Methods in

Molecular Biology 1905

Naoki Tanimizu Editor

Hepatic
Stem Cells
Methods and Protocols
METHODS IN MOLECULAR BIOLOGY

Series Editor
John M. Walker
School of Life and Medical Sciences
University of Hertfordshire
Hatfield, Hertfordshire, AL10 9AB, UK

For further volumes:

http://www.springer.com/series/7651
Hepatic Stem Cells

Methods and Protocols

Edited by

Naoki Tanimizu
Sapporo Medical University, Sapporo, Hokkaido, Japan
Editor
Naoki Tanimizu
Sapporo Medical University
Sapporo, Hokkaido, Japan

ISSN 1064-3745 ISSN 1940-6029 (electronic)

Methods in Molecular Biology
ISBN 978-1-4939-8960-7 ISBN 978-1-4939-8961-4 (eBook)
https://doi.org/10.1007/978-1-4939-8961-4

Library of Congress Control Number: 2018962407

© Springer Science+Business Media, LLC, part of Springer Nature 2019

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,
express 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.

Cover Illustration: Biliary epithelial cells (BECs) labeled with tdTomato proliferate to expand the CK19+ biliary epithelial
tissue structure upon injury, indicating that pre-existing BECs contribute to the “ductular reaction”. (The image related
to Chapter 7 has been provided by Dr. Kenji Kamimoto).

This Humana Press imprint is published by the registered company Springer Science+Business Media, LLC, part of
Springer Nature.
The registered company address is: 233 Spring Street, New York, NY 10013, U.S.A.
Preface

The liver contains two types of epithelial cells, namely, hepatocytes and cholangiocytes
(biliary epithelial cells). Therefore, liver stem/progenitor cells (LPCs) are defined as bipo-
tential cells that can differentiate into hepatocytes and cholangiocytes. Hepatoblasts, fetal
LPCs, split into hepatocytes and cholangiocytes in mid-gestation. Bipotential LPCs also
have been isolated from neonatal liver. On the other hand, it is still controversial whether
bipotential LPCs contribute to maintenance of cellular homeostasis in healthy adult liver.
Recent works demonstrated that both hepatocytes and cholangiocytes are heterogeneous
cell populations and may contain committed progenitors. This book first contains methods
for isolation and expansion of bipotential LPCs from fetal and neonatal liver, for adult
clonogenic cholangiocytes and for hepatocyte progenitors.
The liver has the strong regenerative capability: it has been generally considered that
self-duplication of hepatocytes or cholangiocytes compensates the lost tissue after acute
injuries, whereas LPCs are activated and supply new hepatocytes after chronic injuries.
However, recent works demonstrated that dedifferentiation and lineage conversion of
liver epithelial cells also contribute to liver regeneration. These results highlight that liver
regeneration can be achieved by multiple modes of cellular responses. In the second part,
this book contains experimental methods for identifying and characterizing progenitor cells
involved in liver regeneration in vivo as well as those for investigating molecular mechanisms
regulating progenitor cell-driven liver regeneration.
It is crucial to expand functional hepatocytes and cholangiocytes ex vivo for implement-
ing regenerative medicine such as cell therapy and for establishing drug screening systems.
Recently, novel culture methods have been established to supply hepatocytes and/or
cholangiocytes from pluripotent stem cells as well as somatic terminally differentiated
cells. In addition, it was demonstrated that in vivo conversion of fibrogenic stellate cells
into hepatocytes could be useful to ameliorate liver fibrosis. In the third part, this book
contains culture methods for generating hepatocytes and cholangiocytes from multiple
cellular sources and a technique converting fibroblasts into hepatocytes in vivo.
Liver epithelial cells form three-dimensional (3D) tissue structures, which are indispens-
able for the liver to perform their physiological functions. Facing various types of injuries,
3D hepatic tissue structures are dynamically rearranged to avoid fatal liver damages. Thus,
this book also contains experimental methods for the reconstitution of 3D tissue structures
ex vivo as well as for the investigation of the dynamic structural changes induced after
injuries and during regeneration.
Fibrosis is the major pathology of the liver. Another issue is liver cancers. It is important
to establish a novel strategy to resolve liver fibrosis and to understand etiologies of hepato-
cellular (HCC) and cholangiocellular carcinomas (CC). Therefore, the last part of this book
includes methods resolving hepatic fibrosis by bone marrow transplantation as well as two
novel mouse models of HCCs and CCs.

v
vi Preface

In addition to experimental protocols, each author provides a great introduction sum-

marizing the background of each topic. This book would help researchers understand the
current concept about hepatic stem/progenitor cells and perform basic and translational
works related to liver biology and pathophysiology.

Sapporo, Hokkaido, Japan Naoki Tanimizu

Contents

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
Contributors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix

PART I ISOLATION OF PROGENITOR CELLS

1 Long-Term Culture of Mouse Fetal Hepatic Stem/
Progenitor Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Atsunori Tsuchiya and Shuji Terai
2 Isolation of Bipotential Liver Progenitor Cells from Neonatal
Mouse Liver. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Naoki Tanimizu
3 Identification and Isolation of Clonogenic Cholangiocyte in Mouse. . . . . . . . . . . 19
Bin Li, Craig Dorrell, Pamela S. Canady, and Leslie Wakefield
4 Isolation and Expansion of Rat Hepatocytic Progenitor Cells. . . . . . . . . . . . . . . . . 29
Junichi Kino, Norihisa Ichinohe, Masayuki Ishii, and Toshihiro Mitaka

PART II CHARACTERIZATION OF LIVER PROGENITORS IN VIVO

5 Genetic Lineage Tracing of Biliary Epithelial Cells . . . . . . . . . . . . . . . . . . . . . . . . . . 45

Teresa Rubio-Tomás, Beatriz Aguilar-Bravo, and Pau Sancho-Bru
6 Specific Labeling and Lineage Tracing of Periportal Hepatocytes
Using Two-Step Genetic Recombination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
Nicola de Prisco, Eleanor Stout, and Joan Font-Burgada
7 Analysis for the Heterogeneity of Liver Progenitor Cells . . . . . . . . . . . . . . . . . . . . . 71
Kenji Kamimoto
8 Chemical Screening Using a Zebrafish Model for Liver Progenitor
Cell-Driven Liver Regeneration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Sungjin Ko and Donghun Shin

PART III GENERATION OF HEPATOCYTES, CHOLANGIOCYTES,

AND THEIR PROGENITORS

9 Conversion of Fibroblasts to Hepatocytes In Vitro . . . . . . . . . . . . . . . . . . . . . . . . . . 93

Pengyu Huang, Lulu Sun, Ludi Zhang, and Lijian Hui
10 Conversion of Fibroblasts to Hepatocyte-Like Cells In Vivo . . . . . . . . . . . . . . . . . 103
Guangqi Song, Qinggong Yuan, Zhen Dai, Hsin-Chieh Tsay,
Xizhong Shen, Michael Ott, and Amar Deep Sharma
11 Chemically Induced Liver Progenitors (CLiPs): A Novel Cell Source
for Hepatocytes and Biliary Epithelial Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
Takeshi Katsuda and Takahiro Ochiya

vii
viii Contents

12 Induction of Functional Hepatocytes from Human iPSCs . . . . . . . . . . . . . . . . . . . 131

Taketomo Kido and Yuta Koui
13 Culture System of Bile Duct-Like Cystic Structures Derived from
Human-Inducible Pluripotent Stem Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
Akihide Kamiya, Kazuya Anzai, Kota Tsuruya, and Hiromi Chikada

PART IV RECONSTITUTION OF LIVER TISSUE STRUCTURES

14 Generation of Hepatic Tissue Structures Using Multicellular

Spheroid Culture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
Fumiya Tao, Hirotaka Mihara, and Nobuhiko Kojima
15 Reconstruction of Hepatic Tissue Structures Using Interstitial
Flow in a Microfluidic Device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
Ryo Sudo
16 Generation of Hepatic Organoids with Biliary Structures . . . . . . . . . . . . . . . . . . . . 175
Takeshi Katsuda, Takahiro Ochiya, and Yasuyuki Sakai
17 Analysis for Remodeling of Hepatic Tissue Structures in 3D
During Regeneration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
Kota Kaneko

PART V LIVER INJURY MODELS

18 Canine Liver Fibrosis Model to Assess the Functions of Infused

Autologous Bone Marrow-Derived Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
Taro Takami, Kenji Tani, Yasuho Taura, and Isao Sakaida
19 A Rodent Model for Cell Transplantation of Hepatic Progenitor Cells . . . . . . . . 211
Sei Kakinuma and Akihide Kamiya
20 Mouse Model for Hepatocellular Carcinoma and Cholangiocarcinoma
Originated from Mature Hepatocytes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
Masahiro Yamamoto, Bing Xin, and Yuji Nishikawa
21 Mouse Model for Cholangiocarcinoma from Peribiliary Glands. . . . . . . . . . . . . . . 237
Hayato Nakagawa, Nobumi Suzuki, and Kazuhiko Koike

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
Contributors

BEATRIZ AGUILAR-BRAVO  Institut d’Investigacions Biomèdiques August Pi i Sunyer

(IDIBAPS), Barcelona, Spain
KAZUYA ANZAI  Department of Gastroenterology, Tokai University School of Medicine,
Isehara, Kanagawa, Japan
PAMELA S. CANADY  Oregon Stem Cell Center, Oregon Health and Science University,
Portland, OR, USA
HIROMI CHIKADA  Department of Molecular Life Sciences, Tokai University School of
Medicine, Isehara, Kanagawa, Japan
ZHEN DAI  Department of Gastroenterology, Hepatology and Endocrinology, Hannover
Medical School, Hannover, Germany
NICOLA DE PRISCO  Cancer Biology Program, Fox Chase Cancer Center, Philadelphia, PA,
USA
CRAIG DORRELL  Oregon Stem Cell Center, Oregon Health and Science University,
Portland, OR, USA
JOAN FONT-BURGADA  Cancer Biology Program, Fox Chase Cancer Center, Philadelphia,
PA, USA
PENGYU HUANG  School of Life Science and Technology, Shanghai Tech University, Shanghai,
China
LIJIAN HUI  School of Life Science and Technology, Shanghai Tech University, Shanghai,
China; State Key Laboratory of Cell Biology, CAS Center for Excellence in Molecular Cell
Science, Shanghai Institute of Biochemistry and Cell Biology, Chinese Academy of Sciences,
University of Chinese Academy of Sciences, Shanghai, China
NORIHISA ICHINOHE  Department of Tissue Development and Regeneration, Research
Institute for Frontier Medicine, Sapporo Medical University School of Medicine, Sapporo,
Japan
MASAYUKI ISHII  Department of Tissue Development and Regeneration, Research Institute
for Frontier Medicine, Sapporo Medical University School of Medicine, Sapporo, Japan;
Department of Surgery, Surgical Oncology and Science, Sapporo Medical University School
of Medicine, Sapporo, Japan
SEI KAKINUMA  Department of Gastroenterology and Hepatology, Tokyo Medical and Dental
University, Tokyo, Japan; Department of Liver Disease Control, Tokyo Medical and Dental
University, Tokyo, Japan
KENJI KAMIMOTO  Department of Developmental Biology, Washington University School of
Medicine in St. Louis, St. Louis, MO, USA
AKIHIDE KAMIYA  Department of Molecular Life Sciences, Tokai University School of
Medicine, Isehara, Kanagawa, Japan; Center for Matrix Biology and Medicine, Graduate
School of Medicine, Tokai University, Isehara, Kanagawa, Japan
KOTA KANEKO  Department of Pathology, School of Medicine, University of California, San
Diego, CA, USA
TAKESHI KATSUDA  Division of Molecular and Cellular Medicine, National Cancer Center
Research Institute, Tokyo, Japan
TAKETOMO KIDO  Laboratory of Cell Growth and Differentiation, Institute of Molecular
and Cellular Biosciences, The University of Tokyo, Tokyo, Japan

ix
x Contributors

JUNICHI KINO  Department of Tissue Development and Regeneration, Research Institute for
Frontier Medicine, Sapporo Medical University School of Medicine, Sapporo, Japan;
Tokushima Research Institute, Otsuka Pharmaceutical Co. Ltd., Tokushima, Japan
SUNGJIN KO  Department of Developmental Biology, McGowan Institute for Regenerative
Medicine, Pittsburgh Liver Research Center, University of Pittsburgh, Pittsburgh, PA, USA
KAZUHIKO KOIKE  Department of Gastroenterology, The University of Tokyo, Tokyo, Japan
NOBUHIKO KOJIMA  Department of Life and Environmental System Science, Graduate
School of Nanobioscience, Yokohama City University, Yokohama, Japan
YUTA KOUI  Laboratory of Cell Growth and Differentiation, Institute of Molecular and
Cellular Biosciences, The University of Tokyo, Tokyo, Japan
BIN LI  Oregon Stem Cell Center, Oregon Health and Science University, Portland, OR,
USA
HIROTAKA MIHARA  Department of Life and Environmental System Science, Graduate
School of Nanobioscience, Yokohama City University, Yokohama, Japan
TOSHIHIRO MITAKA  Department of Tissue Development and Regeneration, Research
Institute for Frontier Medicine, Sapporo Medical University School of Medicine, Sapporo,
Japan
HAYATO NAKAGAWA  Department of Gastroenterology, The University of Tokyo, Tokyo, Japan
YUJI NISHIKAWA  Division of Tumor Pathology, Department of Pathology, Asahikawa
Medical University, Asahikawa, Hokkaido, Japan
TAKAHIRO OCHIYA  Division of Molecular and Cellular Medicine, National Cancer Center
Research Institute, Tokyo, Japan
MICHAEL OTT  Department of Gastroenterology, Hepatology and Endocrinology, Hannover
Medical School, Hannover, Germany
TERESA RUBIO-TOMÁS  Institut d’Investigacions Biomèdiques August Pi i Sunyer
(IDIBAPS), Barcelona, Spain
YASUYUKI SAKAI  Institute of Industrial Science, The University of Tokyo, Tokyo, Japan
ISAO SAKAIDA  Department of Gastroenterology and Hepatology, Yamaguchi University
Graduate School of Medicine, Yamaguchi University, Ube, Yamaguchi, Japan
PAU SANCHO-BRU  Institut d’Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS),
Barcelona, Spain; Centro de Investigacion Biomédica en Red de Enfermedades Hepáticas y
Digestivas (CIBERehd), Barcelona, Spain
AMAR DEEP SHARMA  Department of Gastroenterology, Hepatology and Endocrinology,
Hannover Medical School, Hannover, Germany
XIZHONG SHEN  Department of Gastroenterology, Zhongshan Hospital, Fudan University,
Shanghai, China
DONGHUN SHIN  Department of Developmental Biology, McGowan Institute for
Regenerative Medicine, Pittsburgh Liver Research Center, University of Pittsburgh,
Pittsburgh, PA, USA
GUANGQI SONG  Department of Gastroenterology, Zhongshan Hospital, Fudan University,
Shanghai, China
ELEANOR STOUT  Cancer Biology Program, Fox Chase Cancer Center, Philadelphia, PA,
USA
RYO SUDO  Department of System Design Engineering, Keio University, Yokohama, Japan
LULU SUN  State Key Laboratory of Cell Biology, CAS Center for Excellence in Molecular
Cell Science, Shanghai Institute of Biochemistry and Cell Biology, Chinese Academy of
Sciences, University of Chinese Academy of Sciences, Shanghai, China
NOBUMI SUZUKI  Department of Gastroenterology, The University of Tokyo, Tokyo, Japan
 Contributors xi

TARO TAKAMI  Department of Gastroenterology and Hepatology, Yamaguchi University

Graduate School of Medicine, Yamaguchi University, Ube, Yamaguchi, Japan
KENJI TANI  Department of Veterinary Surgery, Joint Faculty of Veterinary Medicine,
Yamaguchi University, Yamaguchi, Yamaguchi, Japan
NAOKI TANIMIZU  Department of Tissue Development and Regeneration, Research Institute
for Frontier Medicine, Sapporo Medical University School of Medicine, Sapporo, Hokkaido,
Japan
FUMIYA TAO  Department of Life and Environmental System Science, Graduate School of
Nanobioscience, Yokohama City University, Yokohama, Japan
YASUHO TAURA  Department of Veterinary Surgery, Joint Faculty of Veterinary Medicine,
Yamaguchi University, Yamaguchi, Yamaguchi, Japan
SHUJI TERAI  Division of Gastroenterology and Hepatology, Graduate School of Medical and
Dental Science, Niigata University, Niigata, Japan
HSIN-CHIEH TSAY  Department of Gastroenterology, Hepatology and Endocrinology,
Hannover Medical School, Hannover, Germany
ATSUNORI TSUCHIYA  Division of Gastroenterology and Hepatology, Graduate School of
Medical and Dental Science, Niigata University, Niigata, Japan
KOTA TSURUYA  Department of Gastroenterology, Tokai University School of Medicine,
Isehara, Kanagawa, Japan
LESLIE WAKEFIELD  Oregon Stem Cell Center, Oregon Health and Science University,
Portland, OR, USA
BING XIN  Division of Tumor Pathology, Department of Pathology, Asahikawa Medical
University, Asahikawa, Hokkaido, Japan
MASAHIRO YAMAMOTO  Division of Tumor Pathology, Department of Pathology, Asahikawa
Medical University, Asahikawa, Hokkaido, Japan
QINGGONG YUAN  Department of Gastroenterology, Hepatology and Endocrinology,
Hannover Medical School, Hannover, Germany
LUDI ZHANG  State Key Laboratory of Cell Biology, CAS Center for Excellence in Molecular
Cell Science, Shanghai Institute of Biochemistry and Cell Biology, Chinese Academy of
Sciences, University of Chinese Academy of Sciences, Shanghai, China
 Part I

Isolation of Progenitor Cells

 Chapter 1

Long-Term Culture of Mouse Fetal Hepatic

Stem/Progenitor Cells
Atsunori Tsuchiya and Shuji Terai

Abstract
Mouse fetal liver includes abundant hepatic stem/progenitor cells (HSPCs). Easy expansion with passage of
HSPCs is necessary to obtain steady data. However, it is often difficult to enrich only HSPCs, and HSPCs
can die when usual trypsin is used for replating. Here, we introduce serum-free long-term culture with
passage of HSPCs using fetal mouse liver without a cell sorter.

Key words Hepatic progenitor cell, Spheroid, Serum-free culture, Colony, Mouse

1 Introduction

Mouse fetal liver includes abundant hepatic stem/progenitor cells

(HSPCs) surrounding hematopoietic cells, endothelial cells, mes-
enchymal cells, and other cells. Owing to the high expansion ability
of HSPCs, it is very convenient to use these cells after expansion for
in vitro and in vivo studies. There are two main problems in long-
term culture of HSPCs. First, while HSPCs have a high expansion
ability, other cells, such as hematopoietic cells, endothelial cells, and
mesenchymal cells, also have high expansion ability. Sometimes
when fetal liver cells are cultured without enrichment with serum,
mesenchymal cells occupy most of the culture dish. Adopting a cell
sorter is a very good approach to isolate only HSPCs by using the
markers CD45 Ter119 c-kit CD29+CD49f+/low [1, 2], epithelial
cell adhesion antigen (EpCAM) [3], CD13 [4, 5], and delta-like
1 homolog (DLK1) [6, 7]. However, culture from single cells is not
easy, and not all laboratories can apply this approach. Second,
HSPCs can die when usual trypsin is used for replating. To over-
come these problems and achieve long-term culture of HSPCs, we
developed an approach involving spheroid culture followed by
two-dimensional (2D) culture with replating using diluted trypsin
[8] (Fig. 1).

Naoki Tanimizu (ed.), Hepatic Stem Cells: Methods and Protocols, Methods in Molecular Biology, vol. 1905,
https://doi.org/10.1007/978-1-4939-8961-4_1, © Springer Science+Business Media, LLC, part of Springer Nature 2019

3
4 Atsunori Tsuchiya and Shuji Terai

Culture Scheme Medium

Dissociated E13.5 fetal liver cells DMEM/F12 (1:1)
+B27+ITS+
For 6 days 20 ng/ml EGF+
20 ng/ml bFGF+
Floating culture 20 ng/ml HGF
Hepatic spheroid formation

Plating culture Monolayer colony formation

DMEM/F12 (1:1)
+B27+ITS+
20 ng/ml EGF+
Subpassage with maintaining With trypsin diluted 20 ng/ml bFGF+
colony formation by KSR and CaC12 10 ng/ml HGF

E13.5 fetal liver cell

Expansion of fetal hepatic stem/progenitor cells Hepatic spheroid

Flow cytometry Ultra-low attachment plate

Immunocytochemistry +OSM etc.
RT-PCR Expanding monolayer colony
Transplantation

Analysis Differentiation Type-1 collagen coated

dish

Fig. 1 Flow diagram presenting our culture system

2 Materials

2.1 Isolation of 1. Mice: A pregnant C57BL/6 mouse at day 13.5 of gestation

HSPCs was used as a source of fetal liver. Other stains were not
attempted for culture.
2. Tweezers.
3. Tissue culture dishes (100 mm, polystyrene).
4. Phosphate-buffered saline (PBS).
5. 1 mL pipette tips.

2.2 For Spheroid 1. Medium: Dulbecco’s Modified Eagle Medium/F12

Culture (Floating (DMEM/F12).
Culture) 2. Supplement: B27 supplement (50), ITS mixture (100), and
HEPES (final concentration: 10 nmol/L) (see Note 1).
3. Antibiotics: Penicillin-streptomycin liquid (see Note 1).
4. Growth factors: Epidermal growth factor (EGF; final concen-
tration 20 ng/mL), hepatocyte growth factor (HGF; final
concentration 20 ng/mL), and basic fibroblast growth factor
(bFGF; final concentration 20 ng/mL).
5. Dish: Six-well ultralow attachment plates (see Note 2).
 Culture of Mouse Fetal HSPCs 5

2.3 For 2D Culture 1. Medium: Same as that in the spheroid culture.

(Plating Culture) 2. Supplement: Same as that in the spheroid culture.
3. Antibiotics: Same as that in the spheroid culture.
4. Growth factors: Same as that in the spheroid culture. It is
possible to reduce the concentration of each growth factor
(20–10 ng/mL).
5. Dish: Six-well collagen type I-coated plates (see Note 2).

2.4 Diluted Trypsin 1. Trypsin (0.25%) 100 mL (final concentration 0.2%).

2. Knockout serum replacement (KSR) 25 mL.
3. CaCl2 (final concentration 1 mmol/L) (see Note 3).
4. Diluted trypsin: Mix 1–3 to make diluted trypsin. Aliquot
15 mL to a Falcon tube, and freeze at 30  C until use.
Warm the Falcon tube at 37  C before use.

3 Methods

3.1 Spheroid Culture 1. Before performing 2D culture, create spheroids using an ultra-
low attachment plate.
2. Kill a pregnant C57BL/6 mouse at day 13.5 of gestation using
isoflurane, and perform cesarean delivery.
3. Remove and dissect fetal livers using tweezers in cold PBS.
4. Dissociate the dissected fetal livers using a 1 mL pipette tip, and
collect the fragments with centrifugation at 500 rpm for 3 min
(see Note 4).
5. Count and plate the collected cells (see Note 5) at a density of
5  105 cells/mL in the medium with growth factors men-
tioned in Subheading 2.2 on a 6-well ultralow attachment plate
(2 mL/well).
6. Add 2 mL of medium and growth factors at days 2 and
4. Hepatic spheroids will form and increase in size for 6 days
(Fig. 2) (see Note 6).

3.2 2D Culture 1. Collect the hepatic spheroids formed in Subheading 3.1 by

centrifugation at 300 rpm for 2 min.
2. Plate the spheroids on a type I collagen-coated plate with the
medium and growth factors mentioned in Subheading 2.2.
Each spheroid will form a colony (Fig. 3a) in the 6-well colla-
gen-coated plate (see Note 7). Hepatic progenitor cells will be
highly enriched in this serum-free condition (see Note 8).
3. Change the medium and growth factors every 3 days.
4. After reaching approximately 80% of the confluency, collect
cells according to the method mentioned below.
6 Atsunori Tsuchiya and Shuji Terai

Fig. 2 Spheroids formed for 6 days by spheroid (floating) culture. The size of
spheroids was approximately 50–250 μm, comprising approximately 20–150
cells

3.3 Replating 1. Remove the medium and wash with PBS.

2. Add 1 mL of diluted trypsin warmed at 37  C to the
6-well collagen-coated plate, and warm the plate at 37  C for
2–5 min.
3. Check the status of the colonies by microscopy every 2–3 min.
If the colonies are partially removed from the plates while
maintaining cell-cell contact, add medium and stop the enzy-
matic digestion/reaction (see Note 9).
4. Press a 2 mL or 5 mL pipette to the plates, and remove
colonies with the strong hydraulic pressure of the pipette
(Fig. 3b).
5. Collect the removed colonies with low-speed centrifugation
(500 rpm for 3 min).
6. Divide the colonies into 2 or 3 aliquots, and replate on newly
prepared collagen-coated plates (Fig. 3c). This procedure can
be repeated to expand the mouse fetal HSPCs for more than
ten passages. A scheme of the culture is shown in Fig. 1 (see
Note 10).
7. Collected colonies can be preserved at 80  C for several
weeks or in liquid nitrogen for years by using
STEM-CELLBANKER®. CELLBANKER® 2 can also be
used (see Note 11).
 Culture of Mouse Fetal HSPCs 7

Fig. 3 Images of the 2D (plating) culture and replating. Colonies are formed from each spheroid (a). Collected
colonies by diluted trypsin (b) are replated onto newly prepared collagen-coated plates (c). During passage,
cell-cell contact was maintained

4 Notes

1. B27 supplement (50), ITS mixture (100), and penicillin-

streptomycin liquid (100) can be purchased from Thermo-
Fisher Scientific.
2. Any type of plate may be appropriate if cells cannot attach to
the plate. We purchased plates from Corning.
3. We made 100 mM CaCl2 solution (100) by dissolving CaCl2
in distilled water and then filtering the solution. The solution
was added to the mixture of trypsin and KSR.
4. Faster centrifugation might be better. However, 500 rpm for
3 min is sufficient to collect dissociated fetal liver cells.
5. During counting, red blood cells obviously distinguished by
size were excluded.
8 Atsunori Tsuchiya and Shuji Terai

6. We examined the efficacy to form spheroids through growth

factor matching. Matching with EGF, HGF, and bFGF was the
most appropriate to obtain many spheroids.
7. Spheroids from each 6-well plate were plated on each well of
6-well collagen type I-coated plates.
8. Hematopoietic cells and mesenchymal cells were included in
the culture. However, most of the cells disappeared after 2–3
passages (confirmed by cell morphology and flow cytometry).
9. If colonies are difficult to remove from the plates, the time for
trypsin treatment can be extended to 10–15 min. Maintaining
cell-cell contact is the most important point for successful
replating.
10. The replating method using diluted trypsin can be employed
for adult mouse hepatic progenitor cells [9, 10].
11. The cell bankers can be purchased from ZENOAQ.

References
1. Suzuki A, Zheng Y, Kondo R et al (2000) 6. Tanimizu N, Nishikawa M, Saito H et al (2003)
Flow-cytometric separation and enrichment of Isolation of hepatoblasts based on the expres-
hepatic progenitor cells in the developing sion of Dlk/Pref-1. J Cell Sci 116:1775–1786
mouse liver. Hepatology 32:1230–1239 7. Tanimizu N, Tsujimura T, Takahide K et al
2. Suzuki A, Zheng YW, Kaneko S et al (2002) (2004) Expression of Dlk/Pref-1 defines a sub-
Clonal identification and characterization of population in the oval cell compartment of rat
self-renewing pluripotent stem cells in the liver. Gene Expr Patterns 5:209–218
developing liver. J Cell Biol 156:173–184 8. Tsuchiya A, Heike T, Fujino H et al (2005)
3. Schmelzer E, Zhang L, Bruce A et al (2007) Long-term extensive expansion of mouse
Human hepatic stem cells from fetal and post- hepatic stem/progenitor cells in a novel
natal donors. J Exp Med 204:1973–1987 serum-free culture system. Gastroenterology
4. Kamiya A, Kakinuma S, Yamazaki Y et al 128:2089–2104
(2009) Enrichment and clonal culture of pro- 9. Tsuchiya A, Heike T, Baba S et al (2007) Long-
genitor cells during mouse postnatal liver term culture of postnatal mouse hepatic stem/
development in mice. Gastroenterology progenitor cells and their relative developmen-
137:1114–1126 1126 e1-14 tal hierarchy. Stem Cells 25:895–902
5. Kakinuma S, Ohta H, Kamiya A et al (2009) 10. Lu WY, Bird TG, Boulter L et al (2015)
Analyses of cell surface molecules on hepatic Hepatic progenitor cells of biliary origin with
stem/progenitor cells in mouse fetal liver. J liver repopulation capacity. Nat Cell Biol
Hepatol 51:127–138 17:971–983
 Chapter 2

Isolation of Bipotential Liver Progenitor Cells from Neonatal

Mouse Liver
Naoki Tanimizu

Abstract
Liver stem/progenitor cells (LPCs) are defined as bipotential cells differentiating into both hepatocytes and
cholangiocytes. The Notch, TGFβ, and Hippo pathways have been implicated in lineage determination of
LPCs during development and regeneration. However, the molecular mechanisms governing the lineage
specification have not been fully elucidated, yet. Epithelial adhesion molecule (EpCAM) is a marker of
cholangiocytes and of LPCs. We found that EpCAM+ cells isolated from neonatal liver contain LPCs that
clonally proliferate and are bipotential in vitro and in vivo. Furthermore, EpCAM+ progenies keep the
capacity of bidirectional differentiation even after long-term culture. These cells are useful to investigate the
molecular mechanisms regulating lineage commitment and epithelial differentiation of LPCs.

Key words Tissue stem cell, Liver, Neonate, Bipotential, Hepatocyte, Cholangiocyte, Liver stem/
progenitor cell, Fluorescence-activated cell sorting, Colony assay, Clonal culture

1 Introduction

Liver stem/progenitor cells (LPCs) are defined as bipotential cells

differentiating into two types of liver epithelial cells, hepatocytes
and cholangiocytes. LPCs have been isolated from normal and
injured livers based on expression of surface antigens including
EpCAM (epithelial adhesion molecule), CD13, and CD133
[1]. However, the genetic lineage tracing done in the past 5 years
suggests that LPCs significantly contribute neither to cellular turn-
over nor regeneration [2, 3]. In contrast, recent results showed that
cholangiocytes supply new hepatocytes when hepatocytes are
almost completely depleted or the proliferative capability of hepa-
tocytes is impaired [4, 5], suggesting that part of cholangiocytes
may have plasticity to act as facultative LPCs. Although the hetero-
geneity of cholangiocytes has been demonstrated [6, 7], it is still
difficult to label those cells in vivo and to prospectively isolate the
sauce of facultative LPCs in normal and injured adult liver.

Naoki Tanimizu (ed.), Hepatic Stem Cells: Methods and Protocols, Methods in Molecular Biology, vol. 1905,
https://doi.org/10.1007/978-1-4939-8961-4_2, © Springer Science+Business Media, LLC, part of Springer Nature 2019

9
10 Naoki Tanimizu

We isolated EpCAM+ cells from the liver between E17 and

12 W after birth and showed that colony-forming cells are enriched
in this fraction [8]. We further found that neonatal EpCAM+ cells
differentiate to hepatocytes and cholangiocytes in vitro and in vivo,
though they gradually lose such differentiation capability during
the subsequent developmental process. It is also possible to estab-
lish bipotential liver progenitor cell lines from neonatal EpCAM+
cells. This chapter includes the protocols for isolating EpCAM+
cells from 1 W mouse liver, expanding them in vitro and inducing
their differentiation to either hepatocyte or cholangiocytes.

2 Materials

2.1 Cell Isolation 1. Isoflurane.

2. Hanks’ balanced salt solution (HBSS).
3. Pre-perfusion solution: 0.5 mM EGTA in HBSS.
4. Twenty-seven-gauge needle.
5. Ten milliliter syringe.
6. Collagenase: 1 mg/mL in HBSS.
7. 6 cm plastic dish.
8. Liver perfusion medium: Liver perfusion medium (Thermo
Fisher Scientific, Cat. No. 17701038) or HBSS
containing EGTA.
9. PBS (+/+):100 mg/mL MgCl2 and 100 mg/mL CaCl2 in
phosphate buffer saline (PBS).
10. Liberase TM (Roche Diagnostics, Basel, Switzerland): It is the
blendzyme of collagenase and thermolysin. Dissolve in PBS,
make aliquots, and store at 80  C. Dilute the stock solution
50 times in PBS (+/+) before use.
11. 70 μm cell strainer.
12. Hemolysis buffer: 10 mM Tris-HCl, 50 mM NH4Cl. Pass
through a 0.22 μm pore filter to sterilize.

2.2 Cell Sorting 1. Antibodies: FITC-conjugated anti-mouse CD329/EpCAM,

APC-conjugated anti-mouse CD31, and APC-conjugated
anti-CD45 and APC-Cy7-conjugated anti-mouse TER119.
2. Wash buffer: Add 2% FBS to PBS. Store at 4  C.
3. Propidium iodide (PI) (1 mg/mL).

2.3 Cell Culture 1. Growth factor-reduced Matrigel® (MG) (BD Biosciences, Bed-
ford, MA): Thaw the bottle of MG on ice, and make aliquots in
1.5 mL tubes. Freeze those tubes in liquid nitrogen, and store
 Isolation of Neonatal Bipotential Liver Progenitors 11

them at 30  C. Thaw MG in a tube on ice before use (see

Note 1).
2. Collagen type IPC (Koken Co., Ltd., Tokyo, Japan).
3. Collagen reconstitution buffer: 2.2% NaHCO3, 0.2 M
HEPES, and 50 mM NaOH. Pass through a 0.22 μm pore
filter to sterilize.
4. 10 Dulbecco’s Modified Eagle’s Medium (DMEM)/Ham’s
F-12 Nutrient Mixture (1:1) (DMEM/F-12) medium.
5. Laminin solution: Thaw laminin-111 in a glass bottle at 4  C.
Transfer it to a 1.5 mL tube, and store at 4  C. Dilute it in PBS
at 10 μg/mL. Add 1 mL diluted solution to 35 mm plate. Wait
about 1 h and remove the solution before plating cells (see
Note 2).
6. Gelatin solution: Autoclave PBS containing 0.1% gelatin. Add
300 μL gelatin solution to each well of 24-well plate. Aspirate
gelatin solution and wash wells with PBS once.
7. Basal medium: Prepare DMEM/F-12 medium supplemented
with 10% fetal bovine serum, 10 mM nicotinamide, and 10 μg/
mL gentamicin.
8. Growth medium: Add 1 insulin/transferrin/selenium (ITS),
1  10 7 M dexamethasone (DEX), 10 ng/mL epidermal
growth factor (EGF), and 10 ng/mL hepatocyte growth factor
(HGF) to the basal medium.
9. Differentiation medium: Add 1 ITS, 1  10 7 M DEX, and
1% dimethyl sulfoxide (DMSO) to the basal medium. Add
10 ng/mL oncostatin M (OSM) or 5% MG before use.
10. Lysis solution for RNA extraction.
11. Cloning ring: Autoclave glass or stainless rings. Apply auto-
claved Vaseline on the bottom of the ring before surrounding a
colony.
12. Trypsin/EDTA: Dissolve 0.25% trypsin in PBS, make aliquots,
and store at 30  C. Dilute the stock solution 5 times with
PBS and add with 1 mM EDTA before use.
13. Blocking solution: Dissolve Block Ace in PBS.
14. Bovine serum albumin (BSA).
15. PFS: Dissolve 0.7% fish gelatin, 0.07% saponin, and 0.02%
sodium azide. Store at 4  C.
16. Paraformaldehyde (PFA) solution: 4% PFA in PBS.
17. Permeabilization buffer: 0.2% Triton X-100 in PBS. Make 10%
Triton X-100 and keep it at room temperature. Dilute the
stock solution by 50 times with PBS.
12 Naoki Tanimizu

3 Methods

3.1 Cell Isolation 1. Anesthetize a mouse by isoflurane. Open the abdomen, and
then cut the diagram and rib to expose the heart.
3.1.1 Collagenase
Perfusion 2. Attach a needle to a 10 mL syringe filled with pre-perfusion
medium that is pre-warmed at 37  C. Insert the needle into the
left ventricle, and inject pre-perfusion medium. Cut the inferior
vena cava after confirming the liver tissue becomes enlarged
due to pressure (Fig. 1). Change the syringe to one filled with
perfusion medium. Inject the perfusion solution slowly.
3. Isolate the liver and place it on a 6 cm plastic dish. Add 5 mL of
HBSS. Hold the liver with a forceps, and scrape off hepatocytes
from undigested liver tissue using another forceps. (Use undi-
gested tissue for further enzymatic digestion. See the next
paragraph.) Pass the cell suspension through a 70 μm cell
strainer.
4. Centrifuge the cell suspension at 50  g for 1 min. Collect the
supernatant and centrifuge it at 350  g for 4 min (see Note 1).
Resuspend the pellet in 1 mL of basal medium.

3.1.2 Enzymatic 1. Mince the undigested tissue after collagenase perfusion. Sus-
Digestion of the Remaining pend tissue pieces in 3 mL of liver perfusion medium, and
Tissue After Collagenase collect them in a 15 mL tube. Centrifuge at 150  g for 3 min.
Perfusion 2. Resuspend the pellet in 1 mL of Liberase TM solution (see
Note 3). Incubate the cell suspension at 37  C for 15 min.
Dissolve the tissue by pipetting (see Note 4). Add 1–2 mL of

Fig. 1 Schematic view of perfusion for a neonatal mouse. A mouse is fixed on a

board with tapes. After exposing the heart and identifying the left ventricle, a
needle is attached to a 10 mL syringe filled with pre-perfusion medium
 Isolation of Neonatal Bipotential Liver Progenitors 13

basal medium to stop enzymatic digestion. Pass the cell sus-

pension through a 70 μm cell strainer.
3. Centrifugation at 50  g for 1 min. Then, centrifuge the
supernatant at 350  g for 4 min to collect dissociated cells.
Resuspend cells in 1 mL of basal medium.

3.1.3 Hemolysis 1. Combine the cell suspensions from Subheading 3.1.1, step 4,
and Subheading 3.1.2, step 3, into a new 15 mL tube. Centri-
fuge at 350  g for 4 min.
2. Resuspend pellet in 1 mL of hemolysis buffer. Incubate at 4  C
for 5 min. Then, add 1–2 mL of basal medium, and pass
through a 70 μm cell strainer. Centrifuge at 350  g for 4 min.
3. Resuspend the pellet in 200 μL of basal medium, and count the
number of cells.

3.2 Cell Sorting 1. Centrifuge the cell suspension at 350  g for 4 min. Resuspend
cells in 100 μL of basal medium.
2. Add 1 μL of anti-CD16/CD32 antibody, and incubate at 4  C
for 20 min to avoid non-specific binding of antibodies by
masking Fcγ receptors. Add ice-cold wash buffer and centrifuge
at 350  g for 4 min.
3. Resuspend cells in 100 μL of basal medium, and then add 1 μL
each of FITC-conjugated anti-mouse EpCAM,
APC-conjugated anti-CD31, APC-conjugated anti-CD45,
and APC-Cy7-conjugated anti-TER119 (see Note 5).
4. For an isotype control, add FITC-conjugated rat IgG,
APC-conjugated rat IgG, and APC-Cy7-conjugated rat IgG.
5. Incubate cells with antibodies at 4  C for 20 min.
6. Add 2 mL of ice-cold wash buffer and centrifuge at 350  g for
4 min. Resuspend cells in 300 μL of basal medium containing
1 μg/mL PI. Pass through 40 μm cell strainer before analysis
on FACS Aria II (BD Biosciences).
7. Select CD31 CD45 TER119 cells in PI singlet cells by
gating on FACS Aria II. Identify and collect EpCAM + cells
in CD31 CD45 fraction into a 5 mL tube (Fig. 2).
8. Transfer the collected cells into a 15 mL tube. Centrifuge at
350  g for 8 min. Resuspend cells and add 7 mL of basal
medium and centrifuge at 350  g for 8 min. Resuspend cells in
1 mL of basal medium.
14 Naoki Tanimizu

Fig. 2 Isolation of EpCAM+ cells by FACS. After gating live singlet CD31 CD45 cells, EpCAM+ cells can be
identified and isolated

3.3 Culture 1. Resuspend cells in growth medium. Plate 5000 cells in 35 mm

dish coated with laminin.
3.3.1 Colony Assay
2. At day 7, fix cells in 4% PFA at 4  C for 15 min. Wash 1 mL of
PBS for 3 times. Permeabilize with 0.7 mL of 0.02% Triton
X-100, and incubate at room temperature for 5 min. Wash cells
with 1 mL of PBS for 2 times. Add 1 mL of Block Ace and
incubate at room temperature for 30 min.
3. Replace blocking solution to 1 mL of PBS containing 1% BSA
and 1 μL each of rabbit anti-mouse cytokeratin 19 (CK19) and
anti-mouse albumin (ALB) antibody. Incubate at 4  C for 4 h.
Wash cells with PBS three times. Add PBS containing 1% BSA,
0.5 μL each of AlexaFluor488-conjugated anti-rabbit IgG and
AlexaFluor555-conjugated anti-goat IgG and 1 μL each of
Hoechst 33342. Incubate at 4  C for 2 h. Wash cells with
PBS three times before examination of the number of cells
and expression of ALB and CK19 under a fluorescence micro-
scope. A typical bipotential colony is shown in Fig. 3a.

3.3.2 Hepatocyte 1. Plate 5000 EpCAM+ cells in a well of 24-well plate coated with
Differentiation gelatin.
2. Replace basal medium to the differentiation medium contain-
ing OSM and incubate for 2 days. Replace the medium to one
containing 5% MG. Keep the culture for additional 3 days.
3. To examine expression of hepatocyte marker genes, dissolve
cells in a lysis solution and extract total RNA. Synthesize cDNA
and perform PCR with gene-specific primers.
4. To examine expression of hepatocyte marker proteins, treat
cells with Subheading 3.3.1, step 2. Incubate cells with rabbit
anti-hepatocyte nuclear factor 4 α (HNF4α) and goat anti-ALB
or with rabbit anti-CCAAT enhancer binding protein α
(C/EBPα) and goat anti-carbamoyl phosphate synthetase I
(CPSI) antibody. Typical images before and after the treatment
of OSM and MG are shown in Fig. 3b.
 Isolation of Neonatal Bipotential Liver Progenitors 15

Fig. 3 Proliferation and hepatocyte differentiation of neonatal EpCAM+ cells. A EpCAM+ cell clonally
proliferates and forms a bipotential colony containing ALB+ hepatocytes and CK19+ cholangiocytes. (a) Part
of EpCAM+ cells expresses ALB and HNF4α before inducing hepatocyte differentiation, whereas they become
HNF4α+ALB+ hepatocytes in the presence of OSM and MG. (b) Bars represent 100 μm

3.3.3 Cholangiocyte 1. Add PBS, collagen reconstitution buffer, 10 DF medium,

Differentiation and collagen IPC at 4:3:3:20 (v:v:v:v) ratio in a 15 mL tube
on ice. Pipette up and down gently to mix thoroughly.
2. Mix collagen gel with MG at 1:1 ratio. Add 80 μL of the mixed
gel to each well of 8-well cover glass chamber. Incubate it at
37  C for 30 min.
3. Plate 1  104 cells in 150 μL of growth medium, and incubate
for 5 min. Add 150 μL of growth medium containing 10% MG.
16 Naoki Tanimizu

4. To examine expression of cholangiocyte marker genes, dissolve

cells in a lysis solution, and extract total RNA. Synthesize
cDNA and perform PCR with gene-specific primers.
To examine expression of cholangiocyte marker proteins, fix
cells in 4% PFA and permeabilize in 0.02% Triton X-100. Wash
cells with PBS once, add PFS, and incubate at room tempera-
ture for 30 min. Incubate cells with PFS containing primary
antibodies such as anti-CK19, anti-integrin beta4, anti-protein
kinase C zeta, and anti-aquaporin 1 antibody followed by Alexa
Fluor dye-conjugated antibodies and Hoechst 33342. Examine
the localization of cholangiocyte markers under a confocal laser
scanning microscope.

3.3.4 Isolation of 1. Keep clonal culture for about 4 weeks.

Bipotential Clones 2. Wash cells with PBS once, and then surround each colony with
a cloning ring. Add 100 μL of trypsin/EDTA, and incubate for
5–10 min at 37  C. Pipette up and down for 3–4 times, and
transfer the cell suspension into 1 mL of basal medium in
15 mL tube. Centrifuge at 350  g for 3 min.
3. Resuspend cells in growth medium, and plate them onto a well
of 24-well plate coated with laminin-111 (see Note 6).

4 Notes

1. It takes about 2 h to dissolve an aliquot of 500 μL MG on ice.

2. The diluted laminin solution can be reused up to three times to
coat tissue culture dishes. We usually recover the laminin solu-
tion and store it in a 15 mL tube at 4  C.
3. For scale-up, undigested tissues can be pooled at this step. The
tissue pieces derived from 2 to 3 mice are collected in a 15 mL
tube and then added with 2 mL of Liberase TM. For digesting
tissue pieces from more than three mice, make another tube for
Liberase digestion and then pooled cell suspensions into one
tube after Liberase digestion.
4. Pipette up and down for 20–30 times to dissociate the most of
cells from connective tissue.
5. Count the number of cells. Use 1 μL of each antibody to label
cells less than 1  107 cells. Increase the quantity of antibodies
according to the amount of cells.
6. The size of culture is gradually increased to 12-well plate and
then 35 mm dish. EpCAM expression is examined by FACS to
acquire EpCAM+ clones. EpCAM+ clones are replated every
4 days by plating 5  104 cells in 35 mm dish coated with
laminin. Hepatocyte and cholangiocyte differentiation can be
 Isolation of Neonatal Bipotential Liver Progenitors 17

induced using the same protocol as the primary EpCAM+ cells,

which are described in Subheadings 3.3.2 and 3.3.3.

Acknowledgments

I thank Dr. Toshihiro Mitaka and Dr. Norihisa Ichinohe for helpful
discussion and Ms. Yumiko Tsukamoto and Ms. Minako Kuwano
for technical assistance. This work is supported by the Ministry of
Education, Culture, Sports, Science and Technology, Japan,
Grants-in-Aid for Scientific Research (C) (25460271,
16 K08716), and Grants-in-Aid for Scientific Research on Innova-
tive Areas “Stem Cell Aging and Disease” (17H05653).

References
1. Miyajima A, Tanaka M, Itoh T (2014) Stem 5. Raven A, Lu WY, Man TY et al (2017) Cholan-
Progenitor cell sin liver development, homeosta- giocytes act as facultative liver stem cells during
sis, regeneration, and reprogramming. Cell Stem impaired hepatocyte regeneration. Nature
Cell 14:261–274 547:350–354
2. Espanol-Suner R, Carpentier R, Van HN et al 6. Kamimoto K, Kaneko K, Kok CY et al (2016)
(2012) Liver progenitor cells yield functional Heterogeneity and stochastic growth regulation
hepatocytes in response to chronic liver injury of biliary epithelial cells dictate dynamic epithe-
in mice. Gastroenterology 143:1564–1575 lial tissue remodeling. Elife 5:e15034
e1567 7. Li B, DorrellC CPS et al (2017) Adult mouse
3. Malato Y, Naqvi S, Schurmann N et al (2011) liver contains two distinct populations of cho-
Fate tracing of mature hepatocytes in mouse langiocytes. Stem Cell Reports 9:478–489
liver homeostasis and regeneration. J Clin Invest 8. Tanimizu N, Kobayashi S, Ichinohe N et al
121:4850–4860 (2014) Downregulation of miR122 by
4. Choi TY, Ninov N, Stainer DY et al (2014) grainyhead-like 2 restricts the hepatocytic differ-
Extensive conversion of hepatic biliary epithelial entiation potential of adult liver progenitor cells.
cells to hepatocytes after near total loss of hepa- Development 141:4448–4456
tocytes in zebrafish. Gastroenterology
146:776–788
 Chapter 3

Identification and Isolation of Clonogenic Cholangiocyte

in Mouse
Bin Li, Craig Dorrell, Pamela S. Canady, and Leslie Wakefield

Abstract
Cholangiocytes are proliferative and are one of the sources for liver progenitor cells. Clonogenic cholan-
giocytes are defined as cells capable of clonally proliferating and differentiating cholangiocytes both in vitro
and in vivo. In this protocol, we describe the method for isolation of primary cholangiocytes from mouse.
To study the heterogeneity of cholangiocytes, we used flow cytometry-based cell sorting to isolate different
subsets of cholangiocytes. Organoid-forming efficiencies from sorted single cells are compared within
different cholangiocyte populations to identify clonogenic cholangiocytes.

Key words Duct cells, ST14, FACS, Organoid

1 Introduction

Liver duct cells, so-called cholangiocytes, are highly proliferative

in vitro and in vivo. In our previous report found in the normal and
injured mouse liver, MIC1-1C3+ liver duct cells are proliferative:
they formed colony in vitro, and gene expression showed MIC1-
1C3+ cells that are expressing Sox9, a progenitor marker [1]. More-
over, MIC1-1C3+ colonies could engraft in the Fah
/
mouse, a
transgenic mouse model for liver injury, and the transplanted liver
cells could replace the damaged host liver [2], indicating the stem/
progenitor capabilities of donor cells. Thus, MIC1-1C3 is used as
an oval cell marker. Lgr5, a member of Wnt signal pathway, has
been previously found as a liver stem cell marker. In contrast to
MIC1-1C3, normal liver doesn’t express Lgr5 [3], whereas, under
chronic liver injury, liver duct cells in mouse become Lgr5+. Those
Lgr5+ cells are proliferative and form liver organoid in vitro. More-
over, transplantation of Lgr5+ organoids could engraft into the
host liver, indicating the stem cell capability [4]. Further report
demonstrated that MIC1-1C3+ duct cells in both normal liver and
pancreas form organoids in vitro [5]. ST14 (suppressor of tumori-
genicity 14 protein) is a marker for pancreatic cancer; however, not

Naoki Tanimizu (ed.), Hepatic Stem Cells: Methods and Protocols, Methods in Molecular Biology, vol. 1905,
https://doi.org/10.1007/978-1-4939-8961-4_3, © Springer Science+Business Media, LLC, part of Springer Nature 2019

19
20 Bin Li et al.

any research has studied ST14 in the liver. In our previous report,
with the cell surface markers MIC1-1C3 [5] and ST14 [3], we
demonstrated that normal mouse cholangiocytes are heteroge-
neous: ST14hiMIC1-1C3+ cells are highly proliferative; they form
single-cell-derived organoid more efficiently than ST14loMIC1-
1C3+ cholangiocytes [3]. Here we will describe the method for
isolating cholangiocytes via liver perfusion. ST14hiMIC1-1C3+
cells in primary liver non-parenchymal cells (NPCs) are directly
sorted into 96-well plate by fluorescence-activated cell sorting
(FACS). In this format of colony-forming assay, clonogenic cho-
langiocytes form single-cell-derived organoids. In addition, orga-
noids cultured for ~3 months are used for the differentiation assay
examining their bidirectional differentiation potential.

2 Materials

2.1 Liver Perfusion 1. Solution 1: 10 mM HEPES and 0.5 mM EGTA in Earle’s

Balanced Salt Solution (EBSS) without Ca2+/Mg2+, pH 7.4
(~30 mL per mouse).
2. Solution 2: 10 mM HEPES in EBSS with Ca2+/Mg2+, pH 7.4
(~20 mL per mouse).
3. Solution 3: 0.1% collagenase II in solution 2 (~50 mL per
mouse) see Note 1.
4. NPC digestion solution: 2.5 mg/mL collagenase IV and
0.1 mg/mL DNase I in solution 2 (~5 mL per mouse).
5. Wash medium: 10% FCS and 0.1 mg/mL DNase in DMEM.
6. 0.05% trypsin.
7. Phosphate-buffered saline (PBS).
8. 70% ethanol.
9. Cotton applicators (sterile).
10. Intravenous (IV) infusion set.
11. 24-gauge catheters.
12. Peristaltic pump.
13. Forceps, scissors (sterile).
14. Tape.
15. 100 μm cell strainer.
16. 60 mm petri dish.
17. Anesthetic solution.
18. Centrifuge, set at 10  C.
19. Mouse: 8-week-old C57B/L6 male mouse. All animal experi-
mentation was conducted in accordance with protocol
IP00445 of the Institutional Review Committee at Oregon
Health and Science University.
 Isolation of Mouse Liver Progenitor Cells 21

2.2 FACS 1. FACS buffer: 3% fetal bovine serum (FBS) in PBS.

2. 5% rat serum in PBS.
3. 5 mL polypropylene round-bottom tube.
4. 1 mg/mL propidium iodide (PI).
5. Antibodies: rat anti-mouse CD26, fluorescein isothiocyanate
(FITC), CD45 Pe-cy7, CD31 Pe-cy7, CD11b Pe-cy7, MIC1-
1C3 [1], 5% rat serum in PBS, rabbit anti-mouse ST14, sec-
ondary anti-rat phycoerythrin (PE), anti-rabbit allophycocya-
nin (APC).
6. Software for FACS data analysis: FlowJo software (FlowJo,
LLC) or any other software that can handle FACS plots.

2.3 Organoid Cell 1. 200 μL multichannel pipette and tips.

Culture 2. 96-well suspension culture plate (Genesee Scientific 25-104).
3. 24-well suspension culture plate (Genesee Scientific 25-102).
4. Matrigel (BD Biosciences) (see Note 2).
5. Organoid culture medium: advanced DMEM/F12 supplied
with 100 ng/mL Noggin, 100 μg/mL Wnt3a, 50 ng/mL
FGF10, 50 ng/mL EGF, 6 μM SB431542, 1.25 mM N-acetyl
cysteine, 10 mM nicotinamide, 1 B27, 1 N-2 supplement,
1 μg/mL R-spondin1, 1 GlutaMAX, 10 mM HEPES, and
penicillin/streptomycin/amphotericin.
6. TrypLE (Gibco).

3 Methods

3.1 Liver Perfusion 1. Sterilize pump tubing (at ~15 mL/min) with 70% ethanol and
then wash with solution 1. Prewarm all solutions to 40  C (see
Note 3).
2. Attach IV infusion set and fill with solution 1 (see Note 4).
3. Anaesthetize the mouse.
4. Immobilize all four limbs with tape on a moisture-absorbing
surface (diaper). Open the abdomen while avoiding
opening the chest. Push intestines to the side with the sterile
cotton applicator, thereby exposing the vessels. Make sure you
identify the inferior vena cava and portal vein (under the liver)
(Fig. 1a, b).
5. Carefully cannulate the portal vein (see Fig. 1c, d) (see Note 5).
Fill the catheter with solution 1 by syringe, attach the IV
tubing to the liquid-filled catheter (avoid introducing any
air), and start the flow of the pump at 4 mL/min. The liver
will fill up with solution and turn brown within a few seconds.
22 Bin Li et al.

Fig. 1 Liver perfusion. (a) Anesthetize the mouse and open the abdomen. (b) Expose the portal vein with cotton
applicator; white arrow shows the portal vein. (c) Cannulate the portal vein; blood will flow through the
catheter. (d) Connect the tubing to the catheter. (e) White arrow shows cutting vena cava. (f) Perfused liver

Quickly cut the vena cava (see Fig.1e). Tape down the catheter
and tubing securely. Start timer.
6. After about 15 mL of solution 1, run solution 2 for 1 min and
finally solution 3 for 10 min (see Note 6).
7. To gauge progress, poke the liver with a cotton applicator after
7 min or so: under-digested liver remains elastic, whereas a
completely digested liver will retain an indentation (Fig. 1f).
8. When the liver is digested, remove the catheter, excise the
gallbladder, and then carefully remove the liver without injur-
ing the capsule if possible.
9. Transfer the liver to a 60 mm petri dish containing solution
3, and proceed to the next perfusion (Fig. 2a). Disrupt the liver
capsules with forceps, and wash out the hepatocytes by swirling
the dish (Fig. 2b). Hepatocytes are collected and filtered
through a 100 μm filter unit into a sterile 50 mL tube that
contains 10 mL of wash medium (Fig. 2c).
10. For studies of non-parenchymal cells (NPC), tissue retained on
the filter is collected and digested further.
11. Hepatocytes are pelleted by spin at 50  g for 2 min at 4  C
(Fig. 2d). Supernatant is collected for NPC and the pellet
resuspended in 50 mL wash media. Centrifuge at 300  g,
5 min for NPC pellet; this is NPC fraction 1.
12. Transfer tissue fragments in step 10 to a 60 mm dish in 5 mL of
NPC digestion solution. Add a magnetic stirring bar, and stir at
37  C for 10 min. See Fig. 2e.
 Isolation of Mouse Liver Progenitor Cells 23

Fig. 2 Liver digestion and FACS sort. (a) Move the liver into a petri dish. (b) Break the liver capsule to
dissociate the cells. (c) Filter the liver cells with cell strainer. (d) Pellet the hepatocytes. (e) Tissue leftover is
under further digestion with NPC solution. (f) Filter the digested NPCs. (g) FACS tube. (h) FACS machine: BD
Influx. (i) 96-well single cell sort

13. After 10 min, a P1000 is used to pipette solution up and down

to assist the breakup of remaining tissue fragments.
14. Stirring is continued for an additional 5–10 min, and disso-
ciated cells are collected by passage through a 40 μm strainer.
This is NPC fraction 2. These should be spun at 300  g and
resuspended in cold wash media.
15. Remaining solid material is transferred to a new dish with 5 mL
of 0.05% trypsin and incubated with stirring for 20 min.
16. A P1000 is used to pipette solution up and down to assist the
breakup of remaining tissue fragments.
17. Newly dissociated cells are collected by passage through a
40 μm strainer. This is NPC fraction 3. These should be
spun at 300  g and resuspended in cold wash medium.
18. Collect all NPC fractions and pellet the cells.

3.2 FACS 1. Count and resuspend the cells with FACS buffer at a density of
1  107 cells/mL. Add 100 μL cells into each control tube
(#1–7). The sample tube (#8) can have 500–1000 μL volume
(see Note 7).
24 Bin Li et al.

Table 1
Antibody labeling for FACS

Tube# 1 2 3 4 5 6 7 8
PI FITC PE PEcy7 APC FM1 FM1 Sort
-APC -PE
MIC1-1C3 (μL) 5 5 50
ST14 (μL) 1 1 10
FACS buffer (μL) 100 100 100 100 100 1000
20 min on ice, wash, pellet cells
Anti-rat PE (μL) 1 1 1 2
Anti-rabbit APC (μL) 1 1 1 2
FACS buffer (μL) 200 200 200 200 400
15 min on ice, wash, pellet cells, blocking tubes #6 and #7 with 5% rat serum for 10 min
FACS buffer (μL) 100 100 100 100 200
FITC CD26 (μL) 1 1 1 2
Pe-cy7 CD45/31/11b (μL) 1 1 1 2
6 6 6 6 6 6 6
Cell number 10 10 10 10 10 10 10 107

2. Add rat anti-mouse MIC1-1C3 antibody (1:20 dilution of

each) to tubes # 3, 6, and 8. Add rabbit anti-mouse ST14
antibody (1:100 dilution) to tubes # 5 and 7.
3. Incubate 20 min on ice.
4. Wash (add 5 mL wash medium to tubes, spin 4 min at 300  g
to pellet, aspirate supernatant), and add 200 μL FACS
medium.
5. Add goat anti-rat F(ab) PE (1:200 dilution) to tubes #3, 6, 7,
and 8 and mouse anti-rabbit APC (1:200 dilution) to tubes #5,
6, 7, and 8.
6. Incubate 15 min on ice.
7. Wash (add 5 mL wash medium to each tube, spin 4 min at
300  g to pellet, aspirate supernatant).
8. Block tubes #6, 7, and 8 with 100 μL and 200 μL 5% rat serum
for 10 min on ice.
9. Add rat anti-mouse CD26 to tubes #2, 6, 7, and 8.
10. Add rat anti-mouse CD45/CD31/CD11b Apc-cy7 (1:100
dilution) to tubes #4, 6, 7, and 8. See Table 1 for antibody
labeling. All steps are on ice.
 Isolation of Mouse Liver Progenitor Cells 25

Fig. 3 FACS gating strategy. (a and b) Cells were sequentially gated based on cell size (forward scatter [FSC]
versus side scatter [SSC]). (a) Depletion of doublets (FSC vs. trigger pulse width) (b). (c) Dead cells and debris
were excluded by PI positivity. Concurrently a combination of CD45, CD31, and CD11b antibodies was used for
depleting blood, endothelium, and Kupffer cells. (d) CD26 (DPPIV) was used for hepatocyte staining. (e) MIC1-
1C3+ cells can be subdivided into two populations: ST14 high (ST14hiM+) and ST14 low (ST14loM+). (f)
Fluorescence minus 1 without secondary antibody APC was used to draw gate for ST14 APC positivity and
negativity (modified from Bin Li et al., 2017 [3])

11. Last wash and pellet the cells. Resuspend cells with PI at
10 μg/mL in FACS buffer. The FACS gates are drawn as
shown in Fig. 3.

3.3 Organoid Cell 1. Precool the 96-well plate, multichannel pipettes, and tips.
Culture 2. Make the 5% Matrigel in fresh organoid culture media before
sorting, using multichannel pipette to add media into 96-well
plate with 100 μL per well. Keep the plate on ice.
3. Sorted cells are directly injected into cold 96-well plate at a
density of 1 cell per well.
4. Carefully flip the plate to help the cells sink into the media. The
plate will be moved into 37  C CO2 incubator immediately.
5. Add more media into organoid culture at day 5. The organoid
should be seen clearly at day 7. The organoid-forming assay
should be done at day 14.
6. For organoid passage: move the organoid with media in each
well into 1.5 mL Eppendorf tube, add the same volume of
26 Bin Li et al.

Fig. 4 Representative picture for single-cell-derived organoid under culture day

14. Upper panel: colony-forming efficiency of single-cell-derived colonies in
96 well. Lower panel: representative picture of single ST14hiM+ and ST14loM+
cell-derived organoid in 96 well. CFU: colony-forming unit. Scale bar ¼ 100 μm
(modified from Bin Li et al., 2017 [3])

TrypLE, pipette with P200, and look through the tube under
microscope. Matrigel dissolves in TrypLE. If organoids are still
stiff, add more TrypLE at room temperature (RT), or put the
tube in 37  C for 3 min.
7. When dissociated small clumps as well as single cell are seen,
add wash medium to wash and spin.
8. The pellet is resuspended in 50 μL Matrigel on ice.
9. Pipette Matrigel droplet on prewarmed 24-well suspension
culture plate at 37  C.
10. After 5 min, add prewarmed organoid medium into plate. The
medium will be changed twice a week. After 14 days culture,
the organoid forming efficiencies are calculated, see Fig. 4.

4 Notes

1. Solutions 1 and 2 are kept at room temperature, and solution

3 is prepared fresh prior to use.
2. Thaw Matrigel overnight on ice.
3. To make sure the whole surgery is at 37  C, set the water bath
to 40  C.
4. Make sure to add liquid to air trap!
 Isolation of Mouse Liver Progenitor Cells 27

5. If the portal vein cannulation failed, cannulate vena cava

instead.
6. At 4 mL/min, there is about a 1 min lag between switching
solutions and them reaching the animal. Once the liver is soft
and not bond, it is the time to stop.
7. Tube #1, PI-only control; #2, FITC-only control; #3, PE-only
control; #4, PEcy7-only control; #5, APC-only control; #6,
florescence minus APC control (second antibody APC only, no
primary ST14); #7, florescence minus PE control (second anti-
body PE only, no primary MIC1-1C3).

