence. J Clin Endocrinol Metab 103:1273–1276
Shen X, Liu Y, Hsu YJ, Fujiwara Y, Kim J, Mao X, Yuan GC, Orkin SH (2008) EZH1 mediates
methylation on histone H3 lysine 27 and complements EZH2 in maintaining stem cell identity
and executing pluripotency. Mol Cell 32:491–502
Simerly RB (1998) Organization and regulation of sexually dimorphic neuroendocrine pathways.
Behav Brain Res 92:195–203
Simerly RB (2002) Wired for reproduction: organization and development of sexually dimorphic
circuits in the mammalian forebrain. Annu Rev Neurosci 25:507–536
Smith JT (2008) Kisspeptin signalling in the brain: steroid regulation in the rodent and ewe. Brain
Res Rev 57:288–298
Smith JT (2013) Sex steroid regulation of kisspeptin circuits. Adv Exp Med Biol 784:275–295
Smith JT, Cunningham MJ, Rissman EF, Clifton DK, Steiner RA (2005) Regulation of Kiss1 gene
expression in the brain of the female mouse. Endocrinology 146:3686–3692
Smith JT, Li Q, Pereira A, Clarke IJ (2009) Kisspeptin neurons in the ovine arcuate nucleus and
preoptic area are involved in the preovulatory luteinizing hormone surge. Endocrinology
150:5530–5538
Smith JT, Shahab M, Pereira A, Pau KY, Clarke IJ (2010) Hypothalamic expression of KISS1 and
gonadotropin inhibitory hormone genes during the menstrual cycle of a non-human primate.
Biol Reprod 83:568–577
Stresemann C, Lyko F (2008) Modes of action of the DNA methyltransferase inhibitors azacytidine
and decitabine. Int J Cancer 123:8–13
Suberbielle E, Sanchez PE, Kravitz AV, Wang X, Ho K, Eilertson K, Devidze N, Kreitzer AC,
Mucke L (2013) Physiologic brain activity causes DNA double-strand breaks in neurons, with
exacerbation by amyloid-β. Nat Neurosci 16:613–621
Svedruzic ZM (2008) Mammalian cytosine DNA methyltransferase Dnmt1: enzymatic mechanism,
novel mechanism-based inhibitors, and RNA-directed DNA methylation. Curr Med Chem
15:92–106
Terasawa E (2001) Luteinizing hormone-releasing hormone (LHRH) neurons: mechanism of
pulsatile LHRH release. Vitam Horm 63:91–129
Terasawa E, Quanbeck CD, Schulz CA, Burich AJ, Luchansky LL, Claude P (1993) A primary cell
culture system of luteinizing hormone releasing hormone (LHRH) neurons derived from fetal
olfactory placode in the rhesus monkey. Endocrinology 133:2379–2390
Terasawa E, Keen KL, Mogi K, Claude P (1999) Pulsatile release of luteinizing hormone-releasing
hormone (LHRH) in cultured LHRH neurons derived from the embryonic olfactory placode of
the rhesus monkey. Endocrinology 140:1432–1441
Terasawa E, Busser BW, Luchansky LL, Sherwood NM, Jennes L, Millar RP, Glucksman MJ,
Roberts JL (2001) Presence of luteinizing hormone-releasing hormone fragments in the rhesus
monkey forebrain. J Comp Neurol 439:491–504
264 J. R. Kurian and E. Terasawa

Terasawa E, Garcia JP, Seminara SB, Keen KL (2018) Role of kisspeptin and neurokinin B in
puberty in female non-humanprimates. Front Endocrinol (Lausanne) 9:148
Tomikawa J, Uenoyama Y, Ozawa M, Fukanuma T, Takase K, Goto T, Abe H, Ieda N, Minabe S,
Deura C, Inoue N, Sanbo M, Tomita K, Hirabayashi M, Tanaka S, Imamura T, Okamura H,
Maeda K, Tsukamura H (2012) Epigenetic regulation of Kiss1 gene expression mediating
estrogen-positive feedback action in the mouse brain. Proc Natl Acad Sci U S A 109:1294–1301
Topaloglu AK, Tello JA, Kotan LD, Ozbek MN, Yilmaz MB, Erdogan S, Gurbuz F, Temiz F,
Millar RP, Yuksel B (2012) Inactivating KISS1 mutation and hypogonadotropic hypogonadism.
N Engl J Med 366:629–635
Toro CA, Wright H, Aylwin CF, Ojeda SR, Lomniczi A (2018) Trithorax dependent changes in
chromatin landscape at enhancer and promoter regions drive female puberty. Nat Commun 9:57
Uenoyama Y, Tomikawa J, Inoue N, Goto T, Minabe S, Ieda N, Nakamura S, Watanabe Y,
Ikegami K, Matsuda F, Ohkura S, Maeda K, Tsukamura H (2016) Molecular and epigenetic
mechanism regulating hypothalamic Kiss1 gene expression in mammals. Neuroendocrinology
103:640–649
Vargas Trujillo M, Kalil B, Ramaswamy S, Plant TM (2017) Estradiol upregulates kisspeptin
expression in the preoptic area of both the male and female Rhesus monkey (Macaca mulatta):
implications for the hypothalamic control of ovulation in highly evolved primates. Neuroendo-
crinology 105:77–89
Wallen K (2005) Hormonal influences on sexually differentiated behavior in nonhuman primates.
Front Neuroendocrinol 26:7–26
Wallen K (2009) The organizational hypothesis: reflections on the 50th anniversary of the publica-
tion of Phoenix, Goy, Gerall, and Young (1959). Horm Behav 55:561–565
Watanabe Y, Uenoyama Y, Suzuki J, Takase K, Suetomi Y, Ohkura S, Inoue N, Maeda KI,
Tsukamura H (2014) Oestrogen-induced activation of preoptic kisspeptin neurones may be
involved in the luteinising hormone surge in male and female Japanese monkeys.
J Neuroendocrinol 26:909–917
Whyte DB, Lawson MA, Belsham DD, Eraly SA, Bond CT, Adelman JP, Mellon PL (1995) A
neuron-specific enhancer targets expression of the gonadotropin-releasing hormone gene to
hypothalamic neurosecretory neurons. Mol Endocrinol 9:467–477
Wu SC, Zhang Y (2010) Active DNA demethylation: many roads lead to Rome. Nat Rev Mol Cell
Biol 11:607–620
Xie H, Hoffmann HM, Meadows JD, Mayo SL, Trang C, Leming SS, Maruggi C, Davis SW,
Larder R, Mellon PL (2015) Homeodomain proteins six3 and six6 regulate gonadotrope-specific
genes during pituitary development. Mol Endocrinol 29:842–855
Yagi H, Deguchi K, Aono A, Tani Y, Kishimoto T, Komori T (1998) Growth disturbance in fetal
liver hematopoiesis of Mll-mutant mice. Blood 92:108–117
Yamaguchi S, Hong K, Liu R, Shen L, Inoue A, Diep D, Zhang K, Zhang Y (2012) Tet1 controls
meiosis by regulating meiotic gene expression. Nature 492:443–447
Yang H, Lin H, Xu H, Zhang L, Cheng L, Wen B, Shou J, Guan K, Xiong Y, Ye D (2014)
TET-catalyzed 5-methylcytosine hydroxylation is dynamically regulated by metabolites. Cell
Res 24:1017–1020
Yeo SH, Clarkson J, Herbison AE (2014) Kisspeptin-gpr54 signaling at the GnRH neuron is
necessary for negative feedback regulation of luteinizing hormone secretion in female mice.
Neuroendocrinology 100:191–197
Yu BD, Hess JL, Horning SE, Brown GA, Korsmeyer SJ (1995) Altered Hox expression and
segmental identity in Mll-mutant mice. Nature 378:505–508
Imprinted Genes and Hypothalamic
Function 10
Michela Pulix and Antonius Plagge

Abstract
Genomic imprinting, a specific type of inherited epigenetic regulation, controls a
number of genes with important functions in the hypothalamus and its associated
neuroendocrinology. Imprinted genes are expressed monoallelically, dependent
on the parental origin of the allele, whereby maternally or paternally inherited
DNA methylation marks determine gene activity. In this chapter, we provide an
overview of imprinted genes with roles in the hypothalamic regulation of whole-
body energy homeostasis, neuroendocrine hormone functions, and circadian
rhythmicity. As an example for imprinted genes that impact on the central
regulation of energy balance, we describe the Gnas locus in more detail, since
it generates parental allele-specific gene products with antagonistic roles in
energy expenditure and sympathetic nervous system activity. Several human
neuroendocrine disorders are caused by defects in imprinted genes as described
here for Prader–Willi Syndrome and Angelman Syndrome. Furthermore, we
outline methodological approaches for the investigation of DNA methylation
marks and allelic gene expression, including bisulfite conversion of DNA and
pyrosequencing. Recent technological advances to resolve methylation and alle-
lic expression at the single-cell level are introduced.

M. Pulix
Cellular and Molecular Physiology, Institute of Translational Medicine, University of Liverpool,
Liverpool, UK
Department of Neuroscience, Sheffield Institute for Translational Neuroscience, University of
Sheffield, Sheffield, UK
A. Plagge (*)
Cellular and Molecular Physiology, Institute of Translational Medicine, University of Liverpool,
Liverpool, UK
e-mail: [email protected]

# Springer Nature Switzerland AG 2020 265

S. Wray, S. Blackshaw (eds.), Developmental Neuroendocrinology, Masterclass in
Neuroendocrinology 9, https://doi.org/10.1007/978-3-030-40002-6_10
266 M. Pulix and A. Plagge

Keywords
Genomic imprinting · Hypothalamus · Gnas · Energy homeostasis · Prader–Willi
syndrome · Angelman syndrome · DNA methylation · Bisulfite conversion of
DNA · Pyrosequencing

10.1 Introduction

The term genomic imprinting refers to the expression of certain genes from only one
of the two parental alleles. This is dependent on the parental origin, i.e., some genes
are monoallelically expressed from the maternal, others from the paternal allele,
while the opposite alleles are silenced. The allelic silencing is already determined in
the parental germ cells through epigenetic “marks” and maintained in the somatic
tissues of the offspring (Fig. 10.1). To maintain this parent-of-origin-specific gene
expression, the epigenetic “marks” are erased in the primordial germline cells of the
developing offspring and reset according to the sex of the offspring. This process
ensures that the maternal or paternal allelic expression, respectively, of an imprinted
gene is inherited from generation to generation. To date, circa 150 imprinted genes
have been confirmed in humans and mice, but imprinted gene expression also exists
in other mammals (http://www.mousebook.org/imprinting-gene-list; http://igc.
otago.ac.nz/home.html). Most imprinted genes occur in conserved clusters, but the
imprinting of some genes is species specific.
Many imprinted genes have essential roles in embryonic development and pla-
cental functions (Plasschaert and Bartolomei 2014). However, they also play impor-
tant roles in the brain, including regulation of neurogenesis, synaptic transmission,
and behavior (Perez et al. 2015; Peters 2014). Within the brain, the hypothalamus
has proven to be the main site of expression of imprinted genes, and from studies of
knockout mouse models as well as human inherited disorders, it is evident that they
influence neuroendocrine pathways, energy homeostasis, circadian rhythm, and
social interactions (Ivanova and Kelsey 2011; Perez et al. 2015). In this chapter,
we provide an overview of the hypothalamic functions of imprinted genes and
highlight a few disease-relevant examples. In addition, we discuss techniques used
in the identification and epigenetic analysis of imprinted genes and indicate currently
open questions and trends in the field.

10.2 Main Features of Imprinted Gene Clusters

The importance of imprinting in mammals became evident from embryological

studies, in which pro-nuclei of newly fertilized mouse eggs were transplanted in
order to generate a genetically diploid bi-maternal (parthenogenetic) or a bi-paternal
(androgenetic) conceptus. The embryos that developed from these concepti died in
utero and showed growth aberration in embryonic and extraembryonic tissues
(Ferguson-Smith 2011). These observations demonstrated that a diploid genome as
10 Imprinted Genes and Hypothalamic Function 267

Maintenance of germline-derived DNA

methylation imprinting marks in somatic cells
De novo
establishment of DNA
methylation
imprinting marks in
germline cells
according to sex (pre-
natally in males; post-
natally in females)

Erasure of DNA
methylation
imprinting marks in
Oocyte primordial germ cells
of embryo
Sperm Embryo

Maintenance of
germline-derived DNA
methylation imprinting
marks during genome-
wide demethylation
events in early embryo

Zygote

Blastocyst

Fig. 10.1 Inheritance cycle of DNA methylation imprinting marks through the germline and
maintenance in somatic tissues. The genomic DNA methylation imprints are established in the
germline cells in a sex-specific manner and transmitted to the zygote, where they resist
reprogramming events and global changes in DNA methylation, which take place after fertilization.
While the genomic imprinting marks are maintained in somatic cells throughout life, they are erased
in the primordial germ cells of the developing embryo and re-established de novo according to the
sex of the organism, to be transmitted to the next generation. Re-establishment of imprinting marks
in gametogenesis occurs prenatally in males and at postnatal stages in females [Adapted from
Bartolomei and Ferguson-Smith (2011)]

such is not sufficient for normal mammalian development, that the two parental
genomes make different functional contributions to embryonic development and that
there must be some kind of “mark,” which mediates the distinct functions of the
maternal and paternal genomes, respectively. Thus, genomic imprinting became a
paradigm that drove research into “epigenetic” mechanisms (Ferguson-Smith 2011).
We know today that those epigenetic “marks” represent methylation of genomic
DNA, histone modifications, and noncoding RNA expression.
In the majority of cases, imprinted genes are typically found in clusters of 3–12
genes spread over 20 Kb–3.7 Mb of DNA (Ferguson-Smith 2011; Peters, 2014) and
268 M. Pulix and A. Plagge

they are usually regulated by the presence of DNA methylation on one parental allele
in so-called Differentially Methylated Regions (DMRs). These regions are
characterized by a high density of CpG dinucleotides (above 55%), termed CpG
islands (CGIs). Here, methylation of cytosines occurs at the C5 position of the
pyrimidine ring and is catalyzed by DNA methyltransferases (DNMTs) (Lyko
2018). A DMR is termed an Imprinting Control Region (ICR), if the differential
methylation is established already in the germ cells and if functional studies have
proven that it controls the imprinting of associated genes within a cluster. Thus, ICRs
regulate gene expression and chromatin structure over long distances (Ferguson-
Smith 2011). Maternally methylated DMRs and ICRs are usually located at
promoters while paternally methylated DMRs and ICRs are located in intergenic
regions (Bartolomei and Ferguson-Smith 2011).
DNA methylation is established during gametogenesis, resists the extensive
genomic reprogramming that occurs after fertilization and is maintained in the
offspring throughout life (Fig. 10.1). For inheritance of imprinted gene expression
to the next generation, the primordial germ cells of an embryo undergo a phase of
erasure of DNA methylation, followed by re-establishment of methylation at DMRs/
ICRs according to the sex of the individual and mediated by DNMT3A and the
accessory protein DNMT3L (Bartolomei and Ferguson-Smith 2011; Plasschaert and
Bartolomei 2014). How the ICRs are recognized within the genome is still unknown,
but it is clear that several factors are involved in the maintenance of DNA methyla-
tion, including ZINC FINGER PROTEIN HOMOLOG 57 (ZFP57) and Develop-
mental pluripotency-associated 3 (Dppa3). Mutations in these genes result in loss of
imprinting at several loci during early embryogenesis, which in turn causes a range
of disease symptoms (Ferguson-Smith 2011; Plasschaert and Bartolomei 2014).
Before discussing techniques and methods used in the analysis of allelic expres-
sion and epigenetic regulation of imprinted genes, we first provide an overview
about the functional roles of imprinted genes in the hypothalamus and, where
applicable, their involvement in human neurodevelopmental disorders. Then, in
the context of the Prader–Willi syndrome (PWS)/Angelman syndrome (AS) imprint-
ing cluster and the Gnas locus, we discuss the structure and organization of regu-
latory elements of imprinted genes.

10.3 Functions of Imprinted Genes in the Hypothalamus

When the first imprinted genes (Igf2, Igf2r, and H19) were identified, they were
found to affect mainly embryonic growth, placental function, and nutrient supply
and demand at the maternal–fetal interface (Ferguson-Smith 2011; Peters 2014;
Plasschaert and Bartolomei 2014). However, it soon became apparent that a number
of imprinted genes were expressed in defined brain regions, and many of them
showed high expression levels specifically in the hypothalamus (Ivanova and Kelsey
2011). Phenotype analyses of knockout mouse models and clinical case studies of
human imprinting disorders revealed that many imprinted genes continue to impact
growth and whole-body energy homeostasis beyond embryonic development and
10 Imprinted Genes and Hypothalamic Function 269

into adulthood via influencing neuroendocrine pathways and the autonomic nervous
system. Other core functions of the hypothalamus, including circadian rhythms,
reproductive functions, and maternal-infant care behavior are also deregulated
upon the deficiency of some imprinted genes. We will first describe the hypotha-
lamic roles of the Gnas and the PWS-AS loci since they have been analyzed in some
detail, before summarizing findings related to other imprinted genes.

10.3.1 Hypothalamic Regulation of Energy Homeostasis by Gnas

and Gnasxl

The Gnas imprinted locus on mouse chromosome 2 and human chromosome

20 comprises a complex arrangement of protein-coding and noncoding transcripts,
alternative promoters and epigenetically controlled regulatory elements (Fig. 10.2a)
(Chen et al. 2011; Peters 2014). The two protein-coding transcripts Gnas and Gnasxl
derive from different first exons, but share exons 2–12. As the open reading frames
are conserved, the two proteins Gαs and XLαs only differ at their NH2-termini, thus
forming alternative versions of the G-protein α-stimulatory subunit. Both can medi-
ate signal transduction from seven transmembrane receptors to adenylate cyclase and
cAMP formation, although XLαs can also stimulate the inositol 1,4,5-trisphosphate
(IP3)/protein kinase C (PKC) second messenger pathway in some cell types
(He et al. 2015). While Gnasxl is exclusively expressed from the paternal allele,
Gnas is biallelically expressed in most tissues, but monoallelically transcribed from
the maternal allele in specific cell types of the brain, renal proximal tubules, thyroid
gland, pituitary somatotroph cells and the ovary (Chen et al. 2011). The third
protein-coding transcript, maternal allele-derived Nesp, generates a neuroendocrine
secretory protein (Nesp55) from an open reading frame that is confined to its
upstream exon, although the RNA is spliced onto downstream exons 2–12 of
Gnas. The regions of differential DNA methylation (ICRs at Nespas-Gnasxl and
exon-1a; DMR at Nesp), the noncoding transcripts Nespas and exon-1a, as well as
the Nesp transcript have a role in the establishment and regulation of genomic
imprinting at the cluster (Plagge 2012). Both ICRs become methylated in the
maternal germ cells, and the methylation on the maternal allele is maintained in
somatic tissues of the offspring. The ICR at Nespas-Gnasxl controls the imprinting
of all transcripts of the locus, while the ICR at exon-1a only regulates the
monoallelic expression of Gnas in those specific tissues mentioned above. The
DNA methylation on the maternal allele inhibits the transcription of Gnasxl, Nespas,
and exon-1a. On the paternal allele, the unmethylated ICRs serve as promoters for
the noncoding RNAs Nespas and exon-1a, which silence in cis the expression of
Nesp and Gnas via unclear mechanisms that involve histone modifications (Mehta
et al. 2015).
Hypothalamic functions have been shown for Gnas and Gnasxl, while behavioral
phenotypes of Nesp knockout mice could not be specifically attributed to a role in
this brain region, although Nesp is expressed in the hypothalamus (Plagge et al.
2005). In humans, GNAS and GNASXL mutations or epigenetic deregulation cause
270 M. Pulix and A. Plagge

A) Nesp55

Gαs

MMMMMMMMM MMMM
Maternal
allele:

Nesp Nespas Gnasxl Exon 1A / Gnas shared Gnas

EXON A/B exon 1 coding exons

Paternal
allele:
MMMMM

( Gαs
)
XLαs

B)

MM MM MMMMM
Maternal
allele:

Mkrn3 Ndn Snrpn /

U-exons / Snord Snord Ube3a
Snurf /
Peg12 Magel2 NPAP1 AS-IC 116 115
PWS-IC
(Frat3)
Paternal
allele:

Fig. 10.2 Simplified schemes of the Gnas and the PWS-AS (Snrpn-Ube3a) imprinted loci. The
features of the maternal and paternal alleles are indicated in red and blue, respectively. Genes or
transcripts are named in the central part. Arrows mark transcription start sites and undulating lines
noncoding RNAs. Open and filled boxes represent noncoding and coding exons, respectively, while
black boxes indicate silenced genes. Differentially methylated regions of DNA (DMRs) are marked
by MMM. (a) For the Gnas locus, the alternatively spliced coding transcripts and proteins are
named above and below the alleles. DMRs at Nespas and Exon 1A (EXON A/B in human) are
established in the maternal germline, while methylation at Nesp occurs during early embryonic
development. Gnas is expressed biallelically, but is silenced on the paternal allele in some tissues
(hatched box). Nesp represents a coding transcript (ORF limited to Nesp exon 2). Nespas is
expressed from the unmethylated imprinting control region (ICR) of the locus. (b) The PWS-AS-
associated Snrpn-Ube3a imprinting cluster contains several genes (represented by single-exon
boxes) that show monoallelic expression in the brain. Ube3a is biallelically expressed in most
tissues, but silenced on the paternal allele in neurons (hatched box). The Snrpn long noncoding
RNA occurs in multiple, variably processed forms, including brain-specific variants that overlap
with Ube3a (interrupted undulating lines). Snord115 and Snord116 represent clusters of C/D box
snoRNAs, which are generated from the Snrpn transcript. The bipartite ICR of the human locus is
indicated as AS-IC and PWS-IC, the latter being conserved in mouse around the main Snrpn start
site. NPAP1 is not present in the murine locus, but Peg12 is imprinted in mice. Somatic DMRs at
Ndn and Mkrn3 are established during development
10 Imprinted Genes and Hypothalamic Function 271

complex neuroendocrine disorders termed Albright’s Hereditary Osteodystrophy

(AHO) and Pseudohypoparathyroidism (PHP), the symptoms of which are caused
by hormone resistance in multiple tissues, including disruption of parathyroid
hormone signaling in the kidney (Table 10.1) (Mantovani et al. 2018; Turan and
Bastepe 2013). Hypothalamus-associated symptoms are related to postnatal growth,
feeding, and regulation of energy balance at the adult stage. Homozygous mutations
of Gnas in mice are embryonic lethal and heterozygous mutations show different
phenotypes, dependent on whether they were maternally or paternally inherited,
whereby the mouse phenotypes largely reproduce the human disease symptoms. Due
to genomic imprinting, maternally inherited mutations in shared exons only affect
the Gnas transcript, but in those tissues, in which it is imprinted, Gαs function will be
completely lost, while a 50% reduction occurs in cells with biallelic expression. On
the other hand, paternally inherited mutations in shared exons will result in loss of
XLαs and a 50% reduction of Gαs in cells with biallelic expression.
Gnas and Gnasxl transcript- and tissue-specific knockout mouse models have
shed some light onto the functions of the two proteins in the hypothalamus and
revealed opposite or antagonistic impacts of Gαs and XLαs on energy balance and
sympathetic nervous system (SNS) activity (Weinstein 2014). Conditional gene
targeting of the Gnas-specific exon 1 and its brain-specific deletion (via Nestin-
Cre strain) on the maternally inherited chromosome causes severe obesity due to
decreased energy expenditure and O2 consumption, lower locomotor activity,
reduced SNS outflow, body temperature, and heart rate (Table 10.1). Glucose
intolerance and insulin resistance were also observed. These symptoms are, how-
ever, not found in the paternally inherited, brain-specific knockout, which show
normal energy homeostasis. These findings indicate genomic imprinting in the brain.
The response to stimulation of the melanocortin receptors (MC3R, MC4R), which
mediate the anorexigenic response downstream of leptin signaling, was found to be
blunted in these mice. However, the maternally inherited Gnas mutation specifically
disrupts the energy expenditure-stimulating functions of melanocortins, while their
food intake-reducing effects remained intact (Chen et al. 2009). The mechanism
(s) underlying these responses are unclear. The decreased energy expenditure in the
mutants has been linked to reduced SNS outflow as levels of norepinephrine and
uncoupling protein 1 (UCP1) in brown adipose tissue (BAT), as well as heart rate
and blood pressure are reduced.
Monoallelic maternal expression of Gnas is found in the paraventricular nucleus
(PVN) of the hypothalamus (Chen et al. 2009). A specific maternal allele knockout
of Gnas in the PVN, using a Sim1-Cre strain, reproduced a much milder obesity
phenotype than observed with the Nestin-Cre line described above suggesting that
imprinted Gαs expression in additional brain regions contributes to the earlier
observed phenotype (Chen et al. 2012). However, it does reproduce the lower
blood pressure and heart rate of the whole-brain knockout. Maternal allele-specific
Gαs is also required in the PVN to mediate melanocortin-induced blood pressure
increases (Li et al. 2016). In contrast to the brain knockout, the SNS stimulation of
BAT, cold-induced thermoregulation as well as the melanocortin induction of
energy expenditure are intact in the PVN-specific Gnas knockout, indicating that
these physiological responses are mediated by other neurons (Chen et al. 2012).
Table 10.1 Functions of imprinted genes in the hypothalamus and related human disorders
272

Hypothalamus-related mouse knockout

Gene Protein/RNA function Expressed allele phenotype Human disorder
Gnas G-protein α-stimulatory subunit Biallelic in most tissues; Maternally inherited mutation: obesity, Albright’s Hereditary
(Gαs); (intracellular signaling) maternal allele in specific reduced energy expenditure, reduced Osteodystrophy (AHO) and
cell types: PVN, DMH, and SNS activity, impaired melanocortin Pseudohypoparathyroidism
subset of MC4R-expressing receptor-mediated stimulation of energy (PHP)
neurons in the brain, expenditure and blood pressure, reduced Hypothalamus-associated
proximal renal tubules, blood pressure and heart rate, decreased symptoms when maternally
thyroid gland, pituitary locomotor activity, glucose intolerance, inherited: obesity, reduced
somatotroph cells, and ovary insulin resistance SNS outflow
Paternally inherited mutation: normal Symptoms related to other
energy homeostasis tissues when maternally
inherited: renal parathyroid
hormone resistance,
hypocalcemia,
hyperphosphatemia, TSH
resistance, GHRH resistance
Parent-of-origin independent
symptoms: short stature,
brachydactyly, heterotopic
ossifications, and other variable
features
Gnasxl Extra large G-protein Paternal Postnatally: growth retardation, Pre- and neonatal symptoms
α-stimulatory subunit (XLαs); hypoglycemia, reduced milk intake, Small for gestational
(intracellular signaling) reduced lipid stores in adipose tissue age/intrauterine growth
Adult stage: decreased adiposity and retardation, low birth weight,
body weight, elevated SNS activity, postnatal growth retardation,
elevated metabolic rate and body feeding problems, low body
temperature, increased blood pressure mass index during childhood
and heart rate, increased glucose
tolerance, and insulin sensitivity
M. Pulix and A. Plagge
 10

Magel2 MAGE family ubiquitin ligase Paternal Postnatally: 10–50% mortality; Prader–Willi syndrome
regulator (endosomal protein diminished oxytocin-induced suckling
sorting and recycling) initiation after birth; reduced neonatal
hypothalamic levels of oxytocin,
arginine-vasopressin, and orexin-A
Adult stage: increased fat mass despite
reduced food intake and orexin-A levels,
reduced lean mass; hypothalamic leptin
resistance; reduced POMC neurons and
projections; rhythmic circadian
expression in SCN and altered/reduced
circadian activity patterns; and abnormal
behavior
Ndn MAGE family ubiquitin ligase Paternal High rate of early neonatal mortality due Prader–Willi syndrome
(Necdin) regulator to respiratory deficiencies (brainstem
Imprinted Genes and Hypothalamic Function

related); fewer hypothalamic oxytocin

and gonadotropin-releasing hormone
neurons; reduced thyrotropin-releasing
hormone levels in PVN and
hypothyroidism; abnormal axon bundles;
and behavioral abnormalities
Snord116 Box C/D small nucleolar RNA Paternal Postnatal and adult growth retardation; Prader–Willi syndrome
reduced adiposity; increased energy
expenditure; resistance to high-fat diet-
induced obesity; improved body
temperature maintenance at 4
C;
behavioral abnormalities
Adult-onset deletion in mediobasal
hypothalamus: hyperphagia and obesity
(continued)
273
Table 10.1 (continued)
274

Hypothalamus-related mouse knockout

Gene Protein/RNA function Expressed allele phenotype Human disorder
Ube3a Ubiquitin-protein ligase E3A; Biallelic in most tissues; Sleep disturbances, circadian rhythm Angelman syndrome
(protein degradation) maternal allele in brain regulation via Bmal1
Dlk1 Protein delta homolog Paternal Hypothalamus-related functions Central precocious puberty;
1, (Trans-membrane and uncertain, potentially involving postnatal increased fat mass (Temple
secreted protein with similarity growth retardation syndrome)
to Notch ligands)
Dio3 Thyroxine 5-deiodinase Paternal (e.g., neonatal Postnatal and adult growth retardation; Potentially involved in Temple
(thyroid hormone catabolism) hypothalamus, tissue- and reduced adiposity; abnormal thyroid syndrome
stage-specific) hormone regulation and hypothalamus–
pituitary–thyroid axis; increased energy
expenditure and locomotion; changes in
hypothalamic gene expression of leptin-
melanocortin, oxytocin, and vasopressin
pathways; abnormal circadian activity;
impaired viability and fertility
Kcnq1 Potassium voltage-gated Maternal No hypothalamus-related phenotype Growth hormone deficiency
channel subfamily KQT described, but range of other symptoms and gingival fibromatosis, QT
member 1 syndrome, auditory and cardiac
symptoms
Peg3 Zinc finger DNA-binding Paternal Reduced body weight throughout life;
protein increased adiposity; leptin resistance;
reduced metabolic rate; lower body
temperature; altered hypothalamic
neuropeptide expression
M. Pulix and A. Plagge
10 Imprinted Genes and Hypothalamic Function 275

Further investigations by the group of L. S. Weinstein excluded the ventromedial

hypothalamus (VMH) and identified the dorsomedial hypothalamus (DMH) as a
major site of imprinted Gαs regulation of energy expenditure (Chen et al. 2017).
Adenovirus-mediated Cre-deletion of Gnas on the maternal allele in the DMH
resulted in a number of changes including obesity, reduced energy expenditure,
heart rate, and locomotor activity, as well as a blunted response to melanocortin-
induced O2 consumption. Baseline activation of BAT was reduced as seen in whole-
brain knockout mice, but cold-induced activation of BAT and maintenance of body
temperature were intact, as was food intake (Chen et al. 2017). Blood pressure was
normal in the DMH-specific deletion of Gαs, in line with previous findings
attributing this Gnas function to the PVN. However, when Gnas was deleted on
the maternal allele in the DMH, stimulation of MC3/4R in the DMH did not lead to
phosphorylation of cAMP response element-binding protein (CREB) (Chen et al.
2017). These findings are consistent with maternal allele-specific expression of Gαs
in the DMH, which is required for basic and melanocortin-induced stimulation of
energy expenditure.
Recently, the importance of Gnas for mediating the melanocortin effects on
energy homeostasis was confirmed through a specific knockout in MC4R-expressing
neurons (Podyma et al. 2018). Homozygous deletion of Gnas in MC4R neurons
results in severe obesity, while maternal allele-specific deficiency leads to a milder
phenotype. Paternal heterozygous knockouts showed normal energy balance.
Reduced energy expenditure and locomotor activity also featured in homozygous
Gnas MC4R-specific knockouts, but the major contributor to the obesity in these
mice was increased food intake, which was not observed in the other maternal allele-
specific hypothalamus knockouts described above (Podyma et al. 2018). The milder
obesity in mice with maternal allele-specific disruption of Gnas in MC4R neurons
was attributable to reduced energy expenditure with food intake being unchanged.
Furthermore, while homozygous knockouts could not maintain body temperature in
a cold environment, heterozygous maternal knockouts of Gnas in MC4R neurons
showed thermoregulation in line with previous observations in PVN- and
DMH-specific deletions (Chen et al. 2012, 2017; Podyma et al. 2018).
In summary, stimulation of SNS outflow appears to be at the heart of the
metabolic functions regulated by Gαs expression from the maternal allele. But it is
also apparent that Gnas expression from the maternal allele regulates various
branches of the SNS, affecting cardiovascular phenotypes via the PVN and energy
expenditure via the DMH, which projects to the raphe pallidus of the medulla and
other SNS-relevant brain areas. Its pathway of regulating cold-induced thermogene-
sis, however, remains as yet unknown (Chen et al. 2012, 2017; Podyma et al. 2018).
Although disruption of MC3/4R signaling contributes substantially to the energy
homeostasis phenotypes of central nervous system-specific Gnas knockouts, it is
currently unclear whether Gαs signaling from other receptors might also contribute.
In contrast to Gnas, the Gnasxl knockout mouse phenotype, as well as case
reports in human, indicate that the extra-large variant XLαs has overtly antagonistic
functions to Gαs in the regulation of energy homeostasis and SNS activity
(Weinstein 2014). Gnasxl is expressed in all major hypothalamic areas (apart from
the VMH), the suprachiasmatic nucleus (SCN), preoptic area, and SNS-associated
276 M. Pulix and A. Plagge

nuclei in the medulla, e.g., the raphe pallidus, nucleus of the solitary tract, as well as
the intermediolateral layer of the spinal cord (Krechowec et al. 2012; Plagge et al.
2004). Expression of Gnasxl in peripheral tissues, for example, muscle tissues,
becomes silenced postnatally with only a few exemptions like the adrenal medulla
(Krechowec et al. 2012; Xie et al. 2006). Newborn XLαs-deficient pups become
growth retarded, take in less milk and display hypoglycemia and lack of lipid
reserves in adipose tissue, all of which results in a high mortality rate in the first
week (Table 10.1; Plagge et al. 2004; Xie et al. 2006). Those Gnasxl knockout mice
that survive to weaning age, develop into fertile adults, but remain underweight and
lean throughout life, despite relatively increased food intake. Furthermore, their
phenotype consists of increased energy expenditure, metabolic rate, body tempera-
ture, blood pressure, and heart rate, which is associated with elevated SNS outflow
(Nunn et al. 2013; Xie et al. 2006). They also show improved glucose tolerance and
insulin sensitivity (Xie et al. 2006). The precise neural pathways through which
Gnasxl regulates SNS outflow and energy homeostasis remain unresolved, as are the
neural receptors that signal via XLαs.
Human disease symptoms related to loss of GNASXL function have been difficult
to ascertain since mutations that specifically disable the first GNASXL exon have not
been identified. However, paternal transmission of mutations in shared downstream
exons 2–13 and maternal uniparental disomies that encompass the GNAS locus on
chromosome 20, which result in two maternal alleles and lack of a paternally inherited
allele, indicate pre-/neonatal symptoms (Table 10.1). These include intrauterine
growth retardation, low birth weight, postnatal growth retardation, and feeding
problems (Genevieve et al. 2005; Kawashima et al. 2018; Mulchandani et al. 2016;
Richard et al. 2013). Although no adult metabolic data of such patients have been
reported, the neonatal symptoms are reminiscent of the Gnasxl knockout mouse
phenotype (Plagge et al. 2004). Taken together, these findings show that XLαs exerts
an important role in neonatal physiology in addition to its regulation of adult energy
homeostasis, which to date has only been confirmed in mice.
In adults, XLαs acts in the brain to inhibit SNS outflow and energy expenditure,
while its variant Gαs functions to stimulate SNS activity and metabolic rate. These
antagonistic roles are correlated on the epigenetic level with genomic imprinting and
monoallelic transcription from opposite parental alleles. Gnasxl is expressed exclu-
sively from the paternal allele, while Gnas is only imprinted in specific hypothalamic
neurons, where it is transcribed from the maternal allele. Why epigenetic regulation
through genomic imprinting of crucial growth and energy homeostasis functions
evolved in mammals is beyond the scope of this chapter, but has been extensively
discussed in the literature (Haig 2004; Peters 2014).

10.3.2 Hypothalamic Functions of the Prader–Willi Syndrome (PWS)

and Angelman Syndrome (AS) Imprinting Cluster

Another example of a cluster of imprinted genes with functions in the hypothalamus

is the Prader–Willi syndrome and Angelman syndrome locus on human
10 Imprinted Genes and Hypothalamic Function 277

chromosome 15 (conserved on mouse chromosome 7) (Buiting et al. 2016; Cassidy

et al. 2012; Peters 2014). The locus comprises five genes with paternal allele-specific
expression (MKRN3, MAGEL2, NDN, NPAP1, and SNURF/SNRPN/SNORD115-
116), and one with maternal allele-specific transcription (UBE3A), although the
murine locus contains paternally expressed Peg12 instead of NPAP1 (Fig. 10.2b).
Loss of function of the paternally expressed genes causes PWS, while loss of
function of the maternally expressed UBE3A, which is imprinted in the brain only,
causes AS. The monoallelic expression of all genes in the cluster is controlled by a
bipartite ICR that encompasses the promoter of the Snurf/Snrpn/Snord115–116 gene
(PWS-IC) and additional sequences further upstream (AS-IC). The PWS-IC is
methylated on the maternal allele, thus silencing the expression of the long Snurf/
Snrpn/Snord115–116 RNA. In neurons, this paternally expressed transcript, which
is also known as Ube3a antisense and can be processed in different ways, extends
into Ube3a exons and is thought to inhibit the expression of Ube3a in cis through
transcriptional interference, thus resulting in intact Ube3a expression only from the
maternal allele (Buiting et al. 2016; LaSalle et al. 2015). On the paternal allele, the
PWS-IC remains unmethylated and has an activation function for all other paternally
expressed genes of the cluster, which involves physical interactions via chromosome
looping (Plagge 2012).
Angelman syndrome constitutes a neurodevelopmental disorder that is mainly
characterized by intellectual disability, microcephaly, speech impairment, happy
demeanor, EEG abnormalities and sleep disturbances (Buiting et al. 2016). Func-
tionally, Ube3a (Ubiquitin-protein ligase 3a) is involved in labeling proteins for
degradation and lack of the enzyme impairs synapse formation and remodeling.
Little is known about hypothalamus-specific roles of Ube3a in the context of AS, but
knockout mice show abnormalities in sleep and circadian rhythm, potentially due to
impaired Ube3a-mediated degradation of the circadian clock protein Bmal1 in the
hypothalamus (Ehlen et al. 2015; Shi et al. 2015). On the other hand, main symptoms
of PWS are postnatal hypotonia, poor suckling, failure to thrive, respiratory
disturbances, hypogonadism, childhood-onset obesity, hyperphagia, cognitive and
behavioral abnormalities, short stature, and growth hormone insufficiency (Cassidy
et al. 2012). Knockout mouse models indicate that all paternally expressed genes
contribute to the range of these symptoms.
The functions of Magel2 (Mage-like protein 2) have been characterized, and it is
highly expressed in many areas of the hypothalamus (Tacer and Potts 2017). On a
molecular level, Magel2 acts as a regulator of RING E3 ubiquitin ligases within a
complex that controls endosomal protein sorting. Deficiency of Magel2 in a hypo-
thalamic cell line impairs protein recycling to the plasma membrane (Tacer and Potts
2017). The physiological functions of Magel2 are indicated by knockout mice,
which show 10–50% pre-weaning mortality (Table 10.1) (Kozlov et al. 2007;
Schaller et al. 2010). This mortality occurs in the first few hours after birth, due to
a deficiency in oxytocin in the pups, which at this stage is required to initiate
suckling activity (Schaller et al. 2010). For surviving knockout mice, a slight
postnatal growth retardation is followed at the adult stage by reduced lean mass,
increased fat mass despite decreased food intake and hypothalamic orexin-A
278 M. Pulix and A. Plagge

(Bischof et al. 2007; Kozlov et al. 2007). It is noteworthy that the observation of
reduced food intake is contrary to PWS symptoms. Hypothalamic POMC neurons
and their projections were found to be deficient in knockout mice (Maillard et al.
2016; Mercer et al. 2013), and leptin resistance might be related to reduced leptin
receptor recycling from endosomes to the plasma membrane (Wijesuriya et al.
2017). The Magel2 protein is also robustly expressed in the SCN and its levels
fluctuate with circadian rhythmicity (Kozlov et al. 2007). Consistent with SCN
function, the knockout mice show overall reduced wheel-running activity and
disturbed activity patterns under constant darkness (Kozlov et al. 2007; Tacer and
Potts 2017).
Similar to Magel2, Necdin (Ndn) constitutes another member of the Mage family
of proteins (Tacer and Potts 2017). Necdin is expressed in a number of brain regions,
and it has roles in neuronal survival, migration, differentiation, and neurite formation
(Table 10.1). Ndn knockout mice have neonatal respiratory problems, which lead to
a substantial rate of mortality shortly after birth (Gerard et al. 1999; Muscatelli et al.
2000). Respiratory disturbances are also observed in PWS and they are due to
defects in the serotoninergic system of the brainstem inspiratory rhythm generator
(Matarazzo et al. 2017). In the hypothalamus, lack of Ndn causes reduced numbers
of oxytocinergic neurons in the PVN and gonadotropin-releasing hormone neurons
in the preoptic area (Miller et al. 2009; Muscatelli et al. 2000). Deficits in the latter
neuron population might be responsible for the hypogonadism observed in PWS.
Necdin knockout mice present symptoms of hypothyroidism due to reduced levels of
thyrotropin-releasing hormone in the PVN (Hasegawa et al. 2012). Furthermore,
Necdin has a role in axon outgrowth as indicated by aberrant axon bundles in the
anterior hypothalamus and other brain regions (Lee et al. 2005). On the molecular
level, Necdin appears to interact with a diverse set of binding partners, including
cytoskeleton-associated proteins as well as transcription factors (Hasegawa et al.
2012; Lee et al. 2005).
The box C/D small nucleolar RNA cluster Snord116 also has a crucial role in
PWS (Cavaille 2017). Patients with microdeletions of this cluster, but intact
NECDIN and MAGEL2 genes, present with PWS symptoms, and knockout mouse
models also show some of these disease characteristics (Bieth et al. 2015; Ding et al.
2008; Skryabin et al. 2007). Postnatal growth retardation is a shared feature in
humans and mice with paternally inherited Snord116 deletions. However, knockout
mice remain lean and underweight throughout life due to increased energy expendi-
ture, while SNORD116-mutant patients develop the typical PWS hyperphagic obe-
sity toward adulthood (Bieth et al. 2015; Ding et al. 2008; Skryabin et al. 2007).
Interestingly, when the Snord116 cluster is deleted in the mediobasal hypothalamus
of adult mice via an adenovirus Cre-recombinase vector injection, the mutant mice
then develop hyperphagia and obesity reminiscent of PWS patients (Polex-Wolf
et al. 2018). These findings highlight physiological differences between humans and
mice and potentially indicate species-specific compensatory abilities to inherited
gene mutations during development.
10 Imprinted Genes and Hypothalamic Function 279

10.3.3 Other Imprinted Genes with Hypothalamic Functions

Although less well characterized, four other imprinted genes have roles in the
hypothalamus, which are briefly summarized here. A cluster of imprinted genes on
mouse chromosome 12 (human chromosome 14) contains two protein-coding genes
with hypothalamic functions, Dlk1 (Protein delta homolog 1) and Dio3 (Thyroxine
5-deiodinase), both expressed from the paternal allele (Table 10.1). The Dlk1
protein has structural similarity to Notch ligands, but its precise signaling mecha-
nism is unknown. Dlk1 is prominently expressed in several hypothalamic nuclei
(arcuate nucleus, PVN, DMH, and SCN), and it has been co-localized in oxytocin,
arginine–vasopressin, orexin, NPY, and dynorphin-positive neurons (Meister et al.
2013; Villanueva et al. 2012). Dlk1 involvement in embryonic development and
adiposity has been well characterized, but its hypothalamus-specific functions are
poorly understood. A mutation of DLK1 in humans leads to central precocious
puberty, possibly via GnRH-neuron regulation, and increased fat mass (Dauber
et al. 2017). Dio3 shows monoallelic expression from the paternal allele in the
neonatal hypothalamus (Martinez et al. 2014). Knockout mice display postnatal
growth retardation, which lasts into adulthood, deregulated thyroid hormone levels
and disturbances in the hypothalamic–pituitary–thyroid axis (Table 10.1). Reduced
adiposity is associated with changes in hypothalamic gene expression of the leptin–
melanocortin pathway, increased energy expenditure, and locomotion (Wu et al.
2017). Activity patterns of knockout mice also showed circadian changes, which
might be associated with Dio3 expression in the SCN. Additional behavioral
abnormalities relating to aggression and maternal care of the knockout mice are
associated with changes in the hypothalamic oxytocin and arginine–vasopressin
systems (Stohn et al. 2018). The third gene is the voltage-gated potassium channel
gene Kcnq1, which is located in a large imprinting cluster on mouse chromosome
7 (human chromosome 11) and is expressed from the maternal allele. It has a crucial
role in the hypothalamic–pituitary growth hormone axis. The potassium channel is
expressed in hypothalamic growth hormone-releasing hormone (GHRH) neurons
and pituitary somatotropes. KCNQ1 mutations in humans cause growth hormone
deficiency and gingival fibromatosis among other symptoms (Table 10.1)
(Tommiska et al. 2017). The fourth imprinted gene associated with hypothalamic
functions is the Paternally expressed gene 3 (Peg3), a Zinc finger-containing
DNA-binding factor that is widely expressed in the hypothalamus (Li et al. 1999).
Peg3 knockout mice have reduced body weight at all stages of life, but are
hypometabolic, show a proportionately increased fat mass, lower body temperature,
leptin resistance and changes in hypothalamic NPY, POMC, MCH, and Orexin
neuropeptide expression (Curley et al. 2005). Whether knockout females show
deficiencies in maternal behavior toward their offspring (e.g., nest building, pup
retrieval, nursing) that are linked to a reduced number of hypothalamic oxytocin
neurons, is controversial (Li et al. 1999; Denizot et al. 2016).
280 M. Pulix and A. Plagge

10.4 Analysis of Genomic Imprinting

10.4.1 Analysis of Methylation

As mentioned in the introduction, the canonical regulatory mechanism of imprinting

is the methylation of DNA at specific CGIs to form DMRs or ICRs. A breakthrough
in the study of DNA methylation was achieved with the technique of bisulfite
conversion of genomic DNA. The treatment with sodium bisulfite promotes the
deamination of cytosine residues, which is followed by desulfonation at elevated pH,
thus leading to conversion into uracil (Fig. 10.3). When a residue is methylated,
however, the deamination cannot take place and the cytosine remains unchanged.
Since only cytosines in single-stranded DNA are susceptible to conversion by
bisulfite, complete DNA denaturation is a critical step. After bisulfite conversion,
the DNA fragment of interest can be amplified by PCR using primers designed to
anneal to either the sense or the antisense DNA strand, which at this point are not
complementary anymore. The PCR primer design needs to be based on the expected
bisulfite-converted sequence, in which methylated cytosines will be retained as C,
and the unmethylated cytosines converted to uracil, i.e., such primers are usually
AT-rich (Fig. 10.3). Furthermore, AT-rich bisulfite PCR amplicons (from originally
unmethylated DNA) can be difficult to amplify using standard Taq-Polymerases and,
therefore, amplicon sizes of less than ~400 bp are recommended. Small amplicon
size is especially important when allelic differences in methylation are investigated
to avoid a bias in amplification of originally methylated over unmethylated alleles,
i.e., to minimize preferential amplification of templates with more balanced nucleo-
tide composition over highly AT-rich templates. The advantages of the bisulfite
method include the efficiency of the chemical conversion, relatively low cost and the
speed of the procedure. Because in one reaction the whole genome can be converted
and investigated in subsequent analyses, this has become the best approach for the
study of DNA methylation.
Following bisulfite conversion and PCR amplification, the DNA fragment of
interest can be further analyzed using several methods. Traditionally, the PCR
product can be cloned into a plasmid and then a sufficiently large number of
individual clones sequenced via Sanger sequencing. This procedure will produce a
representation of the frequency of methylated (preserved) and unmethylated
(converted) cytosines at every CpG dinucleotide position of the DNA amplicon in
the cloned population (Fig. 10.4a). In the analysis of a DMR of an imprinted gene, a
ratio of ~50% methylation would be expected at each CpG site whereby two types of
sequences should be distinguishable: predominantly methylated and predominantly
converted (Fig. 10.4a). If the DNA fragment of interest was isolated from hybrid
offspring of different mouse strains and contains a single nucleotide polymorphism
(SNP), the methylation can be associated with the parental origin of the allele.
An alternative to cloning and sequencing consists of the digestion of a PCR
product with restriction endonucleases that contain CpGs in their recognition sites
such as TaqI (T/CGA) and BstUI (CG/CG) (Fig. 10.4b). This technique, called
COBRA (combined bisulfite restriction analysis) (Eads and Laird 2002), detects
10 Imprinted Genes and Hypothalamic Function 281

NH2

O N O
H

Cytosine HN

NH2 O N
H
CH3
N Uracil

O N
H

5-Methylcytosine

before NaHSO4 treatment

5’-ACCTAAGGTCGTTACCCTACGTTAGGGCGTAGCAACGTGGACGCTATCTAGCTGCATCGAT-3’

converted, PCR-amplified sequence

5’-ATTTAAGGTTGTTATTTTACGTTAGGGTGTAGTAACGTGGATGTTATTTAGTTGTATTGAT-3’

Fig. 10.3 Bisulfite conversion of genomic DNA. The treatment with NaHSO3 induces the deami-
nation of cytosine into uracil residues. This conversion will not happen, if the cytosine is methylated
or hydroxy-methylated in the carbon 5 position. Following bisulfite treatment, the sense and
antisense strands of the original double-stranded DNA are no longer complementary. PCR amplifi-
cation of a strand will now lead to an exchange of uracil residues to thymine and the DNA sequence
will then be characterized by an abundance of A and T. Only the methylated sites will retain
cytosine residues. The primers for bisulfite sequencing need to be complementary to the converted
sequence (orange and green arrows). Black lollipops indicate methylated CpG sites

originally methylated CpGs as they are conserved after bisulfite treatment and PCR
amplification and then cut by the restriction endonuclease. By contrast, bisulfite
conversion of an unmethylated CpG leads to loss of the restriction site. If a CpG is
fully methylated, the PCR product will be completely digested, and vice versa
(Fig. 10.4b). Quantification of partially methylated sites will require blotting of the
gel, hybridization, and signal detection with a probe.
The presence of both cytosine and thymine in the same position following
bisulfite treatment and PCR amplification can indicate partial methylation or poten-
tially incomplete bisulfite conversion. In the bisulfite sequencing approach, incom-
plete conversion can be recognized through the presence of remaining C residues
outside of CpG dinucleotides—all of these should have been converted. A potential
282 M. Pulix and A. Plagge

A) Cloning and Sanger

sequencing of bisulfite PCR Predominantly
products methylated

differentially
methylated

Predominantly
unmethylated

B) Combined bisulfite restriction analysis (COBRA)

BstUI (CGCG) BstUI (CGCG)

Bisulfite-treated gDNA Untreated gDNA

BstUI digest

1000 1000

500 uncut 500

fragment
(Cs converted)
fully cut
fragments

C) Pyrosequencing of bisulfite PCR amplicon

Kidney

72% 92% 79% 54%

40

30

20

10

-10
E S C A A T A T C G A T C C A G A C T A C A T A G C A T A T C G A C A A T C A C A C T C T A T A T A C G A T C
5 10 15 20 25 30 35 40 45 50

Fig. 10.4 Analysis methods of bisulfite-treated and PCR-amplified genomic DNA. (a) Cloning
and Sanger sequencing of bisulfite PCR products. Scheme showing three examples of possible
outcomes for methylation patterns in an amplicon covering a CGI. Each row of dots represents one
cloned PCR product and each dot represents an individual CpG site (black ¼ methylated,
white ¼ unmethylated). The differentially methylated example, which contains methylated and
unmethylated sequences, would be typical of an imprinted DMR. Further SNP analysis would be
required to associate the methylation status with the parental origin of an allele. (b) Combined
bisulfite restriction analysis (COBRA) on untreated and bisulfite-treated genomic DNA from
hippocampus, cerebral cortex, and kidney. Restriction enzyme BstUI recognizes the site CGCG.
When a cytosine is unmethylated in the genomic DNA, it is converted to thymine following bisulfite
treatment and PCR amplification, disrupting the integrity of the restriction sites. In contrast,
10 Imprinted Genes and Hypothalamic Function 283

drawback of bisulfite sequencing is related to cloning biases that can occur in

bacterial strains used for plasmid transformation after ligation of bisulfite PCR
amplicons. In genomic imprinting, a DMR will result in bisulfite amplicons with
different sequences derived from methylated and unmethylated alleles, respectively,
and it is possible that the ligated amplicons are not equally propagated after
transformation into E. coli, thus giving a wrong representation of methylation
frequencies. On the other hand, in the COBRA method appropriate positive controls
for restriction endonuclease functionality need to be included. For this reason, it is
always recommended to confirm any results using at least two independent
approaches.
The best approach to accurately quantify the methylation status is the use of
pyrosequencing of bisulfite-converted genomic DNA (Dejeux et al. 2009). This
method also includes PCR amplification as a first step after bisulfite treatment to
generate fragments of ~100 bp. These are then sequenced by adding one nucleotide
at a time and quantifying its incorporation through a cascade of enzymatic reactions
that result in strictly proportional luciferase-mediated light emission. For example, at
a partially methylated CpG site, incorporation of dGTP (originally methylated C)
and dATP(αS) (originally unmethylated C) are tested consecutively, and the ratio of
light emitted at the two nucleotide steps represents the percentage of methylated
cytosine (Fig. 10.4c). Pyrosequencing is now accepted as the “gold standard” for
quantitative sequencing. Apart from analyzing methylation at specific loci, bisulfite
treatment has now been combined with next-generation sequencing to investigate
the whole genome in specific tissues or cell types, for example, via whole-genome
bisulfite sequencing (WGBS) (Harris et al. 2010). As details of the various
approaches and the importance of avoiding biases at specific steps in the protocols
cannot be covered here, we refer the reader to more specific, recently published work
(Harris et al. 2010; Olova et al. 2018).

10.4.2 Analysis of Monoallelic Expression and Single-Cell Resolution

of Imprinting

The analysis of imprinted genes also requires a proof of preferential or completely

monoallelic gene expression in a parent-of-origin-specific way. To demonstrate this,
exonic SNPs between different mouse strains are used, most commonly between
C57BL/6J and Mus musculus molossinus (JF1) or Mus musculus castaneus (Cast/
EiJ) strains. Tissues or cells from hybrid offspring of reciprocal parental crosses
(e.g., C57BL/6J  JF1 and JF1  C57BL/6J) are isolated, cDNA generated and the
ä

Fig. 10.4 (continued) methylated sequences maintain the cytosine and the restriction enzyme will
cut the PCR amplicon. This example CpG island shows the presence of both methylated and
unmethylated alleles (unpublished data). (c) Bisulfite pyrosequencing analysis of the same CGI as
in (b). A pyrogram sequencing trace of a bisulfite PCR amplicon is shown. The positions
highlighted in light blue represent CpG sites and the degree of methylation at each site is quantified
and displayed as a percentage above the highlighted peaks (unpublished data)
284 M. Pulix and A. Plagge

BRAIN KIDNEY

JF1/B6 B6/JF1 B6/JF1

A: 75% A: 34% A: 52%
G: 25% G: 66% G: 48%

Fig. 10.5 SNP pyrosequencing to quantify allele-specific expression levels of a gene. Brain and
kidney cDNAs were prepared from reciprocal crosses of C57BL/6J (B6) and Japanese fancy mouse
1 (Mus musculus molossinus, JF1), and a SNP-containing amplicon analyzed via pyrosequencing.
Quantification of SNP prevalence indicates ~70% preferential maternal allele-specific expression of
Trappc9 in the brain, but equal biallelic expression in kidney. The reciprocal mouse crosses are
shown as JF1/B6 and B6/JF1, whereby the first position conventionally indicates the maternal
genotype. The SNP position is shaded in light blue

ratio of expression of the SNPs within the gene of interest determined via
pyrosequencing. In the example shown in Fig. 10.5, expression of the Trappc9
gene was found to be predominantly from the maternal allele (~70%) in brain tissue,
while in kidney both SNPs show equal biallelic expression levels.
Recently, parent-of-origin biases of gene expression in the hybrid mouse brain
have been characterized using RNAseq whole-transcriptome methods. In some
cases, specific brain subregions, including the hypothalamus and arcuate nucleus
were analyzed (Babak et al. 2015; Bonthuis et al. 2015; Huang et al. 2017; Perez
et al. 2015). These transcriptome datasets, which include previously unknown
transcripts, can be a good starting point for the investigation of imprinted genes in
the brain. While several genes, including Gnas, Dlk1, and Trappc9, show tissue- or
cell type-specific imprinting (Chen et al. 2009; Ferron et al. 2011; Perez et al. 2015),
it is often unclear whether every cell within a population exhibits the parent-of-
origin-specific, monoallelic expression. In particular, for genes with a bias toward
parental allele-specific expression (e.g., 70% maternal, 30% paternal transcripts,
based on whole tissue lysates) it remains unexplained whether such a ratio reflects
unequal biallelic expression within each cell, or whether it is due to a mixture of cells
with absolute monoallelic expression and cells with equal biallelic expression. This
question is now being addressed with single-cell technologies. There are two main
approaches to the single-cell analysis of allelic gene expression. On the one hand,
probe hybridization onto fixed cells or tissue sections is combined with microscopy
to distinguish between mono- or biallelic expression in situ (Bonthuis et al. 2015;
10 Imprinted Genes and Hypothalamic Function 285

Ginart et al. 2016; Huang et al. 2017). On the other hand, molecular genetics or
genome-wide next-generation sequencing methods are applied on isolated single cell
extracts (Angermueller et al. 2016; Cheow et al. 2015; Clark et al. 2018; Deng et al.
2014; Kelsey et al. 2017). As some of these techniques have been applied to
hypothalamic tissues, we briefly describe them here as promising approaches for
future research.
One of the methods designed to investigate allele-specific expression at the tissue
level is “single nucleotide polymorphism RNA fluorescent in situ hybridization”
(SNP-FISH) (Ginart et al. 2016; Levesque et al. 2013). The method is applied on
tissues or cells from hybrid mice (e.g., C57BL/6J  Cast/EiJ), which harbor several
SNPs in the mRNA of the gene of interest. Three different pools of fluorescently
labeled oligonucleotides are then hybridized simultaneously to fixed samples: (a) a
pool of “guide” probes, which bind to the mRNA of the gene of interest outside of
the SNP regions for general detection of the relevant transcript; (b) two pools of
SNP-specific oligonucleotides with different fluorescent labels, which vary only in a
single nucleotide position to detect the parental origin of the transcript (Fig. 10.6).
These SNP-specific oligos are initially blocked by shorter “mask” oligonucleotides,
leaving the SNP-containing region free to hybridize to the perfectly matching
mRNA. Eventually, the SNP-specific oligos shed their “mask” oligos to fully anneal
to the mRNA of interest. By analyzing co-localization of fluorescent signals from the
“guide” oligos and the SNP-specific oligos, respectively, Ginart and colleagues
(Ginart et al. 2016) were able to demonstrate that, within one tissue, cell populations
with monoallelic maternal expression of the imprinted gene H19 coexisted with
neighboring cell populations that showed biallelic expression.
Another technique used to visualize the imprinted expression of genes in tissues
is RNAscope in situ hybridization. Bonthuis and colleagues (Bonthuis et al. 2015)
adapted the standard RNAscope (Advanced Cell Diagnostics, ACD) method to
investigate actively transcribed alleles of specific genes in individual cell nuclei of
fixed sections from the hypothalamic arcuate nucleus and dorsal raphe nucleus
(Bonthuis et al. 2015). Using intronic oligonucleotide probes, which detect nascent
transcripts in the nucleus before splicing occurs, they were able to determine through
imaging for the presence of one or two hybridization signals (“color substrate dots”)
within a nucleus, whether one or both alleles of a gene were expressed. In contrast to
the SNP-FISH technique, however, the RNAscope method does not provide infor-
mation on the parental origin of monoallelic expression. Also, when used on tissue
sections, the RNAscope approach has an inherent false detection rate problem, since
the plane of sectioning might result in spatial separation of the two alleles in
sequential sections. This would result in single hybridization signals within nuclei
although biallelic expression occurs. This false detection rate needs to be accounted
for by using appropriate control probes for biallelically expressed genes. The authors
estimate that this method has an error rate of ~25% nuclei showing potentially false
monoallelic expression (Bonthuis et al. 2015; Huang et al. 2017). Nevertheless, due
to the high sensitivity of the hybridization probes and the straight forward protocol,
this approach can provide a useful first assessment of the localization of cells with
mono- and biallelic expression within tissues. Especially since it can be combined
 286

allele specific oligonucleotide A mask oligonucleotide

guide oligonucleotide
allele A RNA 5’-TAGCTAGAACT
5’-GCGATCTTGTCGAGCTGTCGATGTCGACT
3’-UGUGAGCUCGAUGCUAGCAUCGAUCUUGACGAUCGAUGGCUAAGAUCGGAUCGUA……GAUCGCCAAGCUCGCUAGAACAGCUCGACAGCUACAGCUGACACAAGCUCUCUUUG-5’

allele specific oligonucleotide B mask oligonucleotide

guide oligonucleotide
5’-TAGCTGGAACT

initial annealing
allele B RNA 5’-GCGATCTTGTCGAGCTGTCGATGTCGACT
3’-UGUGAGCUUGAUGCUAGCAUCGACCUUGACGAUCGAUGGCUAAGAUCGGAUCGUA……GAUCGCCAAGCUCGCUAGAACAGCUCGACAGCUACAGCUGACACAAGCUCUCUUUG-5’

guide oligonucleotide
allele A RNA 5’-TAGCTAGAACTGCTAGCTACCGATTCTA 5’-GCGATCTTGTCGAGCTGTCGATGTCGACT
3’-UGUGAGCUCGAUGCUAGCAUCGAUCUUGACGAUCGAUGGCUAAGAUCGGAUCGUA……GAUCGCCAAGCUCGCUAGAACAGCUCGACAGCUACAGCUGACACAAGCUCUCUUUG-5’

guide oligonucleotide
allele B RNA 5’-TAGCTGGAACTGCTAGCTACCGATTCTA 5’-GCGATCTTGTCGAGCTGTCGATGTCGACT

fully annealed
3’-UGUGAGCUCGAUGCUAGCAUCGACCUUGACGAUCGAUGGCUAAGAUCGGAUCGUA……GAUCGCCAAGCUCGCUAGAACAGCUCGACAGCUACAGCUGACACAAGCUCUCUUUG-5’

allele A
Cytoplasm

transcription sites
non-specific
Nucleus

allele B unclassified RNA

M. Pulix and A. Plagge
 10

Fig. 10.6 Schematic illustration of the SNP-FISH technique for the detection of allele-specific RNA expression in cells and tissue sections. The allelic RNA
variants of a gene of interest are shown with a SNP highlighted in red. A pool of fluorescently labeled “guide” oligonucleotides, which hybridize to non-variant
sequences of the RNA, serve as a general marker for the presence of the transcript. Allele-specific oligonucleotides covering SNP positions (SNV probes) are
labeled with a second and third fluorescent tag, respectively. Initially, only a short stretch of the SNV probe sequence (including the actual SNP position) is free
to hybridize with the target RNA, as the remainder of the oligo is bound by a short complementary “mask” oligo. This allows specific binding of the SNV probes
to the allelic RNA variants. The “mask” oligos are then displaced as the allele-specific oligos fully anneal with their target RNAs. Figure adapted from Levesque
et al. (2013)
Imprinted Genes and Hypothalamic Function
287
288 M. Pulix and A. Plagge

with tissue RNAseq analysis (see below) for determination of parental expression
biases.
Apart from these in situ techniques, molecular genetics approaches have been
established to analyze epigenetic features as well as RNA expression from single-
cell extracts (Angermueller et al. 2016; Cheow et al. 2015; Clark et al. 2018; Deng
et al. 2014; Kelsey et al. 2017). Some methods are designed for a low-cost, focused
investigation of specific loci (Cheow et al. 2015), while recent trends favor genome-
wide next-generation sequencing technologies. RNAseq was first applied by Deng
and colleagues (Deng et al. 2014) to investigate transcriptome-wide allelic expres-
sion in single cells using suitable strain-specific SNPs. Apart from parent-of-origin-
dependent (imprinted) gene expression, they observed that a substantial proportion
of genes (12–24%) are expressed monoallelically in a dynamic, random, and sto-
chastic way (Deng et al. 2014). The most recent technological advances enable a
combined analysis of DNA methylation, transcriptome, and nucleosome accessibil-
ity from single cells (scM&T-seq, scNMT-seq) (Angermueller et al. 2016; Kelsey
et al. 2017). These approaches allow direct correlations to be made between epige-
netic status, chromatin accessibility, and gene expression in a single cell. Although
we cannot cover these methods in detail here [for a comprehensive review see
Kelsey et al. (2017)], we would like to emphasize that consideration needs to be
given to the way single cells can be isolated and the overall number of cells required
for analysis. Most frequently, flow cytometry/fluorescence-activated cell sorting
(FACS) is used to obtain single cells or nuclei (Luo et al. 2017), although
microfluidics or manual collection via micromanipulators can also be applied
depending on the tissue or cell type of interest (Cheow et al. 2015). Furthermore,
and again depending on the complexity of the tissue to be investigated, usually
hundreds or thousands of cells need to be included in an experiment to obtain a
representative dataset for the cell population in question. As single-cell RNAseq and
methylome analyses have already been applied on some brain regions (Luo et al.
2017; Tasic et al. 2016; Zeisel et al. 2015), it will be exciting to see the multi-omics
approaches being adapted to neurobiological questions and specifically to character-
ize the large diversity of hypothalamic cell types in the future.

10.5 Perspectives

Within the brain, the hypothalamus has emerged as a major region in which imprinted
genes are expressed. Here, they exert crucial functions in the regulation of postnatal
development, energy homeostasis, hypothalamus–pituitary, and hypothalamus–thy-
roid neuroendocrine axes, as well as circadian rhythmicity and sleep. It is highly
likely that the list provided in this chapter will be expanded in the near future, as
several additional imprinted genes have been associated with a role in the hypothala-
mus, for example, Neuronatin, Mkrn3, Grb10, and Liz. On a molecular level, the
recent developments in single-cell techniques will shed further light on the epigenetic
regulation of imprinted genes. Many genes display an expression bias toward one
parental allele when analyzed on a whole-tissue level. To clarify whether this is due to
10 Imprinted Genes and Hypothalamic Function 289

a mixture of cell types, some of which showing biallelic others monoallelic expres-
sion, and to understand the neurobiological relevance of this, will be an exciting field
of future research.

Acknowledgment We would like to thank the Wellcome Trust PhD studentship program in
Cellular and Molecular Physiology at the University of Liverpool (099795/Z/12/Z) for research
support.

Key References

Buiting et al. (2016). This review provides an excellent overview about the clinical
symptoms and genetics of Angelman syndrome.
Cassidy et al. (2012). A recommended review summarising the clinical symptoms as
well as the genetics of Prader-Willi syndrome.
Chen et al. (2009). This paper shows for the first time genomic imprinting of Gnas in
the brain and that loss of maternal allele-specific expression of Gnas in the brain is
the cause for an obesity phenotype due to reduced energy expenditure.
Ferguson-Smith (2011) This review provides an excellent overview about the
history of genomic imprinting research and well-studied imprinted genes clusters.
Kelsey et al. (2017). This recent review gives a good introduction into novel single-
cell analysis techniques for epigenetics and gene expression.
Perez et al. (2015) This paper systematically analyzed imprinted gene expression
within the brain transcriptome with excellent detail.
Peters (2014). This review emphasizes the physiological functions of imprinted
genes.
Plagge et al. (2004). This paper identified the physiological functions of the pater-
nally expressed transcript variant Gnasxl of the Gnas locus.