Acknowledgments

The authors thank members in the Markus Grompe Lab at Oregon

Health and Science University for technical supports. The authors
also thank Oregon Health and Science University Flow Cytometry
Core for technical assistance.

References

1. Dorrell C et al (2011) Prospective isolation of a Cell Reports 9:478–489. https://doi.org/10.

bipotential clonogenic liver progenitor cell in 1016/j.stemcr.2017.06.003
adult mice. Genes Dev 25:1193–1203. https:// 4. Huch M et al (2013) In vitro expansion of single
doi.org/10.1101/gad.2029411 Lgr5+ liver stem cells induced by Wnt-driven
2. Azuma H et al (2007) Robust expansion of regeneration. Nature 494:247–250. https://
human hepatocytes in Fah
/
/Rag2
/
/ doi.org/10.1038/nature11826
Il2rg
/
mice. Nat Biotechnol 25:903–910. 5. Dorrell C et al (2014) The organoid-initiating
https://doi.org/10.1038/nbt1326 cells in mouse pancreas and liver are phenotypi-
3. Li B et al (2017) Adult mouse liver contains two cally and functionally similar. Stem Cell Res
distinct populations of cholangiocytes. Stem 13:275–283. https://doi.org/10.1016/j.scr.
2014.07.006
 Chapter 4

Isolation and Expansion of Rat Hepatocytic Progenitor Cells

Junichi Kino, Norihisa Ichinohe, Masayuki Ishii, and Toshihiro Mitaka

Abstract
This protocol showed procedures to isolate and expand small hepatocytes (SHs), as hepatocytic progenitor
cells, from a rat liver. SHs are identified as a subpopulation of mature hepatocytes in a healthy liver. SHs can
proliferate to form colonies in serum-free medium on hyaluronic acid-coated dishes, of which cells show
CD44 positivity (CD44+ SHs). CD44+ SHs can be separated and purified from colonies by using anti-
CD44 antibodies after enzymatic dissociation. CD44+ SHs can proliferate to form colonies on Engelbreth-
Holm-Swarm gel (EHS-gel)-coated dishes in the serum-free medium for a long period and subculture for
several times. Even after the second passage, the cells possess characteristics of hepatocytes such as
expression of albumin and HNF4α. In addition, when the cells are treated with EHS-gel, they can recover
highly differentiated functions of hepatocytes such as glycogen production, CYP activity, and bile secretion.

Key words Small hepatocyte, Serum-free culture, CD44, Hepatocytic function

1 Introduction

Hepatocytes are the predominant cell type in the liver, constituting

over 80% of the liver mass under physiological conditions [1]. The
liver can regenerate after surgical resection or cytotoxic damages.
Following a 2/3 partial hepatectomy (PH), residual hepatocytes
proliferate and replace the lost liver mass [2]. Rat mature hepato-
cytes (MHs) were shown to perform at least 18 cell doublings
[3]. On the other hand, it has been shown that mouse hepatocytes
possess tremendous replication potential by using the technique of
serial transplantations [4, 5]. However, primary hepatocytes are
well known to have poor growth activity in vitro and difficult to
maintain metabolic and xenobiotic functions [6, 7]. Various
attempts have been made over the last several decades to harness
the innate replication potential of hepatocytes ex vivo. Small

All the authors contributed equally to this work.

Naoki Tanimizu (ed.), Hepatic Stem Cells: Methods and Protocols, Methods in Molecular Biology, vol. 1905,
https://doi.org/10.1007/978-1-4939-8961-4_4, © Springer Science+Business Media, LLC, part of Springer Nature 2019

29
30 Junichi Kino et al.

hepatocytes (SHs), hepatocytic progenitor cells, have shown to

possess the highest growth potential in culture [8, 9]. SHs were
isolated from a healthy adult liver of rat and human and proliferated
to form colonies in a serum-free medium on hyaluronic acid (HA)-
coated dishes [10, 11], which may be dependent on CD44 expres-
sion at the cell surface [12]. Although CD44 expression is weaker
than that of a rat, mouse SHs were also identified [13]. The colo-
nies originated from a single cell consisting of approximately 30–40
cells after 8–10 days of plating [10, 12, 14]. Morphology of each
SH is similar to MHs except the size. The colonies consisting of
small-sized cells can be isolated by using collagenase and hyaluron-
idase. CD44+ SHs are isolated and sorted by anti-CD44 antibody,
and then plated on Engelbreth-Holm-Swarm (EHS)-gel-coated
dishes. Attached cells can proliferate to form colonies in the
serum-free medium [15]. The cells possessing CD44 positivity
and the ability of albumin production can be passaged at least
four times. Although the size and other features of most colonies
are uniform in the early culture period after the passage, the fea-
tures of colonies are divided into two types with time in culture; one
is consisted of small-sized cells and the other of large and flattened
ones. In addition, maturation of SHs can be manually induced by
EHS-gel treatment, and the cells recover highly differentiated func-
tions such as glycogen storage, cytochrome P450 (CYP) activity,
and bile secretion. This protocol shows the procedures how to
isolate rat CD44+ SHs maintaining hepatocytic characteristics and
to passage them (Fig. 1).

Fig. 1 Overview of procedures for the isolation and expansion of rat small hepatocytes (SHs). Liver cells are
isolated from a 6–10 week-old male rat. SHs form colonies on hyaluronic acid-coated dish and express CD44
on their cell surfaces. CD44-positive SHs (CD44+ SHs) is purified by magnetic separation using anti-CD44
antibody. CD44+ SHs can expand on EHS-coated dish and be passaged after trypsinization
 Isolation and Expansion of Rat Hepatocytic Progenitor Cells 31

2 Materials

2.1 Isolation of Liver The 6- to 10-week-old F344 rats are used for cell isolation. The
Cells animal is maintained with standard chow and water ad libitum. All
apparatuses used for the collagenase perfusion are sterilized by
autoclave or treatment of 70% ethanol before the experiment starts.
All animal experiments must comply with national and institutional
regulations.
1. Phosphate-buffered saline without Mg2+/Ca2+ (PBS). Store at
room temperature (RT).
2. 500 μg/mL insulin stock solution (1000): Add 100 mg insu-
lin to 100 mL double-distilled water (ddH2O), and subse-
quently add 1.2 mL 1 N HCl (see Note 1). Adjust to 200 mL
with ddH2O and filtrate with a 0.2-μm filter.
3. Pre-perfusion solution: Add approximately 850 mL ddH2O
into a 1000-mL graduated cylinder 100-mL Ca2+, Mg2+-free
HANKS’ balanced salt solution (HBSS), 190 mg ethylene
glycol tetraacetic acid (EGTA), and 1 mL 1000 insulin
stock solution and stir, and then adjust pH to 7.5 with 1 M
NaHCO3. Adjust to 1000 mL with ddH2O and filtrate with a
0.2-μm filter. Store at 4  C until use.
4. Perfusion solution: Add 1 mL 1000 insulin stock solution to
200 mL HBSS. Add collagenase (Wako Pure Chemical Indus-
tries) of 26,000 U to the pre-warmed perfusion solution, and
make 130 U/mL collagenase solution just before a perfusion
(see Note 2).
5. Wash solution: Add 0.5 mL 1000 insulin stock solution,
2 mL penicillin-streptomycin solution (PC-SM, Sigma), and
0.5 mL gentamicin solution (50 mg/mL, Sigma) to 500 mL
HBSS. Store at 4  C until use.
6. Perfusion apparatus.
7. Water bath.
8. Peristatic pump (RP-1000, Tokyo Rika Instruments).
9. Silicon tube: inside diameter, 4.76 mm; outside diameter,
7.94 mm.
10. O2 gas (95% O2 and 5% CO2).
11. Isoflurane and vaporizer.
12. Butterfly needle, 18 gauge.
13. Surgical scissor.
14. Tweezers.
15. Mosquito forceps.
16. Surgical clip, bulldog type.
32 Junichi Kino et al.

17. Surgical thread.

18. Heating lamp.
19. Petri dish (100 mm).
20. Beaker (100 mL).
21. Conical tube (50 mL).
22. Sterilized serological pipette.
23. Pipetting controller.
24. Tabletop centrifuge.
25. Nylon filter net (250 μm).
26. Cell strainer: 70-μm filter (Corning).
27. Neubauer improved hemocytometer.
28. 0.1% trypan blue solution.

2.2 Culture of the 1. 10 mg/mL HA stock solution: Dissolve HA-derived Strepto-

Isolated Liver Cells on coccus equi (SIGMA, Cat No. 53747) in sterilized PBS by
Hyaluronic Acid (HA)- stirring using a magnetic stir bar at 37  C overnight (see Note
Coated Dishes 3). Store HA stock solution at 4  C.
2. Plastic culture dishes, 100, 60, or 35 mm.
3. 10 4 M Dexamethasone stock solution: Add 39.2 mg dexa-
methasone to an autoclaved brown bottle. Add 10 mL absolute
ethanol (10 2 M stock solution) and then dilute by 100-fold
with autoclaved ddH2O. Store at 4  C until use.
4. 10 μg/mL Epidermal growth factor (EGF) stock solution: Add
10 mL autoclaved ddH2O to mouse EGF 100 μg vial (Corn-
ing). Dispense each aliquot in a cryogenic tube and store it at
20  C until use.
5. 1 M Nicotinamide stock solution: Add 12.21 g nicotinamide to
100 mL PBS. Filtrate with a 0.2-μm filter. Store at 4  C
until use.
6. 100 mM Ascorbic acid-2 phosphate (Asc2P) stock solution:
Add 2.90 g Asc2P to100 mL PBS, filtrate with a 0.2-μm filter,
and store in a autoclaved brown bottle at 4  C until use.
7. Culture medium stock solution: Add 15.56 g Dulbecco’s mod-
ified Eagle’s medium/Nutrient Mixture Ham F-12 powder
(DMEM/F12, Sigma), 1.20 g HEPES, 30 mg L-Proline,
PC-SM to a 1000 mL beaker, and ddH2O up to 1000 mL.
Mix the reagents using a magnetic stir bar, add 2.20 g
NaHCO3, adjust to pH 7.6 with 1 N NaOH, and filtrate
with a 0.2-μm filter. Store at 4  C until use.
8. Culture medium: Add 1.67 mL BSA solution (30%), 5.50 mL
nicotinamide stock solution, 5 mL Asc2P stock solution, 5 mL
insulin-transferrin-selenium solution (ITS-X, Thermo Fisher),
0.5 mL EGF stock solution, 0.5 mL dexamethasone stock
 Isolation and Expansion of Rat Hepatocytic Progenitor Cells 33

solution, and 0.5 mL gentamicin solution in 500 mL culture

medium stock solution. Store at 4  C until use.
9. CO2 incubator.

2.3 Isolation of 1. PBS. Store at RT.

CD44-Positive Cells 2. EDTA/PBS: PBS containing 0.02% ethylenediaminetetraace-
from the Cells on tic acid (EDTA). Store at RT.
HA-Coated Dishes 3. Collagenase. Store at 20  C until use.
4. 17,500 U/50 μL Hyaluronidase stock solution: Add hyaluron-
idase to PBS. Adjust to 3500 U/μL, dispense 50 μL aliquot in a
cryogenic tube, and store it at 20  C until use.
5. Separation solution: Add 75 mg collagenase and 50 μL hyal-
uronidase stock solution to 75 mL HBSS, and warm it at 37  C
just before separation.
6. Magnetic stirrer and stirring bar.
7. MACS® buffer: Add 2 mL 500 mM EDTA solution and
8.3 mL 30% BSA solution to 500 mL PBS. Store at 4  C
until use.
8. Mouse anti-rat CD44 monoclonal antibody (BD biosciences,
Cat No. 554869).
9. Microbead-conjugated anti-mouse IgG2a+b (Miltenyi Biotec).
10. MidiMACS™ separator (Miltenyi Biotec).
11. LS column and its plunger (Miltenyi Biotec).
12. Conical tube (15 and 50 mL).
13. Sterilized serological pipette.
14. Pipetting controller.
15. Tabletop centrifuge.

2.4 Culture and 1. EHS-gel (Matrigel®, Growth factor-reduced, Corning). Store

Passage of the CD44- at 20  C until use.
Positive Cells on EHS- 2. Plastic culture dishes, 100 , 60, or 35 mm.
Gel-Coated Dishes 3. Culture medium: the same culture medium mentioned above
(see Subheading 2.2).
4. PBS. Store at RT.
5. PBS containing 0.25% trypsin and 0.01% EDTA. Store at 4  C
until use.
6. Conical tube (15 and 50 mL).
7. Sterilized serological pipette.
8. Pipetting controller.
9. Tabletop centrifuge.
10. Neubauer improved hemocytometer.
34 Junichi Kino et al.

11. 0.1% trypan blue solution.

12. CO2 incubator.

3 Methods

3.1 Isolation of Liver 1. Set up the perfusion apparatus in the warmed water bath at
Cells approximately 39  C. Circulate the pre-perfusion solution in
the perfusion apparatus using the peristatic pump and silicone
tube before the experiment, and bubble in it with O2 gas at a
flow rate of 0.5 L/min.
2. Anesthetize a rat with an inhalation of isoflurane.
3. Cut the abdominal wall with surgical scissors and open the
abdominal cavity.
4. Keep opening the abdominal cavity by grasping the edge of cut
abdominal wall using mosquito forceps.
5. Move the intestine to expose the liver, portal vein, bile duct,
and splenic vein.
6. Ligate a common bile duct and splenic vein with a surgical
thread at the portion nearest the portal vein (see Note 4).
7. Stop the circulation of the pre-perfusion solution and insert a
butterfly needle filled with pre-perfusion solution into the
portal vein at 1.5–2.0 cm from the bifurcation of the portal
vein. Stop the tip of the needle at a position close to bifurcation
of the portal vein, and clamp the needle with a surgical clip.
8. Start the perfusion of the pre-perfusion solution at a flow rate
of 30 mL/min (see Note 5).
9. Cut the inferior vena cava and heart just after starting the
perfusion (see Note 6).
10. When the volume of the pre-perfusion solution becomes small,
add the collagenase to the perfusion solution and then pour it
into the perfusion apparatus.
11. Flow the solution at a flow rate of 15 mL/min. Heat around
the abdominal cavity using a heating lamp during the perfusion
(see Note 7).
12. When the volume of the perfusion solution becomes about
half, change the flow rate at 10mL/min (see Note 8).
13. Stop the flow before air bubbles move into the liver when the
solution flows out from the perfusion apparatus.
14. Cut the liver from the abdominal cavity and transfer it to a
sterilized Petri dish (see Note 9).
15. Prepare a 100-mL beaker with 80-mL wash solution.
 Isolation and Expansion of Rat Hepatocytic Progenitor Cells 35

16. Peel the hepatic capsule carefully, and then, to drop the
digested cells, gently shake the liver in the beaker. Dropped
cells in the Petri dish are also poured to the beaker.
17. Filter the cell suspension through a 250-μm nylon filter net
into a new 100-ml beaker.
18. Filter the cell suspension through a cell strainer with 70-μm
filter into four 50-ml conical tubes using a 25-ml serological
pipette, and then adjust each tube to an equal volume (approx-
imately 40 mL) with wash solution.
19. Centrifuge the tubes at 50  g for 1 min at 4  C.
20. Collect the supernatants and transfer to four new 50-mL coni-
cal tubes (see Note 10), and wash solution is then added to
each tube up to 45 mL. Repeat Steps 19 and 20 twice.
21. Centrifuge the tubes at 45  g for 5 min at 4  C (see Note 11).
22. Aspirate the supernatant of 40 mL, then add new 40-mL wash
solution to each tube, and dissociate the cell pellet by pipetting.
23. Centrifuge the tubes at 150  g for 5 min at 4  C (see
Note 12).
24. Repeat Steps 22 and 23 twice. Aspirate the supernatant as
possible, and then add new 20-mL wash solution to each
tube, and dissociate the cell pellet by pipetting.
25. Gather suspension into two 50-mL conical tubes, and centri-
fuge the tubes at 150  g for 5 min at 4  C.
26. Aspirate the supernatant as possible, and then add new 20-mL
culture medium to each tube, and dissociate the cell pellet by
pipetting.
27. Gather the suspension into one 50-mL conical tube, and cen-
trifuge the tube at 50  g for 5 min at 4  C.
28. Aspirate the supernatant as possible, and then add new 10-mL
culture medium to the tube, and dissociate the cell pellet by
pipetting. Put this tube on ice.

3.2 Culture of the 1. Dilute the HA stock solution to 1 mg/mL with PBS, fill plastic
Isolated Cells on culture dishes, and incubate overnight at 37  C. On the next
HA-Coated Dishes day, aspirate the HA solution and wash once with PBS. Aspirate
PBS and dry in a clean bench for more than 30 min.
2. Mix 0.5 mL of the suspension of the isolated cells and 1.5 mL
trypan blue solution.
3. Count the number of viable and dead cells, which are smaller
than MHs and larger than nonparenchymal cells, using Neu-
bauer improved hemocytometer (see Note 13).
4. Adjust cell density by diluting the suspension with an appropri-
ate volume of the culture medium.
36 Junichi Kino et al.

Fig. 2 Formation of SH colonies on hyaluronic acid (HA)-coated dish. (a) One day after seeding, SHs adhere on
HA-coated dish. (b) SHs proliferate to form colonies at 9 days after plating, of which consist small-sized cells.
(c) All cells in colonies express CD44 on their cell surfaces. Scale bar: 200 μm

5. Wash dry HA-coated dishes with PBS just before cell seeding.
6. Seed the cells (2.0  104 cells/cm2) on the HA-coated dishes.
7. Place the dishes in 5% CO2/95% air-incubator at 37  C for 3 h.
8. Replace the medium with fresh medium (see Note 14).
9. Place the dishes in 5% CO2/95% air-incubator at 37  C for
9 days, and replace the medium every other day (see Fig. 2 and
Note 15).

3.3 Isolation of 1. Wash the SH colonies on the HA-coated dishes with PBS.
CD44-Positive Cells 2. Treat the colonies with 0.02% EDTA/PBS for 5 min (see Note
from the Cells on 16).
HA-Coated Dishes
3. After aspiration of the solution, treat the colonies with the
separation solution for 5 min at 37  C.
 Isolation and Expansion of Rat Hepatocytic Progenitor Cells 37

4. Collect the detached cells from HA-coated dishes into a ster-

ilized 100-mL beaker.
5. Stir the cell suspension using a magnetic stirrer and stir bar for
30 min at 37  C.
6. Collect the cell suspension and centrifuge it at 150  g for
5 min at 4  C.
7. Aspirate the supernatant and suspend the cell pellet in 20 mL
MACS buffer (see Note 17).
8. Centrifuge the cell suspension at 150  g for 5 min at 4  C.
9. Aspirate the supernatant and suspend the cell pellet in 1-mL
mouse anti-rat CD44 antibody solution (1:1000 dilution,
diluted in MACS buffer; see Note 18).
10. Incubated the cell suspension for 60 min on ice.
11. Add 20 mL MACS buffer to the cell suspension and centrifuge
it at 150  g for 5 min at 4  C.
12. Aspirate the supernatant and suspend the cell pellet in 1 mL
microbead-conjugated anti-mouse IgG2a+b solution (1:5 dilu-
tion, diluted in MACS buffer; see Note 19).
13. Incubate the cell suspension for 30 min on ice.
14. Add 20 mL MACS buffer to the cell suspension and centrifuge
it at 150  g for 5 min at 4  C.
15. Aspirate the supernatant and suspend the cell pellet in 2 mL
MACS buffer.
16. Install LS column on a MidiMACS separator, and set a 50-mL
conical tube under the column.
17. Apply 2 mL MACS buffer onto the column.
18. Apply the cell suspension onto the column after the comple-
tion of dripping the filtrate from the column (see Note 20).
19. Apply 2 mL MACS buffer onto the column after the comple-
tion of dripping the filtrate from the column twice.
20. Remove the column from the MidiMACS separator after the
completion of dripping the filtrate from the column.
21. Place the column on the 15-mL conical tube.
22. Add 5 mL MACS buffer to the column and immediately flush
out a fraction of CD44-positive cells using the plunger.

3.4 Coating Dishes 1. Dispense each aliquot of EHS-gel in a cryogenic tube and store
with EHS-Gel it at 20  C until use.
2. Thaw EHS-gel stock solution on ice.
3. Dilute EHS-gel with HBSS on ice and adjust the concentration
to 0.2 mg protein/mL (see Note 21).
38 Junichi Kino et al.

4. Pour the diluted EHS-gel solution into each plastic culture

dish (1 mL/35-mm, 2 mL/60-mm, 5 mL/100-mm dish).
5. Incubate the dishes at 37  C overnight.
6. Aspirate the solution in the dishes and dry them in a clean
bench.
7. Sterilize the dishes with UV irradiation for about 3 min.

3.5 Culture of the 1. Mix 1 volume of the suspension of CD44+ cells and 3 volumes
CD44-Positive (CD44+) of trypan blue solution.
Cells on the EHS-Gel- 2. Count the number of viable and dead cells using a Neubauer
Coated Dishes improved hemocytometer.
3. Centrifuge the cell suspension at 150  g for 5 min at 4  C.
4. Aspirate the supernatant and suspend the cell pellet in an
appropriate volume of the culture medium for adjusting cell
density (see Note 22).
5. Wash dry EHS-gel-coated dishes with PBS just before cell
seeding.
6. Seed the cells on the EHS-gel-coated dishes.
7. Place the dishes in 5% CO2/95% air-incubator at 37  C for 3 h.
8. Replace the medium with fresh culture medium.
9. Place the dishes in 5% CO2/95% air-incubator at 37  C, and
replace the medium every other day (see Fig. 3 and Note 23).

3.6 Passage of the 1. Wash the CD44+ cells on the EHS-gel-coated dishes with PBS.
CD44+ Cells on the 2. Incubate the cells with PBS containing 0.25% trypsin and
EHS-Gel-Coated 0.01% EDTA in 5% CO2/95% air-incubator at 37  C for
Dishes 10 min.
3. Dissociate the cells by pipetting gently.
4. Add the culture medium to the cell suspension to stop the
digestion by trypsin.
5. Centrifuge the cell suspension at 150  g for 5 min at 4  C.

Fig. 3 Growth of sorted CD44+ SHs on an EHS-gel-coated dish. Attached CD44+ SHs can grow to form
colonies on EHS-gel-coated dishes. The same colony was followed until 28 days after the passage. All photos
show the same magnification. Scale bars; 200 μm
 Isolation and Expansion of Rat Hepatocytic Progenitor Cells 39

6. Aspirate the supernatant and suspend the cell pellet in the

culture medium.
7. Mix 1 volume of the suspension of CD44+ cells and 3 volumes
of trypan blue solution.
8. Count the number of viable and dead cells using a Neubauer
improved hemocytometer.
9. Wash dry EHS-gel-coated dishes with PBS just before cell
seeding.
10. Dilute the cell suspension with an appropriate volume of the
culture medium for adjusting cell density (see Note 22).
11. Seed the cells on the EHS-gel-coated dishes.

4 Notes

1. Insulin is dissolved under acidic condition.

2. Collagenase must be dissolved just before the perfusion and
not be shaken vigorously, because its activity is easily decreased.
3. HA should be sterilized with UV irradiation. HA is not dis-
solved easily in PBS at room temperature; therefore, HA-stock
solution should be made until 1 day before coating culture
dishes with HA.
4. If the ligation is loose, it may cause an inadequate digestion by
collagenase.
5. Pre-perfusion is performed for blood removal and calcium
chelating by EGTA in the liver. This step needs for a good
collagenase digestion.
6. The perfusate is flowed out from the cut ends of inferior vena
cava. Be careful not to store the perfusate in the abdominal
cavities. Retention of the collagenase solution causes the early
rupture of hepatic capsules.
7. Heating by a lamp makes collagenase digestion in the liver
better, because collagenase activity is the highest at 37  C.
8. Collagenase can digest the liver tissue well by reducing the flow
rate of perfusion.
9. The liver becomes soft by the collagenase digestion. Be careful
not to disrupt the hepatic capsule during removing the tissues
adjacent to the liver.
10. MHs have higher density than SHs and nonparenchymal cells,
so MHs are removed from the cell suspension by low-speed
and short-time centrifuge.
11. This centrifuge is carried out to remove hematopoietic cells.
40 Junichi Kino et al.

12. This centrifuge is carried out to damage the remaining MHs in

the cell suspension.
13. Count the number of cells as soon as trypan blue solution is
added. Dead cells have nuclei stained with trypan blue. The
viability of cells is about 80–90% when the isolation procedure
is performed successfully. The lower viability may be caused by
an inadequate digestion by collagenase. In this case, modify the
perfusion procedure, e.g., the collagenase activity in the perfu-
sion solution or the time of collagenase digestion by changing
the flow rate of perfusion.
14. Before the medium change, check the adhesion of cells on the
surface of dishes. If the number of attached cells is few, keep the
dishes in the incubator at 37  C for additional 1–2 h. The
adhesion of cells on HA-coated cells is weak on the day of
seeding, so the medium change is performed carefully.
15. During 9 days of culture, SHs formed colonies consisting of
more than 30 cells. When the colonies of SHs are cultured for a
time longer than 9 days, adhesion of colonies on the dishes
becomes weaker.
16. The cells will be detached from the dishes when treated with
0.02% EDTA/PBS, if the colonies weakly adhere on the dishes.
In this case, collect the detached cells in EDTA/PBS and
centrifuge it at 150  g for 5 min at 4  C. The pellet is used
for the subsequent treatment of collagenase (Step 5).
17. After the treatment of collagenase, the procedures should be
performed on ice.
18. One milliliter mouse anti-rat CD44 antibody solution affords
to bind 1  108 cells. If the total cell number is more than this
number, increase the volume of mouse anti-rat CD44 antibody
solution.
19. One milliliter microbead-conjugated anti-mouse IgG2a+b solu-
tion affords to bind 1  108 cells. If the total cell number is
more than this number, increase the volume of mouse anti-rat
CD44 antibody solution.
20. The column can clog with cells in case that the number of cells
is too many. Pipette the suspension above the column in order
to remove the clogging.
21. EHS-gel forms a gel above 10  C. Use pre-cooled pipets, tips,
and tubes when handling EHS-gel.
22. SH growth speed depends on the cell density, the number of
viable cells, which have the potential to attach to the dishes.
The cell density should be more than 1.0  104 cells/cm2 for
expanding SHs.
 Isolation and Expansion of Rat Hepatocytic Progenitor Cells 41

23. The longer a culture period is, the harder the connection
between cells become. The culture period should not be longer
than 4 weeks, if the cells will be passaged (see Subheading 3.6).

References

1. Saxena R, Zucker SD, Crawford JM (2003) 9. Mitaka T, Kojima T, Mizuguchi T, Mochizuki

Anatomy and physiology of the liver. In: Y (1995) Growth and maturation of small
Zakim D, Boyer TD (eds) Hepatology, 4th hepatocytes isolated from adult rat liver. Bio-
edn. Saunders, Philadelphia, pp 3–30 chem Biophys Res Commun 214:310–317
2. Higgins GM, Anderson RM (1931) Experi- 10. Chen Q, Kon J, Ooe H, Sasaki K, Mitaka T
mental pathology of the liver. I: Restoration (2007) Selective proliferation of rat hepatocyte
of the liver of the white rat following partial progenitor cells in serum-free culture. Nat Pro-
surgical removal. Arch Pathol 12:186–202 toc 2:1197–1205
3. Stöcker E, Heine WD (1971) Regeneration of 11. Sasaki K, Kon J, Mizuguchi T, Chen Q, Ooe H,
liver parenchyma under normal and pathologi- Oshima H et al (2008) Proliferation of hepato-
cal conditions. Beitr Pathol 144:400–408 cyte progenitor cells isolated from adult human
4. Overturf K, al-Dhalimy M, Ou CN, livers in serum-free medium. Cell Transplant
Finegold M, Grompe M (1997) Serial trans- 17:1221–1230
plantation reveals the stem-cell-like regenera- 12. Kon J, Ooe H, Oshima H, Kikkawa Y, Mitaka
tive potential of adult mouse hepatocytes. Am J T (2006) Expression of CD44 in rat hepatic
Pathol 151:1273–1280 progenitor cells. J Hepatol 45:90–98
5. Wang MJ, Chen F, Li JX, Liu CC, Zhang HB, 13. Tanimizu N, Ichinohe N, Ishii M, Kino J,
Xia Y et al (2014) Reversal of hepatocyte senes- Mizuguchi T, Hirata K, Mitaka T (2016)
cence after continuous in vivo cell proliferation. Liver progenitors isolated from adult healthy
Hepatology 60:349–361 mouse liver efficiently differentiate to func-
6. Bhatia SN, Underhill GH, Zaret KS, Fox IJ tional hepatocytes in vitro and repopulate liver
(2014) Cell and tissue engineering for liver tissue. Stem Cells 35:2889–2901
disease. Sci Transl Med 6:245sr2 14. Mitaka T, Sato F, Mizuguchi T, Yokono T,
7. Mitaka T (1998) The current status of primary Mochizuki Y (1999) Reconstruction of hepatic
hepatocyte culture. Int J Exp Pathol organoid by rat small hepatocytes and hepatic
79:393–409 nonparenchymal cells. Hepatology 29:111–125
8. Mitaka T, Mikami M, Sattler GL, Pitot HC, 15. Ishii M, Kino J, Ichinohe N, Tanimizu N,
Mochizuki Y (1992) Small cell colonies appear Ninomiya T, Suzuki H et al (2017) Hepatocy-
in the primary culture of adult rat hepatocytes tic parental progenitor cells of rat small hepa-
in the presence of nicotinamide and epidermal tocytes maintain self-renewal capability after
growth factor. Hepatology 16:440–447 long-term culture. Sci Rep 7:46177
 Part II

Characterization of Liver Progenitors In Vivo

 Chapter 5

Genetic Lineage Tracing of Biliary Epithelial Cells

Teresa Rubio-Tomás, Beatriz Aguilar-Bravo, and Pau Sancho-Bru

Abstract
Lineage tracing of liver cells is a powerful tool to understand liver embryonic development, healthy liver cell
homeostasis, tissue repair, and regeneration. Lineage tracing of biliary epithelial cells (BECs) in the adult
liver has been used to assess the contribution of the biliary epithelium to liver injury, regeneration, and
disease. These studies have shown the contribution of BECs to the expansion of ductular reaction (DR) and
liver progenitor cells (LPCs) and eventually the generation of new hepatocytes. Few genetic lineage-tracing
mouse models have been proved to trace BECs. This chapter is focused on lineage tracing of BECs in mouse
models of liver injury and regeneration. First, we mention different existing approaches to trace the biliary
epithelium based on proteins specifically expressed by BECs such as sex-determining region Y-box
9 (SOX9), osteopontin (OPN), and cytokeratin-19 (KRT19). Second, we describe mouse models that
can be used to evaluate cell fate during liver injury and regeneration (i.e., partial hepatectomy (PHx), acute
liver injury models, and chronic liver damage models such as 3,5-diethoxycarbonyl-1,4-dihydro-collidin
(DDC) diet, choline-deficient ethionine-supplemented (CDE) diet, or chronic carbon tetrachloride (CCl4)
administration). Third, we suggest possible readouts to assess BECs fate based on immunofluorescence
analysis.

Key words Lineage tracing, Liver regeneration, Biliary epithelial cells, Liver progenitor cells, Ductular
reaction, Chronic liver injury, Animal models

1 Introduction

In the last years, development of lineage-tracing technology has

allowed the possibility of better understanding cell fate in liver
development as well as in the context of liver injury and regenera-
tion, tissue repair, and cancer. Besides its potential to investigate
stem cell commitment and differentiation, the use of lineage tracing
in adult organs with regenerative capacity has provided important
information to understand tissue cell homeostasis, regeneration,
and disease. In healthy condition the liver is a slow-cycling organ;
however, after liver damage resulting from diverse insults, loss of
liver mass, toxicity, or diseases, the liver rapidly regenerates
recovering the liver cell mass and function. When hepatocyte pro-
liferation is intact, such as in the model of partial hepatectomy

Naoki Tanimizu (ed.), Hepatic Stem Cells: Methods and Protocols, Methods in Molecular Biology, vol. 1905,
https://doi.org/10.1007/978-1-4939-8961-4_5, © Springer Science+Business Media, LLC, part of Springer Nature 2019

45
46 Teresa Rubio-Tomás et al.

(PHx), liver regeneration and recovery of the cell mass is due to

mature cell proliferation, comprising proliferation of hepatocytes
and cholangiocytes but also other non-parenchymal cells. How-
ever, when hepatocyte regeneration and replication is compromised
such as in severe liver injury or in the context of chronic liver
diseases, biliary epithelial cells (BECs) can undergo substantial
morphological, molecular, and functional changes giving rise to
the ductular reaction (DR). This is particularly relevant under
situations comprising biliary injury such as biliary diseases or in
highly cholestatic advanced chronic liver diseases such as alcoholic
cirrhosis and alcoholic hepatitis, in which an important biliary
reaction can be observed [1]. DR is a heterogeneous population
of cells comprising reactive cholangiocytes as well as more imma-
ture cells with a liver progenitor cell (LPC) phenotype character-
ized by the expression of both hepatocyte and biliary markers as
well as stem cell markers [2, 3].
Liver cells have been shown to be highly plastic. Several studies
using lineage-tracing strategies have shown the conversion of bili-
ary cells to mature hepatocytes, as well as the transdifferentiation of
hepatocytes into biliary cells. Moreover, both cell types have been
shown to be able to generate facultative LPCs with the capacity to
proliferate and to differentiate into hepatocyte and cholangiocytes
[4–7]. LPCs have not been identified in the healthy liver, but they
appear and expand after injury, when hepatocytes regenerative
capacity is compromised. Huch et al. used Lgr5-LacZ mice to
prove that this population expresses the stem cell marker Lgr5
[8]. LPCs arise in several chronic liver diseases and animal models,
and its expansion in the parenchymal region correlates with the
degree of liver injury and progression of the disease [9, 10]. LPCs
stain positively for BECs markers and often form clusters with duct-
like and/or cord-like structures.
The important plasticity of the liver and the phenotypic and
functional changes of the BECs reinforce the importance of
lineage-tracing strategies to investigate the origin and fate of cells
of the DR and to understand the contribution of the biliary epithe-
lium to liver regeneration. Moreover, it can be a powerful tool to
investigate the molecular mechanisms underlying the development
of DR and the generation of facultative LPCs [11].
Few genetic lineage-tracing systems have been developed to
trace BECs. Carpentier et al. used sex-determining region Y-box
9 (SOX9)-CreERT to trace embryonic ductal epithelium cells.
They found that these cells gave rise to cholangiocytes lining all
segments of the intrahepatic bile ducts, as well as periportal hepa-
tocytes [12]. Español-Suñer et al. showed that osteopontin (OPN)-
labeled adult liver cells have the potential to give rise to transit-
amplifying progenitor cells and mature hepatocytes in mouse
chronic injury models [13]. Lineage-tracing studies performed
with the BEC marker cytokeratin-19 (KRT19) have shown
 Lineage Tracing of Biliary Epithelium 47

different outcomes depending on the mouse model that was used.

In most animal models (i.e., 3,5-diethoxycarbonyl-1,4-dihydro-
collidin [DDC] diet, choline-deficient ethionine-supplemented
[CDE] diet, partial hepatectomy [PHx] and carbon tetrachloride
[CCl4]), there was no evidence that BECs and other
non-hepatocyte populations contribute significantly to hepatocyte
neogenesis. However, more stringent models inducing hepatocyte
cell arrest showed the generation of hepatocytes from cholangiocytes
[14, 15]. Moreover, lineage-tracing models are also powerful tools
for studying cancer development. Guest et al. used KRT19-CreERT
to target p53 loss in cholangiocytes and observed the appearance of
labeled intrahepatic cholangiocarcinoma in response to chronic
injury [16]. In Rodrigo-Torres et al., we used the hepatocyte nuclear
factor 1-beta (HNF1β) lineage-tracing animal model to trace the
biliary compartment under chronic injury [17]. In this study we
showed the BECs expressing HNF1β give rise to the DR. However,
under the animal models tested, the contribution of HNF1β cells to
newly generated hepatocytes was absent or minimal. The procedures
described in this protocol are based on this study.
Unfortunately lineage-tracing studies have limitations. First,
lack of Cre expression in some target cells could lead to difficulties,
especially when reduced populations are targeted. Second, Cre
expression in non-targeted cell types can complicate its interpreta-
tion and lead to controversy. In order to overcome the limitations
due to poor specificity, recent studies have proposed more robust
genetic lineage-tracing system that incorporates double recombi-
nation based on Dre–rox system to minimize the lack of Cre
specificity [18]. Besides intrinsic technical limitations, it is impor-
tant to mention that animal models do not fully reproduce the
complexity of chronic liver diseases and, therefore, results obtained
from lineage-tracing studies have to be taken cautiously when
translating the findings into human disease.

2 Materials

All chemicals and reagents are commercially available products of

analytical grade.

2.1 Mouse Strains Mice expressing Cre recombinase under the control of a cell-
specific promoter can be used to induce site-specific recombination
2.1.1 Cre-Expressing
between two loxP sites (Fig. 1). The Cre-loxP system can be used
Lines
to generate conditional knockout mice by excising part of a
protein-coding sequence and, therefore, inhibiting its expression.
Cre recombinase sequence can be fused to the estrogen receptor
sequence (CreER), and, therefore, Cre recombinase activity
becomes tamoxifen-inducible. The use of inducible CreER is par-
ticularly interesting in the context of lineage tracing, since expres-
sion of markers may vary along embryonic development, which
48 Teresa Rubio-Tomás et al.

(-) Tamoxifen
ER
Cre
Cytoplasm

ER
Cre

Hnf1β Rosa26
Cre-ERT2 STOP YFP
promoter promoter Nucleus

(+) Tamoxifen

Tm
ER
ER + Tm Cre
Cre Cytoplasm

ER Tm
Cre YFP

Hnf1β Rosa26
Cre-ERT2 YFP
promoter promoter
Nucleus

Tm
ER ER
Cre Cre Tm

Inactive Cre Active Cre Tamoxifen

Fig. 1 Tamoxifen-inducible Cre-loxP system. Cre recombinase sequence is fused to hormone-binding domain
of the estrogen receptor (CreER). CreER expression is under the control of a biliary epithelial cell (BEC)-specific
promoter (e.g., hepatocyte nuclear factor 1-beta [Hnf1β] promoter). In the absence of tamoxifen, CreER
remains inactive in the cytoplasm. In the presence of tamoxifen, CreER activates and translocates to the
nucleus, where it recombines the region flanked by LoxP sites (red triangles). After the release of the stop
codon, yellow fluorescent protein (YFP) is permanently expressed in the cell and its progeny

may lead to confounding results. Different Cre mouse lines can be

used for lineage tracing of BEC cells: Hnf1β-CreER (027681, The
Jackson Laboratory, United States of America) [17, 19]; SOX9-
CreERT2 mice (018829, The Jackson Laboratory, United States of
America) [12, 20]; OPN-iCreERT2 mice [13] and KRT19-CreERT
mice (026925, The Jackson Laboratory, United States of America)
[15, 21].

2.1.2 Reporter Mouse In order to generate lineage-tracing animals, cell-specific CreER

Lines mice should be crossed with mice bearing Cre-inducible Rosa26R
reporters such as LacZ (the gene that codifies for β-galactosidase
[β-GAL] protein) or yellow fluorescent protein (YFP), among
others (see Note 1). Lineage-tracing mice are based on the presence
of a stop codon flanked by LoxP sites located between the promoter
and the reporter gene. When Cre recombinase is expressed, the
 Lineage Tracing of Biliary Epithelium 49

stop codon is excised and the protein sequence located downstream

is expressed. When inducible CreER is used, mice will need to be
treated with tamoxifen to induce Cre-mediated LoxP site recombi-
nation. Several mouse reporter lines are available, among others:
ROSA26RYFP mice (006148, The Jackson Laboratory, United
States of America); ROSA26RlacZ mice (002073, The Jackson
Laboratory, United States of America) and Gt(ROSA)26Sortm4
(ACTB-tdTomato,-EGFP)Luo (007676, The Jackson Laboratory,
United States of America).

2.2 Liver Injury and 1. Tamoxifen.

Regeneration Models 2. Corn oil.
3. Surgical instruments, reagents, and equipment required for
PHx has been previously described [22].
4. 500 mg/kg mice body weight acetaminophen (APAP).
5. 0.5 mL/kg mice body weight CCl4. CCl4 is diluted 1:4 in
corn oil.
6. Standard rodent chow diet containing 0.1% DDC.
7. Standard chow diet.
8. Choline-deficient (MP Biomedicals, United States of America),
ethionine-supplemented (0.15% in water) (CDE) diet.

2.3 Isolation of BEC- 1. EGTA solution: 137 mM NaCl, 5.4 mM KCl, 0.64 mM NaH2-
Derived Cells PO4H2O, 0.85 mM Na2HPO4, 1 mM HEPES, 4.17 mM
NaHCO3, 0.5 mM EGTA, and 5 mM glucose, pH ¼ 7.4.
2. Collagenase solution: 0.5 g/L Collagenase A in 137 mM
NaCl, 5.4 mM KCl, 0.64 mM NaH2PO4H2O, 0.85 mM
Na2HPO4, 1 mM HEPES, 4.17 mM NaHCO3, and 3.8 mM
CaCl2·2H2O, pH ¼ 7.4.
3. Collagenase A (Roche Diagnostics GmbH, Germany).
4. Hank’s Balanced Salt Solution (HBSS) (Biological Industries,
United States of America).
5. 70-μm cell strainer (Miltenyi Biotec, Germany).
6. Pronase (Roche Diagnostics GmbH, Germany).
7. DNase I (Roche Diagnostics GmbH, Germany).
8. Live/Dead cell stain kit (Life Technologies, United States of
America).
9. RLT buffer (Qiagen GmbH, Germany).

2.4 Histological Histochemical procedures, immunostaining, colocalization studies,

and Biochemical and serum analysis are performed using the following materials:
Analysis
1. Antibodies against BEC markers: anti-KRT19 (Developmental
Studies Hybridoma Bank, University of Iowa, United States of
America), anti-epithelial cell adhesion molecule (EpCAM)
50 Teresa Rubio-Tomás et al.

(M0804, DakoCytomation, United States of America, and

552370, BD Pharmingen, United States of America), anti-
HNF1β (Sc-22840 and Sc-7411, Santa Cruz, United States
of America), and anti-SOX9 (AB5535, Millipore Merck,
United States of America).
2. Antibodies against hepatocytes, such as anti-hepatocyte nuclear
factor 4-alpha (HNF4α) (ab181604, Abcam, United Kingdom).
3. Antibodies against neutrophil markers, such as anti-
myeloperoxidase (MPO) (ab9535, Abcam, United Kingdom).
4. Antibodies against macrophage markers, such as anti-F4/80
receptor (MCA497-R, Serotec, United States of America).
5. Antibodies against reporter proteins: anti-β-GAL (A11132,
Invitrogen and ab9361, Abcam, United Kingdom), anti-CRE
[23], and anti-green fluorescent protein (GFP) (ab6673,
Abcam, United Kingdom, and A6455, Invitrogen, United
States of America) (see Note 2).
6. Kits for detection of apoptosis: commercial TUNEL kit
(12156792910—In Situ Cell Death Detection Kit, TMR red,
Merck, United States of America) and caspase activity kit
(G8091, Caspase-Glo 3/7 Assay, Promega, United States of
America).
7. Kits for detection of alanine aminotransferase (ALT), aspartate
aminotransferase (AST), lactate dehydrogenase (LDH), and
alkaline phosphatase (AP) serum levels (ADVIA 1800 Clinical
Chemistry System, Siemens Healthcare, Germany).

3 Methods

3.1 Induction Tamoxifen is administered to induce Cre-mediated recombination.

of Expression
1. Dissolve tamoxifen at 100 mg/mL in corn oil (0.9% NaCl and
of the Reporter
10% absolute ethanol to facilitate sonication).
Proteins by Tamoxifen
2. Treat mice with tamoxifen intraperitoneally every 24 h at a dose
of 75 mg/kg for 5 consecutive days (see Note 3).
3. All animal models of liver injury were started 1 week after the
last tamoxifen treatment.

3.2 Induction of Liver Two-thirds PHx in rodents is a well-established model to study

Injury liver regeneration (Fig. 2a). Recovery of the liver cell mass after
and Regeneration PHx is mediated by processes involving mature cell proliferation,
without extensive necrosis or inflammation. Moreover, as the
3.2.1 Partial regeneration process starts immediately after PHx, this model can
Hepatectomy be precisely controlled. As shown in Rodrigo-Torres et al., after
PHx there was no contribution of HNF1β cells to hepatocytes.
 Lineage Tracing of Biliary Epithelium 51

-1 0 1 or 4
(A) weeks

Tamoxifen 2/3
PHx

-1 0 2
(B) weeks

Tamoxifen 500 mg/kg APAP

-1 0 1
(C) weeks

Tamoxifen 0.5 ml/kg CCl4

-1 0 4
(D) weeks Standard
DDC diet
diet

Tamoxifen

-1 0 3

(E) weeks Standard

diet CDE diet

Tamoxifen

-1 0 1 2 3 4 5 6 7 8
(F) weeks

Tamoxifen CCl4
0.5 ml/kg

Fig. 2 Experimental models of liver regeneration. (a) Partial hepatectomy (PHx), (b) acute acetaminophen
(APAP) treatment, (c) acute carbon tetrachloride (CCl4) administration, (d) 3,5-diethoxycarbonyl-1,4-dihydro-
collidin (DDC) diet, (e) choline-deficient ethionine-supplemented (CDE) diet, and (f) chronic CCl4 administration
52 Teresa Rubio-Tomás et al.

1. Anesthetize mice.
2. Perform two-thirds PHx in mice as previously described [22].
3. Sacrifice 1 week after PHx.

3.2.2 Acute APAP APAP model (Fig. 2b) induces hepatocellular death and release of
Treatment danger-associated molecular patterns (DAMPs) leading to macro-
phages and NKT cells infiltration. This model mimics drug-
induced injury. Mice treated with APAP show a mild induction of
DR expansion in the liver. In Rodrigo-Torres et al., no
BEC-derived hepatocyte generation was observed when assessed
by dual staining of HNF4α/YFP.
1. Fast mice for 8 h with free access to water.
2. Inject intraperitoneally a single dose of 500 mg/kg body
weight APAP diluted in DPBS at 35–40  C.
3. Inject control mice with equivalent amount of DPBS.
4. Sacrifice mice 2 weeks after the injection.

3.2.3 Acute CCl4 Acute treatment with CCl4 (Fig. 2c) is a widely used model of acute
Treatment liver failure caused by toxic substances. CCl4 induces inflammation,
thrichloromethyl radical formation, and production of reactive
oxygen species (ROS). This leads to lipid peroxidation and hepato-
toxicity [24]. As described in Rodrigo-Torres et al., acute CCl4-
treated mouse livers showed a mild induction of DR expansion.
However, no duct-derived hepatocyte generation was observed
based on dual staining of HNF4α/YFP (Fig. 3a).
1. Inject intraperitoneally a single dose of CCl4 at 0.5 mL/kg,
10–25% diluted in corn oil.
2. Inject control mice with equivalent amount of corn oil.
3. Sacrifice 1 week after the injection.

3.2.4 Chronic Liver DDC diet (Fig. 2d) induces liver damage and the expansion of DR
Damage Models and LPCs [25]. This model has been used to study the activation
with DDC Diet and proliferation of LPCs, but most especially it has been used as a
model of primary sclerosing cholangitis and biliary fibrosis. In
addition to DR, the phenotype observed in this model is character-
ized by fibrosis, inflammatory cells infiltration around periportal
areas, cholestasis and expression of inflammatory cytokines.
In Rodrigo-Torres et al., after 4 weeks of DDC diet an impor-
tant DR was observed, which was derived from HNF1β+ cells.
However, there was no contribution of HNF1β+ cells to newly-
formed hepatocytes. During recovery (2 weeks) with standard
chow, DR was reduced, but once again, HNF1β+ cells did not
contribute to hepatocyte generation.
 Lineage Tracing of Biliary Epithelium 53

Fig. 3 Lineage tracing of HNF1β-expressing cells after liver injury. (a) In acute carbon tetrachloride (CCl4) liver
injury model, biliary epithelium cells (BECs) do not give rise to hepatocytes. Hepatocytes express nuclear
hepatocyte nuclear factor 4 alpha (HNF4α) (red) and cells derived from hepatocyte nuclear factor 1-beta
(HNF1β)+ BECs express yellow fluorescent protein (YFP) (green). (b) Image from choline-deficient ethionine-
supplemented (CDE) diet liver injury model. Hepatocytes express HNF4α (green) and cells derived from
HNF1β+ cells express YFP (red). A small number of hepatocytes express the reporter YFP indicating that they
are derived from HNF1β+ cells

1. Feed mice ad libitum with a standard rodent chow diet contain-

ing 0.1% DDC for 4 weeks. To induce an injury-recovery
model, feed mice for 4 weeks with DDC diet followed by
2 weeks of standard chow diet recovery (see Note 4).
2. Feed control animals with standard chow for the entire period.
3. Sacrifice mice when the protocol is completed.

3.2.5 Chronic Liver CDE diet (Fig. 2e) is used as a model of expansion of DR cells.
Damage Models Unlike DDC diet, CDE diet does not induce biliary injury, but
with CDE Diet steatosis and a significant hepatocellular damage lead to failure of
hepatocytes function [26]. CDE model also induces inflammatory
cells proliferation and the expression of pro-inflammatory cytokines
such as tumor necrosis factor alpha (TNF-α) and interleukin 1 beta
(IL-1β).
In Rodrigo-Torres et al., after 3 weeks of CDE treatment, there
was proliferation of DR cells derived from HNF1β+ cells. More-
over, there was a small population of YFP+ hepatocytes in the
periportal area (0.22%). This percentage increased to 1.86% after
2 weeks of recovery (Fig. 3b).
1. Feed mice with CDE diet ad libitum for 3 weeks. To perform
an injury-recovery model, feed the animals for 3 weeks with
CDE diet followed by 2 weeks of standard diet (see Note 4).
2. Feed control animals with standard chow.
3. Sacrifice mice when the protocol is completed.
54 Teresa Rubio-Tomás et al.

3.2.6 Chronic CCl4 Long-term treatment with CCl4 (Fig. 2f) induces hepatotoxicity
Administration leading to fibrosis, bile duct proliferation, and, eventually, hepato-
cellular carcinoma. Chronic CCl4 administration is one of the most
widely used models for fibrosis induction. CCl4 treatment does not
compromise hepatocyte replication. This model shows a limited
expansion of DR.
1. Inject intraperitoneally CCl4 at a dose of 0.5 mL/kg, twice a
week for 8 weeks (see Note 4).
2. Inject control mice with equivalent amount of corn oil.
3. Sacrifice 4 days after the last injection.

3.3 Isolation of BEC- Lineage tracing models allow isolating BEC-derived cells by flow
Derived Cells cytometry cell sorting.
1. Anesthetize mice and cannulate inferior cava vein.
2. Perfuse livers with EGTA solution at a flow rate of 5 mL/min
until liver gets pale.
3. Perfuse with collagenase solution containing Collagenase A
0.5 g/L at a flow rate of 5 mL/min for 11 min.
4. Mince digested liver in Petri dish with ice cold 1 HBSS.
5. Redigest with collagenase solution containing 0.5 g/L Col-
lagenase A, 0.5 g/L pronase and 50 mg/L DNAse I.
6. Stir for 30 min at 37  C.
7. Filter liver through 70-μm cell strainer.
8. Centrifuge at 100  g for 2 min and keep supernatant.
9. Centrifuge supernatant at 300  g for 5 min and keep the
pellet.
10. Sort YFP+ cells by flow cytometry. Use cells from wild type
(WT) mice to establish YFP+ gate. Collect YFP+ cells for fur-
ther analysis.

3.4 Histological and 1. Perform histochemical procedures, immunostaining, and colo-

Biochemical Analysis calization studies using the abovementioned antibodies (see
Note 5).
2. Evaluate apoptosis by TUNEL and caspases activity assays.
3. Analyze hepatic damage by measuring serum levels of ALT,
AST, LDH and AP.
 Lineage Tracing of Biliary Epithelium 55

4 Notes

1. Cre expression may vary among different Cre mice. In addi-

tion, the efficiency of recombination could differ depending on
the combination of CreER-expressing mouse and
Cre-inducible Rosa26R reporter animal used for generating
the lineage-tracing model. Therefore, it is important to evalu-
ate the efficiency of recombination in each of the models
generated.
2. GFP and YFP fluorescence cannot be distinguished by simply
observing the liver sections under the microscope due to auto-
fluorescence of the liver. Specific GFP/YFP signal should be
detected using specific antibodies.
3. A high dose of tamoxifen is known to generate liver damage
that could induce ectopic expression of the gene of interest and
therefore of the reporter. It is important to optimize the
amount of tamoxifen to ensure specific labeling. Intraperito-
neal injection of tamoxifen is the most common administration
method because the amount of administered compound can be
better controlled. Doses ranging from 50 to 100 mg/kg are
usually injected once per day for 5 consecutive days. In
Rodrigo-Torres et al., mice were treated with 75 mg/kg, a
widely used dose. Delivery by oral gavage is also possible. For
oral administration the amount can range from 25 to 175 mg/
kg for 5 consecutive days. Higher doses, ranging from 500 to
1000 mg/kg, can be used in shorter treatments: e.g., in
Rodrigo-Torres et al., mice were treated on day 1 with
1000 mg/kg, on day 3 with 1000 mg/kg, and on day 5 with
500 mg/kg tamoxifen by oral gavage.
4. These protocols require daily monitoring of body weight. For
DDC and CDE diet protocols, pellets could also be weighted
to monitor food intake in each cage. Although a degree of
weight loss is expected, rapid weight loss of 15–20% within a
few days is a commonly accepted humane end point.
5. Immunostaining conditions should be optimized for every
mouse model. Livers from mice fed with DDC diet contain
large amounts of porphyrin crystals. These crystals are auto-
fluorescent, leading to difficulties in immunofluorescent
staining.
56 Teresa Rubio-Tomás et al.

Acknowledgments

We want to thank Daniel Rodrigo-Torres for the advice during

manuscript preparation and the experimental work that this chapter
is based on. This work was supported by grants from the Fondo de
Investigación Sanitaria Carlos III (FIS), co-financed by the Fondo
Europeo de Desarrollo Regional (FEDER), Unión Europea, “Una
manera de hacer Europa” (FIS PI14/00320, PI 17/00673 to
PS-B). PS-B and BA-B are funded by the Instituto de Salud Carlos
III, Miguel Servet (CPII16/00041) and PFIS, respectively, and
co-financed by the Fondo Europeo de Desarrollo Europeo
(FEDER), Unión Europea, “Una manera de hacer Europa.”

References
1. Arias IM et al (2009) The liver: biology and 12. Carpentier R et al (2011) Embryonic ductal
pathobiology, 5th edn. John Wiley & Sons, plate cells give rise to cholangiocytes, periportal
Hoboken hepatocytes, and adult liver progenitor cells.
2. Gouw ASH, Clouston AD, Theise ND (2011) Gastroenterology 141(4):1432–1438.e4
Ductular reactions in human liver: diversity at 13. Español-Suñer R et al (2012) Liver progenitor
the interface. Hepatology 54(5):1853–1863 cells yield functional hepatocytes in response to
3. Okabe M et al (2009) Potential hepatic stem chronic liver injury in mice. Gastroenterology
cells reside in EpCAM+ cells of normal and 143(6):1564–1575.e7
injured mouse liver. Development 136 14. Raven A et al (2017) Cholangiocytes act as
(11):1951–1960 facultative liver stem cells during impaired
4. Kholodenko IV, Yarygin KN (2017) Cellular hepatocyte regeneration. Nature 547
mechanisms of liver regeneration and cell- (7663):350–354
based therapies of liver diseases. Biomed Res 15. Yanger K et al (2014) Adult hepatocytes are
Int 2017:1–17 generated by self-duplication rather than stem
5. Lu W-Y et al (2015) Hepatic progenitor cells of cell differentiation. Cell Stem Cell 15
biliary origin with liver repopulation capacity. (3):340–349
Nat Cell Biol 17(8):971–983 16. Guest RV et al (2014) Cell lineage tracing
6. Pu W et al (2016) Mfsd2a+ hepatocytes repop- reveals a biliary origin of intrahepatic cholan-
ulate the liver during injury and regeneration. giocarcinoma. Cancer Res 74(4):1005–1010
Nat Commun 7:13369 17. Rodrigo-Torres D et al (2014) The biliary epi-
7. Schaub JR, Malato Y, Gormond C, Willenbr- thelium gives rise to liver progenitor cells.
ing H (2014) Evidence against a stem cell ori- Hepatology 60(4):1367–1377
gin of new hepatocytes in a common mouse 18. He L et al (2017) Enhancing the precision of
model of chronic liver injury. Cell Rep 8 genetic lineage tracing using dual recombi-
(4):933–939 nases. Nat Med 23(12):1488–1498
8. Huch M et al (2013) In vitro expansion of single 19. Solar M et al (2009) Pancreatic exocrine duct
Lgr5+ liver stem cells induced by Wnt-driven cells give rise to insulin-producing beta cells
regeneration. Nature 494(7436):247–250 during embryogenesis but not after birth.
9. Sancho-Bru P et al (2012) Liver progenitor cell Dev Cell 17(6):849–860
markers correlate with liver damage and predict 20. Kopp JL et al (2011) Sox9+ ductal cells are
short-term mortality in patients with alcoholic multipotent progenitors throughout develop-
hepatitis. Hepatology 55(6):1931–1941 ment but do not produce new endocrine cells
10. Guy CD et al (2012) Hedgehog pathway acti- in the normal or injured adult pancreas.
vation parallels histologic severity of injury and Development 138:653–665
fibrosis in human nonalcoholic fatty liver dis- 21. Means AL, Xu Y, Zhao A, Ray KC, Gu G
ease. Hepatology 55(6):1711–1721 (2008) A CK19 CreERT knockin mouse line
11. Grompe M (2017) Fah knockout animals as allows for conditional DNA recombination in
models for therapeutic liver repopulation. epithelial cells in multiple endodermal organs.
Springer, Cham, pp 215–230 Genesis 46(6):318–323
 Lineage Tracing of Biliary Epithelium 57

22. Mitchell C, Willenbring H (2008) A reproduc- toxicological model. Crit Rev Toxicol 33
ible and well-tolerated method for 2/3 partial (2):105–136
hepatectomy in mice. Nat Protoc 3 25. Thompson MD, Awuah P, Singh S, Monga
(7):1167–1170 SPS (2010) Disparate cellular basis of
23. Kellendonk C, Tronche F, Casanova E, improved liver repair in β-catenin-overexpres-
Anlag K, Opherk C, Schütz G (1999) Induc- sing mice after long-term exposure to
ible site-specific recombination in the brain. J 3,5-diethoxycarbonyl-1,4-dihydrocollidine.
Mol Biol 285(1):175–182 Am J Pathol 177(4):1812–1822
24. Weber LWD, Boll M, Stampfl A (2003) 26. Itoh T, Miyajima A (2014) Liver regeneration
Hepatotoxicity and mechanism of action by stem/progenitor cells. Hepatology 59
of haloalkanes: carbon tetrachloride as a (4):1617–1626
 Chapter 6

Specific Labeling and Lineage Tracing of Periportal

Hepatocytes Using Two-Step Genetic Recombination
Nicola de Prisco, Eleanor Stout, and Joan Font-Burgada

Abstract
The liver is unmatched in regenerative capacity. However, when exhausted, the liver is predisposed to
various diseases based on injury types and causal agents. Although hepatocytes have been proposed to be
the main source of new hepatocytes during regeneration, the existence of specialized liver stem cells has
been long debated. In mice, oval cells or ductal cells have been postulated as such stem/progenitor pool.
Exhaustive works from different laboratories have shown that in genetically unmodified mice, oval cells, or
by extension ductal cells, only contribute marginally in producing new hepatocytes during liver regenera-
tion, thus indicating that hepatocytes are the main regenerative cell source. In this debated context, we
identified a new population of periportal hepatocytes in the normal mouse liver. These cells we termed
hybrid hepatocytes (HybHP) express low levels of the transcription factor Sox9. Using complementary
lineage tracing tools, we demonstrated that HybHP regenerate the liver after chronic hepatocyte depleting
injuries. Here, we describe the two-step genetic recombination method that allowed us to study HybHP’s
lineage in two established models of liver injury.

Key words Hybrid periportal hepatocytes, Dual recombinase, Lineage tracing, Liver regeneration

1 Introduction

Development of genetic lineage tracing techniques has shed light

on previously unapproachable questions relating to complex tissue
processes, stem cell dynamics, and cancer development. Liver biol-
ogy has been broadly impacted by the implementation of lineage
tracing in the quest to determine which cell populations are respon-
sible for its staggering regeneration capacity. By tracing the lineage
of oval cells, ductal cells, or hepatocytes in wild-type (WT) mice, it
has become evident that the main source of new hepatocytes in a
large variety of mouse models of liver regeneration is hepatocytes
themselves [1–6]. The negligible contribution found from oval/
ductal cells has challenged the notion of a dedicated liver progeni-
tor pool. However, in genetically modified mice that lack β1-integ-
rin in hepatocytes or that overexpress p21 in hepatocytes, bile duct

Naoki Tanimizu (ed.), Hepatic Stem Cells: Methods and Protocols, Methods in Molecular Biology, vol. 1905,
https://doi.org/10.1007/978-1-4939-8961-4_6, © Springer Science+Business Media, LLC, part of Springer Nature 2019

59
60 Nicola de Prisco et al.

cells produce a significant proportion of new hepatocytes after

injury cessation by outcompeting original hepatocytes with blocked
proliferation potential [7]. Recently, we identified a population of
periportal hepatocytes in the normal mouse liver which express low
levels of Sox9, a highly expressed transcription factor in bile duct
cells [8]. These so-called hybrid hepatocytes (HybHP) are quies-
cent during homeostasis. However, we found that in different
mouse models of hepatocyte injury, HybHP are the cells that
produce the majority of the new hepatocytes that repopulate the
liver. In addition, during cholestatic injury, few metaplastic HybHP
fully transdifferentiate to ductal phenotype. Since Sox9-CreErt2 was
targeting both bile duct cells and HybHP and there was no other
tool to specifically label the latter, we developed a new protocol
based on the sequential utilization of two independent recombi-
nases to clonally label HybHP.
For this purpose, we crossed Sox9-CreErt2 with NZG mice
which contain both a loxP-flanked STOP PGK neo cassette
upstream of a nuclear targeted LacZ marker and a downstream
GFP marker, whose expression is prevented by the Frt-flanked
LacZ cassette itself (Fig. 1). The Sox9-CreErt2 transgene expressed
tamoxifen-inducible Cre in both ductal cells and HybHP. Thus,
after administration of tamoxifen to Sox9-CreErt2;NZG heterozy-
gous mice, the floxed STOP cassette was recombined in ductal cells
and HybHP with concomitant expression of nuclear LacZ with no
cells expressing eGFP (Fig. 2a–c). The sequential used adeno-
associated virus (AAV) TBG-FLPo which only transduce and
express the recombinase in hepatocytes. Transduction of
tamoxifen-treated Sox9-CreErt2;NZG heterozygous mice with
AAV-TBG-FLPo resulted in the excision of LacZ cassette in all
hepatocytes and only allowed eGFP expression by HybHP, which
had previously recombined the STOP cassette (Figs. 1 and 3a).
Staining with different markers for ductal-hepatocyte lineages con-
firmed the specificity of HybHP GFP labeling (Fig. 3b, c).
Once clonal HybHP were GFP labeled, these mice were used to
perform lineage tracing in different models of liver injury, confirm-
ing that HybHP play a significant role in repopulating the liver after
chronic hepatocyte damage. This two-step genetic recombination
protocol can be used to label any population of cells lacking a
specific marker if these cells can be defined by the unique intersec-
tion of two independent markers that otherwise are shared with
other cell types.

2 Materials

2.1 Mice Strains 1. Sox9-CreErt2: This strain was described in Kopp et al., 2011
[9]; Sox9-CreErt2 mice express CreErt2 recombinase in all Sox9-
expressing cells, which in the liver are both hybrid periportal
hepatocytes and bile duct cells.
 Lineage Tracing of Hybrid Periportal Hepatocytes 61

PV Tamoxifen
PV AAV-TBG-FLPo
PV

cHP HybHP Bile duct cHP HybHP Bile duct cHP HybHP Bile duct
Cell Cell Cell

CreErt2 - + + - + + - + + +
nLacZ - - - - + + - - + +
eGFP - - - - - - - + - -

Sox9 CreErt2 Sox9 CreErt2 Sox9 CreErt2

pCAG PGKneo nlsLacZ EGFP pCAG nlsLacZ EGFP pCAG EGFP

Fig. 1 Specific labeling of HybHP using NZG reporter. Schematic outlining of the two-step genetic recombina-
tion using NZG reporter. Sox9-CreErt2;NZG mice express inducible Cre in ductal cells and HybHP (left panel).
Tamoxifen treatment induces Cre-mediated recombination of PGK neo cassette with the subsequent expres-
sion of nuclear LacZ in ductal cells and HybHP (pink nuclei, middle panel). When Sox9-CreErt2;NZG mice
previously treated with tamoxifen are injected with adeno- or AAV-FLPo, which is expressed only in
hepatocytes, the recombination of nuclear LacZ and the subsequent expression of eGFP, only in HybHP,
occur (right panel)

2. NZG: Jax 012429 [10] contains loxP-flanked STOP PGK neo

cassette that prevents the expression of a nuclear-targeted LacZ
marker. Furthermore, Frt sites flanking LacZ prevent in turn
the downstream expression of eGFP (Fig. 1).

2.2 Models of Liver 1. Sox9-Cre Ert2;NZG previously treated with tamoxifen and
Injury injected with AAV expressing FLPo were treated with carbon
tetrachloride (CCl4) to induce liver injury and determine the
clonal behavior of HybHP. Mice are treated with a single dose
(acute) of CCl4 or with 6 and 12 injections of CCl4.
2. To induce cholestatic liver injury and see if HybHP can give rise
to true ductal cells, Sox9-Cre Ert2;NZG previously treated with
tamoxifen and with the AAV were fed with diethoxycarbonyl-
1,4-dihydrocollidine (DDC) diet for 6 weeks.
62 Nicola de Prisco et al.

Fig. 2 First recombination step. (a) Sox9-CreErt2;NZG mice (4–6 weeks old) were injected with 100 mg/kg of
tamoxifen. Ductal cells and HybHP were positive for nuclear LacZ (arrowheads and arrows, respectively). (b)
LacZ+ ductal cells were CK19+ (arrowheads) and HybHP LacZ+ were CK19 (arrows). (c) LacZ+ HybHP were
positive for HNF4α (arrows), while LacZ+ ductal cells were HNF4α . Scale bar 20 μm

2.3 Administration 1. Recombinant adenovirus expressing FLP recombinase

of Adenoviral (Ad-CMV-FLPo, Vector Biolabs #1775).
and Adeno-Associated 2. Recombinant AAV-DJ-TBG-FLPo (Vector Biolabs #1728).
Viral Vectors
3. 2 L water bath set at 48  C.
Expressing FLPo
Recombinase
4. Mouse restrainer.
5. 1 mL syringe.
6. 27G needle.
7. 70% ethanol.

2.4 Mouse Weighing 1. Scale.

and Tamoxifen 2. Plastic container or plastic beaker.
Preparation
3. Tamoxifen.
4. Corn oil.
 Lineage Tracing of Hybrid Periportal Hepatocytes 63

Fig. 3 Second recombination step. (a) Sox9-CreErt2;NZG mice were first treated with 100 mg/kg of tamoxifen
and 10 days later were given 20 mg/kg tamoxifen. Two weeks later, mice were given 109 adeno- or 5  1011
AAV-FLPo viral particles, respectively. After 10 days liver sections were obtained, stained, and imaged. Only
some HybHP were GFP+ (green arrows), while some HybHP were not recombined by FLPo recombinase,
remaining nuclear LacZ+ and GFP (white arrow). Ductal cells were GFP and nuclear LacZ+ (arrowhead).
After 10 days some GFP+ HybHP still had residual nuclear LacZ signal. (b) All GFP+ HybHP were CK19 (green
64 Nicola de Prisco et al.

2.5 Mouse Heart 1. DPBS.

Perfusion 2. Heparin: 10 KU/mL heparin in DPBS (see Note 1).
3. Zinc formalin.
4. Form-Zero neutralizing powder (Fisher Scientific).
5. Tris–HCl pH 9.5/DTT solution: 100 mM Tris pH 9.5,
10 mM DTT solution (Teknova).
6. Syringe pumps (NewEra Pumps #NE300): two pumps are
required, one for DPBS-heparin and the other for zinc
formalin.
7. Three female Luers, two male Luers, plastic tube with the inner
diameter of 1.6 mm, and three-way stopcocks (Biorad, #732-
8103) (see Note 2).
8. Formalin absorbent pads (Leica, large pads).
9. Isoflurane.
10. Scissors, clamp, forceps, and circle tip forceps: for mouse
surgery.
11. Surgery tray.
12. Sucrose: 15 and 30% in PBS.
13. Cryomold: 25 mm  20 mm  5 mm.
14. Optimal cutting temperature (OCT) compound.

2.6 1. Cryostat.
Immunofluorescence, 2. Slide Jar and rack.
Antibodies,
3. Staining tray.
and Staining Reagents
4. Citrate buffer: Dilute 10 citrate buffer pH 6.0 (Scytek
#CPL500) with MilliQ water.
5. Phosphate saline buffer (PBS): 137 mM NaCl, 10 mM
Na2HPO4, 1.8 mM KH2PO4, 2.7 mM KCl, pH 7.3–7.5.
6. Donkey serum.
7. Permeabilization solution: Triton X-100 0.1% diluted in PBS.
8. Blocking solution: Tween 0.1%, donkey serum 2% in PBS.
9. Wash solution: Tween 0.1% in PBS.
10. Mounting reagent: Mowiol or an equivalent reagent.
11. Goat anti-HNF-4α (Santa Cruz, sc-6556), 1:50.

Fig. 3 (continued) arrows). (c) All GFP+ HybHP were HNF4α+ (green arrows). Sox9-CreErt2;NZG were given
tamoxifen followed with AAV-FLPo and treated with 1 or 12 (chronic) CCl4 doses. Liver was excised and
analyzed as above. Single GFP+ HybHP were found after only one CCl4 dose (d), while multicellular GFP+
hepatocytes clones expanding from the portal tract (PT) were observed after 12 CCl4 doses (e). Bracketed and
open scale bars: 20 and 50 μm, respectively
 Lineage Tracing of Hybrid Periportal Hepatocytes 65

12. Goat anti-CK19 (Santa Cruz, sc-33,111), 1:100.

13. Mouse anti-LacZ (Promega, Z3781), 1:50.
14. Chicken anti-eGFP (Abcam, Ab13970), 1:800.
15. 1 mg/mL DAPI stock solution.

3 Methods

Sox9 CreErt2 is not viable in homozygosity. This stock is maintained

by breeding heterozygous males with C57/BL6 females.
NZG colony is maintained by breeding homozygous mice.
For experiments, Sox9-CreErt2 males are bred with homozy-
gous NZG females producing offspring 100% heterozygous for the
NZG reporter and 50% for Sox9-CreErt2. Experiments are per-
formed with heterozygous Sox9-CREErt2;NZG mice.

3.1 Tamoxifen Injection of Sox9-CreErt2;NZG mice at postnatal day 10 intraperi-

Preparation toneally at the dose of 100 mg/kg body weight (see Note 3).
and Administration
1. Under a chemical fume hood weight 20 mg of tamoxifen.
2. Dissolve the tamoxifen in 1 mL of sterile corn oil obtaining a
solution of 20 mg/mL (see Note 4).
3. Using a P1000 pipette, dispense 500 μL of tamoxifen directly
into the barrel of the syringe and carefully and slowly pull the
plunger back. Invert the syringe and expel air bubbles.
4. Weigh the mouse to be injected, and inject intraperitoneally the
corresponding volume of the solution to administrate
100 mg/Kg tamoxifen.
5. Maintain the injected mice on absorbent paper for 72 h post
injection then change to a clean cage (see Note 5).

3.2 Administration Transduction of Sox9-CREErt2;NZG mice with Ad-/AVV-DJ-

of Adenovirus or FLPo after a minimum of 2 weeks from tamoxifen administration
Adeno-Associated to wash out tamoxifen. After a minimum wash-out period of
Virus via Tail Vein 2 weeks from virus administration, the mice are ready to lineage
Injection trace HybHP clonally with the model of liver injury of choice.
1. Before the intravenous injection (i.v.), dilute the virus in PBS.
Use 5  1011 genome copies/mouse for AAV or 1  109
plaque-forming units/mouse of adenovirus (see Note 6).
2. Prepare and set a water bath at 48  C.
3. Load the syringe with 200 μL of the solution containing the
adenovirus or the AAV.
4. Place the mouse in the mouse restrainer.
66 Nicola de Prisco et al.

5. Hold the tail in the warm water for 45 s to 1 min to dilate the
tail vein.
6. Inject in one of the two lateral tail veins.

3.3 Intracardiac Carry out all the procedure under chemical hood.
Perfusion and Staining Day 1 (see Note 7)
1. Prepare a 50 mL syringe with 45 mL of DPBS-heparin, and
load the syringe into the pump.
2. Prepare a 50 mL syringe with 45 mL of zinc formalin and load
in the second syringe pump.
3. Prepare the surgery tray with the formalin absorbent pad.
4. Prepare 50 mL conical tube with 30 mL of zinc formalin.
5. Anesthetize the mouse with isoflurane using anesthesia
chamber.
6. Pin the mouse on the formalin absorbing pad in the surgery
tray and spray the abdominal area with 70% ethanol.
7. Using the circle tip forceps and scissors, cut the skin up from
the stomach along the side around the heart, then carefully cut
the peritoneal membrane.
8. Carefully cut the diaphragm and the rib cage both sides of the
sternum. Clamp the sternum, and roll it to expose the heart.
9. Insert the pump needle into the left ventricle.
10. Once the needle is placed in the left ventricle, cut the right
auricle with scissors, and immediately start pumping DPBS-
heparin solution at 5 mL/min.
11. Pump 15–20 mL DPBS-heparin, and once the liver appears
clear, turn off the pump, and switch the valve to zinc formalin.
12. Perfuse the mouse with 20 mL of zinc formalin (see Note 8),
then stop the pump, and remove the needle.
13. Excise the liver, and put it into a 50 mL conical tube with
30 mL zinc formalin, and gently rock it overnight at 4  C.
14. Repeat the procedure for all mice by washing the tubing with
DPBS for one full flush followed by one full flush of DPBS-
heparin solution.
Day 2
15. Discard the zinc formalin, and rinse twice the liver filling the
50 mL conical with PBS, mix, and discard (see Note 9).
16. Wash with PBS for 15 min at room temperature (RT), rocking
gently and repeat this wash twice.
17. Rinse quickly with MilliQ water.
 Lineage Tracing of Hybrid Periportal Hepatocytes 67

18. Using a 10 cm plate as tray, carefully separate each lobe, cut

lobes in half, and put them into a new 15 mL tube.
19. Add 10 mL of Tris–HCl pH 9.5/DTT solution.
20. Rock gently at RT for 2 h.
21. After 2 h discard the Tris–HCl pH 9.5/DTT solution.
22. Do two quick washes with PBS.
23. Fill the tube with 10 mL of 15% sucrose.
24. Rock gently at 4  C overnight.
Day 3
25. Discard 15% sucrose.
26. Add 10 mL of 30% sucrose, rocking gently at 4  C until the
next day.
Day 4
27. Prepare a box with dry ice, OCT compound and standard size
cryomolds.
28. Pour the 30% sucrose solution from a liver sample, then pour
the liver pieces onto a 10 cm plate.
29. Pour a small amount of OCT compound in two standard size
cryomolds, enough to cover the surface.
30. Stack a Kimwipe on a paper towel, place the liver pieces on the
Kimwipe, and dry off with another Kimwipe on top, gently
pressing to collect all the sucrose solution (see Note 10).
31. Place each half of liver lobes into separated cryomolds, with the
convex side of each piece facing down, and cover completely
with OCT compound.
32. Place the cryomold onto dry ice until the OCT is completely
frozen, making sure the mold is completely flat so the OCT
compound does not spill over.
33. Store samples in 80  C.
34. Cut liver sections of 5 μm using a cryostat.

3.4 Immuno- Day 1

fluorescence
1. Set the water bath to 96  C.
2. Pour the 1 citrate buffer into staining jar. Once the water
bath reaches the temperature, place the staining jar with citrate
buffer inside.
3. Place sample slides in a slide rack, and do three PBS washes of
5 min each using an additional staining jar.
4. Perform two quick washes with MilliQ water, then tap the
water off onto a paper towel, and place the slides into citrate
buffer at 96  C for 1 h.
68 Nicola de Prisco et al.

5. After 1 h, take the staining jar out of the water bath and let cool
at room temperature for 45 min.
6. Do two quick washes with PBS.
7. Prepare 100 mL of 0.1% Triton X-100 in PBS, and incubate the
slides for 20 min in a staining jar.
8. After 20 min, do five quick washes with PBS (see Note 11).
9. Process one slide a time, drawing a circle around the tissue
section with Dako pen and then incubate with 150–200 μL
of blocking buffer, carefully avoiding the tissue to dry for
30 min in a staining tray.
10. Prepare appropriate antibody solution by diluting antibodies in
blocking buffer (see Note 12).
11. After 30 min, pour the blocking buffer, and carefully add the
antibodies to each slide sample.
12. Incubate at 4  C overnight.
Day 2
13. Wash slides three times for 5 min each with wash solution using
a staining jar.
14. Incubate each slide with the corresponding secondary antibody
diluted in blocking buffer for 2 h at RT.
15. Wash three times for 5 min with wash solution using a
staining jar.
16. Wash two times for 5 min with PBS.
Optional if reduction of liver autofluorescence is desired; oth-
erwise follow directly to step 20:
17. Wash the slides further with two quick washes with MilliQ
water and one quick wash with 70% ethanol.
18. Incubate slides with 0.1% Sudan Black in 70% ethanol for
20–30 min.
19. After this incubation, thoroughly wash slides with PBS 0.02%
Tween, and using a squeeze bottle wash applying directly the
jet to the tissue being careful not to wipe it out.
20. Incubate each slide with DAPI diluted 1:1000 in PBS (1:100 in
PBS-0.02% Tween if Sudan Black is used) for 10 min for
nuclear staining.
21. Wash quickly the slides using a staining jar with MilliQ water. If
Sudan Black has been used, wash the slides three times for
5 min with PBS 0.02% Tween, followed by two quick washes
with MilliQ water.
22. Tap each slide dry on a paper towel, and then tip the slides up in
a drawer to finish drying and be protected from the light.
23. Once dried, mount slides using 10 μL of Mowiol.
 Lineage Tracing of Hybrid Periportal Hepatocytes 69

4 Notes

1. Add 10 mL of nuclease-free water in 100KU heparin bottle

(Sigma), mix thoroughly, transfer into a 15 mL conical tube,
and store at 4  C. Dilute heparin (1:1000) in DPBS to obtain
the working DPBS-heparin solution.
2. Connect the two syringe pumps (one with DPBS-heparin solu-
tion and the other with zinc formalin) to 1.6 mm plastic tube
using female luers. Connect the two tubes with the three-way
stopcock using the male luers, and finally connect the three-
way stopcock with a 1.6 mm tube with a 26½ needle.
3. Mice can be injected at any age, but in older mice, recombina-
tion becomes less efficient. Consecutive tamoxifen injection to
maximize the number of Sox9-expressing cells that recombine
can be successfully performed.
4. To dissolve properly the tamoxifen, prewarm a thermomixer at
55  C, and put the tube containing corn oil and tamoxifen for
30 min with mixing (speed 800–900 rpm). Remaining crystals
of tamoxifen should dissolve in 30 min.
5. Dispose the bedding into a yellow Chemotherapy Waste bag.
6. Even though we have successfully labeled HybHP with both
adenovirus and AAV, we have observed that adenovirus is
associated with a clear immune reaction, not detected using
AAV, that could be potentially affecting downstream
experiments.
7. All volumes per one mouse.
8. If the perfusion is working properly, the mouse tail will get stiff,
and at the end of the perfusion, the liver should be stiff to the
touch as well.
9. Collect the zinc formalin into a waste bottle, and neutralize it
by adding Form-Zero formaldehyde powder neutralizer. Add
the entire content of the bottle (500 cc) for each 4 L of
formalin waste, mix the solution, and let it stand for
20–25 min to allow the reaction to complete. Dispose the
neutralized formalin into the sanitary sewer.
10. Residual sucrose from the tissue must be removed before
embedding. Sucrose will prevent OCT to contact the tissue
affecting the quality of the histological sectioning.
11. At this point, the slides can be stored in PBS at 4  C for
few days.
12. Prepare the proper antibody dilution in a final minimum vol-
ume of 50 μL of blocking buffer per slide.
70 Nicola de Prisco et al.

Acknowledgments

This work has been supported by CIRM Training Grant II

(TG2-01154) and NIH K99/R00CA191152 grant.

References

1. Rodrigo-Torres D, Affo S, Coll M, Morales- regeneration. J Clin Invest 121

Ibanez O, Millan C, Blaya D, Alvarez-Guaita A, (12):4850–4860. https://doi.org/10.1172/
Rentero C, Lozano JJ, Maestro MA, Solar M, JCI59261
Arroyo V, Caballeria J, van Grunsven LA, 7. Raven A, Lu WY, Man TY, Ferreira-Gonzalez S,
Enrich C, Gines P, Bataller R, Sancho-Bru P O’Duibhir E, Dwyer BJ, Thomson JP, Meehan
(2014) The biliary epithelium gives rise to liver RR, Bogorad R, Koteliansky V, Kotelevtsev Y,
progenitor cells. Hepatology 60 Ffrench-Constant C, Boulter L, Forbes SJ
(4):1367–1377. https://doi.org/10.1002/ (2017) Cholangiocytes act as facultative liver
hep.27078 stem cells during impaired hepatocyte regener-
2. Schaub JR, Malato Y, Gormond C, Willenbr- ation. Nature 547(7663):350–354. https://
ing H (2014) Evidence against a stem cell ori- doi.org/10.1038/nature23015
gin of new hepatocytes in a common mouse 8. Font-Burgada J, Shalapour S, Ramaswamy S,
model of chronic liver injury. Cell Rep 8 Hsueh B, Rossell D, Umemura A, Taniguchi K,
(4):933–939. https://doi.org/10.1016/j.cel Nakagawa H, Valasek MA, Ye L, Kopp JL,
rep.2014.07.003 Sander M, Carter H, Deisseroth K, Verma
3. Tarlow BD, Finegold MJ, Grompe M (2014) IM, Karin M (2015) Hybrid periportal hepa-
Clonal tracing of Sox9+ liver progenitors in tocytes regenerate the injured liver without
mouse oval cell injury. Hepatology 60 giving rise to cancer. Cell 162(4):766–779.
(1):278–289. https://doi.org/10.1002/hep. https://doi.org/10.1016/j.cell.2015.07.026
27084 9. Kopp JL, Dubois CL, Schaffer AE, Hao E, Shih
4. Yanger K, Knigin D, Zong Y, Maggs L, Gu G, HP, Seymour PA, Ma J, Sander M (2011) Sox9
Akiyama H, Pikarsky E, Stanger BZ (2014) + ductal cells are multipotent progenitors
Adult hepatocytes are generated by self- throughout development but do not produce
duplication rather than stem cell differentia- new endocrine cells in the normal or injured
tion. Cell Stem Cell 15(3):340–349. https:// adult pancreas. Development 138
doi.org/10.1016/j.stem.2014.06.003 (4):653–665. https://doi.org/10.1242/dev.
5. Espanol-Suner R, Lemaigre FP, Leclercq IA 056499
(2013) Reply: to PMID 22922013. Gastroen- 10. Yamamoto M, Shook NA, Kanisicak O,
terology 145(1):255–256. https://doi.org/ Yamamoto S, Wosczyna MN, Camp JR, Gold-
10.1053/j.gastro.2013.05.037 hamer DJ (2009) A multifunctional reporter
6. Malato Y, Naqvi S, Schurmann N, Ng R, mouse line for Cre- and FLP-dependent line-
Wang B, Zape J, Kay MA, Grimm D, Willenbr- age analysis. Genesis 47(2):107–114. https://
ing H (2011) Fate tracing of mature hepato- doi.org/10.1002/dvg.20474
cytes in mouse liver homeostasis and
 Chapter 7

Analysis for the Heterogeneity of Liver Progenitor Cells

Kenji Kamimoto

Abstract
Recent technological advances have revealed the heterogeneity of cells and tissues. Existence of heteroge-
neity in hepatic progenitor cells is becoming apparent by various experimental approaches, and here we
describe a series of techniques to investigate the proliferative heterogeneity of these cells. We have
developed a new technique by combining genetic lineage tracking and three-dimensional imaging methods.
The data obtained can be used in statistical analysis to quantitatively investigate the mechanisms underlying
the heterogeneity of hepatic progenitor cells.

Key words Heterogeneity, Cell proliferation, Three-dimensional imaging, Microscopy, Genealogy

tracking, Quantitative analysis

1 Introduction

Recent technological advances in dissecting biological events at the

single-cell level have revealed the existence of heterogeneity in
biological aspects such as gene expression, cellular function, and
proliferative capacity [1, 2].
Hepatic progenitor cells, which are characterized to have bipo-
tential differentiation and high proliferative capacities [3], have also
been hypothesized to be heterogeneous [4]. Its proliferative het-
erogeneity has been reported in many studies using in vitro colony
formation assays [5–7]. Proliferative capacity is examined by mea-
suring the size of a colony derived from a single hepatic progenitor
cell [8]. Cell culture is initiated at a very low density so that their
progenies appear as colonies, and this feature allows retrospective
identification of the original proliferative and differentiation capa-
cities of each progenitor cell. Therefore, colony formation assays
provide useful information on hepatic progenitor cells. However,
colony behavior revealed by this method does not necessarily reflect
its in vivo physiological capacity, since the assay conditions may not
recapitulate the situation around progenitor cells within the tissue
of origin. Here we describe a series of methods for quantitatively

Naoki Tanimizu (ed.), Hepatic Stem Cells: Methods and Protocols, Methods in Molecular Biology, vol. 1905,
https://doi.org/10.1007/978-1-4939-8961-4_7, © Springer Science+Business Media, LLC, part of Springer Nature 2019

71
72 Kenji Kamimoto

Fig. 1 Schematic illustration of the proposed method

evaluating the proliferative capacity of hepatic progenitor cells

in situ.
In vivo investigation of proliferative capacity of hepatic progen-
itor cells is achieved by combining clonal genetic fate tracing with
three-dimensional (3D) microscopy [9] (Fig. 1). Hepatic progeni-
tor cells can be labeled in vivo and traced using a genetic lineage
tracing system [10]. In our proposed method, we utilize Cre-loxP
system, which is a frequently used genetic lineage tracing system in
mouse studies. Thus, cell labeling can be performed at the desired
frequency. We do not need to isolate progenitor cells for the
analysis of cell proliferation because we can label and trace target
cells in vivo. Besides, inducible Cre-loxP system can be tunable.
Thus we can induce cell labeling at the desired frequency. Impor-
tantly, the fate of single cells can be traced by labeling at a frequency
low enough to distinguish each labeled cell. This sparse permanent
label makes it possible to track the fate of individual hepatic pro-
genitor cells under physiological conditions.
Tracing sparsely labeled cells has been applied for the study of
epidermal tissue [11–13], which shows a structure composed of
simple stratified layers. In such a structure, the colony derived from
a single stem cell spreads in 2D, and therefore, its size can be
measured easily. In contrast, the niche of hepatic progenitor cells
is the intrahepatic biliary system, which is the 3D tissue structure
within the liver tissue (Fig. 1). In order to apply the in vivo quanti-
tative clonal cell tracing system to the investigation of hepatic
progenitor cells, we established the high-resolution 3-D imaging
technique visualizing the intrahepatic biliary system at single-cell
level [14].
 Analysis for the Heterogeneity of Liver Progenitor Cells 73

Data is acquired from numerous colonies, each derived from a

single hepatic progenitor cell. By analyzing the data, the mechan-
isms underlying proliferative heterogeneity can be investigated [9].

2 Materials

2.1 Mouse 1. Mouse strain: Prom1-CreERT2. This mouse strain has the
Experiment CreERT2 gene at the Prom1 locus [15] and may be purchased
from the Jackson Laboratory (Stock No: 017743).
2. Mouse strain: R26R-tdTomato. This mouse strain has a loxP-
flanked transcriptional stop fragment inserted upstream of the
red fluorescent protein (tdTomato) gene [16]. The strain is
used in combination with the Cre-expressing mouse strain for
lineage tracing experiments and may be purchased from the
Jackson Laboratory (Stock No: 007909).
3. Mouse strain: Prom1-CreERT2;R26R-tdTomato. Cross
Prom1-CreERT2 with R26R-tdTomato mouse strains to
obtain a Prom1-CreERT2;R26R-tdTomato mouse (see Note
1).
4. Tamoxifen solution: Dissolve 10 mg of tamoxifen in 10 mL
corn oil to obtain a 1 mg/mL solution. It takes several hours to
completely dissolve tamoxifen powder in corn oil. Place the
mixture on an agitator at 37
C, and incubate it until the
powder is dissolved.
5. Gavage needle.
6. Thioacetamide (TAA) solution: Dissolve TAA in water at
300 mg/L. Prepare this solution just before administration
(see Note 2).
7. 26G syringe needle.
8. 10 mL syringe.
9. CO2 cylinder.

2.2 Tissue 1. 4% paraformaldehyde (PFA) in phosphate-buffered saline

Processing (PBS): Dissolve PFA in PBS and store at 4
C.
2. 20% sucrose in PBS: Dissolve sucrose in PBS and store at 4
C.
3. OCT compound.
4. Liquid nitrogen.
5. Cryostat.
6. 5 mL tube.
7. Permeabilization buffer: 3% bovine serum, 0.02% sodium
azide, and 0.2% Triton X-100 in PBS. Store at 4
C.
74 Kenji Kamimoto

8. Nucleus staining solution: Dilute 5 mM SYTOX Green nucleic

acid stain in permeabilization buffer at a 1:100000 ratio (see
Note 3). Prepare the solution just before use.
9. SeeDB solution [14]: 80.2% D-fructose (wt/wt) and 0.5%
1-thioglycerol in water. Prepare just before use (see Note 4).
10. Sample storage solution: 3% bovine serum, 0.02% sodium
azide, and 20% sucrose in PBS.
11. Confocal microscopy: Use a laser confocal microscope
equipped with silicone or oil immersion lens (e.g., microscope,
FV1200; and silicone immersion lens, UPLSAPO30XS) (see
Note 5).

3 Methods

3.1 Mouse 1. Obtain Prom1-CreERT2;R26R-tdTomato mice. Six- to eight

Experiment week-old animals are desirable (see Note 6).
2. Administer 250 μL of tamoxifen solution per 20 g body weight
to the mouse by oral gavage (see Notes 7 and 8) with the
1.0 mL syringe and gavage needle.
3. Start liver injury 2 weeks after tamoxifen administration. For a
TAA liver injury model, use the TAA solution, which can be
administrated as drinking water (see Note 9).
4. Change TAA solution every week to keep it fresh.

3.2 Tissue Fixation The mouse liver should be harvested several weeks after TAA
administration. We recommend analyzing the liver after
1–8 weeks. The experimental period should be selected according
to the experimental design. Detailed information can be inferred
from analyzing a range of injury periods rather than a fixed time
point.
1. Place mouse in chamber for CO2 euthanasia.
2. Flow high concentrations of CO2 until vital signs are lost.
3. Immediately following euthanasia, set the mouse for surgery to
start PFA perfusion (see Note 10).
4. Remove the skin and muscle layers of the abdomen using
surgical scissors to expose the liver.
5. Cut the vena cava to let the blood flow out.
6. Insert the syringe needle into the portal vein, and inject 10 mL
of PBS gently until blood flows out from the liver. Next, inject
10 mL of 2% PFA and then 10 mL of 4% PFA. Make sure the
liver color changes into light brown and the liver becomes hard
(see Note 11).
7. Remove the liver from the mouse body.
 Analysis for the Heterogeneity of Liver Progenitor Cells 75

8. Remove the gallbladder and the extrahepatic bile duct using

surgical scissors.
9. Separate the hepatic lobules using a surgical scissor or knife.

3.3 Tissue 1. Cut the liver into ~1 cm thick slices with a razor blade.
Processing 2. Place the slices in a 5 mL tube filled with 4% PFA, and incubate
at 4
C overnight. Remove the PFA and wash the liver twice
with PBS. Incubate the liver slices in 20% sucrose at 4
C
overnight.
3. Embed the fixed liver slices in the OCT compound. Freeze the
sample with liquid nitrogen. The frozen sample can be stored at
80
C.
4. Place the frozen liver into the sample holder of the cryostat.
5. Slowly warm the frozen liver to melt the surface of the sample
by touching with the hand. Immediately after thawing the
surface of the sample, cut the frozen liver into 200–300 μm
thick sections using the cryostat (see Note 12).
6. Using forceps, place the samples into PBS in a 50 mL tube. At
least 30 liver slices should be obtained from a frozen liver block
with 1 cm in thickness (see Note 13).
7. Wash the sliced samples with PBS three times.
8. Wrap the tube in aluminum foil sheet, or incubate in the dark
to avoid photobleaching of fluorescent dye and the fluorescent
protein.
9. Place the liver samples into a 5 mL tube filled with 2 mL of
permeabilization buffer. Incubate the samples for 30 min on a
rocking device (see Note 14).
10. Aspirate the permeabilization buffer from the tube, and add
2 mL of nucleus staining solution. Incubate the tube overnight
on a rocking device (see Note 15).
11. Wash samples with PBS three times.
12. Remove the PBS completely. Mix samples with 2 mL of SeeDB
in a new 5 mL tube. Incubate the samples overnight on a
rocking device (see Notes 16 and 17).

3.4 3D Imaging 1. Place a SeeDB-processed sample on a 35  60 mm cover glass.

2. Place another cover glass over the sample to hold the liver
sample between the two cover glasses. It is important to place
the cover glass and the sample in parallel. Avoid making air
bubbles between the liver sample and the cover glass (Fig. 2).
3. Attach the slide to the observation stage of the microscope.
Use the sample holding arm to attach the slide to the stage. Use
a sticky tape to attach the slide to the stage if a sample holding
arm is unavailable.
76 Kenji Kamimoto

Fig. 2 Tissue processing workflow

4. It is highly recommended to use silicone or oil immersion lens

to acquire images with high resolution (see Note 18). Place the
oil between the lens and the sample slide. Make sure that the
sample does not move when the stage moves.
5. Operate the confocal microscope according to the manufac-
turer’s instructions. First, observe the fluorescent signal of
labeled cells. The signal from a colony derived from a single
hepatic progenitor is sparsely located (see Note 19). Set the
acquisition range of the XY axis and then the acquisition range
of the Z-axis. Set the step size of the Z-axis to the optimal
minimum size of the objective lens. Set parameters for acquir-
ing images (e.g., acquisition channels, sensitivity of the detec-
tors, scanning speed, etc.). Start acquiring a 3D image of
tdTomato and nuclei.
6. Repeat step 5 to acquire the images of all colonies in a sample.
7. Change the sample, and acquire 3D images until enough data
(>500 colonies) are collected.

3.5 Quantification The 3D image is quantified using image processing software

(Fig. 3). Here, we use Volocity (PerkinElmer) or ImageJ (Fiji) to
quantify the images (see Note 20).

3.5.1 Quantification 1. Load images using the software. All metadata will be imported
by Volocity automatically along with image data.
2. In Measurement mode, detect nuclei using the Find Object
module. The object identified by this step includes all nuclei in
the liver sample (Fig. 3a).
3. Detect labeled cells by applying the Find Object module to the
tdTomato signal. tdTomato is uniformly localized throughout
the cytoplasm, showing whole cells of the labeled colony
(Fig. 3b).
4. Add the Filter population module, and set minimum object
size to remove small objects resulting from noise.
5. Add the Separate touching object module into nuclei objects to
separate incorrectly connected nuclei. This process is necessary
because, in the first step of object identification, two different
nuclei might be mistakenly identified as a single object.
6. Use the Intersect module with nuclei and cell objects. Nuclei of
the labeled colony will be detected (Fig. 3c) (see Note 21).
 Analysis for the Heterogeneity of Liver Progenitor Cells 77

Fig. 3 Quantification process of three-dimensional (3D) images. A 3D image is projected on the 2D plane. (a)
Nuclei. (b) Labeled colony visualized with tdTomato. (c) Identified nuclei of tdTomato-positive cells

7. After following this measurement protocol, you can apply the

analytic protocol to many images all at once to quantify a large
body of data.
8. Quantified data (Fig. 4) can be used in the statistical analysis
(see Note 22).

3.5.2 Quantification by Instead of Volocity, we can use ImageJ/Fiji to quantify the images.
ImageJ (Fiji) An ImageJ plug-in, MorphoLibJ, is required here. Please install
Fiji (https://fiji.sc) [17] and MorphoLibJ (https://github.com/
ijpb/MorphoLibJ/releases) [18] following manufacturer’s
instructions.
1. Load 3D image of nuclei and tdTomato to ImageJ. Serial tiff
images can be imported directly (“File” ! “Import” ! “Image
Sequence. . .”). Alternatively, 3D image files in another format
can be loaded via plug-ins. For example, Olympus image for-
mat can be loaded by “Olympus ImageJ Plugin (https://
imagej.net/OlympusImageJPlugin).”
2. Apply the Gaussian filter to both nuclei channel and tdTomato
channels (“Process” ! “Filters” ! “Gaussian Blur”).
3. Create binary images by setting the fluorescent intensity
threshold (“Image” ! “Adjust” ! “Threshold” (method,
default; dark background, on)). Please adjust threshold value
to properly detect the objects. The optimal threshold value may
change depending on various factors.
4. Merge nuclei and tdTomato channel (“Process” ! “Image
Calculator” (operation, AND)). Set the binary images of nuclei
and tdTomato as input images.
5. Separate touching nuclei by the watershed method (“Plu-
gins” ! “MorphoLibJ” ! “Binary Images” ! “Distance
Transform Watershed 3D” (distances, quasi-Euclidean;
78 Kenji Kamimoto

Fig. 4 Colony size distribution. Liver was analyzed after 0, 2, 4, 6, and 8 weeks of thioacetamide (TAA) injury.
Data are shown as swarm plot and boxplot. Colony size distribution data were reproduced from our previous
study [9]

normalize weights, off; dynamic, 3.00; connectivity, 26)). Seg-

mented 3D images will be produced by this process.
6. In the segmented 3D image, each nucleus has unique label, but
it has not been visualized yet. Visualize these labels by assigning
a unique color to each label (“Plugins” ! “Morpho-
LibJ” ! “Label Images” ! “Labels to RGB”).
7. Visualize result in 3D view (“Plugins” ! “3D viewer”). Make
sure nuclei of tdTomato-positive cells were appropriately iden-
tified. If nuclei identification and segmentation failed, go back
to the segmentation process (step 5). The extent of watershed
segmentation can be optimized by the parameter of
“Dynamic.”
8. Output quantified values. First, select the segmented 3D image
(produced by step 5), and then use the analysis function of
MorphoLibJ
 Analysis for the Heterogeneity of Liver Progenitor Cells 79

(“Plugins” ! “MorphoLibJ” ! “Analyze” ! “Particle Anal-

ysis 3D”). This procedure outputs quantified data table. The
data still contain small non-nucleus objects made by noise.
These noise objects can be removed by thresholding object
volume.

4 Notes

1. The method utilizes an inducible Cre-loxP system to label and

trace hepatic progenitor cells under in vivo conditions.
Although we used the Prom1-CreERT2;R26R-tdTomato
mouse strain, many additional mouse strains are available for
lineage tracing of hepatic progenitor cells [10]. A different
combination of Cre driver and fluorescent reporter strains can
be used rather than the strains described here. Labeling effi-
ciency might change depending on both the Cre driver and
reporter strains. In addition, optimization of tamoxifen dosage
is needed if a mouse strain different from Prom1-CreERT2;
R26R-tdTomato is used.
2. TAA solution slowly degrades even when stored at 4
C. Alter-
natively, a TAA stock solution (300 mg/30 mL) may be
prepared and stored at 20
C in 30 mL aliquots. Dilute one
aliquot of the TAA stock solution (30 mL) in 1 L of water
immediately before use.
3. SYTOX Green nucleic acid stain is used to visualize nuclei of
liver tissue. The reagent emits a strong green fluorescent signal
when bond to DNA. It can be used at very low concentrations,
reducing experimental costs. Hoechst33342, another fluores-
cent reagent, can also be used for nucleus staining, but the
staining protocol should be optimized for each reagent.
4. SeeDB is a tissue-clearing reagent containing D-fructose at a
very high concentration [14]. D-fructose in SeeDB will easily
precipitate if stored at 4
C.
5. Image resolution and depth are highly dependent on the
microscope and lens. The Olympus confocal microscope is
used in this protocol, but another microscope may be used
instead. Multi-photon microscopes can also be used.
6. The mouse age should be selected according to the experimen-
tal design. When examining hepatic progenitor cells at the
developmental stage, 3–4-week-old mice can be used.
7. Tamoxifen dosage is critical for sparse cell labeling. At this
dosage, approximately 0.2% of Prom1-positive hepatic progen-
itor cells will be labeled [9]. The amount of tamoxifen should
be optimized if another mouse strain is used.
80 Kenji Kamimoto

8. Intraperitoneal injection can be used instead of oral gavage

with the same amount of tamoxifen.
9. In this protocol, a chronic liver injury induced by TAA was used
to induce hepatic progenitor cell proliferation. There are also
many other liver injury models that activate hepatic progenitors
[3]. The 3,5-diethoxycarbonyl-1,4-dihydrocollidine diet, car-
bon tetrachloride, choline-deficient, ethionine-supplemented
dietary, and the genetic liver injury models may be used
[19]. However, hepatic progenitor cells exhibit various pheno-
types depending on the type of injury, and the duration of liver
injury varies depending on the model [4]. Therefore, the pro-
tocol should be optimized for each model.
10. Robust tissue fixation is required to keep tissue structure for
later processing protocols. The entire tissue processing step
takes at least 2 days. Liver tissue will be damaged by the process
if fixation is incomplete. Fixation with 4% PFA is therefore
recommended. Hydrophobic reagents such as paraffin, metha-
nol, or acetone should not be used for fixation because they
reduce the fluorescent signal of tdTomato.
11. Perfusion is strongly recommended for fixation for two rea-
sons. First, perfusion is needed to drain blood from the liver
tissue. The liver contains a large amount of blood, which
scatters and absorbs the fluorescence signal. Any remaining
blood deteriorates image quality. Second, perfusion is needed
to inject PFA into the deeper areas of liver tissue. The liver has a
very thick 3D structure, and perfusion is required to efficiently
spread PFA throughout the organ. The color and hardness of
the liver should be monitored when performing a perfusion.
12. It is important to control the temperature for a smooth cut of
the samples. If temperature is too low, the liver becomes fragile
and cracks easily.
13. Many liver sections are prepared by repeating this step to
acquire images of labeled colonies enough for statistical analy-
sis. Because of the sparse labeling, the small number of labeled
hepatic progenitor cell colonies present in each section. Pre-
pare many liver slices at a time. Sliced liver samples can be
stored in the sample storage solution for up to 3 weeks. The
experiment may be interrupted temporarily at this step.
14. If there are more than 100 samples, scale up the experimental
design, and use 15 mL or 50 mL tubes instead of a 5 mL tube.
15. The liver section may break into small pieces if fixation is
incomplete. Insufficiency of PFA fixation may result from
incomplete perfusion of 4% PFA.
16. Do not incubate SeeDB at a temperature lower than 25
C.
Samples in SeeDB should be incubated at 37
C before
 Analysis for the Heterogeneity of Liver Progenitor Cells 81

microscopic examination. D-fructose in SeeDB precipitates

easily, which deteriorates image quality. The sample is stable
for several days, but the fluorescent protein signal slowly decays
in SeeDB.
17. Make sure that the samples became transparent after processing
with SeeDB. Another hydrophilic tissue-clearing reagent such
as ScaleA2 [20] can be used to render the liver sample clear. In
that case, however, it is necessary to optimize protocol details
such as incubation time. A hydrophobic tissue-clearing reagent
should not be used as it easily denatures fluorescent proteins.
The clearing process is very critical for 3D imaging in the
proposed method.
18. Use silicone or oil immersion lens as silicone oil reduces unde-
sired refraction of excitation light. Random refraction should
be minimized for deep imaging. The best magnification for the
objective lens is around 30.
19. If a slide has too many colonies, the sparse labeling failed. The
density of labeled hepatic progenitor cell colonies should be
less than ten per liver slice. If not, the tamoxifen concentration
should be lowered.
20. Many image processing software/plug-ins are available.
Another software can be used instead of the protocol
described here.
21. Hepatic progenitor cells have a high nucleus/cytoplasmic ratio
and are located very close to each other in the colony. Image
processing software sometimes fails to distinguish the individ-
ual nuclei. Therefore, it is recommended to verify the quantifi-
cation results manually. Failure to obtain image quantification
even after parameter optimization may result from insufficient
imaging resolution or an insufficiently low signal/noise ratio.
22. You can infer much hidden information by analyzing quanti-
fied data using statistical methods [9, 11–13]. Although the
count data itself may not be easily interpretable, there must be a
biological mechanism behind the data. By analyzing the distri-
bution of occurrence probability of a phenomenon, it is possi-
ble to estimate what kind of mechanism exists. The mechanism
behind the distribution of count data is estimated by fitting the
count data to various kinds of the probability distribution for
the best fit.
In the case where the known probability distribution cannot
explain it, a new theoretical model is needed. The quantitative
data acquisition method shown here brings further useful
information in combination with such mathematical modeling
and probability distribution.
82 Kenji Kamimoto

References
1. Donati G, Watt FM (2015) Stem cell hetero- 11. Clayton E, Doupé DP, Klein AM et al (2007) A
geneity and plasticity in epithelia. Cell Stem single type of progenitor cell maintains normal
Cell 16:465–476. https://doi.org/10.1016/ epidermis. Nature 446:185–189. https://doi.
j.stem.2015.04.014 org/10.1038/nature05574
2. Etzrodt M, Endele M, Schroeder T (2014) 12. Doupé DP, Klein AM, Simons BD, Jones PH
Quantitative single-cell approaches to stem (2010) The ordered architecture of murine ear
cell research. Cell Stem Cell 15:546–558. epidermis is maintained by progenitor cells
https://doi.org/10.1016/j.stem.2014.10. with random fate. Dev Cell 18:317–323.
015 https://doi.org/10.1016/j.devcel.2009.12.
3. Miyajima A, Tanaka M, Itoh T (2014) Stem/ 016
progenitor cells in liver development, homeo- 13. Driessens G, Beck B, Caauwe A et al (2012)
stasis, regeneration, and reprogramming. Cell Defining the mode of tumour growth by clonal
Stem Cell 14:561–574. https://doi.org/10. analysis. Nature 488:527–530. https://doi.
1016/j.stem.2014.04.010 org/10.1038/nature11344
4. Jelnes P, Santoni-Rugiu E, Rasmussen M et al 14. Ke M-T, Fujimoto S, Imai T (2013) SeeDB: a
(2007) Remarkable heterogeneity displayed by simple and morphology-preserving optical
oval cells in rat and mouse models of stem cell- clearing agent for neuronal circuit reconstruc-
mediated liver regeneration. Hepatology tion. Nat Neurosci 16:1154–1161. https://
45:1462–1470. https://doi.org/10.1002/ doi.org/10.1038/nn.3447
hep.21569 15. Zhu L, Gibson P, Currle DS et al (2009) Pro-
5. Suzuki A, Sekiya S, Onishi M et al (2008) Flow minin 1 marks intestinal stem cells that are
cytometric isolation and clonal identification of susceptible to neoplastic transformation.
self-renewing bipotent hepatic progenitor cells Nature 457:603–607. https://doi.org/10.
in adult mouse liver. Hepatology 1038/nature07589
48:1964–1978. https://doi.org/10.1002/ 16. Madisen L, Zwingman TA, Sunkin SM et al
hep.22558 (2010) A robust and high-throughput Cre
6. Okabe M, Tsukahara Y, Tanaka M et al (2009) reporting and characterization system for the
Potential hepatic stem cells reside in EpCAM+ whole mouse brain. Nat Neurosci
cells of normal and injured mouse liver. Devel- 13:133–140. https://doi.org/10.1038/nn.
opment 136:1951–1960. https://doi.org/10. 2467
1242/dev.031369 17. Schindelin J, Arganda-Carreras I, Frise E et al
7. Dorrell C, Erker L, Schug J et al (2011) Pro- (2012) Fiji: an open-source platform for
spective isolation of a bipotential clonogenic biological-image analysis. Nat Publ Group
liver progenitor cell in adult mice. Genes Dev 9:676–682. https://doi.org/10.1038/
25:1193–1203. https://doi.org/10.1101/ nmeth.2019
gad.2029411 18. Legland D, Arganda-Carreras I, Andrey P
8. Ema H, Morita Y, Yamazaki S et al (2006) (2016) MorphoLibJ: integrated library and
Adult mouse hematopoietic stem cells: purifi- plugins for mathematical morphology with
cation and single-cell assays. Nat Protoc ImageJ. Bioinformatics 32:3532–3534.
1:2979–2987. https://doi.org/10.1038/ https://doi.org/10.1093/bioinformatics/
nprot.2006.447 btw413
9. Kamimoto K, Kaneko K, Kok C et al (2016) 19. Forbes SJ, Newsome PN (2016) Liver regen-
Heterogeneity and stochastic growth regula- eration - mechanisms and models to clinical
tion of biliary epithelial cells dictate dynamic application. Nat Rev Gastroenterol Hepatol
epithelial tissue remodeling. elife 5:e15034. 13:473–485. https://doi.org/10.1038/
https://doi.org/10.7554/eLife.15034.001 nrgastro.2016.97
10. Lemaigre FP (2015) Determining the fate of 20. Hama H, Kurokawa H, Kawano H et al (2011)
hepatic cells by lineage tracing: Facts and pit- Scale: a chemical approach for fluorescence
falls. Hepatology 61:2100–2103. https://doi. imaging and reconstruction of transparent
org/10.1002/hep.27659 mouse brain. Nat Neurosci 14:1481–1488.
https://doi.org/10.1038/nn.2928
 Chapter 8

Chemical Screening Using a Zebrafish Model for Liver

Progenitor Cell-Driven Liver Regeneration
Sungjin Ko and Donghun Shin

Abstract
Following massive hepatocyte ablation in zebrafish, biliary epithelial cells can extensively give rise to
hepatocytes through liver progenitor cells (LPCs). The zebrafish liver injury model is an important system
to elucidate the molecular mechanisms underlying LPC-driven liver regeneration. Here, we describe a
chemical screening method using the zebrafish model for identifying small molecules that can modulate
LPC-driven liver regeneration.

Key words Nitroreductase, Biliary epithelial cells, Hepatocyte ablation, Chemical screen, Liver pro-
genitor cells

1 Introduction

The liver is a highly regenerative organ: it can recover even after

losing 70% of its mass [1]. However, after prolonged or severe liver
injury, as observed in the advanced stages of chronic and acute liver
diseases, the liver loses its ability to regenerate through self-replica-
tion of hepatocytes [2, 3]. This continuing damage in combination
with defective hepatocyte proliferation stimulates excessive inflam-
mation and subsequent wound healing mechanisms, including
fibrotic/cirrhotic responses [1–3]. Currently, liver transplantation
is the only reliable cure to prolong the life of patients with advanced
liver diseases; however, the number of liver donors is not sufficient
to match the increasing number of patients [4]. To alleviate this
bottleneck and to extend patient lives until transplantation, diverse
approaches have been attempted to develop therapeutic options
[5–7]. Although there have been technical improvements, further
progress is still needed for these therapeutics to be considered as
reliable options worthy of testing in clinical trials.
One attractive approach that has recently garnered attention is
to promote the differentiation of activated liver progenitor cells
(LPCs) in diseased livers into either hepatocytes or biliary epithelial

Naoki Tanimizu (ed.), Hepatic Stem Cells: Methods and Protocols, Methods in Molecular Biology, vol. 1905,
https://doi.org/10.1007/978-1-4939-8961-4_8, © Springer Science+Business Media, LLC, part of Springer Nature 2019

83
84 Sungjin Ko and Donghun Shin

cells (BECs) [3, 8]. Given the prevalence of LPCs in diseased

human livers [2, 9], this approach becomes more promising due
to recent findings in mice that activated BEC-derived LPCs are a
key source of new hepatocytes in various liver injury settings with
defective hepatocyte proliferation [9, 10]. While these achieve-
ments are encouraging to elucidate the molecular mechanisms
underlying LPC-driven liver regeneration and to identify drug
targets that can augment this process, the mouse is not an ideal
model animal for large-scale chemical screening due to its high cost
and time constraint.
Given the ease of chemical screening in zebrafish embryos and
larvae [11], we have established a zebrafish liver injury model for
LPC-driven liver regeneration by generating a transgenic line, Tg
(fabp10a:CFP-NTR), that provides a genetic means to ablate dif-
ferentiated hepatocytes [12]. The transgenic fish expresses bacterial
nitroreductase (NTR) enzymes fused with cyan fluorescent proteins
(CFP) specifically in hepatocytes due to the hepatocyte-specific
fabp10a promoter. Since NTR converts the non-toxic prodrug
metronidazole (Mtz) into a cytotoxic drug, treating the fish with
Mtz ablates only NTR-expressing hepatocytes. Following massive
hepatocyte ablation, BECs dedifferentiate into LPCs and subse-
quently differentiate into hepatocytes [12]. Compared to the afore-
mentioned mouse model, this zebrafish model is advantageous for
chemical screening to identify small molecules that affect
LPC-driven liver regeneration [13]: (1) the entire liver regenera-
tion assay takes only 6 days, (2) the regeneration process is homo-
geneous with very small variability among animals, and
(3) hundreds of animals can be manipulated simultaneously. The
zebrafish assay allows for screening ~30 compounds per week. Here
we describe a detailed protocol for chemical screening using the
zebrafish hepatocyte ablation model.

2 Materials

1. Tg(fabp10a:CFP-NTR);Tg(fabp10a:DsRed) adult fish.

2. Mating tanks: large aquaculture tanks with dividers.
3. Fine plastic mesh strainer (tea strainer).
4. Squeeze bottle.
5. Petri dishes: 100
15 mm and 60
15 mm.
6. Egg water: 0.3 g of sea salts “Instant Ocean” in 1 L of distilled
water supplemented with 0.5-mM CaSO4.
7. Disposable polyethylene transfer pipettes.
8. Stereomicroscope.
 Liver Progenitor Cell-Driven Regeneration in Zebrafish 85

9. 5-mM phenylthiourea (PTU) stock solution: 760 mg of PTU

in 1 L of distilled water. Heat the solution to 60  C until
completely dissolved.
10. 24.2-mM tricaine (3-amino benzoic acid ethyl ester, also
known as MS-222) stock solution: 400 mg of tricaine powder
in 100 mL of distilled water, pH 7–7.5.
11. 100% dimethyl sulfoxide (DMSO, molecular grade).
12. Epifluorescence microscope with DsRed and CFP filters.
13. 10-mM Mtz solution: 0.171 g of Mtz in 100 mL of egg water
supplemented with 0.2% DMSO and 0.2-mM PTU. Stir rap-
idly using a magnetic stirrer at room temperature (RT) for
30 min.
14. 12-well tissue culture plates.
15. 15-mL conical tubes.
16. Aluminum foil.
17. 3% methyl cellulose: add 15 g of methyl cellulose into 500 mL
of heated egg water (60  C) and stir with a magnetic stirrer at
RT for 20 min.
18. Superfine eyelash with handle (Ted Pella Inc., Prod No. 113).
19. ImageJ software: it can be obtained at https://imagej.nih.gov/
ij/index.html.
20. Nitrile gloves.
21. System water: water used to raise and maintain zebrafish.

3 Methods

The outline of the screening process is shown in Fig. 1.

3.1 Preparation of 1. In each mating tank, place Tg(fabp10a:CFP-NTR);Tg

Larvae (fabp10a:DsRed) male (3) and female (3) fish that are separated
by sex with a tank divider. To obtain enough embryos for
screening, set up five tanks. Keep tanks at 28  C overnight.
2. On the following morning, replenish the tanks with fresh
system water, and remove the dividers. After 1–2 h, collect
embryos by straining the water containing eggs with a fine
plastic mesh strainer.
3. Rinse the strainer with a fine stream of system water from a
squeeze bottle. Transfer embryos into 100-mm petri dishes
with ~25 mL of fresh egg water. Keep no more than
100 embryos per dish and incubate them at 28  C.
4. About 5–6 h later, under a stereomicroscope, remove unfertil-
ized embryos using a transfer pipette. Replenish with fresh egg
water and keep at 28  C overnight.
86 Sungjin Ko and Donghun Shin

Fig. 1 Cartoon illustrating the entire chemical screening process for identifying compounds that affect
LPC-driven liver regeneration. Red circles denote livers

5. On the next morning around 24-h post-fertilization (hpf),

check the embryos under a stereomicroscope, and remove
dead or deformed embryos using a transfer pipette. After clean-
ing, to inhibit embryo pigmentation, add PTU (final concen-
tration: 0.2 mM) into each petri dish, and raise embryos until
3.3 days post-fertilization (dpf) at 28  C (see Note 1).
6. Anesthetize the embryos by adding tricaine (final concentra-
tion: 0.016%) into the petri dishes (see Note 2). Sort CFP and
DsRed double-positive larvae under an epifluorescence
 Liver Progenitor Cell-Driven Regeneration in Zebrafish 87

microscope. Transfer the double-positive embryos into a new

100-mm petri dish. Remove the egg water containing tricaine
and wash three times with egg water. Then, add fresh egg water
supplemented with 0.2-mM PTU and keep the embryos at
28  C.

3.2 Mtz and 1. Prepare 10-mM Mtz solution (see Subheading 2, item 13, and
Compound Treatments Note 3).
2. Prepare the mixture of each compound and Mtz in a 15-mL
conical tube: add 6 μL of each compound solution (see Note 4)
into 2-mL Mtz solution and vortex (see Note 5). For a control
group, add 6 μL of DMSO into 2-mL Mtz solution (see Note
6).
3. Dispense 10–15 larvae into each well of 12-well tissue culture
plates using a transfer pipette (see Note 5). Remove egg water
from each well and add 2 mL of Mtz/compound mixture. To
prevent the photoinactivation of Mtz and compounds, cover
the plates with aluminum foil, and incubate them at 28  C for
36 h.
4. Prepare each compound solution in egg water: add 6 μL of each
compound solution and 4 μL of DMSO into 2 mL of egg water
supplemented with 0.2-mM PTU and vortex. For a control
group, add 10 μL of DMSO into 2 mL of egg water supple-
mented with 0.2-mM PTU (see Note 5).
5. After 36 h of incubation, to immobilize the larvae, add a lower
concentration of tricaine (final concentration: 0.004%) (see
Note 7). Sort out larvae with well-ablated livers (optional, see
Note 8). Remove the Mtz/compound solution and add 2 mL
of egg water. Gently transfer the embryos into a well of a new
12-well plate using a transfer pipette and swirl the plate twice to
wash the embryos.
6. Remove the egg water and add 2 mL of the compound-
containing egg water. Cover the plates with aluminum foil
and incubate at 28  C for 24 h.

3.3 Image 1. Anesthetize the larvae by adding tricaine (final concentration:

Acquisition and 0.016%) into each well of the 12-well plates. Gently transfer the
Analysis larvae into 60-mm petri dishes filled with 3% methyl cellulose
and using a superfine eyelash (see Subheading 2, item 18);
orient the larvae laterally to clearly reveal the liver.
2. Take a picture using an epifluorescence microscope with a
DsRed filter (Fig. 2).
3. Once images are acquired, the ImageJ software is used to
measure liver area, the integrated density of DsRed, and the
mean background fluorescence, which are used to calculate
corrected total cell fluorescence (CTCF) using the formula
88 Sungjin Ko and Donghun Shin

Fig. 2 Chemical screening identifies SU5416 and U0126 as a negative and a

positive regulator of LPC-driven liver regeneration, respectively. (a) Experimental
scheme illustrating the stages of Mtz and compound treatments and analysis
(arrow). A0h and A36h, ablation for 0 and 36 h, respectively; R24h, regeneration
for 24 h. (b) Epifluorescence images showing fabp10a:DsRed expression in
regenerating larvae at R24h. 2-μM SU5416 and 100-μM U0126 were used.
Scale bar, 100 μm. (c) Quantification of fabp10a:DsRed expression in the
regenerating livers, as shown in (b). One-way ANOVA; *, P < 0.05; error bars,
 SEM

below [14]. CTCF value represents DsRed expression level

(Fig. 2c).
CTCF ¼ Integrated density—(selected area
the mean fluo-
rescence of background readings).

4 Notes

1. PTU is toxic. Use in accordance with appropriate handling

guidelines.
2. Tricaine is toxic. Use in accordance with appropriate handling
guidelines.
3. Prolonged exposure or increased concentration of Mtz is toxic.
Use in accordance with appropriate handling guidelines.
4. The working concentration of each compound can be deter-
mined by treating Tg(fabp10a:CFP-NTR) larvae in a 96-well
plate (three larvae per well) with various concentrations of the
 Liver Progenitor Cell-Driven Regeneration in Zebrafish 89

compound together with 10-mM Mtz; the maximum tolerated

concentration is used for screening.
5. For a well of a 12-well plate, 2 mL is a proper volume (10–15
larvae per well); 5 mL is for a well of a 6-well plate (15–25
larvae per well).
6. Most chemicals are dissolved in DMSO as a concentrated stock.
To prevent any chemical hazards, nitrile gloves are recom-
mended during all screening procedures.
7. Tricaine is metabolized in the liver and can be hepatotoxic; to
avoid aggravating liver damage in the Mtz-treated larvae, use
1/4–1/5 dose of tricaine.
8. (Optional) Remove larvae with partial ablation by examining
them under an epifluorescence microscope with a DsRed filter.
Use larvae with a tiny liver, indicating the full ablation of
hepatocytes. Caution: be aware that some compounds can
suppress NTR/Mtz-induced hepatocyte ablation, preventing
the full ablation. Thus, the effect of these compounds on
LPC-driven liver regeneration cannot be assessed using the
zebrafish hepatocyte ablation model.

Acknowledgments

We thank Jackie Russell, Mehwish Khaliq, and Michael Tsang for

editing the manuscript. This study was supported by an NIH grant
(DK101426) to D.S.

References

1. Michalopoulos GK (2013) Principles of liver Hepatocyte transplantation program: lessons

regeneration and growth homeostasis. Compr learned and future strategies. World J Gastro-
Physiol 3(1):485–513. https://doi.org/10. enterol 22(2):874–886. https://doi.org/10.
1002/cphy.c120014 3748/wjg.v22.i2.874
2. Boulter L, Lu WY, Forbes SJ (2013) Differen- 6. Garg V, Garg H, Khan A, Trehanpati N,
tiation of progenitors in the liver: a matter of Kumar A, Sharma BC, Sakhuja P, Sarin SK
local choice. J Clin Invest 123(5):1867–1873. (2012) Granulocyte colony-stimulating factor
https://doi.org/10.1172/JCI66026 mobilizes CD34(+) cells and improves survival
3. Forbes SJ, Newsome PN (2016) Liver regen- of patients with acute-on-chronic liver failure.
eration - mechanisms and models to clinical Gastroenterology 142(3):505–512.e1.
application. Nat Rev Gastroenterol Hepatol https://doi.org/10.1053/j.gastro.2011.11.
13(8):473–485. https://doi.org/10.1038/ 027
nrgastro.2016.97 7. Alwahsh SM, Rashidi H, Hay DC (2017) Liver
4. Kim WR, Lake JR, Smith JM, Schladt DP, cell therapy: is this the end of the beginning?
Skeans MA, Harper AM, Wainright JL, Snyder Cell Mol Life Sci 75(8):1307–1324. https://
JJ, Israni AK, Kasiske BL (2018) OPTN/ doi.org/10.1007/s00018-017-2713-8
SRTR 2016 annual data report: liver. Am J 8. Lukacs-Kornek V, Lammert F (2017) The pro-
Transplant 18(Suppl 1):172–253. https://doi. genitor cell dilemma: cellular and functional
org/10.1111/ajt.14559 heterogeneity in assistance or escalation of
5. Ibars EP, Cortes M, Tolosa L, Gómez-Lechón liver injury. J Hepatol 66(3):619–630.
MJ, López S, Castell JV, Mir J (2016) https://doi.org/10.1016/j.jhep.2016.10.033
90 Sungjin Ko and Donghun Shin

9. Raven A, Lu WY, Man TY, Ferreira-Gonzalez S, 138:651–679. https://doi.org/10.1016/bs.

O’Duibhir E, Dwyer BJ, Thomson JP, Meehan mcb.2016.10.004
RR, Bogorad R, Koteliansky V, Kotelevtsev Y, 12. Choi TY, Ninov N, Stainier DY, Shin D (2014)
Ffrench-Constant C, Boulter L, Forbes SJ Extensive conversion of hepatic biliary epithe-
(2017) Cholangiocytes act as facultative liver lial cells to hepatocytes after near total loss of
stem cells during impaired hepatocyte regener- hepatocytes in zebrafish. Gastroenterology 146
ation. Nature 547(7663):350–354. https:// (3):776–788. https://doi.org/10.1053/j.
doi.org/10.1038/nature23015 gastro.2013.10.019
10. Lu WY, Bird TG, Boulter L, Tsuchiya A, Cole 13. Ko S, Choi TY, Russell JO, So J, Monga SPS,
AM, Hay T, Guest RV, Wojtacha D, Man TY, Shin D (2015) Bromodomain and extraterm-
Mackinnon A, Ridgway RA, Kendall T, Wil- inal (BET) proteins regulate biliary-driven liver
liams MJ, Jamieson T, Raven A, Hay DC, Ire- regeneration. J Hepatol 64(2):316–325.
dale JP, Clarke AR, Sansom OJ, Forbes SJ https://doi.org/10.1016/j.jhep.2015.10.017
(2015) Hepatic progenitor cells of biliary ori- 14. McCloy RA, Rogers S, Caldon CE, Lorca T,
gin with liver repopulation capacity. Nat Cell Castro A, Burgess A (2014) Partial inhibition
Biol 17(8):971–983. https://doi.org/10. of Cdk1 in G 2 phase overrides the SAC and
1038/ncb3203 decouples mitotic events. Cell Cycle 13
11. Wiley DS, Redfield SE, Zon LI (2017) Chemi- (9):1400–1412. https://doi.org/10.4161/
cal screening in zebrafish for novel biological cc.28401
and therapeutic discovery. Methods Cell Biol
 Part III

Generation of Hepatocytes, Cholangiocytes, and Their

Progenitors
 Chapter 9

Conversion of Fibroblasts to Hepatocytes In Vitro

Pengyu Huang, Lulu Sun, Ludi Zhang, and Lijian Hui

Abstract
Primary hepatocytes are widely used in regenerative medicine, drug metabolism analysis, and in vitro drug
screens. To overcome the shortage of liver donors, several strategies, such as differentiation of pluripotent
stem cells and transdifferentiation from somatic cells, were developed to generate hepatocytes from
alternative sources. Here, we describe in detail lenti-virus-based procedure for direct conversion of
human fibroblasts to hepatocytes (hiHep cells) in vitro. A detailed protocol for preparation of human
fibroblasts from scar tissues is also provided. Based on this protocol, FOXA3, HNF1A, and HNF4A are
introduced into SV40-large-T-antigen-expressing human scar fibroblasts by lenti-virus. It usually takes
about 5–7 days to get epithelial hiHep colonies. SV40-large-T-antigen-expressing hiHep (hiHepLT) cells
are proliferative and can be expanded to a large number for potential uses.