References
Angermueller C, Clark SJ, Lee HJ, Macaulay IC, Teng MJ, Hu TX, Krueger F, Smallwood S,
Ponting CP, Voet T, Kelsey G, Stegle O, Reik W (2016) Parallel single-cell sequencing links
transcriptional and epigenetic heterogeneity. Nat Methods 13:229–232
Babak T, DeVeale B, Tsang EK, Zhou Y, Li X, Smith KS, Kukurba KR, Zhang R, Li JB, van der
Kooy D, Montgomery SB, Fraser HB (2015) Genetic conflict reflected in tissue-specific maps of
genomic imprinting in human and mouse. Nat Genet 47:544–549
Bartolomei MS, Ferguson-Smith AC (2011) Mammalian genomic imprinting. Cold Spring Harb
Perspect Biol 3:a002592
Bieth E, Eddiry S, Gaston V, Lorenzini F, Buffet A, Conte Auriol F, Molinas C, Cailley D,
Rooryck C, Arveiler B, Cavaille J, Salles JP, Tauber M (2015) Highly restricted deletion of
the SNORD116 region is implicated in Prader-Willi Syndrome. Eur J Hum Genet 23:252–255
Bischof JM, Stewart CL, Wevrick R (2007) Inactivation of the mouse Magel2 gene results in
growth abnormalities similar to Prader-Willi syndrome. Hum Mol Genet 16:2713–2719
Bonthuis PJ, Huang WC, Stacher Horndli CN, Ferris E, Cheng T, Gregg C (2015) Noncanonical
genomic imprinting effects in offspring. Cell Rep 12:979–991
290 M. Pulix and A. Plagge

Buiting K, Williams C, Horsthemke B (2016) Angelman syndrome—insights into a rare

neurogenetic disorder. Nat Rev Neurol 12(10):584–593
Cassidy SB, Schwartz S, Miller JL, Driscoll DJ (2012) Prader-Willi syndrome. Genet Med
14:10–26
Cavaille J (2017) Box C/D small nucleolar RNA genes and the Prader-Willi syndrome: a complex
interplay. Wiley Interdiscip Rev RNA 8:e1417
Chen M, Wang J, Dickerson KE, Kelleher J, Xie T, Gupta D, Lai EW, Pacak K, Gavrilova O,
Weinstein LS (2009) Central nervous system imprinting of the G protein G(s)α and its role in
metabolic regulation. Cell Metab 9:548–555
Chen M, Nemechek NM, Mema E, Wang J, Weinstein LS (2011) Effects of deficiency of the G
protein Gsα on energy and glucose homeostasis. Eur J Pharmacol 660:119–124
Chen M, Berger A, Kablan A, Zhang J, Gavrilova O, Weinstein LS (2012) Gsα deficiency in the
paraventricular nucleus of the hypothalamus partially contributes to obesity associated with Gsα
mutations. Endocrinology 153:4256–4265
Chen M, Shrestha YB, Podyma B, Cui Z, Naglieri B, Sun H, Ho T, Wilson EA, Li YQ,
Gavrilova O, Weinstein LS (2017) Gsα deficiency in the dorsomedial hypothalamus underlies
obesity associated with Gsα mutations. J Clin Invest 127:500–510
Cheow LF, Quake SR, Burkholder WF, Messerschmidt DM (2015) Multiplexed locus-specific
analysis of DNA methylation in single cells. Nat Protoc 10:619–631
Clark SJ, Argelaguet R, Kapourani CA, Stubbs TM, Lee HJ, Alda-Catalinas C, Krueger F,
Sanguinetti G, Kelsey G, Marioni JC, Stegle O, Reik W (2018) scNMT-seq enables joint
profiling of chromatin accessibility DNA methylation and transcription in single cells. Nat
Commun 9:781
Curley JP, Pinnock SB, Dickson SL, Thresher R, Miyoshi N, Surani MA, Keverne EB (2005)
Increased body fat in mice with a targeted mutation of the paternally expressed imprinted gene
Peg3. FASEB J 19:1302–1304
Dauber A, Cunha-Silva M, Macedo DB, Brito VN, Abreu AP, Roberts SA, Montenegro LR,
Andrew M, Kirby A, Weirauch MT, Labilloy G, Bessa DS, Carroll RS, Jacobs DC, Chappell
PE, Mendonca BB, Haig D, Kaiser UB, Latronico AC (2017) Paternally inherited DLK1
deletion associated with familial central precocious puberty. J Clin Endocrinol Metab
102:1557–1567
Dejeux E, El abdalaoui H, Gut IG, Tost J (2009) Identification and quantification of differentially
methylated loci by the pyrosequencing technology. Methods Mol Biol 507:189–205
Deng Q, Ramskold D, Reinius B, Sandberg R (2014) Single-cell RNA-seq reveals dynamic,
random monoallelic gene expression in mammalian cells. Science 343:193–196
Denizot AL, Besson V, Correra RM, Mazzola A, Lopes I, Courbard JR, Marazzi G, Sassoon DA
(2016) A novel mutant allele of Pw1/Peg3 does not affect maternal behavior or nursing
behavior. PLoS Genet 12:e1006053
Ding F, Li HH, Zhang S, Solomon NM, Camper SA, Cohen P, Francke U (2008) SnoRNA
Snord116 (Pwcr1/MBII-85) deletion causes growth deficiency and hyperphagia in mice.
PLoS One 3:e1709
Eads CA, Laird PW (2002) Combined bisulfite restriction analysis (COBRA). Methods Mol Biol
200:71–85
Ehlen JC, Jones KA, Pinckney L, Gray CL, Burette S, Weinberg RJ, Evans JA, Brager AJ, Zylka
MJ, Paul KN, Philpot BD, DeBruyne JP (2015) Maternal Ube3a loss disrupts sleep homeostasis
but leaves circadian rhythmicity largely intact. J Neurosci 35:13587–13598
Ferguson-Smith AC (2011) Genomic imprinting: the emergence of an epigenetic paradigm. Nat
Rev Genet 12:565–575
Ferron SR, Charalambous M, Radford E, McEwen K, Wildner H, Hind E, Morante-Redolat JM,
Laborda J, Guillemot F, Bauer SR, Farinas I, Ferguson-Smith AC (2011) Postnatal loss of Dlk1
imprinting in stem cells and niche astrocytes regulates neurogenesis. Nature 475:381–385
Genevieve D, Sanlaville D, Faivre L, Kottler ML, Jambou M, Gosset P, Boustani-Samara D,
Pinto G, Ozilou C, Abeguile G, Munnich A, Romana S, Raoul O, Cormier-Daire V, Vekemans
M (2005) Paternal deletion of the GNAS imprinted locus (including Gnasxl) in two girls
10 Imprinted Genes and Hypothalamic Function 291

presenting with severe pre- and post-natal growth retardation and intractable feeding difficulties.
Eur J Hum Genet 13:1033–1039
Gerard M, Hernandez L, Wevrick R, Stewart CL (1999) Disruption of the mouse necdin gene
results in early post-natal lethality. Nat Genet 23:199–202
Ginart P, Kalish JM, Jiang CL, Yu AC, Bartolomei MS, Raj A (2016) Visualizing allele-specific
expression in single cells reveals epigenetic mosaicism in an H19 loss-of-imprinting mutant.
Genes Dev 30:567–578
Haig D (2004) Genomic imprinting and kinship: how good is the evidence? Annu Rev Genet
38:553–585
Harris RA, Wang T, Coarfa C, Nagarajan RP, Hong C, Downey SL, Johnson BE, Fouse SD,
Delaney A, Zhao Y, Olshen A, Ballinger T, Zhou X, Forsberg KJ, Gu J, Echipare L, O’Geen H,
Lister R, Pelizzola M, Xi Y, Epstein CB, Bernstein BE, Hawkins RD, Ren B, Chung WY, Gu H,
Bock C, Gnirke A, Zhang MQ, Haussler D, Ecker JR, Li W, Farnham PJ, Waterland RA,
Meissner A, Marra MA, Hirst M, Milosavljevic A, Costello JF (2010) Comparison of
sequencing-based methods to profile DNA methylation and identification of monoallelic epige-
netic modifications. Nat Biotechnol 28:1097–1105
Hasegawa K, Kawahara T, Fujiwara K, Shimpuku M, Sasaki T, Kitamura T, Yoshikawa K (2012)
Necdin controls Foxo1 acetylation in hypothalamic arcuate neurons to modulate the thyroid
axis. J. Neurosci 32:5562–5572
He Q, Zhu Y, Corbin BA, Plagge A, Bastepe M (2015) The G protein alpha subunit variant XLαs
promotes inositol 1,4,5-trisphosphate signaling and mediates the renal actions of parathyroid
hormone in vivo. Sci Signal 8:ra84
Huang WC, Ferris E, Cheng T, Horndli CS, Gleason K, Tamminga C, Wagner JD, Boucher KM,
Christian JL, Gregg C (2017) Diverse non-genetic, allele-specific expression effects shape
genetic architecture at the cellular level in the mammalian brain. Neuron 93:1094–1109
Ivanova E, Kelsey G (2011) Imprinted genes and hypothalamic function. J Mol Endocrinol 47:
R67–R74
Kawashima S, Nakamura A, Inoue T, Matsubara K, Horikawa R, Wakui K, Takano K,
Fukushima Y, Tatematsu T, Mizuno S, Tsubaki J, Kure S, Matsubara Y, Ogata T, Fukami M,
Kagami M (2018) Maternal uniparental disomy for chromosome 20: physical and endocrino-
logical characteristics of five patients. J Clin Endocrinol Metab 103:2083–2088
Kelsey G, Stegle O, Reik W (2017) Single-cell epigenomics: recording the past and predicting the
future. Science 358:69–75
Kozlov SV, Bogenpohl JW, Howell MP, Wevrick R, Panda S, Hogenesch JB, Muglia LJ, Van
Gelder RN, Herzog ED, Stewart CL (2007) The imprinted gene Magel2 regulates normal
circadian output. Nat Genet 39:1266–1272
Krechowec SO, Burton KL, Newlaczyl AU, Nunn N, Vlatkovic N, Plagge A (2012) Postnatal
changes in the expression pattern of the imprinted signalling protein XLαs underlie the changing
phenotype of deficient mice. PLoS One 7:e29753
LaSalle JM, Reiter LT, Chamberlain SJ (2015) Epigenetic regulation of UBE3A and roles in human
neurodevelopmental disorders. Epigenomics 7:1213–1228
Lee S, Walker CL, Karten B, Kuny SL, Tennese AA, O’Neill MA, Wevrick R (2005) Essential role
for the Prader-Willi syndrome protein necdin in axonal outgrowth. Hum Mol Genet 14:627–637
Levesque MJ, Ginart P, Wei Y, Raj A (2013) Visualizing SNVs to quantify allele-specific
expression in single cells. Nat Methods 10:865–867
Li L, Keverne EB, Aparicio SA, Ishino F, Barton SC, Surani MA (1999) Regulation of maternal
behavior and offspring growth by paternally expressed Peg3. Science 284:330–333
Li YQ, Shrestha Y, Pandey M, Chen M, Kablan A, Gavrilova O, Offermanns S, Weinstein LS
(2016) G(q/11)α and G(s)α mediate distinct physiological responses to central melanocortins. J
Clin Invest 126:40–49
Luo C, Keown CL, Kurihara L, Zhou J, He Y, Li J, Castanon R, Lucero J, Nery JR, Sandoval JP,
Bui B, Sejnowski TJ, Harkins TT, Mukamel EA, Behrens MM, Ecker JR (2017) Single-cell
292 M. Pulix and A. Plagge

methylomes identify neuronal subtypes and regulatory elements in mammalian cortex. Science
357:600–604
Lyko F (2018) The DNA methyltransferase family: a versatile toolkit for epigenetic regulation. Nat
Rev Genet 19:81–92
Maillard J, Park S, Croizier S, Vanacker C, Cook JH, Prevot V, Tauber M, Bouret SG (2016) Loss
of Magel2 impairs the development of hypothalamic Anorexigenic circuits. Hum Mol Genet
25:3208–3215
Mantovani G, Bastepe M, Monk D, de Sanctis L, Thiele S, Usardi A, Ahmed SF, Bufo R,
Choplin T, De Filippo G, Devernois G, Eggermann T, Elli FM, Freson K, Garcia Ramirez A,
Germain-Lee EL, Groussin L, Hamdy N, Hanna P, Hiort O, Juppner H, Kamenicky P, Knight N,
Kottler ML, Le Norcy E, Lecumberri B, Levine MA, Makitie O, Martin R, Martos-Moreno GA,
Minagawa M, Murray P, Pereda A, Pignolo R, Rejnmark L, Rodado R, Rothenbuhler A,
Saraff V, Shoemaker AH, Shore EM, Silve C, Turan S, Woods P, Zillikens MC, Perez de
Nanclares G, Linglart A (2018) Diagnosis and management of pseudohypoparathyroidism and
related disorders: first international Consensus Statement. Nat Rev Endocrinol 14:476–500
Martinez ME, Charalambous M, Saferali A, Fiering S, Naumova AK, Germain DS, Ferguson-
Smith AC, Hernandez A (2014) Genomic imprinting variations in the mouse type 3 deiodinase
gene between tissues and brain regions. Mol Endocrinol 28:1875–1886
Matarazzo V, Caccialupi L, Schaller F, Shvarev Y, Kourdougli N, Bertoni A, Menuet C,
Voituron N, Deneris E, Gaspar P, Bezin L, Durbec P, Hilaire G, Muscatelli F (2017) Necdin
shapes serotonergic development and SERT activity modulating breathing in a mouse model for
Prader-Willi syndrome. elife 6:e32640
Mehta S, Williamson CM, Ball S, Tibbit C, Beechey C, Fray M, Peters J (2015) Transcription
driven somatic DNA methylation within the imprinted Gnas cluster. PLoS One 10:e0117378
Meister B, Perez-Manso M, Daraio T (2013) Delta-like 1 homologue is a hypothalamus- enriched
protein that is present in orexin-containing neurones of the lateral hypothalamic area.
J Neuroendocrinol 25:617–625
Mercer RE, Michaelson SD, Chee MJ, Atallah TA, Wevrick R, Colmers WF (2013) Magel2 is
required for leptin-mediated depolarization of POMC neurons in the hypothalamic arcuate
nucleus in mice. PLoS Genet 9:e1003207
Miller NL, Wevrick R, Mellon PL (2009) Necdin, a Prader-Willi syndrome candidate gene,
regulates gonadotropin-releasing hormone neurons during development. Hum Mol Genet
18:248–260
Mulchandani S, Bhoj EJ, Luo M, Powell-Hamilton N, Jenny K, Gripp KW, Elbracht M,
Eggermann T, Turner CL, Temple IK, Mackay DJ, Dubbs H, Stevenson DA, Slattery L, Zackai
EH, Spinner NB, Krantz ID, Conlin LK (2016) Maternal uniparental disomy of chromosome 20:
a novel imprinting disorder of growth failure. Genet Med 18:309–315
Muscatelli F, Abrous DN, Massacrier A, Boccaccio I, Le Moal M, Cau P, Cremer H (2000)
Disruption of the mouse Necdin gene results in hypothalamic and behavioral alterations
reminiscent of the human Prader-Willi syndrome. Hum Mol Genet 9:3101–3110
Nunn N, Feetham CH, Martin J, Barrett-Jolley R, Plagge A (2013) Elevated blood pressure, heart
rate and body temperature in mice lacking the XLαs protein of the Gnas locus is due to increased
sympathetic tone. Exp Physiol 98:1432–1445
Olova N, Krueger F, Andrews S, Oxley D, Berrens RV, Branco MR, Reik W (2018) Comparison of
whole-genome bisulfite sequencing library preparation strategies identifies sources of biases
affecting DNA methylation data. Genome Biol 19:33
Perez JD, Rubinstein ND, Fernandez DE, Santoro SW, Needleman LA, Ho-Shing O, Choi JJ,
Zirlinger M, Chen SK, Liu JS, Dulac C (2015) Quantitative and functional interrogation of
parent-of-origin allelic expression biases in the brain. elife 4:e07860
Peters J (2014) The role of genomic imprinting in biology and disease: an expanding view. Nat Rev
Genet 15:517–530
Plagge A (2012) Non-coding RNAs at the Gnas and Snrpn-Ube3a imprinted gene loci and their
involvement in hereditary disorders. Front Genet 3:264
10 Imprinted Genes and Hypothalamic Function 293

Plagge A, Gordon E, Dean W, Boiani R, Cinti S, Peters J, Kelsey G (2004) The imprinted signaling
protein XLαs is required for postnatal adaptation to feeding. Nat Genet 36:818–826
Plagge A, Isles AR, Gordon E, Humby T, Dean W, Gritsch S, Fischer-Colbrie R, Wilkinson LS,
Kelsey G (2005) Imprinted Nesp55 influences behavioral reactivity to novel environments. Mol
Cell Biol 25:3019–3026
Plasschaert RN, Bartolomei MS (2014) Genomic imprinting in development, growth, behavior and
stem cells. Development 141:1805–1813
Podyma B, Sun H, Wilson EA, Carlson B, Pritikin E, Gavrilova O, Weinstein LS, Chen M (2018)
The stimulatory G protein Gsα is required in melanocortin 4 receptor-expressing cells for
normal energy balance, thermogenesis, and glucose metabolism. J Biol Chem
293:10993–11005
Polex-Wolf J, Lam BY, Larder R, Tadross J, Rimmington D, Bosch F, Cenzano VJ, Ayuso E, Ma
MK, Rainbow K, Coll AP, O’Rahilly S, Yeo GS (2018) Hypothalamic loss of Snord116
recapitulates the hyperphagia of Prader-Willi syndrome. J Clin Invest 128:960–969
Richard N, Molin A, Coudray N, Rault-Guillaume P, Juppner H, Kottler ML (2013) Paternal GNAS
mutations lead to severe intrauterine growth retardation (IUGR) and provide evidence for a role
of XLαs in fetal development. J Clin Endocrinol Metab 98:E1549–E1556
Schaller F, Watrin F, Sturny R, Massacrier A, Szepetowski P, Muscatelli F (2010) A single
postnatal injection of oxytocin rescues the lethal feeding behaviour in mouse newborns deficient
for the imprinted Magel2 gene. Hum Mol Genet 19:4895–4905
Shi SQ, Bichell TJ, Ihrie RA, Johnson CH (2015) Ube3a imprinting impairs circadian robustness in
Angelman syndrome models. Curr Biol 25:537–545
Skryabin BV, Gubar LV, Seeger B, Pfeiffer J, Handel S, Robeck T, Karpova E, Rozhdestvensky
TS, Brosius J (2007) Deletion of the MBII-85 snoRNA gene cluster in mice results in postnatal
growth retardation. PLoS Genet 3:e235
Stohn JP, Martinez ME, Zafer M, Lopez-Espindola D, Keyes LM, Hernandez A (2018) Increased
aggression and lack of maternal behavior in Dio3-deficient mice are associated with
abnormalities in oxytocin and vasopressin systems. Genes Brain Behav 17:23–35
Tacer KF, Potts PR (2017) Cellular and disease functions of the Prader-Willi Syndrome gene
MAGEL2. Biochem J 474:2177–2190
Tasic B, Menon V, Nguyen TN, Kim TK, Jarsky T, Yao Z, Levi B, Gray LT, Sorensen SA,
Dolbeare T, Bertagnolli D, Goldy J, Shapovalova N, Parry S, Lee C, Smith K, Bernard A,
Madisen L, Sunkin SM, Hawrylycz M, Koch C, Zeng H (2016) Adult mouse cortical cell
taxonomy revealed by single cell transcriptomics. Nat Neurosci 19:335–346
Tommiska J, Kansakoski J, Skibsbye L, Vaaralahti K, Liu X, Lodge EJ, Tang C, Yuan L,
Fagerholm R, Kanters JK, Lahermo P, Kaunisto M, Keski-Filppula R, Vuoristo S, Pulli K,
Ebeling T, Valanne L, Sankila EM, Kivirikko S, Laaperi M, Casoni F, Giacobini P, Phan-
Hug F, Buki T, Tena-Sempere M, Pitteloud N, Veijola R, Lipsanen-Nyman M, Kaunisto K,
Mollard P, Andoniadou CL, Hirsch JA, Varjosalo M, Jespersen T, Raivio T (2017) Two
missense mutations in KCNQ1 cause pituitary hormone deficiency and maternally inherited
gingival fibromatosis. Nat Commun 8:1289
Turan S, Bastepe M (2013) The GNAS complex locus and human diseases associated with loss-of-
function mutations or epimutations within this imprinted gene. Horm Res Paediatr 80:229–241
Villanueva C, Jacquier S, de Roux N (2012) DLK1 is a somato-dendritic protein expressed in
hypothalamic arginine-vasopressin and oxytocin neurons. PLoS One 7:e36134
Weinstein LS (2014) Role of G(s)α in central regulation of energy and glucose metabolism. Horm
Metab Res 46:841–844
Wijesuriya TM, De Ceuninck L, Masschaele D, Sanderson MR, Carias KV, Tavernier J, Wevrick R
(2017) The Prader-Willi syndrome proteins MAGEL2 and necdin regulate leptin receptor cell
surface abundance through ubiquitination pathways. Hum Mol Genet 26:4215–4230
Wu Z, Martinez ME, St Germain DL, Hernandez A (2017) Type 3 deiodinase role on central
thyroid hormone action affects the leptin-melanocortin system and circadian activity. Endocri-
nology 158:419–430
294 M. Pulix and A. Plagge

Xie T, Plagge A, Gavrilova O, Pack S, Jou W, Lai EW, Frontera M, Kelsey G, Weinstein LS (2006)
The alternative stimulatory G protein α-subunit XLαs is a critical regulator of energy and
glucose metabolism and sympathetic nerve activity in adult mice. J Biol Chem
281:18989–18999
Zeisel A, Munoz-Manchado AB, Codeluppi S, Lonnerberg P, La Manno G, Jureus A, Marques S,
Munguba H, He L, Betsholtz C, Rolny C, Castelo-Branco G, Hjerling-Leffler J, Linnarsson S
(2015) Brain structure. Cell types in the mouse cortex and hippocampus revealed by single-cell
RNA-seq. Science 347:1138–1142
Rhythmic Epigenetics in Neuroendocrine
and Immune Systems 11
Christopher S. Coyle, Elisabetta Tolla, and Tyler J. Stevenson

Abstract
Biological rhythms in neuroendocrine and immune systems are pervasive. Daily
and seasonal changes in day lengths regulate multiple physiological and immuno-
logical parameters in a diverse range of animals. A series of studies have shown
that epigenetic modifications exhibit naturally occurring rhythms across short- and
long-term timescales. In this chapter, we describe daily, estrous and seasonal
oscillations in epigenetic enzymes in neuroendocrine substrates, peripheral repro-
ductive tissues and immune cells (e.g. leukocytes). The predominant focus of
the chapter includes enzymes involved in DNA methylation and histone
modifications, such as DNA methyltransferase and histone deacetylases. The
findings presented herein highlight that epigenetic modifications can be permanent
as well as transient with long-term consequences on the timing of physiological
and behavioural processes. Moreover, the bidirectional interaction between the
immune system and the neuroendocrine nucleus that controls biological rhyth-
micity, the suprachiasmatic nucleus, emphasizes the need to understand rhythmic
changes in epigenetic enzymes and the consequences of disrupted daily and
seasonal rhythms.

Keywords
Methyltransferase · Acetylation · Oscillation · Circadian · Circannual

C. S. Coyle · E. Tolla · T. J. Stevenson (*)

Institute of Biodiversity, Animal Health and Comparative Medicine, University of Glasgow,
Glasgow, UK
e-mail: [email protected]

# Springer Nature Switzerland AG 2020 295

S. Wray, S. Blackshaw (eds.), Developmental Neuroendocrinology, Masterclass in
Neuroendocrinology 9, https://doi.org/10.1007/978-3-030-40002-6_11
296 C. S. Coyle et al.

11.1 Introduction

Biochemical plasticity in epigenetic modifications has significant potential, and

challenges, for our understanding of the mechanistic regulation of neuroendocrine
function. It is now well established that epigenetic modifications, DNA methylation
and histone acetylation, in particular, programme important genomic and chromatin
changes during development (Li 2002) and can exhibit transgenerational inheritance
(Heard and Martienssen 2014). Emerging research findings from multiple
laboratories have revealed that epigenetic modifications exhibit naturally occurring
oscillations across different periods of time. For example, daily and seasonal variation
in epigenetic enzyme expression occurs across the brain and peripheral tissues
(Stevenson 2018). Since epigenetic modifications are an essential component for
the regulation of gene transcription across taxa, rhythmic DNA methylation and
histone acetylation are likely an evolutionarily ancient system (Stevenson 2018). In
this chapter, we will present the historical context that provides the necessary
background information, and then describe the current research in naturally occurring
rhythms in epigenetic modifications. Due to the vast amount of knowledge available,
the chapter will not cover developmental epigenetics or transgenerational epigenetics
and instead focus on rhythmic epigenetics. The chapter will draw from the research
fields of chronobiology, reproductive neuroendocrinology and the neural control of
energy balance. The last section will provide a brief overview of the research
developments in the field of immunochronobiology and highlight the exciting poten-
tial for pharmaceutical reversal of epigenetic modifications for health and disease.

11.2 Historical Perspective on Epigenetics

in the Neuroendocrine Axes

The study of epigenetics originates in the field of developmental biology. Early work
by Conrad Waddington synthesized the concept that interactions between the envi-
ronment and the genetic material during embryogenesis and early life experiences
shaped the mature, complex organism (Waddington 1953). As the field advanced, it
became clear that experience-dependent modifications could alter an individual’s
fitness, and this was subsequently passed to the next generation; commonly referred
to as transgenerational inheritance. Here, the study of mitotically and/or meiotically
heritable changes in gene function was observed independent of changes in the
genome sequence (Riggs and Porter 1996). In parallel, the identities of enzymes
that act to catalyze the biochemical modifications of genome function (i.e. DNA
methylation and histone modification) were discovered to include: histone (de)
acetylases (Taunton et al. 1996) and DNA methyltransferases (Yoder et al. 1997).
Today, the next level of complexity focuses on RNA methylation and non-coding
RNA mechanisms (Guil and Esteller 2009). For the past 20 years, there has been an
exponential growth in our understanding of the environmental and hormonal regula-
tion of epigenetic enzymes (Stevenson 2017a), particularly in the field of cancer
genomics (Esteller 2008) with significantly less attention in the field of
11 Rhythmic Epigenetics in Neuroendocrine and Immune Systems 297

neuroendocrinology. One aim of this chapter is to present the current understanding

of the role of rhythmic DNA methylation and histone acetylation in timing neuroen-
docrine function and outline exciting avenues for future research directions.
The intensive investigation of hormone-driven changes in epigenetic
modifications in neuroendocrine substrates developed during the 1980s. Initial stud-
ies utilized pituitary tumour cell lines collected from Fischer 344 rats (Durrin et al.
1984). Rats that received chronic diethylstilbestrol treatment were shown to have
reduced methylation in the prolactin promoter, and associated increase in prolactin
mRNA expression. Remarkably, the effect of elevated diethylstilbestrol was
localized to the pituitary as the prolactin promoter retained higher methylation in
liver cells. These data indicated that hormones could induce epigenetic modifications
in anatomically localized neuroendocrine substrates. One of the first studies to
demonstrate a clear long-term behavioural effect of epigenetic modification within
neuroendocrine circuits was the observation of experience-dependent effects of rat
maternal licking-and-grooming behaviour (Weaver et al. 2004). High maternal
licking and grooming behaviour was found to significantly reduced DNA methyla-
tion and increased histone acetylation of the nerve growth factor-inducible protein A
(NGFI-A) consensus sequence in the hippocampal glucocorticoid receptor in off-
spring (Weaver et al. 2004). Over the past 35 years, the evidence has expanded to
illustrate that a wide range of hormones can have cellular- and neuroendocrine nuclei-
specific effects of epigenetic modifications during development (Nugent et al. 2015;
also see Chap. 15) and in adulthood (Stevenson 2017a). Indeed, the technological
developments that have significantly advanced genome sequencing have yielded
immense insights into the identification of specific genome motifs and chromatin
that are targets for hormone-driven epigenomic modifications. Since hormones are
facilitators of epigenetic modifications via epigenetic enzymes, here we will focus on
the hormone-dependent regulation of DNA methylation and histone acetylation
epigenetic enzymes in neuroendocrine tissues.

11.3 Function and Distribution of Epigenetic Enzymes

in the Neuroendocrine Axes

Epigenetic changes that impact the genome (i.e. DNA methylation) and chromatin
(i.e. histone acetylation) are induced by a range of enzyme families that show tissue-
specific expression profiles. DNA methylation on the genome template is regulated
by DNA methyltransferases (DNMT) and Ten-Eleven-Translocation methylcytosine
dioxygenase (TET) enzymes (Fig. 11.1a). DNMTs catalyze the removal of methyl
from S-adenosyl methionine (SAM) and preferentially add to cytosine-guanine
(CpG) residues, but also non-CpG residues (Jang et al. 2017). Conversely, TET
enzymes oxidize 5-methylcytosine that is then actively reverted to cytosine through
oxidation and thymine DNA glycosylase (TDG)-mediated base excision repair
(Kohli and Zhang 2013). Both DNMT and TET enzymes have a number of different
isoforms that have distinct structure–function relations that contribute to the devel-
opmental- and cell-specificity of epigenetic modifications. In adults, DNMT1, 3a, and
298 C. S. Coyle et al.

A SAM SAH
B H2O Acetate

DNMT HDAC

euchromatin heterochromatin
TET HAT

CO2+ O2+ CoA Acetyl-CoA

Succinate α-KG

unmethylated gDNA methylated gDNA acetylated histone deacetylated histone

Fig. 11.1 Functional pathways of epigenetic enzymes. (a) DNA methyltransferase (DNMT)
enzymes catalyze the addition of methyl groups, depicted by red dots, on to the genome template.
Ten Eleven Translocation (TET) enzymes initiate a cascade of biochemical changes that include the
oxidation of 5-methylcytosine and promote locus-specific reversal of DNA methylation indicated
by grey dots. (b) Epigenetic enzymes involved in chromatin modifications include histone
deacetylase (HDAC) and histone acetylase (HAT). HDAC triggers the removal of acetyl groups
from histone. HAT enzymes catalyze the removal of acetyl groups from Acetyl-CoA and add acetyl,
indicated by red curvilinear lines, on to histones. Abbreviations: S-adenosyl methionine (SAM), S-
adenosyl homocysteine (SAH), carbon dioxide (CO2), alpha-ketoglutarate (αKG), coenzyme A
(CoA)

3b exhibit wide anatomical distribution with high levels of expression in the brain,
reproductive tissues, and immune glands (e.g. spleen) (Fig. 11.2). The TET1-3
enzymes also show high expression levels in the brain and female reproductive
tissues (i.e. ovary) (Fig. 11.2).
Chromatin modifications include histone acetylation induced by histone
acetylases (HAT/KAT) and the converse histone deacetylase (HDAC) enzymes
that consist of several isoforms (Klose and Bird 2006). HAT enzymes catalyze the
removal of acetyl from acetyl-coenzyme A (CoA) and induce the transformation from
closed heterochromatin into the transcriptional accessible euchromatin states
(Fig. 11.1b). The removal of acetyl groups from histones by HDACs results in
silenced gene transcription and the formation of cellular acetate. Similar to the
enzymes involved in DNA methylation, both HDAC and HAT enzymes show higher
levels of expression in the brain, reproductive tissues and immune glands (Fig. 11.2).

11.4 Predicable Oscillations of Epigenetic Enzymes Indicate

Rhythmic Modifications

The initial evidence to indicate that epigenetic enzymes display oscillations were
derived from high-throughput cDNA microarrays and RNA sequencing platforms
(Stevenson 2018). The current evidence suggests that daily epigenetic enzyme
oscillations are isoform-specific and tissue-dependent. For example, in the adult
11 Rhythmic Epigenetics in Neuroendocrine and Immune Systems 299

Lung Brain
DNMT3a TET2 DNMT1 TET1
HDAC3 TET3 DNMT3a TET2
KAT2a DNMT3b TET3
HDAC1 HAT1
Heart HDAC2 KAT2a
DNMT3a TET1 HDAC3 KAT2b
HDAC3 KAT2b
Spleen
Lymph DNMT1 TET2
DNMT1 HAT1 DNMT3a TET3
DNMT3a DNMT3b HAT1
HDAC1 HDAC1 KAT2a
HDAC3 HDAC3 KAT2b

Liver Ovary
DNMT1 HAT1 DNMT1 TET1
DNMT3b KAT2b DNMT3a TET2
HDAC1 DNMT3b TET3
HDAC2 HDAC2 KAT2a
HDAC3 KAT2b
Adipose
TET1 Uterus
TET2 DNMT1
KAT2a DNMT3a
KAT2b DNMT3b
HDAC1
HDAC2
Testis HDAC3
DNMT1 HAT1
DNMT3a KAT2a
DNMT3b Placenta
HDAC1 DNMT1 TET1
HDAC2 DNMT3a TET2
HDAC3 DNMT3b TET3
HDAC1 HAT1
HDAC2 KAT2a
HDAC3

Fig. 11.2 Distribution and relative abundance of epigenetic enzymes. A representative diagram to
highlight the tissue-specific and expression levels of epigenetic enzymes in multiple tissues.
Epigenetic enzymes presented in tissue boxes were selected based on higher than average Reads
Per Kilobase Million (RPKM) expression levels across all tissues. Most enzymes involved in DNA
methylation and histone modifications are expressed in neuroendocrine substrates (i.e. brain) as
well as reproductive tissues (i.e. testes, ovary and uterus). RPKM data were obtained from
PUBMED and are based on Fagerberg et al. (2014)
300 C. S. Coyle et al.

mouse suprachiasmatic nucleus (SCN), the master neuroendocrine nucleus for the
regulation of daily rhythms (Hastings et al. 2014), DNMT3a is expressed, exhibits
daily rhythms in mRNA levels and serves to maintain the integrity of circadian
oscillations in genome function (Azzi et al. 2014). In the adult male Siberian hamster
there are robust daily oscillations in hypothalamic dnmt3a and hdac4 expression
with peak levels during the dark phase that is followed by a precipitous decline after
lights on (Stevenson 2017b). However, not all epigenetic enzymes exhibit daily
variation in neuroendocrine expression, and include DNMT1, DNMT3b, HDAC1,
HDAC2 and HDAC3. The daily variation in epigenetic enzyme expression
(e.g. DNMT3a) is probably driven by the circadian clock gene aryl hydrocarbon
receptor nuclear translocator-like 1 (BMAL1) proteins as there are 3 BMAL1
binding sites in the DNMT3a promoter and there is robust daily variation in
BMAL1 binding in the DNMT3a promoter (Fig. 11.3a) (Koike et al. 2012).
In addition to daily rhythms, epigenetic enzymes have been shown to exhibit
robust variation across annual scales. Most experimental evidence shows that neu-
roendocrine epigenomic plasticity is a major contributor to the seasonal phenotypic
variation in mammalian and avian species. In the Siberian hamster brain, there is a
significantly higher level of global DNA methylation in the hypothalamus during the
summer breeding conditions compared to the winter non-breeding conditions
(Stevenson and Prendergast 2013). The decrease in global DNA methylation occurs
in parallel to reduced daily dnmt1 and dnmt3a levels (Fig. 11.3b) (Stevenson 2017b).
These data indicate that reduced hypothalamic DNA methylation may be an endog-
enous circannual mechanism for the timing of seasonal life–history transitions
(Helm and Stevenson 2015; Stevenson and Lincoln 2017). There are also significant
seasonal oscillations in DNMT expression in hamster reproductive tissues. During
the breeding period, DNMT3a levels are significantly reduced in the testes and
uterine tissue (Lynch et al. 2016) (Fig. 11.3c), liver (Alvarado et al. 2015), skeletal
muscle (Alvarado et al. 2015) and circulating leukocytes (Stevenson et al. 2014). In
humans, there are also marked seasonal changes in DNA methylation. Blood
samples revealed that in spring and summer months, there were significantly higher
levels of DNA methylation in the promoters for cancer-related genes (RASSF1A and
MGMT) compared to autumn and winter months (Ricceri et al. 2014). The exact
mechanisms for how external information, such as day length, regulates epigenetic
enzyme expression are not well known. In addition to day length, supplementary
cues including temperature and diet (see below), likely contribute to the regulation of
epigenetic enzyme expression in a tissue-dependent manner. The two primary
hormonal candidates that provide an internal code of environmental cues are mela-
tonin and gonadal-derived oestrogen (Stevenson 2017a). In human embryonic
kidney cells (HEK293), the administration of melatonin induced an increase in
dnmt3a expression (Fig. 11.3d). Given that melatonin is a reliable physiological
measurement of ‘night-length’ a proxy for night duration, these data suggest that the
annual change in day length drives melatonin-dependent entrainment of epigenetic
enzyme expression, and DNMT3a in particular. In ovariectomized hamsters, a bolus
injection of estradiol and progesterone (E2P4) significantly reduced dnmt3a expres-
sion (Fig. 11.3e). It is likely that oestrogen actions provide supplementary regulation
11 Rhythmic Epigenetics in Neuroendocrine and Immune Systems 301

A)
DNMT3a promoter

BMAL1 binding sites (E-BOX) ATG

1A 1B 1C 2

DNMT3a exons

B) C)
10 4
LD
SD

dnmt3a expression
mRNA expression

8
3

6
2

1
2

0 0
dnmt1 dnmt3a dnmt3b Testes Ovary Uterus

D) E)
4 6

5
dnmt3a expression
dnmt3a expression

3
4

2 3

2
1
1

0 0
Saline 1nM 10nM 100nM Oil 12hr 24hr

Fig. 11.3 Photoperiodic and hormonal regulation of DNA methyltransferase 3a. DNMT3a is one
epigenetic enzyme that exhibits consistent rhythmic expression in multiple nuclei and tissues. (a)
The dnmt3a promoter has several binding motifs for the circadian clock gene BMAL1. Koike et al.
(2012) confirmed that BMAL1 directly binds to transcriptional regulatory regions for dnmt3a and
shows daily changes with peak activity during the late dark phase. (b) Hypothalamic dnmt1 and
dnmt3a expressions display photoperiod-dependent regulation in which long summer-like day
lengths (LD) stimulate higher levels and short winter-like days (SD) reduce levels. (c) Conversely,
there is an increase in dnmt3a expression in the testes and uterine tissue of SD reproductively
regressed hamsters. The annual change in photoperiod is encoded by high amplitude changes in the
duration of melatonin and plasma oestrogen concentrations. (d) Melatonin was sufficient to induce
dnmt3a expression in HEK293 cells. (e) A single bolus injection of E2P4 to ovariectomized
hamsters significantly reduced uterine dnmt3a expression. Data are adapted from Stevenson
(2017a, b) and Lynch et al. (2016)
302 C. S. Coyle et al.

of enzyme expression that fine-tunes the timing of epigenome modifications. The

precise genomic targets for the annual change in DNA methylation have yet to be
fully elucidated. In mammalian species, there are robust annual changes in the
reproductive neuroendocrine peptides Kisspeptin (Greives et al. 2008; Stevenson
et al. 2012) and RF-related peptide-3 (Greives et al. 2008). Whereas in avian species,
the annual change in reproduction is primarily driven by the photoperiodic regula-
tion of gonadotropin-releasing hormone (GnRH) cells (Stevenson et al. 2013). These
reproductive neuropeptides are excellent candidates for rhythmic epigenetic, and in
particular, the control of seasonal reproduction (Kurian and Terasawa 2013; also see
Chap. 9).

11.5 Rhythmic Epigenetic Enzymes and the Neuroendocrine

Control of the Estrous Cycle

One of the well-described neuroendocrine reproductive biological rhythms across

vertebrates is the female reproductive cycle. Over several decades, the behavioural,
morphological, hormonal, cellular and molecular oscillations have been characterized
in mammalian species (Bronson 1990). The female reproductive cycle serves to
optimize the timing of fertilization and reproductive behaviour (Pfaff et al. 2004).
In rodents, there is typically a 4–5 day estrous cycle that includes complex coordina-
tion of multiple hormonal, morphological and behavioural changes. Rodentia exhibit
four key estrous stages that include diestrous, proestrous, estrous and metestrous
(Westwood 2008). A robust method to monitor the estrous cycle in rats and mice uses
vaginal cytology. A simple vaginal smear can clearly delineate the estrous stage based
on the absence, presence or proportion of neutrophils, small and large nucleated
epithelial cells and anucleated keratinized epithelial cells (Cora et al. 2015). In
parallel, the endometrium of the uterus exhibits large-scale programmed remodelling
that involves cellular proliferation, apoptosis, angiogenesis and leukocyte infiltration
(Evans et al. 1990). During the transition from diestrous–proestrous–estrous, there is
an extensive growth of uterine tissue followed by a subsequent decrease in size
(Lynch et al. 2016; Fig. 11.4a). These morphological and cellular changes are driven
by the local production of sex steroids 17β-estradiol (E2) and progesterone (P4)
(Wood et al. 2007) (Fig. 11.4b).
The epigenetic enzymes dnmt1, dnmt3a and dnmt3b are predominantly localized
in the epithelial layer in ovaries (Ahluwalia et al. 2001), and endometrium cells in the
uterus (Van Kaam et al. 2011), and may function to time fertility (Roy and Matzuk
2006). HDAC1 and HDAC2 are expressed in oocytes, and conditional ovarian
knockouts of hdac1/hdac2 in the ovary causes inhibition of follicle development
past the secondary stage (Ma et al. 2012). Recent studies have examined cyclical
changes in epigenetic enzymes during the estrous cycle and have identified marked
plasticity during the transition from proestrous to estrous stages (Fig. 11.4c). For
example, uterine dnmt3a and dnmt3b enzymes are significantly reduced during late
proestrous and early estrous (Lynch et al. 2016). Uterine dnmt3a has been shown to
be important for decidualization and shows a transient oestrogen-dependent decrease
11 Rhythmic Epigenetics in Neuroendocrine and Immune Systems 303

Uterus
Morphology

Progesterone
Hormone

Estradiol

HDAC2
Epigenetic

DNMT3b
DNMT3a

Diestrus Diestrus Proestrus Estrus

Fig. 11.4 Epigenetic enzymes oscillate in uterine tissues across the estrous cycle. The female
rodent estrous cycle is characterized by robust changes in morphology, plasma hormones and
uterine epigenetic enzymes. (Upper Panel) During the estrous cycle, uterine mass significantly
increases during the transition from proestrous to estrous stages. On the evening of proestrous, there
is a significant increase in the probability and intensity of female reproductive behaviours. The
green bar depicts ‘behavioural estrous’, the period of high proceptivity and receptivity in females.
(Middle Panel) Cycling changes in ovarian hormones estradiol and progesterone, the primary
hormones that induce marked morphological changes in the uterus and transcriptional activation
of several genes involved in the control of reproductive physiology. (Lower Panel) Three epigenetic
enzymes in the uterus, HDAC2, DNMT3a and DNMT3b, were found to be inhibited by estradiol/
progesterone. Data are adapted from Lynch et al. (2016, 2017)
304 C. S. Coyle et al.

in expression (Logan et al. 2013). In the hamster, there is a significant decrease in

dnmt3a during estrous and was significantly reduced in ovariectomized animals in
response to a single bolus of E2P4 (Lynch et al. 2016) (Fig. 11.3e). Parallel changes in
histone modifications are also present as HDAC1/2 are expressed in endometriotic
and endometrial stromal cells and exhibit differential regulation by E2 and combined
E2P4 (Colón-Díaz et al. 2012). Similarly, Jefferson and colleagues (2013) revealed
that diethylstilbestrol inhibited hdac2 mRNA expression; yet uterine hdac1, 2 and
3 did not change at postnatal day 5. To date, the functional significance of estrous
cycling in epigenetic enzymes has not been established.

11.6 Neuroendocrine Control of Daily Energy Balance and Daily

Rhythms in Epigenetic Enzymes

There are multiple systems that control energy intake and expenditure. A complex
interaction of neuroendocrine substrates, adipose tissue, muscle cells, pancreatic
cells and liver cells serve to maintain a homeostatic steady-state condition (Yeo
and Heisler 2012). Daily rhythms in epigenetic enzymes have been characterized in
discrete hypothalamic nuclei and peripheral tissues that suggest a complex
epigenomic reprogramming occurs daily to ensure optimal energy balance (Asher
and Sassone-Corsi 2015). The development of a freely available, online resources
that contain massive datasets based on high-throughput methods (e.g. microarray,
RNA sequencing), such as CircaDB (circadb.hogeneschlab.org), provide a valuable
opportunity to interrogate daily rhythms in epigenetic enzyme RNA expression
across multiple tissues. For example, CircaDB analyzes reveal that brown and
white adipose tissues show daily rhythms in hdac6 and kat2b; which coincidentally
have anti-phase expression waveforms (Pizarro et al. 2013). hdac6 expression
peaked during the mid-point of a 12:12 LD cycle, whereas kat2b peaks during the
dark phase. The enzyme expression patterns here suggest a daily switch in histone
(de)acetylation occurs that is not driven by food intake and instead, reflects an
endogenous timing mechanism. In the liver, tet2 and dnmt3b peak prior to lights
on; which suggest the potential for daily variation in epigenetic tone; high levels of
both methylation and demethylation prior to lights on followed by a subsequent
decline in epigenomic activity during the dark phase. However, other research has
shown that liver dnmt3a peaks during the mid-point of the light phase (Hughes et al.
2009). Conversely, liver hdac5 peaks during the dark phase, and hdac6 peaks during
the light phase.
There is growing evidence that the time of food intake has a remarkable level of
bidirectional interactions between the circadian clock and energy balance (Asher and
Sassone-Corsi 2015). The vast majority of research has examined the contribution of
chromatin remodelling via the activity of sirtuin proteins (SIRT1) within well-
defined hypothalamic nuclei involved in the regulation of metabolism, appetite
and food intake (Asher and Sassone-Corsi 2015). SIRT1 removes histone acetyl
groups via a nicotinamide adenine dinucleotide (NAD+)-dependent manner.
Increased SIRT1 activity in the ventromedial nucleus (VMN) of the hypothalamus
11 Rhythmic Epigenetics in Neuroendocrine and Immune Systems 305

leads to histone deacetylation and can directly interact with the circadian clock gene
CLOCK; and thereby, regulates clock-controlled genes in a histone deacetylation-
dependent manner (Nakahata et al. 2008; Orozco-Solis et al. 2015). The current
evidence suggests that SIRT1 expression specifically within VMN Sf1 neurons is
essential to coordinate circadian food intake and daily energetic state.

11.7 Daily Rhythmic Epigenetic Modifications and Disease

Immune responses are dynamically regulated over daily and seasonal timescales
(Stevenson and Prendergast 2015). This section will give an overview of the ongoing
research regarding the role of epigenetic mechanisms in immune function, focusing
on evidence that indicates reversible modifications. The epigenetic enzymes
described above have been implicated in pathological conditions and progression,
e.g. cancer, and have become a powerful tool in disease treatment (Widschwendter
et al. 2018). One tantalizing conjecture is the potential loss of natural rhythmic
epigenetic modifications as a contributing factor to disease progression. Therefore,
the investigation of natural rhythmic variation in epigenetic modifications can give
clinical researchers insight into the factors that could be involved in enzyme
disruption, as well as aiding in the development of novel reversible treatment
strategies.
It is becoming clear that immune function (e.g. cancer) and biological rhythms
(i.e. circadian) have bidirectional interactions. The increased attention to disrupted
circadian rhythms has resulted in the identification that entrained daily rhythms
contribute to health and well-being (Savvidis and Koutsilieris 2013). Leukocyte
migration from lymph nodes, lymph and plasma shows daily rhythms with peak
CD4+, CD8+ T cells and B cell levels during the light phase in mice (Druzd et al.
2017) and Siberian hamsters (Prendergast et al. 2013). Disruption of circadian
timing induces leukocyte arrhythmia, reduced daily amplitudes in circadian clock
gene and inhibited CD11+ dendritic cells, indicating that a functional SCN is
required for daily variation in immune function (Prendergast et al. 2013). Likewise,
there are distinct seasonal rhythms in plasma leukocytes numbers and leukocyte
transcriptomic profiles in hamsters and humans, respectively (Bilbo et al. 2002;
Banks et al. 2016; Dopico et al. 2015). During winter photoperiods, there are
significantly elevated circulating blood leukocytes, T cells and natural killer cells
(Bilbo et al. 2002). The winter enhancement of immune function and survival is
common across species and serves to increase survival (Nelson 2004). The increase
in many chronic diseases and pathological states has been attributed to not only
disrupted circadian rhythms as described above, but also disruption in seasonal
immune rhythms (Stevenson et al. 2015).
A few studies have examined the daily or seasonal variation of epigenetic enzyme
expression in immune tissues or cells. A massive transcriptomic analysis of seasonal
waveforms of human leukocytes identified that dnmt1, hdac5, -6 and -9 vary across
the year (Dopico et al. 2015). In Siberian hamsters, short photoperiods resulted in
higher dnmt3b leukocyte expression (Stevenson et al. 2014). The increased dnmt3b
306 C. S. Coyle et al.

was associated with higher methylation levels in the proximal promoter of

deiodinase type-3 (dio3), an enzyme involved in thyroid hormone catabolism.
How rhythms in epigenetic enzymes relate to immunoenhancement or immunosup-
pression is not well known, the best evidence can be derived from the research field
in autoimmune diseases and cancer.
Autoimmunity has been associated with DNA hypomethylation in a number of
studies. For instance, in both animal models (Quddus et al. 1993) and human cells
(Richardson et al. 1990), treatment with DNA methylation inhibitor drugs
(e.g. 5-Azacytidine) causes the development of autoimmune conditions equal or
comparable to Systemic Lupus Erythematosus (SLE). SLE progression has been
mainly attributed to reduced DNA methylation in T cells (Richardson et al. 1990)
that causes, in turn, overexpression of adhesion molecules such as lymphocyte
function-associated antigen 1 (LFA-1) (Cornacchia et al. 1988). Increased LFA-1
expression appears to result in immune cell autoreactivity leading to SLE (Kaplan
et al. 2000). The ability to selectively reinstate methylation of LFA-1 in T cells at
specific daily phases may provide one solution to reduce the incidence and progres-
sion of autoimmune diseases such as SLE.
Once the differentiation of immune cells such as T lymphocytes was closely
related to chromatin structure changes (Agarwal and Rao 1998), scientists started to
investigate the precise role of DNMT1 within lymphocyte organization. Naive
lymphocytes undergo a series of differentiations to become either B or T cell, to
express different types of T cell surface receptors (e.g. TCRαβ or TCRγδ) and
lineage markers (e.g. CD4+, CD8+) (Richter et al. 1999). CD4+ cells then are able
to differentiate into regulatory (Treg) cells. An essential role of Treg cells is to
alleviate the immune response and, for instance, prevent an extremely elevated
immune reaction that could lead to autoimmune diseases. Lee et al. (2001) utilized
a Cre/Lox system to generate dnmt1 knockout in thymocytes, and the lack of
DNMT1 enzyme activity resulted in compromised T cell development and survival.
Josefowicz et al. (2009) also examined the functional role of DNMT1 for Treg
cell lineage regulation and Forkhead box P3 ( foxp3), in particular, foxp3 is
important for the stimulation of the transition of CD4+ cells into the Treg cell
lineage (Fontenot et al. 2003). However, when DNMT1 is absent, the foxp3 pro-
moter is hypomethylated and increased FOXP3 in CD8+ cells rather than CD4+
cells. Therefore, the absence of dnmt1 may lead to the progression of CD8+ cells into
Treg cells and thus, an autoimmune response. Therefore, DNMT1 provides a system
for confining foxp3 cell specificity to the CD4+ lineage, ensuring normal Treg cell
development (Josefowicz et al. 2009). It is unknown whether the loss of dnmt1 in
Treg cells leads to the reduction in daily oscillations in cells, endogenous circadian
clock signalling and/or inhibition of rhythmic epigenetic modifications.
Despite the undeniable importance of DNMT1 in immune function, other
methylating enzymes may also play an important role in the daily timing of essential
regulatory genes. DNMT3a and DNMT3b are considered the de novo DNA
methyltransferases, as these enzymes establish new epigenetic patterns in cells
(Okano et al. 1999). Organisms retain functional levels of dnmt3a and dnmt3b
within immune cells well into adulthood (Fitzpatrick and Wilson 2003). Their
11 Rhythmic Epigenetics in Neuroendocrine and Immune Systems 307

activity within the immune response has been explored both in vitro and in vivo
(Okano et al. 1999; Fitzpatrick and Wilson 2003). Similar to DNMT1, their expres-
sion is required for normal development and cell survival, as their loss results in
embryonic lethality (Okano et al. 1999).
It is clear that both genetics and epigenetics play a role in the development of the
majority of neuroendocrine cancers. One the ‘10 hallmarks of cancer’ outlined by
Hanahan and Weinberg (2011) is genomic instability that results from alterations in
the epigenome (Esteller 2008; Hanahan and Weinberg 2011), including both global
hypomethylation (Gaudet et al. 2003) and gene-specific hypermethylation (Herman
and Baylin 2003). High expression of dnmts has been shown in a variety of cancers,
e.g. neuroblastoma, prostate cancer (Kautiainen and Jones 1986). Several studies
have assessed the function of DNMT1 through both pharmacological inhibition and
gene knockdown in vitro and in vivo. DNMT1 is required for development in mice,
as dnmt1 / is lethal (Li et al. 1992). This has been attributed to the enzyme’s
essential function in maintaining DNA methylation patterns in daughter cells
through generations (Lei et al. 1996), a function necessary for normal survival and
growth. Jackson-Grusby et al. (2001) observed that mice dnmt1 / embryonic stem
cell-derived fibroblasts undergo a p53-dependent cell death. p53 is a tumour sup-
pressor gene, involved in inhibiting cell cycle progression when errors are detected
in the genome (Symonds et al. 1994). When methylated, p53 expression is silenced
and results in survival of cancerous cells (Jackson-Grusby et al. 2001). Knox et al.
(2000) monitored the effects of dnmt1 / in human lung cancer cells and found that
hypomethylation due to the absence of dnmt1 also inhibits cell proliferation through
apoptosis. The transcription of other tumour suppressor genes, such as p16 and p21,
was found to be upregulated after using pharmacological inhibitors of DNMT1
enzymatic activity in the human bladder (Fournel et al. 1999) and colon (Rhee
et al. 2000) carcinoma cells, preventing tumour progression. It is intriguing that both
expression of tumour suppressor genes (Miki et al. 2013) and epigenetic enzymes
has been shown to vary in a circadian manner. p53 is able to regulate the cell cycle is
via ‘clocking’ genes, by inhibiting the per2 gene through interaction with BMAL1
and CLOCK (Miki et al. 2013; Fu et al. 2002). Repression of per2 expression has
been associated with overexpression of c-MYC proto-oncogene (cMyc), a gene that
codes for transcription factors involved in cell cycle progression (Pelengaris et al.
2002). cMyc directly binds to DNMT3a to promote cell growth (Brenner et al.
2005). It is then possible that circadian rhythms dictate patterns of certain epigenetic
components, resulting in cancer development. However, the exact mechanisms
underlying the circadian control of epigenetic signals in tumourigenesis are still
largely unknown.
Aberrant expression of HAT and HDAC activity also contributes to the rise and
progression of specific diseases (Jones and Baylin 2007) and include cancer (Cress
and Seto 2000), Alzheimer’s disease (Francis et al. 2009) and mental retardation
(Coffee et al. 1999). HDAC activity has been shown to be linked to the expression of
certain eukaryotic cell cycle regulation factors, such as retinoblastoma (RB) protein
(Brehm et al. 1998). Disrupted interaction between HDAC proteins and RB proteins
prevents the apoptosis of cancerous cells, allowing tumourigenesis (Xu et al. 1993).
308 C. S. Coyle et al.

Direct downregulation of HDACs via chemical inhibitors is also currently being

explored. Use of HDAC inhibitors have been linked to cell cycle arrest in T cell
lymphoma (Olsen et al. 2007) and a reduction in cancerous cells in an acute myeloid
leukaemia mouse model (Bots et al. 2014). Altogether, these data indicate that a
greater understanding of the timing of epigenetic modifications, particularly daily
rhythms can provide significant improvements for healthy immune function and the
treatment of pathological conditions.

11.8 Future Perspectives

It has been suggested that an ‘epigenetic homeostasis’ exists and is needed for
normal cellular function that is characterized by a balance between DNA (de)
methylation as well as histone (de)acetylation (Jones and Baylin 2007). In this
chapter, we have focused on ‘epigenetic plasticity’, the transient modifications to
the genome and/or chromatin template that leads to a functional change in the
neuroendocrine control of reproduction, energy balance and immune function. The
predominant difference between these philosophical positions is that the disruption
of epigenetic homeostasis leads to long-term consequences in cellular function, such
as chronic diseases while the epigenetic plasticity model argues that the short- and
long-term oscillations in epigenetic enzymes maintain cyclical cell- and tissue-
function that is required for naturally occurring, normal daily, estrous and seasonal
rhythms (Stevenson 2018). A greater understanding of the mechanistic and func-
tional basis of the latter can yield significant insights into the genetic control of
animal health and wellness. Below, we outlined two major perspectives for future
research.
The field of epigenetic research is exciting as ongoing findings reveal the massive
potential for basic, translational and applied research. The advent of high-throughput
sequencing platforms such as Whole-Genome Sequencing (WGS), transcriptomic,
proteomic assays has yielded immense information on the complexity of molecular
and cellular signalling pathways in neuroendocrine and immune cells. Moreover, the
chemical modification of genome structure using sodium bisulfite now provides the
capability to identify specific nucleotide residues/motifs that show variation in
epigenetic modifications (i.e. methylation). This chapter has focused on the rhythmic
expression of multiple enzymes involved in epigenetic modifications. The data show
tissue-, cell- and time-dependent oscillations; however, our understanding of the
downstream effects is limited. The next steps are to establish the precise genomic
and chromatin location of daily, estrous and seasonal epigenetic modifications. The
combination of sequencing platforms, such as Illumina or PacBIO, provides a
valuable opportunity to address wherein the genome epigenetic modifications
exhibit rhythmic variation; and thus, novel gene regulatory mechanisms within
tissues and more importantly at a single-cell resolution.
Another advantage of sequencing tools to identify epigenetic modifications is the
development of individual or personalized approaches to medical treatments (Rasool
et al. 2015). Indeed, further investigation into the exact genomic and chromatin
11 Rhythmic Epigenetics in Neuroendocrine and Immune Systems 309

modifications that underlie daily and seasonal plasticity in reproduction, energy

balance and immune function is warranted. The ability to design epigenetic
constructs that selectively (de)methylate target genome regions that is provided
within an optimal daily window has the potential to make significant health and
wellness advancements for both livestock animals and humans.

Key References

Azzi et al. (2014). This study shows that daily oscillations in SCN DNA methylation
regulates circadian clock function.
Bilbo et al. (2002). This is an exciting report on the seasonal regulation of immune
function.
Dopico et al. (2015). This study reveals large annual changes in leukocyte
transcriptomes in humans.
Koike et al. (2012). One key finding identified in this paper were daily rhythms in
BMAL1 binding on the dnmt3a promoter.
Lynch et al. (2016). This report illustrated increased dnmt3a expression was nega-
tively associated with oscillations in fertility.
Nugent et al. (2015). This study discovered sex-differences in neuroendocrine
DNMT3a and a lack of expression facilitated masculinization.
Stevenson and Prendergast. (2013). This study established that rhythmic neuroen-
docrine DNA methylation underlies seasonal timing of reproduction.

References
Agarwal S, Rao A (1998) Modulation of chromatin structure regulates cytokine gene expression
during T cell differentiation. Immunity 9:765–775
Ahluwalia A, Hurteau JA, Bigsby RM, Nephew KP (2001) DNA methylation in ovarian cancer.
II. Expression of DNA methyltransferases in ovarian cancer cell lines and normal ovarian
epithelial cells. Gynecol Oncol 82:299–304
Alvarado S, Mak T, Liu S, Storey KB, Szyf M (2015) Dynamic changes in global and gene-specific
DNA methylation during hibernation in adult thirteen-lined ground squirrels, Ictidomys
tridecemlineatus. J Exp Biol 218:1787–1795
Asher G, Sassone-Corsi P (2015) Time for food: the intimate interplay between nutrition, metabo-
lism and the circadian clock. Cell 161:84–92
Azzi A, Dallmann R, Casserly A, Rehrauer RH, Patrignani A, Maier B, Kramer A, Brown SA
(2014) Circadian behavior is light-programmed by plastic DNA methylation. Nat Neurosci
17:377–382
Banks R, Delibegovic M, Stevenson TJ (2016) Photoperiod- and triiodothyronine-dependent
regulation of reproductive neuropeptides, proinflammatory cytokines, and peripheral physiol-
ogy in Siberian hamsters (Phodopus sungorus). J Biol Rhythm 31:299–307
Bilbo SD, Dhabhar FS, Viswanathan K, Saul A, Yellon SM, Nelson RJ (2002) Short day lengths
augment stress-induced leukocyte trafficking and stress-induced enhancement of skin immune
function. Proc Natl Acad Sci U S A 99:4067–4072
310 C. S. Coyle et al.

Bots M, Verbrugge I, Martin B, Salmon J, Ghisi M, Baker A, Stanley K, Shortt J, Ossenkoppele G,

Zuber J, Rappaport A, Atadja P, Lowe S, Johnstone R (2014) Differentiation therapy for the
treatment of t(8;21) acute myeloid leukemia using histone deacetylase inhibitors. Blood
123:1341–1352
Brehm A, Miska EA, McCance DJ, Reid JL, Bannister AJ, Kouzarides T (1998) Retinoblastoma
protein recruits histone deacetylase to repress transcription. Nature 391:597–601
Brenner C, Deplus R, Didelot C, Loriot A, Viré E, De Smet C, Gutierrez A, Danovi D, Bernard D,
Boon T, Pelicci PG (2005) Myc represses transcription through recruitment of DNA
methyltransferase corepressor. EMBO J 24:336–346
Bronson FH (1990) Mammalian reproductive biology. University of Chicago Press, Chicago
Coffee B, Zhang F, Warren ST, Reines D (1999) Acetylated histones are associated with FMR1 in
normal but not fragile X-syndrome cells. Nat Genet 22:98–101
Colón-Díaz M, Báez-Vega P, García M, Ruiz A, Monteiro JB, Fourquet J, Bayona M, Alvarez-
Garriga C, Achille A, Seto E, Flores I (2012) HDAC1 and HDAC2 are differentially expressed
in endometriosis. Reprod Sci 19:483–492
Cora MC, Looistra L, Travlos G (2015) Vaginal cytology of the laboratory rat and mouse: review
and criteria for the staging of the estrous cycle using stained vaginal smears. Toxicol Pathol
43:776–793
Cornacchia E, Golbus J, Maybaum J, Strahler J, Hanash S, Richardson B (1988) Hydralazine and
procainamide inhibit T cell DNA methylation and induce autoreactivity. J Immunol
140:2197–2200
Cress WD, Seto E (2000) Histone deacetylases, transcriptional control, and cancer. J Cell Physiol
184:1–16
Dopico XC, Evangelou M, Ferreira RC, Guo H, Pekalski M, Smyth DJ, Cooper N, Burren OS,
Fulford AJ, Hennig BJ, Am P, Ziegler AG, Bonifacio E, Wallace C, Todd JA (2015) Wide-
spread seasonal gene expression reveals annual differences in human immunity and physiology.
Nat Commun 6:7000
Druzd D, Matveeva O, Ince L, Harrison U, Wenyan H, Schmal C, Herzel H, Tsang AH,
Kawakami N, Leliavske A, Uhl O, Yao L, Sander LE, Chen CS, Kraus K, de Juan A, Hergenhan
SM, Ehlers M, Koletzko B, Haas R, Solbach W, Oster H, Scheiermann C (2017) Lymphocyte
circadian clocks control lymph node trafficking and adaptive immune responses. Immunity
46:120–132
Durrin LK, Weber JL, Gorksi J (1984) Chromatin structure, transcription, and methylation of the
prolactin gene domain in pituitary tumors of Fischer 344 rats. J Biol Chem 259:7086–7093
Esteller M (2008) Epigenetics in cancer. N Engl J Med 358:1148–1159
Evans GS, Gibson DF, Roberts SA, Hind TM, Potten CS (1990) Proliferative change in the genital
tissues of female ice during the oestrous cycle. Cell Tissue Kinet 23:619–635
Fagerberg L, Hallstrom BM, Oksvold P, Kampf C, Djureinovic D, Odeberg J, Habuka M,
Tahmasebpoor S, Danielsson A, Edlund K, Asplund A, Sjostedt E, Lundberg E, Szigyarto
CA, Skogs M, Takanen JO, Berling H, Tegel H, Mulder J, Nilsson P, Schwenk JM, Lindskog C,
Danielsson F, Mardinoglu A, Sivertsson A, von Feilitzen K, Forsberg M, Zwahlen M, Olsson I,
Navani S, Huss M, Nielsen J, Ponten F, Uhlen M (2014) Analysis of the human tissue specific
expression by genome-wide integration of transcriptomics and antibody-based proteomics. Mol
Cell Proteomics 13:397–406
Fitzpatrick DR, Wilson CB (2003) Methylation and demethylation in the regulation of genes, cells,
and responses in the immune system. Clin Immunol 109:37–45
Fontenot JD, Gavin MA, Rudensky AY (2003) Foxp3 programs the development and function of
CD4+ CD25+ regulatory T cells. Nat Immunol 4:330
Fournel M, Sapieha P, Beaulieu N, Besterman JM, MacLeod AR (1999) Down-regulation of human
DNA-(Cytosine-5) methyltransferase induces cell cycle regulators p16 ink4A and p21WAF/
Cip1 by distinct mechanisms. J Biol Chem 274:24250–24256
Francis YI, Fà M, Ashraf H, Zhang H, Staniszewski A, Latchman DS, Arancio O (2009)
Dysregulation of histone acetylation in the APP/PS1 mouse model of Alzheimer’s disease. J
Alzheimers Dis 18:131–139
11 Rhythmic Epigenetics in Neuroendocrine and Immune Systems 311

Fu L, Pelicano H, Liu J, Huang P, Lee CC (2002) The circadian gene Period2 plays an important
role in tumor suppression and DNA damage response in vivo. Cell 111:41–50
Gaudet F, Hodgson JG, Eden A, Jackson-Grusby L, Dausman J, Gray JW, Leonhardt J, Jaenisch R
(2003) Induction of tumors in mice by genomic hypomethylation. Science 300:489–492
Greives TJ, Kriegsfeld LJ, Bentley GE, Tsutsui K, Demas GE (2008) Recent advances in reproduc-
tive neuroendocrinology: a role for RFamide peptides in seasonal reproduction? Proc Biol Sci
275:1943–1951
Guil S, Esteller M (2009) DNA methylomes, histone codes and miRNAs: tying it all together. Int J
Biochem Cell Biol 41:87–95
Hanahan D, Weinberg R (2011) Hallmarks of cancer: the next generation. Cell 144:646–674
Hastings MH, Brancaccio M, Maywood ES (2014) Circadian pacemaking in cells and circuits of the
suprachiasmatic nucleus. J Neuroendocrinol 26:2–10
Heard E, Martienssen RA (2014) Transgenerational epigenetic inheritance: myths and mechanisms.
Cell 157:95–109
Helm B, Stevenson TJ (2015) Circannual rhythms: history, present challenges, future directions. In:
Numata H, Helm B (eds) Annual, lunar and tidal clocks: patterns and mechanisms of nature’s
enigmatic rhythms. Springer Japan, Tokyo, pp 213–225
Herman JG, Baylin SB (2003) Gene silencing in cancer in association with promoter
hypermethylation. N Engl J Med 349:2042–2054
Hughes ME, DiTacchio L, Hayes KR, Vollmers C, Pulivarthy S, Baggs JE, Panda S, Hogenesch JB
(2009) Harmonics of circadian gene transcription in mammas. PLoS Genet 5:e1000442
Jackson-Grusby L, Beard C, Possemato R, Tudor M, Fambrough D, Csankovszki G, Jaenisch R
(2001) Loss of genomic methylation causes p53-dependent apoptosis and epigenetic deregula-
tion. Nat Genet 27:31–39
Jang HS, Shin WJ, Lee JE, Tae Do J (2017) CpG and non-CpG methylation in epigenetic gene
regulation and brain function. Genes (Basel) 8:148
Jefferson WN, Chevalier DM, Phelps JY, Cantor AM, Padilla-Banks E, Newbold RR, Archer TK,
Kinyamu HK, Williams CJ (2013) Persistently altered epigenetic marks in the mouse uterus
after neonatal estrogen exposure. Mol Endocrinol 27:1666–1677
Jones P, Baylin S (2007) The epigenomics of cancer. Cell 128:683–692
Josefowicz SZ, Wilson CB, Rudensky AY (2009) Cutting edge: TCR stimulation is sufficient for
induction of Foxp3 expression in the absence of DNA methyltransferase 1. J Immunol
182:6648–6652
Kaplan MJ, Beretta L, Yung RL, Richardson BC (2000) LFA-1 overexpression and T cell
autoreactivity: mechanisms. Immunol Investig 29:427–442
Kautiainen TL, Jones PA (1986) DNA methyltransferase levels in tumorigenic and nontumorigenic
cells in culture. J Biol Chem 261:1594–1598
Klose RJ, Bird AP (2006) Genomic DNA methylation: the mark and its mediators. Trends Biochem
Sci 31:89–97
Knox JD, Araujo FD, Bigey P, Slack AD, Price GB, Zannis-Hadjopoulos M, Szyf M (2000)
Inhibition of DNA methyltransferase inhibits DNA replication. J Biol Chem 275:17986–17990
Kohli RM, Zhang Y (2013) TET enzymes, TDG and the dynamics of DNA demethylation. Nature
502:472–479
Koike N, Yoo SH, Huang HC, Kumar V, Lee C, Kim TK, Takahashi JS (2012) Transcriptional
architecture and chromatin landscape of the core circadian clock in mammals. Science
338:349–354
Kurian JR, Terasawa E (2013) Epigenetic control of gonadotropin releasing hormone neurons.
Front Endocrinol 4:61
Lee PP, Fitzpatrick DR, Beard C, Jessup HK, Lehar S, Makar KW, Cherry SR (2001) A critical role
for Dnmt1 and DNA methylation in T cell development, function, and survival. Immunity
15:763–774
Lei H, Oh SP, Okano M, Juttermann R, Goss KA, Jaenisch R, Li E (1996) De novo DNA cytosine
methyltransferase activities in mouse embryonic stem cells. Development 122:3195–3205
312 C. S. Coyle et al.

Li E (2002) Chromatin modification and epigenetic reprogramming in mammalian development.

Nat Rev Genet 3:662–673
Li E, Bestor TH, Jaenisch R (1992) Targeted mutation of the DNA methyltransferase gene results in
embryonic lethality. Cell 69:915–926
Logan PC, Ponnampalam AP, Steiner M, Mitchell MD (2013) Effect of cyclic AMP and estrogen/
progesterone on the transcription of DNA methyltransferase during the decidualization of
human endometrial stromal cells. Mol Hum Reprod 19:302–312
Lynch EWJ, Coyle CS, Lorgen M, Campbell E, Bowman A, Stevenson TJ (2016) Cyclical DNA
methyltransferase 3a expression is a seasonal and oestrus timer in reproductive tissues. Endo-
crinology 157:2469–2478
Lynch EWJ, Coyle CS, Stevenson TJ (2017) Light and hormonal regulation of histone deacetylase
1, 2, 3 in reproductive tissues. Gen Comp Endocrinol 246:194–199
Ma P, Pan H, Montgomery RL, Olson EN, Schultz RM (2012) Compensatory functions of histone
deacetylase 1 (HDAC1) and HDAC2 regulate transcription and apoptosis during mouse oocyte
development. Proc Natl Acad Sci U S A 109:E481–E489
Miki T, Matsumoto T, Zhao Z, Lee CC (2013) p53 regulates Period2 expression and the circadian
clock. Nat Commun 4:2444
Nakahata Y, Kaluzova M, Grimaldi B, Sahar S, Hirayama J, Chen D, Guarente LP, Sassone-Corsi P
(2008) The NAD+-dependent deacetylase SIRT1 modulates CLOCK-mediated chromatin
remodelling and circadian control. Cell 134:329–340
Nelson RJ (2004) Seasonal immune function and sickness responses. Trends Immunol 25:187–192
Nugent BM, Wright CL, Shetty AC, Hodes GE, Lenz KM, Mahurkar A, Russo SJ, Devine SE,
McCarthy MM (2015) Brain feminization requires active repression of masculinization via
DNA methylation. Nat Neurosci 18:690–697
Okano M, Bell DW, Haber DA, Li E (1999) DNA methyltransferases Dnmt3a and Dnmt3b are
essential for de novo methylation and mammalian development. Cell 99:247–257
Olsen EA, Kim YH, Kuzel TM, Pacheco TR, Foss FM, Parker S, Frankel SR, Chen C, Ricker JL,
Arduino JM, Duvic M (2007) Phase IIb multicenter trial of vorinostat in patients with persistent,
progressive, or treatment refractory cutaneous T-cell lymphoma. J Clin Oncol 25:3109–3115
Orozco-Solis R, Ramadori G, Coppari R, Sassone-Corsi P (2015) SIRT1 relays nutritional inputs to
the circadian clock through the Sf1 neurons of the ventromedial hypothalamus. Endocrinol
156:2174–2184
Pelengaris S, Khan M, Evan G (2002) c-MYC: more than just a matter of life and death. Nat Rev
Cancer 10:764
Pfaff DW, Phillips MI, Rubin RT (2004) Hormonal secretions and responses are affected by
biological clocks. In: Principles of hormone/behavior relations. Elsevier Academic Press, San
Diego, pp 181–196
Pizarro A, Hayer K, Lahens NF, Hogenesch JB (2013) CircaDB: a database of mammalian
circadian gene expression profiles. Nucleic Acids Res 41:D1009–D1013
Prendergast BJ, Cable EJ, Patel PN, Pyter LM, Onishi KG, Stevenson TJ, Ruby NF, Bradley SP
(2013) Impaired leukocyte trafficking and skin inflammatory responses in hamsters lacking a
functional circadian system. Brain Behav Immun 32:94–104
Quddus J, Johnson KJ, Gavalchin J, Amento EP, Chrisp CE, Yung RL, Richardson BC (1993)
Treating activated CD4+ T cells with either of two distinct DNA methyltransferase inhibitors,
5-azacytidine or procainamide, is sufficient to cause a lupus-like disease in syngeneic mice. J
Clin Invest 92:38–53
Rasool M, Malik A, Naseer MI, Manan A, Ansari SA, Begum I, Qazi MH, Pushparaji PN,
Abuzenadah AM, Al-Qahtani MH, Kamal MA, Pushparaji PN, Gan SH (2015) The role of
epigenetics in personalized medicine: challenges and opportunities. BMC Med Genet 8:S5
Rhee I, Jair KW, Yen RWC, Lengauer C, Herman JG, Kinzler KW, Vogelstein B, Baylin SB,
Schuebel KE (2000) CpG methylation is maintained in human cancer cells lacking DNMT1.
Nature 404:1003–1007
11 Rhythmic Epigenetics in Neuroendocrine and Immune Systems 313

Ricceri F, Trevisan M, Fiano V, Grasso C, Fasanelli F, Scoccianti C, De Marco L, Tox AG,

Vineis P, Sacerdote C (2014) Seasonality modifies methylation profiles in healthy people. PLoS
One 9:e106846
Richardson B, Scheinbart L, Strahler J, Gross L, Hanash S, Johnson M (1990) Evidence for
impaired T cell DNA methylation in systemic lupus erythematosus and rheumatoid arthritis.
Arthritis Rheum 33:1665–1673
Richter A, Löhning M, Radbruch A (1999) Instruction for cytokine expression in T helper
lymphocytes in relation to proliferation and cell cycle progression. J Exp Med 190:1439–1450
Riggs AD, Porter TN (1996) Overview of epigenetic mechanisms. In: Russo VEA, Martienssen R,
Riggs AD (eds) Epigenetic mechanisms of gene regulation. Cold Spring Harbor Laboratory
Press, Cold Spring Harbor, NY, pp 29–45
Roy A, Matzuk MM (2006) Deconstructing mammalian reproduction: using knockouts to define
fertility pathways. Reproduction 131:207–219
Savvidis C, Koutsilieris M (2013) Circadian rhythm disruption in cancer biology. Mol Med
18:1249–1260
Stevenson TJ (2017a) Environmental and hormonal regulation of neuroendocrine epigenetic
enzymes. J Neuroendocrinol 29(5)
Stevenson TJ (2017b) Circannual and circadian rhythms in hypothalamic DNA methyltransferase
and histone deacetylase in Siberian hamsters (Phodopus sungorus). Gen Comp Endocrinol
243:130–137
Stevenson TJ (2018) Epigenetic regulation of biological rhythms: an evolutionary ancient molecu-
lar timer. Trends Genet 43:90–100
Stevenson TJ, Lincoln GA (2017) Epigenetic mechanisms regulating circannual rhythms, Chap 29.
In: Kumar V (ed) Biological timekeeping: clocks, rhythms and behaviour. Springer, India
Stevenson TJ, Prendergast BJ (2013) Reversible DNA methylation regulates seasonal
photoperiodic time measurement. Proc Natl Acad Sci USA 110:16651–16656
Stevenson TJ, Prendergast BJ (2015) Photoperiodic time measurement and seasonal immunological
plasticity. Front Neuroendocrinol 37:76–88
Stevenson TJ, Hahn TP, MacDougall-Shackleton SA, Ball GF (2012) Gonadotropin-releasing
hormone plasticity: a comparative perspective. Front Neuroendocrinol 33:287–300
Stevenson TJ, Bernard DJ, McCarthy MM, Ball GF (2013) Photoperiod-dependent regulation of
gonadotropin-releasing hormone 1 messenger ribonucleic acid levels in the songbird brain. Gen
Comp Endocrinol 190:81–87
Stevenson TJ, Onishi KG, Bradley SP, Prendergast BJ (2014) Cell-autonomous iodothyronine
deiodinase expression mediates seasonal plasticity in immune function. Brain Behav Immun
36:61–70
Stevenson TJ, Visser ME, Arnold W, Barrett P, Biello S, Dawson A, Denlinger DL, Dominoni D,
Ebling FJ, Elton S, Evans N, Ferguson HM, Foster RG, Hau M, Haydon DT, Hazlerigg DG,
Heideman P, Hopcraft JGC, Jonsson NN, Kronfeld-Schor N, Kumar V, Lincoln GA,
MacLeod R, Martin SAM, Martinez-Bakker M, Nelson RJ, Reed T, Robinson JE, Rock D,
Schwartz WJ, Steffan-Dewenter I, Tauber E, Thackeray SJ, Umstatter C, Yoshimura T, Helm B
(2015) Disrupted seasonal biology impacts health food security and ecosystems. Proc R Soc B
282:1–10
Symonds H, Krall L, Remington L, Saenz-Robles M, Lowe S, Jacks T, Van Dyke T (1994)
p53-dependent apoptosis suppresses tumor growth and progression in vivo. Cell 78:703–711
Taunton J, Hassig CA, Schreiber SL (1996) A mammalian histone deacetylase related to the yeast
transcriptional regulator Rpd3p. Science 272:408–411
Van Kaam KJ, Delvoux B, Romano A, D’Hooghe T, Dunselman GA, Groothuis PG (2011)
Deoxyribonucleic acid methyltransferase and methyl-CpG binding domain proteins in human
endometrium and endometriosis. Fertil Steril 95:1421–1427
Waddington CH (1953) Epigenetics and evolution. Symp Soc Exp Biol 7:186–199
Weaver ICG, Cervoni N, Champagne FA, D’Alessio AC, Sharma S, Seckl JR, Dymov S, Szyf M,
Meaney MJ (2004) Epigenetic programming by maternal behavior. Nat Neurosci 7:847–854
314 C. S. Coyle et al.