Key words Hepatocytes, HiHep cells, Transdifferentiation, Reprogramming

1 Introduction

Most mammalian somatic cells that retain integrated genomes are

supposed to be plastic and could be converted to other cell lineage
after reprogramming. Transdifferentiation, a process of conversion
of one differentiated cell type to another, thus attracts a lot of
interest for its application potential in generating desired cell
types for regenerative medicine. For decades, many efforts have
been made to generate hepatocytes or hepatocyte-like cells by
transdifferentiation. Pancreatic cell, a lineage closely related to the
liver during development, was firstly used as the starting cell type to
generate hepatocytes in vivo and in vitro [1–4]. Recently, more
accessible cell types, such as mesenchymal stem cell and fibroblast,
showed great potential as the starting cell types for transdifferentia-
tion. Previously, we have reported the successful conversion of
mouse and human fibroblasts to hepatocytes by enforced expres-
sion of several hepatic transcription factors [5, 6]. These human-
induced hepatocytes (hiHep cells) expressed many hepatic marker
genes and acquired several hepatic functions, such as glycogen

Naoki Tanimizu (ed.), Hepatic Stem Cells: Methods and Protocols, Methods in Molecular Biology, vol. 1905,
https://doi.org/10.1007/978-1-4939-8961-4_9, © Springer Science+Business Media, LLC, part of Springer Nature 2019

93
94 Pengyu Huang et al.

storage, secretion of albumin, and metabolism of drugs. The

hiHepLT cells also showed the ability of integration into immune-
deficient mouse livers and functions in vivo [6].
In this chapter, we provided an example of protocol to convert
human fibroblasts to expandable hiHepLT cells. The human fibro-
blasts are generated from human scar tissues and immortalized by
enforced expression of SV40 large T antigen before transdifferen-
tiation. SV40-large-T-antigen-expressing fibroblasts are further
infected with lenti-virus expressing FOXA3, HNF1A, and
HNF4A. The medium is then exchanged to hiHep cell culture
medium supplemented with 1% fetal bovine serum (FBS) 2 days
after infection. Usually, epithelial colonies form on the dishes after
5–7 days’ induction (Fig. 1). Co-expression of albumin and AAT is
an important feature of hepatocytes and is usually used to evaluate
the efficiency of generation of hiHepLT cells. Around 10–40% of
SV40-large-T-antigen-expressing fibroblasts can be converted to
albumin and AAT double-positive cells in most cases (Figs. 2 and
3). Compared to hiHepLT cells, fibroblasts are more sensitive to
collagenase digestion. Thus we can use collagenase to eliminate
fibroblasts and enrich hiHepLT cells (Figs. 2 and 3).
Human fibroblasts derived from fetal tissue, foreskin, and scar
tissue are usually used as the starting cells due to their easy accessi-
bility and proliferative potential. Primary human fibroblasts are
capable of proliferation for 20–30 passages under recommended
culture conditions. However, hiHep cells generated from human
fibroblasts usually fail to proliferate. Thus it is not applicable to get
a large number of hiHep cells. To overcome this issue, we intro-
duced SV40 large T antigen into human fibroblasts before trans-
differentiation. HiHepLT cells derived from SV40-large-T-antigen-
expressing fibroblasts acquired the ability to proliferate. Expansion
of hiHepLT cells to the number of around 3
109 has been
achieved for hiHep-cell-based bioartificial liver device [7]. Impor-
tantly, hiHepLT cells of late passage still express hepatic genes and
retain hepatic functions [7]. In addition to fibroblasts, other easy
accessible cells, such as mesenchymal stem cells, exfoliated renal
epithelial cells from urine, and umbilical cells, are also possible cell
sources for hepatic transdifferentiation [8, 9].
Generation of hiHep cells without expression of SV40 large T
antigen is also practicable with a similar protocol. However, hiHep
cells are not proliferative. Only scattered epithelial cells and a few
epithelial clusters could be found after 7–10 days’ induction. More-
over, the transdifferentiation efficiency is largely affected by starting
cell density and multiplicity of the infection (MOI) of the virus.
Thus it is very important to perform pre-experiments to test the
best starting cell density and MOI of the virus. Comparing to
fibroblasts without expression of SV40 large T antigen, SV40-
large-T-antigen-expressing fibroblasts are less sensitive to the start-
ing cell density and MOI during induced hepatic
 Hepatic Transdifferentiation 95

Fig. 1 Representative pictures of human scar fibroblasts and hiHepLT cells. Primary human scar fibroblasts
(passage 3) were generated from human scar tissue and infected with lenti-virus expressing SV40 large T
antigen (left). HiHepLT cells were converted from SV40-large-T-antigen-expressing human scar fibroblasts by
enforced expressions of FOXA3, HNF1A, and HNF4A (right). Scale bar: 500 μm

Fig. 2 Representative pictures of hiHepLT cells stained with albumin and α-1-antitrypsin (AAT) before and after
enrichment. Scale bar: 200 μm

transdifferentiation. It is highly recommended to start with SV40-

Large-T-Antigen-expressing fibroblasts for researchers who have
no experience in hepatic transdifferentiation.
The hepatic transcription factors used for hepatic transdiffer-
entiation were optimized and minimized to FOXA3, HNF1A, and
HNF4A in our previous study [6]. However, several studies also
indicated that different combinations of transcription factors could
achieve similar transdifferentiation processes [5, 6, 10]. After inter-
rogation of the mechanisms of hepatic transdifferentiation,
96 Pengyu Huang et al.

Fig. 3 Albumin and AAT double-positive hiHepLT cells were quantified by flow cytometry before and after
enrichment

inactivation or activation of certain signaling pathways may also be

used for improvement of this process in the future [11].

2 Materials

2.1 Biological Human scar tissues are taken from scar tissues removed during
Material surgical operations after written consent following guidelines
approved by an Ethic Committee.

2.2 Culture Media 1. 293FT cells.

and Reagents 2. The human fibroblast culture (HFC) medium: DMEM/F12
1:1 mixture supplemented with 10% fetal bovine serum (FBS),
recombinant human 5-ng/mL bFGF, 100-nM
β-mercaptoethanol, 1
MEM non-essential amino acids
(MEM NEAA), and 0.5-mM sodium pyruvate (see Note 1).
3. The hiHep cell culture medium: DMEM/F12 1:1 mixture
supplemented with 40-ng/mL recombinant human TGF-α,
40-ng/mL recombinant human EGF, 10-μM dexamethasone,
1
insulin-transferrin-selenium (ITS), 2-g/L galactose, 0.1-
g/L ornithine, 0.3-g/L proline, 0.6-g/L nicotinamide, 0.544-
 Hepatic Transdifferentiation 97

mg/L ZnCl2, 0.75-mg/L ZnSO4·7H2O, 0.2-mg/L

CuSO4·5H2O, 0.025-mg/L MnSO4.
4. 2-mg/mL collagenase IV solution: dissolve 20-mg collagenase
IV in 10-mL PBS.
5. pWPI expression vector.
6. Collagen I solution: Dissolve rat tail-derived collagen I in 0.6%
(v/v) acetic acid solution to 25 mg/L at 4  C overnight.
7. Calcium phosphate solution: 7.5-μg psPAX2, 3-μg pMD2.G,
31-μL 2.5-M CaCl2 in bi-distilled water. The total volume is
adjusted to 250 μL.
8. 2
HEPES solution: 0.8-g NaCl, 0.027-g Na2HPO4·2H2O,
and 1.2-g HEPES in 100-mL bi-distilled water. Calibrate the
pH to 7.05.
9. 0.05% trypsin.
10. 8-mg/mL polybrene.

3 Methods

3.1 Coating Dish 1. Add collagen I solution to cell culture dishes (1.5-μg/cm2
surface).
2. Air-dry the collagen I-coated dishes in a biosafety cabinet for
future use.

3.2 Preparation 1. CDNA of SV40 large T antigen gene, FOXA3, HNF1A, and
of Lenti-Virus HNF4A were inserted into a pWPI expression vector.
2. Seed 1
106 293FT cells on a 100-mm dish, and add 10-mL
DMEM +10% FBS. Culture 293FT cells for 24 h.
3. One hour before transfection, exchange the culture medium
with 6 mL of prewarmed DMEM+10% FBS.
4. Mix 250-μL calcium phosphate solution containing 10-μg
pWPI expression vector and 250-μL 2
HEPES solution by
pipetting. Keep the mixture for 5 min at room temperature
then transfer the mixture to 293FT cells cultured in a
100-mm dish.
5. Incubate the cells for 6 h and exchange the medium with
DMEM +10% FBS.
6. 48 h after transfection, collect supernatant and filtered by a
0.45-μm filter. Store the virus solution in 80  C.

3.3 Preparation 1. Wash human scar tissue with 70% ethanol three times, then
of Human Scar with PBS once.
Fibroblasts 2. Cut the tissue into 2–5-mm3 pieces.
98 Pengyu Huang et al.

3. Place three tissue pieces per 60-mm collagen I-coated dish, and
incubate at 37  C for 1 h in an incubator.
4. Carefully add 5 mL of HFC medium to the dish (see Note 2).
5. Put the dish in an incubator at 37  C for 5 days without
exchange of the medium, and then exchange the medium
every 2 days.
6. Human scar fibroblasts migrate out of the tissue after
2–3 weeks (passage 1).
7. When the cells become confluent, remove HFC medium and
the tissues.
8. Wash once with PBS and trypsinize with 1 mL of 0.05% trypsin
for 15 min.
9. Suspend the cells with 2 mL of HFC medium. Centrifuge the
cells at 150
g for 5 min.
10. Discard the supernatant, resuspend the cells with 5 mL of HFC
medium, and transfer to a 60-mm collagen I-coated dish (pas-
sage 2).
11. When the cells become confluent, passage the cells to new
collagen I-coated dishes (5
105 cells per 60-mm collagen
I-coated dish, passage 3). Exchange HFC medium every
2 days.
12. For the following passages, seed 5
105 cells per 60-mm
collagen I-coated dish and exchange HFC medium every
2 days.

3.4 Immortalization 1. Seed 5
105 human scar fibroblasts (passage 3–5) on a 60-mm
of Human Scar collagen I-coated dish, and add HFC medium to 5 mL.
Fibroblasts 2. The next day, thaw lenti-virus expressing SV40 large T antigen
on ice. Mix lenti-virus (MOI ¼ 2), 5 μL of 8-mg/mL poly-
brene and HFC medium to the volume of 5 mL.
3. Remove the medium from the dish. Add 5 mL of lenti-virus,
polybrene, and HFC medium mixture.
4. After 24 h infection, remove the medium and add 5 mL of fresh
HFC medium.
5. When the cells become confluent, passage the cells to new
collagen I-coated dishes (5
105 cells per 60-mm dish).
6. For the following passages, seed 5
105 cells per 60-mm
collagen I-coated dish and exchange HFC medium every
2 days.

3.5 Conversion 1. Expand human scar fibroblasts expressing SV40 large T antigen
of HSFLT to hiHepLT (HSFLT) to 70–80% confluence.
Cells 2. Discard the medium and wash twice with 2 mL of PBS.
 Hepatic Transdifferentiation 99

3. Trypsinize the cells with 1 mL of 0.05% trypsin at 37  C for

5 min.
4. Add 2 mL of HFC medium to suspend the cells, and transfer
the cells to a 15-mL tube. Count the number of cells.
5. Thaw lenti-virus expressing FOXA3, HNF1A, and HNF4A
on ice.
6. Mix lenti-virus (MOI ¼ 1.25 for each virus, see Note 3), 5 μL
of 8-mg/mL polybrene, 2
105 HSFLT (see Note 3), and
HFC medium to the final volume of 5 mL. Mix gently by
pipetting up and down. Transfer the mixture to a 60-mm
collagen I-coated dish. Incubate the cells at 37  C (day 0).
7. 24 h after infection, remove the medium and add fresh HFC
medium (day 1).
8. The next day, remove the medium. Add 5 mL of hiHep cell
culture medium with 1% FBS (day 2).
9. From day 4 to day 8, exchange cell culture medium with hiHep
cell culture medium with 1% FBS every 2 days. Epithelial
colonies usually form between day 5 and day 7.
10. On day 10, remove cell culture medium, and wash twice with
2 mL of PBS.
11. Trypsinize the cells with 1 mL of 0.05% trypsin at 37  C for
5 min.
12. Suspend the cells with hiHep cell culture medium containing
1% FBS.
13. Transfer 5
105 cells to each 60-mm collagen I-coated dish.
Add hiHep cell culture medium containing 1% FBS to 5 mL.

3.6 Enrichment 1. On day 12, exchange the hiHep cell culture medium with
of hiHepLT Cells 1% FBS.
2. On day 14, aspirate the medium, and wash twice with 2 mL
of PBS.
3. Add 1 mL of 2-mg/mL collagenase IV solution.
4. Incubate the cells for 5–10 min.
5. Carefully monitor the cells under a microscope. When nearly
50% of the cells curl up and start to detach from the dish
surface, gently tap the dish until most of the fibroblasts disso-
ciate, while most of epithelial colonies remain on the dish.
6. Aspirate the supernatant. Wash with 4 mL of PBS to remove
collagenase IV.
7. Add 1 mL of 0.05% trypsin and incubate the cells at 37  C for
3–5 min.
8. Suspend the cells with hiHep cell culture medium containing
1% FBS.
100 Pengyu Huang et al.

9. Transfer 5
105 cells to each 60-mm collagen I-coated dish.

Add 5 mL of hiHep cell culture medium with 1% FBS.
10. Exchange cell culture medium with fresh hiHep cell culture
medium with 1% FBS every 2 days until the cells become
confluent (see Note 4).

3.7 Passage 1. When the cells become confluent, discard the medium, and
of HiHepLT Cells wash twice with 2 mL of PBS.
2. Trypsinize the cells with 1 mL of 0.05% trypsin at 37  C for
5 min.
3. Add 2 mL of hiHep cell culture medium supplemented with
1% FBS.
4. Suspend the cells, and transfer 5
105 cells to each 60-mm
collagen I-coated dish.
5. Add 5 mL of hiHep cell culture medium with 1% FBS.
6. Exchange medium with fresh hiHep cell culture medium with-
out FBS every day.
7. For functional assays, usually use hiHepLT cells 4–8 days after
seeding (see Note 5).

4 Notes

1. The DMEM/F12 1:1 mixture already contains 0.5-mM

sodium pyruvate. Add extra sodium pyruvate to 1 mM in the
final human fibroblast culture medium.
2. The tissue pieces should be attached to the surface of the dish
after adding HFC medium.
3. MOI and starting cell number significantly affect the efficiency
of transdifferentiation. The best MOI and starting cell number
vary for different HSFLT strains. It is highly recommended to
perform pre-experiments to test the best MOI and starting cell
number for each HSFLT strain.
4. We usually perform 1–2 rounds of enrichment to remove most
of the cells that fail to be transdifferentiated.
5. HiHepLT cells will undergo further maturation in FBS-free
hiHep cell culture medium after confluence.

References
1. Scarpelli DG, Rao MS (1981) Differentiation 2. Shen CN, Horb ME, Slack JM, Tosh D (2003)
of regenerating pancreatic cells into Transdifferentiation of pancreas to liver. Mech
hepatocyte-like cells. Proc Natl Acad Sci U S Dev 120(1):107–116
A 78(4):2577–2581 3. Reddy JK, Rao MS, Qureshi SA, Reddy MK,
Scarpelli DG, Lalwani ND (1984) Induction
 Hepatic Transdifferentiation 101

and origin of hepatocytes in rat pancreas. J Cell Li YH, Bao XC, Tse HF, Grillari J, Grillari-
Biol 98(6):2082–2090 Voglauer R, Pei DQ, Esteban MA (2012) Gen-
4. Shen CN, Slack JM, Tosh D (2000) Molecular eration of human induced pluripotent stem
basis of transdifferentiation of pancreas to liver. cells from urine samples. Nat Protoc 7
Nat Cell Biol 2(12):879–887. https://doi. (12):2080–2089. https://doi.org/10.1038/
org/10.1038/35046522 nprot.2012.115
5. Huang P, He Z, Ji S, Sun H, Xiang D, Liu C, 9. Gao Y, Zhang X, Zhang L, Cen J, Ni X, Liao X,
Hu Y, Wang X, Hui L (2011) Induction of Yang C, Li Y, Chen X, Zhang Z, Shu Y,
functional hepatocyte-like cells from mouse Cheng X, Hay DC, Lai D, Pan G, Wei G, Hui
fibroblasts by defined factors. Nature 475 L (2017) Distinct gene expression and epige-
(7356):386–389. https://doi.org/10.1038/ netic signatures in hepatocyte-like cells pro-
nature10116 duced by different strategies from the same
6. Huang P, Zhang L, Gao Y, He Z, Yao D, Wu Z, donor. Stem Cell Reports 9(6):1813–1824.
Cen J, Chen X, Liu C, Hu Y, Lai D, Hu Z, https://doi.org/10.1016/j.stemcr.2017.10.
Chen L, Zhang Y, Cheng X, Ma X, Pan G, 019
Wang X, Hui L (2014) Direct reprogramming 10. Du Y, Wang J, Jia J, Song N, Xiang C, Xu J,
of human fibroblasts to functional and expand- Hou Z, Su X, Liu B, Jiang T, Zhao D, Sun Y,
able hepatocytes. Cell Stem Cell 14 Shu J, Guo Q, Yin M, Sun D, Lu S, Shi Y, Deng
(3):370–384. https://doi.org/10.1016/j. H (2014) Human hepatocytes with drug met-
stem.2014.01.003 abolic function induced from fibroblasts by
7. Shi XL, Gao YM, Yan YP, Ma HC, Sun LL, lineage reprogramming. Cell Stem Cell 14
Huang PY, Ni X, Zhang LD, Zhao X, Ren HZ, (3):394–403. https://doi.org/10.1016/j.
Hu D, Zhou Y, Tian F, Ji Y, Cheng X, Pan GY, stem.2014.01.008
Ding YT, Hui LJ (2016) Improved survival of 11. Ji S, Zhu L, Gao Y, Zhang X, Yan Y, Cen J,
porcine acute liver failure by a bioartificial liver Li R, Zeng R, Liao L, Hou C, Gao Y, Gao S,
device implanted with induced human func- Wei G, Hui L (2017) Baf60b-mediated
tional hepatocytes. Cell Res 26(2):206–216. ATM-p53 activation blocks cell identity con-
https://doi.org/10.1038/cr.2016.6 version by sensing chromatin opening. Cell Res
8. Zhou T, Benda C, Dunzinger S, Huang YH, 27(5):642–656. https://doi.org/10.1038/cr.
Ho JC, Yang JY, Wang Y, Zhang Y, Zhuang Q, 2017.36
 Chapter 10

Conversion of Fibroblasts to Hepatocyte-Like Cells In Vivo

Guangqi Song, Qinggong Yuan, Zhen Dai, Hsin-Chieh Tsay, Xizhong Shen,
Michael Ott, and Amar Deep Sharma

Abstract
In vivo conversion of fibroblasts into hepatocyte-like cells provides one potential approach for the treatment
of liver fibrosis. In our previous study, we showed in vivo conversion of myofibroblasts into induced
hepatocytes (iHeps) by forced expression of four transcription factors in genetic fate-tracing mouse
model of chronic liver disease. These in vivo-generated iHeps showed similar expression profile with
endogenous hepatocytes (eHeps) and also exhibited similar functional characteristics, such as albumin
secretion, urea synthesis, cytochrome activity, and drug responsiveness. Furthermore, the targeted expres-
sion of our reprogramming factors in myofibroblasts attenuated liver fibrosis. Our study suggests that
in vivo reprogramming may open new perspectives for the treatment of diseases such as liver fibrosis.

Key words In vivo reprogramming, Induced hepatocytes (iHeps)

1 Introduction

Myofibroblasts contribute to liver fibrosis by secreting extracellular

matrix in most types of chronic liver diseases. Therefore, we pro-
posed an approach to ameliorate liver fibrosis by converting myofi-
broblasts directly into hepatocytes by introducing transcription
factors. Recently, a novel strategy to convert fibroblasts into
hepatocyte-like cells in vitro by forced expression of four transcrip-
tion factors has been established [1–4]. In our recent study, we
reported effectively trans-reprogramming of myofibroblasts
derived from hepatic stellate cell (HSC) into hepatocyte-like cells
(iHeps) by four transcription factors FOXA3, GATA4, HNF1A,
and HNF4A (4TFs) in vitro. Furthermore, we showed in vivo
converted myofibroblasts into iHeps by forced expression of 4TFs
in genetic fate-tracing mouse model of chronic liver disease
[5]. 4TFs were directly delivered into myofibroblasts using a
p75NTRp-tagged recombinant adenoviral vector (serotype 5),
which was designed to target mouse myofibroblasts through cou-
pling with a peptide (S11-NGFp) [6]. S11-NGFp consisting of a

Naoki Tanimizu (ed.), Hepatic Stem Cells: Methods and Protocols, Methods in Molecular Biology, vol. 1905,
https://doi.org/10.1007/978-1-4939-8961-4_10, © Springer Science+Business Media, LLC, part of Springer Nature 2019

103
104 Guangqi Song et al.

Ad. 4TF S11-NGF peptide p75NTR present on myofibroblasts

B
n ps n ps
ctio f iHe ctio f iHe
nje no nje no
T Fi tio TFi tio
CCl4 injection (twice weekly) .4 tec DDC diet .4 tec
Ad De Ad De
8 weeks 4 weeks

day 0 day 56 day 63 day 93 day 0 day 28 day 35 day 65

Fig. 1 (a) Schematic of in vivo conversion of myofibroblasts into iHeps in LratCre-mT/mG mice using
p75NTRp-tagged recombinant adenoviral vector. In this lineage-tracing model, the endogenous hepatocytes
(eHeps) would express membranous tdTomato, while myofibroblasts and induced hepatocytes (iHeps) derived
from myofibroblasts would express membranous EGFP. (b) Schematic of in vivo iHep generation in CCl4 and
DDC diet-induced liver fibrosis disease mouse model

single chain antibody fragment (S11) directed against the Ad fiber

knob and a nerve growth factor peptide (NGFp) that recognizes
p75NTR on HSCs [6]. The in vivo-generated iHeps were observed
in the liver of the lineage-tracing mice on the 30th day after viral
injection. We sorted these in vivo-generated iHeps and characterize
their function in vitro. They showed similar expression profile to
endogenous hepatocytes (eHeps) and also exhibited similar func-
tional characteristics, such as albumin secretion, urea synthesis,
cytochrome activity, and drug responsiveness. Furthermore, con-
version of myofibroblasts into iHeps in vivo ameliorated liver fibro-
sis. Both CCl4 and DDC diet-induced fibrotic model showed lower
level of liver fibrosis after treatment with adenovirus vector contain-
ing 4TFs. Our study suggested that in vivo trans-reprogramming
may open new perspectives for the treatment of chronic liver
diseases.
In this chapter, we describe our method for conversion of
fibroblasts to iHeps in vivo (Fig. 1). Our strategy requires a line-
age-tracing model and a modified target adenoviral vector system.
LratCre-mT/mG mice have been shown to effectively trace myofi-
broblasts derived from the HSC lineage [7]. The modified adeno-
viral vector coupling with a peptide system has been shown to
effectively target myofibroblasts in vivo [6]. Here, we describe a
generation of iHeps using LratCre-mT/mG mice and p75NTRp-
tagged adenoviral vector in vivo. Additionally, we describe
 Conversion of Fibroblasts to Hepatocyte-Like Cells In Vivo 105

characterization of in vivo-generated iHeps and evaluation of liver

fibrosis level after in vivo conversion.

2 Materials

2.1 Animal Model 1. Gt(ROSA)26Sortm4(ACTB-tdTomato,-EGFP) mice (referred

to as mT/mG).
2. Lrat-Cre mice.
3. 10% carbon tetrachloride (CCl4) dissolved in corn oil.
4. 3,5-Diethoxycarbonyl-1,4-dihydrocollidine (0.1% DDC)-diet.
5. 3 M Vetbond™ No. 1469SB.

2.2 Cell 1. Ad.4TF: The human cDNAs of FOXA3, GATA4, HNF1A,

and Molecular Biology and HNF4A were cloned together into a p75NTRp-tagged
recombinant adenoviral vector (serotype 5, referred to as
Ad.4TF).
2. A peptide (single chain antibody fragment) of the nerve growth
factor (S11-NGFp), which was designed for coupling of
Ad.4TF fiber knobs and binding to the p75 neurotrophin
receptor (p75NTR) present on hepatic stellate cells and
myofibroblasts.
3. pAdEasy1 plasmid.
4. E. coli BJ5183 bacterial cells.
5. HEK293 cells.
6. Caesium chloride (CsCl).
7. Ketamin.
8. Rompun (Bayer).
9. Perfusion solution: 0.5-mM ethylene glycol tetra acetic acid
(EGTA) and 10-mM HEPES in Earle’s Balanced Salt Solution
(EBSS), pH 7.2–7.4.
10. Liberase solution: 100-μg/mL Liberase (Roche Liberase™ TL
Research Grade) and 10-mM HEPES in EBSS, pH 7.2–7.4.
11. Dulbecco’s Modified Eagle’s Medium (DMEM).
12. Fetal calf serum (FCS) (PAN biotech).
13. 100-μm nylon filter.
14. Percoll.
15. Trypan blue.
16. Sorting buffer: 0.5% BSA and 1-mM EDTA in Ca/Mg-
free PBS.
17. Collagen from rat tail tendon (Roche).
106 Guangqi Song et al.

18. Hepatocyte Culture Medium BulletKit (HCM, Lonza),

including 500-mL hepatocyte basal medium (HBM) and
HCM SingleQuots Kit (contains ascorbic acid, 0.5 mL;
BSA-FAF, 10 mL; hydrocortisone, 0.5 mL; human hEGF,
0.5 mL; transferrin, 0.5 mL; insulin, 0.5 mL; GA, 0.5 mL).
19. RNeasy Mini Kit (Qiagen).
20. TRIzol reagent (Invitrogen).
21. Whole Mouse Genome Oligo Microarray v2 (4  44K) (Agi-
lent Technologies).
22. Albumin ELISA Quantitation Set (Bethyl Laboratories).
23. QuantiChrom Urea Assay Kit (BioAssay Systems).
24. P450-Glo CYP1A2, CYP2C9, and CYP3A assay kits
(Promega).
25. Working solution of Oil red O: 90-mg Oil red O in 50-mL 60%
isopropanol.
26. 1% Perid acid.
27. Schiff’s reagent.
28. Methanol/acetone-mix (1:1): The solution should be keep at
20  C.
29. DiI-ac-LDL.
30. Indocyanine green (ICG).
31. Glucagon (Sigma).
32. pCPT-cAMP (Sigma).
33. Insulin.
34. Glucose assay kit (Invitrogen).
35. Phenobarbital.
36. Rifampicin.
37. Agilent Genome Microarray Kits 4  180k arrays (G4125A,
Agilent Technologies).
38. Primers and probes for quantitative real-time PCR (qPCR)
(Tables 1 and 2).

2.3 Histology, 1. Picro-Sirius Red solution: 0.1% direct red 80 plus 0.1% fast
Immunohisto- green dissolved in 1.2% saturated aqueous picric acid solution.
chemistry, and 2. Antibodies for immunofluorescence staining: Albumin
Immunofluorescence (Abcam-19196), MUP (Santa Cruz-21856), FAH (Abcam-
81087), AAT (Abcam-117307), p75-NTR (Abcam-8874),
desmin (Thermo Scientific-RB-9140), HNF4A (Santa Cruz-
6556).
3. Hydroxyproline Assay Kit (Sigma).
 Conversion of Fibroblasts to Hepatocyte-Like Cells In Vivo 107

Table 1
Primers and probes for detection of exogenous expression

Primers

Gene Forward Reverse

HNF4A CATGTACAGCTGCCGGTTCA ACTGGTTCCGCTTATCCTTGTC
Taqman probe Catalog number
GATA4 Hs00171403_m1 Thermo Fisher
HNF1A Hs01551752_m1 Thermo Fisher
FOXA3 Hs00270130_m1 Thermo Fisher

Table 2
Primers and probes for drug responsiveness assay

Primers

Gene Forward Reverse

Cyp1a1 CCTCATGTACCTGGTAACCA AAGGATGAATGCCGGAAGGT
Utg1a1 GGAGGCTGTTAGTGTTCCCT CCGTCCAAGTTCCACCAAAG
Oatp ATGGGATTCCATTCACTGGTTGTA CACGTGCTCCACAGCTGGTTA
Taqman probe Catalog number
Abcc2 Mm00496899_m1 Thermo Fisher

3 Method

3.1 Production 1. For tracing in vivo conversion, a lineage-tracing model

of Lineage-Tracing (LratCre-mT/mG mice) should be generated by crossing
Model 12-week-old Lrat-Cre mice with mT/mG mice. Cells in
LratCre-mT/mG mice express membrane-targeted tdTomato
(mT) before Cre-mediated recombination. The expression of
Cre is promoted by lecithin-retinol acyltransferase (Lrat) pro-
moter, a hepatic stellate cells’ specific promoter. LratCre-mT/
mG mice have been shown as an effectively lineage-tracing
model for myofibroblasts derived from the HSC lineage [7]
(Fig. 2).
2. For inducing liver fibrosis, LratCre-mT/mG mice are injected
with 4 μL/g of 10% CCl4 dissolved in corn oil by intraperito-
neal injection twice per week for 8 weeks (see Note 1).
3. For inducing cholestasis-induced liver fibrosis, LratCre-mT/
mG mice are fed with DDC diet for 4 weeks.
108 Guangqi Song et al.

A Lrat Cre

Cre

pCA mT pA mG pA pCA mG pA

Endogenous Hepatocytes (eHeps) Conversed Hepatocytes (iHeps)

C Control Ad. 4TF

td-Tomato EGFP Merged td-Tomato EGFP Merged

Fig. 2 (a) Cells in LratCre-mT/mG mice express membrane-targeted tdTomato (mT) before Cre-mediated
recombination. Cre expressed under the transcriptional control of the lecithin-retinol acyltransferase (Lrat)
promoter, which is specifically expressed in hepatic stellate cells in liver. Myofibroblasts derived from hepatic
stellate cells express membrane-targeted EGFP. (b) Schematic illustration shows that eHeps can be identified
by tdTomato expression, whereas iHeps are identified by EGFP expression. (c) These confocal figures show the
generation of in vivo iHeps in LratCre-mT/mG mice after injected with Ad.4TF, no reprogrammed cell was
found in control group

3.2 Conversion 1. Ad.4TF expressing the FOXA3, GATA4, HNF1A, and

of Myofibroblasts into HNF4A is generated by homologous recombination following
iHeps In Vivo co-transfection with pAdEasy1 in E. coli BJ5183. Ad.4TF
should be propagated in HEK293 cells, purified by CsCl buoy-
ant density centrifugation, and measured at OD260 [6]. The
production and titration of the Ad.4TF vectors were per-
formed following the method as previously published [8].
2. The peptide (S11-NGFp) (1  108 pfu virus with 100-μg
peptide) should be preincubated with Ad.4TF for 1 h at
37  C in serum-free DMEM before injection into LratCre-
mT/mG mice.
3. One week after cessation of the CCl4 or DDC diet treatment,
p75NTRp-tagged Ad.4TF (Ad.4TF-S11-NGFp) should be
injected into fibrotic LratCre-mT/mG mice. Briefly, mice
were anesthetized using Ketamin and Rompun. The body
cavity was opened, and 1  1010 Ad.4TF-S11-NGFp was
injected via the portal vein using 30G needle (0.3  8 mm).
After injection, press the whole of the portal vein for 1 min, and
then add one drop of tissue adhesive (3 M Vetbond™
 Conversion of Fibroblasts to Hepatocyte-Like Cells In Vivo 109

No. 1469SB) on the surface of the wound. The control group

is injected with empty adenoviral vector.
4. 30 days after injection, liver samples should be collected from
LratCre-mT/mG mice injected with Ad.4TF-S11-NGFp and
empty adenoviral vector for producing frozen sections.
5. Confirmation of in vivo iHep conversed from myofibroblasts
under fluorescence microscope. The eHeps would express
membranous tdTomato, while the iHeps could be detected
by membranous EGFP (Fig. 2).
6. Performing immunofluorescence staining for albumin, MUP,
FAH, AAT, and HNF4A, to further confirm the in vivo con-
version of myofibroblasts into iHeps expressing conventional
hepatocyte markers.

3.3 Isolation 1. Hepatocytes were isolated from mice by a modified two-step

of eHeps and iHeps Liberase perfusion [9]. Briefly, mice were anesthetized using
Ketamin and Rompun. The body cavity was opened, and a
catheter was placed into the portal vein and connected to a
flow pump, which pumped media pre-warmed at 37  C from a
water bath into the catheter, and vena cava was cut. The liver
was first perfused with perfusion solution. Subsequently, Lib-
erase solution was applied for enzymatic digestion of the tissue
at 37  C. After digestion of 10–12 min, the liver was carefully
disconnected, and the tissue was manually disrupted with ster-
ile scissors and a scalpel in DMEM containing 10% FCS. The
suspended hepatocytes were passed through a 100-μm nylon
filter into 50-mL Falcon tubes. The cell suspensions were
centrifuged twice at 300 rpm (or 50  g) for 5 min at 4  C,
and the cell pellet was re-suspended in ice-cold DMEM
medium containing 10% FCS. Cell viability was tested by
trypan blue.
2. Centrifuge the cells for 5 min 300 rpm (or 50  g), 4  C, and
re-suspend resulting pellet in sorting buffer. The concentration
should be in the range from 0.5  106 to 5  106 cells/mL.
Perform fluorescence-activated cell sorting (FACS) to separate
eHeps and iHeps (see Note 2). Collect cells into DMEM
medium containing 10% FCS; tdTomato-positive cells are
eHeps, while EGFP-positive cells are iHeps (Fig. 3, see Note 3).
3. Preparation of the collagen-coated plate. Dissolve collagen
with 0.2% acetic acid to a final concentration of 2 mg/mL.
Add the solution into a plate following 2.5 μL per 1 cm2, and
cover the surface using cell scraper. It is better to stand the
plate coated with collagen in biological safety cabinet more
than 15 min to dry them enough before use.
4. The eHeps and iHeps should be seeded in collagen-coated
plates, cultured in DMEM medium containing 10% FCS,
110 Guangqi Song et al.

Fig. 3 Isolation of in vivo-generated iHeps (EGFP positive) and eHeps (tdTomato positive) by FACS sorting

37  C, 5% CO2. After 3 h, wash cells with PBS and replace the

medium by HCM. To maintain hepatocyte, HCM was
prepared by adding all of the reagents from SingleQuots Kit
into HBM medium. The medium should be changed every
other day.
5. Isolation of mRNA and functional assays should be performed
after 24 h in culture with HCM. Those eHeps isolated from the
same mouse can be used as positive control for iHeps in all
characterization assays. Primary mouse hepatic stellate cells can
be isolated from BALB/c mice according to previously pub-
lished reports [10, 11]. These cells could be cultured in pres-
ence of PDGF to generate myofibroblasts, which can be used as
negative control for iHeps.

3.4 Characterization 1. For analysis of the stability of conversion, total RNA of iHeps
of In Vivo Generated should be isolated. Real-time qPCR can be performed using
iHeps specific primers for four human factors to confirm exogenous
silence. Array-CGH can be performed to test the genomic
integrity of reprogrammed cells using Agilent Genome Micro-
array Kits 4  180k arrays with median overall probe spacing of
about 13 kb. Labeling and hybridization of gDNA should be
performed according to the protocol provided by Agilent.
2. For analysis of the global gene expression profiles, total RNA of
eHeps, iHeps, and myofibroblasts should be isolated from cells
by TRIzol reagent following a standardized protocol. Whole
Mouse Genome Oligo Microarray v2 (4  44K) (Agilent
Technologies) could be used to characterize global gene
 Conversion of Fibroblasts to Hepatocyte-Like Cells In Vivo 111

A C D Col1a1 E Hydroxyproline
Albumin

Hydroxyproline contect
Retaive mRNA
(pg/24h/10000 cells)

expression
iHep eHep

(ug/mg)
Albumin

PAS
Control Ad. 4TF Control Ad. 4TF

iHep eHep F Control Ad. 4TF

Oill red O

B
Urea
HE
(g/24h/10000 cells)
Urea

Sirius Red
ICG

iHep eHep

Fig. 4 (a) Albumin ELISA revealed comparable levels of secreted albumin in iHeps and eHeps. (b) Urea
synthesis showed similar levels in iHeps and eHeps. (c) The iHeps show PAS staining, Oil red O staining, and
ICG uptake. (d–f) Amelioration of liver fibrosis in Ad.4TF-injected mice. (d) Reduced levels of Col1a1 mRNA in
Ad.4TF-injected mice compared to control mice. (e) Hydroxyproline assay showed decreased levels of entire
collagen content, measured in whole liver. (f) H&E and Sirius Red staining showed less fibrosis in Ad.4TF-
injected mice than respective controls

expression profiles of iHeps compared to myofibroblasts and

eHeps.
3. For analysis of albumin secretion, eHeps and iHeps were
washed with PBS and cultivated in HCM without BSA-FAF
from SingleQuots Kit. After 24 h, collect supernatants, and
measure albumin secretion using the mouse albumin ELISA
Quantitation Set according to the manufacturer’s instructions.
Results should be normalized to 10,000 cells at 24 h (Fig. 4a).
4. For analysis of urea production, wash eHeps and iHeps with
PBS, and replace the medium with HCM without BSA-FAF
from SingleQuots Kit. After 24 h, collect supernatants, and
measure urea production using the QuantiChrom Urea Assay
Kit according to the manufacturer’s instructions. Results
should be normalized to 10,000 cells at 24 h (Fig. 4b).
5. For analysis of cytochrome P450 enzymes activities, P450-Glo
CYP1A1, CYP1A2, CYP2C9, and CYP3A assay kits (Promega)
could be used to measure CYP1A1, CYP1A2, CYP2C9, and
CYP3A activities in eHeps and iHeps. Seed cells in collagen-
coated 24 well plates, 2–3  104 per well. Add specific inducers
into medium, incubate for 48–72 h according to manufac-
turer’s instructions (cell-based, non-lytic protocol), and
change medium everyday with inducers. Wash cell using PBS
112 Guangqi Song et al.

and replace by HCM medium without inducers. Add specific

substrates for CYP1A1, CYP1A2, CYP2C9, and CYP3A into
medium separately. After incubation, perform measurements
according to manufacturer’s instructions. The luminescence
signals from the P450-Glo assay were normalized to cell num-
bers (see Note 4).
6. For analysis of the ability to store glycogen, periodic acid-Schiff
(PAS) staining can be performed. Fix eHeps and iHeps with a
cold methanol/acetone-mix for 10 min at 4  C. Add 1% peri-
odic acid, and incubate the at room temperature for 5 min.
Wash cells with water, and incubate cells with Schiff’s reagents
for 15 min; examine the store glycogen under a phase contrast
microscope (Fig. 4c).
7. For analysis of the presence of triglycerides and lipids, eHeps
and iHeps should be fixed using 4% formalin for 30 min. Wash
one time with PBS, add the working solution of Oil red O, and
incubate for 10 min. Remove the working solution and wash
two times with 60% isopropanol. Add the hematoxylin coun-
terstain and incubate for 1 min, washed with PBS and exam-
ined under a phase contrast microscope (Fig. 4c, see Note 5).
8. For analysis of the ability to uptake low-density lipoprotein
(LDL), eHeps and iHeps should be incubated at 37  C for
4 h with 10-μg/mL DiI-Ac-LDL, washed with PBS and
observed by fluorescence microscopy.
9. For analysis of the ability to uptake indocyanine green (ICG),
eHeps and iHeps should be incubated at 37  C for 3 h with
0.1-mg/mL ICG, washed with PBS and examined under a
phase contrast microscope (Fig. 4c).
10. For analysis of the glucose production, eHeps and iHeps
should be treated with either glucagon or insulin. For glucagon
treatment, eHeps and iHeps should be seeded on collagen-
coated plates and cultured in HCM medium with or without
glucagon (100 nM). Similarly, for insulin treatment, cells
should be cultured in HCM medium supplemented with 0.5-
mM pCPT-cAMP and with or without 100-nM insulin. After
2 h incubation, collect the supernatant, and analyze for glucose
production using a glucose assay kit (Invitrogen).
11. For analysis of the drug responsiveness, eHeps and iHeps
should be cultured in the HCM medium with or without
1-mM phenobarbital for 72 h. Total RNA should be isolated
to perform real-time qPCR, and compare the expression level
of Cyp1a1 and Abcc2. And eHeps and iHeps should be
cultured in the HCM medium with or without 1-mM 50-μM
Rifampicin for 72 h. Total RNA should be isolated to perform
real-time qPCR, and compare the expression level of Ugt1a1
and Oatp.
 Conversion of Fibroblasts to Hepatocyte-Like Cells In Vivo 113

3.5 Evaluation 1. For analysis of liver fibrosis level of mice injected with Ad.4TF
of Liver Fibrosis Level and empty adenoviral vector, the expression level of Col1a1 in
After In Vivo the entire liver could be detected by real-time qPCR. The total
Conversion RNA should be isolated from liver tissues that were collected
on the 30th day after viral injection (Fig. 4d).
2. The entire collagen content of a liver could be detected by
hydroxyproline assay. Liver tissues should be collected from
mice injected with Ad.4TF and empty adenoviral vector, and
perform assay following the manual of Hydroxyproline Assay
Kit (Fig. 4e).
3. Paraffine sections of liver tissues from mice injected with
Ad.4TF and empty adenoviral vector can be produced follow-
ing a standardized protocol. For analysis of the fibrotic area,
Sirius Red staining could be performed. Following deparaffini-
zation, the sections should be stained with Picro-Sirius Red
solution and incubated for 60 min. Sections were rinsed with
water, dehydrated and mounted in xylene. Take ten random
photos from each section under a phase contrast microscope.
Fibrotic area could be counted using Image J software (Fig. 4f).
4. For counting the number of fibroblasts in liver tissues, Desmin
and p75-NTR staining can be performed following a standar-
dized immunohistochemistry staining protocol.

4 Notes

1. When you generate liver fibrosis using 10% CCl4, mice should
be monitored carefully. It is highly possible to result in mice’s
death due to overdose injection. You can inject 10% CCl4 two
times per week as shown in Table 3.
2. Conversed iHeps show similar volume to hepatocytes; you
should use 100-μm nylon filter after isolating them from livers.
And iHeps can fall down to the bottom of Falcon tubes in
10 min. Thus, when you perform FACS sorting, invert the
tube gently every 10 min, and use 100-μm nozzle to prevent
cells from blocking the machine.
3. We performed FACS sorting; tdTomato and GFP double-
positive cells could be found in the cells isolated from lineage-
tracing mice. We confirmed that all hepatocytes (and rare
double-positive cells) isolated from control mice are tdTomato
positive. Since LratCre-mTmG model allows specific labeling
of hepatic stellate cells and myofibroblasts with GFP, we believe
that the double-positive cells in a LratCre-mT/mG model may
represent doublets of hepatocytes and myofibroblasts.
4. For the cytochrome P450 enzymes’ activities assay, the
medium should be change every day with inducers. For
114 Guangqi Song et al.

Table 3
Dose of CCl4

Mouse weight 10% CCl4 vol. Mouse weight 10% CCl4 vol. Mouse weight 10% CCl4 vol.
(g) (μL) (g) (μL) (g) (μL)
14 56 23 92 32 128
15 60 24 96 33 132
16 64 25 100 34 136
17 68 26 104 35 140
18 72 27 108 36 144
19 76 28 112 37 148
20 80 29 116 38 152
21 84 30 120 39 156
22 88 31 124 40 160

CYP1A2/Luciferin-1A2 assay, replace HCM medium by

Krebs-Henseleit buffer for incubation of substrate. And, it is
better to perform the measurement immediately after incuba-
tion with substrate. Do not store the supernatant in 20  C or
80  C before measurement.
5. Oil red O working solution is stable for no longer than 2 h. You
can prepare the stock solution by dissolving 300 mg of Oil red
O powder in 100-mL 99% isopropanol. This stock solution is
stable for 1 year in the dark with 4  C temperature. Prepare
work solution by mixing 30 mL of stock solution with 20-mL
deionized water. Filter the working solution completely
through a filter funnel before use.

References
1. Huang PY, He ZY, Ji SY, Sun HW, Xiang D, human fibroblasts to functional melanocytes by
Liu CC, Hu YP, Wang X, Hui LJ (2011) defined factors. Nat Commun 5:5807
Induction of functional hepatocyte-like cells 5. Song G, Pacher M, Balakrishnan A, Yuan Q,
from mouse fibroblasts by defined factors. Tsay HC, Yang D, Reetz J, Brandes S, Dai Z,
Nature 475:386–U142 Putzer BM et al (2016) Direct reprogramming
2. Huang P, Zhang L, Gao Y, He Z, Yao D, Wu Z, of hepatic myofibroblasts into hepatocytes
Cen J, Chen X, Liu C, Hu Y et al (2014) Direct in vivo attenuates liver fibrosis. Cell Stem Cell
reprogramming of human fibroblasts to func- 18:797–808
tional and expandable hepatocytes. Cell Stem 6. Reetz J, Genz B, Meier C, Kowtharapu BS,
Cell 14:370–384 Timm F, Vollmar B, Herchenroder O,
3. Sekiya S, Suzuki A (2011) Direct conversion of Abshagen K, Putzer BM (2013) Development
mouse fibroblasts to hepatocyte-like cells by of adenoviral delivery systems to target hepatic
defined factors. Nature 475:390–393 stellate cells in vivo. PLoS One 8:e67091
4. Yang R, Zheng Y, Li L, Liu S, Burrows M, 7. Mederacke I, Hsu CC, Troeger JS,
Wei Z, Nace A, Herlyn M, Cui R, Guo W Huebener P, Mu X, Dapito DH, Pradere JP,
et al (2014) Direct conversion of mouse and Schwabe RF (2013) Fate tracing reveals hepatic
 Conversion of Fibroblasts to Hepatocyte-Like Cells In Vivo 115

stellate cells as dominant contributors to liver MicroRNA-221 overexpression accelerates

fibrosis independent of its aetiology. Nat Com- hepatocyte proliferation during liver regenera-
mun 4:2823 tion. Hepatology 57:299–310
8. Armendariz-Borunda J, Bastidas-Ramirez BE, 10. Bartneck M, Warzecha KT, Tag CG, Sauer-
Sandoval-Rodriguez A, Gonzalez-Cuevas J, Lehnen S, Heymann F, Trautwein C,
Gomez-Meda B, Garcia-Banuelos J (2011) Weiskirchen R, Tacke F (2015) Isolation and
Production of first generation adenoviral vec- time lapse microscopy of highly pure hepatic
tors for preclinical protocols: amplification, stellate cells. Anal Cell Pathol (Amst)
purification and functional titration. J Biosci 2015:417023
Bioeng 112:415–421 11. Weiskirchen R, Gressner AM (2005) Isolation
9. Yuan Q, Loya K, Rani B, Mobus S, and culture of hepatic stellate cells. Methods
Balakrishnan A, Lamle J, Cathomen T, Mol Med 117:99–113
Vogel A, Manns MP, Ott M et al (2013)
 Chapter 11

Chemically Induced Liver Progenitors (CLiPs): A Novel Cell

Source for Hepatocytes and Biliary Epithelial Cells
Takeshi Katsuda and Takahiro Ochiya

Abstract
Bipotent liver progenitor cells (LPCs) are promising cell sources for cell transplantation therapy in hepatic
disorders as well as biliary dysfunctions. Using a cocktail of small molecules, we recently reported a novel
approach to generate bipotent LPCs, named chemically induced liver progenitors (CLiPs), from adult rat
hepatocytes. In this chapter, we describe a detailed protocol for the induction of rat CLiPs. We first describe
the method to isolate primary rat hepatocytes and then describe how to induce CLiPs from the hepatocytes.
In addition, we describe methods to induce the generated CLiPs to differentiate into hepatocytes and
biliary epithelial cells.

Key words Hepatocyte, Biliary epithelial cell, Liver progenitor cell, CLiP, Reprogramming,
Bipotentiality

1 Introduction

The only current treatment for end-stage liver diseases is liver

transplantation, but its application is limited due to donor
shortages. Thus, there is a strong demand for a novel cell source
to realize regenerative medicine for liver diseases. Accordingly,
researchers have proposed various approaches to supply potential
hepatic cell sources, using embryonic stem cells [1], fetal liver cells
[2], induced pluripotent stem cells [3, 4], iHep cells [5, 6], and
adult liver progenitor cells (LPCs) [7, 8]. These advances have
attracted much attention, but there are still hurdles to be overcome,
such as ethical issues involved with fetal tissue, relatively immature
functionality of the generated hepatic cells compared to mature
hepatocytes (MHs), and poor availability of the adult progenitor
cells. Therefore, further exploration for novel hepatic cell sources is
required to realize liver regenerative medicine.
We recently proposed a novel type of LPCs that can stably
expand in vitro and exhibit bipotentiality [9]. Using a cocktail of
three small molecules Y-27632, A-83-01, and CHIR99021

Naoki Tanimizu (ed.), Hepatic Stem Cells: Methods and Protocols, Methods in Molecular Biology, vol. 1905,
https://doi.org/10.1007/978-1-4939-8961-4_11, © Springer Science+Business Media, LLC, part of Springer Nature 2019

117
118 Takeshi Katsuda and Takahiro Ochiya

(termed YAC), this approach allows reprogramming of rat MHs

into bipotent LPCs without any genetic modification. Thus, we
named these cells chemically induced liver progenitors (CLiPs). Rat
CLiPs can efficiently repopulate injured liver of chronic hepatitis
animal models. Moreover, transplanted CLiPs, at least partly, con-
tribute to ductal structure formation, implying their potential
applicability also to regeneration of the biliary system [9]. Here,
we describe a detailed protocol to generate CLiPs from primary rat
MHs and to evaluate the bipotentiality of the generated CLiPs.

2 Materials

2.1 Animal 1. Wistar rats aged at 5 weeks to 20 weeks (both male and female
and Reagents animals can be used).
2. Pre-perfusion buffer: Dissolve 40 g NaCl, 2.0 g KCl, 0.39 g
NaH2PO4·2H2O, 0.76 g Na2HPO4·12H2O, 4.5 g glucose,
1.9 g EGTA, 3.7 g EDTA, 0.0030 g phenol red, and 11.9 g
HEPES in 500 mL distilled water to prepare 10 stock solu-
tion. Dilute 50 mL of 10 stock solution with 450 mL distilled
water. After autoclave sterilization, add 2.35 mL of 7.5%
NaHCO3 solution and 2 mL of 1 N NaOH solution.
3. 7.5% NaHCO3 solution: Dissolve 7.5 g of NaHCO3 in 100 mL
dH2O. Filter with a 0.22 μm filter, and store at 4  C until use.
Use within 1 month.
4. 0.05% collagenase solution: Dissolve 9.6 g NaCl, 0.48 g KCl,
0.094 g NaH2PO4·2H2O, 0.18 g Na2HPO4·12H2O, 2.86 g
HEPES, 0.888 g CaCl2·2H2O, and 0.42 g NaHCO3 in 1.2 L
of dH2O supplemented with 14.4 mL of 0.05% phenol red
solution. Store this basal solution at 4  C until use. For perfu-
sion, pour 400 mL into a beaker and adjust pH to 7–7.4 with
5 N NaOH and/or 6 N HCl. Then, dissolve 0.02 g trypsin
inhibitor (Sigma-Aldrich) and 0.2 g collagenase (Wako). After
adjusting the pH to 7.6 with 5 N NaOH and/or 6 N HCl,
filter with a 0.22 μm filter, and store at 4  C until use. The
prepared collagenase solution should be used within 1 week.
5. Hepatocyte wash medium: E-MEM.
6. Percoll (GE Healthcare).
7. L-15-based medium for complete Percoll solution: Dissolve
the following reagents in 99 mL of Leibovitz’s L-15 medium
(Gibco, with additives: Sodium pyruvate; L-Glutamine; Phenol
red): 0.2 g bovine serum albumin (BSA) (Sigma-Aldrich) and
0.0429 g HEPES (Sigma-Aldrich). Filter with a 0.22 μm filter,
add 1 mL of antibiotic/antimycotic (Gibco), and store at 4  C
until use.
8. 0.4% trypan blue solution.
 Generation of Chemically Induced Liver Progenitors 119

9. Complete Percoll solution: Prepare complete Percoll solution

in a 50 mL conical tube just before use by mixing 25 mL of
L-15-based medium, 2.4 mL of 10 HBSS (Gibco), and
21.6 mL of Percoll solution.
10. 2 M HEPES solution: Dissolve 47.7 g HEPES in 100 mL
dH2O. Filter with a 0.22 μm filter, and store at 4  C until use.
11. L-proline stock solution: Dissolve 0.3 g of L-proline (Sigma-
Aldrich) in 10 mL of PBS( ). Filter with a 0.22 μm filter, and
store at 4  C until use.
12. 5% BSA stock solution: Dissolve 5 g BSA in 100 mL of PBS( ).
Filter with a 0.22 μm filter. Protect from light and store at 4  C
until use.
13. 10 μg/mL EGF stock solution: Prepare 20 mL of 0.1% BSA
solution by diluting 0.4 mL of 5% BSA with 19.6 mL PBS( ),
and dissolve 0.2 mg EGF (Sigma-Aldrich). Store 500 μL ali-
quots at 20  C until use.
14. 10 4 M dexamethasone: Dissolve 100 mg dexamethasone in
25.4 mL ethanol to make 10 2 M solution and store 500 μL
aliquots at 20  C until use. Dilute 500 μL of 10 2 M solution
with 49.5 mL dH2O. Store the stock solution at 4  C in
the dark.
15. 1 M nicotinamide: Dissolve 12.21 g nicotinamide (Sigma-
Aldrich) in 100 mL PBS( ). Filter with a 0.22 μm filter, and
store at 4  C until use.
16. 100 mM Asc2P stock solution: Dissolve 2.9 g of Asc2P (Wako)
in 100 mL PBS( ). Filter with a 0.22 μm filter. Protect from
light and store at 4  C until use.
17. 5 mM Y-27632: Dissolve 25 mg of Y-27632 (Wako) in
14.8 mL sterilized dH2O. Store 1 mL aliquots at 20  C
until use.
18. 2.5 mM A-83-01: Dissolve 10 mg of A-83-01 (Wako) in
949 μL DMSO to make 25 mM solution and make 100 μL
aliquots for long-term preservation at 20  C. Add 900 μL
DMSO to one vial of 25 mM solution, and store 100 μL
aliquots at 20  C.
19. 15 mM CHIR99021: Add 3.58 mL DMSO into the vial con-
taining 25 mg CHIR99021 (Axon Medchem), and store
100 μL aliquots at 20  C.
20. Small hepatocyte medium (SHM): Remove 27 mL from a
500 mL bottle of DMEM/F12 (with 2.4 g/L NaHCO3 and
L-glutamine, Life Technologies, #11320–082)). The removed
DMEM/F12 can be used to prepare Matrigel-coated plates
(see below). To the remaining DMEM/F12, add 1.25 mL of
2 M HEPES, 500 μL L-proline, and 5 mL antibiotic/
120 Takeshi Katsuda and Takahiro Ochiya

antimycotic. Then, add 250 μL of 5 N NaOH to adjust pH to

approximately 7.5. Then, add 5 mL of 5% BSA, 500 μL of
10 μg/mL of EGF, 5 mL of ITS-X, 500 μL of 10 4 M dexa-
methasone, 5 mL of 1 M nicotinamide, and 5 mL of 100 mM
Asc2P. Store at 4  C.
21. 10 μg/mL mouse oncostatin M (OSM) stock solution: Prepare
2.5 mL of 0.1% BSA/PBS( ) by diluting 50 μL of 5% BSA
with 2.45 mL PBS( ). Dissolve 25 μg mouse OsM (R and D
Systems) in 2.5 mL of 0.1% BSA solution. Store 100 μL ali-
quots at 20  C until use.
22. Matrigel-coated plate: Dilute Matrigel (Corning) at a ratio of
16 μL per mL of DMEM/F12. Add approximately
0.1–0.2 mL/cm2 of the diluted Matrigel solution. Incubate
at 37  C for 20 min. Wash once with an equal volume of SHM
( ) before seeding cells.
23. Hepatic induction medium for Step 1 (HIM-1): Add 100 μL of
10 μg/mL OsM and 500 μL of 10 4 M dexamethasone into
50 mL of SHM + YAC. Final concentrations of OsM and
dexamethasone are 20 ng/mL and 10 6 M, respectively.
24. Hepatic induction medium for Step 2 (HIM-2): Prepare this
medium on ice immediately before use. Mix HIM-1 and Matri-
gel at 7:1 volume ratio. For example, mix 62.5 μL Matrigel and
437.5 μL HIM-1 (500 μL in total) and 125 μL Matrigel and
875 μL HIM-1 (1 mL in total) for one well of a 24- or 12-well
plate, respectively. Use P1000 tips or wide-bore P200 tips
chilled at 4  C or on ice when working with Matrigel.
25. DMEM, high glucose, pyruvate.
26. Cryopreserved MEFs (EmbryoMax® Primary Mouse Embryo
Fibroblasts) (Millipore, catalog number: PMEF-CF).
27. BEC induction medium for Step 1 (BIM-1): BIM-1 is
mTeSR™1 medium (STEMCELL Technologies, catalog
number: ST-05850) supplemented with YAC. Prepare com-
plete mTeSR™1 medium by adding the 100 mL of supple-
ment into 400 mL basal medium, and add each vial of YAC
stock to the solution.
28. BEC induction medium for Step 2 (BIM-2): BIM-2 is BIM-1
supplemented with 2% Matrigel. Prepare this medium imme-
diately before use. Add 1 volume of Matrigel to 49 volumes of
BIM-1.

2.2 Equipment 1. Peristaltic pump (EYELA, Tokyo, Japan, catalog number:

RP-1000).
2. Silicon tube (Ø4.76  7.94 mm2) (EYELA, catalog number:
125540).
 Generation of Chemically Induced Liver Progenitors 121

3. Cannula Ø1.2 mm (plastic outer needle equipped with intra-

venous cannula) (Top corporation, Tokyo, Japan, catalog num-
ber: V1).
4. 4–0 sterilized braid silk suture (Akiyama Medical Mfg,
Co. Ltd., catalog number: DEWB0404).
5. Stainless 60 μm cell strainer (Ikemoto Scientific Technology,
Tokyo, Japan, catalog number: 802–570-01).

3 Methods

3.1 Isolation of Rat Rat primary MHs are isolated using the two-step collagenase per-
Primary MHs fusion method [10]. In the first step, blood is eliminated from the
liver by perfusing the liver with pre-perfusion buffer. In the second
step, the liver is digested by perfusing collagenase solution. Then,
the liver is extracted, minced, and further digested ex vivo with the
remaining collagenase solution. MHs are collected by sequential
low-speed centrifugation at 57  g (see Note 1). This sequential
low-speed centrifugation allows exclusion of nonparenchymal cells.
1. Before starting, place 20 mL E-MEM at room temperature,
which will be used for additional digestion of the extracted
liver.
2. Anesthetize a rat by inhalation of isoflurane vapor. Place a paper
towel cushion on a foam board or cork board, and lay the rat on
the cushion. Sterilize the animal by wiping its fur with 70%
EtOH. Open the abdomen using surgical scissors to expose the
liver.
3. Start drip of pre-perfusion buffer at 25–30 mL/min. Half-cut
the portal vein at the location approximately 1.5 cm from the
bifurcation of the portal vein, and insert the cannula.
4. Immediately after confirming that the color of the liver changes
from reddish brown to yellowish, cut the inferior vena cava.
Perfuse approximately 450 mL pre-perfusion buffer. It takes
approximately 15 min.
5. Pause the pump, transfer the inlet tube to the collagenase
bottle, and resume perfusion. Perfuse approximately
300–350 mL collagenase solution. It takes approximately
10 min.
6. Stop perfusion, and remove the cannula from the portal vein.
Cut the liver from the abdominal cavity, and transfer it to a
100 mm sterile plastic dish. Remove the attaching surrounding
tissue, such as the diaphragm (see Note 2).
7. Add 20 mL E-MEM kept at room temperature and 20 mL of
the remaining collagenase solution. Mince the liver with
122 Takeshi Katsuda and Takahiro Ochiya

surgical scissors into pieces that are smaller than approximately

3 mm in size.
8. Incubate the minced tissue at 37  C in a CO2 incubator for
15 min. After taking out the dish from the incubator, further
dissociate the remaining tissue by gently pipetting the digested
solution with a wide-bore pipette, such as a 25 mL pipette,
several times (see Notes 3 and 4).
9. Remove the undigested tissue fragments by transferring the
digested solution into a 100 mL tube equipped with single-
layered gauze on the top. Discard the single-layered gauze
from the tube top, and transfer the filtered solution into a
100 mL tube with the double-layered gauze (see Notes 5
and 6).
10. Aliquot the filtered solution into two 50 mL tubes. Add
E-MEM into each tube to 50 mL by decantation, and mix by
gentle inversion. Centrifuge the tubes at 57  g for 1 min at
4  C.
11. Aspirate the supernatant, and loosen the cell pellets by gently
tapping the tubes. Add 10 mL E-MEM to each tube, and
resuspend the cell pellet by gently pipetting several times.
Decant E-MEM into each tube to 50 mL, and mix by gentle
inversion. Filter the cell suspension through a double-meshed
stainless cell strainer into new tubes. Centrifuge the tubes at
57  g for 1 min at 4  C.
12. Aspirate the supernatant, and loosen the cell pellet by gently
tapping the tubes. Add 10 mL complete Percoll solution to
each tube, and resuspend the cell pellet by gently pipetting
several times until the cells are suspended homogeneously.
Aliquot the remaining complete Percoll solution to each tube
(approximately 14 mL/tube), and mix by gentle inversion.
Centrifuge the tubes at 57  g for 10 min at 4  C.
13. Aspirate the supernatant, and loosen the cell pellet by gently
tapping the tubes. Add 10 mL E-MEM to each tube, and
resuspend the cell pellet by gently pipetting several times.
Decant E-MEM into each tube to 50 mL, and mix by gentle
inversion. Centrifuge the tubes at 57  g for 2 min at 4  C.
14. Repeat the above wash process (Step 13) once more.
15. Resuspend the cells in 15 mL SHM( ), combine the two
tubes.
16. (Optional) If extensive aggregation disturbs cell counting, fil-
ter the cell suspension with a 40 μm cell strainer before cell
counting. There is no difference in reprogramming efficiency
under microscopic observation between cells isolated with and
without this filtration step.
 Generation of Chemically Induced Liver Progenitors 123

17. Mix 30 μL of cell suspension with 120 μL SHM to prepare a

fivefold-diluted cell suspension. Remove 30 μL from the
diluted cell suspension, and mix it with 30 μL of trypan blue.
Count the viable and dead cell numbers using a hemocytome-
ter. Determine the suspension volume to be used for culture.
18. Seed the isolated MHs at approximately 1  105 cells/cm2. To
obtain homogeneous seeding of MHs, shake the plates vigor-
ously back and forth ten times and then right and left ten times
(see Note 7). Let the plates stand still for 15 min so that MHs
fall down to the bottom of the plate, and then, transfer the
plates into a CO2 incubator.

3.2 Induction Rat CLiPs can be induced from primary MHs upon stimulation
of Primary CLiPs Using with a combination of three small molecules, YAC. An evident
the Small Molecule morphological change compared with YAC-free (YAC( )) culture
Cocktail YAC occurs approximately 4–5 days after plating. As previously reported,
rodent MHs can divide several times in vitro even without YAC, but
their proliferation completely stops thereafter [11, 12]. In contrast,
MHs can continuously proliferate to produce CLiPs during 2-week
culture (Fig. 1). When CLiPs reach 70–100% confluency, harvest
CLiPs to perform hepatic induction or biliary induction, establish
stable CLiPs, or prepare frozen stocks for future use.
1. One day after plating primary MHs, replace the culture
medium with fresh media. Thereafter, renew the medium
every 2–3 days. Given that proliferative YAC(+) cells are bipo-
tential (see the following sections), we designate these cells as
chemically induced liver progenitors (CLiPs).
2. Wash plates with PBS( ) twice.
3. Add 0.05% trypsin-EDTA or TrypLE Express (these two
reagents can be used equivalently for harvesting CLiPs), and
incubate at 37  C in a CO2 incubator until the cells are
detached from the plate (see Notes 8 and 9).
4. Further dissociate the cells physically by pipetting with a P1000
tip to harvest them as completely as possible, and transfer them
to a 15 mL or 50 mL collection tube (see Note 10).
5. Collect the cells remaining in the plate with 5 mL SHM + YAC
supplemented with 5% FBS or SHM + YAC for trypsin-EDTA
or TrypLE Express (TrypLE Express does not require neutrali-
zation of the enzymatic activity).
6. Count the cell number using a hemocytometer, determine the
suspension volume to be used for passage, transfer the required
volume to a new tube, and centrifuge the tube at 200  g for
5 min. Aspirate the supernatant, and loosen the cell pellet by
tapping the tubes. Resuspend the cells with medium to be used
for desired application.
124 Takeshi Katsuda and Takahiro Ochiya

Fig. 1 Morphological change during reprogramming of MHs into CLiPs. MHs actively divide in the presence of
YAC and become smaller in size and exhibit higher nucleus/cytoplasm ratio compared with MHs. This figure is
reproduced from ref. [9]

3.3 Hepatic Hepatic differentiation is performed based on a previously reported

Induction of CLiPs protocol that induced hepatic maturation of fetal liver cells
[13]. This protocol consists of two steps. In the first step, CLiPs
are cultured for 6 days in SHM + YAC supplemented with 20 ng/
mL of OSM and 10 6 M dexamethasone. In the second step, CLiPs
are further cultured for another 2 days in the presence of Matrigel.
Morphological changes in Step 1 are relatively mild. The nucleus/
cytoplasm ratio of hepatic-induced cells (Hep-i(+) cells) is reduced,
and the cytoplasm becomes rich in granules (Fig. 2a). However,
morphological changes in Hep-i(+) cells become more evident in
Step 2. Cell-cell boundaries become clear due to bile canaliculi
formation (Fig. 2a). Hepatic differentiation can be confirmed by
the expression of hepatic marker genes by qRT-PCR (Fig. 2b).
1. Harvest the cells following the steps described in Subheading
3.2, and obtain the required volume of the cell suspension for
hepatic induction. We seed the cells at approximately 5  104
cells/cm2 (1  105 cells/well and 2  105 cells/well for 24-
and 12-well plates, respectively) (see Note 11). Centrifuge the
tube at 200  g for 5 min.
 Generation of Chemically Induced Liver Progenitors 125

A
Hep-i(-) D6 Hep-i(-) D8

D0

Hep-i(+) D6 Hep-i(+) D8
Matrigel
100 mm
overlay
OsM
+ 10-6 M
dexamethasone

B
Alb Ttr G6pc
25 20 200
Relative expression to Actb [a.u.]

20 15 150
15
10 100
10
5 5 50

0 0 0
D0 Hep-i(-) Hep-i(+) D0 Hep-i(-) Hep-i(+) D0 Hep-i(-) Hep-i(+)

Tat Tdo2 Cnx32

7 5 3.5
6 3
4
5 2.5
4 3 2
3 2 1.5
2 1
1
1 0.5
0 0 0
D0 Hep-i(-) Hep-i(+) D0 Hep-i(-) Hep-i(+) D0 Hep-i(-) Hep-i(+)

Fig. 2 Hepatic induction of CLiPs. (a) Morphological change of CLiPs during hepatic induction. Arrowheads
indicate cells that are rich in granules in their cytoplasm. Arrows indicate bile canaliculi formed at the
boundaries between cells that achieve hepatic maturation. (b) qRT-PCR of hepatic marker genes for Hep-i(+),
Hep-i( ), and D0 cells. The data are shown as the mean  S.E.M. of four independent experiments. This
figure is partly reproduced from ref. [9]
126 Takeshi Katsuda and Takahiro Ochiya

2. Aspirate the supernatant, resuspend the cells in 0.5 or 1 mL

SHM + YAC supplemented with 5% FBS, and seed to a well of
collagen-coated 24- or 12-well plates, respectively (see Note
12).
3. One day after seeding, replace the medium with fresh
SHM + YAC, and culture the cells for 2 days.
4. Replace the medium with HIM-1, and continue culturing for
another 6 days. Replace the medium with fresh HIM-1 every
other day.
5. Replace the medium with HIM-2, and continue culturing for
another 2 days (see Notes 13 and 14).

3.4 Biliary Induction CLiPs can be differentiated also into BECs using a two-step proto-
of CLiPs col (Fig. 3a). In Step 1, CLiPs are co-cultured on cell cycle-arrested
mouse embryonic fibroblasts (MEFs) in mTeSR1 + YAC for 6 days.
In Step 2, CLiPs are cultured for another 6 days in 2% Matrigel-
containing mTeSR1 + YAC. CLiP-derived BEC-like cells form
ductal and cystic structures typically observed in Step 2 (Fig. 3b).
CLiP-derived ductal/cystic structures exhibit the capacity to trans-
port water into their luminal space in response to secretin stimula-
tion (Fig. 3c). Following incubation with fluorescein diacetate,
these cells also transport fluorescein to their luminal spaces
(Fig. 3d).
1. One day before harvesting CLiPs, seed cell cycle-arrested
MEFs on 12-well collagen plates at 5  104 cells/well.
2. Harvest the primary CLiPs following the steps described in
Subheading 3.2, and obtain the desired volume of the cell
suspension for biliary induction. We seed the cells at 5  105
cells/well in 12-well plates. Centrifuge the tube at 200  g for
5 min.
3. Aspirate the supernatant, resuspend the cells in 1 mL
SHM + YAC supplemented with 5% FBS, and seed onto the
pre-inoculated MEFs (see Note 12).
4. (Step 1) On the following day, start biliary induction by repla-
cing the medium with BIM-1. Culture the cells for 6 days while
replacing the medium with fresh BIM-1 every other day.
5. (Step 2) Replace the medium with BIM-2. Culture the cells for
another 6 days while replacing the medium with fresh BIM-2
every other day (see Notes 14 and 15).

3.5 Long-Term CLiPs can be cultured stably through continuous passages; how-
Culture of CLiPs ever, the proliferation rates transiently decrease from approximately
Through Continuous passage 2 (P2) to 7 (P7) (see Notes 16 and 17). During long-term
Passages culture, the cells can be frozen for backup or future use (see Sub-
heading 3.6). Primary CLiPs are typically passaged on days 14–16.
 Generation of Chemically Induced Liver Progenitors 127

Fig. 3 Biliary induction of CLiPs. (a) Morphological change of CLiPs during biliary induction. Arrows indicate
ductular structures formed between adjacent cells. (b) Phase-contrast images of BEC-i(+) cells taken before
and 30 min after the addition of secretin. Arrowheads denote enlarged lumens following secretin treatment. (c)
Uptake and excretion of fluorescein diacetate (FD) by the ductular structures of BEC-i(+) cells. Phase-contrast
and fluorescence images were taken immediately after a 15-min incubation in the presence of FD (upper
panel) and after another 30-min incubation in the absence of FD. This figure is partly reproduced from ref. [9]

Subcultured CLiPs are further passaged when they reach 50–90%

confluence between P1 and P2, 30–60% confluence between P3
and P7, and 50–90% confluence thereafter. We typically seed cells at
2  106 viable cells/10 cm dish in earlier passages (P1 to P7).
However, the actual seeded cell number is much lower than
2  106 viable cells/10 cm dish, because the cells in earlier passage
128 Takeshi Katsuda and Takahiro Ochiya

cannot attach efficiently to the plates. Once cell proliferation rates

increase (typically after approximately P5–7), we reduce the seeding
density to 1–10  105 cells/10 cm dish so that they become
subconfluent after 3–7 days of culture.
1. Harvest the cells following the steps described in Subheading
3.2, and obtain the desired volume of the cell suspension for
passage. Centrifuge the tube at 200  g for 5 min.
2. Resuspend the cells in an appropriate volume of SHM + YAC,
seed on a desired plate coated with Matrigel, and culture the
cells in a CO2 incubator. Medium should be replaced every
1–3 days.

3.6 Freeze Stocking 1. Harvest the cells following the steps described in Subheading
of CLiPs 3.2, and obtain a desired volume of the cell suspension for
preparation of frozen stocks (typically 0.5–10  106 cells/
vial). Centrifuge the tube at 200  g for 5 min.
2. Aspirate the supernatant, and resuspend the cells in desired
volume of Cell Banker. The volume of Cell Banker used is
0.5 mL/vial.
3. Aliquot to serum tubes equipped with screw caps. Seal the caps
tightly, and store at 80  C for at least 1 day. Then, transfer
them to liquefied nitrogen.

4 Notes

1. Our centrifugation machine (Kubota 2800) uses 57  g with

the rotation speed at 600 rpm; researchers should adjust their
rotation speed such that they can obtain 50–60  g centrifugal
force.
2. From this step forward, perform all procedures aseptically in a
clean bench or a safety cabinet.
3. If undigested large fragments remain, remove them with ster-
ilized tweezers.
4. When pipetting the MH suspension, a wide-bore pipette
should be used to avoid damaging MHs by extensive shear
stress.
5. When the gauze is clogged with undigested tissue fragments,
remove the clogging materials from the filter using a pipette.
6. From this step forward, all the procedures should be performed
on ice, and all the reagents should be kept at 4  C or on ice.
7. Due to the heavy weight of MHs, be careful to keep the cell
suspension homogeneous. Loading a large amount of the cell
suspension in a pipette and dispensing it to multiple plates/
wells may result in inconsistent seeding density among plates/
 Generation of Chemically Induced Liver Progenitors 129

wells. We typically apply well-mixed undiluted cell suspension

(generally approximately 3–8  106 cells/mL) using a micro-
pipette directly to plates that are already filled with culture
medium. During the cell seeding step, ensure that the cell
suspension is kept homogeneous by periodically inverting
the tube.
8. Make sure that the dissociating reagent completely covers the
culture areas. We typically apply 2 mL of dissociation reagent to
a 100 mm dish and use a similar ratio for other culture scales.
9. CLiPs are relatively resistant to trypsinization, and it typically
takes 10–15 min to detach them from plates.
10. Despite its relatively damaging effect on the cells, this proce-
dure is indispensable. Unlike general cell lines, even if the cells
exhibit round-shaped morphology following trypsinization,
many of them still attach to the plate tightly and can rarely be
harvested only by adding medium onto the plates.
11. The cell density can be reduced to 1–2.5  104 cells/cm2 when
stable CLiPs are used for hepatic induction.
12. To increase the plating efficiency of primary CLiPs on culture
plates, we recommend to use 5% FBS-supplemented
SHM + YAC.
13. Prepare HIM-2 immediately prior to use.
14. Use P1000 or wide-bore 200 μL tips chilled at 4  C or on ice.
15. Prepare BIM-2 immediately prior to use.
16. CLiPs undergo a transient decrease in their proliferative capac-
ity between P2 and P7 (if not they enter complete senescence).
Thus, during this period, it takes 15–40 days until the cells
obtain 30–60% confluence, depending on experimental
batches.
17. Prepare a Matrigel-coated plate before starting passage
(we typically use 35 mm, 60 mm, and 10 cm dishes for cell
maintenance).

Acknowledgments

This work was supported in part by Grant-in-Aid for Young Scien-

tists B (16 K16643) and the Research Program on Hepatitis
(16fk0310505h0005) from the Japan Agency for Medical Research
and Development (AMED).
130 Takeshi Katsuda and Takahiro Ochiya

References
1. Woo DH, Kim SK, Lim HJ et al (2012) Direct 494:247–250. https://doi.org/10.1038/
and indirect contribution of human embryonic nature11826
stem cell-derived hepatocyte-like cells to liver 8. Lu W-Y, Bird TG, Boulter L et al (2015)
repair in mice. Gastroenterology Hepatic progenitor cells of biliary origin with
142:602–611. https://doi.org/10.1053/j. liver repopulation capacity. Nat Cell Biol
gastro.2011.11.0301 17:971–983. https://doi.org/10.1038/
2. Oertel M, Menthena A, Chen Y-Q et al (2008) ncb3203
Purification of fetal liver stem/progenitor cells 9. Katsuda T, Kawamata M, Hagiwara K et al
containing all the repopulation potential for (2017) Conversion of terminally committed
normal adult rat liver. Gastroenterology hepatocytes to culturable bipotent progenitor
134:823–832. https://doi.org/10.1053/j. cells with regenerative capacity. Cell Stem Cell
gastro.2008.01.007 20:41–55. https://doi.org/10.1016/j.stem.
3. Liu H, Kim Y, Sharkis S et al (2011) In vivo 2016.10.007
liver regeneration potential of human induced 10. Seglen PO (1976) Preparation of isolated rat
pluripotent stem cells from diverse origins. Sci liver cells. Methods Cell Biol 13:29–83.
Transl Med 3:82ra39. https://doi.org/10. https://doi.org/10.1016/S0091-679X(08)
1126/scitranslmed.3002376 61797-5
4. Zhu S, Rezvani M, Harbell J et al (2014) 11. Loyer P, Cariou S, Glaise D et al (1996)
Mouse liver repopulation with hepatocytes Growth factor dependence of progression
generated from human fibroblasts. Nature through G1 and S phases of adult rat hepato-
508:93–97. https://doi.org/10.1038/ cytes in vitro. Evidence of a mitogen restriction
nature13020 point in mid-late G1. J Biol Chem
5. Sekiya S, Suzuki A (2011) Direct conversion of 271:11484–11492
mouse fibroblasts to hepatocyte-like cells by 12. Frémin C, Bessard A, Ezan F et al (2009) Mul-
defined factors. Nature 475:390–393. tiple division cycles and long-term survival of
https://doi.org/10.1038/nature10263 hepatocytes are distinctly regulated by extracel-
6. Huang P, He Z, Ji S et al (2011) Induction of lular signal-regulated kinases ERK1 and ERK2.
functional hepatocyte-like cells from mouse Hepatology 49:930–939. https://doi.org/10.
fibroblasts by defined factors. Nature 1002/hep.22730
475:386–389. https://doi.org/10.1038/ 13. Kamiya A, Kojima N, Kinoshita T et al (2002)
nature10116 Maturation of fetal hepatocytes in vitro by
7. Huch M, Dorrell C, Boj SF et al (2013) In extracellular matrices and oncostatin M: induc-
vitro expansion of single Lgr5+ liver stem cells tion of tryptophan oxygenase. Hepatology
induced by Wnt-driven regeneration. Nature 35:1351–1359. https://doi.org/10.1053/
jhep.2002.33331
 Chapter 12

Induction of Functional Hepatocytes from Human iPSCs

Taketomo Kido and Yuta Koui

Abstract
Human-induced pluripotent stem cells (iPSCs) could be a useful source for production of hepatocytes.
Here, we develop protocols to generate iPSC-derived liver progenitor cells, liver sinusoidal endothelial cells
(LSECs), and hepatic stellate cells (HSCs). We also establish long-term two-dimensional co-culture system
to induce functional hepatocytes from iPSC-derived liver cells.

Key words Human-induced pluripotent stem cells, Liver, Liver progenitor cells, Hepatocytes, Liver
sinusoidal endothelial cells, Hepatic stellate cells

1 Introduction

Human-induced pluripotent stem cells (iPSCs) have been gener-

ated from a variety of somatic cells and used as an alternative cell
source for production of different types of cells [1]. Several proto-
cols have been reported for generation of hepatocytes from iPSCs
[2–4]. However, induction of hepatocytes from iPSCs requires
time-consuming multiple processes; also iPSC-derived hepatocytes
exhibit immature phenotypes.
The liver consists of hepatocytes (parenchymal cells) and
non-parenchymal cells, such as liver sinusoidal endothelial cells
(LSECs) and hepatic stellate cells (HSCs). As they establish inter-
cellular communications to make functional liver tissue during
development [5, 6], co-culture of iPSC-derived hepatocytes and
non-parenchymal cells would be useful for production of functional
hepatocytes in vitro.
To establish this co-culture system, we have developed proto-
cols to isolate iPSC-derived liver progenitor cells (LPCs), LSEC
progenitors, and HSC progenitors based on the expression of cell
surface specific markers (Table 1) [7, 8]. iPSC-derived LPCs and
LSEC progenitors could be expanded by several passages and cryo-
preserved. To induce functional hepatocytes from iPSC-derived
LPCs, we have also developed a high-density co-culture system [8].

Naoki Tanimizu (ed.), Hepatic Stem Cells: Methods and Protocols, Methods in Molecular Biology, vol. 1905,
https://doi.org/10.1007/978-1-4939-8961-4_12, © Springer Science+Business Media, LLC, part of Springer Nature 2019

131
132 Taketomo Kido and Yuta Koui

Table 1
Cell surface specific markers on liver cells

iPSC-derived liver cells Specific markers Reference

Liver progenitor cells (LPCs) Carboxypeptidase M (CPM) [7]
Liver sinusoidal endothelial cell (LSEC) CD34/CD31 (PECAM1)/FLK1 (KDR, [8]
progenitors VEGFR2)
Hepatic stellate cell (HSC) progenitors Activate leukocyte cell adhesion molecule [8]
(ALCAM)

2 Materials

1. Cell culture dishes/plates: 6 cm culture dish, 6-well plate,

12-well plate, 24-well plate, 48-well plate, Ultra-Low Attach-
ment 6-well plate (Corning).
2. Accumax Cell Dissociation Solution (Innovative Cell
Technologies).
3. StemPro Accutase Cell Dissociation Reagent (Thermo Fisher
Scientific).
4. FcR Blocking Reagent (Miltenyi Biotec).
5. Cellmatrix Type I-C (Nitta Gelatin).
6. autoMACS Running Buffer (Miltenyi Biotec).
7. Antibodies: Anti-Carboxypeptidase M (CPM) antibody (Med-
ical & Biological Laboratories Co., Ltd.), Purified Mouse IgG1
isotype control (BioLegend), PE Goat anti-mouse IgG anti-
body (BioLegend), FITC conjugated anti-CD34 antibody
(BioLegend), PE conjugated anti-CD31 antibody
(BD Bioscience), APC conjugated anti-CD309 (FLK1) anti-
body (Miltenyi Biotec), FITC mouse IgG1 isotype control
(BD Bioscience), PE mouse IgG1 isotype control (BioLe-
gend), APC mouse IgG1 isotype control (BD Bioscience),
Anti-FITC microbeads (Miltenyi Biotec), Biotin conjugated
anti-CD166 (ALCAM) antibody (Miltenyi Biotec), REA Con-
trol (S)-Biotin (Miltenyi Biotec), Streptavidin-APC
(BD Bioscience).
8. Matrigel solution: 2% Matrigel (Growth Factor Reduced Base-
ment Membrane Matrix, Thermo Fisher Scientific) in RPMI-
1640 medium.
9. RPMI/B27 medium: 1% B-27 Supplement (Thermo Fisher
Scientific) in RPMI-1640 medium.
10. Activin A/CHIR 99021 medium: 100 ng/mL Recombinant
Activin A, 2 μM CHIR 99021 in RPMI/B27 medium.
 Induction of Functional Hepatocytes from Human iPSCs 133

11. Activin A medium: 100 ng/mL Recombinant Activin A in

RPMI/B27 medium.
12. BMP-4/bFGF medium: 20 ng/mL Recombinant Human
BMP-4, 10 ng/mL Recombinant Human FGF-basic (154 a.
a.) in RPMI/B27 medium.
13. HGF medium: 20 ng/mL Recombinant Human HGF in
RPMI/B27 medium.
14. Trypsin/EDTA solution: 0.05% Trypsin and 0.5 mM EDTA
solution in D-PBS ( ).
15. Mouse embryonic fibroblast (MEF) culture medium: 10% fetal
bovine serum (FBS), 1% Penicillin-Streptomycin-Glutamine in
DMEM high glucose.
16. BSA/PBS: 0.3% BSA in D-PBS ( ).
17. Liver progenitor cell (LPC) culture medium: 10% FBS (JRH
Biosciences), 1% Penicillin-Streptomycin-Glutamine (Thermo
Fisher Scientific), 1% Insulin Transferrin, Selenium-X Solution
(Thermo Fisher Scientific), 1% N-2 Supplement (Thermo
Fisher Scientific), 1% MEM Non-Essential Amino Acids Solu-
tion (Thermo Fisher Scientific), 1% L-glutamine (Thermo
Fisher Scientific), 1 mM L-ascorbic acid (Sigma-Aldrich Cor-
poration), 10 mM nicotinamide (Wako Pure Chemical Indus-
tries), 2  10 7 M dexamethasone (Wako Pure Chemical
Industries), 20 ng/mL Recombinant Human HGF (Pepro-
Tech), 10 ng/mL Recombinant Human EGF (PeproTech),
5 μM Y-27632 (Wako Pure Chemical Industries), 5 μM
A83-01 (Tocris) in D-MEM/Ham’s F-12 (Wako Pure Chemi-
cal Industries).
18. EDTA solution: 0.5 mM EDTA in D-PBS ( ).
19. StemPro medium: StemPro-34 SFM (1) (Thermo Fisher
Scientific), StemPro-34 Nutrient Supplement (40) (Thermo
Fisher Scientific).
20. Fibronectin solution: 20 μg/mL human plasma fibronectin in
D-PBS ( ).
21. EC culture medium: EBM-2 Basal Medium (Lonza), EGM-2
SingleQuots (Lonza), 50 ng/mL Recombinant Human VEGF
165 (PeproTech).
22. LSEC culture medium: EBM-2 Basal Medium (Lonza),
EGM-2 SingleQuots (Lonza), 50 ng/mL Recombinant
Human VEGF 165 (PeproTech), 1.5 μM A83-01 (Tocris).
23. HSC culture medium: MSCBM (Lonza), MSCGM Single-
Quots (Lonza), 10 μM Y-27632 (Wako Pure Chemical
Industries).
24. Collagen gel solution: 10% 10  conc. Ham F12 medium
(Nitta Gelatin, KP-7100), 80% Cellmatrix Type I-A (Nitta
134 Taketomo Kido and Yuta Koui

Gelatin, KP-2100), 10% Reconstitution buffer (50 mM

NaOH, 260 mM NaHCO3, 200 mM HEPES) (Nitta Gelatin,
KP-8000).
25. Equipments: Ambient O2/5% CO2 incubator, 4% O2/5% CO2
incubator, autoMACS Pro Separator (Miltenyi Biotec), MoFlo
XDP cell sorter (Beckman Coulter).

3 Methods

3.1 iPSC- 1. Preparation of culture plate: Add 1 mL of Matrigel solution to

Derived LPCs coat each well of a 6-well plate, and incubate the plate for 1 h at
37  C.
3.1.1 Induction of LPCs
from Human iPSCs 2. Preparation of iPSCs: Remove the iPSC culture medium from
each 6 cm culture dish, and wash one time with D-PBS ( ) (see
Note 1).
3. Aspirate D-PBS ( ), add 1 mL of StemPro Accutase Cell
Dissociation Reagent.
4. Incubate cells at 37  C for 5 min. Harvest cells into 15 mL
tube, and add iPSC culture medium for wash.
5. Centrifuge cells for 3 min at 340  g. Resuspend cells in iPSC
culture medium with 10 μM Y-27632.
6. Induction of LPCs from iPSCs: Remove the Matrigel solution
immediately prior to use. Plate the cells at a density of
1–2  104 cells/cm2, and incubate cells overnight at 37  C
with ambient O2/5% CO2 (Day 0) (see Note 2).
7. Replace medium with iPSC culture medium without Y-27632,
and incubate cells overnight at 37  C with ambient O2/5%
CO2 (Day 1) (Fig. 1a).
8. Replace medium with Activin A/CHIR 99021 medium, and
incubate cells for 2 days at 37  C with ambient O2/5% CO2
(Day 2–3). Change medium every day.
9. Replace medium with Activin A medium, and incubate cells for
3 days at 37  C with ambient O2/5% CO2 (Day 4–6) (Fig. 1b).
Change medium every day.
10. Replace medium with BMP-4/bFGF medium, and incubate
cells for 5 days at 37  C with 4% O2/5% CO2 (Day 7–11)
(Fig. 1c). Change medium every day.
11. Replace medium with HGF medium, and incubate cells for
10 days at 37  C with 4% O2/5% CO2 (Day 12–22) (Fig. 1d).
Change medium every day.
 Induction of Functional Hepatocytes from Human iPSCs 135

a b c d

e f

iso CPM

Fig. 1 Generation of LPCs from human iPSCs. (a–d) Morphological changes of iPSCs at different stages of
hepatic differentiation. Scale bar, 100 μm. (e) FCM analysis of CPM expression. (f) Morphology of the CPM+
LPC colonies on MEF feeder cells after 7 days of culture. Scale bar, 100 μm

3.1.2 Isolation and 1. One day before isolation of iPSC-derived LPCs: Plate the
Expansion of iPSC- mitomycin C-treated MEF cells at a density of 5–8  104
Derived LPCs cells/cm2 onto a 0.1% gelatin-coated 12-well plate in MEF
culture medium. Incubate the cells overnight at 37  C with
ambient O2/5% CO2.
2. Remove the HGF medium from each well of 6-well plates, and
wash once with D-PBS ( ).
3. Aspirate D-PBS ( ), add 0.5 mL of Trypsin/EDTA solution.
4. Incubate cells at 37  C for 10–15 min. Harvest cells into 15 mL
tube, and add MEF culture medium for wash.
5. Centrifuge cells for 3 min at 340  g. Resuspend cells in MEF
culture medium, and pass through 70 μm Cell Strainer to
remove the debris.
6. Centrifuge cells for 3 min at 340  g. Resuspend cells in 0.3%
BSA/PBS.
7. Add FcR Blocking Reagent, and incubate the cells for 20 min
on ice.
8. Add anti-Carboxypeptidase M (CPM) antibody, and incubate
the cells for 30 min on ice. Use Purified Mouse IgG1 isotype
control as negative control.
9. Add appropriate volume of 0.3% BSA/PBS for wash.
10. Centrifuge cells for 3 min at 340  g. Resuspend cells in 0.3%
BSA/PBS.
136 Taketomo Kido and Yuta Koui

11. Add PE Goat anti-mouse IgG antibody, and incubate the cells
for 30 min on ice (see Note 3).
12. Wash as in step 9.
13. Add propidium iodide to identify dead cells, and pass through
35 μm Cell Strainer.
14. Isolate iPSC-derived CPM+ LPCs using a MoFlo XDP cell
sorter (Fig. 1e).
15. Centrifuge iPSC-derived CPM+ LPCs for 3 min at 340  g.
Resuspend cells in LPC culture medium.
16. Plate the cells at a density of 1–2  104 cells/cm2 on mitomy-
cin C-treated MEF feeder cells.
17. Incubate cells for 7–10 days at 37  C with ambient O2/5%
CO2. Change medium every day (Fig. 1f) (see Note 4).

3.2 iPSC-Derived 1. Remove the iPSC culture medium from each 6 cm culture dish,
LSECs and wash once with D-PBS ( ).
3.2.1 Induction of LSEC 2. Aspirate D-PBS ( ), add 1 mL of EDTA solution, and incubate
Progenitors from Human cells at 37  C for 5 min.
iPSCs 3. Aspirate EDTA solution, and add 3 mL of iPSC culture
medium. Dissociate cells into small clusters, harvest cells into
15 mL tube, and add iPSC culture medium for wash (see Note
5).
4. Centrifuge cells for 3 min at 340  g. Resuspend cells in
StemPro medium with 10 μM Y-27632 and 2 ng/mL BMP-4.
5. Plate the cell clusters on Ultra-Low Attachment 6-well plates,
and incubate cells overnight at 37  C with 4% O2/5% CO2
(Day 0).
6. Harvest cells into 15 mL tube and centrifuge cells for 2 min at
120  g. Resuspend cells in StemPro medium with 5 ng/mL
Recombinant Activin A, 5 ng/mL Recombinant Human
FGF-basic (154 a.a.), and 30 ng/mL BMP-4 (see Note 6).
7. Replate the cell clusters on Ultra-Low Attachment 6-well
plates, and incubate cells for 3 days at 37  C with 4% O2/5%
CO2 (Day 1–4).
8. Harvest cells into 15 mL tube and centrifuge cells for 2 min at
120  g. Resuspend cells in StemPro medium with 10 ng/mL
Recombinant Human VEGF 165, 5.4 μM SB431542, and
0.5 μM Dorsomorphin dihydrochloride (see Note 6).
9. Replate the cell clusters on Ultra-Low Attachment 6-well
plates, and incubate cells for 2 days at 37  C with 4% O2/5%
CO2 (Day 4–6).
10. Harvest cells into 15 mL tube and centrifuge cells for 2 min at
120  g. Resuspend cells in EC culture medium.
 Induction of Functional Hepatocytes from Human iPSCs 137

11. Replate the cell clusters at a 1:2 split ratio onto a 0.1% gelatin-
coated 6-well plate.
12. Incubate the cells for 7 days at 37  C with 4% O2/5% CO2.
Change medium every other day (see Note 7).

3.2.2 Isolation and 1. Remove the EC culture medium from each well of 6-well
Expansion of iPSC-Derived plates, and wash once with D-PBS ( ).
LSEC Progenitors 2. Aspirate D-PBS ( ), and add 0.5 mL of Trypsin/EDTA
solution.
3. Incubate cells at 37  C for 10 min. Harvest cells into 15 mL
tube, and add MEF culture medium for wash.
4. Centrifuge cells for 3 min at 340  g. Resuspend cells in 0.3%
BSA/PBS, and pass through 70 μm Cell Strainer to remove the
debris.
5. Centrifuge cells for 3 min at 340  g. Resuspend cells in 0.3%
BSA/PBS.
6. Add FcR Blocking Reagent, and incubate the cells for 20 min
on ice.
7. Add FITC conjugated anti-CD34 antibody, PE conjugated
anti-CD31 antibody, and APC conjugated anti-CD309
(FLK1) antibody, and incubate the cells for 30 min on ice.
Use FITC mouse IgG1 isotype control, PE mouse IgG1 iso-
type control, and APC mouse IgG1 isotype control as negative
control.
8. Add appropriate volume of 0.3% BSA/PBS for wash.
9. Centrifuge cells for 3 min at 340  g. Resuspend cells in
autoMACS Running Buffer.
10. Add Anti-FITC microbeads and incubate the cells for 20 min
on ice.
11. Add appropriate volume of autoMACS Running Buffer
for wash.
12. Centrifuge cells for 3 min at 340  g. Resuspend cells in
autoMACS Running Buffer.
13. Enrich FITC+ cells using autoMACS Pro Separator (Miltenyi
Biotec), and centrifuge cells for 3 min at 340  g. Resuspend
cells in 0.3% BSA/PBS.
14. Add propidium iodide to identify dead cells, and pass through
35 μm Cell Strainer.
15. Isolate iPSC-derived CD34+CD31+FLK1+ LSEC progenitors
using a MoFlo XDP cell sorter (Fig. 2).
16. Centrifuge iPSC-derived CD34+CD31+FLK1+ LSEC progeni-
tors for 8 min at 340  g. Resuspend cells in EC culture
medium.
138 Taketomo Kido and Yuta Koui

CD34+ gated

FLK1
5 5
10 10

4 4
10 10

3 3
10 10

2 2
10 10

1 1
10 10

100 100
0 1 2 3 4 5 0 1 2 3 4 5
10 10 10 10 10 10 10 10 10 10 10 10

CD34 CD31

Fig. 2 Generation of LSEC progenitors from human iPSCs. FCM analysis of iPSC-derived LSEC progenitors after
enrichment of CD34+ cells using autoMACS Pro Separator. CD34+ cells were identified (left), and CD31+FLK1+
cells were identified in CD34+ cell fraction

17. Plate the cells at a density of 2  104 cells/cm2 on 12-well

plates coated with Fibronectin solution.
18. Incubate cells for 7 days at 37  C with 4% O2/5% CO2. Change
medium every other day (see Note 8).

3.2.3 Induction of iPSC- 1. Remove the EC culture medium from each well of 12-well
Derived Mature LSECs from plates, and wash once with D-PBS ( ).
LSEC Progenitors 2. Aspirate D-PBS ( ), and add 0.5 mL of Trypsin/EDTA
solution.
3. Incubate cells at 37  C for 10 min. Harvest cells into 15 mL
tube, and add MEF culture medium for wash.
4. Centrifuge cells for 3 min at 340  g. Resuspend cells in LSEC
culture medium.
5. Plate the cells at a density of 1.5  104 cells/cm2 on 12-well
plates coated with Fibronectin solution.
6. Incubate cells for 7 days at 37  C with 4% O2/5% CO2. Change
medium every other day.
7. Passage as in Subheading 3.2.3, steps 1–5.
8. Incubate cells for 7 days at 37  C with 4% O2/5% CO2. Change
medium every other day (see Note 9).
 Induction of Functional Hepatocytes from Human iPSCs 139

3.3 iPSC- Transfer the iPSCs to HSC progenitor differentiation culture as in

Derived HSCs Subheading 3.2.1, steps 1–9.
3.3.1 Induction of HSC
Progenitors from Human
iPSCs

3.3.2 Isolation of HSC 1. Harvest cells into 15 mL tube and centrifuge cells for 3 min at
Progenitors and Induction 340  g.
of iPSC-Derived 2. Aspirate the medium and resuspend cells in Accumax Cell
Mature HSCs Dissociation Solution.
3. Incubate cells at 37  C water bath for 10 min.
4. Add MEF culture medium to inactivate dissociation solution.
5. Centrifuge cells for 3 min at 340  g. Resuspend cells in 0.3%
BSA/PBS, and pass through 70 μm Cell Strainer to remove the
debris.
6. Centrifuge cells for 3 min at 340  g. Resuspend cells in 0.3%
BSA/PBS.
7. Add FcR Blocking Reagent, and incubate the cells for 20 min
on ice.
8. Add Biotin conjugated anti-CD166 (ALCAM) antibody, and
incubate the cells for 30 min on ice. Use REA Control (S)-
Biotin as negative control.
9. Add appropriate volume of 0.3% BSA/PBS for wash.
10. Centrifuge cells for 3 min at 340  g. Resuspend cells in 0.3%
BSA/PBS.
11. Add Streptavidin-APC, and incubate the cells for 20 min
on ice.
12. Wash as in step 9.
13. Add propidium iodide to identify dead cells, and pass through
35 μm Cell Strainer.
14. Isolate iPSC-derived ALCAMhigh HSC progenitors using a
MoFlo XDP cell sorter (Fig. 3).
15. Centrifuge iPSC-derived ALCAMhigh HSC progenitors for
3 min at 340  g. Resuspend cells in HSC culture medium.
16. Plate the cells at a density of 1.5  104 cells/cm2 on Cellmatrix
Type I-C-coated 24-well plates.
17. Incubate cells for 5 days at 37  C with ambient O2/5% CO2.
Change medium every other day (see Note 10).
140 Taketomo Kido and Yuta Koui

5 5
10 10

4 4
10 10

3 3
10 10

2 2
10 10

1 1
10 10

0 0.26 0 36.4
10 10
0 1 2 3 4 5 0 1 2 3 4 5
10 10 10 10 10 10 10 10 10 10 10 10
iso ALCAM

Fig. 3 Generation of HSC progenitors from human iPSCs. FCM analysis of ALCAM expression in iPSC-derived
mesodermal cells (right). The positive gate was defined by the isotype control (left)

3.4 High-Density 1. Remove the medium from LPC culture plates (from Subhead-
Co-culture System by ing 3.1.2, step 17), and wash once with D-PBS ( ).
iPSC-Derived LPCs, 2. Aspirate D-PBS ( ), add 0.5 mL of Trypsin/EDTA solution.
LSECs, and HSCs
3. Incubate cells at 37  C for 10–15 min. Harvest cells into 15 mL
3.4.1 Preparation of tube, and add MEF culture medium for wash.
iPSC-Derived LPCs 4. Centrifuge cells for 3 min at 340  g. Resuspend cells in MEF
culture medium, and pass through 70 μm Cell Strainer to
remove the debris.
5. Centrifuge cells for 3 min at 340  g. Resuspend cells in MEF
culture medium.
6. Plate the cells onto a 0.1% gelatin-coated 10 cm culture dish.
Incubate cells for 30 min to remove adherent MEF cells (see
Note 11).
7. Harvest non-adherent cells (iPSC-derived LPCs) into 15 mL
tube and centrifuge for 3 min at 340  g. Resuspend cells in
LPC culture medium.

3.4.2 Preparation of 1. Remove the medium from each culture plate (Subheading
iPSC-Derived LSECs 3.2.3, step 8, and Subheading 3.3.2, step 17), and wash
and HSCs once with D-PBS ( ).
2. Aspirate D-PBS ( ), and add 0.5 mL of Trypsin/EDTA
solution.
3. Incubate cells at 37  C for 5–10 min. Harvest cells into 15 mL
tube, and add MEF culture medium for wash.
4. Centrifuge cells for 3 min at 340  g. Resuspend cells in LPC
culture medium.
 Induction of Functional Hepatocytes from Human iPSCs 141

3.4.3 Co-culture of iPSC- 1. Preparation of collagen gel-coated plates: Add 0.2 mL of colla-
Derived LPCs, LSECs, gen gel solution to coat each well of a 48-well plate, and
and HSCs incubate the plate for 30 min at 37  C.
2. Plate iPSC-derived LPCs (2  105 cells/well), iPSC-derived
LSECs (2  104 cells/well), and iPSC-derived HSCs (2  104
cells/well) on collagen gel-coated 48-well plates.
3. Incubate cells for 10 weeks at 37  C with ambient O2/5% CO2.
Change medium every other day (see Note 12).

4 Notes

1. When iPSCs reached subconfluence, they can be used for

induction of LPCs.
2. The cell density depends on the growth rate and varies between
the iPSC lines.
3. It is possible to choose different fluorescent-dye conjugated
secondary antibodies.
4. The expanded cells could be cryopreserved for further study.
5. Pipette slowly to prevent excessive dissociation of iPSC clus-
ters. Excessive dissociation results in decrease of differentiation
efficiency.
6. Pipette slowly because the cell clusters break easily.
7. Cell clusters would adhere on the surface of the plate. Then,
the outgrowth of cells would be observed surrounding the
adherent cell clusters.
8. LSEC progenitors would reach confluence and exhibit endo-
thelial cell morphology after 7 days of culture.
9. A83-01 (TGFβRI inhibitor) is a key molecule for inducing
LSEC maturation. After culturing in the presence of A83-01,
the expression levels of the mature LSEC markers, FCGR2B,
STAB2, F8, and LYVE1, would be highly upregulated.
10. Y27632 (ROCK inhibitor) is a key factor for supporting HSC
maturation. iPSC-derived HSCs proliferate and exhibit typical
mature HSC morphology with projections after 5 days of
culture in the presence of Y27632. We can observe vitamin A
droplets in iPSC-derived HSCs after incubation with
vitamin A.
11. After 30 min of incubation, we can observe adherent MEF cells
under the microscope.
12. Measure CYP3A4 activity every week. It would gradually
increase during 10 weeks of culture (Fig. 4).
142 Taketomo Kido and Yuta Koui

CYP3A4 activity
70

Relative CYP3A4 activity

60

(2 weeks = 1.0)
50

40

30

20

10

0
2 3 4 5 6 7 8 9 10 weeks

Fig. 4 CYP3A4 activity. Relative Cytochrome P450 3A4 (CYP3A4) activity in iPSC-
derived liver cells. The results are shown as the mean  SEM of three
independent experiments. The measurements were performed every week

Acknowledgment

This study was supported by CREST program of Japan Science and

Technology Agency, Grants-in-Aid for Scientific Research of Japan
Society for the Promotion of Science, and Japan Agency for Medi-
cal Research and Development (AMED).

References

1. Takahashi K, Tanabe K, Ohnuki M, Narita M, signaling promote the maturation of human plu-
Ichisaka T, Tomoda K, Yamanaka S (2007) ripotent stem cell-derived hepatocytes. Develop-
Induction of pluripotent stem cells from adult ment 140:3285–3296
human fibroblasts by defined factors. Cell 5. Matsumoto K, Yoshitomi H, Rossant J, Zaret KS
131:861–872 (2001) Liver organogenesis promoted by endo-
2. Si-Tayeb K, Noto FK, Nagaoka M, Li J, Battle thelial cells prior to vascular function. Science
MA, Duris C, North PE, Dalton S, Duncan SA 294:559–563
(2010) Highly efficient generation of human 6. Yin C, Evason KJ, Asahina K, Stainier DY
hepatocyte-like cells from induced pluripotent (2013) Hepatic stellate cells in liver develop-
stem cells. Hepatology 51:297–305 ment, regeneration, and cancer. J Clin Invest
3. Takayama K, Inamura M, Kawabata K, 123:1902–1910
Sugawara M, Kikuchi K, Higuchi M, 7. Kido T, Koui Y, Suzuki K, Kobayashi A,
Nagamoto Y, Watanabe H, Tashiro K, Miura Y, Chern EY, Tanaka M, Miyajima A
Sakurai F, Hayakawa T, Furue MK, Mizguchi (2015) CPM is a useful cell surface marker to
H (2012) Generation of metabolically function- isolate expandable bi-potential liver progenitor
ing hepatocytes from human pluripotent stem cells derived from human iPS cells. Stem Cell
cells by FOXA2 and HNF1a transduction. J Rep 5:508–515
Hepatol 57:628–636 8. Koui Y, Kido T, Ito T, Oyama H, Chen SW,
4. Ogawa S, Surapisitchat J, Virtanen C, Ogawa M, Katou Y, Shirahige K, Miyajima A (2017) An
Niapour M, Sugamori KS, Wang S, Tamblyn L, in vitro human liver model by iPSC-derived
Guillemette C, Hoffmann E, Zhao B, Strom S, parenchymal and non-parenchymal cells. Stem
Laposa RR, Tyndale RF, Grant DM, Keller G Cell Rep 9:490–498
(2013) Three-dimensional culture and cAmp
 Chapter 13

Culture System of Bile Duct-Like Cystic Structures Derived

from Human-Inducible Pluripotent Stem Cells
Akihide Kamiya, Kazuya Anzai, Kota Tsuruya, and Hiromi Chikada

Abstract
Inducible pluripotent stem (iPS) cells are multipotent stem cells that are produced by gene transfer of
reprogramming factors to somatic cells. They are thought to be an important source of regenerative
medicine because of their pluripotency and self-renewal ability. Although the liver has high regeneration
ability, continuous death of hepatocytes due to chronic inflammation leads to liver cirrhosis and liver
carcinoma. With regard to such serious liver diseases, liver transplantation is used as a complete cure, but
there is a problem of donor shortage. Therefore, transplantation therapy using liver tissue generated from
stem cells in vitro is expected.
We are developing a system to induce the differentiation of cholangiocytes, one of important
non-parenchymal cells in living liver tissue, from human iPS cells. Bile duct-like cystic structures can be
induced by purifying human iPS cell-derived hepatoblasts expressing hepatic progenitor cell surface markers
and inducing differentiation under appropriate culture conditions. These cells are considered to be useful in
constructing a hepatic organoid that reproduces the liver structure of the living body.

Key words Hepatic progenitor cells, Human iPS cells, Induction of hepatic differentiation, Long-
term proliferation, Purification of progenitor cell

1 Introduction

The liver is the largest organ in the body and contains various
functions for maintaining homeostasis in the body. Hepatocytes
(liver parenchymal cells) express various genes for mature liver
function such as synthesis of serum proteins, metabolism of carbo-
hydrates and lipids, and detoxification of alcohol and drugs. In
addition, several types of non-parenchymal cells are also present in
the liver. Cholangiocytes constitute the intrahepatic bile ducts.
Stellate cells are involved in the accumulation of vitamin
A. Kupffer cells are intrahepatic macrophages, and sinusoidal endo-
thelial cells constitute liver sinusoidal vessels in the liver. It is known
that cell-cell interactions between these non-parenchymal cells and
hepatocytes are important for liver function. Cholangiocytes

Naoki Tanimizu (ed.), Hepatic Stem Cells: Methods and Protocols, Methods in Molecular Biology, vol. 1905,
https://doi.org/10.1007/978-1-4939-8961-4_13, © Springer Science+Business Media, LLC, part of Springer Nature 2019

143
144 Akihide Kamiya et al.

constitute the bile duct system for excretion of bile acid derived
from hepatocytes [1]. Bile acids synthesized by hepatocytes are
released from the apical side of the hepatocyte membrane to the
bile canaliculus and are excreted into the small intestine and gall-
bladder through the intrahepatic bile duct network formed by
cholangiocyte. Therefore, disorders of the intrahepatic bile ducts
such as primary biliary cholangitis lead to poor elimination of bile
acid from the liver and cause various liver injuries.
During embryonic development, the ventral foregut endoderm
begins to form early liver buds by migrating into the septum
transversum (embryonic day 8–9 in mice) [2, 3]. Subsequently,
hepatic buds grow by soluble factor signals from surrounding
tissues, and hepatic progenitor cells (hepatoblasts), which are the
cells that differentiate into the mature hepatocytes and bile duct
cells, grow inside. Hepatoblasts surrounding the portal vein form
ductal plate in the middle of embryogenesis (13–15 days of mouse
embryo) [4]. Part of ductal plate cells lose hepatocytic gene expres-
sion and form a luminal structure of bile duct. Several studies using
model mice describe that the cell-cell interaction between hepato-
blasts and mesenchymal cells around the portal vein and the signal
of soluble factors are important for differentiation from hepato-
blasts to cholangiocytes. For example, the concentration of trans-
forming growth factor β (TGFβ) in the liver during mid-fetal
embryos is high around the portal vein and low as it approaches
the inside of the liver parenchyma. The concentration gradient of
TGFβ induces differentiation of hepatoblasts [5]. In addition,
Notch-Jagged signal is important for bile duct formation process,
and Jagged-1 is the causative gene of Alagille syndrome accompa-
nied by abnormalities of the bile duct system [6, 7]. It is also known
that various transcription regulators are important for bile duct
formation and differentiation. Sal-like protein 4 is a transcription
factor that regulates the fate determination of hepatoblasts and
promotes bile duct differentiation by inducing cholangiocytic
gene expression [8]. It has also been reported that grainyhead-
like 2 regulates bile duct formation via gene expression control of
miR122 [9].
Complex three-dimensional structure of the liver is crucial for
its functions. In order to reproduce the liver functional structure
in vitro, organoid culture method has been applied to co-culture of
hepatocytes, mesenchymal cells, and endothelial cells
[10, 11]. However, these hepatic organoids do not contain the
bile duct structure, and the cytotoxicity of bile acids synthesized
by hepatocytes against cells in these organoids is a potential prob-
lem. Therefore, the construction of a novel hepatic organoid using
a co-culture system containing cholangiocytes is required. Various
groups described about the in vitro culture for induction of cho-
langiocytes. We previously established an in vitro amplification
system of hepatic progenitor cells derived from human-inducible
 Induction of Cholangiocyte from Pluripotent Stem Cells 145

pluripotent stem (iPS) cells by purifying CD13+CD133+ cells after

sequential treatment of cytokines. In addition, bile duct-like cystic
structures can be induced by embedding these hepatic progenitor
cells in extracellular matrix gel and performing three-dimensional
culture [12]. Other groups also reported differentiation system of
human iPS cells into bile ductal cells by changing culture conditions
[13, 14].
Polycystic liver diseases (PLD), which contain multiple cysts in
the liver, are caused by genetic abnormalities of cholangiocytes.
Recent analyses revealed that several genes such as protein kinase
C substrate 80 K-H and SEC63 are involved in the onset of PLD
[15, 16]. Cilium formation is regulated by these genes, but the
molecular mechanism causing the pathological condition of PLD is
still unknown [17]. Advances in genome editing technology have
made it possible to introduce gene mutations in human iPS cells.
Therefore, it is possible to induce mutation in the causative gene of
bile duct disease in human iPS cells, to induce differentiation into
hepatic progenitor cells and cholangiocytic cells. These cells are
useful to reproduce the pathology of PLD or other bile duct-
related diseases in vitro and screen therapeutic drugs. In this chap-
ter, we introduce our culture system inducing bile duct-like tissues
derived from human iPS cells.

2 Materials

2.1 Differentiation The approval of a suitable committee is usually required for animal
of Human iPS Cells into experiments.
Hepatic Progenitor 1. Human iPS cells: HiPS-RIKEN-2F human iPS cells are estab-
Cells lished from human umbilical cord-derived fibroblasts by intro-
duction of Oct3/4, Klf4, Sox2, and c-Myc using a retroviral
system [18]. These cells are cryopreserved in liquid nitrogen
using DAP213 cryopreserve solution (see Note 1).
2. 0.1% gelatin: Gelatin derived from porcine skin is dissociated
with PBS and autoclaved.
3. Fetal bovine serum (FBS): FBS is inactivated by incubation at
55
C for 30 min.
4. Mouse embryonic fibroblast (MEF) culture medium: Dulbec-
co’s modified Eagle’s medium (DMEM) supplemented with
10% FBS and 1 penicillin–streptomycin–glutamine.
5. Phosphate-buffered saline (PBS): 10 PBS without CaCl2 and
MgCl2 is diluted with ultrapure water, autoclaved, and stored
at 4
C.
6. Mitomycin C: A 200 stock solution (2 mg/mL) is prepared
with ultrapure water.
146 Akihide Kamiya et al.

7. Human iPS cell medium: 0.1 mM nonessential amino acids, 1

penicillin–streptomycin–glutamine, 20% knockout serum
replacement, 0.1 mM 2-mercaptoethanol, 5 ng/mL recombi-
nant human basic fibroblast growth factor (FGF) in DMEM/
F12 medium.
8. Differentiation medium: RPMI 1640 containing 2% B27
supplement.
9. Activin A (1000 stock solution): 100 mg/mL recombinant
human activin A in PBS with 0.1% bovine serum protein (BSA).
10. Basic FGF (1000 stock solution): 10 mg/mL recombinant
human basic FGF in PBS with 0.1% BSA.
11. Bone morphogenetic protein (BMP)-4 (1000 stock solu-
tion): 20 mg/mL recombinant human BMP-4 in PBS with
0.1% BSA.
12. Recombinant human hepatocyte growth factor (HGF)
(1000 stock solution): 40 mg/mL HGF in PBS with
0.1% BSA.
13. Staining medium: PBS with 0.3% FBS.
14. Stock propidium iodide solution (1000): 1 mg/mL propi-
dium iodide in PBS. The stock solution is diluted with staining
medium.
15. Antibodies for flow cytometry: see Table 1.

Table 1
List of antibodies used for flow cytometry experiments and immunocytochemistry for human iPS cell
differentiation and cholangiocyte differentiation

Primary antibodies for flow cytometry Clone Source Catalog number

CD13-PE WM15 BD Pharmingen 555394
CD133/1-APC AC133 Miltenyi Biotec 130–090-826
Primary antibodies for immunostaining Dilution Source Catalog number
α-Fetoprotein (AFP) (C3) 1/600 Dako A0008
Keratin-7 (CK7) 1/1000 Dako M7018
β-Catenin 1/1000 BD Pharmingen 610154
F-Actin (Acti-stain 488 phalloidin) 1/150 Cytoskeleton, Inc. PHDG1
Protein kinase Cζ (C-20) 1/500 Santa Cruz sc-216
Integrin α6 (CD49f) (GoH3) 1/500 BD Pharmingen 555734
Secondary antibodies Dilution Source Catalog number
Anti-mouse/Alexa Fluor 555 1/1000 Invitrogen A31570
Anti-rat/Alexa Fluor 488 1/1000 Invitrogen A21208
Anti-rabbit/Alexa Fluor 488 1/1000 Invitrogen A21206
 Induction of Cholangiocyte from Pluripotent Stem Cells 147

2.2 Expansion 1. H-CFU-C medium: 1 Insulin–Transferrin–Selenium X,

of Hepatic Progenitor 10 mM nicotinamide, 2.5 mM HEPES buffer solution, 1
Cells and Induction penicillin–streptomycin–glutamine, 0.1 mM nonessential
of Cholangiocytic amino acids.
Differentiation Using 2. Standard culture medium: 1:1 mixture of hepatic colony-form-
Extracellular Matrices ing unit (H-CFU-C) medium and DMEM with 10% FBS,
107 M dexamethasone, 0.25 μM A-83-01, 10 μM Y-27632,
40 ng/mL recombinant human HGF, and 20 ng/mL recom-
binant human epidermal growth factor (EGF).
3. Extracellular matrix gel solution: A mixture of 40% collagen
type-IA solution (Nitta gelatin) and 40% matrigel in DMEM/
F12 medium.
4. Cholangiocyte culture medium: 2% B27 supplement, 0.25 μM
A-83-01, 10 μM Y-27632, 20 ng/mL EGF, 40 ng/mL HGF,
40 ng/mL recombinant human Wnt-3a, and 100 ng/mL
recombinant human R-spondin 1 in 1:1 mixture of H-CFU-
C medium and DMEM/F-12.

2.3 Characterization 1. RNAiso Plus: Solution for RNA purification.

of Human iPS Cells- 2. ReverTra Ace® qPCR RT Master Mix: Solution for cDNA
Derived Cholangiocytic synthesis.
Cells
3. THUNDERBIRD® Probe qPCR Mix: Solution for
2.3.1 Analyses quantitative PCR.
of Cholangiocytic 4. Universal Library: Detection probes for quantitative PCR. The
Differentiation Using probe numbers and primer sequences for each gene are
RT-PCR provided in Table 2.

2.3.2 Analyses 1. PBS+: PBS containing 1 mM MgCl2 and 1 mM CaCl2.

of Cholangiocytic 2. Collagenase-1: 1000 U/mL high-purity collagenase in PBS+.
Differentiation Using Aliquots are stored at 80
C Thaw and dilute 1:10 in PBS+
Immunocytochemistry before use.
3. 4% paraformaldehyde: Solution for cell fixation.
4. Blocking medium: 5% donkey serum and 0.5% Triton-X100
solution in PBS+. Donkey serum is used after heat inactivation
by incubation at 56
C for 30 min.
5. 40 ,6-Diamidino-2-phenylindole dihydrochloride (DAPI) solu-
tion: A 1000 solution (2 mg/mL) is prepared with methanol.
6. Immunofluorescent mount solution: Dako Fluorescence
Mounting Medium.
148 Akihide Kamiya et al.

Table 2
PCR primers for detection of human gene expression

Human genes Forward primer (50 –30 ) Reverse primer (50 –30 ) Probe number
AFP tgtactgcagagataagtttagctgac tccttgtaagtggcttcttgaac 61
GRHL2 ggacagcacatacagcgaga agccccaactgaagcactc 5
HNF4α attgacaacctgttgcagga cgttggttcccatatgttcc 3
HPRT1 tgaccttgatttattttgcatacc cgagcaagacgttcagtcct 73
Keratin 7 ctgaggctgaagcctggta gggtattccggaggtcgt 64
Keratin 19 gccactactacacgaccatcc caaacttggttcggaagtcat 71
Afp α-feto protein, Grhl grainyhead like transcription factor, HNF hepatocyte nuclear factor, HPRT1 hypoxanthine
phosphoribosyltransferase 1

3 Methods

3.1 Differentiation Methods for maintaining human iPS cells and differentiation into
of Human iPS Cells into hepatic progenitor cells were already shown in “Generation and
Hepatic Progenitor in vitro expansion of hepatic progenitor cells from human iPS cells”
Cells in detail [19].
1. Embryonic day (E) 12 ICR mouse embryos were dissected,
and the head and internal organs are surgically removed. The
torso is minced and dissociated in 0.05% trypsin–EDTA. After
centrifugation and washing steps, cells were inoculated on
culture dishes with MEF culture medium. MEFs are incubated
with 0.01 mg/mL mitomycin C in MEF culture medium at
37
C for 2 h.
2. Human iPS cells are cultured and maintained on mitomycin
C-treated MEFs. The cells are cultured at 37
C in a 5% CO2
incubator. Human iPS cells are proliferated into Semi-
confluent state and used for hepatocytic differentiation.
3. Semi-confluent human iPS cells are cultured in Differentiation
medium supplemented with 100 ng/mL activin A on days
0–4 at 37
C in a 5% CO2 incubator.
4. After activin A stimulation, iPS cells were cultured in Differen-
tiation medium supplemented with 10 ng/mL basic FGF and
20 ng/mL BMP-4 on days 4–8.
5. After FGF and BMP-4 stimulation, iPS cells were cultured in
Differentiation medium supplemented with HGF on days
8–12. These cells are used for the isolation of hepatic progeni-
tor cells.
6. After 12 days of culture, cells are washed with PBS and dis-
sociated with 0.05% trypsin–EDTA (1 mL/60 mm dish). Tryp-
sinized cells are collected into a 15 mL tube and centrifuged
 Induction of Cholangiocyte from Pluripotent Stem Cells 149

(260  g for 5 min). Fluorescent-conjugated antibodies are

added to the cell pellets (Antibody aliquots: APC-conjugated
anti-CD133 and PE-conjugated anti-CD13, Table 1) and
incubated at 4
C for 1 h (see Note 2).
7. After the washing step, cells were stained with propidium
iodide and analyzed using flow cytometer.
8. Doublets are excluded by FSC, SSC, and pulse width. Dead
cells are excluded by propidium iodide staining.
CD13highCD133+ cells are sorted as a hepatic progenitor cell
fraction (see Note 3).

3.2 Culture 1. Mitomycin C-treated MEF cells (2  105 cells/well) are plated
of Hepatic Progenitor onto 0.1% gelatin-coated 12-well plates and maintained in
Cells and Induction MEF culture media the day before sorting. These cells are
of Cholangiocytic incubated for at least 12 h at 37
C in a CO2 incubator.
Differentiation Using 2. Sorted CD13highCD133+ cells (derived from Subheading 3.1)
Extracellular Matrices are cultured on MEF cells. Standard culture medium is used for
cell culture (750–1000 μL per well). Medium is replaced every
3 days.
3. Individual CD13highCD133+ cell clonally proliferates and
forms a hepatic progenitor colony on MEF for 10–14 days of
culture. The colonies are dissociated by 5 min incubation of
0.05% trypsin–EDTA at 37
C. After incubation, 1 mL of
DMEM containing 10% FBS is added, and cells are collected
by the centrifugation at 260  g for 5 min.
4. Cells are replated onto new mitomycin C-treated MEF cells.
Approximately 1/10 to 1/3 of the cells are seeded on individ-
ual well. After 7–10 days of culture, the passaged cells form new
colonies.
5. Several passaged colonies derived from CD13highCD133+ cells
are dissociated using 0.05% trypsin–EDTA and collected in
DMEM containing 10% FBS as shown above. The cell number
is counted.
6. The cells are then combined with an extracellular matrix gel
solution (1500 cells/50 μL extracellular matrix gel/well).
After mixing by pipetting, cells embedded in the gel are
cultured in 24-well culture plates (Fig. 1a, b) (see Note 4).
7. These gel solutions are incubated in 37
C CO2 incubator for
10 min to form gel. Cholangiocyte culture medium (500 μL/
well) is added, and cells are cultured for 10–12 days with
medium changes every 3 days.
8. Cystic structures are formed in the extracellular matrix gel after
14–16 days of culture (Fig. 1c).
150 Akihide Kamiya et al.

Fig. 1 Method for the culture of bile-duct like cysts in extracellular gels. (a) After mixing the cells into
extracellular matrix gel solution, place the solution in the center of the well of the culture dish. (b) After
incubation for 10 min in a CO2 incubator, the gel solution formed gel. (c) Cyst structures derived from human
iPS cells were formed

3.3 Characterization 1. During differentiation steps (derived from Subheading 3.1),

of Human iPS Cells- iPS cells-derived hepatic progenitor cells are washed with PBS,
Derived Cholangiocytic and total RNAs are extracted with RNAiso Plus (1 mL/well of
Cells 6-well culture dish). For the control, total RNA is purified from
non-differentiated iPS cells. Individual extracellular matrix gels
3.3.1 Analyses containing cholangiocytic cysts are lysed in RNAiso Plus
of Cholangiocytic (1 mL/well of 24-well culture dish). Total RNA is purified
Differentiation Using according to the manufacturer’s protocol.
RT-PCR
2. 0.5 μg Total RNA is used for single-strand cDNA synthesis
using ReverTra Ace® qPCR RT Master Mix.
3. The expression levels of marker genes for cholangiocytic cells
(Keratin 19, Keratin 7, and Grhl2) and hepatic cells (AFP and
HNF4α) are examined in normal human iPS cells and differ-
entiated iPS cell-derived cells. The expression of hypoxanthine
 Induction of Cholangiocyte from Pluripotent Stem Cells 151

phosphoribosyltransferase 1 (HPRT1) is used as an internal

control. RT PCR is performed using Roche Universal Library
and THUNDERBIRD® Probe qPCR Mix according to the
manufacturer’s protocol. PCR primers and probes are shown
in Table 2.

3.3.2 Analyses 1. Individual extracellular matrix gels containing cholangiocytic

of Cholangiocytic cysts are washed with PBS and incubated with collagenase
Differentiation Using solution (500 μL/well of 24-well culture dish) for 15 min at
Immunocytochemistry 37
C (see Note 5).
2. After washing with PBS, cells are fixed with 4% paraformalde-
hyde (500 μL/well) for 30 min at room temperature (RT).
3. After the washing step with PBS, cells are permeabilized with
blocking medium (500 μL/well). Cells are incubated for
30 min at RT.
4. The cells are then incubated with diluted primary antibodies
overnight at 4
C. Primary antibodies are diluted with blocking
medium.
5. After the washing step with blocking medium, cells are then
incubated with diluted secondary antibodies diluted with
blocking medium for 3 h at 37
C. In addition, nuclei are
stained with 1 DAPI solution.
6. After the washing step with blocking medium, individual extra-
cellular matrix gels containing cholangiocytic cysts are put on
cover glasses and analyzed under a LSM700 confocal micro-
scope (Fig. 2). In order to avoid the dry of samples, we cover
gel samples using immunofluorescent mount solution.

4 Notes

1. Feeder-free iPS cells are also useful for differentiation into

hepatoblasts and cholangiocytes. We have succeeded cholan-
giocytic cyst formation derived from feeder-free human iPS
cells using Cellartis® iPS Cell to Hepatocyte Differentiation
System.
2. During the trypsinized step of cytokine-stimulated human iPS
cells (Subheading 3.1), MEF feeder cells can also be collected.
These contaminating cells are eliminated in the flow
cytometry step.
3. To centrifuge a very small number of cells, use a specific tube
(low cell adsorption tube STEMFULL™) that can block non-
specific binding to proteins, peptides, and cell surfaces.
4. Matrigel solution and collagen solution have to be kept cold in
order to avoid unnecessary gel forming.
152 Akihide Kamiya et al.