Westwood FR (2008) The female rat reproductive cycle: a practical histological guide to staging.
Toxicol Pathol 36:375–384
Widschwendter M, Jones A, Evans I, Reisel D, Dillner J, Sundstrom K, Steyerberg EW,
Vergouwe Y, Wegwarth O, Rebitschek FG, Siebert U, Sroczynski G, de Beaufort ID, Bolt I,
Cibula D, Zikan M, Bjorge L, Colombo N, Harbeck N, Dudbridge F, Am T, Knoppers BM,
Joly Y, Teschendorff AE, Pashayan N, FORECEE 4C Consortium (2018) Epigenome-based
cancer risk prediction: rationale, opportunities and challenges. Nat Rev Clin Oncol 15:292–309
Wood GA, Fata JE, Watson KLM, Khokha R (2007) Circulating hormones and estrous stage
predict cellular and stromal remodelling in murine uterus. Reproduction 133:1035–1044
Xu HJ, Cairns P, Hu SX, Knowles MA, Benedict WF (1993) Loss of RB protein expression in
primary bladder cancer correlates with loss of heterozygosity at the RB locus and tumor
progression. Int J Cancer 53:781–784
Yeo GS, Heisler LK (2012) Unraveling the brain regulation of appetite: lessons from genetics.
Nat Neurosci 15:1343–1349
Yoder JA, Soman NS, Verdine GL, Bestor TH (1997) DNA (cytosine-T)-methyltransferase in
mouse cells and tissues. Studies with a mechanism-based probe. J Mol Biol 270:385–395
 Part III
Development of Neuroendocrine Circuits
Development of Limbic System
Stress-Threat Circuitry 12
Newton S. Canteras, Dayu Lin, and Joshua G. Corbin

Abstract
How an animal responds to environmental stressors and threats is essential for
survival. These responses are governed through an interconnected circuit in the
brain dubbed the limbic system. Three main structures of the limbic system are
the amygdala, bed nucleus of the stria terminalis (BNST), and hypothalamus.
Together these structures define a central stress-threat circuit. This chapter
describes the most up to date knowledge of how these structures function in
regulating responses to stressors and how each of these structures is formed from
embryonic development. This knowledge of the underlying biology of these
regions is essential to design rational therapeutic approaches for conditions,
such as post-traumatic stress disorder (PTSD) that are characterized by
dysregulation of the stress-threat system. In this chapter, we also describe the
critical unanswered questions in the field of stress-threat research and potential
future research directions.

N. S. Canteras
Department of Anatomy, Institute of Biomedical Sciences, University of São Paulo; São Paulo, SP,
Butanta, Brazil
D. Lin
Neuroscience Institute, New York University School of Medicine, New York, NY, USA
Department of Psychiatry, New York University School of Medicine, New York, NY, USA
Center for Neural Science, New York University, New York, NY, USA
J. G. Corbin (*)
Center for Neuroscience Research, Department of Pediatrics, Children’s Research Institute,
Children’s National Medical Center, Washington, DC, USA
e-mail: [email protected]

# Springer Nature Switzerland AG 2020 317

S. Wray, S. Blackshaw (eds.), Developmental Neuroendocrinology, Masterclass in
Neuroendocrinology 9, https://doi.org/10.1007/978-3-030-40002-6_12
318 N. S. Canteras et al.

Keywords
Stress-threat system · Amygdala · Bed nucleus of stria terminalis ·
Hypothalamus · Embryonic development · Brain circuitry · Defense responses ·
Loss and gain of function · Mouse models · Circuit manipulation

12.1 Introduction

Research into the neural mechanisms underlying defensive responses in animals has
a long history. Animals are naturally selected to protect themselves from threats
associated with the presence of a predator or aggressive conspecifics that evoke
innate fear. Evidence accumulated over the past two decades suggests that the
circuitry that supports fear responses is complex and involves multiple independent
circuits that process different types of fear. In particular, there is good evidence to
support the existence of distinct circuits for fear of predators and fear of aggressive
conspecifics (a member of the same species). In Sect. 12.2 of this chapter, we review
these data and consider these segregated fear circuits and point out the crucial role of
the amygdala, hypothalamus and bed nuclei of the stria terminalis. In Sect. 12.3 of
this chapter, we review the current state of knowledge of how these brain regions are
developmentally formed, and in Sect. 12.4 we describe the critical unanswered
questions in this field and future directions.

12.2 Key Brain Structures of Fear/Threat Circuits

12.2.1 The Amygdala as a Switchboard for Fear

The mammalian amygdala comprises a heterogeneous set of distinct regions, called

nuclei. Accordingly, the amygdalar nuclei have been previously broadly classified
into a cortical division (cortical amygdala, basolateral amygdala (BLA), basomedial
amygdala (BMA), and lateral amygdala (LA)) and a striatal division (medial amyg-
dala (MEA) and central amygdala (CEA)) (Swanson and Petrovich 1998). However,
as more recent studies have revealed a previously unrecognized level of develop-
mental complexity as described in Sect. 12.3 of this chapter, this original simple
“striatal/cortical” classification scheme has been updated.
Neural activity mapping and lesion studies in rodents show that different types of
threat stimuli activate and depend on different parts of the amygdala (Fig. 12.1). In
rodents, olfactory and vomeronasal cues have a crucial role in signaling the presence
of a predator. These signals are conveyed through direct connections from the main
and accessory olfactory bulb, respectively, to the MEA. Predator odors are thought
to work as kairomones, which are semiochemicals that are released by one species
and have a favorable adaptive effect on a different “receiving” species but no
favorable effect on the transmitting species (Wyatt 2014). A number of studies
have shown that the detection of predator odors relies on distinct olfactory
12 Development of Limbic System Stress-Threat Circuitry 319

Fig. 12.1 Parallel circuits mediating innate responses to predators and aggressive conspecifics. See
text for details. Abbreviations: BMAp basomedial amygdalar nucleus, posterior part, BNSTif, pr bed
nuclei stria terminalis, interfascicular nucleus, principal nucleus; LA latereal amygdalar nucleus,
MEApv, pd medial amygdalar nucleus, posteroventral and posterodorsal parts, MPN medial
preoptic nucleus; PAGdm, dl, l periaqueductal gray, dorsomedial, dorsolateral and lateral parts,
PMDdm, vl dorsal premammillary nucleus, dorsomedial and ventrolateral parts, PMV ventral
preamammillary nucleus; VMHdm, vl ventromedial hypothalamic nucleus, dorsomedial and ven-
trolateral parts

subsystems for chemodetection, namely the vomeronasal organ (McGregor et al.

2004), the Grueneberg ganglion (Brechbühl et al. 2013), and subsets of sensory
neurons within the main olfactory epithelium that express trace amine-associated
receptors (TAARs) (Liberles 2015). Nasal detection of different predator odors has
been associated with distinct olfactory subsystems, i.e., 2-phenylethylamine, which
is found in carnivore urine, activates TAAR4 neurons in the main olfactory epithe-
lium (Ferrero et al. 2011; Dewan et al. 2013), cat fur odor activates the vomeronasal
organ (McGregor et al. 2004), and 2-propylthietane, which is extracted from the
stoat (also known as the short-tailed weasel) anal gland, is a Grueneberg ganglion
activator (see Pérez Gómez et al. 2015). Exposure to all of these different types of
predator odors results in increased cFos expression in the posteroventral part of the
medial amygdalar nucleus (MEApv; see Pérez-Gómez et al. 2015). Notably, rats
with lesions in the MEA show a substantial reduction in innate fear responses to a cat
or its odor. Exposure to a live predator also activates two other amygdala nuclei, the
LA and posterior BMA (BMAp) (Martinez et al. 2011). These nuclei receive inputs
from visual and auditory association areas and are likely to integrate non-olfactory
predator-derived sensory cues (see Martinez et al. 2011) (Fig. 12.1). Consistent with
this view, lesions in these nuclei also reduce innate fear responses to a cat (Martinez
et al. 2011). Olfactory and vomeronasal inputs to MEA also have a role in the
detection of conspecifics. Exposure to an aggressive conspecific predominantly
activates the posterodorsal part of the MEA (MEApd; Motta et al. 2009-Key
Reference), suggesting that the processing of olfactory threat cues that associate
with predators and conspecifics is likely to depend on different MEA subnuclei (Fig.
12.1). Interestingly, the MEApd (responsive to an aggressive conspecific) and the
MEApv (responsive to predator odor) are marked by different LIM homeodomain-
containing transcription factors (Lhx6 in the MEApd and Lhx9 in the MEApv),
which may form the basis for hardwiring systems which are differentially responsive
to social and predatory cues (Choi et al. 2005-Key Reference).
320 N. S. Canteras et al.

Box 1 Interesting Information of the Latest Pharmaco- and Optogenetic

Tools for Circuit Manipulation
The mammalian brain is composed on hundreds of regions, billions of
neurons, and trillions of connections. Understanding the behavioral relevance
of any group of cells or its connection is a daunting task. Traditionally,
neuroscientists use physical, electrical, pharmacological, or genetic methods
to perturb the brain (Carter and Shieh 2010). Each of these methods suffer
from various drawbacks, including low spatial and temporal control and poor
cell type selectivity. To achieve more precise control of the cells and their
connections, novel chemogenetic and optogenetic tools were developed over
the last decade by taking inspiration from nature and performing creative
molecular engineering. Since the debut of optogenetic and chemogenetic
probes over 10 years ago, these tools have revolutionized the field of neuro-
science (Luo et al. 2018-Key Reference). They have enabled neuroscientists to
investigate and dissect the neural circuits in a complex brain environment and
gained incredible new insight into the neural generation of complex behaviors,
such as defense.
Pharmacogenetics: As indicated by the names, pharmaco-(also known as
chemo-) and opto-genetic tools use chemicals and light, respectively, to
control the activity of genetically defined neuronal populations or connections.
The most widely used chemogenetic probes are known as DREADDs,
Designer Receptors Exclusive Activated by Designer Drugs (Roth 2016).
DREADDs are virtually mutated G-protein coupled receptors (GPCRs) that
no longer bind to their endogenous ligands but bind small synthetic chemicals.
Depending on the secondary signaling pathways that are coupled to the
DREADDs, activating the DREADDs could either reduce cell excitability
by opening a GIRK channel (Gi-DREADD) or activate the cells by increasing
Ca2+ concentration (Gq-DREADD) or cAMP levels (Gs-DREADD). The
most commonly used excitatory DREADDs are based upon Gq-coupled
human M3 muscarinic receptor (hM3Dq) while Gi-coupled human M4 mus-
carinic receptor (hM4Di) based DREADD is the most commonly used inhibi-
tory type (Armbruster and Roth 2005; Armbruster et al. 2007). Both hM3Dq
and hM4Di were generated via rounds of mutagenesis and screening for
decreased affinity to acetylcholine (the natural ligand of muscarinic receptors)
as well as increased affinity to clozapine-N-oxide (CNO), a biologically inert
small molecule (Armbruster and Roth 2005: Armbruster et al. 2007). In
addition to DREADDs, several other forms of ligand-controlled GPCR
based receptors have also been developed, allowing orthogonal controls of
multiple populations in the same animals (Coward et al. 1998; Vardy et al.
2015; Magnus et al. 2019).
Optogenetics: Nearly all optogenetic probes have been engineered from
natural opsins that selectively respond to specific wavelengths of light and

(continued)
12 Development of Limbic System Stress-Threat Circuitry 321

Box 1 Interesting Information of the Latest Pharmaco- and Optogenetic

Tools for Circuit Manipulation (continued)
translocate ions across the membrane (Fenno et al. 2011). The best known
optogenetic probe is channelrhodopsin-2 (ChR2), a nonspecific cation channel
found in green algae (Boyden et al. 2005). Upon sensing blue light, ChR2
undergoes a conformational change, opening the pore, and allowing H+, Na+,
K+ and Ca2+ ions to passively diffuse across the membrane and depolarize the
cell. When sufficient cation ions influx through ChR2, a spike will be
generated by activating sodium channels. The conformational change of
ChR2 depends on the presence of retinal which happens to be present in
sufficient amount in most vertebrate cells in the form of vitamin A, negating
the need to supply it exogenously (Hegemann and Nagel 2013). The first opsin
developed for inhibiting neural activity was halorhodopsin (NpHR), a light
driven Cl-pump that is expressed by the halobacterium (Schobert and Lanyi
1982; Gradinaru et al. 2010). Upon sensing yellow light, NpHR actively
pumps Cl-ions into the cells causing hyperpolarization of the membrane
potential. A second class of popular inhibitory opsin is archaerhodopsin-3
(Arch), a light driven H+ pump that is found in Halorubrum sodomense
(Chow et al. 2010). Most recently, a third class of inhibitory opsin was
developed by transforming channelrhodopsin into a light-activated chloride
channel (Berndt et al. 2014). To date, a wide range of optogenetic probes with
diverse properties (e.g. excitation wavelength and open and close kinetics) are
available, allowing neuroscientists to control neural activity with unprece-
dented precision.
Delivery methods: The most common way to deliver the chemogenetic or
optogenetic probes is to infect cells with a replication incompetent virus, such
as adeno-associated virus (AAV) that carries the transgene driven by a short
promoter (Luo et al. 2008). To achieve the cell type specificity, conditional
AAV vectors that carry a Cre-dependent transgene is delivered to the trans-
genic animals that express Cre in a specific subset of cells. The specificity can
be further improved based on the connectivity. Retrograde viruses, such as
herpes simplex virus or retrograde AAV, can be delivered to the downstream
regions of the cells of interest so that only Cre-expressing cells with certain
projection patterns are infected by the virus and express the transgene (Neve et
al. 2005; Tervo et al. 2016). Additionally, several transgenic lines that express
optogenetic or chemogenetic probes, either conditionally or constitutively,
have been developed to achieve consistent manipulation across animals and
eliminate the need for viral injection (Madisen et al. 2012; Zhu et al. 2016).
One advantage of optogenetic and chemogenetic probes is that they are
controlled by specific ligands—light or small molecule. As a downside of this
controllability, it requires the experimenter to deliver the ligand exogenously.
Small molecules, such as CNO, are typically delivered through i.p. injection

(continued)
322 N. S. Canteras et al.

Box 1 Interesting Information of the Latest Pharmaco- and Optogenetic

Tools for Circuit Manipulation (continued)
although recent reports suggest that CNO has poor brain-barrier penetrability
and is likely metabolized to clozapine before entering the central nervous
system (Gomez et al. 2017). Because DREADDs are expressed in both cell
bodies and axon terminals, CNO can also be administrated directly to the
regions where the axon terminals reside to block synaptic transmission of a
specific projection (Stachniak et al. 2014). Light delivery is generally achieved
by implantation of a light guide deep into the brain close to the target region.
This is because mammalian brain tissue scatters light exponentially, with only
10% of light intensity remaining at a distance 500 μm from the light source
(Aravanis et al. 2007). During an experiment, an optic fiber is connected to the
implanted light guide to deliver appropriate wavelength light to the target
region. The light can be delivered to either cell bodies or specific projections to
achieve pathway-specific manipulation (Tye and Deisseroth 2012).

12.2.2 Parallel Hypothalamic Paths for Fear

In the hypothalamus, it has been recognized that behavioral control columns, formed
by distinct hypothalamic circuits, are involved in the control of the three basic
classes of goal-oriented behaviors: ingestive, reproductive, and anti-predatory defen-
sive (Swanson 2000). This concept was mostly based on a systematic analysis of
axonal projections from the medial half of the hypothalamus of the rat and was
further corroborated by a number of functional studies (see Swanson 2000). Based
on anatomical tract-tracing experiments of medial hypothalamic zone projections,
this zone has been divided into two networks that show a high degree of intercon-
nection among nuclei (Fig. 12.1, and see Box 1 for description of modern tools now
being employed to understand the link between circuity and behavior). One network
comprises the anterior hypothalamic nucleus (AHN), the dorsomedial part of the
ventromedial nucleus (VMHdm), and the ventrolateral part of the dorsal
premammillary nucleus (PMDvl), which are highly interconnected (Canteras and
Swanson 1992; Canteras et al. 1994; Risold et al. 1994) (Fig. 12.1). Exposure to a
predator or its odor activates the AHN–VMHdm–PMDvl circuit (i.e., the predator-
responsive circuit; Fig. 12.1) (Cezario et al. 2008). Notably, the VMHdm and the
AHN provide a unique type of projection to the PMdvl, with dense bilateral
projection fields (Canteras et al. 1994; Risold et al. 1994). This type of synaptic
arrangement enables the PMdvl to work as an amplifier for the hypothalamic neural
processing of predator-related cues. Corroborating this view, the PMDvl is the most
responsive brain site during exposure to a cat or its odor, and lesions or pharmaco-
logical blockade of the PMD drastically reduces fear responses to a predator or its
odor (see Cezario et al. 2008). The VMHdm, in turn, receives most of the amygdalar
processing of predator-related cues from the MEApv (Canteras et al. 1995) and the
12 Development of Limbic System Stress-Threat Circuitry 323

BMAp (Petrovich et al. 1996) (Fig. 12.1). In this regard, of particular relevance is the
finding that exposure to different types of carnivore odors converge in a pathway
formed by the MEApv and VMHdm (Pérez Gómez et al. 2015). Pharmacogenetic
inhibition in transgenic mice expressing hM4D in steroidogenic factor 1 (SF1)-
expressing neurons in the central and dorsomedial parts of the VMH decreased the
defensive responses of the mice during exposure to a rat but not during exposure to
an aggressive conspecific (Silva et al. 2013). Conversely, optogenetic activation of
SF1-expressing neurons in the same population of VMH cells initiates a range of
behavioral responses that resemble the animal’s natural defensive behaviors (Wang
et al. 2015). The main targets of the VMHdm are the dorsal periaqueductal gray
(PAGd) and the AHN. Accordingly, activation of the pathway from the VMHdm to
the PAG induces immobility, whereas stimulation of the pathway from the VMHdm
to the AHN promotes mobility, running, escape jumping, and avoidance and the
animals tend to avoid the place where they had received the stimulation (Wang et al.
2015). At least part of the complex defensive behaviors that have been described for
the AHN are likely to be mediated through its projections to the PMDvl. The PMDvl
is the main interface between the predator-responsive circuit and the PAG (Motta et
al. 2009-Key Reference). The PMDvl provides strong projections to the dorsomedial
and dorsolateral parts of the PAG, both of which are activated during exposure to a
live predator or its odor and is an important site for the organization of fear responses
to predators (Motta et al. 2009-Key Reference) (Fig. 12.1). Notably, lesions in the
dorsal PAG block the entire range of fear responses to a predator, from flight and
freezing to risk assessment, and is conceivable that the degree of activation of the
dorsal PAG dictates the intensity of the fear response (see Cezario et al. 2008).
The second medial hypothalamic network includes the medial preoptic nucleus
(MPN), the ventrolateral part of the VMH (VMHvl), the tuberal nucleus, and the
ventral premammillary nucleus (PMV), which are also highly interconnected
(Simerly and Swanson 1988; Canteras et al. 1992, 1994). In contrast to predator
exposure, exposure to an aggressive conspecific activates the MPN–VMHvl–PMV
network (the conspecific-responsive circuit; Fig. 12.1), clearly suggesting the exis-
tence of separate circuits responsive to predators and conspecifics (see Motta et al.
2009-Key Reference). In addition, exposure to an aggressive conspecific activates
the dorsomedial part of the PMD (PMDdm), which is distinct from the predator-
activated PMDvl (Fig. 12.1). Of particular relevance, lesions of the PMD block
freezing and passive supine responses to an aggressive conspecific but not, however,
active defense, including boxing and upright defensive postures (Motta et al. 2009-
Key Reference). These data suggest that separate circuits in the PMD are responsible
for fear responses to predators versus fear responses to aggressive conspecifics. The
link between elements of the conspecific-responsive hypothalamic circuit and the
PMDdm is not clear at the moment, and the subfornical region of the lateral
hypothalamic area has been suggested one possible candidate (Goto et al. 2005).
As discussed for the predator-responsive medial hypothalamic circuit, the PMD
serves as an important interface between the conspecific-responsive hypothalamic
circuit and the PAG. However, differently from the PMDvl, the PMDdm projects to
the dorsomedial PAG (PAGdm) and lateral PAG (PAGl) (Fig. 12.1), providing a
324 N. S. Canteras et al.

Table 12.1 Summary of gain and loss of functional studies of the defense circuit
Area (cell type) Manipulation Behavioral change References
VMHdm (SF1+ ChR2 activation Freeze, flight, jump Wang et al. (2015)
cells) Kunwar et al. (2015)
VMHdm(SF1+ DREADDi Reduced defense toward a Silva et al. (2013)
cells) inactivation predator
VMHdm(SF1+ Capase3 mediated Reduced predator Kunwar et al. (2015)
cells) cell ablation avoidance
VMHdm ChR2 activation Freeze, flight, jump Wang et al. (2015)
(SF1)- > AHN
VMHdm ChR2 activation Freeze Wang et al. (2015)
(SF1)- > PAG
AHN ChR2 activation Flight and jump Wang et al. (2015)
(nonselective)
VMHvl DREADDi Reduced social defense Silva et al. (2013)
(nonselective) inactivation
VMHvl (Esr1+) ChR2 activation Increase social defense Wang et al. (2019)
VMHvl (Esr1+) NpHR3.0 Decrease active social Wang et al. (2019)
inactivation defense
PMd Cytotoxic lesion Decrease passive social Motta et al. (2009)-
(nonselective) defense Key Reference
PAGdm Cytotoxic lesion Decrease passive and active Canteras lab,
(nonselective) social defense unpublished results

projection pattern that matches the PAG activation pattern in response to an aggres-
sive conspecific (see Motta et al. 2009-Key Reference). Preliminary findings from
the Canteras laboratory suggest that PAGdm cytotoxic lesions decrease both passive
(freezing and the typical on-the-back position maintained after the resident leaves
them alone) and active forms (an upright position with sparse boxing and dashing
away from the resident) of social defensive responses. In addition to the PMD,
pharmacogenetic inhibition in the VMHvl of mice locally infected with an adeno-
associated virus (AAV) expressing hM4D pharmacogenetic neural inhibition tool
(HA-hM4D) decreased the defensive responses of the mice during exposure to an
aggressive conspecific but not during exposure to a rat (Silva et al. 2013). Moreover,
estrogen receptor+ (Esr1+) cells situated at the anterior portion of the VMHvl are
significantly excited when animals actively defended themselves against conspecific
aggressors. Furthermore, during exposure to a conspecific aggressor, inactivation of
VMHvlEsr1 cells compromised the active forms of social defensive behaviors, such
as dashing away from the aggressor and assuming an upright posture (Wang et al.
2019). A summary of functional studies implicating the VMH in the defensive
behaviors is shown in Table 12.1.
12 Development of Limbic System Stress-Threat Circuitry 325

12.2.3 The BNST and Social and Anti-Predatory Defense

The BNST serves as an important interface between the amygdala and the hypotha-
lamic circuits involved in social and anti-predatory defense (Fig. 12.1). However, the
role the BNST plays in the context of social and anti-predatory defense remains to be
fully investigated. Previous research has established the role of the BNST in
sustained fear as distinct from the role of the amygdala in phasic fear (see Lebow
and Chen 2016; Davis et al. 2010). Accordingly, the amygdala is likely to be
generically involved in the assessment of specific cues related to fearful events,
whereas the BNST would be related to the ability to differentiate between safe and
non-safe situations (Sullivan et al. 2004), causing a sustained state of apprehension
(see Lebow and Chen 2016; Davis et al. 2010). On hodological grounds, the
elements of the posterior division of the BNST fit quite well with the parallel circuits
organizing anti-predatory and social defense. Thus, in the posterior BNST, the
principal nucleus is particularly targeted by the MEApd (Canteras et al. 1995),
which, as previously mentioned, responds to aggressive conspecifics, and projects
to elements of the conspecific-responsive hypothalamic circuit the MPN, VMHvl,
tuberal nucleus, and PMV (Dong and Swanson 2004) (Fig. 12.1). Conversely, the
interfascicular nucleus of the posterior division of the BNST (BNSTif) receives
amygdalar inputs from regions processing predator cues, namely the MEApd and
BMAp (Canteras et al. 1995; Petrovich et al. 1996) and projects to elements of the
predator-responsive hypothalamic circuit, including the AHN, the VMHdm and, to a
lesser degree, the PMD (Dong and Swanson 2004) (Fig. 12.1). Moreover, the
interfascicular and principal nuclei of the BNST receive strong inputs from elements
conspecific- and predator-responsive medial hypothalamic circuits (Simerly and
Swanson 1988; Canteras et al. 1992, 1994; Risold et al. 1994), suggesting that the
BNST has more integrative functions that should be investigated in the context of
social and anti-predatory defense.

12.3 Genetic Regulation of Stress/Threat System Development

As described above many different brain structures and nuclei comprise the stress/
threat system. In this section we cover what is currently understood regarding
development of three central brain regions of this system: (1) the ventromedial
hypothalamic (VMH), paraventricular (PVN), and premammilary (PMN) nuclei of
the hypothalamus, (2) the MEA, and (3) the BNST. These brain structures are highly
tuned to detecting and processing environmental threats, and as such make up a core
interconnected circuit regulating stress/threat responses. While the knowledge of
development of each system remains incomplete and lags behind that of much more
studied brain regions such as the cerebral cortex, a picture is emerging of how each
structure is assembled during embryogenesis and how these developmental
programs may relate to how they function. Below we describe development of
each brain structure individually (Sect. 12.3.1–12.3.3), followed by what is known
about how connectivity is established (Sect. 12.3.4), and conclude with data and
326 N. S. Canteras et al.

Fig. 12.2 BNST and MEA progenitor zones. Schematics of the E13.5 brain at the level of the
telencephalon and diencephalon showing both the overall anatomy and regions of defined progeni-
tor zones known to contribute to either BNST or MEA (A & B). Table below in (C) lists the
transcription factors expressed in each progenitor zone and contributions to either the BNST or
MEA. Abbreviations: CGE caudal ganglionic eminence, CVP caudal-ventral pallium, Di Dienceph-
alon, dLGE dorsal lateral ganglionic eminence, DP dorsal pallium, LGE lateral ganglionic emi-
nence, LP lateral pallium, LV lateral ventricle, MGE medial ganglionic eminence, POA preoptic
area, VP ventral pallium, 3v third ventricle (Illustrations by Katie Sokolowski)

speculation regarding how these developmental genetic programs may link to later
stress/threat behavioral responses in adult animals (Sect. 12.3.5).

12.3.1 Development of the VMH, PVN, and PMN Hypothalamic

Nuclei

The major hypothalamic subnuclei that regulate threat avoidance and detection are
the VMH, the PVN, and the PMN (Gross and Canteras 2012-Key Reference;
Hashikawa et al. 2016-Key Reference; Takahashi 2014) (Fig. 12.1). As development
of these regions are covered in greater detail in the chapters by M. Placzek and M
Towers (Development of the Neuroendocrine Hypothalamus), and J. Michaud
(Development of the Paraventricular Nucleus), here we provide a general overview,
as readers should refer to the above chapters as well as outstanding current reviews
(Bedont et al. 2015; Alvarez-Bolado 2018; Xie and Dorsky 2017) for more details.
The vast majority of neurons in the hypothalamus are generated during embry-
onic development between embryonic day (E)10.5 and E16.5 in mouse (Shimada
12 Development of Limbic System Stress-Threat Circuitry 327

Table 12.2 Key Transcription Factors in Hypothalamic Stress/Threat System Development

Nucleus Key transcription factors Neuronal cell types
VMH Rax, Asc1, Ngn3, Nkx2.1, Fezf1, Nr5a1 ERα+, glutamate+
PVN Sim1, Arnt2, Otp, Brn2(Pou3f2) Avp+, Oxt+, Crh+
PMN Nkx2.1, repression of En1 Da+
Abbreviations: Avp Arginine vasopressin, Crh Corticotropin releasing hormone, DA Dopamine,
ERα+ Estrogen receptor-alpha, Oxt Oxytocin

and Nakamura 1973 and reviewed in Bedont et al. 2015) (Mouse gestation is
approximately 19 days). Hypothalamic progenitor cells are born in the ventricular
zone surrounding the ventral aspect of the third ventricle (di on Fig. 12.2). Dividing
progenitors, spatially distributed along both dorsal-ventral and rostral-caudal axes of
the developing diencephalon, express different combinations of transcription factors
(Ferran et al. 2015; Shimogori et al. 2010), master regulator genes that function in a
combinatorial manner to determine cell fate. Differential transcription factor expres-
sion is one of the early developmental steps in which unspecified neural progenitors
progressively take on neuronal subtype identity prior to achieving their mature
neuronal phenotype. Table 12.2 summarizes the key transcription factors that com-
prise the unique “combinatorial code” for development and specification of major
subclasses of neurons of the VMH, PVN, and PMN (Bedont et al. 2015; Alvarez-
Bolado 2018; Xie and Dorsky 2017; Nouri and Awatramani 2017). These genes are
expressed during the proliferative and/or early postmitotic stages of hypothalamic
neuron development. It is important to note that Table 12.2 represents an incomplete
catalog of genes required for specification of neurons in these nuclei, especially the
PMN, as a number of required transcription factors remain to be identified.

12.3.2 Development of the Medial Amygdala (MEA)

At the anatomical level, the MEA can be subdivided into three distinct subdomains:
Anterior (MEAA), posterio-dorsal (MePd) and posterior-ventral (MePv). At the
neuronal subtype level, the MEA is composed of a mix of three major classes of
neurons: local projecting interneurons (Bian 2013; Guirado et al. 2008; Real et al.
2009), and long-range projecting excitatory and inhibitory output neurons (Bian et
al. 2008; Choi et al. 2005; Keshavarzi et al. 2014). Single-cell RNA-sequencing
studies have revealed that the MEA is populated by at least 16 distinct molecularly
identifiable neuronal subtypes (Wu et al. 2017). Much of this diversity is found
across interneurons, which based on a combination of previously identified molecu-
lar, electrophysiological and morphological criteria, fall into a number of distinct
subtypes (Petilla Interneuron Nomenclature Group et al. 2008). In the MEA, these
include subtypes that are defined by their expression of either somatostatin (Sst),
Calbindin (CB), Calretinin (CR), the serotonin receptor subunit 5Ht3a, or Neuro-
peptide Y (NPY). Interestingly, the MEA is devoid of interneurons marked by
328 N. S. Canteras et al.

expression of parvalbumin (PV) (Carney et al. 2010; Guirado et al. 2008), a subtype
that is found in abundance in the cerebral cortex.
In addition to these local and projection inhibitory neurons, the MEA is
comprised of excitatory output neurons. This mixing of inhibitory and excitatory
output neurons within the same structure is atypical for telencephalic nuclei as output
neurons of telencephalic nuclei are typically either excitatory (e.g., cerebral cortex,
basolateral amygdala) or inhibitory (e.g., basal ganglia), but not both. MEA neurons
can also be identified based on expression of one or more neuropeptides and
neurohormones, factors which mediate the internal state of the animal and appropri-
ate behavioral responses to stressors and threats (Gross and Canteras 2012-Key
Reference; Hashikawa et al. 2018; Li and Dulac 2018; Sokolowski and Corbin
2012; Yang and Shah 2014). How expression of these factors relates to the other
criteria for defining neuronal identity in the MEA remains an area of active investi-
gation. Although less well characterized than their interneuron counterparts, both
inhibitory and excitatory output neurons in the MEA can also be classified according
to distinct intrinsic electrophysiological profiles (Bian 2013; Bian et al. 2008;
Carney et al. 2010; Keshavarzi et al. 2014).

Box 2 This Box Contains Interesting Information on the Techniques for

Fate Mapping Neural Populations
With each technique, cell populations are indelibly marked by a reporter gene
(for example, GFP). Fate of cells are tracked either anatomically via transplan-
tation, via electroporation in which a cell-specific subtype is labeled via
delivery of a GFP-expression vector or via genetic labeling in transgenic
mice (Illustrations by Katie Sokolowski).
12 Development of Limbic System Stress-Threat Circuitry 329

TRANSPLANTATION
E13.5 GFP Donor E13.5 non-GFP Recipient Adult Recipient

ELECTROPORATION
Plasmid E13.5 non-GFP Recipient Adult Recipient

E GFP

TRANSGENIC
F0 Generaon F1 Generaon

Enhancer CRE x ROSA STOP GFP Enhancer CRE ROSA GFP

330 N. S. Canteras et al.

A BNST VP: Pax6, Dbx1

dLGE: Pax6

LGE: Islet 1, Dlx5

MGE: Nkx2.1, Lhx6

POA: Dbx1

CGE: Coup-TFII

B
MEA DP/LP: Emx1

MGE: Nkx2.1, Lhx6

POA: Dbx1, FoxP2

CGE: Coup-TFII

CVP: Ebf3

Di: Foxb1, Otp, Lhx5

Fig. 12.3 Adult BNST and MEA diversity. Coronal sections of the brain at the level of the BNST
(A, green) and MEA (B, blue) show relationship between gene expression patterns in embryonic
progenitor domains and neuronal diversity in each structure. See Fig. 12.1 for abbreviations
(Illustrations by Katie Sokolowski)

The above-described complexity of MEA neuronal cell types may be further under-
stood through the perspective of embryonic development when neurons are first
specified. A relatively significant amount of work studying dynamic gene expression
changes over developmental time, short-and long-term cell tracking experiments
using lipophilic fluorescent dyes, in vivo neuronal labeling, embryonic transplanta-
tion and genetic fate mapping has provided a relatively comprehensive map for
development of diverse MEA subpopulations (Fig. 12.3, Table 12.3, and references
therein, and see Box 2 for description of tracing techniques). From these studies, ~9
of the at least 16 MEA neuronal subclasses can be identified by differential devel-
opmental origin and/or embryonic transcription factor expression (Table 12.3, Figs.
12.2 and 12.3). At the broadest anatomical level, the three major MEA subnuclei
(MEAA, MePD, and MePV) can be individually identified in the postnatal brain by
the differential expression of members of the LIM family of homeodomain encoding
transcription factors, Lhx5, Lhx6, and Lhx9 (Choi et al. 2005-Key Reference; García-
López et al. 2008). The expression of these genes is atypical of developmentally
12 Development of Limbic System Stress-Threat Circuitry 331

Table 12.3 Origins of MEA neuronal diversity

Known
TF
Embryonic Source markers Mature neuronal subtype References
MGE Nkx2.1, Sst + and Calbindin+ Bupesh et al. (2011);
Lhx6 interneurons Carney et al. (2010);
Puelles et al. (2016)
CGE Coup- Inhibitory neurons (likely Kanatani et al. (2015);
TFII including Calretinin+ and Nery et al. (2002); Soma
5HT3a + interneurons) et al. (2009); Touzot et al.
(2016)
POA Dbx1 nNos+, Aromatase+ inhibitory Hirata et al. (2009);
output neurons. A subset Lischinsky et al. (2017)
express ERα and AR
POA?/ Foxp2 Inhibitory output neurons Carney et al. (2010);
Anterior (distinct subclass from Dbx1- Lischinsky et al. (2017)
diencephalon? derived). A subset express ERα
and to a lesser extent, AR.
POA Nkx5.1 Npy + interneurons Gelman et al. (2011)
dLGE Tshz1 Presumptive inhibitory Kuerbitz et al. (2018)
interneurons
Diencephalon/ Foxb1 Unknown subclass of inhibitory Zhao et al. (2008)
Hypothalamus neurons
Lateral and dorsal Emx1 Excitatory neurons Cocas et al. (2009); Gorski
pallium (cerebral et al. (2002)
cortical
primordium)
Caudo-ventral Ebf3 Excitatory neurons Ruiz-Reig et al. (2018)
pallium (CVP)
Diencephalon/ Otp, Excitatory neurons Bupesh et al. (2011); Soma
Hypothalamus Lhx5 et al. (2009); Zhao et al.
(2008)
It has also been proposed that the source of Sst + interneurons is part of a separate subdomain in
the diagonal area of the telencephalon (Dg), which has also been previously referred to alternatively
as the anterior endopeduncular area (AEP), caudal or caudo-ventral MGE (cMGE, cvMGE)
(Bupesh et al. 2011, 2011; Puelles et al. 2000). For simplicity we refer to this region as part of
the MGE.

expressed transcription factors as their expression persists through postnatal stages,

thus providing a putative fate map from embryogenesis to adult stages (with the
major caveat that gene expression patterns might not be fixed within the same
population over time). To overcome this inherent limitation of using gene expression
studies as a proxy for fate mapping, a number of studies have employed more
sophisticated genetic and nongenetic tools to follow MEA neuronal fate over time
(see Box 2).
Embryonic gene expression studies have revealed that in addition to the
LIM-containing transcription factor encoding genes described above, emerging
MEA nuclei can also be identified by expression of a variety of other transcription
332 N. S. Canteras et al.

factor encoding genes such as Tbr1, Foxp2, Nkx2.1, and Otp (Table 12.3, and
references therein). Previous fate mapping studies using Cre-based approaches
(see Box 2) have begun to uncover the relationship between neuronal progenitor
cell identity as defined by transcription factor expression and ultimate MEA neuro-
nal identity. While a full mapping of MEA neuronal diversity remains to be
accomplished and subsequently linked back to developmental origin, there are
clear patterns that have begun to emerge. First, fate mapping work, along with
previous standard birthdating studies, revealed that MEA neuronal specification
occurs within a similar epoch as the hypothalamus, roughly between E10.5 and
E15.5 in the mouse (Soma et al. 2009). Second, the same ventral telencephalic
domains (MGE and CGE), which generate interneurons of the cerebral cortex
(Batista-Brito and Fishell 2009; Xu et al. 2003), are also major sources of interneu-
ron populations for the MEA (Figs. 12.2, 12.3 and Table 12.3, and references
therein). For example, the CGE is the source of calretinin+ neurons and the MGE
is the source of Sst + and Calbindin+ interneurons in the MEA (Kanatani et al. 2015;
Nery et al. 2002; Soma et al. 2009; Touzot et al. 2016). Moreover, similar to the
cerebral cortex, a population of MEA excitatory neurons appears to be derived from
Emx1+ progenitors originating in lateral and dorsal pallial compartments of the
cerebral cortical primordium (Gorski et al. 2002) (Figs. 12.2, 12.3 and Table 12.3).
In contrast to this “shared” origin with other telencephalic brain regions for
populations of MEA interneurons and excitatory neurons, there appear to be multiple
unique sources of origin for MEA neuronal diversity. These include progenitor pools
located embryonic zones that straddle both sides of the telencephalic–diencephalic
border. The telencephalic niches include the preoptic area (POA) (Bulfone et al.
1993) (not be confused with the hypothalamic preoptic area, which by all evidence
does not appear to be derived from the embryonic POA), and likely a region known
as the entopeduncular region (AEP) (Fig. 12.2). These niches are located ventral to
the MGE and CGE, respectively, and are the ventral-most telencephalic domains.
Genetic fate mapping studies have revealed contributions of Nkx5.1+ POA-derived
neurons to the MEA (Although their adult identity was not established) (Gelman et
al. 2011). In addition, more in depth fate mapping studies have revealed Dbx1+ and
Foxp2+ contributions to the MEA from domains at the telencephalic–diencephalic
boundary. As defined by molecular and electrophysiological profiles, Dbx1- and
Foxp2-derived populations generate distinct subclasses of putative inhibitory output
neurons that are distributed across different subnuclei of the MEA (Lischinsky et al.
2017). These developmentally defined subpopulations also contribute to known
neuroendocrine-expressing populations such as those that express Aromatase,
Androgen Receptor (AR), and Estrogen receptor-α (ERα) (Table 12.3). Addition-
ally, as one moves caudally in the telencephalon there is a distinct compartment
where the pallium appears to wrap underneath the CGE, dubbed the caudo-ventral
pallium (CVP) (Ruiz-Reig et al. 2018) (Fig. 12.2). The CVP expresses the transcrip-
tion factor encoding gene, Ebf3, and the secreted factor encoding genes, Gdf10 and
Fgf15. These populations contribute to the MEA and appear to generate a subset of
excitatory neurons. This segregation of excitatory and inhibitory sources for MEA
neuronal diversity follows the current dogma that the subpallial telencephalon is the
12 Development of Limbic System Stress-Threat Circuitry 333

exclusive (or almost exclusive) source of telencephalic inhibitory neurons (both

local interneurons and long-range projecting neurons) and the pallium is the source
of excitatory neurons. Interestingly, and perhaps unique for telencephalic structures,
the MEA also appears to draw neurons from the more distant brain regions such as
the diencephalon; the hypothalamic, and thalamic primordium. Multiple studies
have uncovered waves of long-range migration and settling of diencephalic-derived
populations in the emerging MEA. These populations, which appear to arise primar-
ily from the lateral/parventricular and/or caudal aspects of the hypothalamus, cross
the telencephalic–diencephalic border to reach the MEA (Bupesh et al. 2011; García-
Moreno et al. 2010; Soma et al. 2009; Zhao et al. 2008). Diencephalic-derived
progenitors express in combination or separately, the transcription factor encoding
genes, Otp, Lhx5, or Foxb1 (Fig. 12.2 and Table 12.3). Those expressing Otp or
Lhx5 appear to generate a subset of excitatory neurons of the MEA. Although not yet
formally established, based on their site of origin and embryonic molecular gene
expression patterns these neurons appear to be a distinct population from the dorsal
telencephalic Emx1-derived or CVP-originating excitatory neurons. These genetic
and site of origin differences suggest that excitatory neurons of the MEA, similar to
inhibitory neurons, are likely a highly diverse group composed of subtypes with
different projection patterns, transcriptome profiles, and biophysical properties.
In summary, work from a number of studies described above has revealed that the
MEA is composed of a “patchwork” of neurons generated from spatially distinct
embryonic niches derived from: (1) common sources that generate interneurons
(MGE, CGE) or excitatory neurons (the dorsal telencephalon) destined for the
MEA and other brain regions such as the striatum, cerebral cortex, and hippocampus
and (2) unique niches highly dedicated to the MEA such as the POA and specific
hypothalamic progenitor zones (Figs. 12.2, 12.3 and Table 12.3). The contribution
from numerous different niches likely results in the vast diversity of functionally
distinct neuronal subtypes that populate the MEA, many of which express different
neuropeptides and neurohormones. In addition, neural progenitors that respond to
the secreted patterning factor, Sonic Hedgehog (Shh), differentiate into a sizable
portion of MEA neurons. This likely is reflective of the widespread and earlier
patterning function of Shh in ventralization across the neuraxis (Ericson et al. 1995;
Matise and Wang 2011). In addition, there have been a handful of studies examining
the function of a subset of genes, such as Nkx2.1, Otp, and CoupTF-II, known to
mark MEA progenitor populations (Table 12.2, and references therein). These genes
are individually required for generation and/or migration of subsets of MEA
interneurons (Nkx2.1, CoupTF-II) or excitatory neurons (Otp) (Table 12.4). Thus,
perhaps not surprisingly, genes that mark distinct MEA progenitor subclasses also
appear to be required for their proper development.

12.3.3 Development of the BNST

Located within the ventral telencephalon, the BNST, sometimes also referred to as
“the extended amygdala” is bounded by the globus pallidus and traverses from the
334 N. S. Canteras et al.

Table 12.4 Genes known to function in MEA development

Gene Gene function References
Otp Promotes migration of Lhx5+ presumptive excitatory García-Moreno et al.
neurons from PVH to MeA (2010)
Nkx2.1 Specification of subset of MePD Sst + neurons and Carney et al. (2010)
regulation of Lhx9 and Shh expression
Coup- Via regulation of Nrp2 expression serves as a molecular Kanatani et al. (2015);
TFII switch regulating MeA versus cerebral cortical directed Tang et al. (2012)
migration of a POA-derived presumptive local inhibitory
neurons

septal area to anterior to the hypothalamus (Figs. 12.2 and 12.3). It is composed of up
to 18 distinct subnuclei which are broadly grouped as either part of the anterior or
posterior BNST (Lebow and Chen 2016-Key Reference). As revealed largely from
the work of Larry Swanson and collaborators, the BNST sends and receives
projections from a variety of brain regions, most notably major subnuclei of the
amygdala and hypothalamus (Dong et al. 2001; Thompson and Swanson 1998). At
the neuronal level, the BNST expresses a large variety of neurohormones and
neuropeptides, which points to its central role in modulation of mood, anxiety, and
stress responses. Unfortunately, little work has been undertaken to understand how
this complex and important structure arises, thus BNST development remains a
critical understudied area. However, from a handful of studies, some insight has been
gained. First, classical Brdu birthdating studies from Shirley Bayer, a pioneer along
with Joseph Altman in the use of methodologies to birthdate neurons, showed that
cells that will populate the BNST are born across a rostral to caudal gradient (rostral
earlier) between E13 and E20 in the rat (Bayer 1987). Subsequent studies by George
De Vries have revealed distinct temporal windows within this period for the sexual
dimorphic development of vasopressin (Avp) + and Galanin (Gal+) cells during the
earlier waves of BNST generation (Han and De Vries 1999; al-Shamma and De
Vries 1996). Following Brdu labeling over time the work from Bayer (Bayer 1987)
suggested that BNST cells arise from the medial edge of the MGE near the septum,
and a more dorsal and posterior source located at the approximate level of the CGE
(Note while the term “ganglionic eminences” were first used in the 1970s, it was not
until the early 1990s that the terms MGE, LGE, and CGE were in use). Following
these early studies, more recent gene expression studies carried out primarily from
the laboratories of Loreta Medina and Luis Puelles have extended our understanding
of BNST development (Bupesh et al. 2011; García-López et al. 2008; Puelles et al.
2000). This work, combined with work of others (Hirata et al. 2009; Nery et al. 2002;
Ruiz-Reig et al. 2018; Touzot et al. 2016; Vucurovic et al. 2010), indicates that
patterning of the BNST follows similar rules as MEA development with multiple
sources contributing to the BNST (Table 12.5, and references therein). These
sources likely include the CGE, POA, MGE (AEP), LGE, and perhaps the ventral
pallium (VP) (Figs. 12.2, 12.3 and Table 12.5). However, considering the wide
variety of neuronal cell types that populate the BNST, there are likely other sources
including the dorsal telencephalon as well as extratelencephalic sources (Bupesh et
al. 2011).
12 Development of Limbic System Stress-Threat Circuitry 335

Table 12.5 Known Origins of BNST Neuronal Diversity

Embryonic
origin Cell types/markers References
CGE 5HT3a + presumptive Nery et al. (2002); Ruiz-Reig et al. (2018); Touzot et al.
interneurons (2016); Vucurovic et al. (2010)
POA/VP? Dbx1-derived neurons Hirata et al. (2009)
MGE/AEP Nkx2.1-derived and García-López et al. (2008); Puelles et al. (2000)
Lhx6+ cells
LGE Islet 1+, Dlx5+ cells Bupesh et al. (2011); García-López et al. (2008)
dLGE/VP Pax6+ cells Bupesh et al. (2011)
(VP origin)

12.3.4 Development of Connectivity

The above subsections detail known mechanisms of how diverse neurons of key
threat detection structures are specified during embryonic development. However,
how these circuits become wired together still remains unclear. A number of lines of
evidence suggest that circuitry within this system is laid down relatively early. For
example, studies from the lab of Richard Simerly have shown that circuitry between
the BNST and MEA is already established by postnatal day (P)5 (Hutton et al. 1998).
Moreover, although not experimental evidence, reporter expression studies of limbic
system expressed genes reveal that neuronal tracts of the limbic system can be
observed as early as E15.5 (Fig. 12.4).

12.3.5 Linking Developmental Genetic Programs to Behavioral

Regulation

One aspect of high interest is to understand how the above-described developmental

programs may relate to behavioral regulation. One intriguing idea is that transcrip-
tion factor developmental programs direct stereotypical patterns of circuit formation
in turn regulate different and specific components of hardwired behaviors. Thus,
embryonic patterning may represent a continuum linking development directly to
circuit function and ultimately the behaviors regulated by these circuits. Although
such a link is far from being realized, some evidence within the stress/threat
detection network suggests that developmental patterning mechanisms in mice
directly guide specific behavioral outputs. One such example is the regulation of
stress behavior by the embryonic expressed transcription factor, Otp. At the devel-
opmental genetic level, as described in Sect. 12.3.1, Otp is a key factor implicated in
development of the PVN and the MEA. In the PVN, Otp expression is regulated by
Fezf2, and Otp directly regulates expression of corticotropin releasing hormone
336 N. S. Canteras et al.

Fig. 12.4 Gensat images. DAB immunostained images of GFP expression from BAC driven cre
recombination revealing gene expression of the cell adhesion encoding genes Pchd20, Amigo2 or
Pchd9 by E15.5. Recombination is observed in the olfactory bulb (OB, with asterisk highlighting
AOB), amygdala and hypothalamus, with presumptive projections observed between the OB and
amygdala/hypothalamus (arrow). These images suggest establishment of limbic circuits as by as
early as E15.5. Images courtesy of the Gene Expression Nervous System Atlas (GENSAT) Project,
NINDS Contracts N01NS02331 & HHSN271200723701C to The Rockefeller University
(New York, NY). Abbreviations: Amy amygdala, AOB accessory olfactory bulb, Hyp hypothala-
mus, OB olfactory bulb.

(CRH), a key molecule in modulating stress responses (Amir-Zilberstein et al. 2012;

Bedont et al. 2015; Wircer et al. 2017). Otp mutants display anxiety-like behaviors
and increased CRH expression. Thus, the embryonic patterning gene, Otp, functions
directly in establishment of the stress (HPA) axis. A second example is the apparent
modular function of Dbx1 in development of stress/threat detection circuits in the
arcuate nucleus and lateral hypothalamus. As mentioned above, Dbx1 marks a
subpopulation of neurons destined for the MEA (Figs. 12.2, 12.3 and Table 12.2).
In addition, similar to many other MEA embryonic expressed genes, Dbx1 is
expressed throughout the developing hypothalamus (Sokolowski et al. 2016;
Shimogori et al. 2010). Conditional deletion of Dbx1 selectively in the hypothala-
mus results in a sex-specific deficit in innate fear responses to predator odor
(Sokolowski et al. 2015). Correlating with this lack of innate fear response is a
reduced expression of stress-induced corticosterone, and a lack of heightened neu-
ronal activity in response to stress. Dbx1 conditional mutant mice also do not display
normal stress responses to food deprivation. These behavioral phenotypes are linked
to embryonic misspecification of the lateral and arcuate nucleus hypothalamic
neurons that are known to activate the PVN, without deficits in specification of the
PVN, the site of CRH expression. Thus, by acting during a transient window of
embryogenesis, Dbx1 sets in motion a series of developmental events that are
required for later behavioral and physiological responses to stress. Thus, work to
date has revealed that at least two genes; one likely acting in a region autonomous
manner to specific CRH neurons (Otp) and the other likely acting in a region
nonautonomous manner to specify neurons that feed into the PVN (Dbx1), are
required for specification of the stress/threat axis. This work also highlights the
parallel embryonic genetic processes in which the brain’s stress systems are
specified.
12 Development of Limbic System Stress-Threat Circuitry 337

12.4 Important Areas of Future Investigation

The above subsections describe current (early 2019) knowledge of the circuity and
function of the stress/threat system (Sects. 12.2 and 12.3) and how neurons that
comprise key stress/threat systems are specified from development, with specific
examples of how key transcription factors establish subsystems required for select
behaviors (Sect. 12.4). Despite this knowledge, there remain a number of key open
questions, highlighted in Box 3 below:

Box 3 This Box Contains Interesting Information about Key Questions for
Future Research
Question #1: Do embryonic developmental programs contribute to sex
differences in stress/threat system behaviors, or is this solely governed by
later hormonal effects?
Importance: Many stress linked disorders such as PTSD and anxiety
disorders, as well as behavioral disorders characterized by altered stress
responses, such as autism spectrum disorders, display a sex bias in presenta-
tion and prevalence. In addition to addressing basic biological questions of
how male and female circuitry is established in sexually dimorphic brain
regions such as the BNST, MEA and hypothalamus, answering Question #1
will likely inform and possibly provide better therapeutic approaches to
treating these prevalent conditions.
Question #2: How do environmental stressors affect formation of these
“hardwired” systems?
Importance: While formation of BNST-MEA-Hypothalamic circuits
appears to be laid down primarily via developmental genetic programming
mechanisms, these circuits are highly malleable by environmental influences
during both fetal (embryonic) and early post-natal critical periods. Under-
standing the interplay between genes and environment is a critical unanswered
question that has implications for a variety of human conditions described
above in Question #1.
Question #3: Are there evolutionarily conserved common “blueprints” for
establishing stress/threat system connectivity?
Importance: Normal responses to immediate environmental threats allow
for survival of individuals across all species. This commonality may suggest
that there are evolutionarily conserved mechanisms for development of threat
systems across a diverse species, perhaps via utilization of similar genetic
modules for neuron and circuit specification. If so, this may expand the
repertoire of animal model systems in which to model human disorders
affecting stress/threat systems.
338 N. S. Canteras et al.

Key References

Choi et al. (2005)—This article identifies subcircuits of the medial amygdala that are
genetically identifiable by expression of the LIM-family of transcription factors
and posits that these subcircuits control different innate behaviors.
Gross and Canteras (2012)—This review nicely describes the brain circuits
controlling innate and learned fear.
Hirata et al. (2009)—Using genetic tracing techniques this study revealed that there
are progenitor zones in the developing telencephalon that are dedicated to
generate neuronal diversity in the amygdala.
Hashikawa et al. (2016)—This review eloquently describes the olfactory to deeper
brain circuits that translate odor information to innate behavioral output.
Lebow and Chen (2016)—This review provides an excellent summary the anatomy,
function and role of the bed nucleus of the stria terminalis in psychiatric disorders.
Luo, et al. (2018)—This review provides an very comprehensive description of the
state of the art tools of modern neuroscience.
Motta et al. (2009)—Using a combination of approaches, this study revealed that
there are distinct hypothalamic circuits for processing different fear behaviors.

References
al-Shamma HA, De Vries GJ (1996) Neurogenesis of the sexually dimorphic vasopressin cells of
the bed nucleus of the stria terminalis and amygdala of rats. J Neurobiol 29:91–98
Alvarez-Bolado G (2018) Development of neuroendocrine neurons in the mammalian hypothala-
mus. Cell Tissue Res 375(1):23–39
Amir-Zilberstein L, Blechman J, Sztainberg Y, Norton WHJ, Reuveny A, Borodovsky N, Tahor M,
Bonkowsky JL, Bally-Cuif L, Chen A et al (2012) Homeodomain protein otp and activity-
dependent splicing modulate neuronal adaptation to stress. Neuron 73:279–291
Aravanis AM, Wang LP, Zhang F, Meltzer LA, Mogri MZ, Schneider MB, Deisseroth K (2007) An
optical neural interface: in vivo control of rodent motor cortex with integrated fiberoptic and
optogenetic technology. J Neural Eng 4:S143–S156
Armbruster B, Roth B (2005) Creation of designer biogenic amine receptors via directed molecular
evolution. Neuropsychopharmacology 30:S265–S265
Armbruster BN, Li X, Pausch MH, Herlitze S, Roth BL (2007) Evolving the lock to fit the key to
create a family of G protein-coupled receptors potently activated by an inert ligand. Proc Natl
Acad Sci U S A 104:5163–5168
Batista-Brito R, Fishell G (2009) The developmental integration of cortical interneurons into a
functional network. Curr Top Dev Biol 87:81–118
Bayer SA (1987) Neurogenetic and morphogenetic heterogeneity in the bed nucleus of the stria
terminalis. J Comp Neurol 265:47–64
Bedont JL, Newman EA, Blackshaw S (2015) Patterning, specification, and differentiation in the
developing hypothalamus. Wiley Interdiscip Rev Dev Biol 4:445–468
Berndt A, Lee SY, Ramakrishnan C, Deisseroth K (2014) Structure-guided transformation of
Channelrhodopsin into a light-activated Chloride Channel. Science 3:420–424
Bian X (2013) Physiological and morphological characterization of GABAergic neurons in the
medial amygdala. Brain Res 1509:8–19
Bian X, Yanagawa Y, Chen WR, Luo M (2008) Cortical-like functional organization of the
pheromone-processing circuits in the medial amygdala. J Neurophysiol 99:77–86
12 Development of Limbic System Stress-Threat Circuitry 339

Boyden ES, Zhang F, Bamberg E, Nagel G, Deisseroth K (2005) Millisecond-timescale, genetically

targeted optical control of neural activity. Nat Neurosci 8:1263–1268
Brechbühl J, Moine F, Klaey M, Nenniger-Tosato M, Hurni N, Sporkert F et al (2013) Mouse alarm
pheromone shares structural similarity with predator scents. Proc Natl Acad Sci U S A
110:4762–4767
Bulfone A, Puelles L, Porteus MH, Frohman MA, Martin GR, Rubenstein JL (1993) Spatially
restricted expression of Dlx-1, Dlx-2 (Tes-1), Gbx-2, and Wnt-3 in the embryonic day 12.5
mouse forebrain defines potential transverse and longitudinal segmental boundaries. J Neurosci
13:3155–3172
Bupesh M, Legaz I, Abellán A, Medina L (2011) Multiple telencephalic and extratelencephalic
embryonic domains contribute neurons to the medial extended amygdala. J Comp Neurol
519:1505–1525
Canteras NS, Swanson LW (1992) The dorsal premamillary nucleus: an unusual component of the
mamillary body. Proc Natl Acad Sci U S A 89:10089–10093
Canteras NS, Simerly RB, Swanson LW (1992) Projections of the ventral premamillary nucleus. J
Comp Neurol 324:195–212
Canteras NS, Simerly RB, Swanson LW (1994) Organization of projections from the ventromedial
nucleus of the hypothalamus: a Phaseolus vulgaris leucoagglutinin study in the rat. J Comp
Neurol 348:41–79
Canteras NS, Simerly RB, Swanson L (1995) Organization of projections from the medial nucleus
of the amygdala: a PHAL study in the rat. J Comp Neurol 360:213–245
Carney RSE, Mangin J-M, Hayes L, Mansfield K, Sousa VH, Fishell G, Machold RP, Ahn S, Gallo
V, Corbin JG (2010) Sonic hedgehog expressing and responding cells generate neuronal
diversity in the medial amygdala. Neural Develop. 5:14
Carter M, Shieh JC (2010) Guide to research techniques in neuroscience. Elsevier, Burlington, MA,
pp 1–375
Cezario AF, Ribeiro-Barbosa ER, Baldo MV, Canteras NS (2008) Hypothalamic sites responding
to predator threats – the role of the dorsal premamillary nucleus in unconditioned and
conditioned antipredatory defensive behavior. Eur J Neurosci 28:1003–1015
Choi GB, Dong H-W, Murphy AJ, Valenzuela DM, Yancopoulos GD, Swanson LW, Anderson DJ
(2005) Lhx6 delineates a pathway mediating innate reproductive behaviors from the amygdala
to the hypothalamus. Neuron 46:647–660
Chow BY, Han X, Dobry AS, Qian X, Chuong AS, Li M, Henninger MA, Belfort GM, Lin Y,
Monahan PE, Boyden ES (2010) High-performance genetically targetable optical neural silenc-
ing by light-driven proton pumps. Nature 463:98–102
Coward P, Wada HG, Falk MS, Chan SD, Meng F, Akil H, Conklin BR (1998) Controlling
signaling with a specifically designed Gi-coupled receptor. Proc Natl Acad Sci U S A
95:352–357
Cocas LA, Miyoshi G, Carney RS, Sousa VH, Hirata T, Jones KR, Fishell G, Huntsman MM,
Corbin JG (2009) Emx1-lineage progenitors differentially contribute to neural diversity in the
striatum and amygdala. J Neurosci 29(50):15933–15946. https://doi.org/10.1523/JNEUROSCI.
2525-09.2009. PubMed PMID: 20016109; PubMed Central PMCID: PMC3679174
Davis M, Walker DL, Miles L, Grillon C (2010) Phasic vs sustained fear in rats and humans: role of
the extended amygdala in fear vs anxiety. Neuropsychopharmacology 35:105–135
Dewan A, Pacifico R, Zhan R, Rinberg D, Bozza T (2013) Non redundant coding of aversive
odours in the main olfactory pathway. Nature 497:486–489
Dong HW, Petrovich GD, Swanson LW (2001) Topography of projections from amygdala to bed
nuclei of the stria terminalis. Brain Res Brain Res Rev 38:192–246
Dong HW, Swanson LW (2004) Projections from bed nuclei of the stria terminalis, posterior
division: implications for cerebral hemisphere regulation of defensive and reproductive
behaviors. J Comp Neurol 471(4):396–433. Erratum in: J Comp Neurol. 2004 Jul 5;474
(4):603-4. PubMed PMID: 15022261
340 N. S. Canteras et al.

Ericson J, Muhr J, Jessell TM, Edlund T (1995) Sonic hedgehog: a common signal for ventral
patterning along the rostrocaudal axis of the neural tube. Int J Dev Biol 39:809–816
Fenno L, Yizhar O, Deisseroth K (2011) The development and application of optogenetics. Annu
Rev Neurosci 34:389–412
Ferran JL, Puelles L, Rubenstein JLR (2015) Molecular codes defining rostrocaudal domains in the
embryonic mouse hypothalamus. Front Neuroanat 9:46
Ferrero DM, Lemon JK, Fluegge D, Pashkovski SL, Korzan WJ, Datta SR et al (2011) Detection
and avoidance of a carnivore odor by prey. Proc Natl Acad Sci U S A 108:11235–11240
García-López M, Abellán A, Legaz I, Rubenstein JLR, Puelles L, Medina L (2008) Histogenetic
compartments of the mouse centromedial and extended amygdala based on gene expression
patterns during development. J Comp Neurol 506:46–74
García-Moreno F, Pedraza M, Di Giovannantonio LG, Di Salvio M, López-Mascaraque L, Simeone
A, De Carlos JA (2010) A neuronal migratory pathway crossing from diencephalon to telen-
cephalon populates amygdala nuclei. Nat Neurosci 13:680–689
Gelman D, Griveau A, Dehorter N, Teissier A, Varela C, Pla R, Pierani A, Marín O (2011) A wide
diversity of cortical GABAergic interneurons derives from the embryonic preoptic area. J
Neurosci 31:16570–16580
Gomez JL, Bonaventura J, Lesniak W, Mathews WB, Sysa-Shah P, Rodriguez LA, Ellis RJ, Richie
CT, Harvey BK, Dannals RF, Pomper MG, Bonci A, Michaelides M (2017) Chemogenetics
revealed: DREADD occupancy and activation via converted clozapine. Science 357:503–507
Gorski JA, Talley T, Qiu M, Puelles L, Rubenstein JLR, Jones KR (2002) Cortical excitatory
neurons and glia, but not GABAergic neurons, are produced in the Emx1-expressing lineage. J
Neurosci 22:6309–6314
Goto M, Canteras NS, Burns G, Swanson LW (2005) Projections from the subfornical region of the
lateral hypothalamic area. J Comp Neurol 493:412–438
Gradinaru V, Zhang F, Ramakrishnan C, Mattis J, Prakash R, Diester I, Goshen I, Thompson KR,
Deisseroth K (2010) Molecular and cellular approaches for diversifying and extending
optogenetics. Cell 141:154–165
Gross CT, Canteras NS (2012) The many paths to fear. Nat Rev Neurosci 13:651–658
Guirado S, Real MA, Dávila JC (2008) Distinct immunohistochemically defined areas in the medial
amygdala in the developing and adult mouse. Brain Res Bull 75:214–217
Han TM, De Vries GJ (1999) Neurogenesis of galanin cells in the bed nucleus of the stria terminalis
and centromedial amygdala in rats: a model for sexual differentiation of neuronal phenotype. J
Neurobiol 38:491–498
Hashikawa K, Hashikawa Y, Falkner A, Lin D (2016) The neural circuits of mating and fighting in
male mice. Curr Opin Neurobiol 38:27–37
Hashikawa K, Hashikawa Y, Lischinsky J, Lin D (2018) The neural mechanisms of sexually
dimorphic aggressive behaviors. Trends Genet TIG 34:755–776
Hegemann P, Nagel G (2013) From Channelrhodopsins to Optogenetics. EMBO Mol Med
5:173–176
Hirata T, Li P, Lanuza GM, Cocas LA, Huntsman MM, Corbin JG (2009) Identification of distinct
telencephalic progenitor pools for neuronal diversity in the amygdala. Nat Neurosci 12:141–149
Hutton LA, Gu G, Simerly RB (1998) Development of a sexually dimorphic projection from the
bed nuclei of the stria terminalis to the anteroventral periventricular nucleus in the rat. J
Neurosci 18:3003–3013
Keshavarzi S, Sullivan RKP, Ianno DJ, Sah P (2014) Functional properties and projections of
neurons in the medial amygdala. J Neurosci 34:8699–8715
Kunwar PS, Zelikowsky M, Remedios R, Cai H, Yilmaz M, Meister M, Anderson DJ (2015)
Ventromedial hypothalamic neurons control a defensive emotion state. elife 4:e06633
Kanatani S, Honda T, Aramaki M, Hayashi K, Kubo K, Ishida M, Tanaka DH, Kawauchi T, Sekine
K, Kusuzawa S, Kawasaki T, Hirata T, Tabata H, Uhlén P, Nakajima K (2015) The COUP-TFII/
Neuropilin-2 is a molecular switch steering diencephalon-derived GABAergic neurons in the
developing mouse brain. Proc Natl Acad Sci U S A 112(36):E4985–E4994. https://doi.org/10.
1073/pnas.1420701112. Epub 2015 Aug 24. PubMed PMID: 26305926; PubMed Central
PMCID: PMC4568674
12 Development of Limbic System Stress-Threat Circuitry 341

Kuerbitz J, Arnett M, Ehrman S, Williams MT, Vorhees CV, Fisher SE, Garratt AN, Muglia LJ,
Waclaw RR, Campbell K (2018) Loss of Intercalated Cells (ITCs) in the mouse amygdala of
Tshz1 mutants correlates with fear, depression, and social interaction phenotypes. J Neurosci 38
(5):1160–1177. https://doi.org/10.1523/JNEUROSCI.1412-17.2017. Epub 2017 Dec 18.
PubMed PMID: 29255003; PubMed Central PMCID: PMC5792476
Lebow MA, Chen A (2016) Overshadowed by the amygdala: the bed nucleus of the stria terminalis
emerges as key to psychiatric disorders. Mol Psychiatry 21:450–463
Li Y, Dulac C (2018) Neural coding of sex-specific social information in the mouse brain. Curr
Opin Neurobiol 53:120–130
Liberles SD (2015) Trace amine-associated receptors: ligands, neural circuits, and behaviors. Curr
Opin Neurobiol 34C:1–7
Lischinsky JE, Sokolowski K, Li P, Esumi S, Kamal Y, Goodrich M, Oboti L, Hammond TR,
Krishnamoorthy M, Feldman D et al (2017) Embryonic transcription factor expression in mice
predicts medial amygdala neuronal identity and sex-specific responses to innate behavioral cues.
elife 6:e21012
Luo L, Callaway EM, Svoboda K (2008) Genetic dissection of neural circuits. Neuron 57
(5):634–660
Luo LQ, Callaway EM, Svoboda K (2018) Genetic dissection of neural circuits: a decade of
Progress. Neuron 98:256–281
Madisen L, Mao T, Koch H, Zhuo JM, Berenyi A, Fujisawa S, Hsu YW, Garcia AJ 3rd, Gu X,
Zanella S, Kidney J, Gu H, Mao Y, Hooks BM, Boyden ES, Buzsaki G, Ramirez JM, Jones AR,
Svoboda K, Han X, Turner EE, Zeng H (2012) A toolbox of Cre-dependent optogenetic
transgenic mice for light-induced activation and silencing. Nat Neurosci 15:793–802
Magnus CJ, Lee PH, Bonaventura J, Zemla R, Gomez JL, Ramirez MH, Hu X et al (2019)
Ultrapotent Chemogenetics for research and potential clinical applications. Science 364
(6436):eaav5282. https://doi.org/10.1126/science.aav5282. [Epub ahead of print]
Martinez RC, Carvalho-Netto EF, Ribeiro-Barbosa ER, Baldo MV, Canteras NS (2011) Amygdalar
roles during exposure to a live predator and to a predator-associated context. Neuroscience
172:314–328
Matise MP, Wang H (2011) Sonic hedgehog signaling in the developing CNS where it has been and
where it is going. Curr Top Dev Biol 97:75–117
McGregor IS, Hargreaves GA, Apfelbach R, Hunt GE (2004) Neural correlates of cat odor-induced
anxiety in rats: region-specific effects of the benzodiazepine midazolam. J Neurosci
24:4134–4144
Motta SC, Goto M, Gouveia FV, Baldo MV, Canteras NS, Swanson LW (2009) Dissecting the
brain's fear system reveals the hypothalamus is critical for responding in subordinate conspecific
intruders. Proc Natl Acad Sci U S A 106:4870–4875
Nery S, Fishell G, Corbin JG (2002) The caudal ganglionic eminence is a source of distinct cortical
and subcortical cell populations. Nat Neurosci 5:1279–1287
Neve RL, Neve KA, Nestler EJ, Carlezon WA (2005) Use of herpes virus amplicon vectors to study
brain disorders. BioTechniques 39:381–391
Nouri N, Awatramani R (2017) A novel floor plate boundary defined by adjacent En1 and Dbx1
microdomains distinguishes midbrain dopamine and hypothalamic neurons. Dev Camb Engl
144:916–927
Pérez-Gómez A, Bleymehl K, Stein B, Pyrski M, Birnbaumer L, Munger SD et al (2015) Innate
predator odor aversion driven by parallel olfactory subsystems that converge in the ventromedial
hypothalamus. Curr Biol 25:1340–1346
Petilla Interneuron Nomenclature Group, Ascoli GA, Alonso-Nanclares L, Anderson SA,
Barrionuevo G, Benavides-Piccione R, Burkhalter A, Buzsáki G, Cauli B, Defelipe J et al
(2008) Petilla terminology: nomenclature of features of GABAergic interneurons of the cerebral
cortex. Nat Rev Neurosci 9:557–568
Petrovich GD, Risold PY, Swanson LW (1996) Organization of the projections of the basomedial
nucleus of the amygdala: a PHAL study in the rat. J Comp Neurol 374:387–420
342 N. S. Canteras et al.