Fig. 2 Representative images of bile duct-like cysts derived from human iPS cells. Cysts were stained with
antibodies against AFP, Keratin 7 (CK7), β-catenin, protein kinase Cζ, and integrin α6. F-actin was stained
with phalloidin. Nuclei were counterstained with DAPI

5. During the staining steps, the gels containing cholangiocytic

cysts are easily aspirated using the mechanical aspirator, which
is usually used to remove supernatant. In order to avoid it, we
remove supernatant using a micropipette.

Acknowledgments

This work was supported in part by the Education and Research

Support Center of Tokai University. Some figures have been repro-
duced from [12]. This work was supported by a Grant-in-Aid for
JSPS Fellows and a Grant-in-Aid from the Ministry of Education,
Culture, Sports, Science and Technology, Japan.
 Induction of Cholangiocyte from Pluripotent Stem Cells 153

References
1. Lemaigre FP (2009) Mechanisms of liver 12. Yanagida A, Ito K, Chikada H, Nakauchi H,
development: concepts for understanding Kamiya A (2013) An in vitro expansion system
liver disorders and design of novel therapies. for generation of human iPS cell-derived
Gastroenterology 137(1):62–79 hepatic progenitor-like cells exhibiting a bipo-
2. Douarin NM (1975) An experimental analysis tent differentiation potential. PLoS One 8(7):
of liver development. Med Biol 53(6):427–455 e67541
3. Houssaint E (1980) Differentiation of the 13. Dianat N, Dubois-Pot-Schneider H,
mouse hepatic primordium. I. An analysis of Steichen C, Desterke C, Leclerc P, Raveux A,
tissue interactions in hepatocyte differentia- Combettes L, Weber A, Corlu A, Dubart-
tion. Cell Differ 9(5):269–279 Kupperschmitt A (2014) Generation of func-
4. Lemaigre F, Zaret KS (2004) Liver develop- tional cholangiocyte-like cells from human plu-
ment update: new embryo models, cell lineage ripotent stem cells and HepaRG cells.
control, and morphogenesis. Curr Opin Genet Hepatology 60(2):700–714
Dev 14(5):582–590 14. Sampaziotis F, de Brito MC, Geti I, Bertero A,
5. Clotman F, Jacquemin P, Plumb-Rudewiez N, Hannan NR, Vallier L (2017) Directed differ-
Pierreux CE, Van der Smissen P, Dietz HC, entiation of human induced pluripotent stem
Courtoy PJ, Rousseau GG, Lemaigre FP cells into functional cholangiocyte-like cells.
(2005) Control of liver cell fate decision by a Nat Protoc 12(4):814–827
gradient of TGF beta signaling modulated by 15. Drenth JP, te Morsche RH, Smink R, Bonifa-
Onecut transcription factors. Genes Dev 19 cino JS, Jansen JB (2003) Germline mutations
(16):1849–1854 in PRKCSH are associated with autosomal
6. Lemaigre FP (2008) Notch signaling in bile dominant polycystic liver disease. Nat Genet
duct development: new insights raise new 33(3):345–347
questions. Hepatology 48(2):358–360 16. Davila S, Furu L, Gharavi AG, Tian X, Onoe T,
7. Zhang D, Gates KP, Barske L, Wang G, Lanc- Qian Q, Li A, Cai Y, Kamath PS, King BF,
man JJ, Zeng XI, Groff M, Wang K, Parsons Azurmendi PJ, Tahvanainen P, Kaariainen H,
MJ, Crump JG, Dong PDS (2017) Endoderm Hockerstedt K, Devuyst O, Pirson Y, Martin
Jagged induces liver and pancreas duct lineage RS, Lifton RP, Tahvanainen E, Torres VE,
in zebrafish. Nat Commun 8(1):769 Somlo S (2004) Mutations in SEC63 cause
autosomal dominant polycystic liver disease.
8. Oikawa T, Kamiya A, Kakinuma S, Zeniya M, Nat Genet 36(6):575–577
Nishinakamura R, Tajiri H, Nakauchi H
(2009) Sall4 regulates cell fate decision in 17. Wills ES, Te Morsche RHM, van Reeuwijk J,
fetal hepatic stem/progenitor cells. Gastroen- Horn N, Geomini I, van de Laarschot LFM,
terology 136(3):1000–1011 Mans DA, Ueffing M, Boldt K, Drenth JPH,
Roepman R (2017) Liver cyst gene knockout
9. Tanimizu N, Kobayashi S, Ichinohe N, Mitaka in cholangiocytes inhibits cilium formation and
T (2014) Downregulation of miR122 by Wnt signaling. Hum Mol Genet 26
grainyhead-like 2 restricts the hepatocytic dif- (21):4190–4202
ferentiation potential of adult liver progenitor
cells. Development 141(23):4448–4456 18. Itaba N, Wairagu PM, Aramaki N, Yasui T,
Matsumi Y, Kono Y, Phan AN, Otsu M,
10. Takebe T, Sekine K, Enomura M, Koike H, Kunisada T, Nakamura Y, Okano H, Jeong Y,
Kimura M, Ogaeri T, Zhang RR, Ueno Y, Shiota G (2014) Nuclear receptor gene alter-
Zheng YW, Koike N, Aoyama S, Adachi Y, ation in human induced pluripotent stem cells
Taniguchi H (2013) Vascularized and func- with hepatic differentiation propensity. Hepa-
tional human liver from an iPSC-derived tol Res 44(14):E408–E419
organ bud transplant. Nature 499
(7459):481–484 19. Yanagida A, Nakauchi H, Kamiya A (2016)
Generation and in vitro expansion of hepatic
11. Koui Y, Kido T, Ito T, Oyama H, Chen SW, progenitor cells from human iPS cells. Meth-
Katou Y, Shirahige K, Miyajima A (2017) An ods Mol Biol 1357:295–310
in vitro human liver model by iPSC-derived
parenchymal and non-parenchymal cells. Stem
Cell Rep 9(2):490–498
 Part IV

Reconstitution of Liver Tissue Structures

 Chapter 14

Generation of Hepatic Tissue Structures Using Multicellular

Spheroid Culture
Fumiya Tao, Hirotaka Mihara, and Nobuhiko Kojima

Abstract
The hepatic functions of the hepatocytes in multicellular spheroid (MCS) are lower than those in the liver.
One of the causes is that conventional hepatic MCSs do not reproduce liver-specific microstructures such as
hepatic cord. It is necessary to design the inner structure of hepatic MCSs mimicking a structural feature of
hepatic cord to represent further hepatic functions. Here we introduce a unique method to engineer the
microarchitectures in the MCSs by formation of void spaces or filling of extracellular matrices (ECMs).

Key words Multicellular spheroid, Hepatocyte, Methylcellulose, Microstructure, Alginate hydrogel

bead, Extracellular matrix

1 Introduction

Compared with monolayer cultures, MCSs are useful tools for

mimicking a part of events that occur in biological tissues. How-
ever, the cellular microenvironments of hepatic MCSs are different
from those of liver tissues. Liver tissues contain various microstruc-
tures like hepatic cords. Because these structures are considered to
be responsible for the liver functions, it is desirable to reproduce
the structures inside MCSs. Various methods for creating MCSs
have been reported, as reviewed by Laschke et al. [1]. Nevertheless,
all of these approaches do not generate complex liver tissue struc-
tures in MCSs. Therefore, it is necessary to develop a method that
can design the inner structures of MCSs. We previously reported
several cell aggregation methods using methylcellulose (MC)
medium [2–7]. In this chapter, we introduce a unique method to
gather hepatocytes and hydrogel beads by the MC medium. In
addition, we describe a further application method of the MC
medium in which ECMs are introduced into MCSs. These methods
allow formation of MCSs with microchannel structures and/or
thin layers of ECMs in a few days.

Naoki Tanimizu (ed.), Hepatic Stem Cells: Methods and Protocols, Methods in Molecular Biology, vol. 1905,
https://doi.org/10.1007/978-1-4939-8961-4_14, © Springer Science+Business Media, LLC, part of Springer Nature 2019

157
158 Fumiya Tao et al.

2 Materials

All liquids and plasticware used for cell culture must be sterile.
Regular culture medium is Dulbecco’s Modified Eagle’s Medium
(DMEM) supplemented with 10% fetal bovine serum (FBS) and
100 U/mL penicillin/streptomycin. Phosphate-buffered saline
(PBS) used uncontained calcium and magnesium ions.

2.1 Cell Cultures Choose the appropriated culture medium and centrifugation con-
ditions depending on the type of cells. In this chapter, we introduce
MCS culture using human hepatoma cell line Hep G2 cells (see
Note 1).
1. Polystyrene cell culture dish.
2. Trypsin-EDTA: 0.25 w/v% trypsin-1 mM ethylenediaminete-
traacetic acid (EDTA).
3. Trypan blue solution.
4. Hemocytometer.

2.2 MCS Assembly 1. Methylcellulose (see Note 2).

2.2.1 3% MC Medium 2. Magnetic stirrer and magnetic bar.
3. 250 mL wide-mouthed glass bottle.
4. Cold room (4  C) or 1 L beaker with crushed ice.

2.2.2 Alginate Hydrogel 1. 1.5% sodium alginate solution: sodium alginate in Milli-Q.
Beads Sterilize the solution by autoclave.
2. 5% calcium chloride solution: calcium chloride in Milli-Q.
Sterilize the solution by autoclave.
3. Inkjet system: WaveBuilder, PulseInjector (25 μm-diameter
nozzle), and ink cartridge (see Note 3).
4. 1 mL syringe.
5. 40 μm cell strainer.
6. 0.22 μm syringe filter.
7. Magnetic stirrer and magnetic bar.
8. 100 mm petri dish.
9. 1000 μL micropipette.

2.2.3 Generation of MCS 1. Positive displacement pipette.

and Isolation of MCS from 2. 35 mm petri dish.
the MC Medium
3. 2 μL micropipette.
4. Medium containing diluted Matrigel: thaw appropriate
amount of Matrigel tube overnight at 4  C. Take the Matrigel
tube from 4  C and place them on ice. Then, Matrigel was
 Generation of Hepatic Tissue Structures Using Multicellular Spheroid Culture 159

diluted to about 30 times by using FBS-free DMEM. Store it at

4  C.
5. Ultralow attachment plate.
6. 5 U/mL cellulase solution: cellulase in regular culture
medium.
7. 1000 μL micropipette with a truncated tip.
8. 1.5 mL micro tube.

3 Methods

3.1 3% MC Medium 1. Put 3 g of MC powder in a wide-mouthed bottle with a

(See Note 4) magnetic bar, and sterilize it by autoclaving.
2. Add 100 mL of the regular culture medium to the wide-
mouthed bottle.
3. Mix MC and regular culture medium with a magnetic stirrer at
cold room or in a 1 L beaker with crushed ices overnight (see
Notes 5–7).
4. Store the wide-mouthed bottle at 4  C after MC is completely
dispersed (see Note 8).

3.2 Production Sodium alginate solution can turn into a gel after contacting with
of Alginate Hydrogel calcium ions. To produce alginate hydrogel beads, the inkjet system
Beads is used to discharge droplets of sodium alginate solution into
calcium chloride solution.
1. Pour 5% calcium chloride solution in a 100 mm petri dish with
a magnetic bar, and stir it at about 200 rpm.
2. Filter 1.5% sodium alginate solution using a 0.22 μm syringe
filter to remove small dusts because those may be jammed in a
PulseInjector, and fill an ink cartridge with sodium alginate
solution using a 1 mL syringe.
3. Attach the ink cartridge to a PulseInjector. Discharge sodium
alginate solution to calcium chloride solution at a voltage of
15 V and a frequency of 1000 Hz for several hours (see Note 9).
4. Collect alginate hydrogel beads with calcium chloride solution
in a 50 mL tube using a 1000 μL micropipette, and filtrate it
through a 40 μm cell strainer to remove particles with the
diameter of 40 μm or more.
5. Centrifuge for 3 min at 200  g and aspirate the supernatant.
6. Add 5 mL PBS to the tube and pipette up and down gently.
7. Repeat step 5.
8. Resuspend hydrogel beads with 5 mL regular culture medium,
and take a small aliquot to a hemocytometer.
160 Fumiya Tao et al.

9. Count hydrogel beads.

10. Repeat step 5.
11. Adjust the concentration of hydrogel beads to 2  106 parti-
cles/mL by adding regular culture medium, and store it at 4  C
(see Note 10).

3.3 Generation Figure 1 shows the schematic overview of the aggregation in the
of Normal MCS Using MC medium.
MC Medium
1. Culture Hep G2 cells in polystyrene cell culture dishes. Aspi-
rate the medium and wash in PBS.
2. Remove PBS, and add appropriate volume of trypsin-EDTA
for recovering cells to each dish. Incubate the dish at 37  C for
3 min.
3. Add 10 mL regular culture medium to stop trypsinization.
4. Collect cells into a 15 mL centrifuge tube.

Normal
medium

Cells

3% MC medium

Medium removal Aggregated cells

Fig. 1 Schematic model of the cell aggregation in the 3% MC medium. Cells are shown as red color. Because
of its swelling property, the MC medium started absorbing the regular medium immediately. Finally,
suspended cells were gathered and form a MCS in the MC medium
 Generation of Hepatic Tissue Structures Using Multicellular Spheroid Culture 161

5. Centrifuge for 3 min at 200  g and aspirate the supernatant.

6. Resuspend cells with 5 mL regular culture medium, and take a
small aliquot for cell counting.
7. Mix an equal volume aliquot of both cell suspension and trypan
blue solution, and count live and dead cells using a hemocy-
tometer (see Note 11).
8. Repeat step 5.
9. Adjust the cell number to 2  106 cells/mL by adding regular
culture medium.
10. Pour the MC medium into a petri dish using a positive dis-
placement pipette (see Notes 12 and 13).
11. Inject 1 μL of the cell suspension into the MC medium using a
micropipette (see Note 14).

3.4 Generation Continuously connected hydrogel beads in MCSs formed micro-

of MCS channel structures, which resembles microvasculature. The micro-
with Microchannel channel structures improved the albumin secretion rate and
Structures suppressed the expression of hypoxia-inducible factor-1α in rela-
tively large (≧700 μm in diameter) MCSs composed of Hep G2
cells [4].
1. Follow Subheading 3.3, steps 1–9 to obtain the cell
suspension.
2. Mix equal volumes of the cell suspension and the hydrogel
bead suspension.
3. Pour the MC medium into a petri dish using a positive dis-
placement pipette.
4. Inject 1 μL of the mixture of cells and beads into the MC
medium to assemble MCSs with microchannel structures (see
Note 15) (Fig. 2).

3.5 Generation 1. Follow Subheading 3.3, steps 1–9 to obtain the cell
of MCS with Thin ECM suspension.
Layer (See Note 16) 2. Centrifuge for 3 min at 200  g and aspirate the supernatant.
3. Resuspend cells with medium containing diluted Matrigel
on ice.
4. Pour the MC medium into a petri dish using a positive dis-
placement pipette.
5. Inject 1 μL of cell suspension containing diluted Matrigel into
the MC medium to assemble MCSs composed of cells and
ECM (Fig. 3).
162 Fumiya Tao et al.

A B

: cells 50 µm

Fig. 2 MCSs with microchannel structures. (a) Schematic model of MCSs with
microchannel structures. (b) The hematoxylin and eosin staining with sectioning
of MCSs with microchannel structures. Red line demonstrates microchannel
structures

A B

: cells Red : cells

Green : FITC-collagen Green : FITC-collagen 100 µm

Fig. 3 MCSs with ECM thin layer. (a) Schematic model of MCSs with ECM thin
layer. (b) Image of MCS with thin ECM layer obtained by confocal microscopy.
This figure was used FITC labeled collagen instead of Matrigel to visualize ECM.
Red color shows cells. Green color indicates FITC-labeled collagen

3.6 Isolation of MCS MCSs can be isolated from the MC medium after 1 day (see Note
from the MC Medium 17). Two isolation methods are described below. Choose an appro-
and Further Culture priate isolation method depending on the number of MCSs (see
Note 18). Isolated MCSs can be further cultured by transferring
MCSs to ultralow attachment plates (see Note 19).

3.6.1 Isolation of a Small 1. Take a small amount of the MC medium containing MCSs
Number of MCSs using a micropipette with a truncated tip, and transfer MCSs
to a 1.5 mL micro tube.
2. Add 1 mL PBS to remove the MC medium.
3. Wait for MCSs to sink to the bottom of the tube and aspirate
the supernatant.
4. Add an appropriate volume of regular culture medium.
 Generation of Hepatic Tissue Structures Using Multicellular Spheroid Culture 163

3.6.2 Isolation of a Large 1. Overlay the 5 U/mL cellulase solution on the MC medium at
Number of MCSs 37  C for about 30 min to reduce the viscosity of the medium
(see Notes 20–22).
2. Transfer MCSs from the MC medium using a micropipette
with a truncated tip to an adequate tube.
3. Wash MCSs using PBS to remove the MC medium.
4. Wait until MCSs sink to the bottom of the tube and aspirate the
supernatant.
5. Add an appropriate volume of regular culture medium.

4 Notes

1. The 3% MC medium can generate MCSs using cells described

below: primary human, mouse or rat hepatocytes, human
iPSC-derived hepatocytes, mouse pancreatic β-cell line, pri-
mary mouse or rat pancreatic β-cells, primary mouse or rat
bone marrow cells, primary human vascular endothelial cells,
and primary mouse testicular cells.
2. Product number: M0512, Sigma-Aldrich is especially
recommended.
3. WaveBuilder can adjust discharge conditions of a PulseInjector
with three parameters (drive waveform, number of discharge
per second, and drive voltage). PulseInjector discharges liquid
particles. Ink cartridge is liquid container.
4. The concentration of the MC medium can be changed. Gener-
ally, the cells are less likely to be aggregated, as the concentra-
tion decreases. Maximum concentration of MC is 3% because it
becomes hard to disperse completely.
5. Lower temperatures are better to disperse MC.
6. As MC is dispersed, the viscosity of medium is increased. Check
the bottle occasionally to make sure the bar is continuously
spinning.
7. A powerful magnetic stirrer is recommended.
8. The 3% MC medium is made by a day before use because 1 day
is required to disperse MC. If there are masses of MC, keep
mixing the medium with a magnetic stirrer. A few MC residues
can be dispersed by leaving the bottle at 4  C. The MC medium
is stable when it is stored at 4  C; however, it should be used as
soon as possible.
9. Adjust the voltage and frequency in order to fly uniform dro-
plets from the nozzle. Depending on temperature and humid-
ity, optimum conditions of voltage and frequency are changed.
164 Fumiya Tao et al.

The strobe photography can be used to confirm whether dro-

plets are uniform.
10. The density of hydrogel beads can be changed depending on
the applications. The bead content is able to be changed from
0% to 60% in MCSs. When the bead content becomes greater
than 60%, MCSs are fragile.
11. The live cells are not colored, and the dead cells appear as a blue
color.
12. For example, 35 and 60 mm petri dishes are poured 2 and
4 mL of the MC medium, respectively.
13. When the MC medium is poured into a petri dish, a positive
displacement pipette is used because the 3% MC medium is
highly viscous. If bubbles are present in the MC medium, leave
the dish until all the bubbles disappear.
14. Cells injected into the MC medium are rapidly gathered within
the first 10 min. Although further incubation is effective, the
movement becomes slow and almost stops within 30 min.
15. MCSs holding hydrogel beads have no liquid flow, but materi-
als are exchanged between medium and MSCs in a diffusion
manner because the beads are composed mostly of water. If
material exchange by convection manner is desired, the beads
can be removed by alginate lyase digestion to make void spaces.
To digest beads, add culture medium containing 200 μg/mL
alginate lyase at 37  C for 5 min. Without the beads, however,
MCSs are fragile. Choose stable/diffusion manner or unsta-
ble/convection manner depending on the applications.
16. The tips of the micropipette and the 3% MC medium are
cooled down until the injection procedure to prevent unex-
pected gelation. Injection is performed at room temperature.
17. It is possible to culture for several days in the MC medium.
18. Cellulase treatment can be used to reduce the viscosity of the
MC medium to allow handling of a large number of MCSs.
19. Adjust the number of MCSs to avoid the contact among them,
which causes fusion of MCSs.
20. For example, when 2 mL the MC medium is used, the volume
of cellulase solution is 1 mL.
21. Shake the dish of the MC medium occasionally.
22. When the viscosity is not sufficiently reduced in 30 min, keep
the dishes to extend digestion at 37  C.
 Generation of Hepatic Tissue Structures Using Multicellular Spheroid Culture 165

Acknowledgment

This work was supported by JSPS KAKENHI under Grant Number

24710142 and AMED under Grant Number JP17be034302.

References

1. Laschke MW, Menger MD (2017) Life is 3D: 4. Kojima N, Takeuchi S, Sakai Y (2014) Fabrica-
boosting spheroid function for tissue engineer- tion of microchannel networks in multicellular
ing. Trends Biotechnol 35:133–144 spheroids. Sensor Actuat B Chem 198:249–254
2. Kojima N, Takeuchi S, Sakai Y (2012) Rapid 5. Motoyama W, Sayo K, Mihara H et al (2016)
aggregation of heterogeneous cells and Induction of hepatic tissues in multicellular
multiple-sized microspheres in methylcellulose spheroids composed of murine fetal hepatic
medium. Biomaterials 33:4508–4514 cells and embedded hydrogel beads. Regen
3. Kojima N, Takeuchi S, Sakai Y (2014) Engineer- Ther 3:7–10
ing of pseudoislets: effect on insulin secretion 6. Kamitori S, Ozeki Y, Kojima N (2016)
activity by cell number, cell population, and β-Galactoside-mediated tissue organization dur-
microchannel networks. Transplant Proc ing islet reconstitution. Regen Ther 3:11–14
46:1161–1165 7. Sayo K, Aoki S, Kojima N (2016) Fabrication of
bone marrow-like tissue in vitro from dispersed-
state bone marrow cells. Regen Ther 3:32–37
 Chapter 15

Reconstruction of Hepatic Tissue Structures Using

Interstitial Flow in a Microfluidic Device
Ryo Sudo

Abstract
Construction of three-dimensional (3D) hepatic tissue structures is important for in vitro tissue engineer-
ing of the liver, because 3D culture of hepatocytes is critical for the maintenance of liver-specific functions.
Although conventional 3D culture methods are useful for constructing 3D hepatic tissue structures, the
precise control of culture microenvironments is required to construct more physiological tissues in vitro.
Recent advances in microfluidics technologies have allowed us to utilize microfluidic devices for hepatic cell
culture, which opened the door for creating more physiological 3D culture models of the liver. Here, we
describe the method for the construction of hepatic tissue structures using a microfluidic device which has a
3D gel region with adjacent microchannels. Primary rat hepatocytes are seeded into a microchannel in a
microfluidic device. The cells are then cultured in interstitial flow conditions, which leads to the construc-
tion of 3D tissue structures.

Key words Microfluidic device, Interstitial flow, 3D culture

1 Introduction

Construction of three-dimensional (3D) hepatic tissue structures is

important for in vitro tissue engineering of the liver, because 3D
culture of hepatocytes is critical for the maintenance of liver-specific
functions. Various 3D culture methods have been described for the
construction of liver tissues in vitro [1]. For example, the spheroid
culture is one of the first attempts to culture hepatocytes in a 3D
configuration. In this culture method, hepatocytes are seeded on
the culture surface preventing attachment, which leads to the for-
mation of multicellular aggregates known as hepatocyte spheroids.
3D stacked-up culture is another approach to construct 3D hepatic
tissue structures. In this culture, two-dimensional (2D) cell layers
of hepatocytes are stacked to form hepatic cord-like structures.
Although these 3D culture methods are useful for the construction
of 3D hepatic tissue structures, the precise control of culture

Naoki Tanimizu (ed.), Hepatic Stem Cells: Methods and Protocols, Methods in Molecular Biology, vol. 1905,
https://doi.org/10.1007/978-1-4939-8961-4_15, © Springer Science+Business Media, LLC, part of Springer Nature 2019

167
168 Ryo Sudo

microenvironments is required to construct more physiological

tissues in vitro.
Recent advances in microfluidics technologies have allowed us
to utilize microfluidic devices for cell culture. Compared to con-
ventional culture devices, microfluidic devices have advantages in
terms of the precise control of culture microenvironments
[1–3]. For example, microfluidic devices provide flexibility in
controlling biochemical and biomechanical factors such as growth
factor gradients and stiffness of extracellular matrix in culture. In
addition, the design of microchannels allows to investigate interac-
tions between multiple cell types in close proximity. Furthermore,
these devices allow real-time monitoring of cellular morphogenesis.
Therefore, the use of microfluidic devices for hepatic cells opens the
door for creating more physiological 3D culture models in liver
tissue engineering.
Here, we describe the method for the construction of hepatic
tissue structures using microfluidic devices. Specifically, a micro-
fluidic device, which has a 3D gel region with adjacent microchan-
nels, is used for 3D culture of hepatocytes. Suspension of primary
rat hepatocytes was injected into a microchannel, and the cells are
cultured in interstitial flow conditions. Hepatocytes construct 3D
tissue structures in the interstitial flow conditions.

2 Materials

2.1 Fabrication 1. SYLGARD® 184 silicone elastomer base and the curing agent.
of Microfluidic Devices 2. Plastic cup.
3. Glass rod.
4. Desiccator connected with a vacuum pump.
5. Prepare an SU-8 mold for desired patterning of microchannels
using standard photolithography technique (see ref. [4]).
6. Oven.
7. Biopsy punch.
8. Scotch tape.
9. Plasma cleaner.
10. Culture dish.

2.2 Surface 1. Poly-D-lysine (PDL) solution: 1 mg/mL solution in deionized

Treatment water. Store at 20  C.
of Microchannels 2. Collagen solution: Prepare 20 μL of 10 PBS in a 1 mL tube
and Gel Formation on ice. Add deionized water, 0.5 N NaOH (see Note 1), and rat
in a Microfluidic tail type I collagen solution (see Note 2) to the 1 mL tube on
Device ice, which makes up to 200 μL. Mix the solution gently by
pipetting. Store on ice until use.
 Hepatic Tissue Structures in a Microfluidic Device 169

3. Humid chamber: Pour deionized water into a clean container

(e.g., empty micropipette tip box), and autoclave it. Place the
autoclaved container in a 5% CO2 incubator at 37  C until use.
4. Culture medium: Standard culture medium for hepatocyte
culture can be used (see ref. [5]). DMEM supplemented with
20 mM HEPES, 25 mM NaHCO3, 30 mg/L L-proline,
0.5 mg/L insulin, 10 7 M dexamethasone, 10% FBS, 10 mM
nicotinamide, 1 mM ascorbic acid 2-phosphate, 10 ng/mL
EGF, and antibiotics.

3 Methods

3.1 Fabrication 1. Prepare SYLGARD® 184 silicone elastomer base and the cur-
of Microfluidic Devices ing agent in a 10:1 weight ratio in a disposable plastic cup.
2. Mix the solution, which is poly-dimethyl siloxane (PDMS)
prepolymer, by a glass rod.
3. Place the PDMS prepolymer in a desiccator connected with a
vacuum pump for removing air bubbles in the PDMS
prepolymer.
4. Pour the PDMS prepolymer onto an SU-8 mold to the desired
thickness (e.g., 7 mm), and place them in a desiccator again for
removing air bubbles completely.
5. Place the mold in an oven at 65  C for >3 h to cure the
degassed prepolymer (see Note 3).
6. Cut out and detach the cured PDMS from the mold using a
scalpel.
7. Trim the cured PDMS, and punch the trimmed PDMS devices
to form inlets and outlets of microchannels using a biopsy
punch.
8. Remove small particles from the surface of the PDMS devices
using Scotch tape (see Note 4).
9. Place the PDMS devices in a glass beaker with deionized water
and autoclave them.
10. Place the PDMS devices in a clean container (e.g., empty
micropipette tip box), and autoclave them.
11. Dry the autoclaved PDMS devices in an oven at 65  C.
12. Place the PDMS devices, and autoclaved coverslips in a plasma
cleaner for plasma treatment in air (see Note 5).
13. Bond a PDMS device and a coverslip immediately after air
plasma treatment (Fig. 1).
14. Place the bonded PDMS device and coverslip on a flat surface,
and gently press the PDMS device to enhance adhesion
between the PDMS device and a coverslip.
170 Ryo Sudo

Fig. 1 Plasma bonding. A PDMS device and a coverslip are bonded after air
plasma exposure, resulting in the fabrication of a microfluidic device

15. Place each microfluidic device in a culture dish after plasma

bonding. The following procedures should be carried out
aseptically.

3.2 Surface 1. Fill channels of the microfluidic devices with PDL solution for
Treatment surface coating (see Note 6).
of Microchannels 2. Place the devices in a 5% CO2 incubator at 37  C for >4 h.
and Gel Formation
3. Remove the PDL solution and rinse the microchannels with
in a Microfluidic deionized water twice.
Device
4. Aspirate the microchannels completely and dry them in an oven
at 65  C overnight (see Note 7).
5. Prepare collagen solution on ice.
6. Slowly inject the collagen solution from an outlet of a gel
channel to fill out the gel region using a 10 μL micropipette
(Fig. 2a, b).
7. Place the devices in a humid chamber, and incubate them for
30 min in a 5% CO2 incubator at 37  C to allow gelation (see
Note 8).
8. Gently fill the microchannels with warmed culture medium
(Fig. 2c), and place the device in a culture dish.

3.3 Hepatocyte 1. Prepare cell suspension of hepatocytes on ice, which can be

Seeding into isolated from rats by a two-step collagenase perfusion (see
a Microchannel Chapter 5).
2. Add 20 μL of hepatocyte suspension at a density of 3  106
cells/mL to the outlet of a microchannel (Fig. 3a).
3. Tip the device to flow the cell suspension through the micro-
channel, which allows hepatocytes to attach and stack on the
sidewall of collagen gel, resulting in the formation of multicel-
lular aggregates (Fig. 3b, c) (see Note 9).
 Hepatic Tissue Structures in a Microfluidic Device 171

Fig. 2 Schematic images of a microfluidic device. (a) A microfluidic device before the injection of collagen
solution. (b) A microfluidic device filled with collagen gel. Collagen solution is injected from an outlet of a gel
channel to fill out the gel region. Gelation of the collagen solution results in the formation of two parallel
microchannels for culture medium. (c) A microfluidic device filled with both collagen gel and culture medium.
Microchannels are filled with culture medium after collagen solution is gelled

Fig. 3 Schematic images of a microfluidic device during hepatocyte seeding. (a) Cell suspension is added to
the outlet of a microchannel. (b, c) A microchannel is filled with cell suspension, and the device is tipped on its
side and maintained in the incubator to allow cells to attach and stack on the sidewall of collagen gel

4. Carefully place the devices in a humid chamber, and incubate

them for 30 min in a 5% CO2 incubator to allow cell adhesion.
5. Return the device to horizontal and place each device in a
culture dish.
172 Ryo Sudo

Fig. 4 Reconstruction of hepatic tissue structures in a microfluidic device. (a) A microfluidic device connected
with medium reservoirs. (b) A schematic image of the microchannels with medium reservoirs. A pressure
difference of 5 mmH2O between two microchannels generates interstitial flow across the gel region. (c)
Hepatocytes are stacked and construct multicellular aggregates under interstitial flow. (d) Corresponding
phase-contrast images during hepatocyte tissue formation. Hepatocytes reconstruct hepatic tissue structures
in interstitial flow conditions within 4 days of culture

3.4 Hepatocyte 1. Gently insert medium reservoirs into the microchannel outlets
Culture in Flow (Fig. 4a).
Conditions 2. Add culture medium to the reservoirs to make a pressure
difference of 5 mmH2O between two medium channels across
the gel region (Fig. 4b) (see Note 10).
3. Change culture medium daily (see Note 11).
4. The other cell type (e.g., endothelial cells) can be added to the
other channel for coculture (Fig. 5) (see Note 12).

4 Notes

1. The amount of 0.5 N NaOH is determined to adjust to the

desired pH. We usually adjust pH 5–9. The pH during gelation
is important for the stiffness of collagen gel, which results in
different cellular morphogenesis in the gel (see refs. [6, 7]). For
example, flexible collagen gel promotes sprouting angiogenesis
while no significant effect of the gel stiffness on hepatocyte
morphogensis was observed.
 Hepatic Tissue Structures in a Microfluidic Device 173

Fig. 5 An example of coculture in a microfluidic device. After hepatocytes construct 3D tissue structures, the
other cell type (e.g., endothelial cells) can be added to the other channel

2. We use commercially obtained rat tail type I collagen solution,

the concentration of which is ~4 mg/mL.
3. The temperature of the oven can be increased to shorten the
curing time of PDMS.
4. This cleaning process should be carried out carefully since the
remained small particles on the surface of the PDMS device
disturb adhesion between the PDMS device and a coverslip
during plasma bonding.
5. We usually perform the process of plasma exposure for ~60 s.
However, this exposure time should be optimized depending
on the specification of a plasma cleaner.
6. PDL coating of microchannels is carried out to avoid separa-
tion of collagen gel from channel walls. This process promotes
3D migration of cells in the collagen gel (see ref. [8]).
7. This dry process is important for restoring hydrophobicity of
the channel surface. Hydrophobic channel surfaces are essential
for retaining collagen solution in the desired position during
gel formation described in Subheading 3.2, step 6.
8. The incubation time for gelation should be optimized depend-
ing on the volume of the gel. However, it is important to keep
the incubation time to a minimum for reducing evaporation of
collagen solution even in a humid chamber. Evaporation of
collagen solution leads to insufficient gel volume.
9. The size of multicellular aggregates can be confirmed visually.
10. Hepatocytes form 3D clusters on the sidewall of collagen gel
(Fig. 4c). These cells organize into hepatic tissue structures
within 3–4 days of culture in interstitial flow conditions
(Fig. 4d). We found that the flow direction from the cells to
gel is important for the tissue formation. The pressure differ-
ence of 5 mmH2O can be changed according to the desired
velocity of interstitial flow. However, we confirmed that
174 Ryo Sudo

interstitial flow generated by 5 mmH2O is effective for induc-

ing 3D hepatocyte tissue formation (see ref. [5]).
11. It is important to change culture medium periodically because
the velocity of the interstitial flow decreases over time but is
restored by changing the culture medium (see ref. [5]).
12. Various cell types can be added to the other channel for inves-
tigation of interactions between multiple cell types in close
proximity.

Acknowledgment

This work was supported by JSPS KAKENHI Grant Number

16H03173 to R.S.

References

1. Sudo R (2014) Multiscale tissue engineering for 5. Sudo R, Chung S, Zervantonakis IK,
liver reconstruction. Organogenesis Vickerman V, Toshimitsu Y, Griffith LG,
10:216–224 Kamm RD (2009) Transport-mediated angio-
2. Chung S, Sudo R, Vickerman V, Zervantonakis genesis in 3D epithelial coculture. FASEB J
IK, Kamm RD (2010) Microfluidic platforms 23:2155–2164
for studies of angiogenesis, cell migration, and 6. Yamamura N, Sudo R, Ikeda M, Tanishita K
cell-cell interactions. Ann Biomed Eng (2007) Effects of the mechanical properties of
38:1164–1177 collagen gel on the in vitro formation of micro-
3. Zervantonakis IK, Kothapalli CR, Chung S, vessel networks by endothelial cells. Tissue Eng
Sudo R, Kamm RD (2011) Microfluidic devices 13:1443–1453
for studying heterotypic cell-cell interactions 7. Chung S, Sudo R, Mack PJ, Wan CR,
and tissue specimen cultures under controlled Vickerman V, Kamm RD (2009) Cell migration
microenvironments. Biomicrofluidics 5:13406 into scaffolds under co-culture conditions in a
4. Shin Y, Han S, Jeon JS, Yamamoto K, Zervanto- microfluidic platform. Lab Chip 9:269–275
nakis IK, Sudo R, Kamm RD, Chung S (2012) 8. Chung S, Sudo R, Zervantonakis IK,
Microfluidic assay for simultaneous culture of Rimchala T, Kamm RD (2009) Surface treat-
multiple cell types on surfaces or within hydro- ment induced three dimensional capillary mor-
gels. Nat Protoc 7:1247–1259 phogenesis in a microfluidic platform. Adv
Mater 21:4863–4867
 Chapter 16

Generation of Hepatic Organoids with Biliary Structures

Takeshi Katsuda, Takahiro Ochiya, and Yasuyuki Sakai

Abstract
Incorporation of bile drainage system into engineered liver tissue is an important issue to advance liver
regenerative medicine. Our group reported that three-dimensional (3D) coculture of fetal liver cells (FLCs)
and adult rat biliary epithelial cells (BECs) allows reconstruction of hepatic spheroids that possess bile
ductular structures. In this chapter, we describe the detailed protocol to isolate FLCs and BECs and to
generate the spheroids with bile drainage system using these two types of primary cells.

Key words Biliary epithelial cell, Fetal liver cell, Bile ducts, Hetero-spheroid, 3D culture, Tissue
engineering

1 Introduction

Bile drainage system is indispensable for preparing a transplantable

liver tissue equivalent to serve as an alternative to a liver transplant.
There is increasing evidence that heterotopic transplantation of
liver tissue engineered with adult and fetal hepatocytes, as well as
pluripotent stem cell-derived hepatic cells, is therapeutically benefi-
cial [1–4]. In particular, coculture of hepatic cells with vascular
endothelial cells has dramatically increased the in vivo functionality
of the engineered hepatic tissue after transplantation [2, 3]. How-
ever, the scarcity of the biliary system questions the risk of cholesta-
sis in the engineered hepatic tissue after its transplantation.
Accordingly, researchers have recently attempted to generate biliary
epithelial cells (BECs) from pluripotent stem cells for the potential
use for regenerative medicine [5–7].
In our previous study, we investigated the feasibility of incor-
poration of biliary structures into in vitro engineered liver tissue
[8]. This study demonstrated that adult BECs exhibited the capac-
ity to reconstruct ductular and cystic structures with biliary func-
tionality. Such biliary reconstructive capacity was much higher in
adult BECs than in fetal liver cells (FLCs). This finding, thus,
suggests the important role of committed BECs rather than liver

Naoki Tanimizu (ed.), Hepatic Stem Cells: Methods and Protocols, Methods in Molecular Biology, vol. 1905,
https://doi.org/10.1007/978-1-4939-8961-4_16, © Springer Science+Business Media, LLC, part of Springer Nature 2019

175
176 Takeshi Katsuda et al.

progenitor cells in reconstituting the functional biliary system

in vitro. In this chapter, we describe a step-by-step protocol to
reconstruct hepatic hetero-spheroids possessing biliary structures
using rat FLCs and adult BECs.

2 Materials

2.1 Animal 1. E17.5 pregnant Wistar rats.

and Reagents 2. Isoflurane (Pfizer).
3. E-MEM.
4. Pre-perfusion buffer: Dissolving the following reagents in
500 mL distilled water: 40 g NaCl, 2.0 g KCl, 0.39 g NaH2-
PO4·2H2O, 0.76 g Na2HPO4·12H2O, 4.5 g glucose, 1.9 g
EGTA, 3.7 g EDTA, 0.0030 g phenol red, and 11.9 g HEPES
in 500 mL distilled water to prepare 10
stock solution. Dilut-
ing 50 mL of 10
stock solution with 450 mL distilled water.
After autoclave sterilization, add 2.35 mL of 7.5% NaHCO3
solution and 2 mL of 1 N NaOH solution.
5. 7.5% NaHCO3 solution: Dissolve 7.5 g of NaHCO3 in 100 mL
dH2O. Filter with a 0.22 μm filter, and store at 4  C until use.
Use within 1 month.
6. 0.05% collagenase solution: Dissolve 9.6 g NaCl, 0.48 g KCl,
0.094 g NaH2PO4·2H2O, 0.18 g Na2HPO4·12H2O, 2.86 g
HEPES, 0.888 g CaCl2·2H2O, and 0.42 g NaHCO3 in 1.2 L
of dH2O supplemented with 14.4 mL of 0.05% phenol red
solution. Store this basal solution at 4  C until use. For each
perfusion, pour 400 mL into a beaker, and adjust pH to 7–7.4
with 5 N NaOH and/or 6 N HCl. Then, dissolve 0.02 g
trypsin inhibitor and 0.2 g collagenase. After adjusting the
pH to 7.6 with 5 N NaOH and/or 6 N HCl, filter with a
0.22 μm filter, and store at 4  C until use. The prepared
collagenase solution should be used within 1 week.
7. Bile duct wash medium: Leibovitz L-15 medium (Invitrogen)
supplemented with 20 mM HEPES (Sigma-Aldrich), 1.1 g/L
galactose (Sigma-Aldrich), 30 mg/L L-proline (Sigma-
Aldrich), 0.1 μmol/L insulin (Sigma-Aldrich), 107 M dexa-
methasone (Sigma-Aldrich), and antibiotics (Invitrogen).
8. BEC digestion medium: Leibovitz L-15 medium (Invitrogen)
supplemented with 400 U/mL of collagenase (Wako), 700 U/
mL of hyaluronidase (Sigma-Aldrich), 107 M insulin, and
107 M dexamethasone. Filter with a 0.22 μm filter, and
store at 4  C until use. The prepared solution should be used
within 1 week.
 Generation of Hepatic Organoids with Biliary Structures 177

9. BEC culture medium: DMEM supplemented with insulin-

transferrin-selenium supplement (Invitrogen), 107 M dexa-
methasone, 10 mM nicotinamide (Sigma-Aldrich), 1 mM
ascorbic acid (Wako), 10 ng/mL epidermal growth factor
(EGF; Sigma-Aldrich), 10 ng/mL of hepatocyte growth factor
(HGF; PeproTech), 10% FBS, and antibiotic/antimycotic
solution (100 U/mL penicillin, 100 mg/mL streptomycin,
and 0.25 mg/ mL amphotericin B) (Invitrogen). Filter with a
0.22 μm filter, and store at 4  C until use.
10. FLC digestion solution: PBS supplemented with 0.1% collage-
nase. Filter with a 0.22 μm filter, and store at 4  C until use.
The prepared solution should be used within 1 week.
11. FLC culture medium: Williams’ medium E supplemented with
2 mM Glutamax (Invitrogen), 106 M DEX, 10 ng/mL of
EGF, 20 ng/mL of HGF, 20 ng/mL of fibroblast growth
factor (FGF)-1 (PeproTech), 20 ng/mL of FGF-4 (Pepro-
Tech), 10 ng/mL of oncostatin M (Wako), 107 M insulin,
108 M glucagon (Sigma-Aldrich), 0.5 mM ascorbic acid,
10 mM nicotinamide, antibiotic/antimycotic solution (Invi-
trogen), and 10% FBS.
12. 10% FBS-DMEM (used as BEC washing medium and FLC
decantation medium).
13. Mouse anti-CK19 antibody (Novocastra) (used at 1:50 dilu-
tion for IHC).
14. Goat anti-CK18 antibody (Santacruz) (used at 1:200 dilution
for IHC).
15. 4% paraformaldehyde (PFA) (Wako).

2.2 Equipment 1. Peristaltic pump (EYELA, Tokyo, Japan, catalog number,

RP-1000).
2. Silicon tube (Ø4.76
7.94 mm2) (EYELA, catalog number,
125540).
3. Cannula Ø1.2 mm (plastic outer needle equipped with intra-
venous cannula (Top Corporation, Tokyo, Japan).
4. Incubator shaker (Bioshaker VBR-36 (Taitech)): An equivalent
machine can be used.
5. Autoclaved 100 mL glass conical flask.
6. 40 μm cell strainer (BD).
7. 70 μm cell strainer (BD).
8. 25 mL reservoir (Nunc).
9. Eight-channel pipette (volume range, 10–50 μL, 30–300 μL)
(Thermo).
178 Takeshi Katsuda et al.

10. 96-well, round-bottom plate coated with

2-methacryloyloxyethyl phosphorylcholine (MPC) (NOF
CORPORATION).

3 Methods

3.1 Extraction of the This procedure is performed in parallel with fetal liver extraction
Biliary Trees and (see Subheading 3.2) using one pregnant rat.
Isolation of BECs 1. Anaesthetize a pregnant rat by inhalation of isoflurane vapor.
Place a paper towel cushion on a foam board or cork board, and
lay the rat on the cushion. Sterilize the animal by wiping its fur
with 70% EtOH. Open the abdomen using surgical scissors.
2. Start drip of pre-perfusion buffer at 25–30 mL/min. Half-cut
the portal vein at the location approximately 1.5 cm from the
bifurcation of the portal vein, and insert the cannula.
3. Immediately after confirming that the color of the liver changes
from reddish brown to yellowish, cut the inferior vena cava.
Perfuse approximately 450 mL pre-perfusion buffer (approxi-
mately 15 min).
4. Pause the pump, transfer the inlet tube to the collagenase
bottle, and resume perfusion. Perfuse approximately
300–350 mL collagenase (approximately 10 min).
5. Stop the pump, and remove the cannula from the portal vein.
6. Peel the liver capsule, and gently mash the parenchyma with
round-edged tweezers to release the parenchymal cells from
the tree structures. Gently flush away the digested parenchymal
tissue with E-MEM until white tree structures are visible
(Fig. 1a–e).
7. Resect the tree structures, and place them in bile duct wash
medium. At this moment, the tree structures still hold paren-
chymal cells as noticed by their brownish appearance.
8. Purify the tree structure ex vivo by further removing the
remaining parenchyma until most brownish-colored parts are
removed (Fig. 1f). Gently pressing the tissue with round-edged
tweezers helps releasing of parenchymal cells from the tree
structures.
9. Collect the purified tree fragments into multiple 1.5 mL tubes
so that each tube contain 150–300 μL of fragments (usually
divided into two to four tubes) (see Note 1), and mince them in
the 1.5 mL tubes with surgical scissors into small fragments
(approximately <0.5 mm). From this step, work aseptically in a
clean bench or a safety cabinet.
10. Suspend the minced fragments in 10 mL BEC digestion
medium, and divide it to two 50 mL tubes.
 Generation of Hepatic Organoids with Biliary Structures 179

Fig. 1 Graphical schematic of extraction of intrahepatic bile duct trees. (a) Just after switching from
pre-perfusion buffer to collagenase solution and (b) after completion of collagenase perfusion. Note that the
liver is loosened and the subcapsular tissue is pulpy compared to (a). (c) By peeling the liver capsule and
gently mashing the parenchyma with round-edged tweezers, parenchymal tissue can be liberated from the
tree structure. (d) Flushing the parenchymal tissue with E-MEM by pipetting. Lifting up the rat’s front paws will
help this procedure. (e) White intrahepatic biliary trees can be visible after removal of the parenchymal tissue.
(f) The purified tree structures after mechanically removing the remaining parenchyma using tweezers ex vivo

11. Shake the tubes at 37  C in a shaking incubator for 30 min

while making sure that the content is homogeneously mixed.
12. Filtrate the digested tissue with a 40 μm cell strainer, and
combine dispersed cells from two tubes into a new 50 mL tube.
13. Centrifuge the tube at 800
g for 10 min at 4  C.
180 Takeshi Katsuda et al.

14. Resuspend in 10 mL of 10% FBS-DMEM.

15. Centrifuge the tube at 800
g for 5 min at 4  C.
16. Resuspend in 10 mL of 10% FBS-DMEM.
17. Count the approximate number of viable cells using a hemocy-
tometer and trypan blue solution (see Note 2).
18. Centrifuge the tube at 800
g for 5 min at 4  C.
19. Resuspend the cells in appropriate volume of BEC culture
medium to obtain 0.5–2
106 cells/mL suspension. Keep
on ice until seeding.

3.2 Extraction of While perfusing the adult liver, proceed to fetal liver extraction in
Fetal Livers and parallel.
Purification of FLCs
1. Remove the uterus containing fetuses from the body of the
mother rat, and transfer them to a 10 cm dish filled with PBS.
2. Extract fetuses from the uterus using surgical scissors and
tweezers, and transfer them to a new 10 cm dish filled
with PBS.
3. Extract the livers from the fetuses using tweezers, and transfer
the fetal livers to a new 10 cm dish filled with PBS.
4. Remove the white tissue surrounding the red livers as
completely as possible using tweezers. Divide the purified livers
into 1.5 mL tubes so that each tube contains 150–300 μL (two
to five tubes, dependent on the number of fetuses) (see Note
1), and mince them with surgical scissors into small fragments
(approximately <0.5 mm).
5. Transfer the minced tissue into a 15 mL tube, add PBS to
adjust the total volume to 12 mL, and stand the tube until
the minced tissues fall to the bottom (3–5 min) (see Note 3).
From this step, work aseptically in a clean bench or a safety
cabinet. Carefully discard the supernatant using a pipette, and
add 12 mL PBS.
6. Repeat the wash process in step 5 until the color of the super-
natant changes from red to clear (usually two to three time
repeats are sufficient).
7. Suspend the minced tissue in 2.5 mL/tube PBS, and add
2.5 mL of 0.1% collagenase solution (final concentration:
0.05%).
8. Shake the tube at 37  C in a shaking incubator for 10 min.
Make sure that the content is homogeneously stirred.
9. Let the tube stand for approximately 10 s, and retrieve the
supernatant into a new tube while leaving the remaining frag-
ments (see Note 4).
10. Centrifuge the tube at 200
g for 2 min at 4  C.
 Generation of Hepatic Organoids with Biliary Structures 181

11. Discard the supernatant, and resuspend the cells in 10 mL of

10% FBS-DMEM.
12. Filter the cells with a 70 μm cell strainer into a new 100 mL
tube. Wash the cell strainer by flushing it with fresh 90 mL 10%
FBS-DMEM.
13. Decant the 100 mL cell suspension into an autoclaved glass
conical flask. Let the flask stand at room temperature for
15 min. This decantation step helps removal of blood cells.
14. Carefully discard the upper 70 mL of the suspension, in which
blood cells are enriched. Collect the bottom 30 mL suspension
which is rich in FLCs into a new tube.
15. Count the viable cell number using a hemocytometer and
trypan blue solution.
16. Centrifuge the tube at 200
g for 2 min at 4  C.
17. Resuspend the cells in appropriate volume of FLC culture
medium to obtain 0.5–2
106 cells/mL suspension. Keep
on ice until seeding.

3.3 Formation of To obtain well-developed ductular structures, 50–75% of the mixed

BEC-FLC Hetero- cells should be BECs [8]. Recommended seeding density is
spheroids 1–20
104 cells/well of 96-well plate. In our previous report,
we generated hetero-spheroids by seeding 2
105 cells/well in
total, and we observed formation of ductular networks
(see Note 5).
1. Pipette each cell suspension, and combine BECs and FLCs in a
new tube at a desired mixing ratio.
2. Fill a 25 mL reagent reservoir with the mixed cell suspension.
Seed to a round-bottom MPC-coated 96-well plate using an
eight-channel pipette (see Note 6). Culture the cells at 37  C in a
CO2 incubator. Thereafter, replace the medium with fresh FLC
culture medium every day as described below. To confirm that
BECs and FLCs are successfully harvested, it is recommended to
seed a small portion of remaining cells to collagen-coated plates,
and observe the morphologies of each cell type (Fig. 2).
3. On days 1 and 2, BECs and FLCs generate loose pancake-
shaped aggregates. Since they are still very fragile, carefully
replace the medium: remove approximately 70 μL of the culture
supernatant, and add 100 μL fresh FLC culture medium into
each well slowly not to mix up the cells.
4. Within 3 days of culture, the loose aggregates are compressed to
form compact spheroids (Fig. 3). On day 3, remove cells that do
not contribute to spheroid formation: stir the medium by pipet-
ting with P200 tips (see Note 7), let the spheroids fall to the
bottom, and remove the supernatant containing cells which have
not contributed to spheroid formation.
182 Takeshi Katsuda et al.

Fig. 2 Morphology of FLCs and BECs. (a, b) Phase-contrast images of E17.5 FLCs on day 2 of culture. (c, d)
Phase-contrast images of BECs isolated from the adult rat liver on day 3 of culture. This figure is reproduced
from ref. [8]

5. On days 3–6, fix the spheroids in 4% PFA, embed them in

paraffin blocks, and prepare sections with approximately 5 μm
thickness according to the general method (see Note 8). In the
reconstructed hetero-spheroids, CK18+/CK19+ biliary struc-
tures develop mainly inside the spheroids, while CK18+/
CK19 hepatic structures are located mainly at outer regions
(Fig. 4) (see Note 9).

4 Notes

1. Too much volume in a 1.5 mL tube disturbs efficient mincing.

2. Because it is difficult to completely dissociate biliary fragment
into single cells, it is impossible to determine the precise cell
number.
 Generation of Hepatic Organoids with Biliary Structures 183

Fig. 3 Time course of hetero-spheroid formation. Phase-contrast images of hetero-spheroids with the
composition of “FLC:BEC ¼ 1:1” and “FLC:BEC ¼ 1:3” hetero-spheroids on day 1 (a, b), day 2 (c, d), and
day 3 (e, f). This figure is reproduced from ref. [8]

3. If the minced tissue volume is larger than 500 μL, aliquot the
suspension to multiple tubes so that each tube contains less than
500 μL.
4. White tissue will be visible after digestion. Do not let the tube
(s) stand too long to avoid loss of FLCs.
5. At the cell density of 2
105 cells/well, we observed necrosis at
the central area of the spheroids, most likely because each spher-
oid contains too much cells. We suppose that a lower cell density
184 Takeshi Katsuda et al.

Fig. 4 Histological analyses of hetero-spheroids on day 3. (a, b) HE staining (nuclei, blue; cytoplasm,
red-purple). Arrows and arrowheads indicate duct- and cyst-like structures, respectively. (c, d) IHC of CK19
(brown). Nuclei were counterstained with hematoxylin (blue). Arrowheads indicate continuous ductular net-
works formed by CK19 (+) BECs. (e, f) IHC of CK18 (brown). Nuclei were counterstained with hematoxylin
(blue). Scale bar: 50 μm. CK cytokeratin, HE hematoxylin-eosin, IHC immunohistochemistry. This figure is
reproduced from ref. [8]

would be better, although we have not yet optimized the num-

ber of seeding cells.
6. To avoid letting the cells fall to the bottom, seed cells as soon as
possible.
7. Be careful not to break down the spheroids by touching them
with the P200 tips.
 Generation of Hepatic Organoids with Biliary Structures 185

8. When handing spheroids, we recommend to use nylon biopsy

bags to reduce the risk of small specimen loss.
9. When performing immunohistochemistry, heat-induced anti-
gen retrieval should be performed before antibody reaction.
We usually heat de-paraffinized specimens in 1/200 diluted
ImmunoSaver (Nissin EM Ltd.) at 98  C using an electronic
pot for 45 min.

Acknowledgments

This work was supported in part by Grant-in-Aid for Young Scien-

tists B (16K16643) and the Research Program on Hepatitis
(16fk0310505h0005) from the Japan Agency for Medical Research
and Development (AMED).

References
1. Ohashi K, Yokoyama T, Yamato M et al (2007) 3D bile duct tissue from human pluripotent
Engineering functional two- and three- stem cell-derived spheroids. Stem Cell Rev Rep
dimensional liver systems in vivo using hepatic 12:500–508. https://doi.org/10.1007/
tissue sheets. Nat Med 13:880–885. https:// s12015-016-9657-5
doi.org/10.1038/nm1576 6. Sampaziotis F, De Brito MC, Madrigal P et al
2. Takebe T, Sekine K, Enomura M et al (2013) (2015) Cholangiocytes derived from human
Vascularized and functional human liver from an induced pluripotent stem cells for disease mod-
iPSC-derived organ bud transplant. Nature eling and drug validation. Nat Biotechnol
499:481–484. https://doi.org/10.1038/ 33:845–852. https://doi.org/10.1038/nbt.
nature12271 3275
3. Sakai Y, Yamanouchi K, Ohashi K et al (2015) 7. Dianat N, Dubois-Pot-Schneider H, Steichen C
Vascularized subcutaneous human liver tissue et al (2014) Generation of functional
from engineered hepatocyte/fibroblast sheets cholangiocyte-like cells from human pluripotent
in mice. Biomaterials 65:66–75. https://doi. stem cells and HepaRG cells. Hepatology
org/10.1016/j.biomaterials.2015.06.046 60:700–714. https://doi.org/10.1002/hep.
4. Jiang J, Kojima N, Guo L et al (2004) Efficacy of 27165
engineered liver tissue based on poly-L-lactic 8. Katsuda T, Kojima N, Ochiya T, Sakai Y (2013)
acid scaffolds and fetal mouse liver cells cultured Biliary epithelial cells play an essential role in the
with oncostatin M, nicotinamide, and dimethyl reconstruction of hepatic tissue with a functional
sulfoxide. Tissue Eng 10:1577–1586. https:// bile ductular network. Tissue Eng Part A
doi.org/10.1089/ten.2004.10.1577 19:2402–2411. https://doi.org/10.1089/ten.
5. Tian L, Deshmukh A, Ye Z, Jang YY (2016) TEA.2013.0021
Efficient and controlled generation of 2D and
 Chapter 17

Analysis for Remodeling of Hepatic Tissue Structures in 3D

During Regeneration
Kota Kaneko

Abstract
It has been demonstrated that the liver remarkably alter its tissue structures during regeneration, and
various types of liver stem or progenitor cells locating in specific areas in the liver tissue contribute to
regeneration. Therefore, it is important to analyze the dynamic rearrangement of the liver tissue structures
in 3D for better understanding the process and mechanism of liver regeneration. Here we describe a
macroscopic analysis method to visualize the whole 3D structure of the vasculatures and the biliary tree,
which are dynamically remodeled during regeneration, and a microscopic analysis method to visualize
detailed structures at the cellular level, which can compensate what cannot be detected in the macroscopic
analysis.

Key words Liver regeneration, Liver pathology, 3D tissue structure, 3D imaging, Ductular reaction,
Biliary tree

1 Introduction

Accumulating evidences have shown that the liver regeneration is

supported by various types of stem or progenitor cells locating
differentially in the tissue structure. For example, residential liver
stem/progenitor cells are thought to be located near or inside the
bile duct epithelium [1]. These cells are considered to contribute to
regeneration, when the proliferative ability of mature hepatocytes is
compromised. On the other hand, it has been suggested that
pericentral or periportal hepatocytes have more regenerative poten-
tial than others [2–4]. Because of the diverse locations of these
progenitors, it is necessary to analyze the regenerative processes
correlating with the tissue structure. Moreover, since the liver tissue
is organized in three dimension (3D), 3D analyses rather than
histological analysis with 2D sections are needed for identifying
the spatial localization of liver progenitor cells.
Although the liver is one of the largest organs, it has a repetitive
structure called lobule. The lobules are defined by portal triads and

Naoki Tanimizu (ed.), Hepatic Stem Cells: Methods and Protocols, Methods in Molecular Biology, vol. 1905,
https://doi.org/10.1007/978-1-4939-8961-4_17, © Springer Science+Business Media, LLC, part of Springer Nature 2019

187
188 Kota Kaneko

Macroscopic structure Microscopic structure

Zonation

Portal Bile duct Central

vein vein

Fig. 1 The macroscopic and the microscopic structure of the liver. Bile ducts (green), portal vein (magenta),
central vein (blue), sinusoids (red), hepatic artery (pink), neurons (yellow), and other immune and mesenchy-
mal populations (orange, dark blue, and purple) are shown

the central veins, which are the landmarks to tell the “zonation.”
Therefore the whole liver structure can roughly be regarded as the
collections of the lobules defined by vasculatures (Fig. 1). Around
the portal vein exist bile ducts, hepatic artery, and the peripheral
nerve systems. Importantly, bile ducts dynamically change the
structure during liver injury and regeneration, which is called duct-
ular reaction [5]. Recent studies have shown that bile ducts can
extend their branches as far as to the pericentral area in response to
injury, independently of the vascular structure [6–8]. Together
with the classic view of the continuous epithelial structure of hepa-
tocytes and biliary epithelial cells [1] (Fig. 2), such remodeling of
the bile duct structure needs to be taken into consideration when
assessing any spatial relationships. We recommend using portal
veins, central veins, and bile ducts as basic landmarks to determine
the locations of objects in the liver tissue.
There are many different methods to analyze the 3D architec-
ture of the liver. When we select a particular method, it is necessary
to consider the actual convenience for using the method, technical
limitations of each method (maximum scale, detectable cell types,
compatibility with fluorescent proteins such as GFP, etc.), and the
resolution. Among them, resolution should be particularly impor-
tant to visualize the actual tissue structures. In fact, the analysis of
the biliary tree architecture with an unprecedented high-resolution
analysis revealed the actual heterogeneous structure (Fig. 3) and
the dynamic rearrangements of the bile ducts [6, 9]. In contrast, as
shown in Fig. 3, it is hard to know the correct tissue structures and
their changes by low-resolution analyses. Since only large ducts are
observed in a low-resolution analysis, bile ducts are mostly depicted
in textbooks as a single tube along the vein. Thus, high-resolution
analyses do not only lead to new discoveries but are also crucial for
getting the correct conclusions.
In this chapter, we will introduce macroscopic and microscopic
3D analysis methods, which should preferably be combined
together to cover most of our purposes for analyzing the tissue
 3D Analysis of the Liver Tissue Structure 189

Bile duct Bile canaliculi

Hepatocyte

Portal Central
vein vein

Biliary epithelial cell Canal of Hering

Fig. 2 The continuous epithelial tissue structure of the liver. The epithelial tissue
of the liver consists of hepatocytes and biliary epithelial cells, which are lined up
as a continuous structure, connected at the canal of Hering. Bile juice is
produced by hepatocytes and drained into bile ducts through bile canaliculi.
Hepatocytes near the bile ducts are known to have different characteristics from
those located away from the bile duct

High-resolution Low-resolution

Biliary branching Biliary branching

< >

Fig. 3 Structure of the biliary tree (green) with the portal vein (pink). Bile ducts have heterogeneous structure
consisting of thick tubes running along with the portal vein and fine tubes branching and wrapping around the
portal vein. High-resolution analysis clearly shows the heterogeneous network (left panel), whereas the fine
tubular network is hardly visualized with low-resolution analysis (right panel). Therefore, they possibly lead to
different conclusions

structure during regeneration. Macroscopic analysis visualizes the

whole vasculature and the biliary tree structure in the entire mouse
liver (Fig. 4). Briefly, ink is injected into the biliary tree retrogradely
through the extrahepatic biliary tree using a glass capillary. Then
the harvested liver is made optically transparent using an organic
reagent. Microscopic analysis visualizes all kinds of structures in
high resolution as long as antibodies are available. Objects in thick
190 Kota Kaneko

PV

CV

PV
PV CV

CV

Fig. 4 Macroscopic observation of the liver tissue structure by ink injection. Bile
ducts (black) are running along and wrapping around the portal vein. Portal veins
(PV) and central veins (CV) can be seen with the red blood in them

sections are labeled by immunostaining and imaged by optical

sectioning and reconstituting with a confocal microscope (Fig. 5),
which gives much higher resolution along the Z-axis as compared
to stacking serial sections.