Puelles L, Kuwana E, Puelles E, Bulfone A, Shimamura K, Keleher J, Smiga S, Rubenstein JL

(2000) Pallial and subpallial derivatives in the embryonic chick and mouse telencephalon, traced
by the expression of the genes Dlx-2, Emx-1, Nkx-2.1, Pax-6, and Tbr-1. J Comp Neurol
424:409–438
Puelles L, Morales-Delgado N, Merchán P, Castro-Robles B, Martínez-de-la-Torre M, Díaz C,
Ferran JL (2016) Radial and tangential migration of telencephalic somatostatin neurons
originated from the mouse diagonal area. Brain Struct Funct 221(6):3027–3065. https://doi.
org/10.1007/s00429-015-1086-8. Epub 2015 Jul 19. PubMed PMID: 26189100; PubMed
Central PMCID: PMC4920861
Real MA, Heredia R, Labrador MDC, Dávila JC, Guirado S (2009) Expression of somatostatin and
neuropeptide Y in the embryonic, postnatal, and adult mouse amygdalar complex. J Comp
Neurol 513:335–348
Risold PY, Canteras NS, Swanson LW (1994) Organization of projections from the anterior
hypothalamic nucleus: a Phaseolus vulgaris leucoagglutinin study in the rat. J Comp Neurol
348:1–40
Roth BL (2016) DREADDs for neuroscientists. Neuron 89:683–694
Ruiz-Reig N, Andres B, Lamonerie T, Theil T, Fairén A, Studer M (2018) The caudo-ventral
pallium is a novel pallial domain expressing Gdf10 and generating Ebf3-positive neurons of the
medial amygdala. Brain Struct Funct 223:3279–3295
Schobert B, Lanyi JK (1982) Halorhodopsin is a light-driven chloride pump. J Biol Chem
257:306–313
Shimada M, Nakamura T (1973) Time of neuron origin in mouse hypothalamic nuclei. Exp Neurol
41:163–173
Shimogori T, Lee DA, Miranda-Angulo A, Yang Y, Wang H, Jiang L, Yoshida AC, Kataoka A,
Mashiko H, Avetisyan M et al (2010) A genomic atlas of mouse hypothalamic development.
Nat Neurosci 13:767–775
Silva BA, Mattucci C, Krzywkowski P, Murana E, Illarionova A, Grinevich V, Canteras NS,
Ragozzino D, Gross CT (2013) Independent hypothalamic circuits for social and predator fear.
Nat Neurosci 16:1731–1733
Simerly RB, Swanson LW (1988) Projections of the medial preoptic nucleus: a Phaseolus vulgaris
leucoagglutinin anterograde tract-tracing study in the rat. J Comp Neurol 270:209–242
Sokolowski K, Corbin JG (2012) Wired for behaviors: from development to function of innate
limbic system circuitry. Front Mol Neurosci 5:55
Sokolowski K, Esumi S, Hirata T, Kamal Y, Tran T, Lam A, Oboti L, Brighthaupt S-C, Zaghlula M,
Martinez J et al (2015) Specification of select hypothalamic circuits and innate behaviors by the
embryonic patterning gene dbx1. Neuron 86:403–416
Sokolowski K, Tran T, Esumi S, Kamal Y, Oboti L, Lischinsky J, Goodrich M, Lam A, Carter M,
Nakagawa Y et al (2016) Molecular and behavioral profiling of Dbx1-derived neurons in the
arcuate, lateral and ventromedial hypothalamic nuclei. Neural Develop 11:12
Soma M, Aizawa H, Ito Y, Maekawa M, Osumi N, Nakahira E, Okamoto H, Tanaka K, Yuasa S
(2009) Development of the mouse amygdala as revealed by enhanced green fluorescent protein
gene transfer by means of in utero electroporation. J Comp Neurol 513:113–128
Stachniak TJ, Ghosh A, Sternson SM (2014) Chemogenetic synaptic silencing of neural circuits
localizes a hypothalamus-->midbrain pathway for feeding behavior. Neuron 82:797–808
Swanson LW (2000) Cerebral hemisphere regulation of motivated behavior. Brain Res
886:113–164
Swanson LW, Petrovich GD (1998) What is the amygdala? Trends Neurosci 21:323–331
Sullivan GM, Apergis J, Bush DE, Johnson LR, Hou M, LeDoux JE (2004) Lesions in the bed
nucleus of the stria terminalis disrupt corticosterone and freezing responses elicited by a
contextual but not by a specific cue-conditioned fear stimulus. Neuroscience 128:7–14
Takahashi LK (2014) Olfactory systems and neural circuits that modulate predator odor fear. Front
Behav Neurosci 8:72
12 Development of Limbic System Stress-Threat Circuitry 343

Tervo DGR, Hwang BY, Viswanathan S, Gaj T, Lavzin M, Ritola KD, Lindo S, Michael S,
Kuleshova E, Ojala D, Huang CC, Gerfen CR, Schiller J, Dudman JT, Hantman AW, Looger
LL, Schaffer DV, Karpova AY (2016) A designer AAV variant permits efficient retrograde
access to projection neurons. Neuron 92(2):372–382
Thompson RH, Swanson LW (1998) Organization of inputs to the dorsomedial nucleus of the
hypothalamus: a reexamination with Fluorogold and PHAL in the rat. Brain Res Brain Res Rev
27:89–118
Touzot A, Ruiz-Reig N, Vitalis T, Studer M (2016) Molecular control of two novel migratory paths
for CGE-derived interneurons in the developing mouse brain. Dev. Camb. Engl.
143:1753–1765
Tye KM, Deisseroth K (2012) Optogenetic investigation of neural circuits underlying brain disease
in animal models. Nat Rev Neurosci 13:251–266
Tang K, Rubenstein JL, Tsai SY, Tsai MJ (2012) COUP-TFII controls amygdala patterning by
regulating neuropilin expression. Development 139(9):1630–1639. https://doi.org/10.1242/dev.
075564. PubMed PMID: 22492355; PubMed Central PMCID: PMC3317968
Vardy E, Robinson JE, Li C, Olsen RHJ, DiBerto JF, Giguere PM, Sassano FM, Huang XP, Zhu H,
Urban DJ, White KL, Rittiner JE, Crowley NA, Pleil KE, Mazzone CM, Mosier PD, Song J,
Kash TL, Malanga CJ, Krashes MJ, Roth BL (2015) A new DREADD facilitates the
multiplexed Chemogenetic interrogation of behavior. Neuron 86:936–946
Vucurovic K, Gallopin T, Ferezou I, Rancillac A, Chameau P, van Hooft JA, Geoffroy H, Monyer
H, Rossier J, Vitalis T (2010) Serotonin 3A receptor subtype as an early and protracted marker
of cortical interneuron subpopulations. Cereb Cortex N Y N 1991(20):2333–2347
Wang L, Chen IZ, Lin D (2015) Collateral pathways from the ventromedial hypothalamus mediate
defensive behaviors. Neuron 85:1344–1358
Wang L, Talwar V, Osakada T, Kuang A, Guo Z, Yamaguchi T, Lin D (2019) Hypothalamic
control of conspecific self-defense. Cell Rep 26:1747–1758.e5
Wircer E, Blechman J, Borodovsky N, Tsoory M, Nunes AR, Oliveira RF, Levkowitz G (2017)
Homeodomain protein Otp affects developmental neuropeptide switching in oxytocin neurons
associated with a long-term effect on social behavior. elife 6:e22170
Wu YE, Pan L, Zuo Y, Li X, Hong W (2017) Detecting activated cell populations using single-cell
RNA-Seq. Neuron 96:313–329.e6
Wyatt TD (2014) Pheromones and animal behavior, 2nd edn. Cambridge University Press,
Cambridge, UK
Xie Y, Dorsky RI (2017) Development of the hypothalamus: conservation, modification and
innovation. Dev Camb Engl 144:1588–1599
Xu Q, de la Cruz E, Anderson SA (2003) Cortical interneuron fate determination: diverse sources
for distinct subtypes? Cereb Cortex N Y N 1991(13):670–676
Yang CF, Shah NM (2014) Representing sex in the brain, one module at a time. Neuron
82:261–278
Zhao T, Szabó N, Ma J, Luo L, Zhou X, Alvarez-Bolado G (2008) Genetic mapping of Foxb1-cell
lineage shows migration from caudal diencephalon to telencephalon and lateral hypothalamus.
Eur J Neurosci 28:1941–1955
Zhu H, Aryal DK, Olsen RH, Urban DJ, Swearingen A, Forbes S, Roth BL, Hochgeschwender U
(2016) Cre-dependent DREADD (designer receptors exclusively activated by designer drugs)
mice. Genesis 54:439–446
Organization and Postnatal Development
of Visceral Sensory Inputs 13
to the Neuroendocrine Hypothalamus

Linda Rinaman

Abstract
Sensory signals arising in the body’s internal organs are delivered to the central
nervous system by spinal and vagal afferent neurons. These signals modulate
pituitary hormone release via ascending neural pathways from the caudal
brainstem to the neuroendocrine hypothalamus. Interoceptive (i.e., visceral) feed-
back regarding the moment-to-moment state of gastrointestinal and other physio-
logical systems is critical for shaping and coordinating adaptive neuroendocrine
functions, including hormonal responses to real or perceived homeostatic threats.
This chapter summarizes the functional organization and postnatal development
of visceral sensory inputs to the neuroendocrine hypothalamus in rats and mice,
the two mammalian models in which most of the relevant data have been collected.
A special emphasis is placed on hypothalamic inputs arising from noradrenergic
(NA) and glucagon-like peptide 1 (GLP1) neurons, whose cell bodies occupy the
caudal nucleus of the solitary tract (NTS) and medullary reticular formation (RF).
The axonal projections of NA and GLP1 neurons directly target the hypothalamic
medial preoptic area (mPOA), arcuate nucleus (ARC), paraventricular nucleus
(PVN), and supraoptic nucleus (SON), where adrenergic and GLP1 receptor
signaling modulates neuroendocrine gene expression, pulsatile activity, and pitui-
tary hormone release. NA and GLP1 inputs to the neuroendocrine hypothalamus
mature postnatally in rats and mice, likely contributing to the developmental
maturation of pituitary hormone release in response to both interoceptive and
exteroceptive signals.

Keywords
Fos · Gastrointestinal · Glucagon-like peptide 1 · HPA axis · Noradrenergic ·
Nucleus of the solitary tract · Oxytocin
L. Rinaman (*)
Department of Psychology and Program in Neuroscience, Florida State University, Tallahassee, FL,
USA
e-mail: [email protected]

# Springer Nature Switzerland AG 2020 345

S. Wray, S. Blackshaw (eds.), Developmental Neuroendocrinology, Masterclass in
Neuroendocrinology 9, https://doi.org/10.1007/978-3-030-40002-6_13
346 L. Rinaman

13.1 Introduction

Neuroendocrine systems comprise magnocellular and parvocellular neurons that

control pituitary hormone release. The cell bodies of magnocellular oxytocin
(OT) and vasopressin (AVP) neurons are located within the paraventricular nucleus
(PVN), supraoptic nucleus (SON), and accessory cell groups distributed between
these two hypothalamic nuclei. The axons of magnocellular OT and AVP neurons
terminate within the posterior pituitary, where OT and AVP are released into the
systemic circulation. The PVN also houses parvocellular endocrine neurons whose
axon terminals secrete anterior pituitary hormone-releasing hormones [i.e., cortico-
tropin-releasing hormone (CRH) and vasopressin, thyrotropin releasing hormone
(TRH), and somatostatin (SS)] into the median eminence (ME), located at the base of
the hypothalamus; dopamine (DA), which inhibits prolactin secretion, is similarly
produced by tubero-infundibular (TIDA) neurons. The ME contains a specialized
portal vasculature that conveys these releasing hormones directly into the anterior
pituitary gland at high concentrations. Gonadotropin releasing hormone (GnRH) and
growth hormone-releasing hormone (GHRH) also are secreted into the ME from the
terminals of additional parvocellular endocrine neurons whose cell bodies are
located within the hypothalamic medial preoptic area (mPOA) and arcuate nucleus
(ARC).
Magnocellular and parvocellular endocrine neurons receive direct and relayed
synaptic input from the nucleus of the solitary tract (NTS), located within the caudal
brainstem (Fig. 13.1). The NTS receives interoceptive signals arising in gastrointes-
tinal (GI) and other organ systems, and ascending projections from the NTS to the
hypothalamus provide an interface through which visceral sensory signals influence
the production and release of pituitary hormones. For example, stimulation of GI
vagal sensory afferents in adult rats activates NTS neurons as well as magnocellular
OT neurons and hypophysiotropic CRH neurons, leading to increased plasma
concentrations of OT, adrenocorticotropic hormone (ACTH), and corticosterone.
Conversely, similar stimulation of vagal sensory afferents in 2-day-old rat pups
activates NTS neurons, as in adult rats, but fails to activate hypothalamic endocrine
neurons, which is evidence that ascending projections from the NTS to the neuroen-
docrine hypothalamus are functionally immature. Visceral signals relayed from the
NTS to the hypothalamus also impact the function of parvocellular TRH, GnRH,
GHRH, SS, and DA endocrine neurons in adult rodents; however, the anatomical
and functional maturation of these signaling pathways is relatively unexplored. This
chapter summarizes what is known from rodent models regarding the functional
organization and development of neural pathways that deliver visceral sensory
signals to neuroendocrine cells of the hypothalamus.

13.1.1 Visceral Sensory Inputs to the Spinal Cord

Sensory signals arising within the body’s internal organs (i.e., thoracic, abdominal,
and pelvic viscera) are carried to the CNS by spinal and cranial nerve afferents. In
rats, the cell bodies of spinal visceral sensory afferents occupy the dorsal root ganglia
13 Organization and Postnatal Development of Visceral Sensory Inputs to the. . . 347

Fig. 13.1 Mechanical and chemical signals arising within the stomach, intestines, and other
viscera are relayed to the caudal brainstem by vagal sensory afferents whose cell bodies occupy
the nodose ganglia (NG). The central axons of vagal afferents synapse within the nucleus of the
solitary tract (NTS) in the caudal medulla. Noradrenergic neurons (i.e., immunoreactive for
dopamine beta hydroxylase, DBH) and peptidergic neurons expressing preproglucagon and immu-
noreactive for glucagon-like peptide 1 (GLP1) are located within the NTS, reticular formation (RF),
and ventrolateral medulla (VLM), and carry visceral sensory signals from the caudal brainstem to
the neuroendocrine hypothalamus via the ventral noradrenergic bundle (VNAB). Lower left:
Postnatal changes in innervation of the neuroendocrine hypothalamus. In newborn rats and mice,
DBH- and GLP1-positive axons are sparse within the paraventricular nucleus (PVN) and other
hypothalamic endocrine nuclei, with innervation density increasing progressively to achieve adult-
like levels by the end of the third week postnatal (pink arrow)

(DRG); their axons terminate centrally in laminae I–VII of the spinal cord dorsal
horn and intermediate zone, and also in lamina X around the central canal. A subset
of the second-order spinal sensory neurons that receive visceral sensory inputs also
receive input from somatic sensory DRG neurons; thus, the activity of spinal sensory
neurons is co-modulated by visceral, cutaneous, and proprioceptive stimuli.
348 L. Rinaman

13.1.1.1 Visceral Sensory Pathways from Spinal Cord to Hypothalamus

Convergent somatic and visceral sensory signals are relayed from the spinal cord to
the diencephalon via the anterolateral spinothalamic tract. Separate ascending spinal
pathways appear to be more specialized for visceral sensory signaling; these
pathways include a direct spinosolitary tract pathway that conveys spinal sensory
information to the NTS (Menétrey and dePommery 1991). Ascending pathways
from the spinal cord are viscerotopically organized and provide both direct and
relayed inputs to the hypothalamus. Although the direct spinal inputs to the hypo-
thalamus do not appear to target endocrine neurons, they synapse within regions of
the lateral hypothalamus that do (Palkovits 1999).

13.1.2 Visceral Sensory Inputs to the Caudal Brainstem

Given the apparent lack of direct spinal inputs to the neuroendocrine hypothalamus,
the balance of this chapter will focus on the structure and development of direct
hypothalamic inputs arising from the NTS and medullary RF (Fig. 13.1). In addition
to spinal sensory signals that are relayed via the spinosolitary tract, NTS neurons
receive direct visceral sensory input via two cranial nerves, the glossopharyngeal (IX)
and the vagus (X). The cell bodies of glossopharyngeal and vagal afferent neurons
occupy the inferior glossopharyngeal (aka petrosal) and nodose ganglia, respectively.
The peripheral axons of glossopharyngeal and a subset of vagal afferents monitor
cardiovascular and pulmonary target tissues within the thoracic cavity, while separate
populations of vagal afferents monitor the GI tract and associated digestive viscera,
from the oral cavity through the colon. The centrally projecting axons of
glossopharyngeal and vagal neurons enter the caudal dorsolateral medulla via multi-
ple sensory rootlets to converge within the solitary tract, which conveys visceral
sensory axons along the rostrocaudal length of the NTS. Most of these sensory
afferents form synapses within the caudal visceral portion of the NTS (as opposed
to the more rostral gustatory NTS), but some visceral sensory axons terminate within
the area postrema (AP), a dorsal midline circumventricular organ that lacks a blood-
brain barrier. The AP lies in the floor of the caudal fourth ventricle, immediately
adjacent to the visceral portion of the NTS. AP neurons respond to interoceptive
signals conveyed by hormones, osmolytes, toxins, cytokines, and other circulating
factors and convey these signals to postsynaptic neurons in the NTS, medullary RF,
and pontine parabrachial nucleus (PBN) (Shapiro and Miselis 1985).

13.2 Anatomical Development of Vagal Afferent Inputs

to the NTS

Vagal sensory neurons within the rat nodose ganglia undergo their final mitotic
division between embryonic day (E)10 and E14, with peak cell production on E13.
Even as neural proliferation continues, the peripheral axons of vagal sensory neurons
start to innervate the GI tract, and their central axons begin to penetrate the caudal
13 Organization and Postnatal Development of Visceral Sensory Inputs to the. . . 349

Fig. 13.2 DiI tracing of vagal neurons during embryonic development in rats. DiI crystals were
placed into the developing gastric tube of formalin-fixed embryos harvested on embryonic day (E)
12 or E15. In the left panel, DiI-labeled vagal sensory neurons within the nodose ganglion (NG;
long arrows) are evident, along with their distal axons (asterisk) that are labeled due to their
peripheral association with the gastric tube. However, very few labeled NG neurons have axons
extending centrally into the caudal medulla (short arrows). In the right panel, DiI-labeled motor
neurons within the dorsal motor nucleus of the vagus (DMV) that innervate the gastric tube are
evident, along with the central axons of vagal sensory neurons that bundle within the solitary tract
(tr). Some of these sensory axons emerge to project into the developing NTS (nucleus tractus
solitarii, nucleus of the solitary tract; long arrows). Figure adapted from Rinaman and Levitt (1993)

medulla (Rinaman and Levitt 1993). Results from a study in which crystals of DiI
(1,10 -dioctadecyl-3,3,30 ,30 -tetramethylindocarbocyanine perchlorate) were placed
into the developing digestive tube in intact, formalin-fixed rat embryos (see Box
13.1) revealed that vagal sensory neurons within the nodose ganglia were first
labeled in E12 specimens (Fig. 13.2, NG), evidence that their peripheral axons
first reach the digestive tube at this time point. However, very few of the nodose
neurons labeled at E12 had central axons that entered the caudal medulla, and
labeled fibers within the solitary tract were not identifiable until E13. The density
and rostrocaudal distribution of labeled vagal afferents within the solitary tract
increased markedly between E13 and E14, but labeled vagal afferent axons did not
emerge from the tract until E15, when labeled fibers at all rostrocaudal levels coursed
directly from the solitary tract (Fig. 13.2, tr) into the proliferative ventricular zone of
the alar plate (Fig. 13.2, right, long arrows). Although vagal sensory fibers were not
observed to arborize within the developing NTS until E16, their distribution within
the NTS was remarkably adult-like by E18, approximately 3 days before birth
(Rinaman and Levitt 1993). Thus, the early embryonic development of vagal
sensory inputs to the NTS provides an anatomical substrate for digestive vago-
vagal reflexes (e.g., the gastric accommodation reflex) that already are operational
in newborn rats.
350 L. Rinaman

Box 13.1 Tracing Axonal Projections During Embryonic and Early

Postnatal Development
Two obstacles are immediately apparent when one considers how to
anatomically trace the early development of axonal projection pathways. Many
pathways are established before birth, even in precocial species such as rats and
mice, and neural tracing in live developing embryos presents a formidable
technical challenge. One strategic solution is provided by the use of carbocyanine
dyes such as DiI (1,10 -dioctadecyl-3,3,30 ,30 -tetramethylindocarbocyanine per-
chlorate). DiI travels by diffusion along continuous cellular lipids, fully labeling
the axonal and dendritic processes of mature and immature neurons that are
exposed to the dye anywhere along their plasma membrane (Godement et al.
1987; Rinaman and Levitt 1993; Bouret et al. 2004). Thus, DiI tracing permits
analysis of point-to-point axonal projections not only in living tissue but also in
aldehyde-fixed tissue samples, because dye diffusion along continuous lipid
membranes does not require microtubule-mediated transport. DiI tracing has
been used to reveal the prenatal development of vagal sensory and motor
pathways between the brainstem and digestive tube in aldehyde-fixed embryonic
and early postnatal rat pups (Rinaman and Levitt 1993) (Fig. 13.2). DiI also has
been used to reveal the early postnatal growth of intra-hypothalamic axonal
projections from the ARC to the PVN in fixed slabs of brain tissue dissected
from neonatal rodent pups, and to demonstrate the effect of postnatal leptin
signaling to stimulate axonal growth (Bouret et al. 2004). Generating these
data in embryonic and early neonatal animals would have been an impractical
(and perhaps impossible) feat using survival surgery for targeted delivery of
neural tracer, as typically applied in adult rats and mice.

13.2.1 Functional Development of Vagal Afferent Inputs to the NTS

In addition to displaying functional autonomic vago-vagal reflexes, newborn rats are

behaviorally responsive to at least some vagally mediated satiety signals that are
conveyed by GI vagal afferents to the caudal NTS. A key example of early
behavioral responsiveness to these signals is afforded by considering the effects of
systemically administered cholecystokinin octapeptide (CCK) in newborn rats.
Endogenous CCK is released from intestinal secretory cells during and after food
intake and nutrient absorption, in both adult and newborn animals. CCK acts by
binding to CCK-1 (aka CCK-A) receptors that are abundantly expressed in the GI
tract of newborn rats (Robinson et al. 1987), including expression by GI vagal
sensory afferents. In adult rats, CCK initiates and maintains various digestive
processes and also acts as a satiety signal that suppresses food intake by limiting
meal size through a central, vagally mediated mechanism. Low doses of exogenous
CCK robustly suppress independent milk intake in newborn rats (Robinson et al.
1988), evidence that the neural signaling pathway underlying this behavioral effect
already is functional at birth.
13 Organization and Postnatal Development of Visceral Sensory Inputs to the. . . 351

CCK increases the firing rate of GI vagal afferents through a receptor-mediated

mechanism in both adult and newborn rats (Raybould and Tache 1988); thus,
synthetic CCK is an ideal experimental tool with which to activate vagal sensory
afferents and identify their central postsynaptic targets in rats at different develop-
mental stages. CCK injected intraperitoneally evokes robust expression of Fos (the
protein product of the immediate-early gene c-fos) within the caudal NTS and AP in
both adult rats and newborn pups (Rinaman et al. 1994), with activated neurons
located in specific NTS subregions that receive gastric vagal sensory input. Perhaps
unsurprisingly, given the behavioral responsiveness of newborn rats to CCK, the
distribution of CCK-induced Fos expression within the caudal NTS is virtually
identical in 2-day-old and adult rats. Notably, surgical mid-collicular decerebration
studies in adult rats have revealed that brainstem circuits alone are sufficient for
feeding responses to CCK and other vagal satiety signals (Grill and Norgren 1978).
Thus, the available data indicate that CCK receptor-mediated activation of vagal
afferent inputs to the NTS and subsequent processes for recruitment of brainstem
circuits that underlie CCK-induced hypophagia already are functional in 2-day-old
rats. However, (discussed further, below), newborn rats are decidedly immature in
their hypothalamic Fos and pituitary hormone responses to CCK treatment,
reflecting delayed maturation of synaptic connections between visceral sensory
neurons within the caudal NTS and their hypothalamic targets (Rinaman and
Koehnle 2010).

13.3 Visceral Sensory Stimulation of the Neuroendocrine

Hypothalamus

13.3.1 Responses to Vagal Stimuli in Adult Animals

The ability of visceral sensory stimuli to modulate pituitary hormone release in adult
animals has been appreciated for many years. Early studies demonstrated posterior
pituitary release of AVP and OT in rats after electrical stimulation of the central end
of the severed vagus nerve (Chang et al. 1939; Thieblot et al. 1957), and later work
demonstrated that gastrointestinal or vagal nerve stimulation increases the firing
activity of magnocellular and parvocellular endocrine neurons within the PVN and
SON in anesthetized rats (Jin et al. 1993; Ueta et al. 1993). Some of these
experiments involved mechanical gastric distension or electrical stimulation of
vagal afferents, while others involved pharmacological manipulation of vagal
afferents via systemic administration of CCK or other gut peptides in adult rats.
For example, systemically administered ghrelin (a hormone that is endogenously
secreted from gastric tissue when the stomach is empty) suppresses the firing activity
of GI vagal afferents, whereas CCK increases vagal afferent firing. In adult rodents,
increased gastric release of ghrelin promotes increased release of growth hormone
(GH) from the anterior pituitary. Ghrelin binds to receptors expressed by vagal
sensory neurons, which transport the receptor to their peripheral axon terminals
within the stomach wall (see Fig. 13.1). Blockade of vagal afferent signaling to the
352 L. Rinaman

NTS abolishes ghrelin-induced GH secretion and also blocks the ability of ghrelin to
activate parvocellular GHRH endocrine neurons (Date et al. 2002).

13.3.2 Ascending Visceral Sensory Inputs to the Hypothalamus

in Adult Rats and Mice

Visceral sensory signals arising within the GI tract and other peripheral tissues are
relayed to the neuroendocrine hypothalamus via direct and indirect ascending neural
projection pathways from the NTS and AP in adult rats and mice (Rinaman 2007).
The indirect pathways include projections from AP and NTS neurons that are
relayed through the medullary RF (including the VLM) and pontine PBN before
ascending to the hypothalamus. Projections from AP neurons to the subjacent NTS
and also directly to the VLM and PBN provide a neural route for blood-borne signals
to alter central visceral sensory processing. However, while endocrine neurons
within the hypothalamic PVN, SON, mPOA, and ARC receive direct synaptic
input from catecholaminergic and peptidergic NTS and VLM neurons in adult
rats, they appear to receive little or no direct synaptic input from the pontine PBN
(Saper and Loewy 1980). The PBN instead appears to target primarily pre-auto-
nomic hypothalamic neurons and also provides robust input to the central nucleus of
the amygdala (CeA) and bed nucleus of the stria terminalis (BNST), two highly
interconnected limbic forebrain regions which also receive direct input from the NTS
and VLM (Bienkowski and Rinaman 2013).

13.3.3 Stimulus-Induced Activation of Visceral Sensory Inputs

to the Hypothalamus

As described in previous sections (Sects. 13.2.1 and 13.3.1), exogenous (synthetic)

CCK provides a useful experimental tool with which to manipulate ascending vagal
sensory pathways. In adult rats, systemic administration of CCK increases pituitary
release of ACTH, OT, GH, and thyroid stimulating hormone (TSH) through vagally
mediated, CCK-1 receptor-dependent activation of NTS neurons that provide direct
and relayed input to neuroendocrine cells of the hypothalamus (Fig. 13.1) (Rinaman
2007). Plasma OT levels also are increased by gastric distension, which synergizes
with exogenous CCK to further activate GI vagal afferents (Verbalis et al. 1986).
Systemic CCK increases plasma levels of AVP rather than OT in adult humans,
ferrets, and rhesus macaque monkeys. While the neural circuitry underlying species
differences in CCK-induced posterior pituitary hormone release remains unclear, the
difference is consistent with known species differences in endocrine responses to
stress. Rats and mice (which lack an emetic reflex) release OT along with ACTH
during stress responses, whereas primates and other animals with an emetic reflex
release AVP and ACTH (Verbalis et al. 1987).
Voluntary intake of a large meal by adult rats activates only a subset of the NTS
visceral sensory neurons that are activated after mechanical gastric distension,
13 Organization and Postnatal Development of Visceral Sensory Inputs to the. . . 353

supraphysiological doses of systemic CCK, and other “visceral stressors” that

increase ACTH and OT release (Rinaman 1999a). Nevertheless, the ability of
these visceral stimuli to alter pituitary hormone secretion indicates that visceral
sensory signals impact neuroendocrine hypothalamic cells in adult rats; conversely,
the lack of hypothalamic neural Fos activation and hormonal responses to vagal
stimulation in neonatal rats indicates that the central pathways conveying visceral
signals from the NTS to the hypothalamus are functionally immature at birth
(Rinaman and Koehnle 2010). The following section reviews what is currently
known regarding the chemical phenotypes, anatomical organization, and postnatal
maturation of neural projections from the caudal brainstem NTS and VLM to the
neuroendocrine hypothalamus.

13.3.4 Chemical Phenotypes of Hindbrain Neurons Conveying

Visceral Sensory Signals to the Neuroendocrine
Hypothalamus

13.3.4.1 Noradrenergic (NA) Signaling Pathways

In adult rats, visceral sensory inputs to the hypothalamus feature axonal projections
that arise from medullary NA and adrenergic neurons within the NTS (A2/C2 cell
groups) and VLM (A1/C1 cell groups) (Sawchenko and Swanson 1982; Rinaman
2010). For convenience, these NTS and VLM neurons are hereafter referred to
collectively as NA neurons, because they all are immunoreactive for the NA
synthetic enzyme, dopamine β hydroxylase (DBH). Medullary NA neurons appear
to be glutamatergic, based on their expression of mRNA for vesicular glutamatergic
transporter 2, which can also be visualized immunocytochemically in DBH-positive
terminals within the neuroendocrine hypothalamus (Zheng et al. 2015).
Subpopulations of NA neurons within the NTS and VLM that project to the
hypothalamus also express multiple peptides, including prolactin-releasing peptide
(PrRP) and neuropeptide Y (NPY) (Rinaman 2007).
Endocrine neurons within the mPOA, ARC, PVN, and SON express high levels
of adrenergic receptors (Kaminski and Watts 2012). The caudal NTS and VLM
provide most of the NA input to these hypothalamic nuclei, with a notably smaller
contribution from the locus coeruleus (A6 cell group) (Day et al. 1980). NA
terminals form both symmetric- and asymmetric-type synapses with the dendrites
and somata of OT, AVP, CRF, and TRH neurons. NA signaling promotes activation
of these neurons, thereby increasing hormone release (Nillni 2010). For example,
adrenergic signaling pathways from the caudal medulla to the hypothalamus provide
critical control over vagally mediated activation of the hypothalamic–pituitary–
adrenal (HPA) axis in adult rats, due to synaptic stimulation of CRF neurons
(Kamilaris et al. 1992). In addition to direct synaptic inputs onto the somata and
dendrites of identified endocrine neurons, NA axon terminals form close appositions
with axons in the external layer of the ME in adult rats, suggesting an additional site
for regulating hormone release into the portal system supplying the anterior pituitary.
354 L. Rinaman

In contrast to the stimulatory effects of adrenergic signaling on endocrine CRH,

TRH, AVP, and OT neurons, adrenergic receptor signaling directly and robustly
suppresses the activity of GnRH neurons within the mPOA in adult male and female
mice, in which NA signaling may serve a permissive role in the ability of other
inputs to regulate pulsatile GnRH and LH secretion (Szawka et al. 2013; Briski and
Shrestha 2016). During periods of caloric deficit, for example, NA inputs to the
neuroendocrine hypothalamus provide the primary source of glucoprivic stimulation
that inhibits pituitary LH secretion (Ohkura et al. 2000). Metabolic status is moni-
tored within the NTS, and NA signaling from the NTS suppresses the pre-ovulatory
LH surge (Ibrahim and Briski 2014). Indeed, energy deficiency is a key cause of
reduced frequency or cessation of ovulation in women.
NA projections from the NTS and VLM to the hypothalamus course primarily
through the ventral noradrenergic ascending bundle (VNAB) (Fig. 13.1). Electrical
stimulation of the VNAB elicits CRF secretion into the portal vasculature serving the
anterior pituitary through an adrenergic receptor-mediated mechanism (Plotsky
1987), thereby increasing plasma levels of ACTH. Similarly, direct stimulation of
the NTS or VLM increases PVN neuronal firing and pituitary hormone secretion,
and these effects depend on NA inputs to the PVN (Day et al. 1985). Extracellular
NA content increases in the PVN after systemic administration of CCK, which, as
reviewed above, activates vagal sensory inputs to the NTS and stimulates pituitary
release of several hormones. Systemic CCK activates NA neurons within the NTS
and VLM in rats, ferrets, and rhesus macaques (Rinaman 2007). In rats, NA neurons
activated after CCK administration include those that project directly to the PVN
(Rinaman et al. 1995), and immunotoxin-induced destruction of these NA neurons
attenuates the ability of systemic CCK to activate Fos expression within PVN
endocrine neurons (Rinaman 2010). A CCK-induced increase in NA content
measured within the PVN directly parallels increased plasma levels of ACTH and
OT in rats, consistent with the predominantly excitatory effects of systemic CCK and
synaptic effects of NA on hypothalamic CRF and OT neurons (Verbalis et al. 1991).

13.3.4.2 Glucagon-like Peptide 1 (GLP1) Projections

Separate populations of non-NA hypothalamic-projecting neurons within the caudal
NTS and VLM express immunoreactivity for a variety of neuropeptides (Rinaman
2007), including CCK and GLP1. A subset of CCK-expressing NTS neurons have
ascending axonal projections to the PVN in adult mice (Roman et al. 2017), but the
synaptic targets remain to be determined. Conversely, GLP1 circuits are more well
defined. GLP1 neurons within the caudal NTS and adjacent medullary RF are
glutamatergic (Zheng et al. 2015), similar to NA neurons (Stornetta et al. 2002),
and provide the sole source of GLP1-positive axon terminals within the hypothala-
mus. Similar to ascending NA projections from the NTS and VLM, the axons of
GLP1-positive neurons ascend to the hypothalamus within the VNAB (Fig. 13.1) to
target parvocellular CRH and TRH neurons in the PVN, magnocellular OT neurons
in the PVN and SON, and parvocellular GnRH neurons in the mPOA. GLP1 binding
sites and GLP1 receptor (GLP1-R) mRNA expression is evident in these and other
hypothalamic endocrine nuclei (Goke et al. 1995; Shughrue et al. 1996; Zueco et al.
13 Organization and Postnatal Development of Visceral Sensory Inputs to the. . . 355

1999), and GLP1 terminals synapse on kisspeptin-expressing neurons within the

ARC that regulate GnRH neurons (Heppner et al. 2017). The available data indicate
that GLP1 signaling increases the excitability of endocrine neurons that express
GLP1-R and also acts presynaptically to modulate glutamatergic and GABAergic
synaptic inputs to endocrine neurons (Farkas et al. 2016).
GLP1 neurons have been implicated in neuroendocrine responses to acute stress.
Early studies found that central (but not systemic) infusion of GLP1 or GLP1-R
agonists activates Fos in hypothalamic endocrine neurons and increases plasma
levels of OT, ACTH, and corticosterone in adult rodents (Larsen et al. 1997).
Subsequent studies found that a variety of interoceptive stressors [i.e., systemic
administration of lithium chloride (LiCl), lipopolysaccharide (LPS), or high doses
of CCK] and cognitive stressors (i.e., elevated platform exposure, restraint/immobi-
lization) activate Fos expression in the majority of GLP1 neurons in adult rats,
including GLP1 neurons that project to the hypothalamus (Rinaman 1999a). Further,
central antagonism of GLP1-R suppresses the ability of visceral/interoceptive or
cognitive stressors to activate the HPA axis (Rinaman 1999b; Kinzig et al. 2003).
Thus, both physiological and cognitive threats to homeostasis recruit ascending
GLP1 neural pathways to the neuroendocrine hypothalamus, similar to threat-
induced recruitment of ascending NA pathways (Rinaman 2007). Notably, mechan-
ical gastric distension, supraphysiological doses of CCK, and rapid intake of an
extremely large liquid meal can activate Fos expression in GLP1 neurons (Rinaman
1999a; Kreisler et al. 2014), consistent with evidence that GLP1 neurons respond to
unusually strong (i.e., homeostatically threatening) GI sensory signals (Maniscalco
et al. 2013). Thus, recruitment of hindbrain GLP1 neurons, including those that
innervate hypothalamic neuroendocrine cells, may depend on the intensity, modal-
ity, and/or temporal features of visceral stimulation, and may even differentiate
stressful from non-stressful stimuli.

13.3.5 Postnatal Development of NA and GLP1 Axonal Inputs

to the Neuroendocrine Hypothalamus

Immunocytochemical techniques have demonstrated that NA projections from the

NTS and VLM to the hypothalamus are present at birth. However, the density of
DBH-immunopositive NA terminals increases markedly in rats during the first
3 weeks of postnatal development (Fig. 13.1) (Khachaturian and Sladek
1980; Rinaman 2001), in concert with increasing levels of hypothalamic norepi-
nephrine (Coyle and Axelrod 1971). Additional evidence for delayed neurochemical
maturation of ascending NA projection pathways is offered by analysis of fiber
immunolabeling for PrRP, which is co-expressed by caudal subsets of NA neurons
within the NTS (A2 cell group) and VLM (A1 cell group) that project to the
hypothalamus and limbic forebrain (Chen et al. 1999; Morales et al. 2000). In rats,
PrRP-positive fibers are first visible within the anterior ventral BNST (which is the
main axonal target of hindbrain PrRP neurons) on postnatal day (P)3 (Yano et al.
356 L. Rinaman

2001). By P6, PrRP-positive fibers also are present within the PVN and other
hypothalamic nuclei, but fiber density is markedly sparser at P6 than in adulthood.
It is important to note that an absence or scarcity of neurotransmitter-positive
fibers should not be interpreted as evidence that the axons of NTS and VLM neurons
are not present within the hypothalamus. In one study in which retrograde neural
tracer was microinjected into the PVN in rat pups on the day of birth (P0), the
number of PVN-projecting neurons within the NTS and VLM quantified 24 h later
(on P1) was similar to the number of retrogradely labeled, PVN-projecting neurons
in adult rats (Rinaman 2001). However, a smaller proportion of PVN-projecting
NTS and VLM neurons were DBH-positive in newborn rats compared with adults,
consistent with reduced DBH immunostaining in the neonatal PVN. Conversely,
fiber immunolabeling for phenylethanolamine N-methyltransferase (PNMT, the
enzyme that converts norepinephrine to epinephrine) within the PVN was most
dense at P1, gradually decreasing to adult-like levels by P21. The dynamic temporal
shift in the density of immunolabeling for these catecholaminergic synthetic
enzymes reflects a gradual maturation of noradrenergic and adrenergic signaling to
the hypothalamus in rats during the first 3 weeks of life (Rinaman 2001). It remains
unclear whether the shift in immunolabeling patterns represents parallel increases
and reductions in axonal terminal density arising from PNMT-negative versus
PNMT-positive NA neurons, or whether it represents altered enzyme content within
individual neurons whose axons innervate the neuroendocrine hypothalamus.

13.3.5.1 Postnatal Development of Visceral Sensory Pathways Assessed

in Transgenic Rodent Models
Recent advances in transgenic technology have provided additional tools with which
to assess the development and distribution of axonal projections arising from pheno-
typically identified neurons. In one transgenic mouse model, neurons that express
preproglucagon (PPG, which is cleaved to GLP1, GLP2, glucagon, glicentin,
oxyntomodulin, and other peptides) were engineered to also express yellow fluores-
cent protein (YFP). In these “PPG-YFP” mice, YFP expression completely fills the
somatic, dendritic, and axonal compartments of PPG-expressing neurons, providing a
convenient and chemically specific marker that does not depend on immunolabeling
of GLP1 itself (Llewellyn-Smith et al. 2011). In previously unpublished work, we
observed abundant YFP-positive cell bodies, dendrites, and axons within the caudal
NTS in newborn PPG-YFP mice (i.e., P0), and dual immunolabeling confirmed that
YFP-expressing neurons are immunopositive for GLP1 on P6 (Fig. 13.3). However,
YFP-positive fibers are nearly absent within the neuroendocrine hypothalamus in
newborn mice, with YFP immunolabeling increasing gradually within the PVN over
the first 3 weeks of postnatal development (Fig. 13.4). GLP1 neurons express leptin
receptors in mice, in which leptin signaling stimulates the postnatal development of
axonal outgrowth from the ARC to the PVN (Bouret and Simerly 2007). An ongoing
study is investigating whether the postnatal surge in leptin is also required for normal
development of GLP1 signaling pathways from the NTS to the PVN (J. Biddinger
and R. Simerly, Vanderbilt University Medical Center; personal communication).
GLP1 neurons in rats do not express leptin receptors; instead, leptin receptors in rats
13 Organization and Postnatal Development of Visceral Sensory Inputs to the. . . 357

Fig. 13.3 Immunolabeling for yellow fluorescent protein (YFP) expressed under control of the
preproglucagon (PPG) promoter in transgenic PPG-YFP mice. Top panel: YFP immunoperoxidase
labeling of PPG neurons and processes within the caudal NTS in a newborn mouse (P0). Bottom
panel: green YFP immunofluorescence is co-localized with red GLP1 immunofluorescence to
generate yellow double-labeled neurons within the caudal NTS in a 6-day old PPG-YFP mouse
(P6). C: central canal. Scale bar: 50 μm

are expressed by NA neurons within the caudal NTS, which do not express leptin
receptors in mice (Huo et al. 2008). The potential significance of these species
differences in expression of leptin receptors in the caudal NTS remains unclear, but
it presumably is linked to differential effects of leptin on hindbrain-hypothalamic
circuits in both developing and mature animals.

13.3.5.2 Functional Immaturity of Visceral Sensory Inputs

to the Hypothalamus in Neonatal Rats
The apparent immaturity of NA and GLP1 projections from the caudal medulla to
the hypothalamus in neonatal rats predicts that endocrine neurons are less sensitive
to visceral sensory signals early in life. Indeed, CCK does not activate Fos expres-
sion in the PVN (Fig. 13.5) and does not increase plasma levels of OT in newborn rat
pups (Rinaman et al. 1994), whereas plasma OT levels are increased in response to
stimulation of gastric vagal afferents in 10-day-old rat pups (Nelson et al. 1998).
358 L. Rinaman

Fig. 13.4 YFP-positive

axons arising from nucleus of
the solitary tract (NTS)
neurons gradually appear
within the paraventricular
nucleus (PVN) postnatally.
Immunoperoxidase labeling
of YFP-positive fibers within
the paraventricular nucleus of
the hypothalamus (PVN) in
developing PPG-YFP
transgenic mice. Top panel:
very few YFP-positive fibers
are present within the PVN in
newborn mice (P0). The
density of YFP-positive fibers
increases within the PVN at
later postnatal ages (P6, P13),
with adult-like labeling
evident by the time of
weaning (P20, bottom panel).
Scale bar: 50 μm
13 Organization and Postnatal Development of Visceral Sensory Inputs to the. . . 359

Fig. 13.5 Responses to

visceral signals mature
differentially.
Immunocytochemical
labeling of Fos within the
hypothalamic paraventricular
nucleus (PVN) in neonatal
rats. Top panel: i.p. injection
of cholecystokinin (CCK)
does not increase Fos labeling
within the PVN in 2-day-old
rat pups, although it does so in
adult rats. Figure adapted
from Rinaman et al. (1994).
Bottom panel: s.c. injection of
hypertonic saline
(HS) robustly activates Fos
expression in 2-day-old rat
pups, similar to activation in
adult rats. Scale bar: 50 μm.
Figure adapted from Rinaman
et al. (1997)

Another study investigated the postnatal maturation of central neural Fos responses
to i.p. injection of LiCl, a malaise-inducing agent (Koehnle and Rinaman 2007).
Compared to Fos activation in neonatal control rats injected with physiological
saline, LiCl did not increase Fos in the PVN or other forebrain regions at P0, but
did so at P7 and later, with maximal PVN Fos responses to LiCl achieved by P14.
Comparable results have been obtained examining central Fos responses to an acute
LPS challenge in developing rats (Oladehin and Blatteis 1995). These findings
provide additional evidence that vagal sensory signaling to the neuroendocrine
hypothalamus undergoes significant functional maturation in rats during the first
2–3 weeks of postnatal life (Rinaman and Koehnle 2010).
Given the apparent inability of systemic CCK, LiCl, or LPS to activate hypotha-
lamic neurons in neonatal rats, one might hypothesize that hypothalamic endocrine
neurons are refractory to all sources of excitatory input early in development. This
360 L. Rinaman

hypothesis is challenged by evidence that acute osmotic dehydration (induced by

subcutaneous injection of hypertonic saline) robustly activates Fos expression in
osmosensitive regions of the hypothalamus (Fig. 13.5), including magnocellular
PVN and SON neurons, in both neonatal and adult rats (Rinaman et al. 1997). The
important difference in the ability of interoceptive challenge to activate neuroendo-
crine cells of the hypothalamus seems to be whether or not hypothalamic responses
require ascending visceral sensory inputs from the caudal brainstem. In adult rats,
these ascending inputs are unnecessary for hypothalamic responses to osmotic
dehydration, but they are required for responses to gastric distension, CCK, LPS,
and LiCl. Thus, the lack of hypothalamic Fos activation in neonatal rats treated with
CCK, LiCl, and other visceral stimuli is most likely due to functional immaturity of
ascending projection pathways that relay visceral sensory signals from the NTS and
VLM.

13.4 Postnatal Development and Potential Plasticity of Visceral

Sensory Signaling to the Neuroendocrine Hypothalamus

The findings summarized above (Sect. 13.3.5) support the view that ascending
neural pathways conveying visceral sensory signals from hindbrain to hypothalamus
are functionally immature in neonatal rats. Thus, these circuits may exhibit plasticity
as they assemble during a developmentally defined sensitive period. Highly
specialized mechanisms are crucial for the initial establishment of postsynaptic
specializations during synaptogenesis, and for activity-related changes in synaptic
strength that underlie experience-dependent plasticity. By analogy with other CNS
systems, evoked neural activity within ascending visceral sensory pathways may
shape ongoing synapse formation and stabilization of visceral sensory inputs to the
neuroendocrine hypothalamus during the first few weeks of postnatal life.

13.4.1 The Stress-Hyporesponsive Period (SHRP)

In rats, the first 2 weeks of postnatal life correspond to the so-called SHRP,
characterized by reduced HPA axis responsiveness to stressful stimuli (Walker
et al. 1991). Most of the peripheral and central components of stress-responsive
circuits are capable of being activated in neonatal rats, given sufficient experimental
or natural conditions (Levine 2001). However, HPA axis responsiveness in neonatal
rats is apparently suppressed by maternal factors that generate interoceptive signals
in their developing offspring, including thermal, chemical, and mechanoreceptive
components (Rinaman and Koehnle 2010).
13 Organization and Postnatal Development of Visceral Sensory Inputs to the. . . 361

13.4.2 Early Life Experience Alters Visceral Sensory Inputs

to the Neuroendocrine Hypothalamus

Plasma levels of GI-derived CCK and pituitary-derived GH fall when rat pups are
separated from their dam and increase during and after maternal reunion and feeding
(Kuhn et al. 1979; Weller et al. 1992). Experimental paradigms that reduce/fragment
the early life maternal care received by rat pups exert a significant impact on adult
measures of stress-evoked norepinephrine release within the PVN, modulating
stimulus-induced activation of CRH endocrine neurons and anterior pituitary release
of ACTH (Liu et al. 2000; Walker et al. 2017). Further, CCK-1 receptors are
especially abundant in the upper GI tract in newborn rats (Robinson et al. 1987).
Based on the results of experiments in which early life maternal care was
manipulated together with manipulation of CCK-1 receptor signaling, we have
proposed that the impact of maternal care on central visceral circuits is based, at
least in part, on the activity of GI vagal sensory input to the NTS (Weber et al. 2009;
Rinaman and Koehnle 2010), which in turn regulates the neuroendocrine hypothal-
amus through ascending NA, GLP1, and potentially other neurochemical signaling
pathways.

13.5 Perspectives

Pituitary hormone secretion is modulated by interoceptive feedback signals that are

relayed primarily through the caudal brainstem NTS. NA and GLP1-expressing NTS
neurons project to the hypothalamus, where adrenergic and GLP1R-mediated sig-
naling regulates gene expression and secretory activity in endocrine neurons. Given
the fundamental importance of the caudal NTS as a sensory interface between
internal body organs and the brain, it is striking that we still know relatively little
regarding the developmental maturation of neural signaling pathways from the NTS
to the neuroendocrine hypothalamus or the means by which these circuits might be
programmed by early life experience and other epigenetic influences. In humans,
visceral hypersensitivity and neuroendocrine dysfunctions are recognized as com-
mon symptoms of stress-related disorders such as anxiety and depression, which
feature inappropriate physiological responses to stressful, emotionally triggering
stimuli. Aberrant neuroendocrine and other physiological responses may arise, at
least in part, as a product of postnatal life events that shape the developmental
assembly of brainstem-hypothalamic visceral sensory pathways.

Key References

Bouret and Simerly (2007)—This review article summarizes evidence that the
intrinsic hypothalamic circuits that integrate interoceptive signals (here, leptin
signaling from visceral adipose tissue) undergo a significant amount of postnatal
362 L. Rinaman

development in rodents. The authors present evidence that a natural postnatal

surge in leptin promotes neural outgrowth and synapse formation within the
neuroendocrine hypothalamus.
Coyle and Axelrod (1971)—One of the first publications providing evidence that
noradrenergic inputs to the hypothalamus (which originate primarily from the
hindbrain NTS and VLM) develop gradually during the first 3 weeks postnatal
in rats.
Day et al. (1985)—This report provides compelling evidence in adult rats that
noradrenergic projections from the hindbrain NTS and VLM to the neuroendo-
crine hypothalamus facilitate the activity of neuroendocrine neurons that control
anterior pituitary hormone release.
Kamilaris et al. (1992)—A comprehensive set of experiments demonstrating that
systemic CCK activates vagal sensory inputs to the NTS in rats, promoting
activation of neuroendocrine CRH neurons at the apex of the HPA axis.
Rinaman and Koehnle (2010)—This book chapter summarizes what was known at
the time regarding the anatomical and functional development of ascending
visceral sensory pathways from the hindbrain NTS to the hypothalamus and
other forebrain regions in rats.
Sawchenko and Swanson (1982)—A classic paper and required reading for anyone
interested in understanding the anatomical organization of ascending hindbrain-
hypothalamic circuits. The authors focus on inputs to the neuroendocrine
hypothalamus.
Walker et al. (1991)—A comprehensive analysis of what was previously called the
“stress nonresponsive period” of postnatal development. The authors conclude
that the HPA axis is responsive to certain types of stimuli (including those that do
not depend on ascending visceral sensory pathways to the hypothalamus), and
effectively argue that this early developmental period should instead be called the
“stress hyporesponsive period” (SHRP).

References
Bienkowski MS, Rinaman L (2013) Common and distinct neural inputs to the medial central
nucleus of the amygdala and anterior ventrolateral bed nucleus of stria terminalis in rats.
Brain Struct Funct 218:187–208
Bouret SG, Simerly RB (2007) Development of leptin-sensitive circuits. J Neuroendocrinol
19:575–582
Bouret SG, Draper SJ, Simerly RB (2004) Formation of projection pathways from the arcuate
nucleus of the hypothalamus to hypothalamic regions implicated in the neural control of feeding
behavior in mice. J Neurosci 24:2797–2805
Briski KP, Shrestha PK (2016) Hindbrain estrogen receptor-beta antagonism normalizes reproduc-
tive and counter-regulatory hormone secretion in hypoglycemic steroid-primed ovariectomized
female rats. Neuroscience 331:62–71
Chang HC, Chia KF, Huang JJ, Lim RKS (1939) Vagus-post-pituitary reflex; antidiuretic effect.
Clin J Physiol 14:161–172
Chen C-T, Dun SL, Dun NJ, Chang J-K (1999) Prolactin-releasing peptide-immunoreactivity in A1
and A2 noradrenergic neurons of the rat medulla. Brain Res 822:276–279
13 Organization and Postnatal Development of Visceral Sensory Inputs to the. . . 363

Coyle JT, Axelrod J (1971) Development of the uptake and storage of L-[3H]norepinephrine in the
rat brain. J Neurochem 18:2061–2075
Date Y, Murakami N, Toshinai K, Matsukura S, Niijima A, Matsuo H, Kangawa K, Nakazato M
(2002) The role of the gastric afferent vagal nerve in ghrelin-induced feeding and growth
hormone secretion in rats. Gastroenterology 123:1120–1128
Day TA, Blessing W, Willoughby JO (1980) Noradrenergic and dopaminergic projections to the
medial preoptic area of the rat. A combined horseradish peroxidase/catecholamine fluorescence
study. Brain Res 193:543–548
Day T, Ferguson AV, Renaud L (1985) Noradrenergic afferents facilitate the activity of
tuberoinfundibular neurons of the hypothalamic paraventricular nucleus. Neuroendocrinology
41:17
Farkas I, Vastagh C, Farkas E, Balint F, Skrapits K, Hrabovszky E, Fekete C, Liposits Z (2016)
Glucagon-like peptide-1 excites firing and increases GABAergic miniature postsynaptic
currents (mPSCs) in gonadotropin-releasing hormone (GnRH) neurons of the male mice via
activation of nitric oxide (NO) and suppression of endocannabinoid signaling pathways. Front
Cell Neurosci 10:214
Godement P, Vanselow J, Thanos S, Bonhoeffer F (1987) A study in developing visual systems
with a new method of staining neurones and their processes in fixed tissue. Development
101:697–713
Goke R, Larsen PJ, Mikkelsen JD, Sheikh SP (1995) Distribution of GLP-1 binding sites in the rat
brain: evidence that exendin-4 is a ligand of brain GLP-1 binding sites. Eur J Neurosci
7:2294–2300
Grill HJ, Norgren R (1978) Chronically decerebrate rats demonstrate satiation but not bait shyness.
Science 201:267–269
Heppner KM, Baquero AF, Bennett CM, Lindsley SR, Kirigiti MA, Bennett B, Bosch MA, Mercer
AJ, Ronnekleiv OK, True C, Grove KL, Smith MS (2017) GLP-1R signaling directly activates
arcuate nucleus kisspeptin action in brain slices but does not rescue luteinizing hormone
inhibition in ovariectomized mice during negative energy balance. eNeuro 4(1):
ENEURO.0198-16
Huo L, Gamber KM, Grill HJ, Bjorbaek C (2008) Divergent leptin signaling in proglucagon
neurons of the nucleus of the solitary tract in mice and rats. Endocrinology 149:492–497
Ibrahim BA, Briski KP (2014) Role of dorsal vagal complex A2 noradrenergic neurons in hindbrain
glucoprivic inhibition of the luteinizing hormone surge in the steroid-primed ovariectomized
female rat: effects of 5-thioglucose on A2 functional biomarker and AMPK activity. Neurosci-
ence 269:199–214
Jin Y, Ueta Y, Kannan H, Yamashita H (1993) Synaptic inputs from the stomach to
tuberoinfundibular neurons in the paraventricular nucleus of the hypothalamus in rats. Brain
Res 617:151–154
Kamilaris TC, Johnson EO, Calogero AE, Kalogeras KT, Bernardini R, Chrousos GP, Gold PW
(1992) Cholecystokinin-octapeptide stimulates hypothalamic-pituitary-adrenal function in rats:
role of corticotropin-releasing hormone. Endocrinology 130:1764–1774
Kaminski KL, Watts AG (2012) Intact catecholamine inputs to the forebrain are required for
appropriate regulation of corticotrophin-releasing hormone and vasopressin gene expression
by corticosterone in the rat paraventricular nucleus. J Neuroendocrinol 24:1517–1526
Khachaturian H, Sladek JR (1980) Simultaneous monoaminehistofluorescence and neuropeptide
immunocytochemistry: III. Ontogeny of catecholamine varicosities and neurophysin neurons in
the rat supraoptic and paraventricular nuclei. Peptides 1:77–95
Kinzig KP, D’Alessio DA, Herman JP, Sakai RR, Vahl TP, Figueiredo HF, Murphy EK, Seeley RJ
(2003) CNS glucagon-like peptide-1 receptors mediate endocrine and anxiety responses to
interoceptive and psychogenic stressors. J Neurosci 23:6163–6170
Koehnle T, Rinaman L (2007) Progressive postnatal increases in Fos immunoreactivity in the
forebrain and brainstem of rats after viscerosensory stimulation with lithium chloride. Am J
Physiol Regul Integr Comp Physiol 292:R1212–R1223
364 L. Rinaman

Kreisler AD, Davis EA, Rinaman L (2014) Differential activation of chemically identified neurons
in the caudal nucleus of the solitary tract in non-entrained rats after intake of
satiating vs. non-satiating meals. Physiol Behav 136:47–54
Kuhn CM, Evoniuk G, Schanberg SM (1979) Loss of tissue sensitivity to growth hormone during
maternal deprivation in rats. Life Sci 25:2090–2097
Larsen PJ, Tang-Christensen M, Jessop DS (1997) Central administration of glucagon-like peptide-
1 activates hypothalamic neuroendocrine neurons in the rat. Endocrinology 138:4445–4455
Levine S (2001) Primary social relationships influence the development of the hypothalamic-
pituitary-adrenal axis in the rat. Physiol Behav 73:255–260
Liu D, Caldji C, Sharma S, Plotsky PM, Meaney MJ (2000) Influence of neonatal rearing conditions
on stress-induced adrenocorticotropin responses and norepinephrine release in the hypothalamic
paraventricular nucleus. J Neuroendocrinol 12:5–12
Llewellyn-Smith IJ, Reimann F, Gribble FM, Trapp S (2011) Preproglucagon neurons project
widely to autonomic control areas in the mouse brain. Neuroscience 180:111–121
Maniscalco JW, Kreisler AD, Rinaman L (2013) Satiation and stress-induced hypophagia: examin-
ing the role of hindbrain neurons expressing prolactin-releasing peptide or glucagon-like peptide
1. Front Neurosci 6:199
Menétrey D, dePommery J (1991) Origins of spinal ascending pathways that reach central areas
involved in visceroception and visceronociception in the rat. Eur J Neurosci 3:249–259
Morales T, Hinuma S, Sawchenko PE (2000) Prolactin-releasing peptide is expressed in afferents to
the endocrine hypothalamus, but not in neurosecretory neurones. J Neuroendocrinol
12:131–140
Nelson EE, Alberts JR, Tian Y, Verbalis JG (1998) Oxytocin is elevated in plasma of 10-day-old
rats following gastric distension. Dev Brain Res 111:301–303
Nillni EA (2010) Regulation of the hypothalamic thyrotropin releasing hormone (TRH) neuron by
neuronal and peripheral inputs. Front Neuroendocrinol 31:134–156
Ohkura S, Tanaka T, Nagatani S, Bucholtz DC, Tsukamura H, Maeda K, Foster DL (2000) Central,
but not peripheral, glucose-sensing mechanisms mediate glucoprivic suppression of pulsatile
luteinizing hormone secretion in the sheep. Endocrinology 141:4472–4480
Oladehin A, Blatteis CM (1995) Lipopolysaccharide-induced fos expression in hypothalamic nuclei
of neonatal rats. Neuroimmunomodulation 2:282–289
Palkovits M (1999) Interconnections between the neuroendocrine hypothalamus and the central
autonomic system. Front Neuroendocrinol 20:270–295
Plotsky P (1987) Facilitation of immunoreactive corticotropin-releasing factor secretion into the
hypophysial-portal circulation after activation of catecholaminergic pathways or central norepi-
nephrine injection. Endocrinology 121:924–930
Raybould HE, Tache Y (1988) Cholecystokinin inhibits gastric motility and emptying via a
capsaicin-sensitive vagal pathway in rats. Am J Phys 255:G242–G246
Rinaman L (1999a) Interoceptive stress activates glucagon-like peptide-1 neurons that project to the
hypothalamus. Am J Phys 277:R582–R590
Rinaman L (1999b) A functional role for central glucagon-like peptide-1 receptors in lithium
chloride-induced anorexia. Am J Phys 277:R1537–R1540
Rinaman L (2001) Postnatal development of catecholamine inputs to the paraventricular nucleus of
the hypothalamus in rats. J Comp Neurol 438:411–422
Rinaman L (2007) Visceral sensory inputs to the endocrine hypothalamus. Front Neuroendocrinol
28:50–60
Rinaman L (2010) Hindbrain noradrenergic A2 neurons: diverse roles in autonomic, endocrine,
cognitive, and behavioral functions. Am J Physiol Regul Integr Comp Physiol 00556:02010
Rinaman L, Koehnle TJ (2010) Development of central visceral circuits. In: Blumberg MS,
Freeman JH, Robinson SR (eds) Oxford handbook of developmental behavioral neuroscience.
Oxford University Press, New York, pp 298–321
Rinaman L, Levitt P (1993) Establishment of vagal sensorimotor circuits during fetal development
in rats. J Neurobiol 24:641–659
13 Organization and Postnatal Development of Visceral Sensory Inputs to the. . . 365

Rinaman L, Hoffman GE, Stricker EM, Verbalis JG (1994) Exogenous cholecystokinin activates
cFos expression in medullary but not hypothalamic neurons in neonatal rats. Dev Brain Res
77:140–145
Rinaman L, Hoffman GE, Dohanics J, Le WW, Stricker EM, Verbalis JG (1995) Cholecystokinin
activates catecholaminergic neurons in the caudal medulla that innervate the paraventricular
nucleus of the hypothalamus in rats. J Comp Neurol 360:246–256
Rinaman L, Stricker EM, Hoffman GE, Verbalis JG (1997) Central c-fos expression in neonatal and
adult rats after subcutaneous injection of hypertonic saline. Neuroscience 79:1165–1175
Robinson PH, Moran TH, Goldrich M, McHugh PR (1987) Development of cholecystokinin
binding sites in the rat upper gastrointestinal tract. Am J Phys 252:G529–G534
Robinson PH, Moran TH, McHugh PR (1988) Cholecystokinin inhibits independent ingestion in
neonatal rats. Am J Phys 255:R14–R20
Roman CW, Sloat SR, Palmiter RD (2017) A tale of two circuits: CCK NTS neuron stimulation
controls appetite and induces opposing motivational states by projections to distinct brain
regions. Neuroscience 358:316–324
Saper CB, Loewy AD (1980) Efferent connections of the parabrachial nucleus in the rat. Brain Res
197:291–317
Sawchenko PE, Swanson LW (1982) The organization of noradrenergic pathways from the
brainstem to the paraventricular and supraoptic nuclei in the rat. Brain Res Rev 4:275–325
Shapiro RE, Miselis RR (1985) The central neural connections of the area postrema of the rat.
J Comp Neurol 234:344–364
Shughrue PJ, Lane MV, Merchenthaler I (1996) Glucagon-like peptide-1 receptor (GLP1-R)
mRNA in the rat hypothalamus. Endocrinology 137:5159–5162
Stornetta RL, Sevigny CP, Guyenet PG (2002) Vesicular glutamate transporter DNPI/VGLUT2
mRNA is present in C1 and several other groups of brainstem catecholaminergic neurons.
J Comp Neurol 444:191–206
Szawka RE, Poletini MO, Leite CM, Bernuci MP, Kalil B, Mendonca LB, Carolino RO, Helena
CV, Bertram R, Franci CR, Anselmo-Franci JA (2013) Release of norepinephrine in the
preoptic area activates anteroventral periventricular nucleus neurons and stimulates the surge
of luteinizing hormone. Endocrinology 154:363–374
Thieblot L, Duchene-Marullaz P, Berthelay J (1957) Nouvelle preuve de la libération de principes
posthypophysaires par l’excitation centripète du vague. Ann Endocrinol 18:651–653
Ueta Y, Kannan H, Higuchi T, Negoro H, Yamashita H (1993) CCK-8 excites oxytocin-secreting
neurons in the paraventricular nucleus in rats - possible involvement of noradrenergic pathway.
Brain Res Bull 32:453–459
Verbalis JG, McCann MJ, McHale CM, Stricker EM (1986) Oxytocin secretion in response to
cholecystokinin and food: differentiation of nausea from satiety. Science 232:1417–1419
Verbalis JG, Richardson DW, Stricker EM (1987) Vasopressin release in response to nausea-
producing agents and cholecystokinin in monkeys. Am J Phys 252:R749–R753
Verbalis JG, Stricker EM, Robinson AG, Hoffman GE (1991) Cholecystokinin activates cFos
expression in hypothalamic oxytocin and corticotropin-releasing hormone neurons.
J Neuroendocrinol 3:205–213
Walker CD, Scribner VA, Cascio CS, Dallman MF (1991) The pituitary-adrenocortical system of
neonatal rats is responsive to stress throughout development in a time dependent and stress-
specific fashion. Endocrinology 128:1385–1396
Walker CD, Bath KG, Joels M, Korosi A, Larauche M, Lucassen PJ, Morris MJ, Raineki C, Roth
TL, Sullivan RM, Tache Y, Baram TZ (2017) Chronic early life stress induced by limited
bedding and nesting (LBN) material in rodents: critical considerations of methodology,
outcomes and translational potential. Stress 20:421–448
Weber BC, Manfredo HN, Rinaman L (2009) A potential gastrointestinal link between enhanced
postnatal maternal care and reduced anxiety-like behavior in adolescent rats. Behav Neurosci
123:1178–1184
366 L. Rinaman

Weller A, Corp ES, Tyrka A, Ritter RC, Brenner L, Gibbs J, Smith GP (1992) Trypsin inhibitor and
maternal Reunion increase plasma cholecystokinin in neonatal rats. Peptides 13:939–941
Yano T, Iijima N, Kataoka Y, Hinuma S, Tanaka M, Ibata Y (2001) Developmental expression of
prolactin releasing peptide in the rat brain: localization of messenger ribonucleic acid and
immunoreactive neurons. Dev Brain Res 128:101–111
Zheng H, Stornetta RL, Agassandian K, Rinaman L (2015) Glutamatergic phenotype of glucagon-
like peptide 1 neurons in the caudal nucleus of the solitary tract in rats. Brain Struct Funct
220:3011–3022
Zueco JA, Esquifino AI, Chowen JA, Alvarez E, Castrillon PO, Blazquez E (1999) Coexpression of
glucagon-like peptide-1 (GLP-1) receptor, vasopressin, and oxytocin mRNAs in neurons of the
rat hypothalamic supraoptic and paraventricular nuclei: effect of GLP-1(7-36)amide on vaso-
pressin and oxytocin release. J Neurochem 72:10–16
Astrocytes and Development
of Neuroendocrine Circuits 14
Lydia L. DonCarlos and Julie A. Chowen

Abstract
Astrocytes, one of the most abundant cell types in the hypothalamus, perform a
myriad of functions throughout all stages of development. Radial glia, astroglia-
like progenitor cells, are fundamental for neurogenesis, neuronal migration, axon
extension, and synaptic formation throughout the brain. These progenitor cells
also give rise to mature astrocytes, which are essential for the proper functioning
of the brain at all ages. In the hypothalamus, astrocytes play an important role in
all neuroendocrine systems. Together with tanycytes, specialized glial cells
surrounding the third ventricle of the hypothalamus, astrocytes maintain some
of their developmental potential into adulthood, including their progenitor com-
petence and ability to modulate synaptic plasticity. In this chapter, we will
address what is known regarding the role of astrocytes in hypothalamic develop-
ment. We will briefly cover some of the functions that astrocytes perform in
specific neuroendocrine systems, as well as how the maintenance of certain
developmental capacities during later stages of life is fundamental for systemic
adaptation and homeostatic control. Although the roles of astrocytes in neuroen-
docrine control have received increasing attention in recent years, there is still
much to be learned regarding these fascinating cells. Identification of functional
astrocytic populations with specific markers will allow these cells to be targeted

L. L. DonCarlos
Department of Cell and Molecular Physiology, Stritch School of Medicine, Loyola University
Chicago, Maywood, IL, USA
J. A. Chowen (*)
Department of Pediatrics and Pediatric Endocrinology, Hospital Infantil Universitario Niño Jesús,
Princesa Research Institute La Princesa, Madrid, Spain
Centro de Investigación Biomédica en Red de Fisiopatología de la Obesidad y Nutrición
(CIBEROBN), Instituto de Salud Carlos III, IMDEA Food Institute, CEIUAM+CSIC, Madrid,
Spain
e-mail: [email protected]

# Springer Nature Switzerland AG 2020 367

S. Wray, S. Blackshaw (eds.), Developmental Neuroendocrinology, Masterclass in
Neuroendocrinology 9, https://doi.org/10.1007/978-3-030-40002-6_14
368 L. L. DonCarlos and J. A. Chowen

and more easily manipulated, so that we can begin to learn details about how
astroglia participate in the development of specific hypothalamic circuits.

Keywords
Arcuate nucleus · Astrocytes · Glia · Gliogenesis · Neuroendocrine circuits ·
Paraventricular nucleus · Tanycytes · Suprachiasmatic nucleus · Supraoptic
nucleus

14.1 Introduction

Astrocytes are among the most abundant glial cells in the central nervous system,
with the proportion of astrocytes to other glial cells or to neurons varying depending
on the brain region (Ero et al. 2018). Although astrocytes were early on recognized
as having processes contacting multiple different neurons and communicating with
each other as well (Fig. 14.1), they were relegated to a supporting role for some
years. Research beginning in the 1970s led to increasing recognition of the promi-
nent role of astrocytes in the function of the mature and damaged nervous system,
and, for the purpose of this review, as being key to the development and honing of
neural circuits (Allen and Lyons 2018) including the hypothalamus and neuroendo-
crine systems.
Because they do not generate action potentials, many aspects of astrocytic action
have been difficult to study, and research into their function has often been restricted
by their physiologically inaccessible nature. Overall, astrocytes typically develop
and mature later than neurons, rendering it difficult to study them independent of
developing neuronal circuits. Therefore, our understanding and characterization of
the development and diversity of astrocytes, and how that diversity is determined, is
still in early stages relative to what we know about neurons. However, among the
most informative studies on astrocyte function and plasticity are those that have
examined astrocytes in the hypothalamus, both in development and in response to
manipulations of, for example, gonadal steroid hormones or metabolism. Moreover,
although underappreciated by neuroscientists in general, these studies of astrocytes
in the developing hypothalamus have been both particularly illuminating and yet
particularly enigmatic, as discussed in more detail below.

14.2 Astrocytes in Development

Glial cells include astrocytes, microglia, oligodendrocytes, tanycytes, and NG2 cells.
They are distinguished by expression of specific proteins and based on their mor-
phology. This group of cells is involved in many of the aspects of brain development
that follow early patterning of the nervous system, including neurogenesis, neuronal
migration, and synaptogenesis (Allen and Lyons 2018; Bosworth and Allen 2017).
These cells are intimately involved in neuroprotection (Belanger and Magistretti
14 Astrocytes and Development of Neuroendocrine Circuits 369

Fig. 14.1 Drawing of astrocytes by Santiago Ramón y Cajal. Courtesy of the Cajal Institute, Cajal
Legacy, Spanish National Research Council (CSIC), Madrid, reprinted with permission. This
detailed drawing by Santiago Ramón y Cajal demonstrates how his keen eye observed the intricate
relationship between astrocytes and neurons. Astrocytes (A) can ensheath neuronal cell bodies (B)
with their projections, as well as contacting neuronal processes and cell bodies (a, b) as well as
blood vessels (c)

2009), neurotransmission (Chung et al. 2015; Perea et al. 2009), and adaptation of
the CNS to physiological changes by participating in synaptic reorganization and
modifications in synaptic function (Chowen et al. 2016; Hatton 1997; Sims et al.
2015). They also sustain neurons and neuronal circuits by supplying nutrients and
maintaining extracellular homeostasis (Leloup et al. 2015; Magistretti and Allaman
2018) with intimate contact and communication between astrocytes and neurons
throughout development.
370 L. L. DonCarlos and J. A. Chowen

Neurons
Neuroblasts

Pluripotent Cell

Astrocytes
Radial glia

Fig. 14.2 Schematic representation of cell lines for development of astrocytes and neurons.
Astrocytes are derived from radial glia

Development of the hypothalamus in rodents continues during the early postnatal

period, when many projections and synaptic connections are formed (Altman and
Bayer 1986; Bayer and Altman 1987; Bouret and Simerly 2007). During early brain
development astroglial-like progenitor cells, called radial glia, are generated from
neural stem cells and span the neuroepithelium. The radial glia (Fig. 14.2) are
precursors for both neurons and astrocytes (Malatesta et al. 2000; Noctor et al.
2001), but not all radial glia have the same potential for neuronal or glial production,
and part of this differential potential is likely dependent on their local environment
(Malatesta et al. 2003). In rats, vimentin-reactive radial glia are found widely
distributed throughout the hypothalamus and preoptic area at embryonic day
17, with this vimentin staining beginning to decrease at postnatal day (PND) 0 and
dramatically decreasing by PND3. At PND6, vimentin staining is reported to be
observed only in ependymal cells surrounding the third ventricle and dispersed in the
optic chiasm (Munekawa et al. 2000). Glial fibrillary acidic protein (GFAP) has been
widely used as a marker of astrocytes, although during development radial glial cells
can also express this structural protein, as can tanycytes. Moreover, not all mature
astrocytes express GFAP. Nevertheless, most of the studies to date have used
markers for this protein to follow astroglial development and plasticity in the
hypothalamus. Between gestational day 17.5 and PND0, both GFAP-positive and
GFAP/Ki67 (Ki67 is a protein associated with proliferation) positive cells increase in
14 Astrocytes and Development of Neuroendocrine Circuits 371

the arcuate and supraoptic nuclei (Kim et al. 2016), indicating that both their number
and proliferation increase.
The first GFAP-positive cells in the suprachiasmatic nucleus (SCN) are found on
E15, suggesting the appearance of maturing astrocytes. The number of GFAP-
positive glia increases until P21, when an adult-like pattern of GFAP immunoreac-
tivity is observed (Botchkina and Morin 1995). In the ventrolateral portion of the
SCN, Munekawa and colleagues (Munekawa et al. 2000) reported that GFAP-
positive cells appear at embryonic day 20, increase at PND0, and then rapidly
increase between PND3 and PND4. These developmental changes are influenced
by light, with neuronal inputs from the retinohypothalamic tract being fundamental.
Although GFAP-positive cells are present in the ependymal layer of the arcuate
nucleus by the first postnatal week, consistent with these cells being developing
tanycytes, it is not until the second postnatal week when some GFAP-positive cells
are also located in more lateral regions of the arcuate nucleus, where astrocytes are
known to reside. Later, around the sixth postnatal week, GFAP-positive cells are
found close to arcuate neurons and the number of these cells increases through the
eighth postnatal week (Suarez et al. 1987). This study suggests that in the arcuate
nucleus, GFAP-positive cells that comprise the astrocytic population do not have a
mature cellular organization until after the eighth week of postnatal life. Although
gliogenesis in the hypothalamus (i.e., the proliferation of astrocytes, in this case)
continues through puberty and adulthood (Ahmed et al. 2008; Mohr et al. 2016), the
rate of incorporation decreases with age.

14.2.1 Postnatal Maturation of Astrocytes in the Hypothalamus

Postmitotic maturation of astrocytes includes the onset of, and an increase in,
expression of proteins important for astrocytic functions (e.g., glutamate
transporters) and increased arborization of glial projections. Maturation of astrocytes
is suggested to occur, at least in part, in response to surrounding neuronal activity,
which could be involved in the differential phenotype/functionality of astrocytes in
distinct brain regions. Maturation of astrocytes in areas such as the hippocampus
(Bushong et al. 2004) and cortex (Morel et al. 2014) has been shown to occur from
early postnatal life until approximately one month of age. Postnatal maturation of
astrocytes has been documented in some hypothalamic nuclei, but much is yet to be
learned regarding this process in the hypothalamus as it appears to differ to some
extent from that in other brain areas. For example, in the mouse cerebral cortex,
arborization of astrocytes increases between PND4 and PND26, with this rise being
most significant between PND14 and 26 (Morel et al. 2014). In contrast, in the
ventromedial or dorsomedial nuclei of the hypothalamus, this modification in
astrocyte morphology was not observed during this time period. Moreover, loss of
synaptic inputs from vesicular glutamate transporter 1 (VGluT1) positive neurons
during development had no effect on astrocyte arborization in these hypothalamic
nuclei, while in the cortex astrocyte morphology was altered and resulted in a
reduction in the glial ensheathment of synapses in this brain area, suggesting that
372 L. L. DonCarlos and J. A. Chowen

glutamate signaling is important for these morphological changes in the cortex

(Morel et al. 2014).
In the median eminence, electron microscopic examination of gonadotropin
releasing hormone (GnRH) neuron-containing transplants in adult hypogonadal
mice, which have a spontaneous deletion of GnRH expression, demonstrated that
the transplants sent out GnRH axons that generally followed specialized glial
“channels,” although the presence of astrocytes did not fully account for axon
targeting specificity since the GnRH axons did not home to reactive astrocytes
present at the transplantation site (Silverman et al. 1991). This may suggest that at
least some preoptic/hypothalamic astroglial-independent signals beyond the
channels provided cues to the GnRH axons to home to the ME or that reactive and
nonreactive astrocytes produce different factors (Zamanian et al. 2012).
The proliferation and maturation of astrocytes may differ between males and
females, although studies comparing these processes between the sexes in all areas
of the hypothalamus remain to be performed. The number of GFAP-positive
astrocytes in the SCN differs between males and females during development and
this is suggested to possibly participate in the establishment of the sex differences in
the neuroanatomy of this nucleus (Collado et al. 1995). Glutamate can arrest the cell
cycle of astrocytes and astrocytes in the prenatal preoptic area of male rats are
reported to be more susceptible to inhibition by L-glutamate than those from females
(Hsu et al. 2001). The morphology of astrocytes in the preoptic area differs between
the sexes as early as PND0, with astrocytes from males having more and longer
primary projections than those from females (Amateau and McCarthy 2002). Astro-
cyte morphology in the arcuate nucleus also differs between males and females as
early as PND0, with female astrocytes exhibiting a more biopolar morphology and
those from males are more complex and stellate shaped. Early estrogen actions
underlie these sex differences in both the preoptic area and arcuate nucleus (Amateau
and McCarthy 2002; Mong et al. 1999, 1996; Mong and McCarthy 2002). In
contrast, the response to early sex steroids differs in the VMH, as astrocyte mor-
phology at PND2 was found to be unmodified by differences in sex steroids (Mong
et al. 1999). Moreover, in this study, a clear difference between the detection of
GFAP-positive astrocytes in the arcuate nucleus and VMH at PND2 was reported,
with few cells being found in the VMH compared to the arcuate nucleus. Thus,
future studies are necessary to determine where sex differences in astrocyte number
and morphology occur in the hypothalamus and how this affects the function of these
areas. Moreover, identifying the signals involved in determining these early sex
differences is of great interest, particularly since expression of gonadal steroid
receptors in developing astrocytes appears to occur after most sex differences in
astrocyte morphology and responses have already been established (Lorenz et al.
2005) (see below).
14 Astrocytes and Development of Neuroendocrine Circuits 373

Box 14.1 Does GFAP Serve as a Proxy for Events Occurring

at the Terminals of Astrocytic Processes?
Large-scale studies of astrocyte dynamics in the hypothalamus have necessar-
ily relied on data from light-microscopic investigations, often using changes in
cellular GFAP distribution and expression to infer the impact of the changing
hormonal milieu on the physical modulation of interneuronal communication.
However, GFAP is an intermediate filament protein that is confined to more
proximal processes and to endfeet of those astrocytes that express GFAP,
meaning that GFAP does not demonstrate the entire extent of any given
astrocyte. Therefore, it is unclear at present to what extent GFAP serves to
delineate events taking place at the distal ends of astrocytes. Other methods
illustrating astrocyte morphology have revealed the potential short comings of
GFAP as a proxy for events occurring at astrocytic process terminals, although
each of these other methods has its own shortcomings. Other markers used for
labeling, genetically manipulating, or purifying astrocytes include glutamate
transporter-1 (Glt1), S100 calcium-binding protein B (S100b), aldehyde dehy-
drogenase family 1 member L1 (Aldh1L1), aldolase C (AldoC), and aquaporin 4.

14.3 Astrocyte Diversity

Our understanding and characterization of the development and diversity of

astrocytes, and how that diversity is determined, remains limited relative to what
we know about neurons. Nevertheless, informative studies on hypothalamic
astrocytes have emerged over the years, both in development and in response to
manipulations of gonadal steroid hormones or metabolism. An electron microscopic
study using the H222 antibody [which recognizes estrogen receptor (ER) alpha]
demonstrated that a subpopulation of hypothalamic astrocytes expresses ERs,
indicating molecular heterogeneity (Langub and Watson 1992). A later study,
using confocal microscopy to examine the immunolocalization of androgen
receptors in GFAP-positive cells, confirmed astrocyte heterogeneity, adding that
the expression of steroid receptors changes with age. A subpopulation of astrocytes
in the adult hypothalamus, but not the early postnatal hypothalamus, was shown to
be androgen receptor immunoreactive (AR-ir) (Johnson et al. 2012; Lorenz et al.
2005). Astrocytic responses to metabolic signals are also anatomically specific, as
expression of leptin receptors in these glial cells differs among hypothalamic nuclei,
as well as throughout the brain (Garcia-Caceres et al. 2012). Studies described below
show that morphological responses to hormonal signals vary by region as well.
Therefore, the development of astrocytes in the hypothalamus is likely to be affected
by both intrinsic and local signals that change with age, across regions, and within
regions.
374 L. L. DonCarlos and J. A. Chowen

New transcriptional profiling methods are paving the way for a more precise
study of the developmental changes and physiologic responses in potentially hetero-
geneous populations of astrocytes in the neuroendocrine system. One such method is
Act-seq, a modified single cell RNA-seq method of high throughput transcriptional
profiling (Wu et al. 2017). Act-seq was designed to prevent the artificial upregulation
of immediate early genes (IEGs) that typically occurs in response to proteases used
during conventional cell dissociation. The method employs the transcription inhibi-
tor, actinomycin-D, along with modifications in dissociation methods, to inhibit
transcription during cell dissociation prior to running a single cell RNA-seq assay
and cluster analysis. Three distinct clusters of astrocytes were found in the medial
amygdala. While all three clusters expressed known astrocytic markers, each cluster
also had a distinct transcriptional profile, suggestive of unique functions. By
suppressing dissociation-related upregulation of IEGs, the group was able to show
that the three subsets of astrocytes in the medial amygdala differentially upregulate
IEGs during a physiologic response, in this case, drug-induced seizure.
Similar approaches would be advantageous in studying hypothalamic astrocytes,
in particular because the data obtained would assist the identification of specific
genes of interest in neuroendocrine studies, allowing a better understanding of
astrocyte heterogeneity and what circumstances drive that heterogeneity. These
studies would further provide information to design and use additional methods to
study astrocytes, such as translating ribosome affinity purification (TRAP) for
translational profiling of the different subsets of astrocytes, or regional and morpho-
logical studies following environmental, genetic, or pharmacological manipulations.
In a study by Morel et al. (2017) TRAP analysis followed by RNA-Seq on samples
of somatosensory cortex, hippocampus, thalamus, hypothalamus, nucleus
accumbens, and the caudate-putamen identified both a dorsoventral gradient and
regional differences in expression of translating mRNAs. Hypothalamic astrocytes
were reported to most closely resemble thalamic astrocytes, with a 60% overall
similarity in transcriptional profile, and the transcriptional profiles of astrocytes in
these diencephalic regions were different from those in the telencephalic regions
sampled. It will be of great interest to determine the similarities, or differences,
between astrocytes from specific hypothalamic nuclei under physiologically relevant
paradigms. In the next section, we review the actions of astrocytes in the regulation
of neuroendocrine function, and then discuss what is known about the participation
of astrocytes in the development of neuroendocrine systems, and what molecular
mechanisms mediate these effects.

Box 14.2 Astrocyte Diversity Within Regions

Astrocytes are not a homogeneous population within the central nervous
system or, indeed, within a given region of the nervous system. This is the
case even though astrocytes appear to have properties that are intrinsic to these
cells at the species level (Oberheim et al. 2009). It is known, for example, that

(continued)
14 Astrocytes and Development of Neuroendocrine Circuits 375

Box 14.2 (continued)

human astrocyte precursors transplanted into the cerebral cortex of mice
develop the cellular morphology and functional impact on neurotransmission
that would be expected of astrocytes developing in the human cortex. From
region to region, though, there is a marked difference in astrocytes that can be
demonstrated on multiple levels (Zhang and Barres 2010).
Within a given region of the adult nervous system, astrocytes are heteroge-
neous as well. Illustrating this heterogeneity of neighboring astrocytes are data
on the expression of gonadal steroid hormone receptors. In neurons, only a
subset typically express steroid-specific gonadal steroid hormone receptors in
limbic regions (Morrell et al. 1986). Similarly, steroid receptors are not
expressed in all astrocytes across regions, or within regions. The proportion
of astrocytes within a region that express gonadal steroid receptors or other
proteins may depend on multiple factors, but regardless, the population is
heterogeneous. This constitutes an impediment to understanding how
astrocytes operate within particular regions since biochemical or molecular
methods to date are limited with respect to recognizing that heterogeneity
within a small area. Future studies will focus on identifying these
subpopulations and their specific, related, functions using genetic cell labeling
methods, single cell transcriptomic and proteomic methods, among others, as
discussed in the body of the chapter.

14.4 Astrocytes in Neuroendocrine Function

It has been four decades since hypothalamic astrocytes were first reported to
participate in the response to hormones and the control of specific neuroendocrine
functions (Tweedle and Hatton 1977). These early studies indicated that glial cells
were involved in the neuroendocrine control of osmotic homeostasis, lactation, and
reproduction (Olmos et al. 1989; Perlmutter et al. 1985; Theodosis et al. 1986;
Tweedle and Hatton 1977). Despite these intriguing early observations, glial cells
have remained in the background in neuroendocrine investigation for many years. It
is only recently that they have begun to receive their due attention, and important
advances have been made in understanding how they affect the functioning of
neuroendocrine systems, although there is still much to be learned. Ultimately,
astrocyte development in the preoptic area and hypothalamus may provide a power-
ful model for future investigation into the development of heterogeneity in astrocytes
in general.
376 L. L. DonCarlos and J. A. Chowen

14.4.1 Osmotic Homeostasis

Magnocellular neurons of the paraventricular nucleus (PVN) and supraoptic nucleus

(SON) are fundamental in controlling osmotic homeostasis by regulating the intake
and clearance of water and salts (Stern 2015). These neurons synthesize and release
the neurohypophysial hormones oxytocin and vasopressin into the blood. The
secretion of oxytocin and vasopressin from the neurohypophysis is stimulated by
an increase in plasma osmotic pressure and inhibited by decreased osmotic stimuli.
In addition to receiving information from both peripheral and central osmoreceptors,
these magnocellular neurons also have intrinsic osmosensing capabilities (Bourque
and Oliet 1997; Bourque et al. 1994).
The somas and dendrites of oxytocin and vasopressin neurons are in very close
proximity to one another (Fig. 14.3) and, under basal conditions, thin lamella-like
astrocytic processes separate the dendritic and perikaryal membranes (Tweedle and
Hatton 1977). Oxytocin stimulates the retraction of astrocytic processes in the SON
and PVN, resulting in increased contact between plasma membranes from adjacent
neuronal somas and adjacent dendrites (Marzban et al. 1992; Perlmutter et al. 1985;
Theodosis et al. 1986; Theodosis and Poulain 1984). Structural changes resulting
from magnocellular oxytocin release during lactation or parturition are thought to
participate in the coordinated activation of oxytocin neurons (Wang and Hatton
2007, 2009). These studies demonstrated an inverse relationship between astroglial
coverage of neuronal cell bodies and synaptic inputs to these neurons.
The decrease in astroglial coverage of oxytocin neurons during lactation results in
an increase in extracellular glutamate levels in the space surrounding these neurons,
which switches the function of the presynaptic kainate receptors that control GABA
release from inhibitory synapses on oxytocin neurons (Bonfardin et al. 2010).
Bonfardin and colleagues reported that in non-lactating rats, activation of kainate
receptors facilitates inhibitory GABAergic transmission. The opposite occurs in
lactating rats, where activation of kainate receptors inhibits GABA release. The
restructuring of glial-neuronal contacts results in modifications in synaptic inputs to
oxytocin neurons and in synaptic transmission that increases oxytocin neuronal
activity, culminating in milk ejection in lactating rats (Wang and Hatton 2007,
2009). Dehydration and rehydration also affect modifications in astroglial–neuronal
interactions of oxytocin and vasopressin neurons in the SON and PVN (Marzban
et al. 1992; Perlmutter et al. 1985; Tweedle and Hatton 1977). Moreover, these glial
cells release taurine in response to dehydration, which then activates glycine
receptors on nearby neurons (Deleuze et al. 1998) to regulate extrasynaptic gluta-
mate excitation of N-methyl-D-aspartate (NMDA) receptors (Joe et al. 2014).
Glial cells in the hypothalamic neurohypophysial system are reported to maintain
embryonic features that are important for these structural synaptic modifications.
These astrocytes possess features, such as expression of vimentin and high levels of
polysialyated NCAM, which are found in radial glia. In addition, in the adult,
astrocytes in the SON are also reported to maintain a radial glia-like morphology
(Bonfanti et al. 1993; Theodosis and Poulain 1999).
14 Astrocytes and Development of Neuroendocrine Circuits 377

Soma

Soma
Unsmulated

Soma Astroglial processes

Soma

Soma
Smulated

Soma Astroglial processes

Fig. 14.3 Schematic representation of magnocellular neurons of the paraventricular and supraoptic
nucleus. Under basal or unstimulated conditions, the somas (blue) of these neurons are separated by
astrocytic processes (green). When stimulated (e.g., by oxytocin for oxytocin neurons), astrocytic
processes are retracted and contact between the plasma membranes of neurons is increased

Our understanding of the development of astrocytic function in the neurohypo-

physial system is largely restricted to inference based on changes in astrocyte actions
in adult physiology. It is known, however, that both maternal obesity and interleukin
6 increase astrocyte proliferation in the fetal and neonatal SON (Kim et al. 2016) as
discussed below, but the functional impact of these manipulations on astrocytic
378 L. L. DonCarlos and J. A. Chowen

modulation of oxytocin or vasopressin release by the fetal, neonatal, or adult SON is

not known. A recent study addressed the negative impact of loss of dystrophin-71
(Dp-71) on SON function in mice (Souttou et al. 2019). Dp-71 belongs to a family of
dystrophins (Dps) that comprise components of the extracellular matrix cell receptor,
beta-dystroglycan (β-DG). Dp-71 is found in astrocytes as well as other cell types
within the fetal, neonatal, and adult SON. These results prompted the authors to
evaluate the distribution of Dp-71 specifically, and Dps, more generally, as well as
β-DG at birth and during early postnatal life. Changes were observed in the distribu-
tion of both Dps and β-DG with stage of development. Dps were reported to be in
astrocytic end feet surrounding magnocellular neurons by postnatal day 10, and by
postnatal day 20 were colocalized with β-DG in astrocytic end feet surrounding both
magnocellular neurons and blood vessels. These results, together with those showing
an impact of early metabolic cues on astrocyte development in the hypothalamus,
support a need for more investigation into the impact of physiological cues on
developing astrocytes and how these modify later responses of the hypothalamic
magnocellular system.

14.4.2 Reproduction and Estrogens

Modifications in the physical interactions between neurons and glia are also
involved in the neuroendocrine control of reproduction. Astrocytes in the preoptic
area and arcuate nucleus of the hypothalamus coordinate physiological responses to
estradiol (Garcia-Segura and McCarthy 2004), with these astrocytic responses
regulating the surge of gonadotropin hormone-releasing hormone (GnRH) that
precipitates female ovulation, a function critical for reproduction.
Ovariectomy of nonhuman primates increases the glial coverage of GnRH
neurons in both the preoptic area and medial-basal hypothalamus. This increased
neuronal ensheathment is associated with a decline in the number of synaptic
contacts onto these neurons (Witkin et al. 1991). Gonadal steroid replacement was
shown to at least partially reverse these changes in hypothalamic cellular structure
(Witkin et al. 1991), which agrees with the observation that neuronal-astrocyte
structural organization in the hypothalamus is modified throughout the estrous
cycle in relationship with changes in gonadal steroid levels.
In rodents, hypothalamic astrocytes also undergo morphological modifications in
association with the physiological hormonal variations that occur throughout the
estrous cycle and this is associated with coordinated changes in the number of
synapses (Fernandez-Galaz et al. 1997; Garcia-Segura et al. 2008). Between the
morning and afternoon of proestrus, astroglial processes increase their coverage of
neuronal cell surfaces and GABAergic synaptic terminals are removed from the
soma of arcuate neurons, and this coincides with the luteinizing hormone surge
(Csakvari et al. 2008). These glial processes remain in place, separating the presyn-
aptic terminals from the postsynaptic membrane until the morning of estrus. At this
time the glial processes retract so that by the morning of metestrus, GABAergic
synaptic contacts are reestablished in the arcuate nucleus (Csakvari et al. 2007).
14 Astrocytes and Development of Neuroendocrine Circuits 379

During proestrus when GABAergic inputs to the cell soma are decreased there is a
rise in the number of excitatory synapses on dendritic spines and together this results
in increased firing of arcuate neurons (Csakvari et al. 2007). The synaptic changes
observed in the arcuate nucleus during the estrous cycle are most likely driven by the
proestrus rise in circulating 17β-estradiol levels, as administration of this sex steroid
to ovariectomized rats can mimic the changes observed in cycling rats (Naftolin et al.
1996). A subpopulation of neurons in the arcuate nucleus projects to the median
eminence and controls pituitary hormone secretion and this estrogen-induced synap-
tic remodeling most likely induces functional changes in some neurosecretory
neurons to modulate pituitary hormone secretion (Csakvari et al. 2008), as well as
reorganization of synaptic circuits that control GnRH neurons (Garcia-Segura et al.
2008) and those involved in metabolic control (Gao et al. 2007).
The astrocytic responses in the preoptic area and arcuate nucleus are fascinating
because the estradiol-induced plasticity in morphology of astrocytic processes is
entirely different in the preoptic area from that in the arcuate nucleus, and yet
completely coordinated in ensuring appropriate regulation of female reproduction
(Garcia-Segura and McCarthy 2004). Astrocytes in both regions respond to estradiol
with alterations in extension of processes that affect neurotransmission, but in the
preoptic area, astrocytic processes respond by retracting processes that are
interposed between excitatory glutamatergic inputs and GnRH neurons, stimulating
GnRH release. In contrast, the astrocytic processes in the arcuate nucleus respond to
estradiol with interposition of processes that cut off inhibitory GABAergic synapses,
and further encourage GnRH release in response to high levels of estradiol and then
progesterone. The result is a coordinated increase in excitatory inputs and decrease
in inhibitory inputs to GnRH neurons that supports a sustained and powerful release
of GnRH. But, many questions remain. For example, how do the astrocytes in two
different regions develop the perfectly coordinated, but opposite, morphological
responses to the same stimulus? Does this differential response depend on the
developmental environment or the neuronal environment, since the preoptic area is
derived from telencephalic anlage, whereas the arcuate nucleus is derived from the
diencephalic anlage? How does local heterogeneity of astrocyte gene expression
develop, and how does this affect function?
Evidence indicates that the developmental environment can alter the number of
axosomatic synapses in the arcuate nucleus. In rodents, the number of axosomatic
synapses increases between postnatal days 10 and 20 in both males and females
(Perez et al. 1990). However, from postnatal day 20 onward females have more
axosomatic synapses than males, with this sex difference at least partially dependent
on the neonatal sex steroid environment (Perez et al. 1990). Astrocyte morphology
and expression of GFAP are also modified by the sex steroid environment during this
responsive period of hypothalamic development (Chowen et al. 1995), allowing
astrocytes to intervene in critical synaptic interactions (Garcia-Segura et al. 1995).
380 L. L. DonCarlos and J. A. Chowen

14.4.3 Metabolism

Information regarding the role of astrocytes in metabolic control has expanded

rapidly in recent years, due in part to the increased interest in the search for
treatments of the obesity epidemic. Research on the neuronal circuits involved in
appetite and metabolic control had been largely neglected until the need for thera-
peutics to combat obesity and its comorbidities became a prime interest for drug
development. The increased attention that has been focused on this dramatic health
problem has led to rapid advances in our understanding of how the CNS controls
systemic metabolism. The mechanisms of homeostatic energy control are mainly
coordinated by the hypothalamus, with the primary neuronal populations found in
the arcuate nucleus (Morton et al. 2006). Here, the majority of neuropeptide Y
(NPY) neurons co-express Agouti-related peptide (AgRP) and GABA and exert an
orexigenic effect (Krashes et al. 2013). In opposition, the anorexigenic
proopiomelanocortin (POMC) neurons in this brain area produce melanocyte-
stimulating hormone (MSH) to inhibit appetite and stimulate energy expenditure
(Boston et al. 1997). Astrocyte ensheathment of the somas of these important
metabolic neuronal populations is inversely related to the number of synaptic inputs
that the neurons receive, with this organization being modulated by metabolic
hormones such as leptin and ghrelin (Horvath et al. 2010; Pinto et al. 2004).
Moreover, the basal astroglial/synaptic arrangement in the arcuate nucleus is
suggested to influence the metabolic response of an individual, including their
propensity to become obese on a high fat diet (HFD) (Horvath et al. 2010). That
is, animals resistant to diet-induced obesity have a different synaptic organization/
astrocytic coverage than those animals that become obese on HFD. This hypotha-
lamic architecture involving glial-neuronal contact may result from both genetic and
environmental influences.
Notably, astrocytes in development encourage the formation of functional
synapses (Pfrieger and Barres 1997). For example, postnatal mice undergo a switch
in their method of nutrient intake, initially suckling, and then increasingly beginning
to eat solid food until weaning occurs. This switch involves the development of
synapses onto the key sets of neurons involved in feeding behavior in the arcuate
nucleus. Although the critical neurons are present early on, functional inhibitory
synapses onto neurons that regulate feeding behavior and energy metabolism are not
present during the early postnatal period. The development of astrocytes may
encourage incursion of the necessary glutamatergic axons into the region and
promote synaptogenesis.

14.5 Developmental Influences on Glial Organization

The number of GFAP-positive astrocytes in the arcuate nucleus is greater in rats

exposed to increased energy intake during neonatal life (Fig. 14.4), and these rats
have an increase in the number of primary astrocytic projections (Fuente-Martin
et al. 2012). In this study, no changes in the number of GFAP-positive astrocytes
14 Astrocytes and Development of Neuroendocrine Circuits 381

Maternal obesity
Neonatal overnutrion
Lepn
BDNF
IL6
Increased
number of
astrocytes

Radial glia Astrocytes

3rd ventricle

Arcuate nucleus

Fig. 14.4 Factors changing the number of astrocytes in the arcuate nucleus during fetal and
neonatal development. Several different physiological signals can alter astrocyte numbers at
different times in development. BDNF Brain derived neurotrophic factor, IL6 Interleukin 6

were observed in other brain areas such as the hippocampus or cerebellum,

suggesting a region-specific effect of this early overnutrition.
Maternal obesity increases astrocyte proliferation in both the fetal and neonatal
mouse hypothalamus (Fig. 14.4), with specific increases found in the arcuate nucleus
and the SON (Kim et al. 2016). These authors suggest that the obesity-associated rise
in interleukin (IL) 6 in the maternal circulation could be at least one of the factors
involved in increased numbers of astrocytes in the fetal hypothalamus, as well as the
increased proliferative capacity of hypothalamic astrocytes in the offspring in later
life (Kim et al. 2016). The astroglial coverage and function of POMC neurons in
offspring are modified by maternal high fat intake. In the offspring of dams fed HFD,
the astroglial coverage of POMC neurons in the arcuate nucleus is significantly
increased compared with the offspring of mothers on a normal chow diet (Fuente-
Martin et al. 2012). These highly-ensheathed POMC neurons display a decreased
382 L. L. DonCarlos and J. A. Chowen

frequency of miniature inhibitory postsynaptic currents. The glucose-sensing capac-

ity of highly ensheathed POMC neurons in HFD-exposed offspring was also
affected, as their action potential frequency was decreased in response to lower
glucose concentrations, a response that was not observed in offspring of normal
chow fed mothers (Fuente-Martin et al. 2012).

14.5.1 Leptin Influences on Metabolic System Development

and Glial Organization

Leptin, a cytokine produced mainly by adipose tissue, acts in the hypothalamus to

inhibit food intake. This adipokine also influences the development of metabolic
circuits, increasing the projections from metabolic neurons in the arcuate nucleus to
the PVN (Bouret et al. 2004), which may be partially mediated by astrocyte-
dependent actions of leptin. Leptin induces astrogenesis in the developing hypothal-
amus (Ahima et al. 1999; Rottkamp et al. 2015), while removal of the leptin receptor
during early postnatal life from these glial cells reduces their proliferation and the
number of GFAP-positive cells at postnatal day 21, but does not affect the number of
hypothalamic neurons (Rottkamp et al. 2015). Leptin levels increase in neonatally
overfed animals during the critical period for the development of metabolic circuits
(Argente-Arizon et al. 2016; Fuente-Martin et al. 2012), and the rise in circulating
leptin levels could underlie the increase in the number of hypothalamic astrocytes in
these animals.

14.5.2 Astrocytes and BDNF in Metabolic Circuits

Brain derived neurotrophic factor (BDNF) is important for the development of

projections of POMC and AgRP neurons from the arcuate nucleus to the PVN and
dorsomedial nucleus of the hypothalamus (DMH), although the number of POMC or
AgRP neurons in the arcuate nucleus were unchanged in BDNF genetic knockout
mice (Liao et al. 2015). The lack of BDNF specifically reduced projections from
POMC neurons of the arcuate nucleus to neurons of the DMH and from AgRP
neurons of the arcuate nucleus to neurons located in the lateral portion of the PVN.
Astrocytes produce BDNF (Wu et al. 2004) and activation of the melanocortin
4 receptor (MC4R) induces production of this growth factor (Caruso et al., 2012;
Ramirez et al. 2015). BDNF is also involved in synaptic plasticity in metabolic
circuits (Zhang et al. 2017) and regulates inhibitory synapse formation during brain
development (Elmariah et al. 2005).
Other growth factors, such as ciliary neurotrophic factor, (CNTF) have been
reported to induce generation of metabolically active neurons in the arcuate nucleus
of the adult mouse (Kokoeva et al. 2005), although these results are controversial
(Borg et al. 2016). Astrocytes produce and respond to CNTF (Muller et al. 2007;
Poluch and Juliano 2007), with CNTF treatment inducing radial glia to differentiate
into astrocytes, disrupting neuron migration (Poluch and Juliano 2007).
14 Astrocytes and Development of Neuroendocrine Circuits 383

Many questions remain regarding the interaction of astrocytes and neurons and
how they control appetite and systemic metabolism. For example, astrocytic contact
with both POMC and NPY neurons is increased in obese subjects, in association
with a decrease in the number of somatic synaptic inputs in both populations of
neurons (Horvath et al. 2010). However, on POMC neurons this decrease is due to a
decline in inhibitory inputs, whereas on NPY neurons it is the result of a reduction in
stimulatory inputs, which together promote an overall balance toward an anorexi-
genic signal. How is this specific rewiring coordinated? What is the mechanism of
cross talk between astroglia and neurons that generates this specific synaptic
rewiring, both during development and in adults? Are the astrocytes that surround
these two neuronal populations different, even though they both respond to leptin? Is
there cross talk between astrocytes and neurons during development that ensures that
the required functional circuits in each hypothalamic area are formed, although the
signals involved remain elusive? The hypothalamic circuits controlling each neuro-
endocrine system or axis are often studied and spoken of as if they function
independently, but there must also be precise communication between these circuits
to maintain systemic homeostasis. Astrocytes and radial glia are involved in the
generation of neurons and their projections. How they determine the precise organi-
zation of the neuroendocrine hypothalamus remains largely unknown.
Studies using genetically modified mice can be used to provide evidence for how
neurons affect astrocytes. For example, Gnasxl-deficient mice, which lack a
G-protein alpha stimulatory subunit that regulates energy balance, are lean and
hypermetabolic (Holmes et al. 2016). In these animals, expression of GFAP in the
adult hypothalamus was greatly diminished and the numbers of astrocytes and a
subpopulation of tanycytes (alpha) were also diminished. In juvenile animals as late
as postnatal day 15, however, the numbers of astrocytes and tanycytes were not
different in mutants. Since Gnasxl is expressed only in neurons, this study provides
tentative evidence that late development of astrocytes in the hypothalamus is
modulated by neuronal activity, although it is also possible that any effect of these
mutants on astrocytes may be secondary to undernutrition.

14.6 The Role of Tanycytes in Neuroendocrine Regulation

Tanycytes, specialized radial glia located in the ventral wall of the third ventricle,
have been proposed to be a source of neural progenitor cells in the adult hypothala-
mus (Haan et al. 2013; Lee et al. 2012; Robins et al. 2013). Tanycytes are classified
into subtypes depending on their location and function, with some studies indicating
that all subtypes have neural progenitor properties. However, other studies suggest
that α2-tanycytes generate β-tanycytes in vivo and that only α2-tanycytes are
capable of generating proliferative neurospheres in vitro (Robins et al. 2013).
Tanycytes have also been proposed to participate in the neuroendocrine control of
reproduction and metabolism (Ojeda et al. 2003; Prevot et al. 2018) and to modulate
hypothalamic control of the thyroid axis (Muller-Fielitz et al. 2017). Tanycytes have
end feet that extend into the VMH, arcuate nucleus, and median eminence
384 L. L. DonCarlos and J. A. Chowen

(Rodriguez et al. 2005) and can be retracted. They contribute to the regulation of
GnRH release, as their retractable end feet are interlaced with GnRH synaptic
terminals and the portal vasculature in the median eminence (Baroncini et al.
2007; Ojeda et al. 2008; Prevot et al. 2018). Extension of tanycytic processes to
ensheath GnRH terminals prevents release of GnRH into the portal vasculature.
These tanycytic processes retract during the preovulatory stage of the estrous cycle,
which permits the contact of GnRH terminals with portal capillaries for neuropeptide
release (Bellefontaine et al. 2011; Prevot et al. 1999). More than 30 years ago, Sergio
Ojeda hypothesized that glial cells released substances that modulated the release of
GnRH to the portal vasculature (Ojeda 1994). These signals are now known to
include prostaglandin E2, transforming growth factor beta, and nitric oxide (NO),
which stimulate and coordinate GnRH neuronal firing (Bellefontaine et al. 2011;
Clasadonte et al. 2011; Melcangi et al. 1995). More recently, tanycytic were shown
to produce semaphorin7A which binds to plexinC1 and β1-integrin (Parkash et al.
2015). The levels of semaphorin7A are modulated by ovarian steroids, and the
signaling of semaphorin 7A is necessary for normal estrous cycling and fertility. It
is proposed that semaphorin7A acts through is receptors in a coordinated fashion to
induce expansion of tanycytic end feet and promote the retraction of GnRH nerve
terminals from the pericapillary space (Parkash et al. 2015).

14.7 Perspectives

In sum, astrocytes have been difficult to study because: (1) they are not electrically
active and do not generate action potentials; (2) they have divergent morphologies
(including small nuclei and irregular processes); and (3) there is a lack of specific
antibodies or other markers that identify astrocytic subpopulations.
The lack of astrocyte-specific reagents is certainly problematic. For example,
GFAP is found in astrocytes and progenitor cells and not all astrocytes express
GFAP. Moreover, GFAP immunoreactivity does not allow full visualization of
astrocyte morphology, and the extent of GFAP staining within astrocytes may be
less indicative of morphological changes than of changes in GFAP expression. Other
difficulties relate to the problem of visualizing astrocytes, especially at the same time
as neurons since they are often found in different planes, making it difficult to
simultaneously label and visualize the nuclei of both cell types.
Controversies in studies of astrocytes have also arisen recently, even for issues
that were previously considered to be somewhat resolved. For example, the idea that
there are co-transmitters released by astrocytes that assist neuronal neurotransmis-
sion has been called into question (Wolosker et al. 2016, 2017). Does the concept of
the tripartite synapse hold true, or is this functionally restricted to developmental
stages? Are calcium waves between astrocytes an artifact of culture conditions,
rather than an in vivo event? Should the umbrella term “astrocyte” be eliminated
in favor of more specific terms that refer to distinct subtypes of this class of glia? Do
astrocytes deliver glucose, lactate or both to neurons?
14 Astrocytes and Development of Neuroendocrine Circuits 385

Few studies to date have used optogenetic techniques to interrogate astrocyte

function in the hypothalamus. In one study, channelrhodopsin-2 was activated
specifically in astrocytes of the posterior hypothalamus of mice following injections
of a recombinant adeno-associated virus in which the GFAP promoter drove expres-
sion of channelrhodopsin and a reporter gene. Optogenetic activation of astrocytes
increased sleep during the active (lights-off) phase (Pelluru et al. 2016). Optogenetic
stimulation of medial-basal hypothalamic astrocytes was shown to suppress food
intake in a frequency-dependent manner (Sweeney et al. 2016). These methods will
likely provide additional tools for dissecting the role of astrocytes in neuroendocrine
development, especially once more specific markers of astrocyte subpopulations
have been identified.
As of this writing, many perspectives about astrocytes remain in flux partly
because significant developmental changes in their competence are now recognized.
Astrocytes are typically derived from young animals for in vitro studies and these
cells do not necessarily behave as do those from adults (Verkhratsky and Nedergaard
2018), or as young astrocytes in vivo for that matter. The renewed interest in
astrocytes in neuroendocrine systems will undoubtedly advance our understanding
of these fascinating cells, both in their developmental capacities and in their lifelong
contribution to regulating physiological processes.

Key References

Ahmed et al. (2008). Challanged the dogma of that pubertal hormones, as opposed to
sex steroids during embryological/neonatal development did not affect
neurogenesis.
Bouret et al. (2004). Seminal study in understanding how early metabolic/hormonal
changes can have long-term effects on metabolism.
Malatestaet al. (2003). Important contributions in order to understand radial glia.
Mong et al. (1996). One of the first studies to demonstrate sex differences in glial
cells in the hypothalamus.
Morel et al. (2017). Important advances in understanding differences in astrocytes in
different brain regions.
Naftolin et al. (1996). An early study associating synaptic remodeling in the hypo-
thalamus with a physiological outcome.
Perlmutter et al. (1985). A seminal study in understanding the physiological out-
come of interactions between neurons and glia in the hypothalamus.
Pfrieger et al. (1997). A clear advance in undestanding the role of astrocytes in
synaptic transmission.
Prevot et al. (2018). A recent update on the implications of tanycytes in neuroendo-
crine control.
Theodosis et al. (1986). A seminal study of neuron-glial interactions in physiology.
386 L. L. DonCarlos and J. A. Chowen

References
Ahima RS, Bjorbaek C, Osei S, Flier JS (1999) Regulation of neuronal and glial proteins by leptin:
implications for brain development. Endocrinology 140:2755–2762
Ahmed EI, Zehr JL, Schulz KM, Lorenz BH, DonCarlos LL, Sisk CL (2008) Pubertal hormones
modulate the addition of new cells to sexually dimorphic brain regions. Nat Neurosci
11:995–997
Allen NJ, Lyons DA (2018) Glia as architects of central nervous system formation and function.
Science 362:181–185
Altman J, Bayer SA (1986) The development of the rat hypothalamus. Adv Anat Embryol Cell Biol
100:1–178
Amateau SK, McCarthy MM (2002) Sexual differentiation of astrocyte morphology in the devel-
oping rat preoptic area. J Neuroendocrinol 14:904–910
Argente-Arizon P, Ros P, Diaz F, Fuente-Martin E, Castro-Gonzalez D, Sanchez-Garrido MA,
Barrios V, Tena-Sempere M, Argente J, Chowen JA (2016) Age and sex dependent effects of
early overnutrition on metabolic parameters and the role of neonatal androgens. Biol Sex Differ
7:26
Baroncini M, Allet C, Leroy D, Beauvillain JC, Francke JP, Prevot V (2007) Morphological
evidence for direct interaction between gonadotrophin-releasing hormone neurones and
astroglial cells in the human hypothalamus. J Neuroendocrinol 19:691–702
Bayer SA, Altman J (1987) Development of the preoptic area: time and site of origin, migratory
routes, and settling patterns of its neurons. J Comp Neurol 265:65–95
Belanger M, Magistretti PJ (2009) The role of astroglia in neuroprotection. Dialog Clin Neurosci
11:281–295
Bellefontaine N, Hanchate NK, Parkash J, Campagne C, de Seranno S, Clasadonte J, d’Anglemont
de Tassigny X, Prevot V (2011) Nitric oxide as key mediator of neuron-to-neuron and
endothelia-to-glia communication involved in the neuroendocrine control of reproduction.
Neuroendocrinology 93:74–89
Bonfanti L, Poulain DA, Theodosis DT (1993) Radial glia-like cells in the supraoptic nucleus of the
adult rat. J Neuroendocrinol 5:1–5
Bonfardin VD, Fossat P, Theodosis DT, Oliet SH (2010) Glia-dependent switch of kainate receptor
presynaptic action. J Neurosci 30:985–995
Borg ML, Reichenbach A, Lemus M, Oldfield BJ, Andrews ZB, Watt MJ (2016) Central adminis-
tration of the ciliary neurotrophic factor analogue, axokine, does not play a role in long-term
energy homeostasis in adult mice. Neuroendocrinology 103:223–229
Boston BA, Blaydon KM, Varnerin J, Cone RD (1997) Independent and additive effects of central
POMC and leptin pathways on murine obesity. Science 278:1641–1644
Bosworth AP, Allen NJ (2017) The diverse actions of astrocytes during synaptic development. Curr
Opin Neurobiol 47:38–43
Botchkina GI, Morin LP (1995) Ontogeny of radial glia, astrocytes and vasoactive intestinal peptide
immunoreactive neurons in hamster suprachiasmatic nucleus. Brain Res Dev Brain Res
86:48–56
Bouret SG, Simerly RB (2007) Development of leptin-sensitive circuits. J Neuroendocrinol 19
(8):575–582
Bouret SG, Draper SJ, Simerly RB (2004) Trophic action of leptin on hypothalamic neurons that
regulate feeding. Science 304:108–110
Bourque CW, Oliet SH (1997) Osmoreceptors in the central nervous system. Annu Rev Physiol
59:601–619
Bourque CW, Oliet SH, Richard D (1994) Osmoreceptors, osmoreception, and osmoregulation.
Front Neuroendocrinol 15:231–274
Bushong EA, Martone ME, Ellisman MH (2004) Maturation of astrocyte morphology and the
establishment of astrocyte domains during postnatal hippocampal development. Int J Dev
Neurosci 22:73–86
14 Astrocytes and Development of Neuroendocrine Circuits 387

Caruso C, Carniglia L, Durand D, Gonzalez PV, Scimonelli TN, Lasaga M (2012) Melanocortin
4 receptor activation induces brain-derived neurotrophic factor expression in rat astrocytes
through cyclic AMP-protein kinase A pathway. Mol Cell Endocrinol 348:47–54
Chowen JA, Busiguina S, Garcia-Segura LM (1995) Sexual dimorphism and sex steroid modula-
tion of glial fibrillary acidic protein messenger RNA and immunoreactivity levels in the rat
hypothalamus. Neuroscience 69:519–532
Chowen JA, Argente-Arizon P, Freire-Regatillo A, Frago LM, Horvath TL, Argente J (2016) The
role of astrocytes in the hypothalamic response and adaptation to metabolic signals. Prog
Neurobiol 144:68–87
Chung WS, Welsh CA, Barres BA, Stevens B (2015) Do glia drive synaptic and cognitive
impairment in disease? Nat Neurosci 18:1539–1545
Clasadonte J, Poulain P, Hanchate NK, Corfas G, Ojeda SR, Prevot V (2011) Prostaglandin E2
release from astrocytes triggers gonadotropin-releasing hormone (GnRH) neuron firing via EP2
receptor activation. Proc Natl Acad Sci U S A 108:16104–16109
Collado P, Beyer C, Hutchison JB, Holman SD (1995) Hypothalamic distribution of astrocytes is
gender-related in Mongolian gerbils. Neurosci Lett 184:86–89
Csakvari E, Hoyk Z, Gyenes A, Garcia-Ovejero D, Garcia-Segura LM, Parducz A (2007) Fluctua-
tion of synapse density in the arcuate nucleus during the estrous cycle. Neuroscience
144:1288–1292
Csakvari E, Kurunczi A, Hoyk Z, Gyenes A, Naftolin F, Parducz A (2008) Estradiol-induced
synaptic remodeling of tyrosine hydroxylase immunopositive neurons in the rat arcuate nucleus.
Endocrinology 149:4137–4141
Deleuze C, Duvoid A, Hussy N (1998) Properties and glial origin of osmotic-dependent release of
taurine from the rat supraoptic nucleus. J Physiol 507(Pt 2):463–471
Elmariah SB, Oh EJ, Hughes EG, Balice-Gordon RJ (2005) Astrocytes regulate inhibitory synapse
formation via Trk-mediated modulation of postsynaptic GABAA receptors. J Neurosci
25:3638–3650
Ero C, Gewaltig MO, Keller D, Markram H (2018) A cell atlas for the mouse brain. Front
Neuroinform 12:84
Fernandez-Galaz MC, Morschl E, Chowen JA, Torres-Aleman I, Naftolin F, Garcia-Segura LM
(1997) Role of astroglia and insulin-like growth factor-I in gonadal hormone-dependent synap-
tic plasticity. Brain Res Bull 44:525–531
Fuente-Martin E, Garcia-Caceres C, Granado M, de Ceballos ML, Sanchez-Garrido MA, Sarman B,
Liu ZW, Dietrich MO, Tena-Sempere M, Argente-Arizon P, Diaz F, Argente J, Horvath TL,
Chowen JA (2012) Leptin regulates glutamate and glucose transporters in hypothalamic
astrocytes. J Clin Invest 122:3900–3913
Gao Q, Mezei G, Nie Y, Rao Y, Choi CS, Bechmann I, Leranth C, Toran-Allerand D, Priest CA,
Roberts JL, Gao XB, Mobbs C, Shulman GI, Diano S, Horvath TL (2007) Anorectic estrogen
mimics leptin’s effect on the rewiring of melanocortin cells and Stat3 signaling in obese
animals. Nat Med 13:89–94
Garcia-Caceres C, Fuente-Martin E, Argente J, Chowen JA (2012) Emerging role of glial cells in
the control of body weight. Mol Metab 1:37–46
Garcia-Segura LM, McCarthy MM (2004) Minireview: Role of glia in neuroendocrine function.
Endocrinology 145:1082–1086
Garcia-Segura LM, Duenas M, Busiguina S, Naftolin F, Chowen JA (1995) Gonadal hormone
regulation of neuronal-glial interactions in the developing neuroendocrine hypothalamus. J
Steroid Biochem Mol Biol 53:293–298
Garcia-Segura LM, Lorenz B, DonCarlos LL (2008) The role of glia in the hypothalamus:
implications for gonadal steroid feedback and reproductive neuroendocrine output. Reproduc-
tion 135(4):419–429
Haan N, Goodman T, Najdi-Samiei A, Stratford CM, Rice R, El Agha E, Bellusci S, Hajihosseini
MK (2013) Fgf10-expressing tanycytes add new neurons to the appetite/energy-balance
regulating centers of the postnatal and adult hypothalamus. J Neurosci 33:6170–6180
388 L. L. DonCarlos and J. A. Chowen

Hatton GI (1997) Function-related plasticity in hypothalamus. Annu Rev Neurosci 20:375–397

Holmes AP, Wong SQ, Pulix M, Johnson K, Horton NS, Thomas P, de Magalhaes JP, Plagge A
(2016) Reductions in hypothalamic Gfap expression, glial cells and alpha-tanycytes in lean and
hypermetabolic Gnasxl-deficient mice. Mol Brain 9:39
Horvath TL, Sarman B, Garcia-Caceres C, Enriori PJ, Sotonyi P, Shanabrough M, Borok E,
Argente J, Chowen JA, Perez-Tilve D, Pfluger PT, Bronneke HS, Levin BE, Diano S, Cowley
MA, Tschop MH (2010) Synaptic input organization of the melanocortin system predicts diet-
induced hypothalamic reactive gliosis and obesity. Proc Natl Acad Sci U S A 107:14875–14880
Hsu C, Hsieh YL, Ho ML, Hsu HK, Yu JY (2001) Sexually dimorphic effect of glutamate treatment
on cell cycle arrestment of astrocytes from the preoptic area of neonatal rats. Dev Neurosci
23:399–405
Joe N, Scott V, Brown CH (2014) Glial regulation of extrasynaptic NMDA receptor-mediated
excitation of supraoptic nucleus neurones during dehydration. J Neuroendocrinol 26:35–42
Johnson RT, Schneider A, DonCarlos LL, Breedlove SM, Jordan CL (2012) Astrocytes in the rat
medial amygdala are responsive to adult androgens. J Comp Neurol 520:2531–2544
Kim DW, Glendining KA, Grattan DR, Jasoni CL (2016) Maternal obesity leads to increased
proliferation and numbers of astrocytes in the developing fetal and neonatal mouse hypothala-
mus. Int J Dev Neurosci 53:18–25
Kokoeva MV, Yin H, Flier JS (2005) Neurogenesis in the hypothalamus of adult mice: potential
role in energy balance. Science 310:679–683
Krashes MJ, Shah BP, Koda S, Lowell BB (2013) Rapid versus delayed stimulation of feeding by
the endogenously released AgRP neuron mediators GABA, NPY, and AgRP. Cell Metab
18:588–595
Langub MC Jr, Watson RE Jr (1992) Estrogen receptor-immunoreactive glia, endothelia, and
ependyma in guinea pig preoptic area and median eminence: electron microscopy. Endocrinol-
ogy 130:364–372
Lee DA, Bedont JL, Pak T, Wang H, Song J, Miranda-Angulo A, Takiar V, Charubhumi V,
Balordi F, Takebayashi H, Aja S, Ford E, Fishell G, Blackshaw S (2012) Tanycytes of the
hypothalamic median eminence form a diet-responsive neurogenic niche. Nat Neurosci 15
(6):700–702
Leloup C, Allard C, Carneiro L, Fioramonti X, Collins S, Penicaud L (2015) Glucose and
hypothalamic astrocytes: more than a fueling role? Neuroscience 323:110–120
Liao G-Y, Bouyer K, Kamitakahara A, Sahibzada N, Wang C-H, Rutlin M, Simerly RB, Xu B
(2015) Brain-derived neurotrophic factor is required for axonal growth of selective groups of
neurons in the arcuate nucleus. Mol Metab 4(6):471–482
Lorenz B, Garcia-Segura LM, DonCarlos LL (2005) Cellular phenotype of androgen receptor-
immunoreactive nuclei in the developing and adult rat brain. J Comp Neurol 492:456–468
Magistretti PJ, Allaman I (2018) Lactate in the brain: from metabolic end-product to signalling
molecule. Nat Rev Neurosci 19:235–249
Malatesta P, Hartfuss E, Gotz M (2000) Isolation of radial glial cells by fluorescent-activated cell
sorting reveals a neuronal lineage. Development 127:5253–5263
Malatesta P, Hack MA, Hartfuss E, Kettenmann H, Klinkert W, Kirchhoff F, Gotz M (2003)
Neuronal or glial progeny: regional differences in radial glia fate. Neuron 37:751–764
Marzban F, Tweedle CD, Hatton GI (1992) Reevaluation of the plasticity in the rat supraoptic
nucleus after chronic dehydration using immunogold for oxytocin and vasopressin at the
ultrastructural level. Brain Res Bull 28:757–766
Melcangi RC, Galbiati M, Messi E, Piva F, Martini L, Motta M (1995) Type 1 astrocytes influence
luteinizing hormone-releasing hormone release from the hypothalamic cell line GT1-1: is
transforming growth factor-beta the principle involved? Endocrinology 136:679–686
Mohr MA, Garcia FL, DonCarlos LL, Sisk CL (2016) Neurons and glial cells are added to the
female rat anteroventral periventricular nucleus during puberty. Endocrinology 157:2393–2402
Mong JA, McCarthy MM (2002) Ontogeny of sexually dimorphic astrocytes in the neonatal rat
arcuate. Brain Res Dev Brain Res 139:151–158
14 Astrocytes and Development of Neuroendocrine Circuits 389

Mong JA, Kurzweil RL, Davis AM, Rocca MS, McCarthy MM (1996) Evidence for sexual
differentiation of glia in rat brain. Horm Behav 30:553–562
Mong JA, Glaser E, McCarthy MM (1999) Gonadal steroids promote glial differentiation and alter
neuronal morphology in the developing hypothalamus in a regionally specific manner. J
Neurosci 19:1464–1472
Morel L, Higashimori H, Tolman M, Yang Y (2014) VGluT1+ neuronal glutamatergic signaling
regulates postnatal developmental maturation of cortical protoplasmic astroglia. J Neurosci
34:10950–10962
Morel L, Chiang MSR, Higashimori H, Shoneye T, Iyer LK, Yelick J, Tai A, Yang Y (2017)
Molecular and functional properties of regional astrocytes in the adult brain. J Neurosci
37:8706–8717
Morrell JI, Krieger MS, Pfaff DW (1986) Quantitative autoradiographic analysis of estradiol
retention by cells in the preoptic area, hypothalamus and amygdala. Exp Brain Res 62: 343–354
Morton GJ, Cummings DE, Baskin DG, Barsh GS, Schwartz MW (2006) Central nervous system
control of food intake and body weight. Nature 443:289–295
Muller A, Hauk TG, Fischer D (2007) Astrocyte-derived CNTF switches mature RGCs to a
regenerative state following inflammatory stimulation. Brain 130:3308–3320
Muller-Fielitz H, Stahr M, Bernau M, Richter M, Abele S, Krajka V, Benzin A, Wenzel J, Kalies K,
Mittag J, Heuer H, Offermanns S, Schwaninger M (2017) Tanycytes control the hormonal
output of the hypothalamic-pituitary-thyroid axis. Nat Commun 8:484
Munekawa K, Tamada Y, Iijima N, Hayashi S, Ishihara A, Inoue K, Tanaka M, Ibata Y (2000)
Development of astroglial elements in the suprachiasmatic nucleus of the rat: with special
reference to the involvement of the optic nerve. Exp Neurol 166:44–51
Naftolin F, Mor G, Horvath TL, Luquin S, Fajer AB, Kohen F, Garcia-Segura LM (1996) Synaptic
remodeling in the arcuate nucleus during the estrous cycle is induced by estrogen and precedes
the preovulatory gonadotropin surge. Endocrinology 137:5576–5580
Noctor SC, Flint AC, Weissman TA, Dammerman RS, Kriegstein AR (2001) Neurons derived from
radial glial cells establish radial units in neocortex. Nature 409:714–720
Oberheim NA, Takano T, Han X, He W, Lin JHC, Wang F, Qiwu Xu Q, Wyatt JD, Pilcher W,
Ojemann JG, Ransom BR, Goldman SA, Nedergaard M (2009) Uniquely hominid features of
adult human astrocytes. J Neurosci 29(10):3276
Ojeda SR (1994) The neurobiology of mammalian puberty: has the contribution of glial cells been
underestimated? J NIH Res 6:51–56
Ojeda SR, Prevot V, Heger S, Lomniczi A, Dziedzic B, Mungenast A (2003) Glia-to-neuron
signaling and the neuroendocrine control of female puberty. Ann Med 35:244–255
Ojeda SR, Lomniczi A, Sandau US (2008) Glial-gonadotrophin hormone (GnRH) neurone
interactions in the median eminence and the control of GnRH secretion. J Neuroendocrinol
20:732–742
Olmos G, Naftolin F, Perez J, Tranque PA, Garcia-Segura LM (1989) Synaptic remodeling in the
rat arcuate nucleus during the estrous cycle. Neuroscience 32:663–667
Parkash J, Messina A, Langlet F, Cimino I, Loyens A, Mazur D, Gallet S, Balland E, Malone SA,
Pralong F, Cagnoni G, Schellino R, De Marchis S, Mazzone M, Pasterkamp RJ, Tamagnone L,
Prevot V, Giacobini P (2015) Semaphorin7A regulates neuroglial plasticity in the adult hypo-
thalamic median eminence. Nat Commun 6:6385
Pelluru D, Konadhode RR, Bhat NR, Shiromani PJ (2016) Optogenetic stimulation of astrocytes in
the posterior hypothalamus increases sleep at night in C57BL/6J mice. Eur J Neurosci
43:1298–1306
Perea G, Navarrete M, Araque A (2009) Tripartite synapses: astrocytes process and control synaptic
information. Trends Neurosci 32:421–431
Perez J, Naftolin F, Garcia Segura LM (1990) Sexual differentiation of synaptic connectivity and
neuronal plasma membrane in the arcuate nucleus of the rat hypothalamus. Brain Res
527:116–122
390 L. L. DonCarlos and J. A. Chowen

Perlmutter LS, Tweedle CD, Hatton GI (1985) Neuronal/glial plasticity in the supraoptic dendritic
zone in response to acute and chronic dehydration. Brain Res 361:225–232
Pfrieger FW, Barres BA (1997) Synaptic efficacy enhanced by glial cells in vitro. Science
277:1684–1687
Pinto S, Roseberry AG, Liu H, Diano S, Shanabrough M, Cai X, Friedman JM, Horvath TL (2004)
Rapid rewiring of arcuate nucleus feeding circuits by leptin. Science 304(5667):110–115
Poluch S, Juliano SL (2007) A normal radial glial scaffold is necessary for migration of
interneurons during neocortical development. Glia 55:822–830
Prevot V, Croix D, Bouret S, Dutoit S, Tramu G, Stefano GB, Beauvillain JC (1999) Definitive
evidence for the existence of morphological plasticity in the external zone of the median
eminence during the rat estrous cycle: implication of neuro-glio-endothelial interactions in
gonadotropin-releasing hormone release. Neuroscience 94:809–819
Prevot V, Dehouck B, Sharif A, Ciofi P, Giacobini P, Clasadonte J (2018) The versatile tanycyte: a
hypothalamic integrator of reproduction and energy metabolism. Endocr Rev 39:333–368
Ramirez D, Saba J, Carniglia L, Durand D, Lasaga M, Caruso C (2015) Melanocortin 4 receptor
activates ERK-cFos pathway to increase brain-derived neurotrophic factor expression in rat
astrocytes and hypothalamus. Mol Cell Endocrinol 411:28–37
Robins SC, Stewart I, McNay DE, Taylor V, Giachino C, Goetz M, Ninkovic J, Briancon N,
Maratos-Flier E, Flier JS, Kokoeva MV, Placzek M (2013) alpha-Tanycytes of the adult
hypothalamic third ventricle include distinct populations of FGF-responsive neural progenitors.
Nat Commun 4:2049
Rodriguez EM, Blazquez JL, Pastor FE, Pelaez B, Pena P, Peruzzo B, Amat P (2005) Hypothalamic
tanycytes: a key component of brain-endocrine interaction. Int Rev Cytol 247:89–164
Rottkamp DM, Rudenko IA, Maier MT, Roshanbin S, Yulyaningsih E, Perez L, Valdearcos M,
Chua S, Koliwad SK, Xu AW (2015) Leptin potentiates astrogenesis in the developing
hypothalamus. Mol Metab 4:881–889
Silverman RC, Gibson MJ, Silverman AJ (1991) Relationship of glia to GnRH axonal outgrowth
from third ventricular grafts in hpg hosts. Exp Neurol 114:259–274
Sims RE, Butcher JB, Parri HR, Glazewski S (2015) Astrocyte and neuronal plasticity in the
somatosensory system. Neural Plast 2015:732014
Souttou S, Benabdesselam R, Siqueiros-Marquez L, Sifi M, Deliba M, Vacca O, Charles-Messance-
H, Vaillend C, Rendon A, Guillonneau X, Dorbani-Mamine L (2019) Expression and localiza-
tion of dystrophins and beta-dystroglycan in the hypothalamic supraoptic nuclei of rat from birth
to adulthood. Acta Histochem 121:218–226
Stern JE (2015) Neuroendocrine-autonomic integration in the paraventricular nucleus: novel roles
for dendritically released neuropeptides. J Neuroendocrinol 27:487–497
Suarez I, Fernandez B, Bodega G, Tranque P, Olmos G, Garcia-Segura LM (1987) Postnatal
development of glial fibrillary acidic protein immunoreactivity in the hamster arcuate nucleus.
Brain Res 465:89–95
Sweeney P, Qi Y, Xu Z, Yang Y (2016) Activation of hypothalamic astrocytes suppresses feeding
without altering emotional states. Glia 64:2263–2273
Theodosis DT, Poulain DA (1984) Evidence for structural plasticity in the supraoptic nucleus of the
rat hypothalamus in relation to gestation and lactation. Neuroscience 11:183–193
Theodosis DT, Poulain DA (1999) Contribution of astrocytes to activity-dependent structural
plasticity in the adult brain. Adv Exp Med Biol 468:175–182
Theodosis DT, Montagnese C, Rodriguez F, Vincent JD, Poulain DA (1986) Oxytocin induces
morphological plasticity in the adult hypothalamo-neurohypophysial system. Nature
322:738–740
Tweedle CD, Hatton GI (1977) Ultrastructural changes in rat hypothalamic neurosecretory cells and
their associated glia during minimal dehydration and rehydration. Cell Tissue Res 181:59–72
Verkhratsky A, Nedergaard M (2018) Physiology of astroglia. Physiol Rev 98:239–389
Wang YF, Hatton GI (2007) Interaction of extracellular signal-regulated protein kinase 1/2 with
actin cytoskeleton in supraoptic oxytocin neurons and astrocytes: role in burst firing. J Neurosci
27:13822–13834
14 Astrocytes and Development of Neuroendocrine Circuits 391

Wang YF, Hatton GI (2009) Astrocytic plasticity and patterned oxytocin neuronal activity: dynamic
interactions. J Neurosci 29:1743–1754
Witkin JW, Ferin M, Popilskis SJ, Silverman AJ (1991) Effects of gonadal steroids on the
ultrastructure of GnRH neurons in the rhesus monkey: synaptic input and glial apposition.
Endocrinology 129:1083–1092
Wolosker H, Balu DT, Coyle JT (2016) The rise and fall of the D-serine-mediated gliotransmission
hypothesis. Trends Neurosci 39:712–721
Wolosker H, Balu DT, Coyle JT (2017) Astroglial versus neuronal D-serine: check your controls!
Trends Neurosci 40:520–522
Wu H, Friedman WJ, Dreyfus CF (2004) Differential regulation of neurotrophin expression in basal
forebrain astrocytes by neuronal signals. J Neurosci Res 76:76–85
Wu YE, Pan L, Zuo Y, Li X, Hong W (2017) Detecting activated cell populations using single-cell
RNA-seq. Neuron 96:313–329.e316
Zamanian JL, Xu L, Foo LC, Nouri N, Zhou L, Giffard RG, Barres BA (2012) Genomic analysis of
reactive astrogliosis. J Neurosci 32(18):6391–6410
Zhang Y, Barres BA (2010) Astrocyte heterogeneity: an underappreciated topic in neurobiology.
Curr Opin Neurobiol 20:588–594
Zhang Y, Reichel JM, Han C, Zuniga-Hertz JP, Cai D (2017) Astrocytic process plasticity and
IKKbeta/NF-kappaB in central control of blood glucose, blood pressure, and body weight. Cell
Metab 25:1091–1102.e1094
Origins of Sex Differentiation of Brain
and Behavior 15
Margaret M. McCarthy

Abstract
Sex differences in adult brain and behavior are often established early in devel-
opment when the brain is remarkably immature. In adults sex differences take on
many forms including latent variables, dimorphisms, frequency, and more.
Androgens and estrogens from the developing male fetal testis masculinize the
brain, so that physiology and behavior are in synchrony and harmony with
gonadal phenotype. This differentiation process occurs during a critical window
in males, but females remain sensitive to exogenous steroid treatment for a longer
developmental period. The cellular mechanisms of sexual differentiation are
diverse but often involve inflammatory signaling molecules and immune cells.
This may have consequences for the higher vulnerability of males to neurological
and neuropsychiatric disorders of development.

Keywords
Amygdala · Androgens · Estrogens · Hypothalamus · Neuroimmune · Preoptic
area

15.1 Introduction

The brain is a reproductive organ, essential for both the control of gamete production
by the gonads and the physiology and behaviors required for achieving fertilization,
gestation, parturition, and parenting. By definition, these are different in males and
females since sex is defined by gamete size, production, and delivery. But determin-
ing how the brain drives the different physiologies and behaviors of males and
females is confounded by the inherent complexity of the nervous system and its wide

M. M. McCarthy (*)
Department of Pharmacology, University of Maryland School of Medicine, Baltimore, MD, USA
e-mail: [email protected]

# Springer Nature Switzerland AG 2020 393

S. Wray, S. Blackshaw (eds.), Developmental Neuroendocrinology, Masterclass in
Neuroendocrinology 9, https://doi.org/10.1007/978-3-030-40002-6_15
394 M. M. McCarthy

range of functions, from regulating heart beat to composing a symphony. Many of

these functions are the same, or essentially the same, in males and females, but many
are not. Moreover, many things influence how a brain develops, including nutrition,
genetics, injury, nurturing, stress, and environmental complexity. Sex differences
may appear under some circumstances or diminish under others. Males and females
may exhibit the same response, but the underlying neural mechanisms of that
response may be profoundly different. What is true in one species might not be
true in another. For these reasons, it is essential to be precise in discussions of sex
differences in physiology and behavior, but it is also possible to strive to identify
general principles that broadly apply.
This chapter will focus on the role of testicular hormones in driving masculiniza-
tion of brain and behavior, with an emphasis on rodents, both because it is what we
know the most about as it has been the preferred model system for over 50 years, and
because it provides a frame work for the appropriate study of more complex
organisms, including humans.

15.2 Historical Overview

The story of sex differences in the brain is one of on-again off-again. Unbeknownst
to many students of behavioral neuroendocrinology, in the 1960s and 1970s it was
considered radical to say there was any difference in the brains of males and females,
so radical that the first two studies in mammals were published in Science (Pfaff
1966; Raisman and Field 1971), and both found only very small and seemingly
insignificant differences, thereby confirming that male and female brains were
largely the same. This was prior to a full understanding of the neural control of
gonadotropin secretion from the pituitary, and indeed even the relationship between
the brain and the “master gland” was misunderstood until the 1950s (Melmed 2011).
In the 1960s, Charles Barraclough made the surprising observation that treating
newborn female rat pups with high doses of androgen resulted in a sterile adult
(Barraclough and Gorski 1961). He assumed this was the result of a direct effect on
the immature ovary, but through systematic investigation determined that it was
actually the result of dysregulation in the brain, and resulted from the loss of the
ability to release a surge of luteinizing hormone from the pituitary. This illustrated
for the first time that hormones early in development could “program” the brain
differently in males and females, but there were no clear neuroanatomical
underpinnings to the effect.
The story changed dramatically when studies on song birds revealed robust
differences in the nuclei in the brain that control song. This was intuited by the
fact that males sing complex melodies while females mostly tweet, but the fact that
this behavioral difference so clearly mapped onto neuroanatomy had not been
previously demonstrated (Nottebohm and Arnold 1976). The observation in birds
reopened investigation into mammals and several neuroanatomical sex differences
related to reproduction were subsequently identified. Among these are the sexually
dimorphic nucleus of the preoptic area (SDN-POA), the spinal nucleus of the
15 Origins of Sex Differentiation of Brain and Behavior 395

bulbocavernosus (SNB), and the anteroventral periventricular nucleus (AVPV).

Because each of these differences was directly tied to a reproductive endpoint, either
behavior or gonadotropin secretion, everyone was comfortable. Yes, male and
female brains differ, the thinking was, but only within the narrow confines of
reproduction. A paradigm shift came in the 1990s when Bruce McEwen and
colleagues published a series of papers showing effects of estrogens on the density
of dendritic spines on hippocampal pyramidal neurons (Gould et al. 1990). At that
time the hippocampus was the focus of an enormous swath of neuroscience for its
role in learning and memory, as well as regulating the stress axis. As a result, the
effect that McEwen and colleagues reported drew tremendous interest, and not
without some initial skepticism. Eventually the evidence of hormonal modulation
of hippocampal function was irrefutable and while not a sex difference per se, the
profound effect of female hormones on a nonreproductive related neuroanatomical
endpoint suggested that the influence of sex could be more widespread in the
nervous system than previously thought. This may or may not be the reason that
about the same time there was a constriction in neurophysiological studies to almost
exclusive use of male subjects (Beery and Zucker 2010). As a result, there was little
thought given to sex differences by most neuroscientists, and our increased under-
standing of the brain over the next two decades has been largely built around the
model male.
Despite the neglect of sex as an important variable in the majority of neuroscience
research, the field of behavioral neuroendocrinology continued to flourish with
steady progress made in understanding the cellular and molecular mechanisms by
which steroids impact both the developing brain and adult physiology and behavior.

Box 15.1 Scientific Societies Dedicated to Hormones and Behavior

In response to the increasing awareness of the widespread impact of hormones
on the brain, a new scientific society was formed in 1996 called the Society for
Behavioral Neuroendocrinology (SBN). The journal Hormones and Behavior
was adopted as its official journal, a publication founded in 1969 and predating
the era of neuroscience. Ten years on from the founding of SBN saw formation
of a second scientific society, the Organization for the Study of Sex
Differences (OSSD), which broadly addresses all aspects of biology that differ
in males and females and formed a new journal, Biology of Sex Differences.

Box 15.2 NIH Policy on SABV

The last few years have seen a paradigm shift, the inclusion of sex as a
biological variable in all NIH- funded preclinical research in the US (Clayton
and Collins 2014), as well as efforts at Canadian and European funding
agencies to increase the inclusion of female subjects in research, now referred

(continued)
396 M. M. McCarthy

Box 15.2 (continued)

to as SABV for Sex as a Biological Variable. The intent was to remove the
impact of gender, which includes the impact of cultural and societal
expectations, and focuses only on that which is biological. Whether and how
this will transform our understanding of the brain and all its mysteries is
unknown, but early signs suggest that dogmas will fall, and new and unex-
pected vistas will appear.

15.3 All Sex Differences Are Not Created Equal

In science, semantics matter. Precision in word use is essential for accurately

conveying meaning in our principle mode of communication, writing. Often the
greatest slippage in the precise use of language is with the most common words. In
the realm of sex differences research, there are two mistakes that are particularly
frequent. First is the use of the word gender in discussion of research on animals.
Gender is a human construct that combines self-perception and others’ perceptions
of one’s sex. We cannot ask animals what sex they think they are, nor what sex they
think others are, and so they do not have gender. Thus, any discussion of gender
differences should be restricted to humans. The second common mistake is the
overuse of the term “sexually dimorphic” to describe any and all sex differences.

(a) Dimorphic is a compound word consisting of di ¼ two, and morph ¼ form, so

two forms. If something is sexually dimorphic this means it has two different
forms in males and females, which does happen, but not very often. The most
notable would be the control of LH release from the pituitary which is not only
pulsatile in males and females but also cyclical in females (and in seasonally
breeding males and females), including a large surge for inducing ovulation.
Mating behavior is also sexually dimorphic, involving mounting and
intromitting in males and solicitations and lordosis in females. But many people
use the term sexually dimorphic for what is really a sex difference, often a very
small one. The descriptor “sex difference” is the appropriate term to apply to any
continuous variable that on average differs between males and females. In many
instances, there is overlap in the response or endpoint in males and females, but
the average is significantly different. Ironically, the sexually dimorphic nucleus
of the POA, or SDN, is arguably the most historically important sex difference
in the brain, but it is really just a sex difference. Both males and females have an
SDN, it is just much larger in males. Indeed, it is so much larger there is very
little overlap between the sexes, making it reasonable to call it “dimorphic.”
(b) Sometimes sex differences are latent, meaning they are not readily visible or
only appear after a challenge or in a specific context. Learning is one such
example, with males performing better under stress and female performance
15 Origins of Sex Differentiation of Brain and Behavior 397

deteriorating. Tracey Shors and colleagues determined this was correlated with a
change in the synaptic profile of hippocampal neurons known to be associated
with that particular form of learning (Shors 2006). An additional means by
which sex differences may be hidden is by appearing the same in presentation in
the two sexes but differing in prevalence. Alzheimer’s disease is the same
disease in men and women, but more frequently diagnosed in women, and not
just because they live longer. Conversely, stuttering is more common in boys
than girls. This type of sex difference has been referred to as a population sex
difference.
(c) Yet another form that sex differences take is in the existence of convergence,
also referred to as compensation and first articulated by Geert De Vries (2004).
In this instance, males and females are at a different baseline or set point, but
under the appropriate circumstance can converge on the same endpoint. Some-
times this occurs because of a lack of the necessary physiology in one sex or the
other. For instance, nurturing behavior toward young by adult males is not a
general feature of the behavioral repertoire of most rodent species, including
toward their own progeny, whereas females who have given birth universally
care for their offspring. Pregnancy and parturition involve marked physiological
and hormonal changes that are critical to inducing maternal behavior. In
mammals, males are deprived of this experience, both physiologically and
hence behaviorally as well, with one notable exception. Males in one particular
species of monogamous voles have evolved a unique vasopressinergic circuitry
that promotes paternal behavior, meaning they have compensated for the lack of
exposure to the hormones of pregnancy by generating a new independent
regulatory circuit. Catherine Woolley has articulated a related but distinct
principle that she also refers to as latent sex differences. A latent sex difference
is one that is not immediately apparent. For instance, measures of synapse
function of hippocampal pyramidal neurons are the same in males and females,
but the underlying cellular mechanisms mediating synaptic plasticity are highly
distinct, involving separate signal transduction pathways or different receptors
all together (Huang and Woolley 2012). It is essential to conceptualize the
various ways a sex difference can manifest when attempting to unravel and
understand the origins of either a developmental or adult response profile
(Fig. 15.1).