2 Materials

2.1 Macroscopic 1. Mice.

Analysis 2. Hematocrit tubes for making capillaries (Fig. 6): Microhema-
tocrit capillary tubes (22-362-566; Fisher Scientific) work well,
though any other glass tube might work.
3. Syringe: Syringes with smaller volumes (typically around 1 mL)
are better for controlling the air pressure.
4. Tube to connect the glass capillary with a syringe: Cut a but-
terfly needle to get a tube with a socket to connect with a
syringe (Fig. 6).
5. Ink: Black waterproof fountain pen ink, SPC-200 (Platinum
Japan), is suitable for the visualization (see Note 1).
6. PBS.
7. Ethanol.
8. BABB: Mix 500 mL of benzyl benzoate and 250 mL of benzyl
alcohol (see Note 2).
9. 20% sucrose in PBS.
10. O.C.T. compound.
 3D Analysis of the Liver Tissue Structure 191

Fig. 5 Microscopic 3D reconstitution of the liver tissue structure. This image shows the biliary epithelial cells
(white) in the CDE diet model liver. Veins (pink) can be reconstituted by inverting the autofluorescence of the
tissue. Since this method is based on immunostaining, it can show subcellular structures of any types of liver
cells as long as antibodies are available

2.2 Microscopic 1. Frozen liver tissue (see Note 3).

Analysis 2. Antibody reaction solution: Dissolve antibodies in PBS contain-
ing 5% BSA, 0.1% Triton X-100, and 0.01% sodium azide. This
reagent can be reused until the antibodies are consumed.
192 Kota Kaneko

Hematocrit tube Melt

Stretch

break
Connect

butterfly needle
Syringe
Cut

“Capillary”

Fig. 6 Glass capillary for injection. Ink can be loaded into the capillary and
injected into the bile duct by manipulating the syringe

The concentration of the antibodies should be similar to that for

a regular immunostaining or half of that. The total volume
needed for one thick section is usually 500–1000 μL.
3. Washing buffer: PBS containing 0.1% Triton X-100.
4. Fixation buffer: 4% PFA in PBS (pH 7.4) or 2% PFA 15% picric
acid in PBS (pH 7.4).
5. Denaturing buffer: PBS containing 5% β-mercaptoethanol
(ME) and 0.1% Triton X-100.
6. Counter staining solution: PBS containing 0.1% Triton X-100
with dyes such as DAPI or Hoechst.
 3D Analysis of the Liver Tissue Structure 193

7. Clearing/mounting reagent: Choose one from 2,2-

0
-thiodiethanol (TDE), saturated sucrose solution in PBS, or
SeeDB [10].
8. Confocal microscope.

3 Methods

3.1 Macroscopic 1. Prepare capillaries (Fig. 6) with different thicknesses by melting

Analysis the hematocrit tube in the middle with a gas burner or a candle
flame and pulling the both ends apart to stretch the tube (see
Note 4). Once getting the thinner part in the middle of the
tube, break the part to make a capillary for insertion into the
extrahepatic bile duct.
2. Euthanize the mouse, and open the abdominal cavity. Expose
the liver, the portal vein, and the extrahepatic bile duct.
3. Identify the junction between the extrahepatic bile duct tube
and the intestine (Fig. 7a). Select a capillary with a proper
thickness that fits the bile duct tube (see Note 5).
4. Connect the glass capillary to a tube from a butterfly needle
connected to a syringe (Fig. 6). Load the capillary with the ink
(see Note 6).
5. Cut the junction between the bile duct and the intestine to
make a hole for insertion of the capillary (Fig. 7b).
6. Pull the bile duct tube with sharp tweezers without blocking
the hole, and insert the capillary loaded with the ink (Fig. 7c)
(see Note 7).
7. Inject the ink very slowly while carefully observing the surface
of the liver. The gallbladder will turn black first, when the
capillary is successfully inserted into the bile duct. Stop the
injection, when tiny black dots appear on the surface of the
liver (see Note 8).
8. Harvest the liver and immerse into PBS in a petri dish. Place the
liver between cotton pads in the PBS to fix the liver posture, if
desired.
9. Place the petri dish on a shaker and gradually add ethanol to the
PBS in a petri dish. Increase the ethanol concentration until
finally replacing the whole PBS with ethanol (see Note 9). For
instance, PBS for 1 min, 15% EtOH in PBS for 10 min, 30%
EtOH in PBS for 10 min, 40% EtOH in PBS for 20 min, 60%
EtOH in PBS for 30 min, 80% EtOH in PBS for 30 min, and
100% EtOH for 30 min three times.
10. After complete dehydration, immerse the liver in BABB.
Refresh the BABB solution until the liver becomes completely
transparent (see Notes 2 and 10). With a whole liver in 50 mL
194 Kota Kaneko

a Identify the extrahepatic bile duct Extrahepatic bile duct

and the junction to the intestine
in
te
st
in
e

Portal vein

intstine

b Cut the junction

c Insert the capilary Tweezers

Capillary

Fig. 7 Procedures of the ink injection into the biliary tree. (a) Extrahepatic bile duct (arrowheads) is running
from the liver to the intestine on the side of the portal vein. The bile duct has a bulge (arrow and dashed line) at
the junction with the intestine. (b) Cutting the bulge at the junction. (c) Insertion of the capillary into the bile
duct tube

tube, the treatment in 20 mL BABB for 1 h is repeated three

times.
11. Observe and take pictures of the liver (see Note 11).
12. The liver after step 11 can be immersed back in ethanol and
then PBS, 20% sucrose in PBS and be embedded in
O.C.T. compound and frozen for sectioning and
immunostaining.
13. Successful ink injection can be evaluated by the coexistence of
biliary markers and ink on the liver sections.
 3D Analysis of the Liver Tissue Structure 195

3.2 Microscopic 1. Make sections of the liver in around 200 μm thickness by

Analysis cryostat (see Note 12). Directly drop the frozen section in the
cold PBS in 50 mL tube, and unroll the section by gently
inverting the tube. If the tissue was embedded in
O.C.T. compound, the tissue section is released from the com-
pound during this step.
2. Substitute the PBS with 25 mL of cold washing buffer. Rock
the tube for 1 h at 4  C.
3. Discard the liquid, leaving 1 mL. With this 1 mL buffer,
transfer the liver section to 5 mL tube or proper sized/shaped
container. Discard all the liquid and add the fixation buffer and
incubate for overnight at 4  C, if the tissue is not prefixed. For
prefixed tissues, add the denaturing buffer for antigen
retrieval/permeabilization, and incubate for 1 h at 37  C.
4. Transfer the section to 50 mL tube with the liquid, discard the
liquid, and wash the section with 25 mL of cold washing buffer
for more than 1 h three times and final wash for overnight,
rocking.
5. Discard the liquid, leaving 1 mL. With this 1 mL buffer,
transfer the liver section to 5 mL tube or proper sized/shaped
container. Discard all the liquid and add the antibody reaction
solution.
6. Incubate for more than 2 days at 4  C, rocking the container
(see Note 13). Occasional manual rocking and flipping the
section gives better results.
7. Transfer the antibody reaction solution to another tube for
reusing next time. Add 1 mL of cold washing buffer to the
section and transfer them to 50 mL tube. Wash the section
three times in 25 mL of cold washing buffer with rocking for
more than 1 h.
8. Repeat steps 5–7 for the reaction with secondary antibodies.
Then wash the section with cold washing buffer overnight.
9. Discard the liquid, leaving 1 mL. With this 1 mL buffer,
transfer the liver section to 5 mL tube or proper sized/shaped
container. Discard all the liquid and add the counter-staining
solution.
10. Wash the section with cold washing buffer half an hour three
times.
11. Discard the liquid, leaving 1 mL. With this 1 mL buffer,
transfer the liver section on to a glass slide. Remove the liquid
with pipette, and add 50 μL of the clearing/mounting reagent.
Replace the liquid with the clearing/mounting reagent until
the section becomes evenly clear.
196 Kota Kaneko

12. Cover the section with a glass cover slip (see Note 14), and
acquire images with a confocal microscope.

4 Notes

1. The main component of this ink is “carbon black,” fine parti-

cles of amorphous carbon dissolved in a detergent. When
injected into the biliary tree, the carbon particles get stuck
inside the biliary tube, while the solvent in the ink leaks out
to the portal veins and the bile canaliculi. This makes the biliary
tube filled with carbon particles. Thus, any other ink with the
similar components may work, but the particle size of the
carbon black, which differs according to the manufacturing, is
crucial for getting the optimal result.
2. BABB melts plastics except polypropylene. Therefore glass or
polypropylene should be used to contain BABB.
3. Liver tissues are fixed or non-fixed before frozen depending on
the antibody used.
4. Once the glass melts and becomes very unsolid on the fire,
displace the tube away from the flame to let it slightly solidify
for around 0.2–1 s. When the melted part becomes “resilient”
or “viscous,” pull apart both ends of the tube. The thickness
and length of the capillary are determined by the solidity of the
melted part on the moment it is stretched. Longer capillary
part with the similar thickness as the extrahepatic bile duct is
desirable. Since the thickness of the extrahepatic bile duct
differs with each mouse, it is important to prepare capillaries
with different thicknesses.
5. It is important to use a capillary with a proper thickness that fits
the thickness of the biliary tube. If the capillary is too thin
and/or sharp, it easily penetrate through tissues so that it
would get into wrong places or break the wall of the extrahe-
patic bile duct.
6. Ink should be carefully loaded on to the bottom half (or 80%)
of the capillary/tube without ever passing the half (or 80%)
point. Otherwise the capillary/tube will be coated with the
dark ink, and the surface of the loaded ink becomes difficult
to recognize. This surface is an important indicator of how fast
the ink is moving while injecting it into the biliary tree. To
prevent the overload, level the air pressure inside the syringe to
zero or slightly positive before taking the capillary out of the
ink. This prevents the suction of the air into the capillary, which
happens much quicker than the suction of the ink.
7. When inserting the capillary into the bile duct tube, the air
pressure inside the syringe should be zero or slightly positive to
 3D Analysis of the Liver Tissue Structure 197

avoid loading of air or any tissue fragments on the capillary tip.

Poking tissues randomly with the capillary during the insertion
into the bile duct may insert tissue fragments inside the capil-
lary. This can cause clogs inside the intrahepatic biliary tree
during ink injection, preventing uniform ink loading through-
out the entire liver.
8. While injecting the ink, adjusting the posture of the liver may
help the well-distributed injection over the entire liver. Exces-
sive ink load leads to leakage of the ink (carbon particle) to the
bile canaliculi and especially portal veins. The amount of the ink
needed for one adult mouse is usually around 15–20 μL, but it
depends on every individual, and the injection amount should
be determined based on observation of the liver surface during
injection. Tiny black dots appearing on the liver surface are a
sign to stop injecting.
9. Rapid increase of the ethanol concentration will make strongly
dehydrated parts which block penetration of the liquid into
other parts, leading to an insufficient dehydration. It is impor-
tant to make sure the color of the whole liver is gradually
turning to white evenly as dehydration proceeds. Sufficient
dehydration is curial to clear the tissue by BABB.
10. Used BABB contaminated with ethanol can be reused for
initial steps of clearing in step 10.
11. Blood remained in the vasculature visualizes the portal veins
and the central veins by the transparent red color. The portal
veins and the central veins can easily be distinguished by the
extrahepatic bile duct entering the liver along with the portal
vein. Vasculature can also be stained with perfusion of any ink
before the ink injection into the biliary tree. However, strong
stains on the vasculature make the observation difficult, due to
the dense existence of the vasculature in the whole liver.
12. Higher cutting temperature makes it easier to get thick sec-
tions. The temperature depends on the cryosectioning
machine used, but it is usually optimal around 15  C.
13. Incubation with primary antibodies can be performed longer
(3 days to 1 week) for better results, while incubation with
secondary antibodies may be optimal with just 2 days.
14. Insert something (glass cover slips work fine) between the
cover slip and the glass slide near the edge of the cover slip to
avoid pressing the thick section by the cover slip.

Acknowledgments

The experiments to establish these methods were supported in part

by Japan Society for the Promotion of Science KAKENHI grants
198 Kota Kaneko

22118006 (to Dr. Atsushi Miyajima), 24112507 (to Dr. Tohru

Itoh), 24590342 (to T.I.), and 26253023 (to A.M.) as well as by
CREST from the Japan Science and Technology Agency (to A.M.).
K. K. was a research fellow at the Japan Society for the Promotion
of Science.

References
1. Miyajima A, Tanaka M, Itoh T (2014) Stem/ reactions in human livers. Hepatology (Balti-
progenitor cells in liver development, homeo- more, MD) 39(6):1739–1745. https://doi.
stasis, regeneration, and reprogramming. Cell org/10.1002/hep.20130
Stem Cell 14(5):561–574. https://doi.org/ 6. Kaneko K, Kamimoto K, Miyajima A, Itoh T
10.1016/j.stem.2014.04.010 (2015) Adaptive remodeling of the biliary
2. Font-Burgada J, Shalapour S, Ramaswamy S, architecture underlies liver homeostasis. Hepa-
Hsueh B, Rossell D, Umemura A, Taniguchi K, tology (Baltimore, MD) 61(6):2056–2066.
Nakagawa H, Valasek Mark A, Ye L, Kopp https://doi.org/10.1002/hep.27685
Janel L, Sander M, Carter H, Deisseroth K, 7. Kamimoto K, Kaneko K, Kok CYY, Okada H,
Verma Inder M, Karin M (2015) Hybrid peri- Miyajima A, Itoh T (2016) Heterogeneity and
portal hepatocytes regenerate the injured liver stochastic growth regulation of biliary epithe-
without giving rise to cancer. Cell 162 lial cells dictate dynamic epithelial tissue remo-
(4):766–779. https://doi.org/10.1016/j.cell. deling. eLife 5:e15034
2015.07.026 8. Nagahama Y, Sone M, Chen X, Okada Y,
3. Tanimizu N, Ichinohe N, Yamamoto M, Yamamoto M, Xin B, Matsuo Y, Komatsu M,
Akiyama H, Nishikawa Y, Mitaka T (2017) Suzuki A, Enomoto K, Nishikawa
Progressive induction of hepatocyte progenitor Y. Contributions of hepatocytes and bile duct-
cells in chronically injured liver. Sci Rep ular cells in ductular reactions and remodeling
7:39990. https://doi.org/10.1038/ of the biliary system after chronic liver injury.
srep39990 Am J Pathol 184(11):3001–3012. https://doi.
4. Wang B, Zhao L, Fish M, Logan CY, Nusse R org/10.1016/j.ajpath.2014.07.005
(2015) Self-renewing diploid Axin2(+) cells 9. Tanimizu N, Kaneko K, Itoh T, Ichinohe N,
fuel homeostatic renewal of the liver. Nature Ishii M, Mizuguchi T, Hirata K, Miyajima A,
524(7564):180–185. https://doi.org/10. Mitaka T (2016) Intrahepatic bile ducts are
1038/nature14863 developed through formation of homogeneous
5. Roskams TA, Theise ND, Balabaud C, continuous luminal network and its dynamic
Bhagat G, Bhathal PS, Bioulac-Sage P, Brunt rearrangement in mice. Hepatology (Balti-
EM, Crawford JM, Crosby HA, Desmet V, more, MD) 64(1):175–188. https://doi.org/
Finegold MJ, Geller SA, Gouw AS, 10.1002/hep.28521
Hytiroglou P, Knisely AS, Kojiro M, Lefko- 10. Ke M-T, Fujimoto S, Imai T (2013) SeeDB: a
witch JH, Nakanuma Y, Olynyk JK, Park YN, simple and morphology-preserving optical
Portmann B, Saxena R, Scheuer PJ, Strain AJ, clearing agent for neuronal circuit reconstruc-
Thung SN, Wanless IR, West AB (2004) tion. Nat Neurosci 16:1154. https://doi.org/
Nomenclature of the finer branches of the bili- 10.1038/nn.3447
ary tree: canals, ductules, and ductular
 Part V

Liver Injury Models

 Chapter 18

Canine Liver Fibrosis Model to Assess the Functions

of Infused Autologous Bone Marrow-Derived Cells
Taro Takami, Kenji Tani, Yasuho Taura, and Isao Sakaida

Abstract
Liver cirrhosis is the end stage of chronic liver disease, and the only radical treatment for decompensated
liver cirrhosis is still liver transplantation. Development of effective regenerative therapy for liver cirrhosis is
an urgent task. Before human clinical trials can be considered, the safety and efficacy of any planned
protocol must be confirmed in medium-to-large animals. Therefore, we have developed a novel canine
liver fibrosis model for proof of concept (POC) studies.

Key words Liver cirrhosis, Canine model, Mesenchymal stem cells, Carbon tetrachloride

1 Introduction

Liver cirrhosis (LC) is the end stage of chronic liver disease caused
by viral hepatitis, excess consumption of alcohol, obesity, and so
on. Although the liver is an organ with a high regenerative capacity,
chronic inflammation results in the development of cirrhosis, and
dense fibrosis ultimately halts hepatocyte proliferation. Currently,
the only radical treatment available for decompensated LC is liver
transplantation. However, liver transplantation has its own issues,
including donor shortage, immunorejection, and medical cost. To
overcome the issues related to liver transplantation, we aimed to
develop a liver regeneration therapy for decompensated LC patients
using autologous bone marrow cells (BMCs) [1]. In a murine LC
model created by repeated carbon tetrachloride (CCl4) treatment
[2], we previously confirmed that syngeneic non-cultured whole
BMCs that were infused through the tail vein had repopulated the
damaged liver and produced matrix metalloproteinases (MMPs),
ultimately reducing liver fibrosis [3]. Based on these evidences, we
conducted a clinical study and reported the clinical efficacy and
safety of “autologous bone marrow cell infusion therapy (ABMi
therapy)” as liver regeneration therapy for decompensated LC

Naoki Tanimizu (ed.), Hepatic Stem Cells: Methods and Protocols, Methods in Molecular Biology, vol. 1905,
https://doi.org/10.1007/978-1-4939-8961-4_18, © Springer Science+Business Media, LLC, part of Springer Nature 2019

201
202 Taro Takami et al.

[4–6]. However, the therapy involves bone marrow

(BM) aspiration under general anesthesia. Therefore, we aimed to
develop a less invasive therapy that uses cultured autologous
BM-derived mesenchymal stem cells (BMSCs) aspirated under
local anesthesia. We previously demonstrated that the peripheral
infusion of cultured BMSCs reduced liver fibrosis in an immunode-
ficient cirrhotic murine model [7] and thioacetamide (TAA)-
induced cirrhotic murine model [8]. However, before human clin-
ical trials can be considered, the safety and efficacy of the protocol
must be confirmed in medium-to-large animals. Therefore, we
aimed to develop a canine liver fibrosis model by CCl4 treatment.
It was difficult to make a canine take CCl4 orally, because of a keen
nose. We have finally established a novel canine liver fibrosis model
by repeated direct CCl4 injections into the stomach to demonstrate
the safety and efficacy of infusion of autologous BMCs for the
treatment of LC [9]. Here, we describe the method of creating
this canine liver fibrosis model.

2 Materials

2.1 Gastric Catheter 1. Beagles (1–2 years old): These canines were housed in our
Placement animal facility and treated in accordance with the university’s
animal care guidelines. This study was approved by our Institu-
tional Ethics Committee (approval number; 21-033).
2. Infusion port: A P-U Celsite port (Toray Medical Co. Ltd.,
Tokyo) consisting of a catheter coated with a heparinized
hydrophilic material and a Celsite port.
3. 18 G Teflon IV catheter (6-French P-U catheter).
4. Anesthesia machine: For delivering positive pressure
ventilation.
5. Propofol 1% injection: The loading dose of propofol is about
7 mg/kg body weight (BW).
6. Lepetan: Intravenous Lepetan (buprenorphine) is administered
at a dose of 10 μg/kg BW to provide effective pain relief.
7. Isoflurane: Anesthesia is maintained with isoflurane in oxygen.
The end-tidal isoflurane concentration is monitored and main-
tained between 1.4 and 2.8%.
8. Endobronchial tube for general anesthesia.
9. Chlorhexidine acetate.
10. 2% chlorhexidine gluconate in 70% isopropyl alcohol.
11. #21 scalpel.
12. Right-angled forceps.
13. Babcock forceps.
 Canine Liver Fibrosis Model 203

14. Suture material (3-0 PDS).

15. Povidone-iodine.
16. Upper gastrointestinal endoscope (Olympus Co., Tokyo).

2.2 Aspiration of 1. Medetomidine.

Bone Marrow Fluid, 2. Atipamezole hydrochloride.
Culture, and Cell
3. 16 G biopsy needle.
Infusion of Bone
Marrow Derived- 4. 5- and 20-mL syringe.
Mesenchymal Stem 5. Heparin solution.
Cells (BMSCs) 6. Saline.
7. Culture medium: 10% fetal bovine serum and 100 μg/mL
gentamicin in Dulbecco’s Modified Eagle Medium (DMEM).
8. Cell detachment solution: TrypLE Select (Thermo Fisher Sci-
entific Inc.).
9. Phosphate-buffered saline (PBS).
10. Culture flask: T-75 flasks.

2.3 Induction of Liver 1. Carbon tetrachloride (CCl4).

Fibrosis 2. Corn oil.
3. Winged Surecan port needle, 22 G  25 mm.

2.4 Liver Biopsy 1. Medetomidine.

2. Atipamezole hydrochloride.
3. Isoflurane: Anesthesia is maintained with isoflurane in oxygen.
The end-tidal isoflurane concentration is monitored and main-
tained between 1.4 and 2.8%.
4. 0.5% lidocaine for local anesthesia.
5. Povidone-iodine.
6. 16 G biopsy needle (Aragon Medical Devices, Plano, TX).
7. Ultrasound system: This consists of an ultrasound machine
(Aplio MX, Canon Medical Systems Co., Tokyo), probe
(PVT-745BTV convex electronic scanning probe, Canon Med-
ical Systems Co., Tokyo), and needle biopsy adaptor (UAGV-
029A, Canon Medical Systems Co., Tokyo).

2.5 Indocyanine 1. Indocyanine green.

Green (ICG) Test 2. ICG meter A-30 (Fuchu Giken, Inc., Tokyo).
3. 10 unit/mL heparin: 1000 units of heparin are mixed with
100 mL of physiological saline.

2.6 Sirius Red 1. Sirius red solution (1.3% aqueous picric acid/0.1% direct red
Staining 80).
204 Taro Takami et al.

2. 0.5% acetic acid.

3. 100% ethanol.
4. Xylene.

3 Methods

3.1 Gastric Catheter 1. Under general anesthesia, the canine is restrained in the right
Placement lateral decubitus position and shaved over a wide area from the
posterior margin of the scapula to the anterior margin of the
femur and from the linea alba (midline) to around 2 cm left of
the vertebral body.
2. Wearing sterile gloves, the operator picks up a chlorhexidine
acetate scrub in a piece of gauze and uses it to scrub the
operating field, foaming it thoroughly, and disinfects the area
with cotton soaked in 2% chlorhexidine gluconate in 70%
isopropyl alcohol. The operating field is then sprayed with
povidone-iodine and allowed to dry naturally, after which it is
draped.
3. An approximately 3-cm-long skin incision is made with a #21
scalpel at a point 2 cm caudal to the last rib, blunt dissection of
the subcutaneous tissue is performed using right-angled for-
ceps, and the abdominal external oblique muscle is grasped
with Babcock forceps and exposed. Right-angled forceps are
used to divide the muscles bluntly, and the abdominal internal
oblique muscle is similarly divided to approach the abdominal
cavity.
4. An assistant operates the upper gastrointestinal endoscope and
inflates the stomach, assisting the operator by making it easier
to grasp the serosal side of the stomach with Babcock forceps.
Part of the stomach is exposed extracorporeally and the stom-
ach is fixed to the abdominal wall by placing single ligatures of
suture material (3-0 PDS) at four places on the gastric wall;
then, the abdominal external oblique muscle around the loca-
tion is grasped with the Babcock forceps.
5. The catheter is inserted into the stomach using a sheath, guide-
wire, and dilator (Fig. 1a). Endoscopy is used to confirm that
the catheter tip does not extend past the cardiac orifice into the
esophagus (Fig. 1b).
6. An approximately 3-cm-long skin incision is made slightly to
the left of the eighth to tenth thoracic vertebrae, and blunt
dissection of the subcutaneous tissues is performed to create a
bed for the port (Fig. 1c). An approximately 1-cm-long skin
incision is made at a point exactly midway between the sites of
port placement and catheter insertion, and right-angled for-
ceps are passed through this incision to grasp the catheter and
guide it under the skin to the port placement site.
 Canine Liver Fibrosis Model 205

Fig. 1 Canine liver fibrosis model. (a) Infusion port: A P-U Celsite port. (b) The catheter tip in the stomach was
confirmed using endoscopy. (c) The port was placed in the back of the neck. Arrow shows the place of the port

7. The length of the catheter is adjusted, and the catheter is

connected to the port via a device that joins and anchors the
catheter and port.
8. The port is placed under the skin and subcutaneous tissue and
anchored by suturing it to the subcutaneous tissue with 3-0
PDS. The skin incisions are then sutured with single ligatures of
3-0 PDS.

3.2 Aspiration of 1. After intramuscular injection of 20 μg/kg medetomidine, the

Bone Marrow Fluid, canine is restrained in the right lateral decubitus position, and
Culture, and Cell an area approximately 4 cm in diameter in the center of the left
Infusion of BMSCs shoulder is shaved and disinfected.
2. The greater tubercle of the humerus is palpated, and after the
aspiration site has been decided, a tiny skin incision is made
with an 18 G needle.
3. After the humerus has been firmly immobilized with the right
hand, a 16 G biopsy needle is inserted into the greater tubercle.
Once it has passed through the cortical bone, the inner needle
is withdrawn, and the bone marrow is aspirated using a 5-mL
syringe filled with 1 mL of heparin.
4. The 16 G biopsy needle is withdrawn and manual compression
is applied for hemostasis, and then atipamezole hydrochloride
is intramuscular injected finally.
5. To culture bone marrow derived-mesenchymal stem cells
(BMSCs), the aspirated BM fluid is seeded into T-75 flasks
and cultured in culture medium in a 5% CO2 incubator at
37  C.
6. After 2 days of incubation, nonadherent cells are removed
during medium replacement. The culture medium is changed
every 2 days for 2–3 weeks.
7. Expanded BMSCs are detached using cell detachment solution
and then are collected. BMSCs are washed with PBS and resus-
pended in 20 mL saline. BMSCs are injected through the
peripheral vein slowly (around 2–5 mL/min) using 20-mL
syringe.
206 Taro Takami et al.

3.3 Induction of Liver 1. CCl4 is diluted 1:1 in corn oil before its intragastric injection
Fibrosis through the catheter.
2. A winged Surecan port needle is inserted into the subcutaneous
port, and the mixture of carbon tetrachloride and corn oil is
injected into the stomach (see Note 1).
3. Injections are repeated for 10 weeks using the implanted cath-
eter (high-dose period, 1.0 mL/kg BW once a week and
0.5 mL/kg BW once a week for 6 weeks; low-dose period,
0.25 mL/kg BW twice a week for 4 weeks) to induce liver
fibrosis [9]. After the infusion of BMSCs, low-dose CCl4 is
further administrated for 4 weeks.
4. The CCl4 injections are made on the first and fourth day of
each week.

3.4 Liver Biopsy 1. After intramuscular injection of 20 μg/kg medetomidine, the

canines are placed in the left lateral decubitus position.
2. The skin over the biopsy site is shaved, followed by disinfection
with povidone-iodine.
3. The local anesthetic (0.5% lidocaine) is injected into a small
area of the skin and tissues over part of the liver, and
ultrasonography-guided liver biopsies are carried out using a
16 G biopsy needle (Aragon Medical Devices, Plano, TX) (see
Note 2).
4. After removing the biopsy needle, the puncture site is manually
compressed, and then atipamezole hydrochloride is intramus-
cular injected finally.
5. Some samples are fixed in 4% paraformaldehyde overnight and
used for the evaluation of liver fibrosis by Sirius red staining
(Fig. 2a–c), while others are stored at 80  C for microarray
analysis and real-time quantitative polymerase chain reaction
(RT-qPCR) testing.

3.5 Indocyanine 1. A 22 G indwelling venous cannula is inserted into the internal

Green (ICG) Test to jugular vein and blood is collected. This venous route is kept
Assess the Liver open with physiological saline (see Notes 3–5).
Functions 2. ICG is prepared immediately prior to its administration.
3. Following the collection of blood for the zero-time sample,
0.1 mg ICG/kg is administered via the peripheral vein, and
blood is collected 5, 10, 15, 20, and 30 min later from the
maintained venous route (see Note 3).
4. Blood samples are centrifuged.
5. Serum is harvested from blood samples and analyzed for ICG
content using an ICG meter (Fig. 2d) (see Note 5).
 Canine Liver Fibrosis Model 207

20 25

15 20

15
10
10
5
5

0 0

Fig. 2 Kinetics of liver fibrosis and half-life time of indocyanine green (ICG) testing after the infusion of cultured
autologous bone marrow derived-mesenchymal stem cells (BMSCs) in the canine liver fibrosis model. (a, c)
The amount of fibrosis, as seen with Sirius red staining, was increased in the control group (after continuous
carbon tetrachloride treatment without the autologous BMSC infusion). (b, c) A decreased level of fibrosis was
seen at 4 weeks after the autologous BMSC infusion, even though carbon tetrachloride treatment was also
continued. (d) The half-life of ICG was significantly shorter at 4 weeks in the BMSC-infused group compared
with 0 weeks in the same group and 4 weeks in the control group. Every value and bar shown indicates
mean  SD. This figure was reproduced with permission from ref. [9]

3.6 Sirius Red 1. Dewax and hydrate paraffin sections.

Staining 2. Stain liver fibrosis with Sirius red solution for 15–20 min.
3. Wash with acetic acid for 1 min.
4. Dehydrate with 100% ethanol for 5 min.
5. Clear in xylene for 10 min and then mount in a resinous
medium.
208 Taro Takami et al.

4 Notes

1. When CCl4 is administered, the port and catheter are flushed

with water to confirm the absence of resistance before admin-
istration. The catheter is finally flushed out with water to
ensure that no food residue becomes stuck in the catheter.
2. The needle insertion angle for liver biopsy is usually 44 .
3. Around 5 mL of blood is sampled after ICG administration.
4. Animals are fasted from 12 h before ICG tests and liver biopsy.
5. The amount of serum used for the ICG meter is 500 μL
each time.

Acknowledgments

The development of this model was supported by the Japanese

highway program for the realization of regenerative medicine of
the Japan Science and Technology Agency (JST) and the Japan
Agency for Medical Research and Development (AMED) and by
JSPS KAKENHI grant numbers JP17H04162 to I.S. and
JP17K09428 to T.T. We also thank the members involved in this
model, Luiz Fernando Quintanilha, Takashi Matsuda, Bruno Diaz
Paredes, Yuki Aibe, Ryo Sasaki, Tatsuro Nishimura, Toshihiko
Matsumoto, Tsuyoshi Ishikawa, Naoki Yamamoto, Yasuhisa
Oishi, Ikuyo Yoshida, Koushiro Ozono, Takafumi Kanematsu,
Katsuhiro Kesamaru, Hironori Morisaki, Miku Yamashita, Yutaro
Kitaura, Ko Nakasumi, Kazuaki Igari, Shotaro Eto, Kyosuke
Hidari, Masahiro Imagawa, Tsunehiro Isozaki, Kentaro Itoh, Har-
uhi Sakai, Mai Sakai, Hiroya Inoue, Fumiya Yamane, Masato
Hiyama, and Saori Shinzato.

References

1. Takami T, Terai S, Sakaida I (2012) Stem cell 4. Terai S, Ishikawa T, Omori K, Aoyama K,
therapy in chronic liver disease. Curr Opin Gas- Marumoto Y, Urata Y, Yokoyama Y, Uchida K,
troenterol 28(3):203–208. https://doi.org/10. Yamasaki T, Fujii Y, Okita K, Sakaida I (2006)
1097/MOG.0b013e3283521d6a Improved liver function in patients with liver cir-
2. Terai S, Sakaida I, Yamamoto N, Omori K, rhosis after autologous bone marrow cell infusion
Watanabe T, Ohata S, Katada T, Miyamoto K, therapy. Stem Cells 24(10):2292–2298. https://
Shinoda K, Nishina H, Okita K (2003) An doi.org/10.1634/stemcells.2005-0542. pii:
in vivo model for monitoring trans- 2005-0542
differentiation of bone marrow cells into func- 5. Kim JK, Park YN, Kim JS, Park MS, Paik YH, Seok
tional hepatocytes. J Biochem 134(4):551–558 JY, Chung YE, Kim HO, Kim KS, Ahn SH, Kim
3. Sakaida I, Terai S, Yamamoto N, Aoyama K, do Y, Kim MJ, Lee KS, Chon CY, Kim SJ, Terai S,
Ishikawa T, Nishina H, Okita K (2004) Trans- Sakaida I, Han KH (2010) Autologous bone mar-
plantation of bone marrow cells reduces CCl4- row infusion activates the progenitor cell compart-
induced liver fibrosis in mice. Hepatology 40 ment in patients with advanced liver cirrhosis. Cell
(6):1304–1311. https://doi.org/10.1002/ Transplant 19(10):1237–1246. https://doi.org/
hep.20452 10.3727/096368910X506863
 Canine Liver Fibrosis Model 209

6. Kim JK, Kim SJ, Kim Y, Chung YE, Park YN, Kim Terai S, Sakaida I (2014) Canine mesenchymal
HO, Kim JS, Park MS, Sakaida I, Kim DY, Lee JI, stem cells show antioxidant properties against
Ahn SH, Lee KS, Han KH (2017) Long-term thioacetamide-induced liver injury in vitro and
follow-up of patients after autologous bone mar- in vivo. Hepatol Res 44(10):E206–E217.
row cell infusion for decompensated liver cirrho- https://doi.org/10.1111/hepr.12204
sis. Cell Transplant 26(6):1059–1066. https:// 9. Matsuda T, Takami T, Sasaki R, Nishimura T,
doi.org/10.3727/096368917X694778 Aibe Y, Paredes BD, Quintanilha LF,
7. Tanimoto H, Terai S, Taro T, Murata Y, Matsumoto T, Ishikawa T, Yamamoto N,
Fujisawa K, Yamamoto N, Sakaida I (2013) Tani K, Terai S, Taura Y, Sakaida I (2017) A
Improvement of liver fibrosis by infusion of canine liver fibrosis model to develop a therapy
cultured cells derived from human bone marrow. for liver cirrhosis using cultured bone marrow-
Cell Tissue Res 354(3):717–728. https://doi. derived cells. Hepatol Commun 1(7):691–703.
org/10.1007/s00441-013-1727-2 https://doi.org/10.1002/hep4.1071
8. Quintanilha LF, Takami T, Hirose Y, Fujisawa K,
Murata Y, Yamamoto N, Goldenberg RC,
 Chapter 19

A Rodent Model for Cell Transplantation of Hepatic

Progenitor Cells
Sei Kakinuma and Akihide Kamiya

Abstract
Hepatic progenitor cells are defined as cells exhibiting potency for active proliferation and capacity for
bipotential differentiation into hepatocytes and cholangiocytes. To prove the capacity of target cells for
terminal differentiation and reconstitution of organs, cell transplantation models have been widely used in
previous studies, including those involving the liver. Here we describe a protocol for transplantation of
hepatic progenitor cells using retrorsine pretreatment and partial hepatectomy. This transplantation assay
reveals the potential for reconstitution of hepatocytes in recipient livers by primary hepatic progenitor cells.
Donor cells are detected as a colony composed of 5–10 mature hepatocytes.

Key words Hepatoblasts, Hepatocytes, Intrasplenic injection, Partial hepatectomy, Retrorsine, Trans-
plantation, Differentiation

1 Introduction

Stem cells are defined as cells exhibiting both the capacity for self-
renewal and the ability to undergo multilineage differentiation
[1]. Somatic stem cells in the liver (i.e., hepatic stem cells) are
defined as cells capable of self-renewal and terminal differentiation
into hepatocytes and cholangiocytes. Liver lobules are constituted
of a variety of cells, including hepatocytes, cholangiocytes, hepatic
stellate cells (Ito cells), Kupffer cells, sinusoidal endothelial cells,
and pit cells, among others. Both hepatocytes and cholangiocytes
are derived from hepatoblasts in the fetal liver [2]. Hepatic progen-
itor cells are also capable of bipotential differentiation and active
proliferation but lack the capacity for self-renewal [3]. Hepatic
stem/progenitor cells can be enriched in mouse mid-gestational
fetal hepatic cell fractions based on several cell surface markers,
including Dlk [4], E-cadherin [5], Liv2 [6], and CD13 [7]. Rat
Dlk+ or Thy-1+ cells isolated from mid-gestational fetal liver exhibit
characteristics expected for hepatic stem/progenitor cells [8]. In
addition to enrichment in fetal liver, hepatic stem/progenitor cells

Naoki Tanimizu (ed.), Hepatic Stem Cells: Methods and Protocols, Methods in Molecular Biology, vol. 1905,
https://doi.org/10.1007/978-1-4939-8961-4_19, © Springer Science+Business Media, LLC, part of Springer Nature 2019

211
212 Sei Kakinuma and Akihide Kamiya

can also be enriched in mouse hepatic cells based on the expression

of several molecules, including CD13+/CD49f+/CD133+ cells in
neonatal liver [9], MIC1-1C3+/CD133+ cells in adult healthy liver
[10], and Lgr5+ cells in adult damaged liver [11].
To verify the capacity of target cells for terminal differentiation
and organ reconstitution, cell transplantation models involving
various organ systems including blood (bone marrow), neuron,
retina, intestine, and liver have been utilized. The objective of
transplantation experiments with hepatic progenitor cells is to
demonstrate the capacity for hepatocytic or cholangiocytic recon-
stitution by donor cells. In this section, we focus on the reconstitu-
tion of hepatocytes by transplantation of hepatic progenitor cells in
rodent models. It is essential to inhibit proliferation of intrinsic
mature hepatocytes for successful reconstitution by donor cells,
because little or no engrafted donor cells are detected in a healthy
liver or regenerating liver after simple partial hepatectomy in
rodents. Previous studies have demonstrated that several strains of
genetically modified mice can be used as recipients for hepatic cell
transplantation, including transgenic (Tg) mice carrying an albu-
min promoter linked to mouse urokinase-type plasminogen activa-
tor [12] and fumarylacetoacetate hydrolase-deficient mice
[13]. Intrinsic hepatocytes are continuously damaged in those
mice, whereas transplanted donor cells are not damaged and can
actively proliferate in the recipient liver. Alternatively, pretreatment
of retrorsine also attenuates hepatocyte proliferation after partial
hepatectomy, which allows donor cells to proliferate in the recipient
liver.
To detect the donor cells in recipient livers, donor cells are
labeled with marker proteins, such as green fluorescent protein
(GFP), other fluorescent proteins, or β-galactosidase. The ratio of
donor cells to recipient cells is assessed by histological analysis or
secreted marker proteins in recipient serum. However, it is possible
that overexpression of marker proteins can induce the rejection of
donor cells through an immune response against marker
proteins [14].
In this chapter, we describe the methods for transplantation of
primary hepatic progenitor cells after retrorsine treatment and 70%
partial hepatectomy in mice [7, 15]. This method can be used for
wild-type and nude mice as recipients. As representative data, the
results of transplantation of Dlk+ fetal hepatic progenitor cells
derived from heterozygous GFP-Tg mice are presented.

2 Materials

2.1 Preparation 1. Donor mice (C57BL/6 background, see Note 1).

of Donor Cells 2. Sterile phosphate buffered saline (PBS), calcium-free.
 Transplantation of Hepatic Progenitor Cells 213

3. EGTA buffer (pH 7.5): Hanks’ balanced salt solution (without

magnesium and calcium) supplemented with 10 mM HEPES
and 0.5 mM EGTA [16]. Weigh 8.0 g NaCl, 0.4 g KCl,
0.121 g Na2HPO4·12H2O, 0.06 g KH2PO4, 0.35 g
NaHCO3, 1.0 g glucose, 2.383 g HEPES, and 0.19 g EGTA,
and transfer to the beaker. Add distilled water to a volume of
900 mL. Mix and adjust pH to 7.5 in a total volume of 1 L with
distilled water, and filter through a 0.22 μm bottle-top filter.
Store at 4  C.
4. Collagenase buffer (pH 7.5): Hanks’ balanced salt solution
(with magnesium sulfate and 5 mM calcium chloride, see
Note 2) supplemented with 10 mM HEPES [16]. Weigh
8.0 g NaCl, 0.4 g KCl, 0.121 g Na2HPO4·12H2O, 0.06 g
KH2PO4, 0.35 g NaHCO3, 1.0 g glucose, 2.383 g HEPES,
0.2 g, 0.2 g MgSO4·7H2O, and 0.735 g CaCl2·2H2O, and
transfer to the beaker. Add distilled water to a volume of
900 mL. Mix and adjust pH to 7.5 in a total volume of 1 L
with distilled water, and filter through a 0.22 μm bottle-top
filter. Store at room temperature.
5. Collagenase solution: Dissolve 50 mg collagenase in collage-
nase buffer in a total volume of 100 mL (final concentration,
0.05%). Filter through a 0.22-μm syringe filter. The solution is
dispensed into 15-mL centrifuge tube for a single use, and
store at 20  C. Thaw the solution prior to use.
6. Staining medium: PBS supplemented with 3% fetal calf serum
(FCS).
7. Dulbecco’s Modified Eagle medium (DMEM) supplemented
with 10% FCS.
8. NH4Cl solution: Dissolve 4.012 g NH4Cl in distilled water in a
total volume of 1 L (final concentration, 75 mM). Filter
through a 0.22-μm bottle-top filter.
9. Rat anti-mouse Dlk antibody (see Note 3).
10. Magnetic cell selection system: Magnet beads conjugated with
secondary antibody for rat IgG and appropriate system for each
magnetic bead protocol, including positive selection column
and magnetic tube holder (see Notes 4 and 5).
11. Small scissors.
12. Fine small forceps.
13. Nylon mesh with 70-μm pores.

2.2 Recipient Mice 1. Mice. C57BL/6 mice or nude mice (male or female) are used
and Pretreatment in this protocol. Pretreatment of recipient mice starts at 3 or
Before Transplantation 4 weeks old.
214 Sei Kakinuma and Akihide Kamiya

2. 20 mg/mL retrorsine stock solution: Dissolve 40 mg retro-

rsine in 100% ethanol in a total volume of 2 mL. Store at
20  C (see Note 6).
3. Small scissors.
4. Small forceps.
5. Surgical suture.
6. Anesthetic drug (as permitted in your institution).

3 Methods

3.1 Preparation 1. Fetal livers derived from mid-gestational fetal mice (embryonic
of Donor Cells (Hepatic day 12.5–14.5) are harvested using fine small forceps and
Progenitor Cells placed in sterile PBS. The livers are then transferred to EGTA
Derived from Fetal buffer in a sterile Petri dish and minced using small scissors at
Mice) room temperature.
2. Transfer the minced livers gently using a pipette into a 15 mL
centrifuge tube. Incubate the minced livers in EGTA buffer at
37  C for 5 min.
3. Centrifuge the collection tube at 150–200  g for 3 min (see
Note 7). Remove the supernatant and tap the centrifugation
tube gently to suspend the cell pellet.
4. Suspend with 0.05% collagenase solution (see Note 8). Incu-
bate the cells at 37  C for 15 min. After incubation, tap the
centrifugation tube vigorously (see Note 9).
5. Add DMEM with 10% FCS (see Note 10), and centrifuge the
collection tube at 150–200  g for 3 min. Discard the
supernatant.
6. Suspend the pellet in 75 mM NH4Cl solution (see Note 11),
and incubate the cells at 4  C for 7 min.
7. Tap the collection tube (see Note 12), and add DMEM with
10% FCS to stop the hemolysis. Centrifuge the collection tube
at 150–200  g for 3 min.
8. Discard supernatant and suspend the cells in staining medium.
Cell suspensions are filtered using a cell strainer with a 70-μm
filter (see Note 13).
9. Incubate the cells with anti-Dlk antibody at 4  C for 30 min.
After incubation, cells are suspended in staining medium and
centrifuged at 150–200  g for 3 min (see Note 4).
10. Discard supernatant and suspend the cells in staining medium.
Incubate the cells with secondary antibody and treat the cells
according to the manufacturer’s instructions for each magnetic
selection system.
 Transplantation of Hepatic Progenitor Cells 215

11. Suspend the harvested cells in DMEM with 10% FCS. Cell
suspensions are filtered using nylon mesh with 70-μm pores.
12. The volume of cell suspension should be adjusted to
0.1–0.15 mL/mouse (see Note 14). Harvested cells should
be transplanted into recipient mice as soon as possible.

3.2 Pretreatment 1. To inhibit replication of intrinsic hepatocytes, retrorsine is used

of Recipient Mice in this protocol [7, 15]. Dilute retrorsine stock solution with
sterile PBS (see Note 6) and adjust the final concentration of
retrorsine to 4 mg/mL in 20% ethanol/PBS.
2. Intraperitoneally (i.p.) inject 60 mg/kg retrorsine to
recipient mice.
3. Treat recipient mice weekly with retrorsine (60 mg/kg i.p.) for
3 weeks (see Note 15).
4. Three days after the last retrorsine injection, partial hepatec-
tomy and cell transplantation procedures are performed.

3.3 Partial 1. Perform 70% hepatectomy under general anesthesia using the
Hepatectomy classic procedure described by Higgins and Anderson [17].
2. Briefly, after the confirmation of deep anesthesia, cut the skin
and peritoneum at midline of the abdomen.
3. Ligate the left medial lobe and right medial lobe (see Note 16)
of the liver proximally using surgical silk sutures. Confirm a
change of liver color, indicating effective ischemia of the ligated
lobes.
4. Remove the left medial lobe and right medial lobe (see Note
17).
5. Ligate the left lateral lobe of the liver proximally using surgical
silk sutures. Confirm a change of liver color, indicating effective
ischemia of the ligated lobes.
6. Remove the left lateral lobe.

3.4 Cell 1. Loosely ligate the inferior pole of the spleen using surgical silk
Transplantation sutures (see Note 18).
2. Inject an aliquot of cells suspended in 0.1–0.15 mL of DMEM
supplemented with 10% FCS into the splenic pulp from the
inferior pole of the spleen using a 30-gauge needle.
3. Tightly ligate the inferior pole of the spleen proximal to the
injection point (see Note 19).
4. Suture the skin and peritoneum using surgical sutures.
5. Sacrifice mice 3–12 weeks after transplantation, at which
time donor cells can be clearly detected in recipient livers
(Figs. 1 and 2).
216 Sei Kakinuma and Akihide Kamiya

Fig. 1 Transplanted E13.5 Dlk+ hepatic progenitor cells exhibit liver-repopulating activity in vivo. Dlk+ cells
derived from E13.5 fetal livers were transplanted into nude mice treated with retrorsine before partial
hepatectomy. Mice were analyzed 4 weeks after transplantation. Representative fluorescent stereoscopic
images of recipient liver are shown. The recipient liver contained many GFP+ signals. Bright positive signals
(arrows) and dull positive signals (arrow heads) were detected, with variation in brightness ascribed to
variation in proximity to the liver surface

Fig. 2 Microscopic view of GFP+ donor cells. GFP+ donor cells were detected as a colony composed of 5–10
cells (dashed circle). Donor-derived cells formed colonies producing GFP (green) and albumin (red), indicating
that such cells engrafted as parenchymal cells
 Transplantation of Hepatic Progenitor Cells 217

4 Notes

1. Donor cells should be tagged with an appropriate marker

protein. Heterozygous GFP-Tg mice and heterozygous
LacZ/ROSA26-Tg mice have been widely used in previous
studies. Donor cells derived from such Tg mice in recipient
livers are easily detected using a stereoscopic (fluorescent)
microscope or by immunohistological analysis. However, it is
possible that overexpression of the marker protein can induce
the rejection of donor cells by triggering a recipient immune
response against the marker protein. Such immune rejection
can be avoided in immunocompromised recipients, such as
nude or SCID mice.
2. The final concentration of calcium chloride (5 mM) is higher
than that of normal Hanks’ balanced salt solution, whereas the
final concentration of magnesium sulfate is the same as that of
normal Hanks’ balanced salt solution.
3. Other appropriate primary antibodies can be used for positive
selection of hepatic progenitor cells, such as EpCAM, CD13,
and CD133. Negative selection (for hematopoietic cells) can
be combined with the positive selection to increase the purity
of hepatic progenitor cells.
4. A fluorescence-activated cell sorter (FACS) can be used instead
of a magnetic selection system. However, it is possible for
FACS-sorted cells to be damaged by sheath pressure and pro-
longed sorting if the target for the final yield of hepatic pro-
genitors is more than 1  105. If target cells are to be purified
by FACS, cell sorting should be done after negative selection
(for hematopoietic cells) using the magnetic system in order to
save time.
5. A centrifuge tube (15 mL) with low adsorption properties is
appropriate for collection of hepatic progenitor cells.
6. Dilute 20 mg/mL retrorsine solution fivefold with PBS ( )
before injection to recipient mice.
7. To decrease hematopoietic cells in cell suspensions, cells are
centrifuged at a relatively low speed compared with blood cell
separation.
8. Adjust the volume of collagenase solution depending on the
volume of liver cells. When 8–10 livers derived from embryonic
day (E) 13.5 mice are harvested, liver cells are suspended with
10 mL collagenase solution.
9. To dissociate the liver cells, tap the tube more vigorously at this
step than at previous steps.
218 Sei Kakinuma and Akihide Kamiya

10. To inactivate collagenase, serum should be added to the cell

suspension.
11. Adjust the volume of NH4Cl solution depending on the vol-
ume of liver cells. When 8–10 livers derived from E13.5 mice
are harvested, liver cells are suspended with 7 mL NH4Cl
solution.
12. To induce hemolysis of red blood cells, tap the tube vigorously
at this step.
13. To avoid contamination with dead cell cluster, cell suspension
should be filtered at this step.
14. The engraftment rate of hepatic progenitor cells in recipient
livers is relatively low compared with that of mature hepato-
cytes. Donor cells are easily detected after transplantation of
1  104 adult hepatocytes, whereas it is very difficult to detect
the donor cells in the recipient liver after transplantation of less
than 1  104 hepatic progenitor cells. If the number of har-
vested cells is less than 1  104 cells, in vitro expansion of such
cells should be considered. If the number of harvested cells is
more than 1  105 cells, primary hepatic progenitor cells can
be detected using a stereoscopic microscope.
15. Retrorsine treatment should be started when mice are
3–4 weeks old to inhibit proliferation of intrinsic hepatocytes.
It is possible that retrorsine treatment is relatively ineffective
against older mice.
16. The liver should be ligated gently. The gallbladder is situated
between the left medial lobe and right medial lobe. Remove the
gallbladder at the same time as liver resection.
17. When ischemia of the ligated lobes is sufficient, there will be
little bleeding from the surface of scission.
18. Do not tightly ligate the sutures at this step.
19. To avoid the backflow of the injected cell suspension, tightly
ligate with silk sutures (loosely ligated at the previous step) just
after injection of cell suspension.

Acknowledgments

This work was supported in part by Grants-in-Aid for Scientific

Research from the Ministry of Education, Culture, Sports, Science
and Technology in Japan, Japan Agency for Medical Research and
Development, and the Japanese Society of Gastroenterology. We
thank Prof. H. Nakauchi (the University of Tokyo), Prof.
A. Miyajima (the University of Tokyo), and Dr. N. Tanimizu (Sap-
poro Medical University) for the helpful discussion.
 Transplantation of Hepatic Progenitor Cells 219

References
1. Rashid T, Kobayashi T, Nakauchi H (2014) development in Mice. Gastroenterology 137
Revisiting the flight of Icarus: making human (3):1114–1126
organs from PSCs with large animal chimeras. 10. Dorrell C, Erker L, Schug J, Kopp JL, Canaday
Cell Stem Cell 15(4):406–409 PS, Fox AJ, Smirnova O, Duncan AW, Fine-
2. Shiojiri N, Lemire JM, Fausto N (1991) Cell gold MJ, Sander M, Kaestner KH, Grompe M
lineages and oval cell progenitors in rat liver (2011) Prospective isolation of a bipotential
development. Cancer Res 51(10):2611–2620 clonogenic liver progenitor cell in adult mice.
3. Kakinuma S, Nakauchi H, Watanabe M (2009) Genes Dev 25(11):1193–1203
Hepatic stem/progenitor cells and stem-cell 11. Huch M, Dorrell C, Boj SF, van Es JH, Li VS,
transplantation for the treatment of liver dis- van de Wetering M, Sato T, Hamer K, Sasaki N,
ease. J Gastroenterol 44(3):167–172 Finegold MJ, Haft A, Vries RG, Grompe M,
4. Tanimizu N, Nishikawa M, Saito H, Clevers H (2013) In vitro expansion of single
Tsujimura T, Miyajima A (2003) Isolation of Lgr5+ liver stem cells induced by Wnt-driven
hepatoblasts based on the expression of Dlk/ regeneration. Nature 494(7436):247–250
Pref-1. J Cell Sci 116. (Pt 9:1775–1786 12. Rhim JA, Sandgren EP, Palmiter RD, Brinster
5. Nitou M, Sugiyama Y, Ishikawa K, Shiojiri N RL (1995) Complete reconstitution of mouse
(2002) Purification of fetal mouse hepatoblasts liver with xenogeneic hepatocytes. Proc Natl
by magnetic beads coated with monoclonal Acad Sci USA 92(11):4942–4946
anti-e-cadherin antibodies and their in vitro 13. Overturf K, Al-Dhalimy M, Manning K, Ou
culture. Exp Cell Res 279(2):330–343 CN, Finegold M, Grompe M (1998) Ex vivo
6. Watanabe T, Nakagawa K, Ohata S, hepatic gene therapy of a mouse model of
Kitagawa D, Nishitai G, Seo J, Tanemura S, Hereditary Tyrosinemia Type I. Hum Gene
Shimizu N, Kishimoto H, Wada T, Aoki J, Ther 9(3):295–304
Arai H, Iwatsubo T, Mochita M, Watanabe T, 14. Maeda H, Shigoka M, Wang Y, Fu Y, Wesson
Satake M, Ito Y, Matsuyama T, Mak TW, Pen- RN, Lin Q, Montgomery RA, Enzan H, Sun Z
ninger JM, Nishina H, Katada T (2002) (2014) Disappearance of GFP-positive hepato-
SEK1/MKK4-mediated SAPK/JNK signaling cytes transplanted into the liver of syngeneic
participates in embryonic hepatoblast prolifer- wild-type rats pretreated with retrorsine. PLoS
ation via a pathway different from NF-kappaB- One 9(5):e95880
induced anti-apoptosis. Dev Biol 250 15. Dan YY, Riehle KJ, Lazaro C, Teoh N,
(2):332–347 Haque J, Campbell JS, Fausto N (2006) Isola-
7. Kakinuma S, Ohta H, Kamiya A, Yamazaki Y, tion of multipotent progenitor cells from
Oikawa T, Okada K, Nakauchi H (2009) Ana- human fetal liver capable of differentiating
lyses of cell surface molecules on hepatic stem/ into liver and mesenchymal lineages. Proc
progenitor cells in mouse fetal liver. J Hepatol Natl Acad Sci U S A 103(26):9912–9917
51(1):127–138 16. Kost DP, Michalopoulos GK (1991) Effect of
8. Oertel M, Menthena A, Chen YQ, Teisner B, 2% dimethyl sulfoxide on the mitogenic prop-
Jensen CH, Shafritz DA (2008) Purification of erties of epidermal growth factor and hepato-
fetal liver stem/progenitor cells containing all cyte growth factor in primary hepatocyte
the repopulation potential for normal adult rat culture. J Cell Physiol 147(2):274–280
liver. Gastroenterology 134(3):823–832 17. Higgins G, Anderson R (1931) Experimental
9. Kamiya A, Kakinuma S, Yamazaki Y, Nakauchi pathology of the liver I: restoration of the liver
H (2009) Enrichment and clonal culture of of the white ret following partial surgical
progenitor cells during mouse postnatal liver removal. Arch Pathol 12:186–202
 Chapter 20

Mouse Model for Hepatocellular Carcinoma and

Cholangiocarcinoma Originated from Mature Hepatocytes
Masahiro Yamamoto, Bing Xin, and Yuji Nishikawa

Abstract
Liver cancer consists of two main histological subtypes, hepatocellular carcinoma and cholangiocarcinoma,
both of which have poor prognosis. Therefore, in searching for new therapeutic targets, adequate mouse
models to develop and validate therapeutic strategies are urgently needed. Although there are mouse
models of liver cancer, each model has shortcomings. To overcome these shortcomings, a mouse model
using a hydrodynamic tail vein injection and the Sleeping Beauty transposon was developed. By inducing
stable expression of oncogenes in mouse hepatocytes in vivo, the model can easily induce liver cancer with
specific characteristics that depend on the oncogenes used to induce carcinogenesis. Here, we describe the
details of the methods to induce hepatocellular carcinoma or cholangiocarcinoma from mouse hepatocytes.

Key words Sleeping Beauty transposon, Hydrodynamic tail vein injection, Hepatocellular carcinoma,
Intrahepatic cholangiocarcinoma, Mouse model, Liver tumor, Liver cancer

1 Introduction

Liver cancer is the second most common cause of cancer death

worldwide, and its prognosis is very poor (the ratio of mortality to
incidence is 0.95) [1]. Liver cancer consists of two main histological
subtypes, hepatocellular carcinoma (HCC) and intrahepatic cho-
langiocarcinoma (CC). The risk factors associated with HCC
include hepatitis B virus (HBV) and hepatitis C virus (HCV),
alcohol, nonalcoholic fatty liver disease, diabetes, environmental
toxin exposure, and obesity [2, 3]. The incidence of CC is highest
in Northern Thailand, where infection with liver flukes (Clonorchis
sinensis and Opisthorchis viverrini) leads to chronic biliary injury
[4]. In regions without liver fluke infection, the main etiologies of
CC are primary sclerosing cholangitis and HBV or HCV infection
[4]. Lifestyle-related risk factors, such as smoking, alcohol, and
diabetes, also contribute to CC [5].

Naoki Tanimizu (ed.), Hepatic Stem Cells: Methods and Protocols, Methods in Molecular Biology, vol. 1905,
https://doi.org/10.1007/978-1-4939-8961-4_20, © Springer Science+Business Media, LLC, part of Springer Nature 2019

221
222 Masahiro Yamamoto et al.

The therapies for HCC are surgical resection, liver transplanta-

tion, ablation, transarterial chemoembolization, and tyrosine kinase
inhibitors [6]. Sorafenib, a multiple kinase inhibitor, is used as
standard drug therapy for HCC and has some survival benefit [7],
but its effects remain limited [8, 9]. Most patients with CC are
found in the advanced stage and have poor prognosis despite
treatment with current standard chemotherapy [10]. Even for
patients who undergo surgical resection, prognosis is still poor
due to the high recurrence rate [4]. Therefore, appropriate animal
models for liver cancer are necessary to identify therapeutic targets
and perform preclinical studies.
Several mouse models of liver cancer are currently available,
including chemical carcinogenesis models, genetically engineered
mouse models, and xenograft models of human cancer cell lines
[11]. The chemical carcinogenesis model is a classical and widely
used model, but the genetic abnormalities in these models are
usually unknown, and the duration of tumor development is long.
Although genetically engineered mouse models are useful models
to examine the roles of the modified genes, these models are costly
and time-consuming [12]. Furthermore, the important caveat of
genetically engineered mouse models is that mutations typically
occur in all hepatocytes, while only tumor cells harbor mutations
in human cancer. Xenograft models are frequently used to evaluate
chemotherapeutic drugs, but the cell lines sometimes do not reflect
the major population of the original tumors because of selection
during culture [13]. Because xenograft models of human cancer
cell lines require immunocompromised mice, the effects of host
immunity against tumor cells are neglected in this model.
To overcome these limitations of mouse models, a mouse
model with the combination of a hydrodynamic tail vein injection
(HTVi) and Sleeping Beauty (SB) transposon-mediated somatic
integration of genes has been developed and widely used
[12]. This model can induce liver tumors in immunocompetent
mice in a short period by known oncogenes. The method can cause
long-term expression of oncogenes in mouse hepatocytes and
induce liver tumors in vivo.
The standard procedure of HTVi involves the rapid (5–7 s)
injection of a large volume (8–10% of body weight) of a plasmid
solution via the mouse lateral tail vein [14, 15]. During injection,
the solution first accumulates in the right atrium; then, the solution
flows retrogradely into the hepatic vein and hepatic sinusoids. The
pressure in the sinusoid pushes the plasmid solution into hepato-
cytes through fenestrated hepatic sinusoidal endothelial cells. By
the hydrodynamic mechanism, HTVi can deliver the plasmid spe-
cifically to hepatocytes [16–19]. The limitation of the HTVi
method for cancer research is that gene expression is only transient,
peaking 6 h after transfection and rapidly decreasing to a level less
of than 1/1000 in 3 days [14].
 Mouse Models of Liver Cancers from Mature Hepatocytes 223

a b
1. Cut transposon content
Plasmid for expression of SB transposase
SB transposase
"
SB transposase

"
Promoter SB transposase polyA Promoter Gene of Interest polyA
ITR ITR

Plasmid containing SB transposon cassette 2. Paste into host genome

SB transposase SB transposase
Promoter Gene of Interest polyA
ITR ITR Promoter Gene of Interest polyA

Fig. 1 The Sleeping Beauty (SB) transposon system. (a) Components of the Sleeping Beauty transposon
system. The SB transposase expression plasmid and a plasmid containing SB transposon cassette. ITR
inverted terminal repeat. (b) Integration of the transposon cassette into a host genome by the SB transposase.
SB transposase transfers a gene in the transposon cassette into a host genome in a cut-and-paste manner

Transposons were discovered in the 1940s by Barbara McClin-

tock in the maize genome, and they are known as jumping genes.
The Human Genome Project revealed that approximately 45% of
the human genome is transposon-derived, but the transposons are
inactive [20]. Zoltan Ivics modified a transposon and transposase of
the salmonid fish to work in mammalian cells and named the
modified transposon the SB transposon [21]. The SB transposon
system consists of two components, a transposase-expressing vector
and a plasmid with the SB transposon cassette, which contains a
promoter, a gene to be expressed, and a poly A sequence that is
flanked by inverted terminal repeats (ITRs) (Fig. 1a). When these
components are introduced into mouse hepatocytes by the HTVi
method, the induced SB transposase recognizes ITRs on the trans-
poson cassette and then cuts the content and pastes it into the
mouse hepatocyte genome (Fig. 1b), leading to stable expression
of the oncogenes. Integration occurs at any site containing A:T/T:
A sequences [22]. With this method, several liver cancer models
have been developed that have different phenotypes that are remi-
niscent of HCC, CC, and hepatoblastoma, depending on the genes
used for tumor induction (Table 1).
The cellular origins of HCC and CC are still debated
[23, 24]. HCC and CC are thought to be derived from malignant
transformation of corresponding parenchymal cells; hepatocytes and
intrahepatic bile ducts. Recent evidence has indicated that HCC and
CC are derived from transformation of stem/progenitor cells or
mature cells subjected to reprogramming, which results in stem/
progenitor features [23, 25]. Furthermore, studies on animal mod-
els have revealed that CC can originate from mature hepatocytes by
transdifferentiation [26–29]. A genome-wide approach in human
HCC and CC revealed that CC arising in chronic hepatitis, which is
also a risk factor for HCC [30], shows similar gene alteration profiles
224 Masahiro Yamamoto et al.

Table 1
Sleeping Beauty transposon-mediated liver cancer models

Histology Genes Strain Reference

G12V
HCC myrAKT/HRAS C57BL/6 [34]
HRASG12V/c-Myc C57BL/6 [34]
myrAKT/c-Myc C57BL/6 [34]
c-MYC/MCL1 FVB/N [35]
β-cateninS33Y/KRASG12V FVB/N [36]
β-cateninS45Y/KRASG12V
β-cateninS33Y/hMet FVB/N [37]
β-cateninS45Y/hMet
myrAKT/c-Met FVB/N [38]
myAKT/NRASG12V FVB/N [39]
myrAKT/Spry2Y55F FVB/N [40]
myrAKT/hMet FVB/N, C57BL/6 [41]
myrAKT FVB/N [42]
hMet/Spry2Y55F FVB/N [43]
NRASG12V/BMI1 FVB/N [44]
ΔN90-β-catenin/Spry2Y55F FVB/N [45]
hMET/ΔN90-β-catenin FVB/N [46]
Hepatocellular adenoma myrAKT/ΔN90-β-catenin FVB/N, C57BL/6 [41]
hMet/Ccnd1 FVB/N [47]
CC myrAKT/YAPS127A C57BL/6 [28, 48]
myrAKT/N1ICD FBV/N [27]
Biliary cholangioadenoma myrAKT/N1ICD C57BL/6 [28]
myrAKT/N2ICD
Mixed HCC and CC PIK3CAH1047R/YAPS127A FVB/N [29]
NRASG12V FVB/N(Arf
/
) [49]
Hepatoblastoma YAPS127A/c-Myc C57BL/6 [28]
ΔN90-β-catenin/YAPS127A FVB/N [50]
c-Myc FVB/N [51]
Undifferentiated tumor HRASG12V/shTrp53 C57BL/6 [52]
HCC hepatocellular carcinoma, CC cholangiocarcinoma

to HCC, suggesting that HCC and CC with chronic hepatitis arise

from same cellular origin, such as liver progenitor cells [31].
In this chapter, we describe mouse HCC and CC models that
originated from mature hepatocytes in vivo by the HTVi and SB
transposon system.