15.4 Critical and Sensitive Periods for Sexual Differentiation

The brain is designed to respond to the environment and reprogram itself in

accordance with what it encounters, i.e., learn. The reprogramming that occurs
during development is particularly important as it is anticipatory of future events
and circumstances. This often means that one developmental path is followed at the
398 M. M. McCarthy

Fig. 15.1 Categories of sex differences. There are multiple ways in which the sexes can differ and
these begin with whether the dependent variable (endpoint) is the same or different. Left. When the
endpoint is different, it may be sexually dimorphic, meaning it takes on a different form in males
versus females, or it may be the average measure of the endpoint is significantly different in the two
sexes. A third possibility is that the two sexes are similar at baseline but diverge to different
endpoints upon a challenge or with aging. Right. When the endpoint is the same it can be the result
of different pathways or mechanisms, which are referred to as latent variables. Alternatively, the
sexes may be divergent at baseline but converge to the same endpoint following challenge or evolve
a mechanism to achieve the same endpoint via distinct pathway (therefore involving latent
variables). Lastly, the endpoint may be the same in males and females but occur more frequently
in the population in one sex versus the other

exclusion of another. It also usually means the decision point is restricted to a critical
period during which the relevant environmental stimulus is more salient and the
biological processes mediating fate determination are in synchrony. One of the best
examples is the coinciding events of eye opening and innervation of the lateral
geniculate in altricial mammals. Light is the essential external stimulus, and coordi-
nated neuronal activity is the essential internal biological response. The result is a
functioning visual system. If light is withheld during the critical period, the animal
will be forever blind. Going one step further, if an animal is raised during the critical
period in an environment consisting of only vertical lines, no horizontal lines, it will
be forever blind to horizontal lines. In this way, the nervous system fine tunes its
responses during the critical period, and those responses are then permanently
embedded in the brain.
In the case of the visual system, the environmental signal is external light. But
internal signals can also drive critical periods of developmental programming. The
amount and type of nutrition experienced prenatally programs future metabolomics.
Similarly, the amount and type of stress before and shortly after birth impact adult
behavioral and physiological phenotypes associated with anxiety, emotionality, and
activation of the hypothalamic–pituitary–adrenal axis. In both these instances, males
and females may differ in sensitivity or other aspects of critical period programming,
15 Origins of Sex Differentiation of Brain and Behavior 399

Fig. 15.2 Critical and sensitive periods for sexual differentiation in rodents and humans. The
critical period for sexual differentiation begins with the onset of testicular androgen production by
the developing fetus in rodents and humans. This is deemed a critical period because if androgens
levels do not rise at this time in males, the brain will fail to be masculinized. The sensitive period is
defined as that time during which a female can be treated with exogenous steroid (testosterone or
estrogen, or other masculinizing agents) and has the potential to be phenotypically converted to the
male for a particular endpoint. In rodents this extends into the first postnatal week but may vary by
endpoint. In humans, we do not know when or if there is an analogous sensitive period

but they both are exposed to the same salient internal or external stimuli. The case of
sexual differentiation of the brain is unique, however, in that only one sex is exposed
to the essential agent that drives the developmental program down a particular path.
That essential agent is testosterone from the testis of the fetal male (Fig. 15.2). Here
the goal is to differentiate male from female in order to assure that adult neurophysi-
ology and behavior is in synchrony with gonadal phenotype. Thus, males need to
continuously release small pulses of LH from the pituitary to maintain spermatogen-
esis and they need to seek out and mate with females and, in some species, compete
with other males and/or defend territories and vital resources. Females, by contrast,
must be capable of a massive surge in LH release in order to induce ovulation. They
must also seek out and mate with males, but usually preferably only one, and then
gestate, deliver, and nurture the offspring. Nurturing is a complex combination of
nesting, lactating, retrieving, grooming, and defending the young, each of which
involves distinct neural circuits and remain incompletely understood. This complex
divergence in physiology and behavior is programmed into the brain beginning prior
to birth of the individual, despite the actual endpoints not being manifest until
adulthood.
400 M. M. McCarthy

15.5 Masculinization, Feminization, and Defeminization

Up to this point, the process of masculinization has been emphasized because it is the
key “differentiation” event in sexual differentiation, meaning it is the process by
which the male is differentiated from the female. The feminization of both brain and
body is the developmental trajectory that occurs in the absence of testis and androgen
production, meaning an ovary is not required for female development (But it is
obviously essential to adult reproductive capacity). This does not mean that femini-
zation of the brain is not an active process, it surely is, but it is much harder to discern
what it is in the absence of some third “neither male nor female” phenotype. There is
some evidence of a later critical period in female brain development, during the
second postnatal week, that may involve estrogen production by the ovaries (Bakker
and Baum 2008), but unfortunately, this concept is not fully developed.
Perhaps the best angle on discerning what is feminization, is its polar opposite,
defeminization, which is an active process whereby the female phenotype is removed
(Fig. 15.3). This phenomenon is best illustrated, and perhaps limited to, the sex-typic
mating behavior seen in rats and mice where males show mounting of sexually
receptive females which respond with lordosis, a posture that allows the male to
intromit his penis. Since feminization is the default, the neural circuitry of lordosis
comes as preinstalled software. Removing that programming is achieved via defemi-
nization which is also driven by androgens aromatized to estrogens in developing

Fig. 15.3 Masculinization, feminization and defeminization. There are three distinct processes that
occur during sexual differentiation of the brain to regulate adult mating behavior. Upper. Mascu-
linization is induced by gonadal steroids of the fetal male testis to organize the neural substrate to be
responsive to androgens in adulthood and promote male mating behavior. Defeminization is also a
steroid induced process occurring during the same time as masculinization, and occurs via distinct
cellular mechanisms that result in erasing the innate program of feminization. As a result, sex
differences in adult mating behavior are maximized. Lower. Feminization occurs in the absence of
gonadal steroids and programs the neural substrate to respond to cyclical estrogens and progestins in
adulthood to promote female mating behavior. These events occur during distinct periods of
development, although the parameters surrounding a sensitive period for feminization remain
much less well defined than those for masculinization
15 Origins of Sex Differentiation of Brain and Behavior 401

males but via distinct cellular mechanisms (Schwarz and McCarthy 2008). Why such
a system has evolved is a mystery, and whether it applies outside the context of sex
behavior is debatable, but it tells us there are multiple independent processes occur-
ring simultaneously in the developing brain that assure as little overlap as possible
between males and females for certain key reproductive functions. Notably, there is
no parallel process of demasculinization in females, and when masculinization is
blocked in males it is best referred to as dysmasculinization, since it is the disruption
of a normal process, not itself a normal process.

15.6 Testicular Steroids Drive Sexual Differentiation

Everyone is aware that puberty is a time of marked divergence for males and females
in terms of the hormonal milieu. But many are not as aware that as early as fetal life
there is a similar marked divergence created when the fetal testis becomes steroido-
genic and synthesizes sufficiently high levels of androgens that the steroid enters the
circulation and gains access to the fetal brain. In rodents, this process begins during
the last days of an approximately 3-week gestation and extends into the first hours
after birth (Weisz and Ward 1980). In human males, fetal steroidogenesis starts early
in the second trimester and is waning by birth but followed by a later active period
postnatally that is sometimes referred to as “mini-puberty” (Kuiri-Hanninen et al.
2014). The opening of the critical period for sexual differentiation of the brain is
operationally defined by the onset of testicular androgen production. Artificially
elevating androgens prior to the naturally occurring period of steroidogenesis is
largely without effect, as the essential biological processes are not yet in play. The
closing of the critical period is another matter. In males, the critical period closes
when the surge of androgen production ends. Once the system is fully exposed to
testosterone the programming process is initiated and cannot be undone, meaning
that within one day of birth masculinization is largely complete and the critical
period closed. Females, however, remain sensitive to the programming potential of
testosterone well into the postnatal period, up to a week (Fig. 15.2). If injected with
either testosterone or estradiol, the masculinization process is initiated, reflecting a
sensitive period as opposed to a critical period (McCarthy et al. 2018). This
seemingly lucky accident of timing in rodents provides an excellent tool for studying
the mechanisms of steroid-mediated programming, by essentially using the female
as a platform for exogenously induced masculinization. In nonhuman primates,
however, the end of the critical and the sensitive period are both prenatal (Wallen
and Baum 2002), confounding our ability to replicate mechanistic findings in
rodents and instead forcing us to consider other routes for translational insight,
which will be discussed below. Moreover, since we cannot experiment on humans,
we do not know if there is a postnatal sensitive period in addition to the prenatal
critical period.
The rodent model has many advantages, but there are confounds just as there are
in any “model.” The ability to induce the masculinization process in newborn
females is an important strength but since the naturally occurring process in males
402 M. M. McCarthy

begins in utero, the experimental females are in fact older when treatment occurs. At
this stage of development, even a few days is a lot of time for changes to occur. It is
possible to treat pregnant dams with steroids and thereby begin the exposure of
females in utero, but the endocrine milieu of late pregnancy is both robust and
dynamic in preparation for birth. Introducing exogenous steroids against this back-
drop is unavoidably confounded. In early studies, beginning with the iconic Phoenix
et al. (1959) study, androgen treatment of pregnant guinea pigs not only
masculinized the sexual behavior of female offspring but it also masculinized the
genitalia, an obvious confound that needs to be avoided if one wishes to understand
the masculinization of the brain in isolation from the body. Lastly, rodents give birth
to litters, meaning multiple offspring develop in two uterine horns where they are
packed in like peas in a pod. Evidence of just how sensitive the developing brain is to
gonadal steroids is found in the intrauterine position (IUP) phenomenon whereby
females who develop between two males are slightly masculinized and males who
develop between two females are slightly less masculinized relative to littermates
that sit between mixed-sex or same sex pups. The shift in masculinization is due to
androgen production of fetal males being so profuse that some is transferred to
nearest neighbors. This is a naturally occurring source of individual variation in
phenotype that is independent of genetics and is often overlooked. Endpoints that are
impacted by IUP include adult hormonal levels, aggression, sexual behavior, and
susceptibility to endocrine disruption. Limited evidence in humans suggests that
females from mixed-sex twins exhibit some evidence of masculinization (Ryan and
Vandenbergh 2002).

15.7 Impacts of Critical and Sensitive Periods Are Delayed

The famous geneticist Dobszhansky is often quoted for saying “Nothing in biology
makes sense except in the light of evolution.” Equally true would be “Nothing in
neuroscience makes sense except in the light of development.” The brain is the
slowest developing organ of the body and is designed to be impacted by the local
environment and experience as it matures. A fundamental goal of neuroscience is to
understand how the brain controls complex adult behaviors. We want to know how
the brain determines that now is the time to feed, now to run, now to mate, now to
flee, now to fight, and now to sleep. All of these decisions are made against a
backdrop of current context and past experience, but they are also impacted by how
the brain was wired from the beginning, during development. Development is a
simple word for a dynamic multifactorial robust process that is also sensitive to
perturbations, sometimes only during a specific window or time period. More
importantly, in development what happens first affects what happens next, which
affects what happens next and so on in an increasingly constrained one-way trajec-
tory. Development cannot be undone. It is the pervasive underpinning of all that
follows.
Sexual differentiation of the brain is a developmental process but one of the more
astonishing things about it is the timing, meaning how early. In humans our best
15 Origins of Sex Differentiation of Brain and Behavior 403

estimate of the critical period for sexual differentiation is early second trimester to
mid- to late-third trimester. This means the brain is being organized for behaviors
that will not be expressed for a decade or more, with all the intervening maturation of
this most complex organ. In rodents the critical period is more clearly defined due to
our ability to experiment, beginning with the onset of testicular hormones during the
last days of gestation and lasting until the first few days of life. But here too the brain
is remarkably immature during the process of differentiation. So how is it that these
early organizational events endure until adulthood?
The conceptualization of sexual differentiation has evolved along with our
expanding views of the brain and its capabilities. In the 1970s and 1980s the mature
brain was still considered essentially immutable, with no capacity for neurogenesis,
regeneration, or even new synapse formation. The brain was often interpreted as
analogous to a computer, being hard wired and programmed. Consistent with that, the
impact of sexual differentiation was referred to as building the “neuroarchitecture,”
establishing a “blue print” and “constructing” a circuit. But we now know the brain is
highly mutable, even plastic, with new neurons being born well into adulthood and
synapses coming and going with astonishing frequency. At first this view makes it
even more challenging to fathom how a developmental process in a brain that is
essentially a gelatinous mass can be programmed to regulate later complex social
behaviors. Here a second revolution in our thinking about the brain comes into play,
the role of epigenetic modifications.

15.8 Neuroepigenetics

Epigenetics literally means “above the genome” and thus refers to modifications to
the genetic code that impact gene expression. The dominant forms of epigenetic
modifications are changes to the histones that are integral to the chromatin around
the DNA and the addition of methyl groups to the nucleotides themselves. There are
a variety of changes to the histones but the most prevalent is the addition or removal
of acetyl and/or methyl groups to lysine residues. Changes to the DNA itself are
largely restricted to the addition of methyl groups and this occurs mostly but not
exclusively on cytosines proximal to guanines, so-called CpGs. The notion of
epigenetics has been with us for a long time, dating back to the geneticist
Waddington in the 1940s who invoked the term to explain how the same DNA in
every cell could lead to divergent cell states. There needed to be a mechanism by
which some genes would be permanently repressed while others were expressed.
This is required in order for a hepatocyte to remain a hepatocyte while an osteoclast
stays an osteoclast.
Epigenetics was rediscovered in the nervous system as a means to explain how
behavioral phenotypic variation could be transferred from dam to pup. More specifi-
cally, Michael Meaney and colleagues determined that rat dams that were highly
attentive to their pups produced adult females that were also highly attentive to their
pups. That this was not genetic in the classic sense was proven by cross fostering
pups born to low attentive dams to highly attentive dams and observing that they
404 M. M. McCarthy

were highly attentive as adults. Epigenetic modifications in the promoter region of

the glucocorticoid receptor were identified as a mediator of this transfer in behavioral
phenotype (Szyf et al. 2007). Many subsequent studies have revealed important roles
for epigenetics in behaviors ranging from learning and memory to drug addiction. A
central principle emerged that when it comes to epigenetics, the brain is different.
Rather than the tight irreversible epigenetic control exerted during cell fate determi-
nation, epigenetics of the nervous system is far more pliable and appears designed to
respond to experience and the local environment. This is presumably a means for the
brain to predict the future and thereby prepare itself. The term neuroepigenetics was
coined in recognition of the seemingly unique rules and regulations regarding
epigenetics in the brain (Sweatt 2013).
Sexual differentiation of the brain is also a case of predicting the future, as adult
physiology and behavior need to be in synch with gonadal phenotype. Since gonadal
differentiation is determined by the presence or absence of the gene Sry, and NOT the
Y chromosome per se, the brain matches itself to the gonads instead of chromosome
complement. This can be achieved epigenetically. Moreover, steroid hormone
receptors are transcription factors that interact with the DNA both directly and
indirectly and form complexes with other proteins that possess enzymatic activity
that regulates histone modifications. These are often in the category of co-activators,
such as SRC-1, and co-repressors, such as NCor (Smith and O’Malley 2004). Given
the intimate relationship between steroids and the genome, it was natural to postulate
epigenetic programming as the basis for the enduring impact of sexual differentiation.
There are two general approaches to exploring epigenetic contributions to sex
differences in the brain. First is to pick a key enzyme or epigenetic modification and
then measure it in males versus females. This approach has detected sex differences
in some histone modifications, as well as activity of the enzymes methylating DNA,
the DNA methyltransferases (DNMTs) (McCarthy and Rissman 2014). Second is to
use pharmacological or genetic manipulations of the same key enzymes and see if
they impact outcomes differently in males versus females. The impacted outcomes
may be neuroanatomical, physiological, or behavioral. This approach was used to
detect a critical role for histone modifications in the onset of puberty (Lomniczi et al.
2013) and for DNA methylation in masculinization of the preoptic area (Nugent
et al. 2015).
The POA is a key brain region regulating male sexual behavior (Hull and
Dominguez 2007). If the POA is lesioned or pharmacologically inactivated, males
lose all interest in sex. If the POA is stimulated, males lose all interest in anything but
sex. The synaptic profile of POA dendrites is established within the first few days of
life, endures to adulthood and correlates with the frequency and intensity of male
mating behavior in a standardized test (Wright et al. 2008). In an effort to understand
how the synaptic density pattern is maintained across the lifespan, the methylation of
DNA in the POA was quantified and found to be greater in females (Nugent et al.
2015). This was determined to be the result of greater activity of the DNMTs. Males
have lower enzymatic activity due to inhibition from estrogens derived from testicu-
lar androgens. To determine if this methylation was functionally significant, neonatal
females were injected with a drug into the POA that demethylates the DNA and
15 Origins of Sex Differentiation of Brain and Behavior 405

briefly prevents further methylation during the first few postnatal days and then
raised to adulthood. In order to determine if their behavior was masculinized, the
females were ovariectomized and implanted with silastic capsules that release
testosterone. Once the circulating levels of testosterone in these females were on
par with that normally seen in males, they were tested for mating behavior in a
controlled setting using test females who were sexually receptive.
Intriguingly, these females were fully masculinized, behaving just as normal
males do, and examination of their brains at completion of the experiment revealed
the synaptic pattern was the same as that seen in males as well. Collectively these
data suggest that epigenetic modifications produced by steroid-mediated DNMT
activity during the critical period help to maintain a dendritic architecture that is
conducive to male-typic mating behavior in adulthood (Fig. 15.4). Of particular
interest, this is a steroid driven process but does not involve interaction of the steroid
and its receptor with the DNA that is being modified, it is more of an epigenetic
modification. The mechanism by which estradiol modulates DNMT activity is
unknown but there is no change in either the mRNA or the protein levels of
DNMT1, DNMT3, and DNMT3a. Steroids never cease to amaze.
Critical periods have a beginning and an end. The beginning is usually associated
with the onset of some stimulus, such as in the case of sexual differentiation, the
onset of steroidogenesis by the fetal testis. But the ends of critical periods are much
more elusive, usually being defined as that developmental time point at which the
brain no long responds to the requisite stimulus. But what determines the loss of
responsiveness? The observation of a change in DNA methylation offered a clue. If
hormone-responsive genes are epigenetically silenced at a particular developmental
time point, that could close the window of the critical/sensitive period. Indeed, this
does appear to be the case since in the aforementioned study. If females were treated
with the DNA demethylating drug after the normal close of the critical/sensitive
period, they were still masculinized as evidenced by adult brain and behavior.
Importantly, if females were treated with a masculinizing dose of steroid at the
same time point, there was no effect, indicating a loss of steroid responsiveness due
to DNA methylation. Thus, the close of the critical/sensitive period for sexual
differentiation may be mediated by epigenetic repression of hormone-responsive
genes that promote masculinization (Nugent et al. 2015).
In addition to the epigenetic effects of steroids described above, there are
examples of direct gene transcription mediated sexual differentiation. The amygdala
is a brain region critical to social behavior in general and a mediator of sex
differences in juvenile play behavior in particular. In males the cells of the amygdala
express more of the neuropeptide vasopressin, a sex difference that is programmed
developmentally by androgens. If a particular DNA methyl-binding protein called
MeCP2 is reduced during the critical period in males, the adult vasopressin level is
also reduced, indicating a neonatal epigenetic programming of adult gene expression
(Forbes-Lorman et al. 2012).
Given the central role of steroid receptors in both developmental and adult
sensitivity to gonadal hormones, it is reasonable to also ask if the genes coding for
estrogen receptors alpha and beta or progesterone receptor are also epigenetically
406 M. M. McCarthy

Fig. 15.4 Epigenetic programming of masculinization. Upper, males. The process of masculini-
zation of brain and behavior occurs weeks, months to decades prior to the expression of the
phenotype. One mechanism by which this early life programming is maintained is via epigenetic
modifications of the DNA in which methyl groups are added to cytosine nucleotides or via changes
to the surrounding chromatin, both of which impart long-term regulatory functions on gene
expression. In the developing POA elevated gonadal steroids in males inhibit the DNMT enzymes
that transfer a methyl group to cytosines, resulting in reduced DNA methylation compared to
females. Adult male mounting behavior is illustrated. Lower, females. The higher level of methyla-
tion of the female genome in the POA inhibits the gene expression profile essential for adult
expression of male sex behavior and may also provide the means by which the sensitive period is
closed. Adult female lordosis behavior and maternal behavior are illustrated

modified. There are studies of the promoter regions of all these receptors and there
are sex differences, but there is little agreement about the direction and magnitude of
the amount of methylation or, even in some cases, precisely which CpGs are
differentially methylated (McCarthy et al. 2017). As more studies are conducted, it
is becoming increasingly clear that the epigenetics of the brain truly are highly
malleable and that a sex difference can exist at one point in time in one brain region,
disappear and reappear at another time at different CpG’s within the same gene.
Moreover, some hormonally induced epigenetic changes do not appear until long
after hormone exposure. In what might be called an “epigenetic echo,” two indepen-
dent studies have found hormonally induced changes in DNA methylation that were
not apparent within days of steroid treatment in neonates but did appear in juveniles
or adults (Ghahramani et al. 2014; Nugent et al. 2010).
15 Origins of Sex Differentiation of Brain and Behavior 407

How this happens at a mechanistic level remains entirely unknown but highlights
one of the biggest challenges in our quest to understand epigenetic underpinnings of
sexual differentiation—our inability to repeatedly monitor changes in the same
animal. Many things that we seek to measure as neuroendocrinologists can be
measured repeatedly, for example, circulating corticosterone, or LH release, body
weight, even neuronal activity via indwelling electrodes. But epigenetic
modifications to the DNA and surrounding chromatin can only be assessed ex vivo.
Moreover, most techniques involve a mixture of cell types and thereby contribute
noise to the output and uncertainty to the interpretation. Lastly, the approaches with
the highest fidelity, such as deep sequencing, are not as widely available for most
laboratories as other techniques. Fortunately, all of these problems are improving in
the right direction and so while neuroepigenetics is still at the frontier of neurosci-
ence, it is a worthy frontier to explore.

15.9 Neuroinflammation as a Mediator of Sexual

Differentiation

Steroids drive masculinization of the developing brain, but how? Many investigators
have explored the mechanistic basis of sexual differentiation for many years, but it is
fair to say that no clear unifying theory emerged from the effort and a general
conclusion of “it’s complicated” pervaded the field. The majority of work focused on
neurotransmitter systems for the obvious reason that they are central to the control of
behavior. Disruption of a particular neurotransmitter system will often impair normal
masculinization, but there was no documented incidence of being able to induce
masculinization in the absence of steroids. This stalemate was ended with the
discovery that the key masculinizing agent in at least one case—the dendritic
synaptic profile of the POA and adult mating behavior—was not a neurotransmitter
at all, but instead the membrane derived inflammatory signaling molecule, prosta-
glandin E2, also called PGE2 (Amateau and McCarthy 2004). Prostaglandins are
synthesized from arachidonic acid in the cell membrane by the cyclooxygenase
enzymes, COX1 and COX2. Males have higher levels of both enzymes in the
POA, leading to higher PGE2 production, which binds to the EP2 and EP4 receptors,
G-protein coupled receptors that activate adenylate cyclase and cAMP production.
Subsequent activation of protein kinase A (PKA) and phosphorylation of key
subunits of glutamate receptors induces clustering of glutamate receptors at the
membrane and the formation and stabilization of dendritic spines (Lenz et al.
2011). Many cells throughout the body are capable of synthesizing prostaglandins,
but of particular note are microglia, the brain’s innate immune system. These
macrophage-lineage cells are resident in the brain and they tile throughout, meaning
they spread themselves out so that each cell occupies a limited domain that does not
overlap with other microglia but the entire brain is in contact with at least one
microglial cell. They also respond to and produce prostaglandins, and, do so more in
males than females when located within the POA. Indeed, the microglia of the POA
are essential for normal masculinization to occur. If they are either depleted or
408 M. M. McCarthy

inhibited, the male brain develops without the capacity for male sexual behavior
(Lenz et al. 2013; VanRyzin et al. 2018).
The discovery that an innate immune cell is essential for normal masculinization
of brain and behavior was a surprise. But this surprise was surpassed with the
additional discovery that another innate immune cell, the mast cell, which is derived
from precursors made in the bone marrow, is rather the true mastermind behind the
immune-driven process of masculinization (Lenz et al. 2018). Mast cells are not
ubiquitous throughout the brain, instead being clustered in the meninges, and located
deep in the neuropil of the POA. Importantly, there are more in the male POA than in
the female. Remarkably, these cells number only in the tens, but when stimulated to
release histamine they signal to neighboring microglia which then increase produc-
tion of prostaglandins and the entire cascade leading to masculinization of brain and
behavior is initiated (Fig. 15.5). If the mast cells are prevented from releasing
histamine, the train is stopped and the brain dysmasculinized. Blocking the degran-
ulation of mast cells requires pharmacological treatment. Evidence for a more
nuanced role for mast cells came from inducing an allergic reaction in pregnant
dams and observing that the mast cells of the POA in the offspring were more
activated. Moreover, in adulthood the females were shifted towards a masculinized
behavioral profile while the male offspring were less masculinized (Lenz et al.
2019). In both cases the behavior of the males and females was within the normal
range, if that range includes both sexes, suggesting this is a potential source of
natural variation in behavior. In summary, there are multiple cell types involved in
the process of masculinizing the POA, and at least two of them are non-neuronal
cells; microglia and mast cells.
Microglia have proven to be equally important to another sexually differentiated
behavior, juvenile social play, but involving a completely different mechanism. Play
in rodents, sometimes called “rough-and-tumble” play is a complex social behavior
that may involve dyads of the same or different sex, or could involve large groups of
six or more animals (Argue and McCarthy 2015). The medial amygdala has long
been known as a key brain region in the neural circuitry of play and appears to be the
critical node for driving a sex difference in which males play more intensely and
frequently than females. How this occurs, however, was unknown until recently
when it was discovered that the number of astrocytes in the medial amygdala is
tightly controlled and is different between males and females, there being more in
females. This is not because more astrocytes are born in the female amygdala,
however, instead it is because the microglia of this brain region in males phagocy-
tose recently born astrocytes, thereby eliminating a portion of the population before
it has time to mature (VanRyzin et al. 2019). In this instance, there is no role for
prostaglandins or histamine but instead another class of membrane derived signaling
molecules, the endocannabinoids. These short-lived local signaling molecules are
found at a higher concentration in the developing amygdala of males and promote
phagocytosis by microglia. The end result is a neural circuitry that is more conducive
to rough-and-tumble play in males.
15 Origins of Sex Differentiation of Brain and Behavior 409

Fig. 15.5 Inflammatory signaling molecules and masculinization. In the preoptic area (POA),
sexual differentiation of synaptic density begins with hormonal regulation (by testosterone, T, or
estradiol, E2) of gene expression for the cyclooxygenase enzymes (COX-2) which make prosta-
glandin E2 (PGE2), an inflammatory signaling molecule. However additional PGE2 from the
microglia (brown cells), the brain’s own immune system, is also required for the full masculiniza-
tion process. The production of PGE2 by microglia is initiated by histamine release from resident
mast cells which are more numerous in males and thereby drive higher PGE2 production which
induces glutamate release from neighboring astrocytes. The glutamate binds to AMPA receptors
which have been induced to cluster at the membrane surface by the actions of protein kinase A
which is activated by PGE2 binding to G-protein coupled receptors on neurons. The activation of
the AMPA receptors causes the formation of dendritic spine synapses. Thus at least four cell types
are involved in the masculinization of the preoptic area, and two of them are of non-neuronal origin
but instead are components of the immune system

15.10 Perspectives

Sex differences are established during the process of sexual differentiation which
occurs weeks, months and in some cases years before the physiology and behavior
are manifest. Epigenetic programming both establishes and maintains sex
differences but also provides for reversibility across the life span. Sex differences
come in a variety of forms and understanding the nature of any particular one is
essential to proper interpretation and experimental design. The mechanisms of
sexual differentiation are diverse but often include inflammatory signaling molecules
and immune cells. The discovery that the immune profile of portions of the
410 M. M. McCarthy

developing male brain is elevated, or, more activated, compared to the female, is still
highly novel but is leading to new vistas of exploration. Re-analyses of published
transcriptomic data in human fetal cortex implicated greater expression of inflam-
matory pathways in males, and higher still in adult autistic males (Werling et al.
2016). This has led to speculation of a causal connection between the elevated
inflammatory state of the developing male brain and the higher risk for
neurodevelopmental psychiatric disorders such as autism spectrum disorders,
schizophrenia and attention deficit and hyperactivity disorders (McCarthy 2018).
Importantly, the researchers conducting the transcriptomic analysis in humans were
greatly aided in their ability to interpret their findings as genuine in light of the basic
science research done in animals. This is an important way in which so-called “blue
sky” research can have an impact.
We have likely seen just the tip of the proverbial iceberg of sex differences in
adult brain and behavior that have origins in development. Identifying the cellular
and molecular means by which early differences are established is essential to
understanding how they manifest in adulthood. Attention to developmental pro-
cesses as they are occurring is essential to elucidating the mechanisms by which sex
differences are established and ultimately maintained.

Key References

Amateau and McCarthy (2004)—This was the first report in which masculinization
of a female brain was achieved by a downstream signaling molecule from steroid
action, as opposed to treating directly with steroids. This opened a new era in
understanding of the novel mechanisms by which steroids can impact the brain,
including through the enhancement of immune system signaling.
Beery and Zucker (2010)—An in depth analyses of a large number of publications
across a wide swatch of biomedical research revealed the over reliance on male
animals, especially in neuroscience. Almost six times as many studies used
exclusively male rats and mice. Subsequent to this publication the NIH mandated
the incorporation of sex as a biological variable in all preclinical research.
De Vries (2004)—This minireview proposed the novel idea that sometimes the two
sexes evolve mechanisms to try and be more similar, not more different. Using
the example of the biparental vole, DeVries highlights how males have evolved a
new neural circuit that promotes parenting and is active in the absence of the
hormones of pregnancy and parturition that are so important to females.
Nottebohm and Arnold (1976)—Arguably the first robust neuroanatomical sex
difference discovered in a vertebrate brain, this land mark publication led to
increased scrutiny of the mammalian brain and the discovery of multiple sex
differences in the size of various brain regions and subnuclei.
Nugent et al. (2015)—One of a small number of studies that revealed sex differences
in the epigenome and in this case demonstrates functional significance for the
sexual differentiation of sexual behavior.
15 Origins of Sex Differentiation of Brain and Behavior 411

Werling et al. (2016)—Secondary analyses of the transcriptome from human fetal

postmortem cortical tissue found that more genes associated with inflammation,
activated astrocytes and microglia are up regulated in the male compared to the
female. Preclinical studies in animals were essential to interpretation of these
findings as indicative of normal healthy brain development in males.

References
Amateau SK, McCarthy MM (2004) Induction of PGE(2) by estradiol mediates developmental
masculinization of sex behavior. Nat Neurosci 7(6):643–650
Argue KJ, McCarthy MM (2015) Characterization of juvenile play in rats: importance of sex of self
and sex of partner. Biol Sex Differ 6:16
Bakker J, Baum MJ (2008) Role for estradiol in female-typical brain and behavioral sexual
differentiation. Front Neuroendocrinol 29(1):1–16. https://doi.org/10.1016/j.yfrne.2007.06.001
Barraclough CA, Gorski RA (1961) Evidence that the hypothalamus is responsible for androgen-
induced sterility in the female rat. Endocrinology 68:68–79
Beery AK, Zucker I (2010) Sex bias in neuroscience and biomedical research. Neurosci Biobehav
Rev 35:565–572. https://doi.org/10.1016/j.neubiorev.2010.07.002
Clayton JA, Collins FS (2014) Policy: NIH to balance sex in cell and animal studies. Nature 509
(7500):282–283
De Vries GJ (2004) Minireview: sex differences in adult and developing brains: compensation,
compensation, compensation. Endocrinology 145(3):1063–1068
Forbes-Lorman RM, Rautio JJ, Kurian JR, Auger AP, Auger CJ (2012) Neonatal MeCP2 is
important for the organization of sex differences in vasopressin expression. Epigenetics 7
(3):230–238. https://doi.org/10.4161/epi.7.3.19265
Ghahramani NM, Ngun TC, Chen PY, Tian Y, Krishnan S, Muir S et al (2014) The effects of
perinatal testosterone exposure on the DNA methylome of the mouse brain are late-emerging.
Biol Sex Differ 5:8. https://doi.org/10.1186/2042-6410-5-8
Gould E, Woolley CS, Frankfurt M, McEwen BS (1990) Gonadal steroids regulate dendritic spine
density in hippocampal pyramidal cells in adulthood. Neuroscience 10:1286–1291
Huang GZ, Woolley CS (2012) Estradiol acutely suppresses inhibition in the hippocampus through
a sex-specific endocannabinoid and mGluR-dependent mechanism. Neuron 74(5):801–808.
https://doi.org/10.1016/j.neuron.2012.03.035
Hull EM, Dominguez JM (2007) Sexual behavior in male rodents. Horm Behav 52(1):45–55.
https://doi.org/10.1016/j.yhbeh.2007.03.030
Kuiri-Hanninen T, Sankilampi U, Dunkel L (2014) Activation of the hypothalamic-pituitary-
gonadal axis in infancy: minipuberty. Horm Res Paediatr 82(2):73–80. https://doi.org/10.
1159/000362414
Lenz KM, Wright CL, Martin RC, McCarthy MM (2011) Prostaglandin E regulates AMPA
receptor phosphorylation and promotes membrane insertion in preoptic area neurons and glia
during sexual differentiation. PLoS One 6(4):e18500. https://doi.org/10.1371/journal.pone.
0018500
Lenz KM, Nugent BM, Haliyur R, McCarthy MM (2013) Microglia are essential to masculinization
of brain and behavior. J Neurosci 33(7):2761–2772
Lenz KM, Pickett LA, Wright CL, Davis KT, Joshi A, McCarthy MM (2018) Mast cells in the
developing brain determine adult sexual behavior. J Neurosci 38(37):8044–8059. https://doi.
org/10.1523/JNEUROSCI.1176-18.2018
Lenz KM, Pickett LA, Wright CL, Galan A, McCarthy MM (2019) Prenatal allergen exposure
perturbs sexual differentiation and programs lifelong changes in adult social and sexual behav-
ior. Sci Rep 9:4837
412 M. M. McCarthy

Lomniczi A, Loche A, Castellano JM, Ronnekleiv OK, Bosch M, Kaidar G et al (2013) Epigenetic
control of female puberty. Nat Neurosci 16(3):281–289. https://doi.org/10.1038/nn.3319
McCarthy MM (2018) Sex differences in neuroimmunity as an inherent risk factor.
Neuropsychopharmacology 44(1):38-44. https://doi.org/10.1038/s41386-018-0138-1
McCarthy MM, Rissman EF (2014) Epigenetics of reproduction. In: Plant T, Zeleznik AJ (eds)
Knobil & Neill’s physiology of reproduction, vol 2, 4th edn. Academic, London, pp 2439–2501
McCarthy MM, Nugent BM, Lenz KM (2017) Neuroimmunology and neuroepigenetics in the
establishment of sex differences in the brain. Nat Rev Neurosci 18(8):471–484. https://doi.org/
10.1038/nrn.2017.61
McCarthy MM, Herold K, Stockman SL (2018) Fast, furious and enduring: sensitive versus critical
periods in sexual differentiation of the brain. Physiol Behav 187:13–19. https://doi.org/10.1016/
j.physbeh.2017.10.030
Melmed S (2011) The pituitary, 3rd edn. Academic, London
Nottebohm F, Arnold AP (1976) Sexual dimorphism in vocal control areas of the songbird brain.
Science 194:211–213
Nugent BM, Schwarz JM, McCarthy MM (2010) Hormonally mediated epigenetic changes to
steroid receptors in the developing brain: implications for sexual differentiation. Horm Behav.
https://doi.org/10.1016/j.yhbeh.2010.08.009
Nugent BM, Wright CL, Shetty AC, Hodes GE, Lenz KM, Mahurkar A et al (2015) Brain
feminization requires active repression of masculinization via DNA methylation. Nat Neurosci
18(5):690–697. https://doi.org/10.1038/nn.3988
Pfaff DW (1966) Morphological changes in the brains of adult male rats after neonatal castration.
J Endocrinol 36:415–416
Phoenix CH, Goy RW, Gerall AA, Young WC (1959) Organizing action of prenatally administered
testosterone proprionate on the tissues mediating mating behavior in the female guinea pig.
Endocrinol 65:369–382
Raisman G, Field PM (1971) Sexual dimorphism in the preoptic area of the rat. Science
173:731–733
Ryan BC, Vandenbergh JG (2002) Intrauterine position effects. Neurosci Biobehav Rev 26
(6):665–678
Schwarz JM, McCarthy MM (2008) Steroid-induced sexual differentiation of the brain: multiple
pathways, one goal. J Neurochem 105:1561–1572
Shors TJ (2006) Stressful experience and learning across the lifespan. Annu Rev Psychol 57:55–85
Smith CL, O’Malley BW (2004) Coregulator function: a key to understanding tissue specificity of
selective receptor modulators. Endocr Rev 25(1):45–71
Sweatt JD (2013) The emerging field of neuroepigenetics. Neuron 80(3):624–632. https://doi.org/
10.1016/j.neuron.2013.10.023
Szyf M, Weaver I, Meaney M (2007) Maternal care, the epigenome and phenotypic differences in
behavior. Reprod Toxicol 24(1):9–19. https://doi.org/10.1016/j.reprotox.2007.05.001
VanRyzin JW, Pickett LA, McCarthy MM (2018) Microglia: driving critical periods and sexual
differentiation of the brain. Dev Neurobiol 78(6):580–592. https://doi.org/10.1002/dneu.22569
VanRyzin JW, Marquardt AE, Argue KJ, Vecchiarelli HA, Ashton SE, Arambula SE, Hill MN,
McCarthy MM (2019) Microglial phagocytosis of newborn cells is induced by
endocannabinoids and sculpts sex differences in juvenile rat social play. Neuron 102:435–449
Wallen K, Baum MJ (2002) Masculinization and defeminization in altricial and precocial
mammals: comparative aspects of steroid hormone action. In: Pfaff D (ed) Hormones brain
and behavior, vol 5. Academic, London, pp 385–424
Weisz J, Ward IL (1980) Plasma testosterone and progesterone titers of pregnant rats, their male and
female fetuses and neonatal offspring. Endocrinology 106:306–313
Werling DM, Parikshak NN, Geschwind DH (2016) Gene expression in human brain implicates
sexually dimorphic pathways in autism spectrum disorders. Nat Commun 7:10717. https://doi.
org/10.1038/ncomms10717
Wright CL, Burks SR, McCarthy MM (2008) Identification of prostaglandin E2 receptors
mediating perinatal masculinization of adult sex behavior and neuroanatomical correlates.
Dev Neurobiol 68:1406–1419
Development and Modulation of Female
Reproductive Function by Circadian Signals 16
Neta Gotlieb, Jacob Moeller, and Lance J. Kriegsfeld

Abstract
Female reproductive success requires multifaceted, temporally coordinated hor-
mone secretion. The circadian timing system is central to this complex neuroen-
docrine regulation, with circadian disruption associated with irregular ovulatory
cycles, reduced fertility, increased miscarriage rates, and anomalous fetal devel-
opment. Because living in the modern world is associated with relatively chronic
circadian disruption due to limited sunlight exposure during the day and exposure
to artificial light at night, this issue is of broad concern. The master mammalian
circadian pacemaker in the suprachiasmatic nucleus communicates monosynap-
tically to neuroendocrine cells in the brain to mediate the hormonal events
required for ovulation and pregnancy maintenance. Likewise, maternal hormones
cross the placenta to drive rhythmic fetal processes critical for typical develop-
ment. The present overview describes the means by which the circadian timing
system integrates with the reproductive axis to regulate female reproductive
functioning. Likewise, the negative consequences of circadian disruptions on
female reproductive health, and the mechanisms underlying these deleterious
outcomes, are considered.

N. Gotlieb
Department of Psychology, University of California, Berkeley, Berkeley, CA, USA
J. Moeller
Graduate Group in Endocrinology, University of California, Berkeley, Berkeley, CA, USA
L. J. Kriegsfeld (*)
Department of Psychology, University of California, Berkeley, Berkeley, CA, USA
Department of Integrative Biology, University of California, Berkeley,, Berkeley, CA, USA
Graduate Group in Endocrinology, University of California, Berkeley, Berkeley, CA, USA
The Helen Wills Neuroscience Institute, University of California, Berkeley, Berkeley, CA, USA
e-mail: [email protected]

# Springer Nature Switzerland AG 2020 413

S. Wray, S. Blackshaw (eds.), Developmental Neuroendocrinology, Masterclass in
Neuroendocrinology 9, https://doi.org/10.1007/978-3-030-40002-6_16
414 N. Gotlieb et al.

Keywords
Biological rhythms · Circadian · Hormone · Reproduction

16.1 Introduction

Coordinated timing of neuroendocrine events is fundamental for successful female

reproduction across mammalian species, including humans (Mahoney 2010;
Williams and Kriegsfeld 2012; Kriegsfeld et al. 2018; Boden et al. 2013; Gamble
et al. 2013; Simonneaux et al. 2017). Each phase of the female reproductive cycle,
from ovulation, pregnancy, and fetal development, to parturition, requires specifi-
cally timed patterns of hormone secretion regulated by the circadian system
(McEwen et al. 1987; Blaustein et al. 1994; Kriegsfeld et al. 2002; Egli et al.
2004; Mong and Pfaff 2003). Disruptions to circadian timing have marked, negative
consequences for female reproductive health. For example, women with irregular
work schedules or frequent travel across time zones experience abnormal menstrual
cycles (Lawson et al. 2011; Wang et al. 2016), reduced fertility (Ahlborg et al. 1996;
Fernandez et al. 2016), and increased miscarriage rates (Nurminen 1998; Lawson
et al. 2012; Fernandez et al. 2016). Analogously, in rodents, ablation of the master
circadian clock in the brain, blocking relevant clock output signals, or disrupting the
genes driving circadian clock function at the cellular level, lead to pronounced
deficits in ovulation and reproductive success (Miller et al. 2004; Nunez and Stephan
1977; Wiegand and Terasawa 1982; van der Horst et al. 1999; Chu et al. 2013).
The circadian system confers a selective advantage by allowing organisms to
anticipate rhythmic and predictable environmental change and adjust their physiol-
ogy and behavior accordingly. Although circadian rhythms are endogenously
generated, to effectively synchronize internal timing with the external environment,
exposure to sunlight during the day and darkness at night entrains (synchronizes)
these rhythms to environmental time (Hughes et al. 2015). Unfortunately, a major
consequence of contemporary lifestyles and technological advancements omnipres-
ent in the modern world is increased exposure to sun-free environments during the
day and artificial lighting at night. This combination results in an incongruence
between the endogenous circadian timing system and the external environment,
leading to chronic and pervasive “jet lag” in the modern world (Bedrosian and
Nelson 2013; Fonken and Nelson 2014). Such concerns have attracted the attention
of the medical community, with the American Medical Association adopting a
policy statement on the dangers of light at night for health and reproductive
functioning (Stevens et al. 2013).
Because the majority of studies to date have concentrated on the role of circadian
rhythms in female reproductive function, the present chapter focuses on this sex.
Although, it is worth noting that circadian disruption negatively affects semen
quality and sperm numbers (Cagnacci et al. 1999; Xie et al. 2018) and compromises
fertility in men (Deng et al. 2018) with analogous results seen in mouse models in
which the genes regulating the circadian clockwork are knocked out (Alvarez et al.
16 Development and Modulation of Female Reproductive Function by Circadian. . . 415

2008; Schoeller et al. 2016). Herein, we consider all phases of the female reproduc-
tive cycle, including ovulation, mating, pregnancy and fetal development, and
parturition and describe how the circadian system integrates with the reproductive
axis to mediate reproductive success. Likewise, we consider development in relevant
systems and circuits where applicable and the negative consequences of circadian
disruption during each stage of reproduction.

16.2 The Circadian Timing System

The circadian system consists of a master brain clock in the suprachiasmatic nucleus
(SCN) of the anterior hypothalamus that is synchronized to environmental time via a
direct retinal pathway. As indicated previously, although circadian rhythms are
endogenously generated, to be adaptive for an organism and anticipate environmen-
tal changes across the day, these endogenous cycles are entrained to environmental
time. Light entrains the SCN via retinal communication from rods and cones that, in
turn, target specialized, intrinsically photosensitive retinal ganglion cells containing
the photopigment, melanopsin (Hattar et al. 2002; Berson et al. 2002; Lucas et al.
2003; Panda et al. 2002). Successively, the SCN uses neural, diffusible, and auto-
nomic communication to convey timing information to the whole organism (Butler
et al. 2010; Buijs et al. 2013).
Circadian rhythms are a cell-autonomous property, with circadian rhythms being
generated by an autoregulatory transcription–translation feedback loop consisting of
clock genes and their protein products (Takahashi 2015; Honma 2018). The core
feedback loop begins in the morning with the clock protein, CLOCK, binding to
BMAL1 to drive the transcription of the Period (Per1 and Per2) and Cryptochrome
(Cry1 and Cry2) genes. Over the course of the day, Per and Cry transcripts are
translated into their respective proteins that inevitably feed back to the cell nucleus to
repress CLOCK: BMAL1-mediated transcription until the next morning when
transcription resumes. Whereas the Clock gene is constitutively expressed, an
additional feedback loop driven by the CLOCK: BMAL1 complex regulates
Bmal1 transcription through repression by Rev-erb/ and transcriptional activation
via retinoic acid receptors (RORs) (Fig. 16.1). Circadian timekeeping is a ubiquitous
property of cells throughout the brain and body, with virtually all cells exhibiting
circadian timekeeping (Honma 2018). Even when the SCN is isolated in culture, the
master clock maintains indefinite circadian rhythms at the tissue level due to unique
coupling among independent oscillators. In contrast, in the absence of master clock
communication or other entraining stimuli, extra-SCN brain loci and peripheral
organs exhibit loss of rhythmicity after several cycles (Yamazaki et al. 2000; Abe
et al. 2002). This loss of rhythmicity in extra-SCN systems results from loss of
coupling among cellular oscillators having slightly different periods (Welsh et al.
2004).
416 N. Gotlieb et al.

Fig. 16.1 The molecular clockwork. A simplified model of the intracellular mechanisms respon-
sible for mammalian circadian rhythm generation. The process begins when CLOCK and BMAL1
proteins dimerize to drive the transcription of the Per (Per1 and Per2) and Cry (Cry1 and Cry2)
genes. Throughout the day, PER and CRY proteins rise within the cell cytoplasm. When levels of
PER and CRY reach a threshold, they form heterodimers, feed back to the cell nucleus, and
negatively regulate CLOCK: BMAL1-mediated transcription of their own genes. Levels of PER are
regulated by casein kinase 1 epsilon (CK1ε) which phosphorylates these proteins and marks them
for degradation, thereby appropriately delaying negative feedback. Whereas Clock is constitutively
expressed, a secondary feedback loop drives the transcription of ROR and Rev-Erbα that, in turn,
induce rhythms in Bmal1 transcription through stimulatory and inhibitory actions on ROR response
elements (RRE) in the Bmal1 promotor, respectively. Clock-controlled genes are tissue-specific
genes that are produced rhythmically by the CLOCK:BMAL1 complex but are not part of the
clockwork mechanism (i.e., do not feed back onto the clockwork)

16.3 The Hypothalamic–Pituitary–Gonadal Axis (HPG)

At the pinnacle of the HPG axis, hypothalamic gonadotropin-releasing hormone

(GnRH) neurons send axonal projections to the median eminence. Here, GnRH is
released from nerve terminals into the portal vasculature to trigger the secretion of
the gonadotropins, follicle stimulating hormone (FSH) and luteinizing hormone
(LH), from the anterior pituitary. In turn, LH and FSH act on the gonads to stimulate
16 Development and Modulation of Female Reproductive Function by Circadian. . . 417

sex steroid (e.g., progesterone, estradiol, and testosterone) synthesis and secretion
and gametogenesis, respectively. Sex steroids and gonadotropins feed back to the
hypothalamus and anterior pituitary to regulate HPG axis activity. Upstream of the
GnRH system, two key neuropeptides were identified around the turn of the millen-
nium that have pronounced inhibitory and stimulatory actions on GnRH neurons,
namely RFamide-related peptide 3 and kisspeptin, respectively. The discovery and
significance of these neuropeptides are described below.

16.3.1 RFamide-Related Peptide 3

In a search for novel peptides in the Arg-Phe-NH2 (i.e., R¼Arginine,

F¼Phenylalanine; termed RFamide peptides) family, Tsutsui and colleagues discov-
ered a novel RFamide peptide that inhibited gonadotropin release from cultured quail
pituitaries (Tsutsui et al. 2000). Because this hypothalamic neuropeptide specifically
inhibited the gonadotropins without affecting other pituitary hormones, they named
this peptide gonadotropin-inhibitory hormone (GnIH) (Tsutsui et al. 2000). In birds,
the GnIH precursor encodes one GnIH and two GnIH-related peptides (GnIH-RP-1
and GnIH-RP-2) (Satake et al. 2001; Tsutsui et al. 2009). In mammals, the neuropep-
tide precursor cDNA encodes three peptides (known as RFamide-related peptides
[RFRPs]), RFRP-1, -2 and -3), with RFRP-3 thought to be the ortholog of avian GnIH
(Tsutsui et al. 2009). Studies across mammalian species have found pronounced roles
for this neuropeptide in regulating reproductive function.
The receptor for GnIH/RFRP-3 is a G-protein coupled receptor (GPR), the
formerly orphaned GPR147 [also called NPFF1 (Bonini et al. (2000)]. GPR147
is most commonly coupled to an inhibitory G-protein (Gαi), with GnIH/RFRP3
suppressing adenylate cyclase activity (Hinuma et al. 2000; Shimizu and Bedecarrats
2010). However, in some instances, GPR147 is coupled to Gαs or Gαq proteins,
where this differential coupling may account for disparity in the effects of RFRP3
based on sex or reproductive status [described further below (Gouarderes et al.
2007)].
In most rodent species, RFRP3-cell bodies are localized exclusively to the
dorsomedial hypothalamus (DMH; reviewed in (Henningsen et al. 2017; Kriegsfeld
et al. 2018; Ubuka and Tsutsui 2018). In mammals, RFRP-3-immunoreactive (ir)
neuronal fibers are widely distributed in the diencephalon, mesencephalon, and
limbic structures (Smith et al. 2008; Kriegsfeld et al. 2006; Yano et al. 2003),
providing a direct synaptic mechanism for broadly affecting neurophysiology and
behavior. Across mammalian species, including humans, RFRP-3 generally
suppresses GnRH and gonadotropin secretion via direct actions on GnRH cells and
potentially at the level of the pituitary (Clarke et al. 2008; Johnson et al. 2007;
Kriegsfeld et al. 2006; George et al. 2017; Ducret et al. 2009; Wu et al. 2009);
reviewed in Ubuka and Tsutsui (2018).
Under some circumstances, RFRP-3 stimulates the reproductive axis. For exam-
ple, male Siberian hamsters (Phodopus sungorus) that typically breed during long,
summer-like days exhibit an increase in LH concentrations following central
418 N. Gotlieb et al.

administration of RFRP-3 when maintained in short-day conditions (Ubuka et al.

2012). Additionally, male Syrian hamsters maintained in long days exhibit elevated
LH and FSH in response to RFRP-3 administration (Ancel et al. 2012). Likewise, in
striped hamsters (Cricetulus barabensis), the relationship between RFRP-3 and
GnRH expression depends on sex and developmental/reproductive status (Zhao
et al. 2014). Finally, RFRP3 stimulates LH secretion in male mice but inhibits LH
release in female mice when estradiol concentrations are high at the time of the LH
surge, but has no effect during diestrus or in ovariectomized females with low
estradiol concentrations provided exogenously (Ancel et al. 2017). These findings
suggest that the effects of RFRP-3 on gonadotropins may depend on the species,
season, sex, and developmental status. Likewise, these findings suggest that
GPR147 couples differentially with Gαi, Gαs, or Gαq based on these same variables
(Gouarderes et al. 2007).

16.3.2 Kisspeptin

Kisspeptin is also a member of the RFamide family of peptides. Kisspeptin was

initially known as metastin, a tumor metastasis suppressor gene in human melanoma
and breast carcinoma, discovered in 1996 in Hershey, PA (Lee et al. 1996; Lee and
Welch 1997; Kotani et al. 2001). The gene for metastin was named Kiss1, with the
“SS” representing suppressor sequence and “Ki” added as an homage for the home
town’s famous Hershey kisses (Lee et al. 1996). The role of kisspeptin as a positive
regulator of the HPG axis was discovered in 2003 in individuals exhibiting
hypophysiotropic hypogonadism. These individuals were found to have a mutation
in G-protein coupled receptor (GPR54), the cognate receptor for Kiss1 gene protein
products (de Roux et al. 2003; Seminara et al. 2003). These individuals exhibit low
or absent circulating LH, fail to undergo puberty and are reproductively incompetent
as adults. When the authors created GPR54-deficient mice, these animals exhibited
an analogous phenotype (de Roux et al. 2003; Seminara et al. 2003). The protein
product of the Kiss1 gene is now commonly referred to as kisspeptin.
The Kiss1 gene encodes a family of kisspeptin peptides, beginning with a
precursor polypeptide of 145 amino acids, that is cleaved into a 54 amino acid
protein (named kisspeptin-54) that can be further cleaved into 10, 13, and 14 amino
acid proteins, all of which share an RFamide sequence on their C-terminus, are
biologically active, and share a similar affinity to GPR54 (Kotani et al. 2001; Muir
et al. 2001; Ohtaki et al. 2001).
Kisspeptin directly stimulates the secretion of GnRH and LH across mammalian
species, including humans, via direct actions on GnRH neurons (Trevisan et al. 2018;
Saedi et al. 2018; Yeo and Colledge 2018). In women with hypothalamic amenor-
rhea, acute administration of kisspeptin stimulates gonadotropin release (Jayasena
et al. 2009, 2010), indicating that failure of kisspeptin signaling contributes to
reproductive cessation is these women. Both pituitary cells and GnRH neurons
express GPR54; however, whether kisspeptin directly stimulates gonadotropin
release from the pituitary requires further investigation. For example, although
16 Development and Modulation of Female Reproductive Function by Circadian. . . 419

prominent expression of GPR54 is detected in human pituitary (Muir et al. 2001;

Ohtaki et al. 2001), and peripheral administration of kisspeptin increases plasma LH
concentrations in some cases (Thompson et al. 2004), the effects of kisspeptin on
cultured pituitary cells are equivocal (Matsui et al. 2004; Thompson et al. 2004;
Navarro et al. 2005a, b).
Kisspeptin cell bodies are concentrated in the anteroventral periventricular
(AVPV) and the arcuate (Arc) nuclei of the hypothalamus in rodents (Smith et al.
2005a, b; Clarkson et al. 2008; Clarkson and Herbison 2006) and in the preoptic
area (POA) and infundibulum in humans (Hrabovszky et al. 2010). A large percent-
age of kisspeptin cells express estrogen receptor (ER)-α (Smith et al. 2005a),
thereby mediating positive (AVPV) and negative (Arc) feedback effects of estradiol
in females (Smith et al. 2005a, 2006; Adachi et al. 2007; Williams et al. 2011;
Mittelman-Smith et al. 2012).
A number of studies suggest that Arc kisspeptin cells serve as critical components
of estradiol negative feedback and the GnRH pulse generator. Arc kisspeptin neurons
co-express two additional peptides, neurokinin B (NKB) and dynorphin (Burke et al.
2006; Goodman et al. 2007, 2014; Navarro et al. 2009; Rance 2009; Wakabayashi
et al. 2010; Okamura et al. 2017). Together, this population is referred to as KNDy
(pronounced “candy”) neurons. These triple phenotype cells communicate in an
autosynaptic feedback manner in which KNDy cells project to other KNDy cells
(Foradori et al. 2002; Burke et al. 2006; Krajewski et al. 2010). Local secretion of
NKB stimulates kisspeptin release from KNDy cells onto GnRH neurons
(Wakabayashi et al. 2013). Mutations in the NKB gene (i.e., Tac3) or its receptor
(Tacr3) lead to hypogonadotropic hypogonadism (Guran et al. 2009; Topaloglu et al.
2009; Young et al. 2010). In contrast to NKB, dynorphin inhibits Arc kisspeptin
secretion and is thought to participate in the termination of each pulse (Goodman et al.
2013; Ruka et al. 2013).

16.4 Circadian Control of Ovulation

In species that ovulate spontaneously, the preovulatory LH surge that initiates

ovulation is ubiquitous [rodents (Levine and Ramirez 1982; Mahoney et al. 2004;
Moline et al. 1981; Sarkar et al. 1976), sheep (Moenter et al. 1992), rhesus macaques
(Pau et al. 1993), and women (Cahill et al. 1998; Edwards et al. 1980; Elkind-Hirsch
et al. 1984)]. The LH surge occurs in early morning in women (Kerdelhue et al.
2002) and diurnal rodents (Mahoney et al. 2004) but in early evening in nocturnal
rodents (Christian and Moenter 2010), coordinating the time of maximal fertility
with activity. During most of the rodent estrous cycle, gonadotropin concentrations
are low due to the negative feedback effects of gonadal steroids (reviewed in
Christian and Moenter 2010). At the time of ovulation, however, positive feedback
effects of estradiol serve a permissive role in initiating the LH surge by a timed,
neural signal from the SCN (i.e., positive feedback) (Karsch et al. 1997; Levine
1997; Levine and Ramirez 1982; Williams and Kriegsfeld 2012). This “switch” in
estradiol action results from the coordinated timing of positive and negative
420 N. Gotlieb et al.

regulators of the reproductive axis by the circadian timing system. The requirement
for both high estradiol concentrations and a circadian timing signal to initiate the LH
surge ensures appropriate oocyte maturation and maximal sexual motivation coin-
cide with the time of ovulation.
As indicated previously, lesions of the SCN or the severing of connections
between the SCN and the preoptic area results in acyclicity in rodents (Brown-
Grant and Raisman 1977; Wiegand et al. 1980; Wiegand and Terasawa 1982).
Furthermore, genetic impairments of the molecular clockwork also disrupt estrous
cycles and the LH surge in mice (Khan and Kauffman 2012; Chu et al. 2013;
Dolatshad et al. 2006; Miller et al. 2004). Although the SCN can support behavioral
rhythms via a diffusible signal, the SCN-derived signal that generates the LH surge
is neural; SCN-lesioned female hamsters receiving fetal SCN transplants that do not
form neural connections with the host brain exhibit rhythmic behavior but not an
LH surge (Lehman et al. 1987; Silver et al. 1996; Meyer-Bernstein et al. 1999). As
described further below, both direct SCN-GnRH neuronal communication and
indirect connections from the SCN to estrogen receptor-α (ERα)-expressing cell
phenotypes that positively and negatively regulate the GnRH system integrate
circadian and estradiol signaling to initiate the preovulatory LH surge and ovulation
(Fig. 16.2).

16.4.1 Monosynaptic SCN Regulation of the GnRH System in LH

Surge Generation

Two major neuropeptidergic SCN cell phenotypes have been implicated in

the generation of the preovulatory LH surge and ovulation in rodents, neurons
synthesizing the neuropeptide vasoactive intestinal peptide (VIP) in the
retinorecipient SCN core and vasopressin (AVP) in the dorsal SCN (Moore et al.
2002). Both VIP and AVP exhibit peak release around the time of the LH surge
(Schwartz et al. 1983; Kalsbeek et al. 1995; Francl et al. 2010) and central injections
of either VIP or AVP receptor antagonists attenuate LH surge amplitude (Funabashi
et al. 1999; Palm et al. 1999; Harney et al. 1996; Sun et al. 2012). SCN VIPergic
neurons project monosynaptically to GnRH neurons in the POA (Van der Beek et al.
1997; Horvath et al. 1998). Several lines of evidence suggest a role for this direct
pathway in LH surge generation. First, concomitant with the LH surge, FOS, is
preferentially expressed in GnRH neurons receiving VIPergic input on the afternoon
of proestrus (van der Beek et al. 1994). Furthermore, in vitro, VIP stimulates GnRH
secretion (Zhao and Kriegsfeld 2009; Samson et al. 1981; Ohtsuka et al. 1988), neural
firing (Christian and Moenter 2008; Piet et al. 2016), and intracellular calcium
release (Piet et al. 2016). Finally, VIPergic projections from the SCN to GnRH
cells are sexually dimorphic, emerging during the pubertal transition into adulthood
(Kriegsfeld et al. 2002), with female rats demonstrating higher VIPergic innervation
than males (Horvath et al. 1998). Notably, GnRH neurons do not express ERα, the
estrogen receptor mediating positive and negative feedback effects of estradiol. This
led researchers to search for estradiol-responsive systems upstream of the GnRH
16 Development and Modulation of Female Reproductive Function by Circadian. . . 421

Fig. 16.2 Circadian control of the preovulatory LH surge and ovulation. Model of interactions
among the circadian, RFRP-3, and kisspeptin systems in control of the preovulatory LH surge and
ovulation. At the time of the LH surge, the SCN coordinates estradiol positive feedback (through
kisspeptin cell activation) with removal of estradiol negative feedback (through suppression of
RFRP-3 cells in the DMH). GnRH, RFRP-3, and AVPV kisspeptin cells all exhibit rhythms in
clock protein or gene expression to further temporal precision in this circuit. See text for further
details

system that integrate circadian and estradiol signaling (Herbison and Theodosis 1992;
Hrabovszky et al. 2000; Wintermantel et al. 2006). As described below, RFRP-3 and
kisspeptin have emerged as key, estrogen-responsive cell phenotypes upstream of the
GnRH system that mediate circadian-controlled estrogen negative and positive
feedback, respectively.

16.4.2 Integration of Circadian and Estrogenic Signaling Upstream

of the GnRH System

Early findings pointed to the AVPV as a likely neural locus regulating estradiol
positive feedback and the LH surge. ERα expressing neurons the AVPV send
monosynaptic projections to GnRH neurons, express FOS at the time of the LH
surge, and AVPV lesions result in the loss of estrous cyclicity in intact and
ovariectomized/estradiol-treated rats (Wiegand and Terasawa 1982; Wiegand et al.
1980; Ronnekleiv and Kelly 1988; Gu and Simerly 1997; Le et al. 1999). However,
422 N. Gotlieb et al.

until relatively recently, the cell phenotype participating in estradiol positive feed-
back was unknown.
Following the discovery of kisspeptin, it was soon shown that ERα is expressed
in the majority of AVPV and Arc kisspeptin cells (Smith et al. 2005a; Dubois et al.
2015), with AVPV kisspeptin cells mediating estradiol positive feedback and the
Arc population negative feedback and GnRH pulsatility (Millar et al. 2010; Moore
et al. 2018). The SCN sends monosynaptic AVPergic projections to AVPV estrogen-
responsive kisspeptin cells (Fig. 16.2) that, in turn, project to kisspeptin-receptor-
expressing GnRH neurons in the POA to positively drive the LH surge (Vida et al.
2010; Chassard et al. 2015; Dubois et al. 2015; Piet et al. 2015; Simonneaux and
Bahougne 2015; Smarr et al. 2013; Smith et al. 2006; Williams et al. 2011; Yip et al.
2015). Significantly, the increased neural firing of kisspeptin cells in response to
AVP is estrogen dependent, consistent with a role of kisspeptin cells in integrating
estrogenic and circadian signaling (Piet et al. 2015).
In SCN-lesioned rats and Clock mutant mice, central administration of AVP
produces surge-like levels of LH (Palm et al. 1999; Miller et al. 2006). Importantly,
exogenous AVP only stimulates LH release when administered in the afternoon, the
timepoint when the preovulatory LH surge occurs in rodents (Palm et al. 2001).
Given this time-dependent sensitivity, we asked if time dependence of the surge was
due to daily changes in sensitivity of kisspeptin cells to AVP stimulation, daily
changes in GnRH cell sensitivity to kisspeptin stimulation, or a combination of both
processes. In Syrian hamsters, AVPV kisspeptin cells do not demonstrate time-
dependent changes in response to AVP stimulation, whereas GnRH neurons exhibit
daily changes in their responsiveness to kisspeptin, suggestive of autonomous
circadian timekeeping in cells downstream of the SCN (Williams et al. 2011).
Indeed, circadian oscillations in core clock genes that drive circadian rhythms at
the cellular level are found both in vitro and in vivo in GnRH cells (Chappell et al.
2003; Olcese et al. 2003; Zhao and Kriegsfeld 2009). Immortalized GnRH neurons
exhibit circadian rhythms in responsiveness to VIP and kisspeptin stimulation,
further indicating that the GnRH system maintains circadian timing potentially as
a mechanism mediating daily changes in responsiveness to upstream signaling (Zhao
and Kriegsfeld 2009). Despite not exhibiting daily changes to AVP stimulation,
kisspeptin cells express the clock gene Per1 and the AVPV exhibits sustained
circadian rhythms in clock gene expression in cultured AVPV explants (Jacobs
et al. 2016; Chassard et al. 2015). Whether these sustained rhythms in AVPV
kisspeptin cells confer daily changes in responsiveness to upstream neurochemicals
other than AVP remains to be determined.
A specific role for ARC kisspeptin cells in the generation of the preovulatory LH
surge has not been established. However, several findings suggest a role for Arc
kisspeptin cells in this process. For example, ablation of ARC kisspeptin cells results
in abnormal estrous cycles and LH surges, indicating a potential role in generating
the LH surge (Helena et al. 2015; Mittelman-Smith et al. 2016). Furthermore, this
population expresses receptors for both AVP and VIP (Lukas et al. 2010; Mounien
et al. 2006; Ronnekleiv et al. 2014), and AVP and VIP increase intracellular calcium
in subsets of Arc kisspeptin neurons in sexually dimorphic manner (Schafer et al.
2018).
16 Development and Modulation of Female Reproductive Function by Circadian. . . 423

As mentioned previously, prior to the LH surge and ovulation, estradiol acts via
negative feedback to maintain gonadotropins at low concentrations. Several lines of
evidence indicate a role for RFRP-3 cells in integrating circadian and estrogenic
signals to mediate estradiol negative feedback. First, RFRP-3 cells exhibit a high
activational state (as measured by FOS expression) during diestrus, reduced activity
around the time of the LH surge, and increased activity soon thereafter (Gibson et al.
2008). This pattern of timing is mediated by the SCN, with both AVP/VIPergic SCN
cell terminals forming close appositions with RFRP-3 cells in Syrian hamsters
(Gibson et al. 2008). Likewise, RFRP-3 neurons express ERα and increase their
activity in response to estradiol injections at the time of the LH surge (Kriegsfeld
et al. 2006; Molnar et al. 2011; Iwasa et al. 2012; Poling et al. 2012). Additionally,
analogous to effects seen for GnRH cells, VIP suppresses RFRP-3 cellular activity
around the time of the LH surge, but not prior to the surge, suggesting that removal
of RFPR-3-mediated estradiol negative feedback is accomplished via time-
dependent sensitivity to VIP signaling (Russo et al. 2015). This time-dependent
sensitivity is associated with rhythmic clock protein expression in RFRP-3 cells
(Russo et al. 2015). Additionally, RFRP-3 cell terminals form close appositions with
GnRH cell soma and terminals at median eminence, both of which express Gpr147
(Kriegsfeld et al. 2006; Johnson et al. 2007; Bentley et al. 2010; Kriegsfeld et al.
2010). The same impact of RFRP-3 is observed in both an estradiol surge implant
model (Gibson et al. 2008) and during the afternoon of proestrus (Henningsen et al.
2017). Finally, in mice, RFRP-3 similarly inhibits LH secretion when administered
at the time of the preovulatory LH surge, but not during diestrus (Ancel et al. 2017).
Together, these findings suggest that RFRP-3 cells integrate estrogen and circadian
signaling to time the removal of estradiol negative feedback with stimulation of the
LH surge.

16.5 Implications for Circadian Rhythms in Pregnancy Success

and Fetal Development

16.5.1 Circadian Timing and Pregnancy Maintenance

The reproductive system is exposed to an array of rhythmic hormonal secretions

during pregnancy (Magiakou et al. 1996; Tamura et al. 2008) and disruptions to
maternal timing negatively impact pregnancy success and fetal development. In one
early study using a chronic jet lag model in which mice were repeatedly subjected to
a 6 h phase advance or delay of the light:dark cycle, jet-lagged animals exhibited a
dramatic reduction in full-term pregnancy success (Summa et al. 2012). Likewise,
female mice maintained in 22 or 26 h light:dark cycles, cycles to which they cannot
entrain, exhibit decreased mating behavior and experience higher rates of fetal
resorption, reduced embryo weights, and delayed development (Endo and Watanabe
1989).
Prolactin, a key hormone regulating pregnancy maintenance, is under strict
circadian control in rodents. In mice and rats, for example, prolactin exhibits twice
424 N. Gotlieb et al.

daily surges during pregnancy (a diurnal and nocturnal surge) that maintains the
viability of the corpora lutea (CL) and the secretion of progesterone in the first half of
pregnancy. Around midpregnancy, these surges cease and placental lactogens main-
tain progesterone secretion for the remainder of gestation (Robertson and Friesen
1981; Bertram et al. 2010; Freeman et al. 1974; Shiu et al. 1973; Strauss et al. 1996).
The SCN regulates prolactin release through the pacing of inhibiting and stimulating
factors for this hormone, principally dopamine (DA) and oxytocin, respectively
(Bertram et al. 2010; Freeman et al. 2000; Mai et al. 1994). This regulation likely
occurs via VIPergic projections from the SCN, as both arcuate tuberoinfundibular
DA (TIDA) neurons and paraventricular nucleus (PVN) oxytocin neurons are
innervated by VIPergic fibers originating in the SCN (Arey and Freeman 1992;
Teclemariam-Mesbah et al. 1997; Gerhold et al. 2001; Egli et al. 2004; Poletini et al.
2010). In addition, VIP antisense oligonucleotides aimed at the SCN abolish prolac-
tin surges in rats (Harney et al. 1996). As would be expected given this mechanism
of control, SCN lesions abolish prolactin surges (Kawakami et al. 1980; Pan and
Gala 1985). Likewise, prolactin surges entrain to light:dark cycles and free run in
constant darkness (Bethea and Neill 1979; Yogev and Terkel 1980). Finally, knock-
down of essential clock genes (Per1, Per2, and Clock) in the SCN abolish prolactin
surges (Poletini et al. 2007), and mice lacking a functional Clock gene exhibit
reduced concentrations of progesterone and marked pregnancy failure, suggestive
of disrupted prolactin timing (Miller et al. 2004). These findings point to the critical
role of precisely timed prolactin secretion for the maintenance of pregnancy.
In addition to SCN regulation of TIDA and oxytocin neurons, kisspeptin cells
project to TIDA neurons and kisspeptin administration increases prolactin release
via dopaminergic cell inhibition (Szawka et al. 2010; Sawai et al. 2012, 2014; Saedi
et al. 2018). Furthermore, kisspeptin neurons express prolactin receptors, and
hyperprolactinemia causes reduced activity of hypothalamic kisspeptin cells,
indicating reciprocal communication between kisspeptin and TIDA neurons (Saedi
et al. 2018; Sonigo et al. 2012). As with TIDA and oxytocin regulation by the SCN,
arcuate kisspeptin cells express VIP receptor and VIP administration alters their
cellular activity (Schafer et al. 2018), suggesting that the SCN likely regulates TIDA
neurons directly and indirectly, via kisspeptin cells through VIPergic signaling.
Whether or not disruptions to kisspeptin cell timing affect prolactin release and
pregnancy success remains to be determined.
In contrast to rodents, the pituitary gland (or pituitary hormones) is not required
for the initiation and maintenance of pregnancy in humans. In the first 8 weeks of
pregnancy, the CL is maintained by human chorionic gonadotropin (hCG) produced
by the trophoblast (a layer of tissue that later forms part of the placenta), with
placental progesterone sufficient to maintain pregnancy thereafter (Malassine et al.
2003). Although humans do not exhibit a twice daily prolactin surge and, progester-
one secretion is maintained via the placenta, circadian disruptions profoundly reduce
pregnancy success as seen in rodents (Gamble et al. 2013; Bonzini et al. 2011;
Knutsson 2003; Lin et al. 2011). Given that the placenta is responsible for fetal-
maternal nutrient exchange and maintains rhythmic clock gene and hormonal
expression, it is likely that circadian disruption negatively influences pregnancy
16 Development and Modulation of Female Reproductive Function by Circadian. . . 425

success in humans via disruptions to this structure, a hypothesis that has not yet been
explored.

16.5.2 Molecular Clockwork and Pregnancy Success

Molecular clockwork plays a key role in regulating pregnancy success, with

anomalies in core clock genes resulting in marked female reproductive deficits.
For example, particular single nucleotide polymorphisms in the Bmal1 or Clock
genes are associated with a higher rate of miscarriage in women (Kovanen et al.
2010; Hodzic et al. 2018). In animal models, mice with mutations or genetic ablation
of clock genes (Clock, Bmal1, Per1, and Per2) exhibit deficiencies in implantation,
pregnancy maintenance, parturition, and have higher rates of fetal resorptions (i.e.,
miscarriage) (Pilorz and Steinlechner 2008; Alvarez et al. 2008; Boden et al. 2010;
Ratajczak et al. 2009; Miller and Takahashi 2013). These mouse models also exhibit
shorter duration prolactin surges or failure to initiate prolactin release following
mating and pregnancy (Poletini et al. 2010). Likewise, Clock mutant mice exhibit
reduced concentrations of progesterone, higher rates of fetal resorptions, and fewer
full-term pregnancies (Miller et al. 2004; Miller and Takahashi 2013; Csapo and
Wiest 1969), all likely due to altered prolactin secretion.
Notably, although clock gene knockout and mutant mice exhibit reduced fertility,
these animals lack the gene of interest systemically and permanently (i.e., clock gene
expression is absent in all tissues throughout development). Hence, it is not possible
to determine whether the reproductive deficits originate through circadian disruption
in specific reproductive tissues and/or during specific time points. For example, it is
possible that lacking essential clock genes throughout development leads to abnor-
mal establishment of circuits required for mating and pregnancy maintenance (i.e.,
organizational effects). It is equally possible that a dysfunctional cellular clockwork
results in reproductive deficits through abnormal timing in these same circuits in
adulthood (i.e., activational effects). Moreover, even when reproductive deficits do
not manifest in knockout mice, it is possible that compensatory mechanisms permit
typical mating and pregnancy success. Studies applying approaches that allow
temporal and spatial specificity of genetic manipulations would help to disambiguate
developmental and postpubertal roles of the circadian clockwork (e.g., Bmal1
knockdown specifically in the SCN following puberty).

16.5.3 Circadian Timing in Early in Fetal Development

The developing embryo is exposed to a circadian environment as early as its initial

migration from the oviduct into the uterus. The oviduct provides essential nutrients
and growth factors to the developing embryo, and although studies examining
oviduct rhythmicity are sparse, findings suggest that the oviducts regulate the
embryonic environment via the rhythmic expression of clock genes as well as
plasminogen activator inhibitor 1 (Kennaway et al. 2003; Johnson et al. 2002;
426 N. Gotlieb et al.

Hastings et al. 2007), a clock-controlled gene that is involved in oviduct activity

during peri-implantation and is thought to play a role in embryo protection (Kouba
et al. 2000). The uterus also exhibits rhythmic expression of clock genes, in both
nonpregnant and pregnant females, throughout different stages of pregnancy and
under constant conditions (Hastings et al. 2007; Johnson et al. 2002; Ratajczak et al.
2010, 2012; Resuehr et al. 2007; Nakamura et al. 2010; Akiyama et al. 2010).
Finally, another key structure in pregnancy maintenance and regulation, the pla-
centa, exhibits rhythmic expression of clock genes as well as glucocorticoid receptor
and glucocorticoid metabolic enzyme expression (Akiyama et al. 2010; Waddell
et al. 2012; Wharfe et al. 2011). The placenta is responsible for maternal–fetal
exchanges, secreting hormones to maintain gestation and promote the health of the
fetus. Taken together, the developing conceptus is exposed to a rhythmic environ-
ment from its first day, long before developing its own autonomous clock. As
suggested below, this rhythmic environment is likely critical for normal fetal and
postnatal development (Fig. 16.3).

16.5.4 Maternal–Fetal Rhythm Synchronization

Circadian disruptions that alter maternal endocrine timing signals can impair
maternal–fetal synchronization and fetal development (Fig. 16.3). A major maternal
signal providing photoperiodic information to the embryo is melatonin, a hor-
mone secreted from the maternal pineal gland at darkness. Melatonin crosses the
placenta and influences the embryo’s physiological rhythms and development
(Schenker et al. 1998; Williams et al. 1991; Okatani et al. 1998). Melatonin receptors
are widespread in the fetal nervous system (including the SCN), as well as in
peripheral organs, beginning early in fetal development (Drew et al. 1998; Thomas
et al. 2002; Rivkees and Reppert 1991; Vanecek 1988; Torres-Farfan et al. 2008).
Likewise, the placenta expresses melatonin receptors, and melatonin plays a role in
placenta development (Tamura et al. 2008). Nighttime melatonin secretion increases
throughout pregnancy (Nakamura et al. 2001), and melatonin rhythms in maternal
blood are mirrored in fetal circulation (Kennaway et al. 1996; Okatani et al. 1998).
In addition to melatonin, fetal rhythms are also entrained by maternal cortisol and
body temperature rhythms (Seron-Ferre et al. 2001), and possibly feeding times
(Novakova et al. 2010) (Fig. 16.3).
Because the pineal gland matures only after birth and the developing fetus and
newborn do not produce their own melatonin, maternal melatonin signaling is
conveyed to offspring through the placenta (in utero) and milk (after birth) and is
required for fetal/newborn rhythms. Thus, it is not surprising that maternal circadian
disruptions are reflected in fetal rhythms. In nonhuman primates, for example,
maternal light exposure disrupts rhythmic expression of fetal clock genes that can
be rescued by maternal melatonin administration (Torres-Farfan et al. 2006). Like-
wise, suppression of maternal melatonin by constant light exposure during preg-
nancy is associated with intrauterine growth restriction (IUGR), lower
concentrations and altered rhythms of cortisol, modified mRNA expression of
16 Development and Modulation of Female Reproductive Function by Circadian. . . 427

Fig. 16.3 Maternal–fetal rhythm synchronization. The developing fetus is exposed to an array of
time cues from its mother. The main signal providing rhythmic information is melatonin, secreted
from the mother’s pineal gland at darkness and crossing the placenta to the embryo. Additional
entraining signals include maternal cortisol, body temperature rhythms, and feeding times. In
addition, the developing embryo is exposed to a rhythmic environment via clock genes that are
rhythmically expressed in maternal reproductive tissues, including the oviduct, uterus, and placenta.
This rhythmic environment is likely critical for normal fetal and postnatal development. Circadian
disruptions that alter maternal endocrine timing signals can impair maternal–fetal synchronization
and fetal development. Clocks indicate rhythmic expression of clock genes

clock genes and clock-controlled genes in the fetal adrenal gland, and aberrant
adrenal response to adrenocorticotropic hormone (ACTH) in rats. Melatonin admin-
istration during the subjective night rescues all of the above (Mendez et al. 2012). In
in vitro fertilization (IVF) treatments, melatonin promotes embryo development
(Ishizuka et al. 2000; Tian et al. 2010; Sampaio et al. 2012). Likewise, melatonin
administration prior to IVF treatments and throughout pregnancy is associated with
improved pregnancy outcomes; fertility rates are 50% higher in melatonin treated
IVF cycles (Rizzo et al. 2010; Unfer et al. 2011). These findings raise questions
regarding the static nature of the environment that the conceptus is typically exposed
428 N. Gotlieb et al.

to in cultured IVF conditions and whether a rhythmic environment that better mimics
the in vivo milieu would increase IVF success rates. Likewise, the timing of IVF
embryo implantation may be important to treatment success, but this remains to be
determined. Nonetheless, the dependence of fetal/early postnatal rhythms on mater-
nal environment further underscores the importance of maintaining circadian health
during pregnancy and lactation.

16.5.5 Circadian Disruption and Fetal/Offspring Development

Because almost every aspect of female reproduction is regulated by the circadian

system, it is not surprising that circadian disruption is associated with a host of
negative reproductive outcomes. For example, shift work during pregnancy is
associated with preterm birth, low birth weight, small for gestational age births,
and increased risk of miscarriages (Gamble et al. 2013; Bonzini et al. 2011;
Knutsson 2003; Lin et al. 2011). Taken together with findings indicating the
prominence of circadian rhythms during pregnancy, these findings indicate that
environmental temporal disruption can perturb endogenous timing and have marked
negative consequences for developing progeny.
In animal models, the impact of chronic jetlag has been shown to negatively
impact a host of mental and physical health parameters in both adults and developing
offspring. For example, circadian disruption through twice weekly, 6 h advances of
the light:dark cycle, decreases hippocampal cell proliferation and neurogenesis by
>50% and leads to pronounced deficits in learning and memory in adult hamsters
(Gibson et al. 2010). Although not examined, it is likely that animals can recover
from these deficits after a period of time. However, in developing offspring, the
impact of circadian disruption appears to be permanent, with the impact of in utero
circadian disruption lasting until adulthood likely by modifying neural development.
For example, fetal hippocampal clock and clock-controlled gene rhythms are
suppressed by maternal exposure to constant light, leading to deficits in spatial
memory in these offspring as adults (Vilches et al. 2014). These adult deficits are
associated with dampened circadian rhythms of hippocampal clock and clock-
controlled genes. Intriguingly, these deficits can be rescued by providing pregnant
dams with rhythmic, exogenous melatonin (Vilches et al. 2014), pointing to a
potential for clinical interventions for women unable to escape circadian disruption
during pregnancy. Additionally, in utero and/or early life circadian disruption
impairs the ability of mice to elicit maternal care, even when crossed-fostered to
non-disrupted dams, and leads to adult deficits in social behavior and anxiety (Smarr
et al. 2017). Taken together, these findings indicate that circadian disruption
can have long lasting, likely permanent, impact on offspring neurobehavioral
development.
16 Development and Modulation of Female Reproductive Function by Circadian. . . 429

16.6 Circadian Control of Birth

The length of gestation is determined by multiple clocks, including a principal clock

monitoring fetal development and additional clock-regulated mechanisms governing
labor onset and parturition timing. Fetal membrane senescence is thought to initiate
the signaling cascades leading to parturition via inflammatory processes that increase
uterine sensitivity to uterotonins, including prostaglandins and oxytocin (reviewed
in Menon et al. 2016). For the sake of the present review, we focus on the circadian
regulation of parturition via maternal neuroendocrine pathways, although parturition
is a process regulated by multiple clocks at multiple levels (i.e., mother, fetus, and
placenta) (Fig. 16.3).
As different species adapted to specific temporal niches over the course of
evolution, a selective advantage was gained by initiating parturition during the
daytime or nighttime dependent of selective pressures. Because delivering offspring
in the home den is safer than in the open areas for prey species, many diurnal species
have evolved to initiate parturition at nighttime and nocturnal species in daytime.
Circadian timing of labor onset or parturition has been reported in mice, rats,
hamsters, sheep, pigs, horses, and primates, including humans (Cooperstock et al.
1987a, b; Ngwenya and Lindow 2004; Boer et al. 1975; Jensen and Bobbitt 1967;
Zahn and Hattensperger 1993; Germain et al. 1993; Moore et al. 1994; Takayama
et al. 2003; Olcese 2012; Olcese et al. 2013; Chaney et al. 2018; Reppert et al. 1987).
In rats, the length of pregnancy is influenced by the light regimen in which animals
are maintained, with longer day lengths leading to later parturition (i.e., gestational
day 23 instead of 22) (Bosc and Nicolle 1980). Clock genes are likely involved in
timing parturition, as mice lacking a functional molecular clockwork fail to either
enter labor or have prolonged and nonproductive parturition, resulting in resorption
of the fully developed embryos (Miller et al. 2004). Furthermore, as described
previously, Clock mutant mice experience higher rates of fetal resorptions and
fewer pregnancies reaching term.
In humans, labor onset and parturition tend to cluster between the late night and
early morning, exhibiting a trough in the late afternoon (Kaiser and Halberg 1962;
Olcese 2012; Olcese et al. 2013; Gamble et al. 2013). A similar circadian pattern has
been reported in the timing of parturition among indigenous populations living in
rural areas (Chaney et al. 2018) and in the onset of preterm labors (Vatish et al.
2010). Underlying the circadian rhythm of labor and birth may be a rhythm in uterine
myometrial activity, with contractions peaking at nighttime (Germain et al. 1993;
Moore et al. 1994; Zahn and Hattensperger 1993). Moreover, the nocturnal surge in
uterine activity in the last trimester of gestation can predict preterm deliveries, with
women who deliver prematurely losing these nocturnal surges weeks before birth
(Germain et al. 1993). Together, these findings suggest that the mechanisms under-
lying parturition are under circadian regulation across species.
The species-specific difference in circadian phase of parturition likely results
from antiphase uterine cell responses to endocrine signals. In humans, uterine
contractions exhibit a peak at night and a trough in the morning (Zahn and
Hattensperger 1993; Germain et al. 1993). This pattern may result from motility-
430 N. Gotlieb et al.

enhancing factors, such as oxytocin, estrogens, and prostaglandins, being higher at

nighttime (Zahn and Hattensperger 1993; Seron-Ferre et al. 1993), and from the
actions of nighttime melatonin secretion from the pineal gland (Tamura et al. 2008;
Kivela et al. 1989). In a proof of principle study, late-term pregnant women exposed
to bright light at night exhibited suppression in plasma melatonin concentrations that
was associated with a reduction in contraction intensity (Olcese et al. 2013), likely
resulting in the potentiation of oxytocin’s actions on the uterus (Sharkey et al. 2009).
Whether nighttime melatonin potentiates the actions of Pitocin (synthetic oxytocin)
in women induced into labor, potentially achieving a faster and safer process, has not
been explored.
In contrast to humans, melatonin exhibits a negative (tocolytic) effect on
oxytocin-induced uterine contractility in nocturnal rodents (Dominguez Rubio
et al. 2014; Takayama et al. 2003; Olcese 2012; Reppert et al. 1987). Thus, although
oxytocin stimulates uterine contractions in both diurnal and nocturnal species, and
melatonin peaks at night in both cases, this hormone exhibits opposite effects on the
myometrium. Rats in which the source of melatonin is eliminated via pinealectomy
deliver their pups independent of time of day. However, when melatonin is
administered to pinealectomized females at the onset of the dark phase, the circadian
rhythm of parturition is rescued. If melatonin is administered at the onset of the light
phase or in a constant release manner (i.e., via capsules), the circadian pattern of
deliveries is not regained (Takayama et al. 2003). This finding suggests that melato-
nin signaling is required at the appropriate time to initiate birth, likely due to
interactions with other regulators of this complex process (e.g., oxytocin).

16.7 Conclusions and Considerations

The circadian timing system participates in all aspects of female reproduction, from
ovulation to childbirth. Disruptions to circadian timing have marked negative
consequences for ovulation, pregnancy success and maintenance, and offspring
development. The modern world makes circadian disruption virtually inescapable,
underscoring the importance of developing safe and effective strategies to maximize
circadian health in the face of pervasive circadian disruption. Likewise, educating
couples about the importance of circadian health represents an important consider-
ation for health professionals providing counseling during family planning.

Key References

de Roux et al. (2003)—The authors investigated individuals with hypophysiotropic

hypogonadism and mice lacking the GPR54 gene, establishing a critical role for
this G-protein coupled receptor, the cognate receptor for Kiss1 gene protein
products (i.e., kisspeptin), as critical for reproductive function.
16 Development and Modulation of Female Reproductive Function by Circadian. . . 431

Endo and Watanabe (1989)—This article describes the effects of non-24 h days on
female reproduction in mice. The authors found that mice maintained on a 22 or
26 h light:dark cycle exhibit decreased mating behavior and experience higher
rates of fetal resorption, reduced embryo weights, and delayed development.
Kriegsfeld et al. (2006)—Based on the discovery of gonadotropin-inhibitory hor-
mone (GnIH) in birds, the authors provided the first, broad characterization of
GnIH (also known as RFamide-related peptide-3) in mammals (i.e., rats, mice
and hamsters), establishing a more general modulatory role for this neuropeptide
across species.
Mendez et al. (2012)—This article demonstrated a role for melatonin in fetal
development and rhythm generation/synchronization. The authors established
that exposure of pregnant rats to constant light during the second half of gestation
negatively affects fetal development, with embryos exhibiting intrauterine growth
retardation and disrupted adrenal clock genes and clock-controlled genes as well
as altered corticosterone rhythms. However, when mothers received melatonin
during their subjective night, reproductive function was rescued.
Miller et al. (2006)—This article examined reproduction of Clock mutant mice,
establishing an important role for circadian rhythms in female reproductive
function. The authors found that Clock mutant mice exhibit disruptions in estrous
cyclicity and pregnancy maintenance, including higher rates of fetal resorptions
and fewer pregnancies reaching full term.
Seminara et al. (2003)—This article employed complementary approaches in
humans and mice to investigate the onset of puberty. The authors found that
mutations in the G-protein coupled receptor, GPR54, the cognate receptor
kisspeptin, caused idiopathic hypophysiotropic hypogonadism in both mice and
humans. This finding was key to identifying a critical role for kisspeptin in
reproductive functioning.
Takayama et al. (2003)—This article demonstrated a role for melatonin in timing
parturition. The authors found that female rats lacking melatonin deliver pups
throughout the day and night, unlike melatonin-proficient rats who enter parturi-
tion during the day. Melatonin administration to rats exclusively during the dark
phase restored parturition to daytime.
Tsutsui et al. (2000)—This article identified a hypothalamic peptide that inhibits
gonadotropin release in a vertebrate pituitary and named it gonadotropin-
inhibitory hormone (GnIH). This finding represented a discovery with broad
implications for understanding reproductive axis regulation.

References
Abe M, Herzog ED, Yamazaki S, Straume M, Tei H, Sakaki Y, Menaker M, Block GD (2002)
Circadian rhythms in isolated brain regions. J Neurosci 22(1):350–356
Adachi S, Yamada S, Takatsu Y, Matsui H, Kinoshita M, Takase K, Sugiura H, Ohtaki T,
Matsumoto H, Uenoyama Y, Tsukamura H, Inoue K, Maeda K (2007) Involvement of
anteroventral periventricular metastin/kisspeptin neurons in estrogen positive feedback action
on luteinizing hormone release in female rats. J Reprod Dev 53(2):367–378
432 N. Gotlieb et al.

Ahlborg G Jr, Axelsson G, Bodin L (1996) Shift work, nitrous oxide exposure and subfertility
among Swedish midwives. Int J Epidemiol 25(4):783–790
Akiyama S, Ohta H, Watanabe S, Moriya T, Hariu A, Nakahata N, Chisaka H, Matsuda T,
Kimura Y, Tsuchiya S, Tei H, Okamura K, Yaegashi N (2010) The uterus sustains stable
biological clock during pregnancy. Tohoku J Exp Med 221(4):287–298
Alvarez JD, Hansen A, Ord T, Bebas P, Chappell PE, Giebultowicz JM, Williams C, Moss S,
Sehgal A (2008) The circadian clock protein BMAL1 is necessary for fertility and proper
testosterone production in mice. J Biol Rhythm 23(1):26–36. https://doi.org/10.1177/
0748730407311254
Ancel C, Bentsen AH, Sebert ME, Tena-Sempere M, Mikkelsen JD, Simonneaux V (2012) Stimu-
latory effect of RFRP-3 on the gonadotrophic axis in the male Syrian hamster: the exception
proves the rule. Endocrinology 153(3):1352–1363. https://doi.org/10.1210/en.2011-1622
Ancel C, Inglis MA, Anderson GM (2017) Central RFRP-3 stimulates LH secretion in male mice
and has cycle stage-dependent inhibitory effects in females. Endocrinology 158(9):2873–2883.
https://doi.org/10.1210/en.2016-1902
Arey BJ, Freeman ME (1992) Activity of vasoactive intestinal peptide and serotonin in the
paraventricular nucleus reflects the periodicity of the endogenous stimulatory rhythm regulating
prolactin secretion. Endocrinology 131(2):736–742. https://doi.org/10.1210/endo.131.2.
1639019
Bedrosian TA, Nelson RJ (2013) Influence of the modern light environment on mood. Mol
Psychiatry 18(7):751–757. https://doi.org/10.1038/mp.2013.70mp201370
Bentley GE, Tsutsui K, Kriegsfeld LJ (2010) Recent studies of gonadotropin-inhibitory hormone
(GnIH) in the mammalian hypothalamus, pituitary and gonads. Brain Res 1364:62–71. https://
doi.org/10.1016/j.brainres.2010.10.001
Berson DM, Dunn FA, Takao M (2002) Phototransduction by retinal ganglion cells that set the
circadian clock. Science 295(5557):1070–1073. https://doi.org/10.1126/science.1067262
Bertram R, Helena CV, Gonzalez-Iglesias AE, Tabak J, Freeman ME (2010) A tale of two rhythms:
the emerging roles of oxytocin in rhythmic prolactin release. J Neuroendocrinol 22(7):778–784.
https://doi.org/10.1111/j.1365-2826.2010.02012.x
Bethea CL, Neill JD (1979) Prolactin secretion after cervical stimulation of rats maintained in
constant dark or constant light. Endocrinology 104(4):870–876. https://doi.org/10.1210/endo-
104-4-870
Blaustein JD, Tetel MJ, Ricciardi KH, Delville Y, Turcotte JC (1994) Hypothalamic ovarian steroid
hormone-sensitive neurons involved in female sexual behavior. Psychoneuroendocrinology 19
(5–7):505–516
Boden MJ, Varcoe TJ, Voultsios A, Kennaway DJ (2010) Reproductive biology of female Bmal1
null mice. Reproduction 139(6):1077–1090. https://doi.org/10.1530/REP-09-0523
Boden MJ, Varcoe TJ, Kennaway DJ (2013) Circadian regulation of reproduction: from gamete to
offspring. Prog Biophys Mol Biol 113(3):387–397. https://doi.org/10.1016/j.pbiomolbio.2013.
01.003
Boer K, Lincoln DW, Swaab DF (1975) Effects of electrical stimulation of the neurohypophysis on
labour in the rat. J Endocrinol 65(2):163–176
Bonini JA, Jones KA, Adham N, Forray C, Artymyshyn R, Durkin MM, Smith KE, Tamm JA,
Boteju LW, Lakhlani PP, Raddatz R, Yao WJ, Ogozalek KL, Boyle N, Kouranova EV, Quan Y,
Vaysse PJ, Wetzel JM, Branchek TA, Gerald C, Borowsky B (2000) Identification and
characterization of two G protein-coupled receptors for neuropeptide FF. J Biol Chem 275
(50):39324–39331. https://doi.org/10.1074/jbc.M004385200
Bonzini M, Palmer KT, Coggon D, Carugno M, Cromi A, Ferrario MM (2011) Shift work and
pregnancy outcomes: a systematic review with meta-analysis of currently available epidemio-
logical studies. BJOG 118(12):1429–1437. https://doi.org/10.1111/j.1471-0528.2011.03066.x
Bosc MJ, Nicolle A (1980) Influence of photoperiod on the time of parturition in the rat. I. Effect
of the length of daily illumination on normal or adrenalectomized animals. Reprod Nutr Dev
20(3A):735–745
16 Development and Modulation of Female Reproductive Function by Circadian. . . 433

Brown-Grant K, Raisman G (1977) Abnormalities in reproductive function associated with the

destruction of the suprachiasmatic nuclei in female rats. Proc R Soc Lond B Biol Sci 198
(1132):279–296
Buijs RM, Escobar C, Swaab DF (2013) The circadian system and the balance of the autonomic
nervous system. Handb Clin Neurol 117:173–191. https://doi.org/10.1016/B978-0-444-53491-
0.00015-8
Burke MC, Letts PA, Krajewski SJ, Rance NE (2006) Coexpression of dynorphin and neurokinin B
immunoreactivity in the rat hypothalamus: morphologic evidence of interrelated function within
the arcuate nucleus. J Comp Neurol 498(5):712–726. https://doi.org/10.1002/cne.21086
Butler MP, Kriegsfeld LJ, Silver R (2010) Circadian regulation of endocrine functions. In: Pfaff D,
Etgen A (eds) Hormones, brain and behavior, 2nd edn. Academic, New York, pp 473–507
Cagnacci A, Maxia N, Volpe A (1999) Diurnal variation of semen quality in human males. Hum
Reprod 14(1):106–109
Cahill DJ, Wardle PG, Harlow CR, Hull MG (1998) Onset of the preovulatory luteinizing hormone
surge: diurnal timing and critical follicular prerequisites. Fertil Steril 70(1):56–59
Chaney C, Goetz LG, Valeggia C (2018) A time to be born: variation in the hour of birth in a rural
population of Northern Argentina. Am J Phys Anthropol 166(4):975–978. https://doi.org/10.
1002/ajpa.23483
Chappell PE, White RS, Mellon PL (2003) Circadian gene expression regulates pulsatile
gonadotropin-releasing hormone (GnRH) secretory patterns in the hypothalamic GnRH-
secreting GT1-7 cell line. J Neurosci 23(35):11202–11213
Chassard D, Bur I, Poirel VJ, Mendoza J, Simonneaux V (2015) Evidence for a putative circadian
Kiss-clock in the hypothalamic AVPV in female mice. Endocrinology 156(8):2999–3011.
https://doi.org/10.1210/en.2014-1769
Christian CA, Moenter SM (2008) Vasoactive intestinal polypeptide can excite gonadotropin-
releasing hormone neurons in a manner dependent on estradiol and gated by time of day.
Endocrinology 149(6):3130–3136. https://doi.org/10.1210/en.2007-1098
Christian CA, Moenter SM (2010) The neurobiology of preovulatory and estradiol-induced gonadotro-
pin-releasing hormone surges. Endocr Rev 31(4):544–577. https://doi.org/10.1210/er.2009-0023
Chu A, Zhu L, Blum ID, Mai O, Leliavski A, Fahrenkrug J, Oster H, Boehm U, Storch KF (2013)
Global but not gonadotrope-specific disruption of Bmal1 abolishes the luteinizing hormone
surge without affecting ovulation. Endocrinology 154(8):2924–2935. https://doi.org/10.1210/
en.2013-1080
Clarke IJ, Sari IP, Qi Y, Smith JT, Parkington HC, Ubuka T, Iqbal J, Li Q, Tilbrook A, Morgan K,
Pawson AJ, Tsutsui K, Millar RP, Bentley GE (2008) Potent action of RFamide-related peptide-3 on
pituitary gonadotropes indicative of a hypophysiotropic role in the negative regulation of gonado-
tropin secretion. Endocrinology 149(11):5811–5821. https://doi.org/10.1210/en.2008-0575
Clarkson J, Herbison AE (2006) Postnatal development of kisspeptin neurons in mouse hypothala-
mus; sexual dimorphism and projections to gonadotropin-releasing hormone neurons. Endocri-
nology 147(12):5817–5825. https://doi.org/10.1210/en.2006-0787
Clarkson J, d’Anglemont de Tassigny X, Moreno AS, Colledge WH, Herbison AE (2008)
Kisspeptin-GPR54 signaling is essential for preovulatory gonadotropin-releasing hormone
neuron activation and the luteinizing hormone surge. J Neurosci 28(35):8691–8697. https://
doi.org/10.1523/JNEUROSCI.1775-08.2008
Cooperstock M, England JE, Wolfe RA (1987a) Circadian incidence of labor onset hour in preterm
birth and chorioamnionitis. Obstet Gynecol 70(6):852–855
Cooperstock M, England JE, Wolfe RA (1987b) Circadian incidence of premature rupture of the
membranes in term and preterm births. Obstet Gynecol 69(6):936–941
Csapo AI, Wiest WG (1969) An examination of the quantitative relationship between progesterone
and the maintenance of pregnancy. Endocrinology 85(4):735–746. https://doi.org/10.1210/
endo-85-4-735
de Roux N, Genin E, Carel JC, Matsuda F, Chaussain JL, Milgrom E (2003) Hypogonadotropic
hypogonadism due to loss of function of the KiSS1-derived peptide receptor GPR54. Proc Natl
Acad Sci USA 100(19):10972–10976. https://doi.org/10.1073/pnas.1834399100
434 N. Gotlieb et al.

Deng N, Kohn TP, Lipshultz LI, Pastuszak AW (2018) The relationship between shift work and
Men’s health. Sex Med Rev 6(3):446–456. https://doi.org/10.1016/j.sxmr.2017.11.009
Dolatshad H, Campbell EA, O’Hara L, Maywood ES, Hastings MH, Johnson MH (2006) Devel-
opmental and reproductive performance in circadian mutant mice. Hum Reprod 21(1):68–79.
https://doi.org/10.1093/humrep/dei313
Dominguez Rubio AP, Sordelli MS, Salazar AI, Aisemberg J, Bariani MV, Cella M, Rosenstein
RE, Franchi AM (2014) Melatonin prevents experimental preterm labor and increases offspring
survival. J Pineal Res 56(2):154–162. https://doi.org/10.1111/jpi.12108
Drew JE, Williams LM, Hannah LT, Barrett P, Abramovich DR (1998) Melatonin receptors in the
human fetal kidney: 2-[125I]iodomelatonin binding sites correlated with expression of Mel1a
and Mel1b receptor genes. J Endocrinol 156(2):261–267
Dubois SL, Acosta-Martinez M, DeJoseph MR, Wolfe A, Radovick S, Boehm U, Urban JH, Levine
JE (2015) Positive, but not negative feedback actions of estradiol in adult female mice require
estrogen receptor alpha in kisspeptin neurons. Endocrinology 156(3):1111–1120. https://doi.
org/10.1210/en.2014-1851
Ducret E, Anderson GM, Herbison AE (2009) RFamide-related peptide-3, a mammalian
gonadotropin-inhibitory hormone ortholog, regulates gonadotropin-releasing hormone neuron
firing in the mouse. Endocrinology 150(6):2799–2804. https://doi.org/10.1210/en.2008-1623
Edwards RG, Steptoe PC, Fowler RE, Baillie J (1980) Observations on preovulatory human ovarian
follicles and their aspirates. Br J Obstet Gynaecol 87(9):769–779
Egli M, Bertram R, Sellix MT, Freeman ME (2004) Rhythmic secretion of prolactin in rats: action
of oxytocin coordinated by vasoactive intestinal polypeptide of suprachiasmatic nucleus origin.
Endocrinology 145(7):3386–3394. https://doi.org/10.1210/en.2003-1710
Elkind-Hirsch K, Ravnikar V, Tulchinsky D, Schiff I, Ryan KJ (1984) Episodic secretory patterns
of immunoreactive luteinizing hormone-releasing hormone (IR-LH-RH) in the systemic circu-
lation of normal women throughout the menstrual cycle. Fertil Steril 41(1):56–61
Endo A, Watanabe T (1989) Effects of non-24-hour days on reproductive efficacy and embryonic
development in mice. Gamete Res 22(4):435–441. https://doi.org/10.1002/mrd.1120220409
Fernandez RC, Marino JL, Varcoe TJ, Davis S, Moran LJ, Rumbold AR, Brown HM, Whitrow MJ,
Davies MJ, Moore VM (2016) Fixed or rotating night shift work undertaken by women:
implications for fertility and miscarriage. Semin Reprod Med 34(2):74–82. https://doi.org/10.
1055/s-0036-1571354
Fonken LK, Nelson RJ (2014) The effects of light at night on circadian clocks and metabolism.
Endocr Rev 35, 648–670. https://doi.org/10.1210/er.2013-1051
Foradori CD, Coolen LM, Fitzgerald ME, Skinner DC, Goodman RL, Lehman MN (2002)
Colocalization of progesterone receptors in parvicellular dynorphin neurons of the ovine
preoptic area and hypothalamus. Endocrinology 143(11):4366–4374. https://doi.org/10.1210/
en.2002-220586
Francl JM, Kaur G, Glass JD (2010) Regulation of vasoactive intestinal polypeptide release in the
suprachiasmatic nucleus circadian clock. Neuroreport 21(16):1055–1059. https://doi.org/10.
1097/WNR.0b013e32833fcba4
Freeman ME, Smith MS, Nazian SJ, Neill JD (1974) Ovarian and hypothalamic control of the daily
surges of prolactin secretion during pseudopregnancy in the rat. Endocrinology 94(3):875–882.
https://doi.org/10.1210/endo-94-3-875
Freeman ME, Kanyicska B, Lerant A, Nagy G (2000) Prolactin: structure, function, and regulation
of secretion. Physiol Rev 80(4):1523–1631. https://doi.org/10.1152/physrev.2000.80.4.1523
Funabashi T, Aiba S, Sano A, Shinohara K, Kimura F (1999) Intracerebroventricular injection of
arginine-vasopressin V1 receptor antagonist attenuates the surge of luteinizing hormone and
prolactin secretion in proestrous rats. Neurosci Lett 260(1):37–40
Gamble KL, Resuehr D, Johnson CH (2013) Shift work and circadian dysregulation of reproduc-
tion. Front Endocrinol 4:92. https://doi.org/10.3389/fendo.2013.00092
George JT, Hendrikse M, Veldhuis JD, Clarke IJ, Anderson RA, Millar RP (2017) Effect of
gonadotropin-inhibitory hormone on luteinizing hormone secretion in humans. Clin Endocrinol
86(5):731–738. https://doi.org/10.1111/cen.13308
16 Development and Modulation of Female Reproductive Function by Circadian. . . 435

Gerhold LM, Horvath TL, Freeman ME (2001) Vasoactive intestinal peptide fibers innervate
neuroendocrine dopaminergic neurons. Brain Res 919(1):48–56
Germain AM, Valenzuela GJ, Ivankovic M, Ducsay CA, Gabella C, Seron-Ferre M (1993)
Relationship of circadian rhythms of uterine activity with term and preterm delivery. Am J
Obstet Gynecol 168(4):1271–1277
Gibson EM, Humber SA, Jain S, Williams WP 3rd, Zhao S, Bentley GE, Tsutsui K, Kriegsfeld LJ
(2008) Alterations in RFamide-related peptide expression are coordinated with the preovulatory
luteinizing hormone surge. Endocrinology 149(10):4958–4969. https://doi.org/10.1210/en.
2008-0316
Gibson EM, Wang C, Tjho S, Khattar N, Kriegsfeld LJ (2010) Experimental ‘jet lag’ inhibits adult
neurogenesis and produces long-term cognitive deficits in female hamsters. PLoS One 5(12):
e15267. https://doi.org/10.1371/journal.pone.0015267
Goodman RL, Lehman MN, Smith JT, Coolen LM, de Oliveira CV, Jafarzadehshirazi MR,
Pereira A, Iqbal J, Caraty A, Ciofi P, Clarke IJ (2007) Kisspeptin neurons in the arcuate nucleus
of the ewe express both dynorphin A and neurokinin B. Endocrinology 148(12):5752–5760.
https://doi.org/10.1210/en.2007-0961
Goodman RL, Hileman SM, Nestor CC, Porter KL, Connors JM, Hardy SL, Millar RP, Cernea M,
Coolen LM, Lehman MN (2013) Kisspeptin, neurokinin B, and dynorphin act in the arcuate
nucleus to control activity of the GnRH pulse generator in ewes. Endocrinology 154
(11):4259–4269. https://doi.org/10.1210/en.2013-1331
Goodman RL, Coolen LM, Lehman MN (2014) A role for neurokinin B in pulsatile GnRH
secretion in the ewe. Neuroendocrinology 99(1):18–32. https://doi.org/10.1159/000355285
Gouarderes C, Mazarguil H, Mollereau C, Chartrel N, Leprince J, Vaudry H, Zajac JM (2007)
Functional differences between NPFF1 and NPFF2 receptor coupling: high intrinsic activities of
RFamide-related peptides on stimulation of [35S]GTPgammaS binding. Neuropharmacology
52(2):376–386. https://doi.org/10.1016/j.neuropharm.2006.07.034
Gu GB, Simerly RB (1997) Projections of the sexually dimorphic anteroventral periventricular
nucleus in the female rat. J Comp Neurol 384(1):142–164
Guran T, Tolhurst G, Bereket A, Rocha N, Porter K, Turan S, Gribble FM, Kotan LD, Akcay T,
Atay Z, Canan H, Serin A, O’Rahilly S, Reimann F, Semple RK, Topaloglu AK (2009)
Hypogonadotropic hypogonadism due to a novel missense mutation in the first extracellular
loop of the neurokinin B receptor. J Clin Endocrinol Metab 94(10):3633–3639. https://doi.org/
10.1210/jc.2009-0551
Harney JP, Scarbrough K, Rosewell KL, Wise PM (1996) In vivo antisense antagonism of
vasoactive intestinal peptide in the suprachiasmatic nuclei causes aging-like changes in the
estradiol-induced luteinizing hormone and prolactin surges. Endocrinology 137(9):3696–3701.
https://doi.org/10.1210/endo.137.9.8756535
Hastings M, O’Neill JS, Maywood ES (2007) Circadian clocks: regulators of endocrine and
metabolic rhythms. J Endocrinol 195(2):187–198. https://doi.org/10.1677/JOE-07-0378
Hattar S, Liao HW, Takao M, Berson DM, Yau KW (2002) Melanopsin-containing retinal ganglion
cells: architecture, projections, and intrinsic photosensitivity. Science 295(5557):1065–1070.
https://doi.org/10.1126/science.1069609
Helena CV, Toporikova N, Kalil B, Stathopoulos AM, Pogrebna VV, Carolino RO, Anselmo-
Franci JA, Bertram R (2015) KNDy neurons modulate the magnitude of the steroid-induced
luteinizing hormone surges in ovariectomized rats. Endocrinology 156(11):4200–4213. https://
doi.org/10.1210/en.2015-1070
Henningsen JB, Ancel C, Mikkelsen JD, Gauer F, Simonneaux V (2017) Roles of RFRP-3 in the
daily and seasonal regulation of reproductive activity in female Syrian hamsters. Endocrinology
158(3):652–663. https://doi.org/10.1210/en.2016-1689
Herbison AE, Theodosis DT (1992) Localization of oestrogen receptors in preoptic neurons
containing neurotensin but not tyrosine hydroxylase, cholecystokinin or luteinizing hormone-
releasing hormone in the male and female rat. Neuroscience 50(2):283–298
436 N. Gotlieb et al.

Hinuma S, Shintani Y, Fukusumi S, Iijima N, Matsumoto Y, Hosoya M, Fujii R, Watanabe T,

Kikuchi K, Terao Y, Yano T, Yamamoto T, Kawamata Y, Habata Y, Asada M, Kitada C,
Kurokawa T, Onda H, Nishimura O, Tanaka M, Ibata Y, Fujino M (2000) New neuropeptides
containing carboxy-terminal RFamide and their receptor in mammals. Nat Cell Biol 2
(10):703–708. https://doi.org/10.1038/35036326
Hodzic A, Lavtar P, Ristanovic M, Novakovic I, Dotlic J, Peterlin B (2018) Genetic variation in the
CLOCK gene is associated with idiopathic recurrent spontaneous abortion. PLoS One 13(5):
e0196345. https://doi.org/10.1371/journal.pone.0196345
Honma S (2018) The mammalian circadian system: a hierarchical multi-oscillator structure for
generating circadian rhythm. J Physiol Sci 68(3):207–219. https://doi.org/10.1007/s12576-018-
0597-5
Horvath TL, Cela V, van der Beek EM (1998) Gender-specific apposition between vasoactive
intestinal peptide-containing axons and gonadotrophin-releasing hormone-producing neurons in
the rat. Brain Res 795(1–2):277–281
Hrabovszky E, Shughrue PJ, Merchenthaler I, Hajszan T, Carpenter CD, Liposits Z, Petersen SL
(2000) Detection of estrogen receptor-beta messenger ribonucleic acid and 125I-estrogen
binding sites in luteinizing hormone-releasing hormone neurons of the rat brain. Endocrinology
141(9):3506–3509. https://doi.org/10.1210/endo.141.9.7788
Hrabovszky E, Ciofi P, Vida B, Horvath MC, Keller E, Caraty A, Bloom SR, Ghatei MA, Dhillo
WS, Liposits Z, Kallo I (2010) The kisspeptin system of the human hypothalamus: sexual
dimorphism and relationship with gonadotropin-releasing hormone and neurokinin B neurons.
Eur J Neurosci 31(11):1984–1998. https://doi.org/10.1111/j.1460-9568.2010.07239.x
Hughes S, Jagannath A, Hankins MW, Foster RG, Peirson SN (2015) Photic regulation of clock
systems. Methods Enzymol 552:125–143. https://doi.org/10.1016/bs.mie.2014.10.018
Ishizuka B, Kuribayashi Y, Murai K, Amemiya A, Itoh MT (2000) The effect of melatonin on
in vitro fertilization and embryo development in mice. J Pineal Res 28(1):48–51
Iwasa T, Matsuzaki T, Murakami M, Kinouchi R, Osugi T, Gereltsetseg G, Yoshida S, Irahara M,
Tsutsui K (2012) Developmental changes in the mammalian gonadotropin-inhibitory hormone
(GnIH) ortholog RFamide-related peptide (RFRP) and its cognate receptor GPR147 in the rat
hypothalamus. Int J Dev Neurosci 30(1):31–37. https://doi.org/10.1016/j.ijdevneu.2011.10.003
Jacobs DC, Veitch RE, Chappell PE (2016) Evaluation of immortalized AVPV- and arcuate-
specific neuronal kisspeptin cell lines to elucidate potential mechanisms of estrogen
responsiveness and temporal gene expression in females. Endocrinology 157(9):3410–3419
Jayasena CN, Nijher GM, Chaudhri OB, Murphy KG, Ranger A, Lim A, Patel D, Mehta A, Todd C,
Ramachandran R, Salem V, Stamp GW, Donaldson M, Ghatei MA, Bloom SR, Dhillo WS
(2009) Subcutaneous injection of kisspeptin-54 acutely stimulates gonadotropin secretion in
women with hypothalamic amenorrhea, but chronic administration causes tachyphylaxis. J Clin
Endocrinol Metab 94(11):4315–4323. https://doi.org/10.1210/jc.2009-0406
Jayasena CN, Nijher GM, Abbara A, Murphy KG, Lim A, Patel D, Mehta A, Todd C,
Donaldson M, Trew GH, Ghatei MA, Bloom SR, Dhillo WS (2010) Twice-weekly administra-
tion of kisspeptin-54 for 8 weeks stimulates release of reproductive hormones in women with
hypothalamic amenorrhea. Clin Pharmacol Ther 88(6):840–847. https://doi.org/10.1038/clpt.
2010.204
Jensen GD, Bobbitt RA (1967) Changing parturition time in monkeys (Macaca nemestrina) from
night to day. Lab Anim Care 17(4):379–381
Johnson MH, Lim A, Fernando D, Day ML (2002) Circadian clockwork genes are expressed in the
reproductive tract and conceptus of the early pregnant mouse. Reprod Biomed Online 4
(2):140–145
Johnson MA, Tsutsui K, Fraley GS (2007) Rat RFamide-related peptide-3 stimulates GH secretion,
inhibits LH secretion, and has variable effects on sex behavior in the adult male rat. Horm Behav
51(1):171–180. https://doi.org/10.1016/j.yhbeh.2006.09.009
Kaiser IH, Halberg F (1962) Circadian periodic aspects of birth. Ann N Y Acad Sci 98:1056–1068
16 Development and Modulation of Female Reproductive Function by Circadian. . . 437

Kalsbeek A, Buijs RM, Engelmann M, Wotjak CT, Landgraf R (1995) In vivo measurement of a
diurnal variation in vasopressin release in the rat suprachiasmatic nucleus. Brain Res 682
(1–2):75–82
Karsch FJ, Bowen JM, Caraty A, Evans NP, Moenter SM (1997) Gonadotropin-releasing hormone
requirements for ovulation. Biol Reprod 56(2):303–309
Kawakami M, Arita J, Yoshioka E (1980) Loss of estrogen-induced daily surges of prolactin and
gonadotropins by suprachiasmatic nucleus lesions in ovariectomized rats. Endocrinology 106
(4):1087–1092. https://doi.org/10.1210/endo-106-4-1087
Kennaway DJ, Goble FC, Stamp GE (1996) Factors influencing the development of melatonin
rhythmicity in humans. J Clin Endocrinol Metab 81(4):1525–1532. https://doi.org/10.1210/
jcem.81.4.8636362
Kennaway DJ, Varcoe TJ, Mau VJ (2003) Rhythmic expression of clock and clock-controlled
genes in the rat oviduct. Mol Hum Reprod 9(9):503–507
Kerdelhue B, Brown S, Lenoir V, Queenan JT Jr, Jones GS, Scholler R, Jones HW Jr (2002) Timing
of initiation of the preovulatory luteinizing hormone surge and its relationship with the circadian
cortisol rhythm in the human. Neuroendocrinology 75(3):158–163
Khan AR, Kauffman AS (2012) The role of kisspeptin and RFamide-related peptide-3 neurones in
the circadian-timed preovulatory luteinising hormone surge. J Neuroendocrinol 24(1):131–143.
https://doi.org/10.1111/j.1365-2826.2011.02162.x
Kivela A, Kauppila A, Leppaluoto J, Vakkuri O (1989) Serum and amniotic fluid melatonin during
human labor. J Clin Endocrinol Metab 69(5):1065–1068. https://doi.org/10.1210/jcem-69-5-1065
Knutsson A (2003) Health disorders of shift workers. Occup Med (Lond) 53(2):103–108
Kotani M, Detheux M, Vandenbogaerde A, Communi D, Vanderwinden JM, Le Poul E,
Brezillon S, Tyldesley R, Suarez-Huerta N, Vandeput F, Blanpain C, Schiffmann SN,
Vassart G, Parmentier M (2001) The metastasis suppressor gene KiSS-1 encodes kisspeptins,
the natural ligands of the orphan G protein-coupled receptor GPR54. J Biol Chem 276
(37):34631–34636. https://doi.org/10.1074/jbc.M104847200
Kouba AJ, Burkhardt BR, Alvarez IM, Goodenow MM, Buhi WC (2000) Oviductal plasminogen
activator inhibitor-1 (PAI-1): mRNA, protein, and hormonal regulation during the estrous cycle
and early pregnancy in the pig. Mol Reprod Dev 56(3):378–386. https://doi.org/10.1002/1098-
2795(200007)56:3<378::AID-MRD8>3.0.CO;2-B
Kovanen L, Saarikoski ST, Aromaa A, Lonnqvist J, Partonen T (2010) ARNTL (BMAL1) and
NPAS2 gene variants contribute to fertility and seasonality. PLoS One 5(4):e10007. https://doi.
org/10.1371/journal.pone.0010007
Krajewski SJ, Burke MC, Anderson MJ, McMullen NT, Rance NE (2010) Forebrain projections of
arcuate neurokinin B neurons demonstrated by anterograde tract-tracing and monosodium
glutamate lesions in the rat. Neuroscience 166(2):680–697. https://doi.org/10.1016/j.neurosci
ence.2009.12.053
Kriegsfeld LJ, Silver R, Gore AC, Crews D (2002) Vasoactive intestinal polypeptide contacts on
gonadotropin-releasing hormone neurones increase following puberty in female rats. J
Neuroendocrinol 14(9):685–690
Kriegsfeld LJ, Mei DF, Bentley GE, Ubuka T, Mason AO, Inoue K, Ukena K, Tsutsui K, Silver R
(2006) Identification and characterization of a gonadotropin-inhibitory system in the brains of
mammals. Proc Natl Acad Sci USA 103(7):2410–2415. https://doi.org/10.1073/pnas.
0511003103
Kriegsfeld LJ, Gibson EM, Williams WP 3rd, Zhao S, Mason AO, Bentley GE, Tsutsui K (2010)
The roles of RFamide-related peptide-3 in mammalian reproductive function and behaviour. J
Neuroendocrinol 22(7):692–700. https://doi.org/10.1111/j.1365-2826.2010.02031.x
Kriegsfeld LJ, Jennings KJ, Bentley GE, Tsutsui K (2018) Gonadotrophin-inhibitory hormone and
its mammalian orthologue RFamide-related peptide-3: discovery and functional implications for
reproduction and stress. J Neuroendocrinol 30(7):e12597. https://doi.org/10.1111/jne.12597
Lawson CC, Whelan EA, Lividoti Hibert EN, Spiegelman D, Schernhammer ES, Rich-Edwards JW
(2011) Rotating shift work and menstrual cycle characteristics. Epidemiology 22(3):305–312.
https://doi.org/10.1097/EDE.0b013e3182130016
438 N. Gotlieb et al.

Lawson CC, Rocheleau CM, Whelan EA, Lividoti Hibert EN, Grajewski B, Spiegelman D, Rich-
Edwards JW (2012) Occupational exposures among nurses and risk of spontaneous abortion.
Am J Obstet Gynecol 206(4):327.e321–328. https://doi.org/10.1016/j.ajog.2011.12.030
Le WW, Berghorn KA, Rassnick S, Hoffman GE (1999) Periventricular preoptic area neurons
coactivated with luteinizing hormone (LH)-releasing hormone (LHRH) neurons at the time of
the LH surge are LHRH afferents. Endocrinology 140(1):510–519. https://doi.org/10.1210/
endo.140.1.6403
Lee JH, Welch DR (1997) Suppression of metastasis in human breast carcinoma MDA-MB-435
cells after transfection with the metastasis suppressor gene, KiSS-1. Cancer Res 57
(12):2384–2387
Lee JH, Miele ME, Hicks DJ, Phillips KK, Trent JM, Weissman BE, Welch DR (1996) KiSS-1, a
novel human malignant melanoma metastasis-suppressor gene. J Natl Cancer Inst 88
(23):1731–1737
Lehman MN, Silver R, Gladstone WR, Kahn RM, Gibson M, Bittman EL (1987) Circadian
rhythmicity restored by neural transplant. Immunocytochemical characterization of the graft
and its integration with the host brain. J Neurosci 7(6):1626–1638
Levine JE (1997) New concepts of the neuroendocrine regulation of gonadotropin surges in rats.
Biol Reprod 56(2):293–302
Levine JE, Ramirez VD (1982) Luteinizing hormone-releasing hormone release during the rat
estrous cycle and after ovariectomy, as estimated with push-pull cannulae. Endocrinology 111
(5):1439–1448
Lin YC, Chen MH, Hsieh CJ, Chen PC (2011) Effect of rotating shift work on childbearing and
birth weight: a study of women working in a semiconductor manufacturing factory. World J
Pediatr 7(2):129–135. https://doi.org/10.1007/s12519-011-0265-9
Lucas RJ, Hattar S, Takao M, Berson DM, Foster RG, Yau KW (2003) Diminished pupillary light
reflex at high irradiances in melanopsin-knockout mice. Science 299(5604):245–247. https://
doi.org/10.1126/science.1077293
Lukas M, Bredewold R, Neumann ID, Veenema AH (2010) Maternal separation interferes with
developmental changes in brain vasopressin and oxytocin receptor binding in male rats.
Neuropharmacology 58(1):78–87. https://doi.org/10.1016/j.neuropharm.2009.06.020
Magiakou MA, Mastorakos G, Rabin D, Margioris AN, Dubbert B, Calogero AE, Tsigos C,
Munson PJ, Chrousos GP (1996) The maternal hypothalamic-pituitary-adrenal axis in the
third trimester of human pregnancy. Clin Endocrinol 44(4):419–428
Mahoney MM (2010) Shift work, jet lag, and female reproduction. Int J Endocrinol 2010:813764.
https://doi.org/10.1155/2010/813764
Mahoney MM, Sisk C, Ross HE, Smale L (2004) Circadian regulation of gonadotropin-releasing
hormone neurons and the preovulatory surge in luteinizing hormone in the diurnal rodent,
Arvicanthis niloticus, and in a nocturnal rodent, Rattus norvegicus. Biol Reprod 70
(4):1049–1054. https://doi.org/10.1095/biolreprod.103.021360
Mai LM, Shieh KR, Pan JT (1994) Circadian changes of serum prolactin levels and
tuberoinfundibular dopaminergic neuron activities in ovariectomized rats treated with or with-
out estrogen: the role of the suprachiasmatic nuclei. Neuroendocrinology 60(5):520–526.
https://doi.org/10.1159/000126789
Malassine A, Frendo JL, Evain-Brion D (2003) A comparison of placental development and
endocrine functions between the human and mouse model. Hum Reprod Update 9(6):531–539
Matsui H, Takatsu Y, Kumano S, Matsumoto H, Ohtaki T (2004) Peripheral administration of
metastin induces marked gonadotropin release and ovulation in the rat. Biochem Biophys Res
Commun 320(2):383–388. https://doi.org/10.1016/j.bbrc.2004.05.185
McEwen BS, Jones KJ, Pfaff DW (1987) Hormonal control of sexual behavior in the female rat:
molecular, cellular and neurochemical studies. Biol Reprod 36(1):37–45
Mendez N, Abarzua-Catalan L, Vilches N, Galdames HA, Spichiger C, Richter HG, Valenzuela GJ,
Seron-Ferre M, Torres-Farfan C (2012) Timed maternal melatonin treatment reverses circadian
disruption of the fetal adrenal clock imposed by exposure to constant light. PLoS One 7(8):
e42713. https://doi.org/10.1371/journal.pone.0042713
16 Development and Modulation of Female Reproductive Function by Circadian. . . 439

Menon R, Bonney EA, Condon J, Mesiano S, Taylor RN (2016) Novel concepts on pregnancy
clocks and alarms: redundancy and synergy in human parturition. Hum Reprod Update 22
(5):535–560. https://doi.org/10.1093/humupd/dmw022
Meyer-Bernstein EL, Jetton AE, Matsumoto SI, Markuns JF, Lehman MN, Bittman EL (1999)
Effects of suprachiasmatic transplants on circadian rhythms of neuroendocrine function in
golden hamsters. Endocrinology 140(1):207–218. https://doi.org/10.1210/endo.140.1.6428
Millar RP, Roseweir AK, Tello JA, Anderson RA, George JT, Morgan K, Pawson AJ (2010)
Kisspeptin antagonists: unraveling the role of kisspeptin in reproductive physiology. Brain Res
1364:81–89. https://doi.org/10.1016/j.brainres.2010.09.044
Miller BH, Takahashi JS (2013) Central circadian control of female reproductive function. Front
Endocrinol 4:195. https://doi.org/10.3389/fendo.2013.00195
Miller BH, Olson SL, Turek FW, Levine JE, Horton TH, Takahashi JS (2004) Circadian clock
mutation disrupts estrous cyclicity and maintenance of pregnancy. Curr Biol 14(15):1367–1373.
https://doi.org/10.1016/j.cub.2004.07.055
Miller BH, Olson SL, Levine JE, Turek FW, Horton TH, Takahashi JS (2006) Vasopressin
regulation of the proestrous luteinizing hormone surge in wild-type and clock mutant mice.
Biol Reprod 75(5):778–784. https://doi.org/10.1095/biolreprod.106.052845
Mittelman-Smith MA, Williams H, Krajewski-Hall SJ, Lai J, Ciofi P, McMullen NT, Rance NE
(2012) Arcuate kisspeptin/neurokinin B/dynorphin (KNDy) neurons mediate the estrogen
suppression of gonadotropin secretion and body weight. Endocrinology 153(6):2800–2812.
https://doi.org/10.1210/en.2012-1045
Mittelman-Smith MA, Krajewski-Hall SJ, McMullen NT, Rance NE (2016) Ablation of KNDy
neurons results in hypogonadotropic hypogonadism and amplifies the steroid-induced LH surge
in female rats. Endocrinology 157(5):2015–2027. https://doi.org/10.1210/en.2015-1740
Moenter SM, Brand RC, Karsch FJ (1992) Dynamics of gonadotropin-releasing hormone (GnRH)
secretion during the GnRH surge: insights into the mechanism of GnRH surge induction.
Endocrinology 130(5):2978–2984
Moline ML, Albers HE, Todd RB, Moore-Ede MC (1981) Light-dark entrainment of proestrous LH
surges and circadian locomotor activity in female hamsters. Horm Behav 15(4):451–458
Molnar CS, Kallo I, Liposits Z, Hrabovszky E (2011) Estradiol down-regulates RF-amide-related
peptide (RFRP) expression in the mouse hypothalamus. Endocrinology 152(4):1684–1690.
https://doi.org/10.1210/en.2010-1418
Mong JA, Pfaff DW (2003) Hormonal and genetic influences underlying arousal as it drives sex and
aggression in animal and human brains. Neurobiol Aging 24(Suppl 1):S83–S88. discussion
S91-2
Moore TR, Iams JD, Creasy RK, Burau KD, Davidson AL (1994) Diurnal and gestational patterns
of uterine activity in normal human pregnancy. The Uterine Activity in Pregnancy Working
Group. Obstet Gynecol 83(4):517–523
Moore RY, Speh JC, Leak RK (2002) Suprachiasmatic nucleus organization. Cell Tissue Res 309
(1):89–98. https://doi.org/10.1007/s00441-002-0575-2
Moore AM, Lucas KA, Goodman RL, Coolen LM, Lehman MN (2018) Three-dimensional
imaging of KNDy neurons in the mammalian brain using optical tissue clearing and multiple-
label immunocytochemistry. Sci Rep 8(1):2242. https://doi.org/10.1038/s41598-018-20563-2
Mounien L, Bizet P, Boutelet I, Gourcerol G, Fournier A, Vaudry H, Jegou S (2006) Pituitary
adenylate cyclase-activating polypeptide directly modulates the activity of proopiomelanocortin
neurons in the rat arcuate nucleus. Neuroscience 143(1):155–163. https://doi.org/10.1016/j.
neuroscience.2006.07.022
Muir AI, Chamberlain L, Elshourbagy NA, Michalovich D, Moore DJ, Calamari A, Szekeres PG,
Sarau HM, Chambers JK, Murdock P, Steplewski K, Shabon U, Miller JE, Middleton SE,
Darker JG, Larminie CG, Wilson S, Bergsma DJ, Emson P, Faull R, Philpott KL, Harrison DC
(2001) AXOR12, a novel human G protein-coupled receptor, activated by the peptide KiSS-1. J
Biol Chem 276(31):28969–28975. https://doi.org/10.1074/jbc.M102743200
440 N. Gotlieb et al.

Nakamura Y, Tamura H, Kashida S, Takayama H, Yamagata Y, Karube A, Sugino N, Kato H

(2001) Changes of serum melatonin level and its relationship to feto-placental unit during
pregnancy. J Pineal Res 30(1):29–33
Nakamura TJ, Sellix MT, Kudo T, Nakao N, Yoshimura T, Ebihara S, Colwell CS, Block GD
(2010) Influence of the estrous cycle on clock gene expression in reproductive tissues: effects of
fluctuating ovarian steroid hormone levels. Steroids 75(3):203–212. https://doi.org/10.1016/j.
steroids.2010.01.007
Navarro VM, Castellano JM, Fernandez-Fernandez R, Tovar S, Roa J, Mayen A, Barreiro ML,
Casanueva FF, Aguilar E, Dieguez C, Pinilla L, Tena-Sempere M (2005a) Effects of KiSS-1
peptide, the natural ligand of GPR54, on follicle-stimulating hormone secretion in the rat.
Endocrinology 146(4):1689–1697. https://doi.org/10.1210/en.2004-1353
Navarro VM, Castellano JM, Fernandez-Fernandez R, Tovar S, Roa J, Mayen A, Nogueiras R, Vazquez
MJ, Barreiro ML, Magni P, Aguilar E, Dieguez C, Pinilla L, Tena-Sempere M (2005b) Characteri-
zation of the potent luteinizing hormone-releasing activity of KiSS-1 peptide, the natural ligand of
GPR54. Endocrinology 146(1):156–163. https://doi.org/10.1210/en.2004-0836
Navarro VM, Gottsch ML, Chavkin C, Okamura H, Clifton DK, Steiner RA (2009) Regulation of
gonadotropin-releasing hormone secretion by kisspeptin/dynorphin/neurokinin B neurons in the
arcuate nucleus of the mouse. J Neurosci 29(38):11859–11866. https://doi.org/10.1523/
JNEUROSCI.1569-09.2009
Ngwenya S, Lindow SW (2004) 24 hour rhythm in the timing of pre-labour spontaneous rupture of
membranes at term. Eur J Obstet Gynecol Reprod Biol 112(2):151–153
Novakova M, Sladek M, Sumova A (2010) Exposure of pregnant rats to restricted feeding schedule
synchronizes the SCN clocks of their fetuses under constant light but not under a light-dark
regime. J Biol Rhythm 25(5):350–360. https://doi.org/10.1177/0748730410377967
Nunez AA, Stephan FK (1977) The effects of hypothalamic knife cuts on drinking rhythms and the
estrus cycle of the rat. Behav Biol 20(2):224–234
Nurminen T (1998) Shift work and reproductive health. Scand J Work Environ Health 24(Suppl
3):28–34
Ohtaki T, Shintani Y, Honda S, Matsumoto H, Hori A, Kanehashi K, Terao Y, Kumano S,
Takatsu Y, Masuda Y, Ishibashi Y, Watanabe T, Asada M, Yamada T, Suenaga M, Kitada C,
Usuki S, Kurokawa T, Onda H, Nishimura O, Fujino M (2001) Metastasis suppressor gene
KiSS-1 encodes peptide ligand of a G-protein-coupled receptor. Nature 411(6837):613–617.
https://doi.org/10.1038/35079135
Ohtsuka S, Miyake A, Nishizaki T, Tasaka K, Tanizawa O (1988) Vasoactive intestinal peptide
stimulates gonadotropin-releasing hormone release from rat hypothalamus in vitro. Acta
Endocrinol (Copenh) 117(3):399–402
Okamura H, Yamamura T, Wakabayashi Y (2017) Mapping of KNDy neurons and immunohisto-
chemical analysis of the interaction between KNDy and substance P neural systems in goat. J
Reprod Dev 63(6):571–580. https://doi.org/10.1262/jrd.2017-103
Okatani Y, Okamoto K, Hayashi K, Wakatsuki A, Tamura S, Sagara Y (1998) Maternal-fetal
transfer of melatonin in pregnant women near term. J Pineal Res 25(3):129–134
Olcese J (2012) Circadian aspects of mammalian parturition: a review. Mol Cell Endocrinol 349
(1):62–67. https://doi.org/10.1016/j.mce.2011.06.041
Olcese J, Domagalski R, Bednorz A, Weaver DR, Urbanski HF, Reuss S, Middendorff R (2003)
Expression and regulation of mPer1 in immortalized GnRH neurons. Neuroreport 14
(4):613–618. https://doi.org/10.1097/01.wnr.0000063508.18654.3c
Olcese J, Lozier S, Paradise C (2013) Melatonin and the circadian timing of human parturition.
Reprod Sci 20(2):168–174. https://doi.org/10.1177/1933719112442244
Palm IF, Van Der Beek EM, Wiegant VM, Buijs RM, Kalsbeek A (1999) Vasopressin induces a
luteinizing hormone surge in ovariectomized, estradiol-treated rats with lesions of the
suprachiasmatic nucleus. Neuroscience 93(2):659–666
Palm IF, van der Beek EM, Wiegant VM, Buijs RM, Kalsbeek A (2001) The stimulatory effect of
vasopressin on the luteinizing hormone surge in ovariectomized, estradiol-treated rats is time-
dependent. Brain Res 901(1–2):109–116
16 Development and Modulation of Female Reproductive Function by Circadian. . . 441

Pan JT, Gala RR (1985) Central nervous system regions involved in the estrogen-induced afternoon
prolactin surge. II. Implantation studies. Endocrinology 117(1):388–395. https://doi.org/10.
1210/endo-117-1-388
Panda S, Sato TK, Castrucci AM, Rollag MD, DeGrip WJ, Hogenesch JB, Provencio I, Kay SA
(2002) Melanopsin (Opn4) requirement for normal light-induced circadian phase shifting.
Science 298(5601):2213–2216. https://doi.org/10.1126/science.1076848
Pau KY, Berria M, Hess DL, Spies HG (1993) Preovulatory gonadotropin-releasing hormone surge
in ovarian-intact rhesus macaques. Endocrinology 133(4):1650–1656
Piet R, Fraissenon A, Boehm U, Herbison AE (2015) Estrogen permits vasopressin signaling in
preoptic kisspeptin neurons in the female mouse. J Neurosci 35(17):6881–6892. https://doi.org/
10.1523/JNEUROSCI.4587-14.2015
Piet R, Dunckley H, Lee K, Herbison AE (2016) Vasoactive intestinal peptide excites GnRH
neurons in male and female mice. Endocrinology 157(9):3621–3630. https://doi.org/10.1210/
en.2016-1399
Pilorz V, Steinlechner S (2008) Low reproductive success in Per1 and Per2 mutant mouse females due
to accelerated ageing? Reproduction 135(4):559–568. https://doi.org/10.1530/REP-07-0434
Poletini MO, McKee DT, Kennett JE, Doster J, Freeman ME (2007) Knockdown of clock genes in
the suprachiasmatic nucleus blocks prolactin surges and alters FRA expression in the locus
coeruleus of female rats. Am J Phys Endocrinol Metab 293(5):E1325–E1334. https://doi.org/10.
1152/ajpendo.00341.2007
Poletini MO, Kennett JE, McKee DT, Freeman ME (2010) Central clock regulates the cervically
stimulated prolactin surges by modulation of dopamine and vasoactive intestinal polypeptide
release in ovariectomized rats. Neuroendocrinology 91(2):179–188. https://doi.org/10.1159/
000254379
Poling MC, Kim J, Dhamija S, Kauffman AS (2012) Development, sex steroid regulation, and
phenotypic characterization of RFamide-related peptide (Rfrp) gene expression and RFamide
receptors in the mouse hypothalamus. Endocrinology 153(4):1827–1840. https://doi.org/10.
1210/en.2011-2049
Rance NE (2009) Menopause and the human hypothalamus: evidence for the role of kisspeptin/
neurokinin B neurons in the regulation of estrogen negative feedback. Peptides 30(1):111–122.
https://doi.org/10.1016/j.peptides.2008.05.016
Ratajczak CK, Boehle KL, Muglia LJ (2009) Impaired steroidogenesis and implantation failure in
Bmal1-/- mice. Endocrinology 150(4):1879–1885. https://doi.org/10.1210/en.2008-1021
Ratajczak CK, Herzog ED, Muglia LJ (2010) Clock gene expression in gravid uterus and extra-
embryonic tissues during late gestation in the mouse. Reprod Fertil Dev 22(5):743–750. https://
doi.org/10.1071/RD09243
Ratajczak CK, Asada M, Allen GC, McMahon DG, Muglia LM, Smith D, Bhattacharyya S, Muglia
LJ (2012) Generation of myometrium-specific Bmal1 knockout mice for parturition analysis.
Reprod Fertil Dev 24(5):759–767. https://doi.org/10.1071/RD11164
Reppert SM, Henshaw D, Schwartz WJ, Weaver DR (1987) The circadian-gated timing of birth
in rats: disruption by maternal SCN lesions or by removal of the fetal brain. Brain Res
403(2):398–402
Resuehr D, Wildemann U, Sikes H, Olcese J (2007) E-box regulation of gonadotropin-releasing
hormone (GnRH) receptor expression in immortalized gonadotrope cells. Mol Cell Endocrinol
278(1–2):36–43. https://doi.org/10.1016/j.mce.2007.08.008
Rivkees SA, Reppert SM (1991) Appearance of melatonin receptors during embryonic life in
Siberian hamsters (Phodopus sungorous). Brain Res 568(1–2):345–349
Rizzo P, Raffone E, Benedetto V (2010) Effect of the treatment with myo-inositol plus folic acid
plus melatonin in comparison with a treatment with myo-inositol plus folic acid on oocyte
quality and pregnancy outcome in IVF cycles. A prospective, clinical trial. Eur Rev Med
Pharmacol Sci 14(6):555–561
Robertson MC, Friesen HG (1981) Two forms of rat placental lactogen revealed by radioimmuno-
assay. Endocrinology 108(6):2388–2390. https://doi.org/10.1210/endo-108-6-2388
442 N. Gotlieb et al.

Ronnekleiv OK, Kelly MJ (1988) Plasma prolactin and luteinizing hormone profiles during the
estrous cycle of the female rat: effects of surgically induced persistent estrus. Neuroendocrinol-
ogy 47(2):133–141. https://doi.org/10.1159/000124903
Ronnekleiv OK, Fang Y, Zhang C, Nestor CC, Mao P, Kelly MJ (2014) Research resource: gene
profiling of G protein-coupled receptors in the arcuate nucleus of the female. Mol Endocrinol 28
(8):1362–1380. https://doi.org/10.1210/me.2014-1103
Ruka KA, Burger LL, Moenter SM (2013) Regulation of arcuate neurons coexpressing kisspeptin,
neurokinin B, and dynorphin by modulators of neurokinin 3 and kappa-opioid receptors in adult
male mice. Endocrinology 154(8):2761–2771. https://doi.org/10.1210/en.2013-1268
Russo KA, La JL, Stephens SB, Poling MC, Padgaonkar NA, Jennings KJ, Piekarski DJ, Kauffman
AS, Kriegsfeld LJ (2015) Circadian control of the female reproductive axis through gated
responsiveness of the RFRP-3 system to VIP signaling. Endocrinology 156(7):2608–2618.
https://doi.org/10.1210/en.2014-1762
Saedi S, Khoradmehr A, Mohammad Reza JS, Tamadon A (2018) The role of neuropeptides and
neurotransmitters on kisspeptin/kiss1r-signaling in female reproduction. J Chem Neuroanat
92:71–82. https://doi.org/10.1016/j.jchemneu.2018.07.001
Sampaio RV, Conceicao S, Miranda MS, Sampaio Lde F, Ohashi OM (2012) MT3 melatonin
binding site, MT1 and MT2 melatonin receptors are present in oocyte, but only MT1 is present
in bovine blastocyst produced in vitro. Reprod Biol Endocrinol 10:103. https://doi.org/10.1186/
1477-7827-10-103
Samson WK, Burton KP, Reeves JP, McCann SM (1981) Vasoactive intestinal peptide stimulates
luteinizing hormone-releasing hormone release from median eminence synaptosomes. Regul
Pept 2(4):253–264
Sarkar DK, Chiappa SA, Fink G, Sherwood NM (1976) Gonadotropin-releasing hormone surge in
pro-oestrous rats. Nature 264(5585):461–463
Satake H, Hisada M, Kawada T, Minakata H, Ukena K, Tsutsui K (2001) Characterization of a
cDNA encoding a novel avian hypothalamic neuropeptide exerting an inhibitory effect on
gonadotropin release. Biochem J 354(Pt 2):379–385
Sawai N, Iijima N, Takumi K, Matsumoto K, Ozawa H (2012) Immunofluorescent histochemical
and ultrastructural studies on the innervation of kisspeptin/neurokinin B neurons to
tuberoinfundibular dopaminergic neurons in the arcuate nucleus of rats. Neurosci Res 74
(1):10–16. https://doi.org/10.1016/j.neures.2012.05.011
Sawai N, Iijima N, Ozawa H, Matsuzaki T (2014) Neurokinin B- and kisspeptin-positive fibers as
well as tuberoinfundibular dopaminergic neurons directly innervate periventricular hypophyseal
dopaminergic neurons in rats and mice. Neurosci Res 84:10–18. https://doi.org/10.1016/j.
neures.2014.05.002
Schafer D, Kane G, Colledge WH, Piet R, Herbison AE (2018) Sex- and sub region-dependent
modulation of arcuate kisspeptin neurons by vasopressin and vasoactive intestinal peptide. J
Neuroendocrinol:e12660. https://doi.org/10.1111/jne.12660
Schenker S, Yang Y, Perez A, Acuff RV, Papas AM, Henderson G, Lee MP (1998) Antioxidant
transport by the human placenta. Clin Nutr 17(4):159–167
Schoeller EL, Clark DD, Dey S, Cao NV, Semaan SJ, Chao LW, Kauffman AS, Stowers L, Mellon
PL (2016) Bmal1 is required for normal reproductive behaviors in male mice. Endocrinology
157(12):4914–4929. https://doi.org/10.1210/en.2016-1620
Schwartz WJ, Coleman RJ, Reppert SM (1983) A daily vasopressin rhythm in rat cerebrospinal
fluid. Brain Res 263(1):105–112
Seminara SB, Messager S, Chatzidaki EE, Thresher RR, Acierno JS Jr, Shagoury JK, Bo-Abbas Y,
Kuohung W, Schwinof KM, Hendrick AG, Zahn D, Dixon J, Kaiser UB, Slaugenhaupt SA,
Gusella JF, O’Rahilly S, Carlton MB, Crowley WF Jr, Aparicio SA, Colledge WH (2003) The
GPR54 gene as a regulator of puberty. N Engl J Med 349(17):1614–1627. https://doi.org/10.
1056/NEJMoa035322
Seron-Ferre M, Ducsay CA, Valenzuela GJ (1993) Circadian rhythms during pregnancy. Endocr
Rev 14(5):594–609. https://doi.org/10.1210/edrv-14-5-594
16 Development and Modulation of Female Reproductive Function by Circadian. . . 443

Seron-Ferre M, Riffo R, Valenzuela GJ, Germain AM (2001) Twenty-four-hour pattern of cortisol

in the human fetus at term. Am J Obstet Gynecol 184(6):1278–1283
Sharkey JT, Puttaramu R, Word RA, Olcese J (2009) Melatonin synergizes with oxytocin to
enhance contractility of human myometrial smooth muscle cells. J Clin Endocrinol Metab 94
(2):421–427. https://doi.org/10.1210/jc.2008-1723
Shimizu M, Bedecarrats GY (2010) Activation of the chicken gonadotropin-inhibitory hormone
receptor reduces gonadotropin releasing hormone receptor signaling. Gen Comp Endocrinol
167(2):331–337. https://doi.org/10.1016/j.ygcen.2010.03.029
Shiu RP, Kelly PA, Friesen HG (1973) Radioreceptor assay for prolactin and other lactogenic
hormones. Science 180(4089):968–971
Silver R, LeSauter J, Tresco PA, Lehman MN (1996) A diffusible coupling signal from the
transplanted suprachiasmatic nucleus controlling circadian locomotor rhythms. Nature 382
(6594):810–813. https://doi.org/10.1038/382810a0
Simonneaux V, Bahougne T (2015) A multi-oscillatory circadian system times female reproduc-
tion. Front Endocrinol (Lausanne) 6:157. https://doi.org/10.3389/fendo.2015.00157
Simonneaux V, Bahougne T, Angelopoulou E (2017) Daily rhythms count for female fertility. Best
Pract Res Clin Endocrinol Metab 31(5):505–519. https://doi.org/10.1016/j.beem.2017.10.012
Smarr BL, Gile JJ, de la Iglesia HO (2013) Oestrogen-independent circadian clock gene expression
in the anteroventral periventricular nucleus in female rats: possible role as an integrator for
circadian and ovarian signals timing the luteinising hormone surge. J Neuroendocrinol 25
(12):1273–1279. https://doi.org/10.1111/jne.12104
Smarr BL, Grant AD, Perez L, Zucker I, Kriegsfeld LJ (2017) Maternal and early-life circadian
disruption have long-lasting negative consequences on offspring development and adult behav-
ior in mice. Sci Rep 7(1):3326. https://doi.org/10.1038/s41598-017-03406-4
Smith JT, Cunningham MJ, Rissman EF, Clifton DK, Steiner RA (2005a) Regulation of Kiss1 gene
expression in the brain of the female mouse. Endocrinology 146(9):3686–3692. https://doi.org/
10.1210/en.2005-0488
Smith JT, Dungan HM, Stoll EA, Gottsch ML, Braun RE, Eacker SM, Clifton DK, Steiner RA
(2005b) Differential regulation of KiSS-1 mRNA expression by sex steroids in the brain of the
male mouse. Endocrinology 146(7):2976–2984. https://doi.org/10.1210/en.2005-0323
Smith JT, Popa SM, Clifton DK, Hoffman GE, Steiner RA (2006) Kiss1 neurons in the forebrain as
central processors for generating the preovulatory luteinizing hormone surge. J Neurosci 26
(25):6687–6694. https://doi.org/10.1523/JNEUROSCI.1618-06.2006
Smith JT, Coolen LM, Kriegsfeld LJ, Sari IP, Jaafarzadehshirazi MR, Maltby M, Bateman K,
Goodman RL, Tilbrook AJ, Ubuka T, Bentley GE, Clarke IJ, Lehman MN (2008) Variation in
kisspeptin and RFamide-related peptide (RFRP) expression and terminal connections to
gonadotropin-releasing hormone neurons in the brain: a novel medium for seasonal breeding
in the sheep. Endocrinology 149(11):5770–5782. https://doi.org/10.1210/en.2008-0581
Sonigo C, Bouilly J, Carre N, Tolle V, Caraty A, Tello J, Simony-Conesa FJ, Millar R, Young J,
Binart N (2012) Hyperprolactinemia-induced ovarian acyclicity is reversed by kisspeptin
administration. J Clin Invest 122(10):3791–3795. https://doi.org/10.1172/JCI63937
Stevens RG, Brainard GC, Blask DE, Lockley SW, Motta ME (2013) Adverse health effects of
nighttime lighting: comments on American Medical Association policy statement. Am J Prev
Med 45(3):343–346. https://doi.org/10.1016/j.amepre.2013.04.011
Strauss JF 3rd, Martinez F, Kiriakidou M (1996) Placental steroid hormone synthesis: unique
features and unanswered questions. Biol Reprod 54(2):303–311
Summa KC, Vitaterna MH, Turek FW (2012) Environmental perturbation of the circadian clock
disrupts pregnancy in the mouse. PLoS One 7(5):e37668. https://doi.org/10.1371/journal.pone.
0037668
Sun Y, Shu J, Kyei K, Neal-Perry GS (2012) Intracerebroventricular infusion of vasoactive
intestinal peptide rescues the luteinizing hormone surge in middle-aged female rats. Front
Endocrinol (Lausanne) 3:24. https://doi.org/10.3389/fendo.2012.00024
444 N. Gotlieb et al.