2 Materials

2.1 HTVi 1. Plasmids

(a) pT2-C-Luc/PGK-SB13 (Addgene#20207, Addgene).
(b) pT3-EF1α-myrAKT-HA (Addgene#31789, Addgene).
 Mouse Models of Liver Cancers from Mature Hepatocytes 225

(c) pT3-EF1α-FLAG-YAPS127A (Addgene#86497,

Addgene).
(d) pT3-EF1α-3xFLAG-N2ICD [28].
(e) Mice: Male C57BL/6 mice, 6–10 weeks old (see Note 1).
2. 2.5-mL syringe with a Luer Lock.
3. 25G  100 (25 mm) or 27G  1/200 (13 mm) needle (see Note 2).
4. Anesthetic: Isoflurane.
5. Inhalational anesthesia apparatus for small animals.
6. 70% ethanol.
7. Gauze.

2.2 Tissue Sampling 1. Phosphate-buffered saline (PBS).

2. Peristaltic pump.
3. Scale to weight mouse (~15–40 g) and mouse liver (~3–15 g).
4. Digital camera.
5. 4% paraformaldehyde/PBS.
6. Liquid nitrogen.
7. Freezer (
80  C or less).

2.3 Embedding 1. Tissue processor.

Tissue into Paraffin 2. Ethanol.
3. Xylene.
4. Paraffin.
5. Embedding center.
6. Tissue mold.
7. Microtome.
8. Slide glass (see Note 3).

2.4 Hematoxylin 1. Mayer’s hematoxylin.

and Eosin Stain 2. 1% eosin Y solution (eosin Y diluted to 1% with distilled water).
3. Cover glass (24  40 mm).
4. Mounting media; Entellan new (Merck).

2.5 Immunohisto- 1. 30% H2O2.

chemistry 2. Methanol.
3. Antigen retrieval solution; Target Retrieval solution, pH 6.0
(DAKO).
4. Electric thermo that can keep water at 98  C.
5. 0.01% Tween 20 dissolved in PBS (PBST).
226 Masahiro Yamamoto et al.

6. Blocking reagent; SuperBlock (PBS) Blocking Buffer (Thermo

Fisher Scientific).
7. Secondary antibody; DAKO EnVision System.
8. Diaminobenzidine (DAB) solution; ImmPACT™ DAB (Vec-
tor Laboratories).

3 Methods

3.1 HTVi 1. Purify the plasmids with the endotoxin-free Maxiprep kit (Qia-
gen) or equilibrium. Because endotoxin causes adverse or toxic
effects on animals, endotoxin must be removed [32]. A high
concentration of endotoxin reduces the transfection efficiency
in eukaryotic cells [33].
2. For integrating two oncogenic genes, the molar ratio of SB
transposase-expressing plasmid and two plasmids carrying
oncogenic genes is 1:2:2, and the total amount of plasmid is
25 μg (see Note 4).
3. Dissolve the plasmids in 2.5 mL of Ringer solution or saline in a
sterile manner.
4. Expose the mice to anesthesia with isoflurane.
5. Gather 2.5 mL of the plasmid solution with a 2.5-mL syringe
with a Luer Lock.
6. Attach a 25G needle to the syringe and remove any air bubbles
from the syringe.
7. Warm the tail of the mouse with hot water (at 40–50  C) to
dilate the lateral tail vein, and clean the insertion site with 70%
ethanol.
8. Insert the needle into the lateral tail vein, and rapidly infuse the
solution in a continuous manner within 8 s.
9. Press the insertion site with gauze until bleeding stops.
10. Successful injection leads to short-term gene expression in
about 30% of hepatocytes.

3.2 Hepatocellular 1. Inject a combination of pT3-EF1α-myrAKT-HA and

Carcinoma Induced by pT3-EF1α-FLAG-HRASG12V with the SB13-expressing plas-
Activation of the PI3 mid by the HTVi method.
Kinase and RAS-MAP 2. Weigh the mice once per week, and carefully check the mice
Kinase Pathways twice per week to determine whether they have abdominal
(Fig. 2) distension or health problems. Abnormal weight gain or loss
indicates the development of liver tumors. Usually, 6–8 weeks
after the HTVi, the mice die from liver tumors. However, the
timing differs depending on the mouse strain, sex, and maturity
of HTVi skill.
 Mouse Models of Liver Cancers from Mature Hepatocytes 227

Fig. 2 Hepatocellular carcinoma model induced by myrAKT and HRASG12V. (a) Gross appearance of the liver at
2, 4, and 6 weeks after the HTVi. At 6 weeks, multinodular liver tumors develop. Scale bar, 10 mm. (b)
Hematoxylin and eosin (HE) staining of the liver. At 2 weeks, tumor cells with abundant lipid accumulation in
their cytoplasm appear. At 4 weeks, clusters of tumor cells devoid of lipid accumulation (arrowheads) emerge.
At 6 weeks, HCCs with scant lipid accumulation develop. The upper panels are enlarged in the lower panels.
Scale bar, 50 μm

3. To sample liver tissue, under deep anesthesia, perfuse the liver

with PBS via the portal vein by a peristaltic pump and excise the
liver. Weigh the whole liver, and calculate the ratio with respect
to the body weight. Take macroscopic pictures of the liver with
a digital camera, cut the main parts of each lobe of the liver, and
fix them with phosphate-buffered 4% paraformaldehyde (4%
PFA/PBS) overnight at 4  C for a paraffin-embedded section.
For biochemical analysis, enucleate liver tumors and dice them
into small pieces. Then, snap-freeze them using liquid nitro-
gen. Keep them at
80  C or less until use.
228 Masahiro Yamamoto et al.

4. The tumors are macroscopically multinodular. Histologically,

the tumors are typical hepatocellular carcinoma. In the early
lesion (2 weeks), the tumor cells are lipid-rich; then, in the
process of tumor development, the tumor cells lose lipid and
proliferate more actively by the action of spontaneously upre-
gulated Myc protein [34].

3.3 Cholangioc- 1. Inject a combination of pT3-EF1α-myrAKT-HA and

arcinoma Induced by pT3-EF1α-FLAG-YAPS127A plasmids with the SB13-
Activation of the PI3 expressing vector by the HTVi method.
Kinase Pathway 2. The observation methods are the same as above. Usually,
and Hippo-Yap 6–8 weeks after the HTVi, mice die from liver tumors.
Pathway (Fig. 3) 3. The earliest change in the AKT/YAP model is lipid accumula-
tion in transformed hepatocytes by activation of the PI3 kinase
pathway. At 2 weeks, spindle-shaped cells that are devoid of
lipid accumulation proliferate. The cells gradually form ductu-
lar structures with an increasing fibrous stroma. The immuno-
histochemical markers for cholangiocytes, such as cytokeratin
19 (CK19) and grainyhead-like 2 (Grhl2), are expressed in the
cells.

3.4 Biliary 1. Inject a combination of pT3-EF1α-myrAKT-HA and

Cystadenoma Induced pT3-EF1α-3xFLAG-N2ICD with the SB13-expressing plas-
by Activation of the PI3 mid by the HTVi method.
Kinase and Notch 2. The observation methods are the same as above. Usually,
Pathways (Fig. 4) 6–8 weeks after HTVi, multicystic tumors develop. In this
model, mice can tolerate severe abdominal distension.
3. Microscopically, the tumors show multicysts lined with cholan-
giocytic tumor cells that invade less into the surrounding tissue.

3.5 Embedding 1. After incubating tissues in 4% PFA/PBS at 4  C overnight,

Tissues in Paraffin transfer the tissues to 70% ethanol and keep the samples at 4  C
Wax and Cutting Thin until tissue processing.
Sections 2. Use a tissue processor to incubate the tissues as follows:
(a) 70% ethanol, 1 h once.
(b) 80% ethanol, 1 h once.
(c) 90% ethanol, 1 h once.
(d) 95% ethanol, 1 h once.
(e) 100% ethanol, 1 h three times.
(f) Xylene, 1 h three times.
(g) Paraffin wax, 1 h three times at 62  C.
3. Remove tissues from paraffin wax and transfer them to the
embedding center.
 Mouse Models of Liver Cancers from Mature Hepatocytes 229

Fig. 3 Cholangiocarcinoma model induced by myrAKT and YAPS127A. (a) Gross appearances of the liver at 2, 4,
and 6 weeks after the HTVi. Tumors diffusely infiltrate into the liver. Scale bar, 10 mm. (b) Microscopic
appearances of the AKT/YAP tumors. HE stain, Sirius red stain, and immunohistochemistry for cytokeratin
19 (CK19) and grainyhead-like 2 (Grhl2) at 2, 4, and 6 weeks after the HTVi. Scale bar, 50 μm. Reproduced
from Ref. 28 with permission from Elsevier
230 Masahiro Yamamoto et al.

Fig. 4 Biliary cystadenoma model induced by myrAKT and N2ICD. (a) Gross appearances of the liver 6 weeks
after the HTVi. Scale bar, 10 mm. (b) HE staining of the tumors. Scale bar, 50 μm

4. Dispense paraffin wax into an embedding mold and let it sit on

a heated surface.
5. Place tissues at an appropriate orientation in the embedding
mold, place a cassette on top of the tissue mold, and move the
embedding mold to a cold surface of the embedding center to
solidify the wax.
6. Use a microtome to cut 3 μm sections from the paraffin block.
Place the section into a tissue-sectioning bath at 50  C for
2–3 s, and mount it onto an appropriate slide glass (see Note 3).
7. Keep slides at 50  C overnight to allow sections to adhere to
the slides.

3.6 Hematoxylin 1. Deparaffinize sections in a deparaffinization series as in the

and Eosin Staining following incubation protocol:
(a) 100% xylene, 10 min three times.
(b) 50% xylene/50% ethanol, 5 min once.
(c) 100% ethanol, 1 min three times.
(d) 90% ethanol, 1 min once.
(e) 80% ethanol, 1 min once.
(f) 70% ethanol, 1 min once.
2. Wash sections with tap water for 3 min.
3. Incubate sections in Mayer’s hematoxylin solution for 20 min.
4. Wash sections with running tap water for 10 min.
5. Wash sections with 50% ethanol for 30 s.
6. Incubate sections in eosin solution for 10 min.
7. Dehydrate sections in 70%, 80%, and 90% ethanol for 30 s each,
and then move them to 100% ethanol for 2 min three times.
8. To clear, immerse slides in xylene for 5 min four times.
9. Mount the cover glass with mounting media.
10. Dry the slides until the mounting medium solidifies.
 Mouse Models of Liver Cancers from Mature Hepatocytes 231

Table 2
Primary antibodies for immunohistochemical markers for hepatocellular and cholangiocellular
differentiation

Marker for Antigen Company Order no. Species Dilution

Hepatocellular differentiation HNF4αP1 Perseus Proteomics K9218 Mouse 1:200
MUP1 Nordic-MUbio GAM/MUP Goat 1:500
Cholangiocellular differentiation CK19 Developed in Dr. Atsushi Rabbit 1:4000
Miyajima’s Laboratory, Tokyo
University, Japan
Sox9 Millipore AB5535 Rabbit 1:20,000
Grhl2 Sigma A35809 Rabbit 1:1000
HNF4αP1 hepatocyte nuclear factor α isoform P1, MUP1 major urinary protein 1, Sox9 SRY (sex-determining region Y)-
box 9

3.7 Immunohisto- The antibodies and their optimal dilution for specification of hepa-
chemistry of Markers tocellular and cholangiocellular differentiation are listed in Table 2.
for Hepatocytic The staining procedure can be performed at room temperature
and Cholangiocytic unless otherwise specified.
Differentiation 1. Deparaffinize sections as described in Subheading 3.6.
2. Immerse sections in 3% H2O2 dissolved in methanol for 10 min
to block endogenous peroxidase.
3. Briefly wash sections with PBST once.
4. Heat sections in a preheated Target Retrieval solution at 98  C
for 40 min in an electric thermos.
5. Remove sections from the electric thermos and cool them to
room temperature.
6. Circle the tissues with a hydrophobic pen.
7. Block the tissues with blocking reagent for 30 min in a humidi-
fier chamber. If the source of the primary antibody is mouse,
please see Note 5 to avoid secondary antibody to binding to
mouse serum.
8. Incubate the tissues with primary antibody dissolved in the
blocking reagent for 1 h at room temperature or overnight at
4  C.
9. Wash the slide with PBST for 5 min twice.
10. Incubate the tissues with secondary antibody for 30 min.
11. Wash the slide with PBST for 5 min twice.
12. Visualize with the DAB solution. During visualization, check
the slide under a microscope. When optimal staining is
obtained, stop the reaction by putting the slides into tap water.
13. Counterstain slides in Mayer’s hematoxylin solution for 1 min.
14. Wash slides with running tap water for 10 min.
232 Masahiro Yamamoto et al.

15. Rehydrate and clear the slides and then mount a cover glass on
them as described in Subheading 3.6.
16. Dry slides until the mounting medium solidifies.

4 Notes

1. Other strains (e.g., FVB, Balb/c, C3H, etc.) can be used.

However, susceptibility to tumor development may be differ-
ent among strains.
2. For smaller mice, a 27G  1/200 (13 mm) needle, instead of a
25G needle, can be used.
3. For immunohistochemistry, use slide glasses with a special coat
(e.g., Silan) to prevent detachment of tissue sections during the
heating step. For usual staining without a heating step in the
procedures, glass slides without a special coat are usually
sufficient.
4. For a HTVi of three oncogenic genes, the molar ratio of SB
transposase-expressing plasmid and three plasmids carrying
oncogenic genes is 1:2:2:2, and the total amount of plasmid
is 36 μg. However, the ratio and total amounts of plasmid can
be changed depending on the purpose of the experiment.
5. For primary antibodies developed in mice, use a staining kit for
mouse tissues, such as a Histofine Mouse Stain Kit (Nichirei).

References
1. Ferlay J, Soerjomataram I, Dikshit R, Eser S, 5. Palmer WC, Patel T (2012) Are common fac-
Mathers C, Rebelo M, Parkin DM, Forman D, tors involved in the pathogenesis of primary
Bray F (2015) Cancer incidence and mortality liver cancers? A meta-analysis of risk factors
worldwide: sources, methods and major pat- for intrahepatic cholangiocarcinoma. J Hepatol
terns in GLOBOCAN 2012. Int J Cancer 57(1):69–76. https://doi.org/10.1016/j.
136(5):E359–E386. https://doi.org/10. jhep.2012.02.022
1002/ijc.29210 6. Forner A, Reig M, Bruix J (2018) Hepatocel-
2. Altekruse SF, Henley SJ, Cucinelli JE, lular carcinoma. Lancet 391:1301–1314.
McGlynn KA (2014) Changing hepatocellular https://doi.org/10.1016/s0140-6736(18)
carcinoma incidence and liver cancer mortality 30010-2
rates in the United States. Am J Gastroenterol 7. Llovet JM, Ricci S, Mazzaferro V, Hilgard P,
109(4):542–553. https://doi.org/10.1038/ Gane E, Blanc JF, de Oliveira AC, Santoro A,
ajg.2014.11 Raoul JL, Forner A, Schwartz M, Porta C,
3. El-Serag HB, Kanwal F (2014) Epidemiology Zeuzem S, Bolondi L, Greten TF, Galle PR,
of hepatocellular carcinoma in the United Seitz JF, Borbath I, Haussinger D, Giannaris T,
States: where are we? Where do we go? Hepa- Shan M, Moscovici M, Voliotis D, Bruix J,
tology 60(5):1767–1775. https://doi.org/10. SHARP Investigators Study Group (2008)
1002/hep.27222 Sorafenib in advanced hepatocellular carci-
4. Razumilava N, Gores GJ (2014) Cholangiocar- noma. N Engl J Med 359(4):378–390.
cinoma. Lancet 383(9935):2168–2179. https://doi.org/10.1056/NEJMoa0708857
https://doi.org/10.1016/S0140-6736(13) 8. Cheng AL, Kang YK, Chen Z, Tsao CJ, Qin S,
61903-0 Kim JS, Luo R, Feng J, Ye S, Yang TS, Xu J,
 Mouse Models of Liver Cancers from Mature Hepatocytes 233

Sun Y, Liang H, Liu J, Wang J, Tak WY, Pan H, Ther 15(12):2063–2069. https://doi.org/10.
Burock K, Zou J, Voliotis D, Guan Z (2009) 1038/sj.mt.6300314
Efficacy and safety of sorafenib in patients in 18. Sawyer GJ, Rela M, Davenport M,
the Asia-Pacific region with advanced hepato- Whitehorne M, Zhang X, Fabre JW (2009)
cellular carcinoma: a phase III randomised, Hydrodynamic gene delivery to the liver: theo-
double-blind, placebo-controlled trial. Lancet retical and practical issues for clinical applica-
Oncol 10(1):25–34. https://doi.org/10. tion. Curr Gene Ther 9(2):128–135
1016/S1470-2045(08)70285-7 19. Bonamassa B, Hai L, Liu D (2011) Hydrody-
9. Inuzuka T, Nishikawa H, Sekikawa A, namic gene delivery and its applications in
Takeda H, Henmi S, Sakamoto A, Saito S, pharmaceutical research. Pharm Res 28
Kita R, Kimura T, Osaki Y, Kudo M (2011) (4):694–701. https://doi.org/10.1007/
Complete response of advanced hepatocellular s11095-010-0338-9
carcinoma with multiple lung metastases trea- 20. Lander ES, Linton LM, Birren B, Nusbaum C,
ted with sorafenib: a case report. Oncology 81 Zody MC, Baldwin J, Devon K, Dewar K,
(Suppl 1):152–157. https://doi.org/10. Doyle M, FitzHugh W, Funke R, Gage D,
1159/000333279 Harris K, Heaford A, Howland J, Kann L,
10. Valle J, Wasan H, Palmer DH, Cunningham D, Lehoczky J, LeVine R, McEwan P,
Anthoney A, Maraveyas A, Madhusudan S, McKernan K, Meldrim J, Mesirov JP,
Iveson T, Hughes S, Pereira SP, Roughton M, Miranda C, Morris W, Naylor J, Raymond C,
Bridgewater J, ABC Trial Investigators (2010) Rosetti M, Santos R, Sheridan A, Sougnez C,
Cisplatin plus gemcitabine versus gemcitabine Stange-Thomann Y, Stojanovic N,
for biliary tract cancer. N Engl J Med 362 Subramanian A, Wyman D, Rogers J,
(14):1273–1281. https://doi.org/10.1056/ Sulston J, Ainscough R, Beck S, Bentley D,
NEJMoa0908721 Burton J, Clee C, Carter N, Coulson A,
11. Heindryckx F, Colle I, Van Vlierberghe H Deadman R, Deloukas P, Dunham A,
(2009) Experimental mouse models for hepa- Dunham I, Durbin R, French L, Grafham D,
tocellular carcinoma research. Int J Exp Pathol Gregory S, Hubbard T, Humphray S, Hunt A,
90(4):367–386. https://doi.org/10.1111/j. Jones M, Lloyd C, McMurray A, Matthews L,
1365-2613.2009.00656.x Mercer S, Milne S, Mullikin JC, Mungall A,
12. Chen X, Calvisi DF (2014) Hydrodynamic Plumb R, Ross M, Shownkeen R, Sims S,
transfection for generation of novel mouse Waterston RH, Wilson RK, Hillier LW,
models for liver cancer research. Am J Pathol McPherson JD, Marra MA, Mardis ER, Fulton
184(4):912–923. https://doi.org/10.1016/j. LA, Chinwalla AT, Pepin KH, Gish WR, Chis-
ajpath.2013.12.002 soe SL, Wendl MC, Delehaunty KD, Miner
13. Landgraf M, McGovern JA, Friedl P, Hutma- TL, Delehaunty A, Kramer JB, Cook LL, Ful-
cher DW (2018) Rational design of mouse ton RS, Johnson DL, Minx PJ, Clifton SW,
models for cancer research. Trends Biotechnol Hawkins T, Branscomb E, Predki P,
36:242–251. https://doi.org/10.1016/j. Richardson P, Wenning S, Slezak T,
tibtech.2017.12.001 Doggett N, Cheng JF, Olsen A, Lucas S,
Elkin C, Uberbacher E, Frazier M, Gibbs RA,
14. Liu F, Song Y, Liu D (1999) Hydrodynamics- Muzny DM, Scherer SE, Bouck JB, Sodergren
based transfection in animals by systemic EJ, Worley KC, Rives CM, Gorrell JH, Metzker
administration of plasmid DNA. Gene Ther 6 ML, Naylor SL, Kucherlapati RS, Nelson DL,
(7):1258–1266. https://doi.org/10.1038/sj. Weinstock GM, Sakaki Y, Fujiyama A,
gt.3300947 Hattori M, Yada T, Toyoda A, Itoh T,
15. Zhang G, Budker V, Wolff JA (1999) High Kawagoe C, Watanabe H, Totoki Y, Taylor T,
levels of foreign gene expression in hepatocytes Weissenbach J, Heilig R, Saurin W,
after tail vein injections of naked plasmid DNA. Artiguenave F, Brottier P, Bruls T, Pelletier E,
Hum Gene Ther 10(10):1735–1737. https:// Robert C, Wincker P, Smith DR, Doucette-
doi.org/10.1089/10430349950017734 Stamm L, Rubenfield M, Weinstock K, Lee
16. Yokoo T, Kamimura K, Abe H, Kobayashi Y, HM, Dubois J, Rosenthal A, Platzer M,
Kanefuji T, Ogawa K, Goto R, Oda M, Suda T, Nyakatura G, Taudien S, Rump A, Yang H,
Terai S (2016) Liver-targeted hydrodynamic Yu J, Wang J, Huang G, Gu J, Hood L,
gene therapy: Recent advances in the tech- Rowen L, Madan A, Qin S, Davis RW, Feder-
nique. World J Gastroenterol 22 spiel NA, Abola AP, Proctor MJ, Myers RM,
(40):8862–8868. https://doi.org/10.3748/ Schmutz J, Dickson M, Grimwood J, Cox DR,
wjg.v22.i40.8862 Olson MV, Kaul R, Raymond C, Shimizu N,
17. Suda T, Liu D (2007) Hydrodynamic gene Kawasaki K, Minoshima S, Evans GA,
delivery: its principles and applications. Mol Athanasiou M, Schultz R, Roe BA, Chen F,
234 Masahiro Yamamoto et al.

Pan H, Ramser J, Lehrach H, Reinhardt R, (8):2911–2915. https://doi.org/10.1172/

McCombie WR, de la Bastide M, Dedhia N, JCI63212
Blocker H, Hornischer K, Nordsiek G, 28. Yamamoto M, Xin B, Watanabe K, Ooshio T,
Agarwala R, Aravind L, Bailey JA, Bateman A, Fujii K, Chen X, Okada Y, Abe H, Taguchi Y,
Batzoglou S, Birney E, Bork P, Brown DG, Miyokawa N, Furukawa H, Nishikawa Y
Burge CB, Cerutti L, Chen HC, Church D, (2017) Oncogenic determination of a broad
Clamp M, Copley RR, Doerks T, Eddy SR, spectrum of phenotypes of hepatocyte-derived
Eichler EE, Furey TS, Galagan J, Gilbert JG, mouse liver tumors. Am J Pathol
Harmon C, Hayashizaki Y, Haussler D, 187:2711–2725. https://doi.org/10.1016/j.
Hermjakob H, Hokamp K, Jang W, Johnson ajpath.2017.07.022
LS, Jones TA, Kasif S, Kaspryzk A, Kennedy S, 29. Li X, Tao J, Cigliano A, Sini M, Calderaro J,
Kent WJ, Kitts P, Koonin EV, Korf I, Kulp D, Azoulay D, Wang C, Liu Y, Jiang L, Evert K,
Lancet D, Lowe TM, McLysaght A, Demartis MI, Ribback S, Utpatel K,
Mikkelsen T, Moran JV, Mulder N, Pollara Dombrowski F, Evert M, Calvisi DF, Chen X
VJ, Ponting CP, Schuler G, Schultz J, (2015) Co-activation of PIK3CA and Yap pro-
Slater G, Smit AF, Stupka E, Szustakowki J, motes development of hepatocellular and cho-
Thierry-Mieg D, Thierry-Mieg J, Wagner L, langiocellular tumors in mouse and human
Wallis J, Wheeler R, Williams A, Wolf YI, liver. Oncotarget 6(12):10102–10115.
Wolfe KH, Yang SP, Yeh RF, Collins F, Guyer https://doi.org/10.18632/oncotarget.3546
MS, Peterson J, Felsenfeld A, Wetterstrand KA,
Patrinos A, Morgan MJ, de Jong P, Catanese 30. Lee CH, Chang CJ, Lin YJ, Yeh CN, Chen MF,
JJ, Osoegawa K, Shizuya H, Choi S, Chen YJ, Hsieh SY (2009) Viral hepatitis-associated
Szustakowki J, International Human Genome intrahepatic cholangiocarcinoma shares com-
Sequencing Consortium (2001) Initial mon disease processes with hepatocellular car-
sequencing and analysis of the human genome. cinoma. Br J Cancer 100(11):1765–1770.
Nature 409(6822):860–921. https://doi.org/ https://doi.org/10.1038/sj.bjc.6605063
10.1038/35057062 31. Fujimoto A, Furuta M, Totoki Y, Tsunoda T,
21. Ivics Z, Hackett PB, Plasterk RH, Izsvak Z Kato M, Shiraishi Y, Tanaka H, Taniguchi H,
(1997) Molecular reconstruction of Sleeping Kawakami Y, Ueno M, Gotoh K, Ariizumi S,
Beauty, a Tc1-like transposon from fish, and Wardell CP, Hayami S, Nakamura T, Aikata H,
its transposition in human cells. Cell 91 Arihiro K, Boroevich KA, Abe T, Nakano K,
(4):501–510 Maejima K, Sasaki-Oku A, Ohsawa A,
Shibuya T, Nakamura H, Hama N, Hosoda F,
22. Izsvak Z, Ivics Z (2004) Sleeping beauty trans- Arai Y, Ohashi S, Urushidate T, Nagae G,
position: biology and applications for molecu- Yamamoto S, Ueda H, Tatsuno K, Ojima H,
lar therapy. Mol Ther 9(2):147–156. https:// Hiraoka N, Okusaka T, Kubo M, Marubashi S,
doi.org/10.1016/j.ymthe.2003.11.009 Yamada T, Hirano S, Yamamoto M, Ohdan H,
23. Marquardt JU, Andersen JB, Thorgeirsson SS Shimada K, Ishikawa O, Yamaue H,
(2015) Functional and genetic deconstruction Chayama K, Miyano S, Aburatani H,
of the cellular origin in liver cancer. Nat Rev Shibata T, Nakagawa H (2016) Whole-
Cancer 15(11):653–667. https://doi.org/10. genome mutational landscape and characteri-
1038/nrc4017 zation of noncoding and structural mutations
24. Sell S (2002) Cellular origin of hepatocellular in liver cancer. Nat Genet 48(5):500–509.
carcinomas. Semin Cell Dev Biol 13 https://doi.org/10.1038/ng.3547
(6):419–424 32. Kim MJ, Ahituv N (2013) The hydrodynamic
25. Oikawa T (2016) Cancer Stem cells and their tail vein assay as a tool for the study of liver
cellular origins in primary liver and biliary tract promoters and enhancers. Methods Mol Biol
cancers. Hepatology 64(2):645–651. https:// 1015:279–289. https://doi.org/10.1007/
doi.org/10.1002/hep.28485 978-1-62703-435-7_18
26. Sekiya S, Suzuki A (2012) Intrahepatic cholan- 33. Butash KA, Natarajan P, Young A, Fox DK
giocarcinoma can arise from Notch-mediated (2000) Reexamination of the effect of endo-
conversion of hepatocytes. J Clin Invest 122 toxin on cell proliferation and transfection effi-
(11):3914–3918. https://doi.org/10.1172/ ciency. BioTechniques 29(3):610–614,
JCI63065 616, 618–619
27. Fan B, Malato Y, Calvisi DF, Naqvi S, 34. Xin B, Yamamoto M, Fujii K, Ooshio T,
Razumilava N, Ribback S, Gores GJ, Chen X, Okada Y, Watanabe K, Miyokawa N,
Dombrowski F, Evert M, Chen X, Willenbring Furukawa H, Nishikawa Y (2017) Critical role
H (2012) Cholangiocarcinomas can originate of Myc activation in mouse hepatocarcinogen-
from hepatocytes in mice. J Clin Invest 122 esis induced by the activation of AKT and RAS
 Mouse Models of Liver Cancers from Mature Hepatocytes 235

pathways. Oncogene 36(36):5087–5097. 42. Calvisi DF, Wang C, Ho C, Ladu S, Lee SA,
https://doi.org/10.1038/onc.2017.114 Mattu S, Destefanis G, Delogu S,
35. Mendez-Lucas A, Li X, Hu J, Che L, Song X, Zimmermann A, Ericsson J, Brozzetti S,
Jia J, Wang J, Xie C, Driscoll PC, Tschaharga- Staniscia T, Chen X, Dombrowski F, Evert M
neh DF, Calvisi DF, Yuneva M, Chen X (2017) (2011) Increased lipogenesis, induced by
Glucose catabolism in liver tumors induced by AKT-mTORC1-RPS6 signaling, promotes
c-MYC can be sustained by various PKM1/ development of human hepatocellular carci-
PKM2 ratios and pyruvate kinase activities. noma. Gastroenterology 140(3):1071–1083.
Cancer Res 77(16):4355–4364. https://doi. https://doi.org/10.1053/j.gastro.2010.12.
org/10.1158/0008-5472.CAN-17-0498 006
36. Tao J, Zhang R, Singh S, Poddar M, Xu E, 43. Lee SA, Ladu S, Evert M, Dombrowski F, De
Oertel M, Chen X, Ganesh S, Abrams M, Murtas V, Chen X, Calvisi DF (2010) Synergis-
Monga SP (2017) Targeting beta-catenin in tic role of Sprouty2 inactivation and c-Met
hepatocellular cancers induced by coexpression up-regulation in mouse and human hepatocar-
of mutant beta-catenin and K-Ras in mice. cinogenesis. Hepatology 52(2):506–517.
Hepatology 65(5):1581–1599. https://doi. https://doi.org/10.1002/hep.23681
org/10.1002/hep.28975 44. Xu CR, Lee S, Ho C, Bommi P, Huang SA,
37. Tao J, Xu E, Zhao Y, Singh S, Li X, Couchy G, Cheung ST, Dimri GP, Chen X (2009) Bmi1
Chen X, Zucman-Rossi J, Chikina M, Monga functions as an oncogene independent of
SP (2016) Modeling a human hepatocellular Ink4A/Arf repression in hepatic carcinogene-
carcinoma subset in mice through coexpression sis. Mol Cancer Res 7(12):1937–1945.
of met and point-mutant beta-catenin. Hepa- https://doi.org/10.1158/1541-7786.MCR-
tology 64(5):1587–1605. https://doi.org/10. 09-0333
1002/hep.28601 45. Lee SA, Ho C, Roy R, Kosinski C, Patil MA,
38. Hu J, Che L, Li L, Pilo MG, Cigliano A, Tward AD, Fridlyand J, Chen X (2008) Inte-
Ribback S, Li X, Latte G, Mela M, Evert M, gration of genomic analysis and in vivo trans-
Dombrowski F, Zheng G, Chen X, Calvisi DF fection to identify sprouty 2 as a candidate
(2016) Co-activation of AKT and c-Met trig- tumor suppressor in liver cancer. Hepatology
gers rapid hepatocellular carcinoma develop- 47(4):1200–1210. https://doi.org/10.1002/
ment via the mTORC1/FASN pathway in hep.22169
mice. Sci Rep 6:20484. https://doi.org/10. 46. Tward AD, Jones KD, Yant S, Cheung ST, Fan
1038/srep20484 ST, Chen X, Kay MA, Wang R, Bishop JM
39. Ho C, Wang C, Mattu S, Destefanis G, Ladu S, (2007) Distinct pathways of genomic progres-
Delogu S, Armbruster J, Fan L, Lee SA, sion to benign and malignant tumors of the
Jiang L, Dombrowski F, Evert M, Chen X, liver. Proc Natl Acad Sci U S A 104
Calvisi DF (2012) AKT (v-akt murine thy- (37):14771–14776. https://doi.org/10.
moma viral oncogene homolog 1) and N-Ras 1073/pnas.0706578104
(neuroblastoma ras viral oncogene homolog) 47. Patil MA, Lee SA, Macias E, Lam ET, Xu C,
coactivation in the mouse liver promotes rapid Jones KD, Ho C, Rodriguez-Puebla M, Chen
carcinogenesis by way of mTOR (mammalian X (2009) Role of cyclin D1 as a mediator of
target of rapamycin complex 1), FOXM1 (fork- c-Met- and beta-catenin-induced hepatocarci-
head box M1)/SKP2, and c-Myc pathways. nogenesis. Cancer Res 69(1):253–261.
Hepatology 55(3):833–845. https://doi.org/ https://doi.org/10.1158/0008-5472.CAN-
10.1002/hep.24736 08-2514
40. Wang C, Delogu S, Ho C, Lee SA, Gui B, 48. Zhang S, Song X, Cao D, Xu Z, Fan B, Che L,
Jiang L, Ladu S, Cigliano A, Dombrowski F, Hu J, Chen B, Dong M, Pilo MG, Cigliano A,
Evert M, Calvisi DF, Chen X (2012) Inactiva- Evert K, Ribback S, Dombrowski F, Pascale
tion of Spry2 accelerates AKT-driven hepato- RM, Cossu A, Vidili G, Porcu A, Simile MM,
carcinogenesis via activation of MAPK and Pes GM, Giannelli G, Gordan J, Wei L,
PKM2 pathways. J Hepatol 57(3):577–583. Evert M, Cong W, Calvisi DF, Chen X (2017)
https://doi.org/10.1016/j.jhep.2012.04.026 Pan-mTOR inhibitor MLN0128 is effective
41. Stauffer JK, Scarzello AJ, Andersen JB, De against intrahepatic cholangiocarcinoma in
Kluyver RL, Back TC, Weiss JM, Thorgeirsson mice. J Hepatol 67(6):1194–1203. https://
SS, Wiltrout RH (2011) Coactivation of AKT doi.org/10.1016/j.jhep.2017.07.006
and beta-catenin in mice rapidly induces forma- 49. Carlson CM, Frandsen JL, Kirchhof N, McIvor
tion of lipogenic liver tumors. Cancer Res 71 RS, Largaespada DA (2005) Somatic integra-
(7):2718–2727. https://doi.org/10.1158/ tion of an oncogene-harboring Sleeping
0008-5472.CAN-10-2705 Beauty transposon models liver tumor
236 Masahiro Yamamoto et al.

development in the mouse. Proc Natl Acad Sci 51. Chow EK, Fan LL, Chen X, Bishop JM (2012)
U S A 102(47):17059–17064. https://doi. Oncogene-specific formation of chemoresis-
org/10.1073/pnas.0502974102 tant murine hepatic cancer stem cells. Hepatol-
50. Tao J, Calvisi DF, Ranganathan S, Cigliano A, ogy 56(4):1331–1341. https://doi.org/10.
Zhou L, Singh S, Jiang L, Fan B, 1002/hep.25776
Terracciano L, Armeanu-Ebinger S, 52. Ju HL, Ahn SH, Kim DY, Baek S, Chung SI,
Ribback S, Dombrowski F, Evert M, Chen X, Seong J, Han KH, Ro SW (2013) Investigation
Monga SPS (2014) Activation of beta-catenin of oncogenic cooperation in simple liver-
and Yap1 in human hepatoblastoma and induc- specific transgenic mouse models using nonin-
tion of hepatocarcinogenesis in mice. Gastro- vasive in vivo imaging. PLoS One 8(3):
enterology 147(3):690–701. https://doi.org/ e59869. https://doi.org/10.1371/journal.
10.1053/j.gastro.2014.05.004 pone.0059869
 Chapter 21

Mouse Model for Cholangiocarcinoma from Peribiliary

Glands
Hayato Nakagawa, Nobumi Suzuki, and Kazuhiko Koike

Abstract
A good mouse model is mandatory for elucidating carcinogenic mechanisms and identifying the cellular
origin of cancer. Although the lack of an appropriate mouse model has hampered investigation of extrahe-
patic cholangiocarcinoma (ECC), we recently established a novel mouse model of biliary injury-related
ECC by ductal cell-specific activation of Kras and deletion of transforming growth factor (TGF) β receptor
type 2 and E-cadherin. Using this mouse model, we identified that peribiliary glands, which are considered
a biliary epithelial stem cell niche, are potential cellular origins of ECC. Furthermore, we established an
extrahepatic biliary organoid-derived xenograft cholangiocarcinoma (CC) model by lentiviral induction of
Cre in organoids. This organoid system recreated the in vivo conditions and facilitated analysis of carcino-
genesis. In this chapter, we describe the protocol used to establish our mouse model of ECC derived from
peribiliary glands and our extrahepatic biliary organoid-derived xenograft model of CC.

Key words Extrahepatic cholangiocarcinoma, Biliary organoid, Peribiliary gland, Inflammation, E-

cadherin

1 Introduction

Cholangiocarcinoma (CC) is a highly malignant cancer, and its

incidence is increasing worldwide [1]. CC is commonly classified
into intrahepatic and extrahepatic CC (ICC and ECC, respectively)
according to anatomical location, and ECC is further divided into
hilar and distal CC [2]. As 10–20% of CC cases occur in the
presence of chronic biliary injury and inflammation, such as primary
sclerosing cholangitis, liver flukes, and hepatitis viral infection,
chronic inflammation plays a key role in CC development [1].
Importantly, recent studies using next-generation sequencing tech-
nology have provided comprehensive mutation profiling of CC,
identifying an altered mutation spectrum depending on anatomical
subtype and underlying etiology [3, 4].
Recent genetic lineage-tracing studies have provided new
insights into the cellular origin of CC. Although CC has been

Naoki Tanimizu (ed.), Hepatic Stem Cells: Methods and Protocols, Methods in Molecular Biology, vol. 1905,
https://doi.org/10.1007/978-1-4939-8961-4_21, © Springer Science+Business Media, LLC, part of Springer Nature 2019

237
238 Hayato Nakagawa et al.

considered to originate from biliary epithelial cells (BECs), mature

hepatocytes also give rise to ICC via activation of Notch signaling
[5, 6]. In regard to ECC, the anatomical and immunohistochemi-
cal analyses revealed that peribiliary glands (PBGs), clusters of
epithelial cells residing in the submucosal compartment of extrahe-
patic bile ducts (EHBDs), form epithelial networks within the walls
of EHBDs and are a potential stem cell niche of BECs; thus, PBGs
have attracted attention as a potential origin of ECC [7, 8].
A good mouse model is necessary to identify the cellular origin
of cancer. Several mouse models of ICC have been established and
have contributed to our understanding of the molecular carcino-
genic mechanisms of ICC. However, because most of these models
were generated by liver-specific genetic manipulation or hepatotox-
ins, CC development was limited in the intrahepatic and perihilar
areas [5, 9–11]. The lack of an appropriate mouse model has been a
considerable barrier to the study of ECC.
Recently, we established a new mouse model of ECC by ductal
cell-specific gene manipulation using K19CreERT mice, in which
tamoxifen-inducible Cre ERT was knocked into the endogenous
K19 locus [12]. Based on recent human studies showing
frequent genetic alterations of Ras and transforming growth factor
(TGF)β/SMAD signaling pathways in ICC and ECC, mice exhibit-
ing ductal cell-specific Kras activation and TGFβ receptor type
2 (TGFβR2) inactivation were generated by crossing LSL-
KrasG12D, Tgfbr2 flox/flox, and K19CreERT mice (KT-K19CreERT).
KT-K19CreERT mice showed only mild hyperplasia of BECs in the
extrahepatic bile duct; however, additional loss of E-cadherin by
crossing KT-K19CreERT mice with CDH1 flox/flox mice (KTC-
K19CreERT) resulted in development of periductal infiltrating-type
ECC, the progression of which depended on biliary epithelial
injury-induced inflammatory and regenerative responses. Time-
course analysis of EHBDs after tamoxifen administration revealed
that recombined BECs lining the bile duct lumen had died and
detached as a result of E-cadherin loss, whereas recombined cells
survived in the PBGs, which provide a cancer stem cell niche. Cell
lineage tracing suggested that PBGs are the cellular origin of ECC.
Furthermore, we established an extrahepatic biliary organoid-
derived xenograft CC model by lentiviral induction of Cre into
organoids derived from the EHBD of Cre-negative KTC mice.
Interestingly, detachment of BECs was also seen in Cre+ KTC
organoids, which was similar to that seen in the in vivo model.
Therefore, this EHBD organoid system recreated the in vivo con-
ditions and facilitated analysis of carcinogenesis induced by genetic
manipulations in BECs separately from the tumor microenviron-
ment. Using these novel mouse and organoid models, we showed
that dying BECs release interleukin-33, which stimulated inflam-
mation and subsequent regeneration of the EHBD by PBGs via
type 2 innate lymphoid cells, eventually leading to ECC [12]. Thus,
 Mouse Model for Cholangiocarcinoma from Peribiliary Glands 239

these models are useful tools for investigating carcinogenic

mechanisms, cellular origin, and potential therapeutic targets
of ECC.
Here, we describe the protocols used to establish our mouse
model of ECC derived from PBGs and our extrahepatic biliary
organoid-derived xenograft CC model.

2 Materials

2.1 In Vivo Model 1. KTC-K19CreERT mice: Cross K19CreERT, LSL-KrasG12D+/ ,

of ECC Tgfbr2 flox/flox, and CDH1 flox/flox mice [13–15]. All mice were
of the C57BL/6 genetic background.
2. Tamoxifen: 20 mg/mL in corn oil containing 10% ethanol.
Mix for 30 min using a tube rotor.

2.2 Extrahepatic 1. Cre-negative KTC mice: Crossing LSL-KrasG12D+/ ,

Biliary Organoid- Tgfbr2 flox/flox, and CDH1 flox/flox mice.
Derived Xenograft CC 2. Nude mice.
Model
3. Phosphate-buffered saline (PBS)
4. Digestion buffer: 4% fetal calf serum, 1 mg/mL collagenase D,
0.5 g/mL dispase, 40 μg/mL DNase I, and 1 mM EDTA
in RPMI.
5. Organoid culture medium: Advanced Dulbecco’s Modified
Eagle Medium/F-12 supplemented with B27 (1:50) and N2
(1:100), 1 mM N-acetylcysteine, 10 mM nicotinamide,
50 ng/mL epidermal growth factor, 1 μg/mL Rspo1,
100 ng/mL Noggin, 100 ng/mL fibroblast growth factor
10, 10% Wnt3a-conditioned medium, and penicillin/
streptomycin.
6. The ROCK inhibitor Y27632 (final 10 μM).
7. Matrigel.
8. Cell Recovery Solution (BD Biosciences).
9. 29-gauge insulin syringe.
10. Puromycin (final 5 μg/mL).

3 Methods

3.1 In Vivo Model 1. Tamoxifen (200 mg/kg) was administered by oral gavage to
of ECC KTC-K19CreERT mice (7–8 weeks old) for 3 consecutive days.
This protocol can induce gene recombination (~40%) in all
parts of the biliary tree, including the EHBD, PBGs, GB,
perihilar bile duct (BD), and intrahepatic bile duct.
2. On day 7, surface BECs started to detach from the BD wall.
240 Hayato Nakagawa et al.

Fig. 1 Histological findings of extrahepatic bile duct (EHBD) in KTC-K19CreERT mice. (a) Hematoxylin and eosin
(H&E)-stained images of EHBDs and PBGs from Cre-negative control and KTC-K19CreERT mice at 10 days after
tamoxifen administration (scale bar: upper panels, 200 μm; lower panels, 50 μm). Right panel shows enlarged
and dysplastic peribiliary glands (PBGs). (b) H&E-stained images of EHBDs from Cre-negative control and
KTC-K19CreERT mice at 3.5 weeks after tamoxifen administration (scale bar: left two panels, 500 μm; right
panel, 50 μm)

3. On day 10, many shedding BECs were observed in the BD

lumen, and inflammatory cells had infiltrated the subepithelial
connective tissue. The PBGs were enlarged and morphologi-
cally dysplastic, accompanied by mitotic figures (Fig. 1a).
4. On day 15, although the shedding of surface BECs had sub-
sided, the cancerous glands had spread through the subepithe-
lial area, accompanied by a fibrotic response.
5. On day 20, most of the histological features of periductal
infiltrating-type ECC were apparent (Fig. 1b) (see Note 1).
6. All KTC-K19CreERT mice died within 4 weeks of tamoxifen
administration mainly from liver failure or respiratory failure
caused by lung adenocarcinoma.
 Mouse Model for Cholangiocarcinoma from Peribiliary Glands 241

3.2 Extrahepatic 1. Isolate the EHBD tissue from Cre-negative KTC mice and
Biliary Organoid- place tissue in a 10-cm petri dish filled with PBS.
Derived Xenograft CC 2. Remove the adhered interstitial and pancreatic tissue from the
Model EHBD using forceps with gentle touching.
3.2.1 Culture of EHBD 3. Cut the EHBD tissue down the middle and divide it into two
Organoids parts.
4. Place the EHBD tissue in a 1.5-mL tube containing 1 mL
digestion buffer.
5. Incubate for 30 min at 37  C.
6. Gently disperse the EHBD tissue by pipetting.

7. Centrifuge at 300  g for 5 min at 4 C. Discard the
supernatant.
8. Place the 1.5-mL tube containing EHBD cells on ice and add
1 mL cold PBS. Vortex for 10 s.

9. Centrifuge at 300  g for 5 min at 4 C. Discard the
supernatant.
10. Place the 1.5-mL tube containing EHBD cells on ice. Add
80 μL ice-cold Matrigel and mix by pipetting using a cold
200-μL tip with a large-end diameter (cut off end of the tip)
(see Note 2). Do not generate bubbles while mixing.
11. Seed EHBD cells in a 40 μL drop in the center of each well of a
24-well plate.
12. Carefully place the 24-well plate in a 37  C incubator for
10 min.
13. Add 500 μL culture medium supplemented with ROCK inhib-
itor to each well.
14. Organoid formation was observed 1–2 days after plating.
EHBD organoids formed a ball-like structure and were com-
posed of a single layer of the biliary epithelium (Fig. 2a, b).
15. Refresh the culture medium every 2–3 days and passage the
EHBD organoids every 5–10 days.
16. When passaging organoids, remove the culture medium and
add 1 mL Cell Recovery Solution to dissolve the Matrigel.
17. Place the solution in a 1.5-mL tube and incubate 10 min on ice
to precipitate the organoids by gravity. Remove the upper
portion of the supernatant (~700 μL), which contains dead
cells and debris.
18. Add 700 μL Cell Recovery Solution and suspend the
organoids well.
19. Pass the solution through a 29-gauge insulin syringe once or
twice.
242 Hayato Nakagawa et al.

Fig. 2 Extrahepatic biliary organoid-derived xenograft CC model. (a) Micrograph of biliary organoids cultured
from EHBDs of mice (scale bar: 250 μm). (b) H&E-stained images of EHBD organoids from Cre-negative KTC
mice infected with Cre-expressing or control lentivirus (scale bar: 50 μm). (c) H&E-stained image of an
organoid-derived tumor (scale bar: 50 μm)


20. Centrifuge at 600  g for 5 min at 4 C. Discard the
supernatant.
21. Add 400 μL cold PBS and suspend well.
22. Transfer approximately 800 organoid fragments to a new 1.5-
mL tube on ice.

23. Centrifuge at 600  g for 5 min at 4 C. Discard the
supernatant.
24. Add 160 μL Matrigel and suspend well.
25. Seed 40 μL organoid solution into each well of a 24-well plate.
26. Carefully place the 24-well plate in a 37  C incubator for
10 min.
27. Add 500 μL culture medium supplemented with ROCK inhib-
itor into each well.
28. Place the plate in a 37  C incubator.

3.2.2 Induction of Gene 1. Resolve Matrigel and wash organoids in the same way with
Recombination in EHBD Subheading 3.2.1, steps 16–23.
Organoids Using a 2. Add 15 μL lentivirus solution (titer 1  107 transduction
Cre-Expressing Lentivirus units/mL) and suspend well.
3. Incubate 30 min on ice.
 Mouse Model for Cholangiocarcinoma from Peribiliary Glands 243

4. Add 160 μL Matrigel and suspend well.

5. Seed 40 μL organoid solution into each well of a 24-well plate.
6. Carefully place the 24-well plate in a 37  C incubator for
10 min.
7. Add 500 μL culture medium supplemented with ROCK inhib-
itor into each well.
8. Place the plate in a 37  C incubator.
9. After 2–3 days, replace the culture medium with 500 μL
medium containing the selection drug (e.g., 5 μg/mL
puromycin).
10. After 2 weeks of selection, confirm gene recombination by
polymerase chain reaction or Western blot. Organoids with
successful recombination of LSL-KrasG12D, Tgfbr2 flox/flox, and
CDH flox/flox alleles will exhibit shedding BECs (Fig. 2b).

3.2.3 Extrahepatic Biliary 1. Culture EHBD organoids from Cre-negative KTC mice and
Organoid-Derived induce gene recombination as described above.
Xenograft CC Model 2. Seed 100 organoid fragments in each well of a 24-well plate.
3. After 2 days, remove the medium.
4. Add 1 mL cold PBS to each well and mix with the Matrigel
containing organoids.
5. Place the solution in a 1.5-mL tube on ice.
6. Centrifuge at 600  g for 5 min at 4  C.
7. Discard the supernatant and leave approximately 100 μL Matri-
gel/PBS mixture containing organoids.
8. Add 100 μL Matrigel/PBS mixture (1:1) and suspend well. At
this stage, 200 ul suspension remains.
9. Inject 100 μL of the organoid suspension subcutaneously in
two sites at the back of nude mice using an ice-cold 29-gauge
insulin syringe.
10. After 2 months, approximately two-thirds of the organoids will
have formed tumors with a similar histology to that of ECC in
KTC-K19CreERT mice (Fig. 2c).

4 Notes

1. ECC is induced more stably in male mice (100%) than in female

mice (80%).
2. Matrigel should be handled on ice as much as possible to
prevent it from gelling.
244 Hayato Nakagawa et al.

Acknowledgments

This study was supported by the Japan Society for the Promotion of
Science KAKENHI grant number 15K09039, Astellas Foundation
for Research on Metabolic Disorders, Nakayama Cancer Research
Institute, Okinaka Memorial Institute for Medical Research,
Research Grant of the Princess Takamatsu Cancer Research Fund,
and Foundation for Promotion of Cancer Research in Japan.

References
1. Rizvi S, Gores GJ (2013) Pathogenesis, diag- 7. Dipaola F, Shivakumar P, Pfister J, Walters S,
nosis, and management of cholangiocarci- Sabla G, Bezerra JA (2013) Identification of
noma. Gastroenterology 145(6):1215–1229. intramural epithelial networks linked to peri-
https://doi.org/10.1053/j.gastro.2013.10. biliary glands that express progenitor cell mar-
013 kers and proliferate after injury in mice.
2. Malhi H, Gores GJ (2006) Cholangiocarci- Hepatology 58(4):1486–1496. https://doi.
noma: modern advances in understanding a org/10.1002/hep.26485
deadly old disease. J Hepatol 45(6):856–867. 8. Lanzoni G, Cardinale V, Carpino G (2016)
https://doi.org/10.1016/j.jhep.2006.09.001 The hepatic, biliary, and pancreatic network of
3. Chan-On W, Nairismagi ML, Ong CK, Lim stem/progenitor cell niches in humans: a new
WK, Dima S, Pairojkul C, Lim KH, McPherson reference frame for disease and regeneration.
JR, Cutcutache I, Heng HL, Ooi L, Chung A, Hepatology 64(1):277–286. https://doi.org/
Chow P, Cheow PC, Lee SY, Choo SP, Tan IB, 10.1002/hep.28326
Duda D, Nastase A, Myint SS, Wong BH, 9. Xu X, Kobayashi S, Qiao W, Li C, Xiao C,
Gan A, Rajasegaran V, Ng CC, Nagarajan S, Radaeva S, Stiles B, Wang RH, Ohara N,
Jusakul A, Zhang S, Vohra P, Yu W, Huang D, Yoshino T, LeRoith D, Torbenson MS, Gores
Sithithaworn P, Yongvanit P, Wongkham S, GJ, Wu H, Gao B, Deng CX (2006) Induction
Khuntikeo N, Bhudhisawasdi V, Popescu I, of intrahepatic cholangiocellular carcinoma by
Rozen SG, Tan P, Teh BT (2013) Exome liver-specific disruption of Smad4 and Pten in
sequencing identifies distinct mutational pat- mice. J Clin Invest 116(7):1843–1852.
terns in liver fluke-related and non-infection- https://doi.org/10.1172/JCI27282
related bile duct cancers. Nat Genet 45 10. Ikenoue T, Terakado Y, Nakagawa H,
(12):1474–1478. https://doi.org/10.1038/ Hikiba Y, Fujii T, Matsubara D, Noguchi R,
ng.2806 Zhu C, Yamamoto K, Kudo Y, Asaoka Y,
4. Nakamura H, Arai Y, Totoki Y, Shirota T, Yamaguchi K, Ijichi H, Tateishi K,
Elzawahry A, Kato M, Hama N, Hosoda F, Fukushima N, Maeda S, Koike K, Furukawa Y
Urushidate T, Ohashi S, Hiraoka N, (2016) A novel mouse model of intrahepatic
Ojima H, Shimada K, Okusaka T, Kosuge T, cholangiocarcinoma induced by liver-specific
Miyagawa S, Shibata T (2015) Genomic spec- Kras activation and Pten deletion. Sci Rep
tra of biliary tract cancer. Nat Genet 47 6:23899. https://doi.org/10.1038/
(9):1003–1010. https://doi.org/10.1038/ srep23899
ng.3375 11. Guest RV, Boulter L, Kendall TJ, Minnis-
5. Sekiya S, Suzuki A (2012) Intrahepatic cholan- Lyons SE, Walker R, Wigmore SJ, Sansom
giocarcinoma can arise from Notch-mediated OJ, Forbes SJ (2014) Cell lineage tracing
conversion of hepatocytes. J Clin Invest 122 reveals a biliary origin of intrahepatic cholan-
(11):3914–3918. https://doi.org/10.1172/ giocarcinoma. Cancer Res 74(4):1005–1010.
JCI63065 https://doi.org/10.1158/0008-5472.CAN-
6. Fan B, Malato Y, Calvisi DF, Naqvi S, 13-1911
Razumilava N, Ribback S, Gores GJ, 12. Nakagawa H, Suzuki N, Hirata Y, Hikiba Y,
Dombrowski F, Evert M, Chen X, Willenbring Hayakawa Y, Kinoshita H, Ihara S, Uchino K,
H (2012) Cholangiocarcinomas can originate Nishikawa Y, Ijichi H, Otsuka M, Arita J,
from hepatocytes in mice. J Clin Invest 122 Sakamoto Y, Hasegawa K, Kokudo N,
(8):2911–2915. https://doi.org/10.1172/ Tateishi K, Koike K (2017) Biliary epithelial
JCI63212 injury-induced regenerative response by IL-33
 Mouse Model for Cholangiocarcinoma from Peribiliary Glands 245

promotes cholangiocarcinogenesis from peri- signaling in cooperation with active Kras

biliary glands. Proc Natl Acad Sci U S A 114 expression. Genes Dev 20(22):3147–3160.
(19):E3806–E3815. https://doi.org/10. https://doi.org/10.1101/gad.1475506
1073/pnas.1619416114 15. Nakagawa H, Hikiba Y, Hirata Y, Font-
13. Means AL, Xu Y, Zhao A, Ray KC, Gu G Burgada J, Sakamoto K, Hayakawa Y,
(2008) A CK19(CreERT) knockin mouse line Taniguchi K, Umemura A, Kinoshita H,
allows for conditional DNA recombination in Sakitani K, Nishikawa Y, Hirano K,
epithelial cells in multiple endodermal organs. Ikenoue T, Ijichi H, Dhar D, Shibata W,
Genesis 46(6):318–323. https://doi.org/10. Akanuma M, Koike K, Karin M, Maeda S
1002/dvg.20397 (2014) Loss of liver E-cadherin induces scle-
14. Ijichi H, Chytil A, Gorska AE, Aakre ME, rosing cholangitis and promotes carcinogene-
Fujitani Y, Fujitani S, Wright CV, Moses HL sis. Proc Natl Acad Sci U S A 111
(2006) Aggressive pancreatic ductal adenocar- (3):1090–1095. https://doi.org/10.1073/
cinoma in mice caused by pancreas-specific pnas.1322731111
blockade of transforming growth factor-beta
 INDEX

A Extrahepatic Cholangiocarcinoma
(ECC) ...............................................237–240, 243
Alginate hydrogel beads ...................................... 158–160
Animal models..........................46, 47, 50, 118, 222, 223 F
B Fetal liver cells (FLCs) ............................... 3, 7, 117, 124,
175, 179, 181–183
Bile ducts .....................46, 143, 144, 188–190, 223, 238
Fluorescence Activated Cell Sorting (FACS) ..13, 14, 16,
Biliary epithelial cells (BECs) ....................... v, 45–55, 83, 20–25, 109–111, 217
84, 125, 175, 176, 178, 180–182, 188, 189, 191,
237–240, 243 H
Biliary organoid...................................238–240, 242, 243
Biliary tree ................................. 178–180, 188, 189, 194, Hepatic progenitor cells .......................... 5, 8, 71, 72, 78,
196, 197, 239 80, 81, 144–146, 148–150, 211
Bipotential ..............................................v, 9–17, 122, 211 Hepatic stellate cells (HSCs) .............................. 104, 105,
Bipotentiality ................................................................. 117 108, 110, 111, 131, 139–141, 211
Hepatoblasts .............................................v, 144, 151, 211
C Hepatocellular carcinoma .....................54, 221, 224, 227
Hepatocyte ........................................ 4, 9, 22, 29, 45, 59,
Canine model ....................................................... 201–208 83, 93, 103, 117, 131, 143, 157, 167, 175, 187,
Carbon tetrachloride (CCl4)............... 47, 49, 51–53, 61,
201, 222, 237
80, 104, 105, 108, 111, 114, 201–203, 206–208 Hepatocyte ablation ..................................................84, 89
CD44 ..................................................... 30, 33, 36–38, 40 Heterogeneity....................................................... 9, 71–81
Cell proliferation .................................46, 50, 72, 80, 128
Human Induced Pluripotent Stem Cells (Human Ips
Chemically Induced Liver Progenitors Cells) ............................... 145, 146, 148, 150–152
(CLiPs)......................................... 31, 34, 117–129 Hybrid periportal hepatocytes........................................ 60
Chemical screen ........................................................83–89 Hydrodynamic tail vein injection
Cholangiocytes .............................v, 9, 10, 15, 16, 19–27,
(HTVi) .....................................222–227, 229, 232
46, 47, 143–146, 149, 151, 211, 227
Chronic liver injury ...................................................19, 80 I
Clonal culture.................................................................. 16
Colony ................................ 5, 10, 11, 14–16, 19, 20, 26, Induced hepatocytes (Iheps) .............. 103–105, 108–112
38, 65, 71, 72, 76–78, 81, 146, 149, 216 Induction of hepatic differentiation............................. 124
Colony assay .................................................................... 14 Inflammation .................................... 50, 52, 83, 201, 237
Interstitial flow .............................................................. 167
D Intrahepatic cholangiocarcinoma ................................. 221
In vivo reprogramming........................................ 104, 108
Differentiation...................................... 10, 11, 14–16, 20,
45, 71, 83, 124, 135, 139, 141, 144–146, L
148–151, 212, 231
Duct cells ................................................... 19, 59, 60, 144 Lineage tracing ..... 9, 45–55, 59–69, 104, 105, 111, 237
Ductular reaction (DR) ......................................... 46, 188 Liver3, 9, 19, 29, 45, 59, 72, 83, 93, 103, 117, 131, 143,
157, 167, 175, 187, 201, 211, 221, 237
E Liver cancer .......................................................v, 221–224
Liver cirrhosis (LC).............................................. 201, 202
E-cadherin ............................................................ 211, 238 Liver progenitor cells (LPCs) ....................................... 136
Extracellular matrix ........... 103, 145, 146, 149–151, 168
Liver regeneration .................. v, 46, 59, 83–89, 187, 201

Naoki Tanimizu (ed.), Hepatic Stem Cells: Methods and Protocols, Methods in Molecular Biology, vol. 1905,
https://doi.org/10.1007/978-1-4939-8961-4, © Springer Science+Business Media, LLC, part of Springer Nature 2019

247
 HEPATIC STEM CELLS: METHODS AND PROTOCOLS
248 Index
Liver sinusoidal endothelial cells (LSECs)..................131, P
136, 138, 140, 141
Liver stem/progenitor cell (LPCs) ....................... v, 9–17, Partial hepatectomy (PHx)........................ 45, 47, 48, 50,
46, 71–81, 83, 117, 131, 134, 135, 140, 141, 52, 212, 215, 216
176, 187, 224 Peribiliary glands (PBGs) ........................... 237, 238, 240
Liver tumor .......................................................... 223, 227
Q
Long-term proliferation ........................................ 54, 125
Quantitative analysis .............................................. 71, 206
M
R
Mesenchymal stem cells ........................94, 203, 205, 207
Methylcellulose (MC).........................157–161, 163, 164 Reprogramming .................. 93, 103, 104, 118, 124, 223
Microfluidic device............................................... 167–174 Retrorsine .....................................................212, 214–218
Microscopy .....................................................74, 112, 162
Microstructures ............................................................. 157 S
Mouse .........................5, 6, 8–17, 19–27, 29, 32, 33, 37,
Sleeping Beauty transposon................................. 223, 224
40, 47, 49, 52, 55, 60, 62, 64–66, 69, 72–74, 78,
Small hepatocyte (SHs) ..................................... 30, 36, 38
84, 93, 103, 110, 114, 125, 132, 137, 144, 145,
Spheroid..................................4–6, 8, 167, 176, 181–184
148, 163, 177, 189, 192, 196, 197, 211, 212,
Suppressor of tumorigenicity 14 protein
215, 221–232, 237–243
(ST14)...............................................19, 21, 24–27
Mouse model................................v, 19, 59, 84, 103, 105,
221, 238, 239 T
Multicellular spheroid (MCS) ............................. 157–164
3D culture ............................................................ 167, 168
N Three-dimensional imaging............................................ 77
3D tissue structures ...................................v, 72, 168, 173
Nitroreductase (NTR) ...............84, 85, 88, 89, 105, 111
Tissue engineering ............................................... 167, 168
Tissue stem cell ............................................................... 59
O
Transdifferentiation................................... 46, 93–95, 100
Organoid ............................................19–21, 24, 26, 144, Transplantation .................................. v, 29, 83, 117, 175,
175–185, 242 201, 211–218, 222