Szawka RE, Ribeiro AB, Leite CM, Helena CV, Franci CR, Anderson GM, Hoffman GE, Anselmo-
Franci JA (2010) Kisspeptin regulates prolactin release through hypothalamic dopaminergic
neurons. Endocrinology 151(7):3247–3257. https://doi.org/10.1210/en.2009-1414
Takahashi JS (2015) Molecular components of the circadian clock in mammals. Diabetes Obes
Metab 17(Suppl 1):6–11. https://doi.org/10.1111/dom.12514
Takayama H, Nakamura Y, Tamura H, Yamagata Y, Harada A, Nakata M, Sugino N, Kato H
(2003) Pineal gland (melatonin) affects the parturition time, but not luteal function and fetal
growth, in pregnant rats. Endocr J 50(1):37–43
Tamura H, Nakamura Y, Terron MP, Flores LJ, Manchester LC, Tan DX, Sugino N, Reiter RJ
(2008) Melatonin and pregnancy in the human. Reprod Toxicol 25(3):291–303. https://doi.org/
10.1016/j.reprotox.2008.03.005
Teclemariam-Mesbah R, Kalsbeek A, Pevet P, Buijs RM (1997) Direct vasoactive intestinal
polypeptide-containing projection from the suprachiasmatic nucleus to spinal projecting hypo-
thalamic paraventricular neurons. Brain Res 748(1–2):71–76
Thomas L, Purvis CC, Drew JE, Abramovich DR, Williams LM (2002) Melatonin receptors in
human fetal brain: 2-[(125)I]iodomelatonin binding and MT1 gene expression. J Pineal Res 33
(4):218–224
Thompson EL, Patterson M, Murphy KG, Smith KL, Dhillo WS, Todd JF, Ghatei MA, Bloom SR
(2004) Central and peripheral administration of kisspeptin-10 stimulates the hypothalamic-
pituitary-gonadal axis. J Neuroendocrinol 16(10):850–858. https://doi.org/10.1111/j.1365-
2826.2004.01240.x
Tian XZ, Wen Q, Shi JM, Liang W, Zeng SM, Tian JH, Zhou GB, Zhu SE, Liu GS (2010) Effects of
melatonin on in vitro development of mouse two-cell embryos cultured in HTF medium. Endocr
Res 35(1):17–23. https://doi.org/10.3109/07435800903539607
Topaloglu AK, Reimann F, Guclu M, Yalin AS, Kotan LD, Porter KM, Serin A, Mungan NO, Cook
JR, Imamoglu S, Akalin NS, Yuksel B, O’Rahilly S, Semple RK (2009) TAC3 and TACR3
mutations in familial hypogonadotropic hypogonadism reveal a key role for Neurokinin B in the
central control of reproduction. Nat Genet 41(3):354–358. https://doi.org/10.1038/ng.306
Torres-Farfan C, Rocco V, Monso C, Valenzuela FJ, Campino C, Germain A, Torrealba F,
Valenzuela GJ, Seron-Ferre M (2006) Maternal melatonin effects on clock gene expression in
a nonhuman primate fetus. Endocrinology 147(10):4618–4626. https://doi.org/10.1210/en.
2006-0628
Torres-Farfan C, Valenzuela FJ, Mondaca M, Valenzuela GJ, Krause B, Herrera EA, Riquelme R,
Llanos AJ, Seron-Ferre M (2008) Evidence of a role for melatonin in fetal sheep physiology:
direct actions of melatonin on fetal cerebral artery, brown adipose tissue and adrenal gland. J
Physiol 586(16):4017–4027. https://doi.org/10.1113/jphysiol.2008.154351
Trevisan CM, Montagna E, de Oliveira R, Christofolini DM, Barbosa CP, Crandall KA, Bianco B
(2018) Kisspeptin/GPR54 system: what do we know about its role in human reproduction? Cell
Physiol Biochem 49(4):1259–1276. https://doi.org/10.1159/000493406
Tsutsui K, Saigoh E, Ukena K, Teranishi H, Fujisawa Y, Kikuchi M, Ishii S, Sharp PJ (2000) A
novel avian hypothalamic peptide inhibiting gonadotropin release. Biochem Biophys Res
Commun 275(2):661–667. https://doi.org/10.1006/bbrc.2000.3350
Tsutsui K, Saigoh E, Yin H, Ubuka T, Chowdhury VS, Osugi T, Ukena K, Sharp PJ, Wingfield JC,
Bentley GE (2009) A new key neurohormone controlling reproduction, gonadotrophin-
inhibitory hormone in birds: discovery, progress and prospects. J Neuroendocrinol 21
(4):271–275. https://doi.org/10.1111/j.1365-2826.2009.01829.x
Ubuka T, Tsutsui K (2018) Comparative and evolutionary aspects of gonadotropin-inhibitory
hormone and FMRFamide-like peptide systems. Front Neurosci 12:747. https://doi.org/10.
3389/fnins.2018.00747
Ubuka T, Inoue K, Fukuda Y, Mizuno T, Ukena K, Kriegsfeld LJ, Tsutsui K (2012) Identification,
expression, and physiological functions of Siberian hamster gonadotropin-inhibitory hormone.
Endocrinology 153(1):373–385. https://doi.org/10.1210/en.2011-1110
16 Development and Modulation of Female Reproductive Function by Circadian. . . 445

Unfer V, Raffone E, Rizzo P, Buffo S (2011) Effect of a supplementation with myo-inositol plus
melatonin on oocyte quality in women who failed to conceive in previous in vitro fertilization
cycles for poor oocyte quality: a prospective, longitudinal, cohort study. Gynecol Endocrinol 27
(11):857–861. https://doi.org/10.3109/09513590.2011.564687
van der Beek EM, van Oudheusden HJ, Buijs RM, van der Donk HA, van den Hurk R, Wiegant VM
(1994) Preferential induction of c-fos immunoreactivity in vasoactive intestinal polypeptide-
innervated gonadotropin-releasing hormone neurons during a steroid-induced luteinizing hor-
mone surge in the female rat. Endocrinology 134(6):2636–2644. https://doi.org/10.1210/endo.
134.6.8194489
Van der Beek EM, Horvath TL, Wiegant VM, Van den Hurk R, Buijs RM (1997) Evidence for a
direct neuronal pathway from the suprachiasmatic nucleus to the gonadotropin-releasing hor-
mone system: combined tracing and light and electron microscopic immunocytochemical
studies. J Comp Neurol 384(4):569–579
van der Horst GT, Muijtjens M, Kobayashi K, Takano R, Kanno S, Takao M, de Wit J, Verkerk A,
Eker AP, van Leenen D, Buijs R, Bootsma D, Hoeijmakers JH, Yasui A (1999) Mammalian
Cry1 and Cry2 are essential for maintenance of circadian rhythms. Nature 398(6728):627–630.
https://doi.org/10.1038/19323
Vanecek J (1988) The melatonin receptors in rat ontogenesis. Neuroendocrinology 48(2):201–203.
https://doi.org/10.1159/000125008
Vatish M, Steer PJ, Blanks AM, Hon M, Thornton S (2010) Diurnal variation is lost in preterm
deliveries before 28 weeks of gestation. BJOG 117(6):765–767. https://doi.org/10.1111/j.1471-
0528.2010.02526.x
Vida B, Deli L, Hrabovszky E, Kalamatianos T, Caraty A, Coen CW, Liposits Z, Kallo I (2010)
Evidence for suprachiasmatic vasopressin neurones innervating kisspeptin neurones in the
rostral periventricular area of the mouse brain: regulation by oestrogen. J Neuroendocrinol 22
(9):1032–1039. https://doi.org/10.1111/j.1365-2826.2010.02045.x
Vilches N, Spichiger C, Mendez N, Abarzua-Catalan L, Galdames HA, Hazlerigg DG, Richter HG,
Torres-Farfan C (2014) Gestational chronodisruption impairs hippocampal expression of
NMDA receptor subunits Grin1b/Grin3a and spatial memory in the adult offspring. PLoS
One 9(3):e91313. https://doi.org/10.1371/journal.pone.0091313
Waddell BJ, Wharfe MD, Crew RC, Mark PJ (2012) A rhythmic placenta? Circadian variation,
clock genes and placental function. Placenta 33(7):533–539. https://doi.org/10.1016/j.placenta.
2012.03.008
Wakabayashi Y, Nakada T, Murata K, Ohkura S, Mogi K, Navarro VM, Clifton DK, Mori Y,
Tsukamura H, Maeda K, Steiner RA, Okamura H (2010) Neurokinin B and dynorphin A in
kisspeptin neurons of the arcuate nucleus participate in generation of periodic oscillation of
neural activity driving pulsatile gonadotropin-releasing hormone secretion in the goat. J
Neurosci 30(8):3124–3132. https://doi.org/10.1523/JNEUROSCI.5848-09.2010
Wakabayashi Y, Yamamura T, Sakamoto K, Mori Y, Okamura H (2013) Electrophysiological and
morphological evidence for synchronized GnRH pulse generator activity among Kisspeptin/
neurokinin B/dynorphin a (KNDy) neurons in goats. J Reprod Dev 59(1):40–48
Wang Y, Gu F, Deng M, Guo L, Lu C, Zhou C, Chen S, Xu Y (2016) Rotating shift work and
menstrual characteristics in a cohort of Chinese nurses. BMC Womens Health 16(1):24. https://
doi.org/10.1186/s12905-016-0301-y
Welsh DK, Yoo SH, Liu AC, Takahashi JS, Kay SA (2004) Bioluminescence imaging of individual
fibroblasts reveals persistent, independently phased circadian rhythms of clock gene expression.
Curr Biol 14(24):2289–2295. https://doi.org/10.1016/j.cub.2004.11.057
Wharfe MD, Mark PJ, Waddell BJ (2011) Circadian variation in placental and hepatic clock genes
in rat pregnancy. Endocrinology 152(9):3552–3560. https://doi.org/10.1210/en.2011-0081
Wiegand SJ, Terasawa E (1982) Discrete lesions reveal functional heterogeneity of suprachiasmatic
structures in regulation of gonadotropin secretion in the female rat. Neuroendocrinology 34
(6):395–404. https://doi.org/10.1159/000123335
446 N. Gotlieb et al.

Wiegand SJ, Terasawa E, Bridson WE, Goy RW (1980) Effects of discrete lesions of preoptic and
suprachiasmatic structures in the female rat. Alterations in the feedback regulation of gonado-
tropin secretion. Neuroendocrinology 31(2):147–157. https://doi.org/10.1159/000123066
Williams WP 3rd, Kriegsfeld LJ (2012) Circadian control of neuroendocrine circuits regulating
female reproductive function. Front Endocrinol 3:60. https://doi.org/10.3389/fendo.2012.00060
Williams LM, Martinoli MG, Titchener LT, Pelletier G (1991) The ontogeny of central melatonin
binding sites in the rat. Endocrinology 128(4):2083–2090. https://doi.org/10.1210/endo-128-4-2083
Williams WP 3rd, Jarjisian SG, Mikkelsen JD, Kriegsfeld LJ (2011) Circadian control of kisspeptin
and a gated GnRH response mediate the preovulatory luteinizing hormone surge. Endocrinol-
ogy 152(2):595–606. https://doi.org/10.1210/en.2010-0943
Wintermantel TM, Elzer J, Herbison AE, Fritzemeier KH, Schutz G (2006) Genetic dissection of
estrogen receptor signaling in vivo. Ernst Schering Found Symp Proc 1:25–44
Wu M, Dumalska I, Morozova E, van den Pol AN, Alreja M (2009) Gonadotropin inhibitory
hormone inhibits basal forebrain vGluT2-gonadotropin-releasing hormone neurons via a direct
postsynaptic mechanism. J Physiol 587(Pt 7):1401–1411. https://doi.org/10.1113/jphysiol.
2008.166447
Xie M, Utzinger KS, Blickenstorfer K, Leeners B (2018) Diurnal and seasonal changes in semen
quality of men in subfertile partnerships. Chronobiol Int 35(10):1375–1384. https://doi.org/10.
1080/07420528.2018.1483942
Yamazaki S, Numano R, Abe M, Hida A, Takahashi R, Ueda M, Block GD, Sakaki Y, Menaker M,
Tei H (2000) Resetting central and peripheral circadian oscillators in transgenic rats. Science
288(5466):682–685
Yano T, Iijima N, Kakihara K, Hinuma S, Tanaka M, Ibata Y (2003) Localization and neuronal
response of RFamide related peptides in the rat central nervous system. Brain Res 982
(2):156–167
Yeo SH, Colledge WH (2018) The role of Kiss1 neurons as integrators of endocrine, metabolic, and
environmental factors in the hypothalamic-pituitary-gonadal axis. Front Endocrinol 9:188.
https://doi.org/10.3389/fendo.2018.00188
Yip SH, Boehm U, Herbison AE, Campbell RE (2015) Conditional viral tract tracing delineates the
projections of the distinct Kisspeptin neuron populations to gonadotropin-releasing hormone
(GnRH) neurons in the mouse. Endocrinology 156(7):2582–2594. https://doi.org/10.1210/en.
2015-1131
Yogev L, Terkel J (1980) Effects of photoperiod, absence of photic cues, and suprachiasmatic
nucleus lesions on nocturnal prolactin surges in pregnant and pseudopregnant rats. Neuroendo-
crinology 31(1):26–33. https://doi.org/10.1159/000123046
Young J, Bouligand J, Francou B, Raffin-Sanson ML, Gaillez S, Jeanpierre M, Grynberg M,
Kamenicky P, Chanson P, Brailly-Tabard S, Guiochon-Mantel A (2010) TAC3 and TACR3
defects cause hypothalamic congenital hypogonadotropic hypogonadism in humans. J Clin
Endocrinol Metab 95(5):2287–2295. https://doi.org/10.1210/jc.2009-2600
Zahn V, Hattensperger W (1993) Circadian rhythm of pregnancy contractions. Z Geburtshilfe
Perinatol 197(1):1–10
Zhao S, Kriegsfeld LJ (2009) Daily changes in GT1-7 cell sensitivity to GnRH secretagogues that
trigger ovulation. Neuroendocrinology 89(4):448–457. https://doi.org/10.1159/000192370
Zhao L, Zhong M, Xue HL, Ding JS, Wang S, Xu JH, Chen L, Xu LX (2014) Effect of RFRP-3 on
reproduction is sex- and developmental status-dependent in the striped hamster (Cricetulus
barabensis). Gene 547(2):273–279. https://doi.org/10.1016/j.gene.2014.06.054
Glossary

5-Methylcytosine A genomic cytosine nucleotide that has been methylated.

Amygdala A brain region considered essential to a variety of social and emotional
behaviors and which shows sex differences in anatomy and physiology.
Androgens A class of steroid hormones that includes testosterone and dihydrotes-
tosterone. Testosterone can be aromatized into estradiol but dihydrotestosterone
cannot.
Anisotropic growth Growth and differentiation from a progenitor population that is
significantly different (greater) than growth/differentiation from adjacent
progenitors.
Anorexigenic Inhibiting appetite and promoting energy expenditure.
Anterior lobe of the pituitary gland or adenohypophysis Contains pituitary
hormone-producing cell types.
Anterioventral periventricular nucleus (AVPV) A small subnucleus in the
preoptic area which is larger in females and projects to the GnRH neurons that
control ovulation.
Antiphase When comparing two or more oscillations, the peak of one oscillation
corresponds to the trough of another.
Arcuate nucleus An aggregation of neurons in the mediobasal hypothalamus,
which is located adjacent to the third ventricle and the median eminence;
orexigenic, appetite-stimulating; anorexigenic, appetite-suppressing.
Aromatase An enzyme that converts testosterone into estradiol. It is found at high
concentrations in the ovary and in discrete regions of the brain.
Astrocytes A type of glia dispersed broadly throughout the brain that play impor-
tant roles in modulating neuronal health and activity.
Autosynaptic feedback A chemical or electrical synapse formed by the axon of a
neuron on its own dendrites or soma.
bHyp (basal Hypothalamic) cells A proliferating cell population that develops
from RDVM cells, gives rise to anterior, mammillary and infundibular
progenitors, and is retained through life. Potentially harbours a neural stem cell.
Bisulfite conversion of genomic DNA Treatment with sodium bisulfite to promote
the deamination of cytosine residues, which is followed by desulfonation at
elevated pH, thus leading to conversion into uracil. Methylated cytosines are
resistant to bisulfite conversion.

# Springer Nature Switzerland AG 2020 447

S. Wray, S. Blackshaw (eds.), Developmental Neuroendocrinology, Masterclass in
Neuroendocrinology 9, https://doi.org/10.1007/978-3-030-40002-6
448 Glossary

Caudal brainstem Including the medulla (medulla oblongata), aka the hindbrain.
Houses the area postrema (AP), nucleus of the solitary tract (NTS), dorsal motor
nucleus of the vagus (DMV), medullary reticular formation (RF), and ventrolat-
eral medulla (VLM).
Cell lineage All the cells born from one dividing cell (or from one specific kind of
dividing cell) form a “cell lineage.”
Cellular central oscillator Activity of the core circadian oscillator at the cellular
(as opposed to organ or organismal) level.
CHIP-seq High-throughput sequencing method combining chromatin immunopre-
cipitation technique with deep sequencing. It allows the identification of specific
DNA sequences bound to proteins of interest in vivo. Benefits include single
nucleotide resolution, relatively low amount of DNA input, and high dynamic
range for quantification.
Chronobiology The study of endogenous biological rhythms and the entrainment
by environmental cues. Also referred to as biochronometry.
Chronotype Intrinsic, individual differences in the synchrony between light cycles
and activity rhythms. Manifests as “morningness” (phase-advanced) and
“eveningness” (phase-delayed sleep/wake rhythms).
Circadian An endogenous behavioural, biochemical, or physiological oscillation
that approximates a 24-h period.
Circadian disruption An incongruence between endogenous circadian rhythms
and external time.
Circadian entrainment This occurs when rhythmic physiological or behavioral
events match their period to that of an environmental oscillation. Entraining
signals can include light, food, and neurotransmitters.
Circadian phase synchrony and stabilization Circadian phase synchrony occurs
when cellular clocks are synchronized and run in phase, typically in response to
an external cue. Circadian phase stabilization indicates that this synchrony is
robust and reproducible over many phases.
Circadian rhythms Endogenously generated behavioral, physiological, endocrine,
metabolic, and/or molecular rhythms having a ~24-h period that is synchronized
to environmental time and persist in the absence of environmental cues, but can
be entrained by light, food and other external cues.
Clock gene A gene that is critical for generating circadian timekeeping at the
cellular level.
Collateral inhibition Neural networks in which inhibitory (usually GABAergic)
neurons inhibit the activity of synaptically connected neurons.
Conditional mutant A transgenic mouse mutant line engineered so that gene A
(gene of interest) will become inactive only in a certain specific tissue domain
where it is co-expressed with gene B (“Cre-driver gene”).
Core circadian oscillator A transcriptional feedback loop in which increased
expression of Per and Cry proteins inhibits activity of Bmal1 and Clock. It is
active in virtually all mammalian cells.
Glossary 449

Corticotropes Anterior lobe cells that produce adrenocorticotropin (ACTH) from

pro-opiomelanocortin (POMC).
CpG island DNA sequence of more than 200 bp with >50% GC-content and a
CpG dinucleotide observed/expected ratio of >60%.
CPHD Combined Pituitary Hormone Deficiency (CPHD) is multiple pituitary
hormone deficiencies, usually involving GH and at least one other hormone.
Critical period A developmental window during which endogenous or exogenous
stimuli impact the trajectory of growth. Examples of such stimuli are light or
hormones, both of which direct changes in synaptic circuitry that are essential to
normal development of the visual system and reproductive physiology and
behavior, respectfully. Exposure to the stimuli must occur during the critical
period for normal development to ensue.
Cyclooxygenase enzymes (COX-1 and COX-2) Enzymes which convert
arachidonic acid derived from lipid membranes into short levied precursors of
prostaglandins, which are synthesized by specific enzymes for each prostaglandin.
Cyclopia A birth defect characterized by the presence of one single eye, positioned
in the middle of the face. Cyclopia is caused by the failure of the developing
forebrain to separate into the left and right hemispheres and accompanies some
kinds of holoprosencephaly.
De novo motif analysis A computational method to identify conserved sequence
features in large amounts of CHIP-seq data.
Defeminization A biological process by which the default neural underpinnings of
female physiology and behavior is removed from the male brain.
Developmental regulatory gene A gene encoding a transcription factor or a signal-
ing protein, which is expressed during development in a specific moment in
specific patterns, and is able to regulate expression of other genes and control
patterning and morphogenesis of specific regions.
Diencephalon The posterior region of the forebrain. Gives rise to thalamus and
hypothalamus.
Differentially methylated region (DMR) CpG island with different levels of DNA
methylation on its two alleles.
DNA methylation Addition of a methyl group to DNA, which affects gene activity
without alteration of the DNA sequence. DNA methylation is found primarily on
the cytosine of cytosine-guanine (CpG) dinucleotides.
DNA methyltransferase These are a family of enzymes that are effectors of DNA
methylation and catalyze either de novo or maintenance methylation of hemi-
methylated DNA. There are 3 active DNMTs in mammalian species, DNMT1,
DNMT3A, and DNMT3B. These enzymes place a methyl group onto cytosines
that are proximal to guanines in the DNA, thereby changing the probability that
the surrounding genome will be transcribed.
Dysmasculinization The disruption of the normal process of masculinization,
which can therefore only occur in males.
Embryonic development The period of an animal’s development in utero (also
referred to as fetal development in higher vertebrates).
450 Glossary

Emetic reflex The vomiting reflex, present in most mammalian species (including
humans), but absent in non-emetic rodents (e.g., rats and mice).
Endocannabinoids A class of signaling molecules derived from membrane lipids
and that bind to the same receptors as the active compounds in marijuana.
Endogenous Naturally occurring, e.g., hormones released from endocrine tissues.
Entrainment Synchronization of endogenously generated circadian rhythms to the
external environment.
Epigenetic modification Biochemical change to the genome or chromatin template
that alters the conformational structure and induces either a permissive or an
inhibitory state for gene transcription. Examples include DNA methylation,
histone acetylation and histone methylation.
Epigenetic regulator Chromatin remodeling enzymes and enzymes that place or
remove methylation and acetylation marks on histones bound to chromatin,
altering access of transcription factors to bind DNA.
Epigenetics It refers to modifications to the DNA that impact gene expression but
do not change the genetic code (i.e., G, C, A & T). These can include direct
methylation of the DNA and indirect via modifications of the surrounding
histones that impact accessibility of transcription factors to the DNA.
Epithelial to mesenchymal-like transition Epithelial cells lose their polarity,
release cell–cell adhesion, change shape and acquire migratory properties, a
process that occurs in development of many organs and in cancer.
Estradiol A sex steroid of the estrogen family, synthesized in, and secreted from,
the ovary.
Estrogens A class of steroid hormones that includes estradiol, estriol, and estrone.
In adulthood estradiol is produced in high quantities by the ovary. During
development estradiol is produced in the brain by being derived from testosterone
in a process called aromatization. In rodents, estrogens are masculinizing
hormones during development.
Exogenous Artificially sourced, e.g., drugs, synthetic peptides, and other chemical
agents.
Exteroceptive Sensory signals arising from the environment (e.g., stimuli con-
veyed to the central nervous system through visual or auditory pathways) rather
than from within the body.
Fast neurotransmitter Fast-acting neurotransmitters such as glutamate (excit-
atory) and gamma-aminobutyric acid (GABA) (inhibitory), modulate their target
cells in under one millisecond by directly activating ligand-gated ion channels.
Monoamines and neuropeptides, which act via G-protein coupled receptors
(GPCRs), act more slowly.
Fate mapping Any technique allowing for the labeling of a cell lineage, i.e., the
labeling of one specific kind of dividing cell and all the cells that derive from it.
Floor plate A tissue domain, that acts of a source of morphogens, occupying the
ventral midline of the developing neural tube and formed by specific cells (not
neuroepithelial cells) with essential functions in neural tube fate specification.
The ventral midline of the most rostral part of the neural tube (corresponding to
Glossary 451

the hypothalamus) shows important differences with the floor plate elsewhere and
usually receives a different name (see Hypothalamic floor plate, Rostral
diencephalic ventral midline).
Follicle stimulating hormone A hormone secreted by the anterior pituitary that
stimulates the growth of ovarian follicles and sperm maturation.
Folliculostellate cells Non-endocrine cells in the anterior pituitary with star-like
cytoplasmic processes.
Forebrain Consists of structures derived from the telencephalon and diencepha-
lon. It is rostral to the mesencephalon and contains the centers involved in the
control of sensorimotor, autonomic and endocrine functions, in emotional behav-
ior, as well as in cognitive functions such as learning and memory.
Fos The protein product of the immediate-early gene, c-fos. Fos protein accumulates
within the nuclei of neurons and other cells due to increased intracellular calcium.
Increased neuronal Fos labeling (compared to baseline levels) is interpreted as
evidence that the neuron was stimulated/activated (e.g., received increased excit-
atory synaptic input).
Free run A self-sustained circadian rhythm that is not synchronized to
environmental time.
GABA Gamma-aminobutyric acid—The primary inhibitory neurotransmitter in the
adult brain and a secreted factor that regulates cell function during development.
GAD Glutamic acid decarboxylase—Enzyme that converts glutamate to GABA
(two forms GAD65 and GAD67).
GATA2 Transcription factor that drives thyrotrope and gonadotrope cell fate.
Genomic imprinting Parent-of-origin-dependent, monoallelic gene expression
caused by epigenetic marks (i.e., DNA methylation) set in the parental germ
cells and maintained in the somatic cells of the offspring.
GHRH-neurons Arcuate nucleus neurons that stimulate linear growth; Kiss1-
neurons, arcuate nucleus neurons that directs reproduction.
Gliogenesis Generation of new glial cells from progenitor cells.
Glycoprotein hormones Bioactive LH, FSH, and TSH are composed of specific
beta subunits and a common alpha subunit, CGA or chorionic gonadotropin
alpha. Both alpha and beta subunits are glycosylated.
GnRH Gonadotropin-releasing hormone—A hormone secreted into anterior pitui-
tary portal system that stimulates the release of gonadotropins from the anterior
pituitary into the general circulation and as such regulates reproductive function.
Gonadotropes Anterior lobe cells that produce follicle-stimulating hormone (FSH)
and luteinizing hormone (LH).
Gonadotropin-inhibitory hormone An RFamide peptide that is inhibitory to the
synthesis and secretion of gonadotropins from the anterior pituitary; a ligand for
GPR147 also called RFamide-related peptide-3 (RFRP-3) in mammals.
Gonadotropins Hormones synthesized and secreted from the anterior pituitary that
regulate sex steroid production and gametogenesis—follicle stimulating hormone
and luteinizing hormone, respectively.
HESX1 Transcription factor involved in development of the eye, forebrain, and
Rathke’s pouch.
452 Glossary

Histamine An amino acid that acts as a signaling molecule in inflammatory

reactions and in the nervous system to promote arousal. It can be synthesized
and released by neurons in a neurotransmitter fashion or from mast cells following
degranulation.
Histone acetylase Acetylate-specific lysine residues in histone substrates.
Histone deacetylase Remove acetyl residues from histone substrates.
Histone modification A covalent posttranslational modification to histone proteins
resulting in methylation, phosphorylation, acetylation, ubiquitylation, and
sumoylation. These modifications result in chromatin structure to alter gene
expression levels.
Histones Proteins that form the basis of the nucleosomes, tightly wound compact
units of DNA that impact the accessibility of transcription factors to the DNA.
The addition of modifications to histone “tails”, determine how tightly or loosely
the DNA is coiled within the nucleosomes.
Holoprosencephaly A congenital malformation of the brain and face, sometimes
resulting in cyclopia. One major cause of holoprosencephaly is genetic mutations
affecting the Shh signaling pathway.
Homology Structures are proposed to be homologous in different animals when-
ever it is thought that the same structures were present in their common ancestor,
being not necessary a functional similarity.
Hypophagia Reduced food intake.
Hypophysiotropic Acting on the anterior pituitary gland (hypophysis).
Hypothalamic basal plate The Shh expression domain forming a lateral band in
the developing forebrain.
Hypothalamic floor plate The portion of the floor plate corresponding to the
presumptive hypothalamus. Unlike the floor plate in the rest of the neural tube,
hypothalamic floor plate cells proliferate; they also express markers other than
those typical of the floor plate elsewhere.
Hypothalamus A brain region critical to the control of the anterior pituitary and
reproductive behaviors.
Hypothalamus Pituitary Gonadal (HPG) axis Neuroendocrine network that
consists of functional interactions between the hypothalamus, anterior pituitary
gland, and gonads. The decapeptide gonadotropin-releasing hormone (GnRH) is
the hypothalamic peptide controlling reproduction that stimulates the secretion of
the anterior pituitary hormones—luteinizing hormone (LH) and follicle-
stimulating hormone (FSH), which ultimately controls the production of gonadal
steroids and gametogenesis.
IGHD Isolated Growth Hormone Deficiency (IGHD) is a deficiency in circulating
GH without deficiency of other pituitary hormones.
Imprinting Control Region (ICR) DMR that is already established in the germ
cells, maintained in somatic cells of the offspring and has been proven to control
the imprinting of neighboring genes within a cluster.
Interoceptive Sensory signals arising within the body (e.g., from cardiovascular,
gastrointestinal, and endocrine systems) rather than from the environment.
Glossary 453

Intrinsically photosensitive retinal ganglion cells (ipRGCs) Opn4-expressing

retinal ganglion cells that send postsynaptic projections to the SCN, as well as
other hypothalamic regions that regulate sleep. They are the sole source of light
input to the SCN.
Ki67 Protein associated with cell proliferation.
Kisspeptin An RFamide peptide encoded by the Kiss1 gene that is an essential,
stimulatory signal residing upstream of the GnRH system. A ligand for GPR54
and formerly known as metastin.
Lactotropes Anterior lobe cells that produce prolactin (PRL).
Lateral Progenitor Domain In the context of this review, the tissue domain
expressing Shh as a lateral band in the presumptive hypothalamus. It is the
neuroepithelial domain where the mammillary and tuberal regions are generated
and receives also the name of hypothalamic basal plate.
LHX3 Transcription factor necessary for expansion of Rathke’s pouch and func-
tional overlap with LHX3.
LHX4 Transcription factor necessary for expansion of Rathke’s pouch.
Lordosis A posture adopted by female rodents in response to mounting from a male
if they are sexually receptive. The inward dorsoflexion of the spine led to the use
of the term lordosis.
LSD1 Lysine-specific demethylase, officially known as Kdm1a, implicated in
differentiation of somatotropes and lactotropes.
Luteinizing hormone (LH) A hormone secreted by the anterior pituitary that
stimulates sex steroid production in males and females and facilitates ovarian
follicle development and ovulation in females.
Mast cells Innate immune cells which are found throughout the body but include a
brain resident population. They contain large secretory granules which upon
degranulation release their contents into the surround. The contents include
histamine as well as serotonin and other signaling molecules. They also make
and release cytokines and prostaglandins.
Melanopsin A photosensitive pigment and member of the G-protein coupled
receptor opsin family expressed in intrinsically photosensitive retinal ganglion
cells involved in transmitting light information to entrain the circadian system.
Melanotropes Intermediate lobe cells that produce melanocyte-stimulating hor-
mone (MSH) from POMC.
Meninges A membranous structure around the brain that forms a barrier with the
periphery but is also a gateway for specific trafficking of molecules and cells.
Methylcytosine dioxygenase A family of enzymes that initiates the molecular
cascade to demethylate DNA via the conversion of methylated cytosine residues
into a hydromethylated state.
Microglia Specialized cells derived from macrophages and that constitute the
brains innate immune system. Despite the name, microglia are not related to the
other glia in the brain, astrocytes and oligodendrocytes.
454 Glossary

Morphogenetic field Higher order module, specified by a particular combination of

developmental regulatory genes, that gives rise to a particular region. It
constitutes a major unit of development and evolution, and the natural compari-
son character for homology considerations.
Neural precursor An immature postmitotic neuron.
Neural progenitor A proliferating cell that gives rise to neurons.
Neural Shh Shh secreted by the neural tube, as opposed to prechordal Shh, which is
secreted by the prechordal plate.
Neuroepithelial Shh(+) progenitors A dynamic progenitor pool that develops
from bHyp cells, whose size is constantly maintained through generation and
differentiation.
Neuroepithelium Cells of the neural tube that have not yet initiated neurogenesis.
Non-rapid eye movement (NREM) sleep Also known as quiescent sleep.
Comprises ~80% of sleep time in young adult humans. During NREM sleep,
the brain waves as measured by electroencephalographic recording are low
frequency and high amplitude, and the sleeper is relatively still, but maintains
muscle tone. Dreaming is infrequent.
NR5A1 Transcription factor that drives gonadotrope cell fate.
OE Main olfactory epithelium that projects to the olfactory bulb.
Orexigenic Promoting appetite and inhibiting energy expenditure.
Oscillation A recurrent event that is characterized by its period, frequency, ampli-
tude and phase.
PAX7 Pioneering transcription factor that drives melanotrope cell fate and opens
chromatin, creating access for other transcription factors to bind.
Period The length of time to complete one cycle of a rhythm.
Phase-advanced Characterized by early waking and sleeping times.
Phase-delayed Characterized by late waking and sleeping times.
Pituicytes Glial-like cells in the posterior pituitary that assists the axons producing
oxytocin and vasopressin Posterior Lobe of the Pituitary Gland or Neurohypoph-
ysis: derived from neural ectoderm. Contains axon terminals for oxytocin and
vasopressin and pituicytes.
PITX1 Transcription factor expressed in early pituitary development, enriched in
thyrotropes and gonadotropes, with overlapping function with PITX2.
PITX2 Transcription factor important for expansion of Rathke’s pouch and later,
for expression of Gata2, Nr5a1, and Pou1f1.
Placental lactogen A placental hormone of the somatotropin family that modifies
the maternal metabolic state to supply energy to the developing fetus.
POMC-neurons Arcuate nucleus neurons that suppress appetite; AgRP-neurons,
arcuate nucleus neurons that stimulate appetite.
POU1F1 Pituitary-specific transcription factor necessary for promotion of the
thyrotrope, somatotrope, and lactotrope cell fates.
Prechordal mesendoderm A fan-shaped structure that forms anterior to the noto-
chord, and underlies the ventral midline of the diencephalon. Posteriorly, it is
composed of axial prechordal mesoderm, and anteriorly, it is composed of a mix
of axial prechordal mesoderm and endoderm.
Glossary 455

Prechordal plate The portion of the axial mesoderm underlying the presumptive
hypothalamus. Shh from the prechordal plate is required for hypothalamic
specifications.
Preoptic Area (POA) A brain region critical to the control of male sexual behavior
and female maternal behavior. It is also the site of the sexually dimorphic nucleus
(SDN), and a variety of other sex differences in anatomy and physiology.
Primary wake-promoting neurons Neurons that rapidly (i.e., within seconds to
minutes) induce wakefulness following activation.
Progenitor cell A proliferative cell that can undergo a limited number of cell
divisions, and gives rise to a limited type of differentiated cells.
Progenitor domain A restricted region of the early neural tube, which generates
the cells for a certain subdivision of the central nervous system.
PROP1 Pituitary-specific transcription factor that marks anterior and intermediate
lobe progenitors and is required for expansion of the Pou1f1 lineage.
Prostaglandin E2 (PGE2) A membrane derived signaling molecule usually
associated with inflammation but which also has normal roles in brain develop-
ment, including the promoting of synapse formation in the male POA.
Protein kinase-A (PKA) An enzyme which phosphorylates other proteins to
modify their function, such as subunits of glutamate receptors, which impacts
how they traffic to the membrane.
Puberty Developmental period when an individual becomes capable of sexual
reproduction. It is marked by genital organ maturation and development of
secondary sexual characteristics.
PVN Paraventricular nucleus of the hypothalamus—Heterogeneous grouping of
cells in the hypothalamus known to function in the regulation of blood pressure,
appetite, stress responses, sexual function, cardiac output, and organismal
growth.
Rapid eye movement (REM) sleep Comprises ~20% of sleep time in young adult
humans. During REM sleep, brain waves as measured by electroencephalo-
graphic recording are typically high frequency and low amplitude, and the sleeper
exhibits flaccid paralysis, with exception of the ocular muscles, which move
rapidly. Dreaming is frequent and vivid.
Rathke’s pouch An invagination of oral ectoderm that forms the anterior and
intermediate lobes of the pituitary gland.
Retinohypothalamic tract Axonal projections of the SCN to the hypothalamus.
RFamide peptide A peptide that contains Arg-Phe-NH2 motif at its C-terminus.
RFRP-3 An RFamide peptide synthesized and secreted from the hypothalamus that
inhibits the release of gonadotropins from the anterior pituitary; mammalian
ortholog of gonadotropin-inhibitory hormone (GnIH).
Rhythmic epigenetics The study of the biochemical, physiological and environ-
mental regulation of epigenetic enzyme oscillations and the functional signifi-
cance of predictable changes in epigenetic modifications across different
biological periods (e.g. seasonal).
456 Glossary

RNA-seq High-throughput sequencing method for mapping and quantifying

transcriptomes. Benefits include: Simultaneous mapping of transcribed regions,
dynamic range of gene expression level (>8000-fold), ability to distinguish
between isoforms and allelic expression, and a relatively low cost for
transcriptomes of large genomes.
Rostral diencephalic ventral midline (RDVM) cells A subset of medial neural
plate cells that constitute the precursors of the hypothalamic floor plate. They are
a transient embryonic population that expresses Shh and will upregulate Fgf10 to
develop into bHyp cells under the influence of prechordal mesendoderm-derived
signals.
scRNA-seq A method to define the expression profiles of individual cells by
performing next generation sequencing at single cell levels.
Seasonal rhythms A biological rhythm that is entrained to a 12-month period.
Sensitive period A developmental window during which the brain is responsive to
exogenous stimuli that can alter the trajectory of growth. Sensitive periods often
overlap with critical periods.
Sensory processing The term used to describe how the brain converts sensory
information (e.g., olfactory cues) into an appropriate behavioral response.
Sexual dimorphism A sex-related difference in structure, function, and/or
behavior.
Sexually dimorphic nucleus of the preoptic area (SDN-POA) A collection of
neurons in a subnucleus that is 3–5 times larger in males than females, it was the
first major neuroanatomical sex difference reported in a mammal.
SHRP (stress-hyporesponsive period) In rodents, an early postnatal developmen-
tal period in which stress-induced activation of the hypothalamic–pituitary–
adrenal (HPA) axis is suppressed.
Skeleton photoperiod An external light-dark cycle in which a long period of light
(the day phase) is not present, and instead short periods of light mark the
beginning and end of the day phase.
Sleep homeostasis The amount of daily sleep time required by an organism. A
sleep deficit elicits a compensatory increase in the intensity and duration of sleep,
while excessive sleep reduces sleep propensity. NREM and REM homeostasis are
regulated separately. There is considerable age and species-dependent variation
in sleep homeostasis.
Sleep pressure The consists of the sum of two components: (S), a drive to sleep that
increases with time spent awake and is dissipated by time spent asleep and (C), a
drive to sleep that is under direct control of the central circadian oscillator in the
suprachiasmatic nucleus.
Somatotropes Anterior lobe cells that produce somatotropin or growth hormone
(GH).
SOX2 Transcription factor in pituitary stem cells.
Specific Cre lines Mouse lines expressing the Cre recombinase in specific cell
types, which are utilized to delete genes of interests in specific cell types.
Glossary 457

Spinal nucleus of the bulbocavernosus (SNB) A collection of motor neurons in

the spinal cord that controls the penis. The motor neurons are largely absent or
substantially diminished in females.
Stem cell A proliferative cell that can self-renew indefinitely and give rise to many
different types of differentiated cells.
Stress responses Behavioral and physiological (endocrine, autonomic) responses
to stress.
Stress Referring to stimuli or conditions that threaten (or are perceived to threaten)
physiological homeostasis and well-being; such stimuli and conditions are
“stressful.”
Stress-threat system Brain nuclei that regulate an animal’s responses to stressors
and external threats.
Suprachiasmatic nucleus (SCN) Small, paired hypothalamic nuclei (of about
20,000 neurons) found dorsal to the optic chiasm that act as the regulator of
physiological circadian rhythms in mammals. The SCN is directly regulated by
light, relayed from the retina via ipRGCs. It is the master circadian pacemaker in
mammals.
Synaptic plasticity Changes in synaptic architecture or strength in response to
environmental or physiologic cues.
Tanycytes Specialized cells that are a type of ependymal (ventricle lining) cell with
process that extend into the basal hypothalamus; among the hypothesized
functions are sensing both circulating and cerebrospinal fluid content and
regulating neuroendocrine peptide release (including physically impeding axonal
contacts).
Telencephalon The most anterior major division of the neural tube, and anterior
region of the forebrain. Gives rise to cortex, hippocampus, amygdala, and septal
nuclei.
Testicular hormones They are steroids derived from the testis, predominantly
androgens which are then aromatized to estrogens in the brain or act directly on
androgen receptors (AR).
Tet (Ten-eleven translocation) enzymes Oxidize the methyl group of
5-methylcytosine (5mC) to yield 5-hydroxymethylcytosine (5hmC). There are
3 Tet enzymes, Tet 1, Tet 2, and Tet 3.
Thyrotropes Anterior lobe cells that produce thyrotropin (Thyroid-Stimulating
Hormone, TSH).
TPIT Transcription factor that drives corticotrope and melanotrope cell fate.
Transcription factors Proteins that bind to regulatory regions of DNA in a highly
specific manner and participate in a complex with other proteins and enzymes to
transcribe the DNA into RNA.
Ventricular zone The region of the developing brain lining the ventricles where
neurons are born.
VMH Ventromedial nucleus of the hypothalamus—Heterogeneous grouping of
cells in the hypothalamus known to function in the regulation of appetite, fear,
thermoregulation, social and sexual behaviors (also referred to as VMN).
458 Glossary

VNO Vomeronasal organ—Putative site of origin for GnRH neurons and origina-
tion of olfactory fibers to the accessory olfactory bulb.
ZEB1 Transcription factor implicated in differentiation of lactotropes from
somatotropes.
ZEB2 Transcription factor involved in epithelial to mesenchymal-like transition of
pituitary progenitors.
Zona incerta (ZI) A bilateral elongated nucleus in dorsal hypothalamus. It
regulates a diverse range of physiological processes, including encoding sleep
pressure.
Index

A Anterior pituitary (AP), 6, 21, 346, 351, 353,

AAV9 virus, 223 354
Abrogates, 144 Anterior pituitary gland, 346
A1/C1 cell groups, 353 Anterolateral spinothalamic tract, 348
A2/C2 cell groups, 353 Anteroventral periventricular nucleus (AVPV),
A1 cell group, 355 209, 213, 248–253, 395
A2 cell group, 355 Antihistamines, 106
Accessory cell groups, 346 Antisense morpholino oligonucleotides, 73
Acetylation, 251 Anxiety, 361
Acetylation of H3, 222 Apoptosis, 90
Acroterminal region, 33 Apterous, 139–141
Adenohypophysis, 75 Arborization of astrocytes, 371
Adenosine, 109 Arborization of glial, 371
Adolescence, 112 Arcuate, 71, 378
Adrenergic, 353, 356, 361 Arcuate neurons, 379
Adrenergic receptors, 353, 354 Arcuate nucleus (ARC), 38, 72, 79, 213,
Adrenocorticotropic hormone (ACTH), 131, 248–254, 346, 350, 352, 355, 356, 371,
346, 352–355, 361 378–383
Agouti-related peptide (AgRP), 84 Arcuate nucleus (ARC) neurons, 208
Alar, 33 Area postrema (AP), 348, 351, 352
Alar plate, 349 Assay for Transposase-Accessible Chromatin
Albright’s Hereditary Osteodystrophy (AHO), sequencing (ATAC-seq), 210
271 Astrocyte arborization, 371
α-melanocyte stimulating hormone (αMSH), Astrocyte ensheathment, 380
78, 84 Astrocyte heterogeneity, 373, 374
α2-tanycytes, 383 Astrocyte morphology, 371, 379
Amenorrhea, 418 Astrocyte proliferation, 381
Anamnio-amniotic transition, 80 Astrocytes, 85, 368–371, 383, 384, 408
Anamniote tetrapods, 80 Astrocytes and development of neuroendocrine,
Androgen receptor immunoreactive (AR-ir), 368–375
373 Astrocytes in metabolic control, 380
Androgen receptors in GFAP-positive cells, Astrocyte-specific reagents, 384
373 Astrocyte subpopulations, 385
Angelman syndrome (AS), 268 “Astrocyte” umbrella, 384
Anterior, 8, 33 Astrocytic end feet, 378
Anterior and intermediate lobes, 129 Astrocytic erbB4 receptors, 210
Anterior forebrain, 33 Astrocytic modulation of oxytocin or
Anterior hypothalamic nucleus, 71 vasopressin release, 377
Anterior neural ridge, 50 Astrocytic processes, 377, 379

# Springer Nature Switzerland AG 2020 459

S. Wray, S. Blackshaw (eds.), Developmental Neuroendocrinology, Masterclass in
Neuroendocrinology 9, https://doi.org/10.1007/978-3-030-40002-6
460 Index

Astrocytic subpopulations, 384 Cerebral cortex, 332

Astroglial coverage, 376, 381 c-fos, 351
Astroglial–neuronal interactions, 376 Chemodetection, 319
Astroglial processes, 378 Childhood, 112
Attention deficit and hyperactivity Cholecystokinin octapeptide (CCK), 350–355,
disorders, 410 357, 360, 361
Autism spectrum disorders, 97, 410 Chromatin Immunoprecipitation (CHIP), 210
Autoregulatory transcription–translation Chromatin Immunoprecipitation with
feedback loop, 415 sequencing (CHIP-seq), 93, 158, 210,
Axial mesoderm, 32 217, 219, 221, 222, 224
Chronobiology, 296
Chronotypes, 112
B Cilia, 33
Background adaptation, 78 Ciliary neurotrophic factor, (CNTF), 382
Basal, 33 Circadian, 300, 304–307
Basal astroglial/synaptic arrangement, 380 Circadian disruptions, 424, 426, 428
Basic-helix-loop-helix (bHLH), 85 Circadian oscillations, 300
Bed nucleus of the stria terminalis (BNST), Circadian rhythms, 266, 414, 415, 423–428
352, 355 Circadian system, 414, 428
β-catenin, 113, 133, 147 Circadian timing, 425
Beta-dystroglycan (β-DG), 378 Circadian timing system, 420
β-tanycytes, 383 Clash of nomenclatures, 33
Binding to C/EBPα, 147 Clock, 110, 422, 425, 429
Binds β-catenin, 144 Clock genes, 424
Birth, 429–430 Cluster analysis, 374
Birthdate labeling, 186 Co-activators, 404
Birth-dating techniques, 73 “Cocaine and amphetamine regulated
Bisulfite, 280 transcript” (CART), 84
Black, 107 Co-expression network analysis, 227
Blood-brain barrier, 348 Cognitive stressors, 355
Blood–brain interfaces, 7 Collateral inhibition, 111
Bmal1, 425 Combined pituitary hormone deficiency
BMP signaling, 113 (CPHD), 133
Bone, 84 Comorbidity, 199
Bone morphogenetic proteins (BMPs), 14, 85, Conditional mutant mice, 336
132 Congenital hypopituitarism, 133
Brain derived neurotrophic factor (BDNF), 381, Controversies in studies of astrocytes, 384
382 Core circadian oscillator, 109
Brown, 93 Co-repressors, 404
Corpora lutea (CL), 424
Corticosterone, 346, 355
C Corticotropes, 131
Caloric deficit, 354 Corticotropin-releasing factor (CRF), 85, 353,
Cardiovascular, 348 354
Catecholaminergic, 352, 356 Corticotropin-releasing hormone (CRH), 346,
Caudal brainstem, 346 361
Caudal medulla, 348, 349, 353, 357 Coverage of neuronal cell, 378
CCK-1 (aka CCK-A) receptors, 350 CpG island (CGI), 239
CCK-1 receptors, 352, 361 CRISPR/Cas9 system, 73
Cell fate, 327 Critical period, 398, 401
Cellular oscillators, 119 Cross talk, 94
Central canal, 347 Cry, 110
Central nucleus of the amygdala (CeA), 352 Cyclooxygenase enzymes, 407
Cephalic flexure, 69 Cyclopia, 50
Index 461

D Early enhancer, 147

Darkening of the skin, 78 Early life experience, 361
Data alignment, 216 Early sex steroids, 372
Databases, 229 EGF, 132
DAVID, 224, 225 Embryo, 425
Deep sequencing, 210 Embryonic, 350
Defeminization, 400 Embryonic day, 348
Dehydration, 376 Embryos, 74
Dehydration and rehydration, 376 Emetic, 352
Depression, 361 Emetic reflex, 352
Developmental boundaries, 194 Eminence, 84
Developmental expression patterns, 70 ENCODE Consortium, 224
Development of astrocytes, 370 Endocannabinoids, 408
Dicer, 245 Endocrine disrupters, 68
Diencephalic, 113 Endogenous CCK, 350
Diencephalic nuclei, 184 Endogenously, 351
Diet-induced obesity, 380 Energy balance, 383
Differentially methylated regions (DMRs), 268 Energy control, 380
Digestive viscera, 348 Energy deficiency, 354
Dil, 349, 350 Energy expenditure, 84, 271
Dio3 (Thyroxine 5-deiodinase), 279 Energy homeostasis, 268
Distal-less (DLX 1 and 2), 190 Energy metabolism, 380
Divergent morphologies, 384 Enrichment algorithms, 224
Diversity of astrocytes, 373 Enrichment analysis, 224
Dividing progenitors, 327 Ensheath GnRH terminals, 384
Dlk1 (Protein delta homolog 1), 279 Ensheath neuronal cell bodies, 369
Dll4, 79 Epigenetic, 115, 266, 361
Dlx, 70 Epigenetic chromatin regulators, 97
Dlx genes, 71 Epigenetic control, 208
Dlx2, 89 Epigenetic marks, 222
DNA demethylation, 240, 242, 254 Epigenetic mechanisms, 208, 221
DNA methylation, 221, 250, 268 Epigenetic modifications, 296, 297, 305, 306,
DNA methyltransferases (DNMT), 297, 306 308
DNMT enzymes, 239 Epigenetic repression, 222
DNMT1, 405 Epinephrine, 356
DNMT3, 405 Estradiol, 379
DNMT3a, 405 Estradiol-induced plasticity, 379
Dopamine (DA), 346, 424 Estrogen, 208, 212
Dopamine β hydroxylase (DBH), 353, 355, 356 Estrogen actions, 372
Dorsolateral hypothalamus, 112 Estrogen-induced synaptic remodeling, 379
Dorsal root ganglia (DRG), 346, 347 Estrogen receptor α, 79
Dorsomedial hypothalamus (DMD), 213 Estrogen receptor (ER), 373
Dorsomedial nucleus of the hypothalamus Estrous cycles, 378, 379, 384, 419, 420
(DMH), 382 Estrous cyclicity, 222, 223
Dwarfism, 97 Exogenous CCK, 350, 352
Dynorphin, 208 Experience-dependent plasticity, 360
Dysmasculinization, 401 External layer of the ME, 353
Dystrophin-71 (Dp-71), 378 Eye, 49

E F
EAP1, 209 FastQC, 217
Earlier puberty, 210 Fate-mapping, 37
462 Index

Feeding behavior, 380 GI vagal sensory, 361

Ferrets, 352, 354 GLI transcription factors, 32
Fertility, 197 Gli2, 144
Fetal development, 423–428 GliA, 42
Fibroblast growth factors (FGFs), 14, 85, 132 Glial coverage, 378
5-methylcytosine, 297 Glial ensheathment of synapses, 371
Floor plate, 50 Glial excitatory inputs, 208
Fluorescent reporter, 192 Glial inputs, 207
Follicle-stimulating hormone (FSH), 131, 416 Glial processes retract, 378
Food intake, 84 Glicentin, 356
Footprints, 158 Gliogenesis, 371
Formaldehyde-Assisted Isolation of Regulatory GliR, 42
Elements (FAIRE-seq), 210 Global Run-On sequencing (GRO-seq), 210
Fos, 351, 353–355, 357, 359, 360 Glossopharyngeal (IX), 348
FoxG1, 71 GLP1 receptor (GLP1-R), 354, 355
Frogs and mammals, 69 Glucagon, 356
Functional Significance, 197–198 Glucagon-like peptide 1 (GLP1), 354–361
Fushi tarazu, 149 Glucagon-like peptide 2 (GLP2), 356
Glucoprivic stimulation, 354
Glutamic acid decarboxylases 1 and 2 (GAD
G aka GAD67 and GAD65), 188
GABAergic neurons, 111 Gnas, 268
GABAergic populations, 78 Gnasxl, 269–276
GABA gradients, 194 GnRH nerve terminals, 384
GABAA receptors, 189 GnRH neuronal network, 208
GABAB receptors, 189 GnRH neurons, 207, 355, 378, 379
Gamma-aminobutyric acid (GABA), 182 GnRH pulsatility, 422
Ganglionic eminences, 334 GnRH pulse generator, 213
Gastric distension, 351, 352, 355, 360 GnRH terminals, 384
Gastrin-related peptide (Grp), 111, 115 Gonadal steroid, 378
Gastrointestinal (GI), 346, 355, 361 Gonadal steroid hormones, 373
Gata3, 148 Gonadal steroid receptors, 372
GATAD1, 227 Gonadotropes, 131
Gender, 396 Gonadotropin-releasing hormone (GnRH), 187,
Gene families, 73 194, 195, 207, 346, 354, 416, 420, 421
GeneMANIA, 224–226 Gonadotropin-releasing hormone
Gene networks, 73, 209–210 (GnRH)-neurons, 84
Gene network visualization, 224–229 G protein-coupled receptors, 189
Gene ontology database, 229 Growth hormone (GH), 84, 131, 351, 352, 361
Gene ontology (GO), 224, 225, 227, 229 Growth hormone-releasing hormone (GHRH),
Gene sets, 224 84, 346
Genetic lineage-tracing, 13 Gαs, 269
Genome, 217
Genome editing, 73
Genomic imprinting, 266 H
GFAP-positive astrocytes, 380 HES1, 190
GFAP-positive astrocytes in the arcuate nucleus Hesx1, 162
and VMH, 372 Heterogeneity of arcuate neurons, 84
GFAP-positive astrocytes in the SCN, 372 Heterogeneous populations of astrocytes, 374
GFAP-positive cells, 371, 382 Hindbrain, 360
Ghrelin, 351, 352, 380 Hippo, 132
GI tract, 348, 350 Histamine, 408
GI vagal afferents, 350–352 Histaminergic neurons, 106
Index 463

Histone acetylases, 298 Interoceptive, 346, 348, 355, 360

Histone acetylation, 251 Intrahypothalamic diagonal, 49
Histone acetyltransferase, 251 Intrauterine position (IUP) phenomenon, 402
Histone deacetylase (HDAC), 252, 296, 298 In vitro fertilization (IVF), 427
Histone deacetylation, 250 Isl1, 70
Histone eraser KDM1A, 222 Isolated growth hormone deficiency
Histone marks, 221 (IGHD), 133
Histone PTMs, 221 Isolated Xenopus nervous system, 73
Histones, 403
Holoprosencephaly, 49
Homeodomain, 144 J
Homeostasis, 22, 355 Jet lag, 414, 423
Homeotic genes, 190
Hormonal variations, 378
Hormone, 75 K
HPA axis, 360 Kiss1, 84, 209, 210, 213, 222, 223, 229
Human chorionic gonadotropin (hCG), 68 Kiss1 enhancer region, 223
Human LIN28B gene, 210 Kiss1 expression, 221
Humans, 352, 361 Kiss1 mRNA, 222
Hypersomnia, 106 Kisspeptin, 208, 223, 252, 418, 422, 424
Hypocretin, 121 Kisspeptin neurons, 209, 221
Hypogonadotropic hypogonadism, 247, 248 Kiss1 promoter, 221
Hypophagia, 351 KNDy, 419
Hypophysiotropic CRH, 346 KNDy neurons, 208
Hypophysis, 49 Knockout, 266
Hypothalamic basal plate, 37
Hypothalamic development, 69, 182–199
Hypothalamic dopaminergic cells, 74 L
Hypothalamic neuroepithelium, 115 Labor, 429
Hypothalamic neurogenesis, 112 Lactation, 376
Hypothalamic patterning, 190 Lactation/parturition, 376
Hypothalamus, 182, 207, 213, 218, 268 Lactotropes, 131
Hypothalamus–Pituitary–Gonadal (HPG) axis, LacZ, 87
212, 416 Lamella-like astrocytic processes, 376
Hypothalamus–pituitary–thyroid axis, 68 Lamina terminalis, 72
Lamina X, 347
Larvae, 74
I Larval development, 73
Immunohistochemistry, 73 Larval stages, 74
Imprinted gene, 266 Latent sex differences, 397
Imprinting control region (ICR), 268 Lateral ganglionic eminence, 59
Increased energy intake, 380 Lateral hypothalamus, 348
Infancy, 208, 209 Lateral hypothalamus area (LHA), 85
Inferior glossopharyngeal, 348 Lateral Progenitor Domain, 37
Infundibulum, 17, 38, 72 Lean, 276
Initiation of puberty, 208 Lef1, 113
In situ hybridization, 73 Leptin, 356, 357, 380, 382, 383
Insulin-like growth factor 1 (IGF1), 84 Leptin induces astrogenesis, 382
Interactomics databases, 229 Leptin receptors, 356, 357, 373
Interleukin (IL) 6, 381 Leptin signaling, 350
Interleukin 6 increase astrocyte proliferation, Let7 miRNAs, 210
377 LH pulsatility, 208
Internal clock, 78–79 LH pulses, 208
464 Index

LH surge, 418–422 Metabolic response, 380

Lhx3 knockout, 161 Metabolism, 373, 383
Library preparation, 213–221, 223 Metabolomics, 398
Limbic forebrain, 352, 355 Metamorphosis, 74
LIN28b, 210 Metamorphosis phase, 68
Lin28b-let7 posttranscriptional regulatory Methylcytosine dioxygenase, 297
pathway, 210 Methyltransferase, 239, 240, 242, 244, 247,
Lineage, 37 253, 296
Lineage-tracing, 20 Mice, 71, 113, 162, 350, 352, 354, 356, 357
Lipopolysaccharide (LPS), 355, 359, 360 Microglia, 368, 407
Lithium chloride (LiCl), 355, 359, 360 Micro-RNAs, 210, 245
Locus coeruleus (A6 cell group), 353 Migration, 185
Locus Control Region (LCR), 159 Mini-puberty, 212, 401
Lordosis, 396 Molecular clockwork, 420, 425, 429
Luteinizing hormone (LH), 131, 207, 208, 212, Molecular genetics, 194–197
354, 394, 416 Molecular switch, 209
Monoallelic, 269
Monoallelic expression, 277
M Morphological modifications, 378
MAGEL2, 277 Morphological responses, 379
Magnocellular, 6, 351 Morphology of astrocytes, 372
Magnocellular and parvocellular Mouse, 115, 356
neurosecretory cell populations, 71
Magnocellular neurons, 377, 378
Magnocellular OT, 346, 354 N
Magnocellular system, 378 NA neurons, 353
Male rhesus monkeys, 222 Narcoleptic patients, 106
Mammillary (M) regions, 71 NDN, 277
Mammillary, 8, 9, 33 Negative feedback, 109
Mast cell, 408 Neonatal, 350
Master clock, 109 Neural plate, 35
Maternal, 360 Neural precursors, 113, 120
Maternal care, 361 Neural progenitor, 113
Maternal high fat intake, 381 Neural Shh, 46
Maternal obesity, 381 Neural stem cells, 85
Matrin-3, 147 Neural tube flexure, 69
Maturation of astrocytes, 371 Neuraxis, 113
MCH-positive LH neurons, 108 Neurodevelopmental psychiatric disorders, 410
MeCP2, 405 Neuroendocrine, 5
Medial preoptic area (mPOA), 346, 352, 354 Neuroendocrine functions, 375
Medial Progenitor Domain, 33 Neuroepigenetics, 404
Medial basal hypothalamus (MBH), 222 Neurogenesis, 115, 368
Median, 6 Neurogenic, 13
Median eminence (ME), 72, 213, 346 Neurohypophysis, 72, 75
Medullary RF, 348, 352, 354 Neurokinin B (NKB), 208
Melanocyte-stimulating hormone, 131 Neuromeres, 71
Melanopsin, 415 Neuronal ensheathment, 378
Melanotropes, 131 Neuronal migration, 368
Melanotropic cells, 78 Neuronal-astrocyte structural organization, 378
Melatonin, 426, 427, 430 Neurons, 5
Meninges, 408 Neuron-specific enhancer, 238
Menstrual cycles, 414 Neuropeptides, 75
Metabolic circuits, 382 Neuropeptide Y (NPY), 78, 84, 353
Index 465

Neuropil, 408 Oxyntomodulin, 356

Neurosecretory preoptic area, 54 Oxytocin (OT), 346, 351–355, 357, 424
Neurotransmitter, 407
Newborn rats, 350
Next-generation sequencing (NGS), 216 P
NG2 cells, 368 Paramagnetic dynabeads, 218
NGS technologies, 221–223 Paraventricular, 71
Nkx2, 70 Paraventricular area (Pa), 71, 75
Nkx2.1, 71, 85, 87 Paraventricular nucleus (PVN), 48, 196, 271,
Nkx2.2, 71 346, 350–352, 354, 356, 357, 359–361,
Node, 50 376, 382
Nodose, 349 Pars nervosa, 75
Nodose ganglia, 348, 349 Parturition, 429
Non-rapid eye movement (NREM), 106 Parvocellular, 6, 351
Noradrenergic (NA), 353–357, 361 Parvocellular CRH, 354
Norepinephrine, 355, 356, 361 Parvocellular endocrine neurons, 346
Notch, 132 Parvocellular GHRH, 352
Notch2, 144 Parvocellular GnRH, 354
Notch/Delta signaling, 113 Paternal behavior, 397
Notochord, 32, 113 Paternally expressed gene 3 (Peg3), 279
Nr1d1, 110 Pax7 / , 162
NR5A1, 162 Pax7+ cells, 79
NTS visceral sensory, 352 PCR, 280
Nuclear Receptor Co-Repressor (NCoR), 142 PCR artifacts, 217
Nuclei, 6 Penk, 111
Nucleogenesis, 115 Peptidergic NTS, 352
Nucleus of the solitary tract (NTS), 346, 348– Per, 110
357, 360, 361 Peripubertal girls, 208
Periventricular nuclei, 71
Phagocytosis, 408
O Pharmacology, 194–197
Obese, 383 Phase-advanced, 112
Obesity, 271, 380 Phase-delayed, 112
Obesity-associated, 381 Phase synchrony and stabilization, 111
Oligodendrocytes, 59, 85, 368 Phenylethanolamine N-methyltransferase
Opsin-positive neurons, 75 (PNMT), 356
Optic disk, 47 Pineal gland, 426, 430
Optogenetic activation of astrocytes, 385 Pituicytes, 7, 23, 131
Optogenetic stimulation, 385 Pituitary organoids, 162
Optogenetic techniques to interrogate Pitx1, 161
astrocyte, 385 Placenta, 426
Orexinergic, 108 Plasticity, 360
Orexinergic neurons, 106 Polycomb, 210, 223, 225, 226
Orexins, 75 Polycomb Group (PcG), 221
Organizing signals, 85 Pontine parabrachial nucleus (PBN), 348, 352
Organotypic cultures, 193 Portal system, 75
Orthopedia (Otp), 70, 75, 190 Portal vasculature, 346, 354
Oscillations, 296, 298, 300, 302, 306, 308 Positive feedback, 208
Osmotic dehydration, 360 Posterior lobe, 129
Osmotic homeostasis, lactation, and Posterior lobe of the pituitary, 6
reproduction, 375 Posterior pituitary, 346, 351, 352
OTx2, 70 Postmitotic maturation, 371
Ovarian steroids, 384 Postnatal, 271, 350
Oviduct, 425 Postnatal maturation of astrocytes, 371
Ovulation, 354, 419–423 Postsynaptic melanotrope plasticity, 78
466 Index

Posttranscriptional repressor, 210 Q

POU1F1, 162 qPCR, 218
PPG-YFP mice, 356 Quality control, 216, 217
Prader–Willi syndrome (PWS), 199, 268
Pre-autonomic hypothalamic neurons, 352
Prechordal plate, 35 R
Precise organization, 383 Radial glia, 187, 370, 376, 382, 383
Precision nuclear Run-On sequencing Radial glial cells, 370
(PRO-seq), 210 Radial glial fibers, 195
Pregnancy, 423–429 Rapid eye movement (REM), 106
Pregnant women, 68 Rat, 212
Premammillary, 40 Rathke’s, 17
Prenatal, 350 Rathke’s pouch, 49
Preoptic, 33 Rats, 346, 349–357, 359–361
Preoptic area, 378, 379 Rax, 87, 191
Preoptic/anterior, 9 Regulatory gene expression patterns, 71
Preoptic region, 71 Regulatory genes, 70
Pre-ovulatory LH surge, 354 Reproduction, 378, 379, 383
Preovulatory surge, 209 Reproductive biology, 69
Preproglucagon (PPG), 356 Restructuring of glial-neuronal contacts, 376
Presynaptic SMIN plasticity, 78 Retinoic acid, 132
Primary astrocytic projections, 380 Retractable end feet, 384
Primary wake-promoting neurons, 106 Retraction of astrocytic processes, 376
Progenitor cells, 8, 383 Retrochiasmatic Area, 38
Progenitor pool, 85 RFamide-related peptide, 417
Progenitor zones, 87 RFamide-related peptides (RFRPs)-3, 423
Projections, 369 Rhesus macaque monkeys, 352
Prolactin, 84, 131, 346, 423 Rhesus macaques, 212, 222, 354
Prolactin-releasing peptide (PrRP), 355 Rhythmic epigenetics, 296
Proliferation, 382 Ribo-Zero rRNA removal, 215
The proliferation and maturation of RNA Integrity Number (RIN), 213
astrocytes, 372 RNAscope, 285
Proliferation of astrocytes, 371 RNA-seq, 210, 213–217, 221, 222, 284
Proliferative capacity, 381 Rodents, 208, 346, 351, 355
Proneural TFs, 85 Role of astrocytes, 385
Pro-opiomelanocortin (POMC), 78, 84, 131 Rora, 110
Prophet of Pit1 (PROP1), 162 Rorb, 110
Prosomeres, 70 Rostral Diencephalic Ventral Midline
Prosomeres P1–P3, 71 (RDVM), 33
Prosomeric model, 69–73
Protein kinase A (PKA), 407
Protein–DNA interactions, 218 S
p.S83P, 144 Sanger sequencing, 280
Pseudohypoparathyroidism (PHP), 271 SATB1, 147, 148
Pubertal timing, 208 Satiety signals, 350, 351
Puberty, 207 Schizophrenia, 410
Pulmonary, 348 Seasonal, 296, 300, 305, 308, 309
Pulsatile GnRH release, 208 Seasonal adaptation, 75
Pulsatile LH release, 208 Seasonal epigenetic modifications, 308
Purple, 136 Seasonal oscillations, 300
PVN neuronal firing, 354 Secondary sexual characteristics, 208
Pyrosequencing, 283 Secreted factor, 182
Sensitive period, 360, 401
Index 467

Sensory cues, 319 Suprachiasmatic melanotrope-inhibiting

Sex differences, 372, 379, 394 neurons (SMINs), 78
Sex differences in astrocyte morphology, 372 Suprachiasmatic nucleus (SCN), 78, 109, 371,
Sex differences in astrocyte number and 415, 419, 420, 422, 424
morphology, 372 Supramammillary, 40
Sex steroid environment, 379 Supramammilary nucleus, 106
Sex steroids, 372, 379 Supraoptic, 71
Sex-selective effects, 198 Supraoptic nucleus (SON), 346, 351–354, 360,
Sexually dimorphic nucleus of the preoptic area 376–378, 381
(SDN-POA), 394 Sympathetic nervous system (SNS), 271
Sexual maturation, 198 Synaptic organization/astrocytic coverage, 380
Signaling, 113 Synaptic plasticity in metabolic circuits, 382
Single-cell, 283–288 Synaptogenesis, 368, 380
Single cell RNA-seq, 21 SynCAM1, 210
Single cell RNA-seq assay, 374
Single cell sequencing, 162
Single nucleotide polymorphism RNA T
fluorescent in situ hybridization TAC3, 210, 222
(SNP-FISH), 285 Tachykinin receptors, 208
Single nucleotide polymorphism (SNP), 225, Tadpole, 74
280 Tanycyte end feet, 384
Six3, 70 Tanycytes, 7, 23, 38, 368, 370, 371, 383
Skeleton photoperiods, 119 Tanycytic, 384
Sleep, 208 Tanycytic processes, 384
Sleep pressure, 106 Tanycytic processes retract, 384
Sleep-promoting neurons, 107 TBX19, 162
Sleep/wake transitions, 106, 109 Telencephalic, 12
SNURF/SNRPN/SNORD115-116, 277 Telencephalic–diencephalic border, 332
Solitary tract, 348, 349 Telencephalic neuroepithelium, 113
Somatostatin (SS), 85, 346 Telencephalic Progenitor Domain, 37
Somatotropes, 131 Testicular hormones, 394
Sonic hedgehog (Shh), 14, 71, 85, 113 Tet enzymes, 242–245
Sox2 deficiency impairs, 161 Thermogenesis, 85
Specific Cre lines, 96 Three wave hypothesis, 41
Specific rewiring, 383 Thyroid axis, 383
Specific synaptic rewiring, 383 Thyroid-stimulating hormone (TSH), 131, 352
Specification of hypothalamic postmitotic Thyrotropes, 131
populations, 75 Thyrotropin releasing hormone (TRH), 75, 78,
Spinal cord dorsal horn, 347 346, 353, 354
Spinal nucleus of the bulbocavernosus Timetable of development, 69
(SNB), 394 Tissue explants, 193
Spinosolitary tract, 348 Topological relationships, 69
Split-Cre system, 96 Transcriptional profile, 374
Stem cell, 25 Transcriptional profiles of astrocytes, 374
Stress, 352, 355, 360 Transcriptional repressors, 221
Stressors, 355 Transcription factor binding, 221
Stress-related physiology and behaviors, 198 Transcription factors (TFs), 85, 112, 327
Stress responses, 352 Transcript quantitation, 216
Structural changes, 376 Trans-synaptic inhibitory control, 209
Subparaventricular zone, 71 TRG network, 209
Subsets of astrocytes, 374 TRH cells, 79
Suprachiasmatic, 71 Trimethylation, 222
Suprachiasmatic hypothalamus, 78–79 Trithorax, 222, 223, 226
468 Index

Trithorax complex, 210 Ventrolateral medulla (VLM), 352–356, 360

Tuberal, 9, 33 Ventromedial hypothalamus (VMH), 213
Tuberal region, 71 Ventromedial nucleus (VMN), 38, 79, 85, 195,
Tuberal region molecular expression pattern, 79 383
Tuberal region of anurans, 79 Vesicular glutamatergic transporter, 353
Tuberoinfundibular dopamine (TIDA)-neurons, Vgf, 111
84, 346 Vip, 111, 115
Tuberomammillary nucleus (TMN), 106 Visceral/interoceptive, 355
Tumor-Related Genes (TRGs), 209 Visceral sensory, 351, 352, 357, 360–361
Tyrosine hydroxylase (TH), 78, 84 Visceral sensory signals, 346, 353
Visceral stimuli, 353, 360
Visceral stressors, 353
U Visualizing astrocytes, 384
UBE3A, 277 Voltage-gated potassium channel
Undernutrition, 383 gene Kcnq1, 279
Upper-echelon genes, 209
Urocortin 1, 78
Uterus, 426 W
Wakefulness, 106
Wake-promoting neurons, 108
V Whole brain imaging, 73
Vagal, 348, 353 Whole-genome bisulfite sequencing
Vagal afferent firing, 351 (WGBS), 283
Vagal afferent neurons, 348 Wingless/integrins (Wnts), 14, 85, 132
Vagal afferents, 348, 349, 351, 357 Women, 354
Vagally mediated, 353
Vagal sensory, 349–351, 354, 359
Vagal sensory afferents, 346 X
Vagal sensory neurons, 348, 349 Xenopus laevis frogs, 68
Vago-vagal reflexes, 349, 350 XLαs, 269
Vagus nerve, 351
Vagus (X), 348
Vascular networks, 23
Vasoactive intestinal peptide (VIP), 420, 423 Y
Vasopressin (AVP), 111, 115, 346, 351–354, Yellow fluorescent protein (YFP), 356
405, 420
Vasopressinergic, 397
Ventral floor plate, 113 Z
Ventral forebrain, 33 ZEB1, 148
Ventral noradrenergic ascending bundle Zeb2, 144
(VNAB), 354 Zebra fish, 113
Ventricular zone, 85, 349 Zona incerta (ZI), 109