GLIA 61:24–36 (2013)

The ‘‘Quad-Partite’’ Synapse: Microglia-Synapse

Interactions in the Developing and Mature CNS
DOROTHY P. SCHAFER, EMILY K. LEHRMAN, AND BETH STEVENS*
Department of Neurology, F.M. Kirby Neurobiology Center, Children’s Hospital, Harvard Medical School, Boston, Massachusetts

KEY WORDS Perry, 2009). While these rapid and robust responses to
glia; pruning; plasticity; neurotransmission; cortex; hippo- CNS injury and disease make microglia interesting can-
campus; retinogeniculate didates for developing diagnostic and therapeutic strat-
egies, these qualities also make microglia very difficult
to study in the context of the healthy brain. Culture
ABSTRACT
models such as dissociated cells and brain slices can
Microglia are the resident immune cells and phagocytes of
our central nervous system (CNS). While most work has mimic injury resulting in ‘‘activation’’ of microglia, a
focused on the rapid and robust responses of microglia dur- graded cellular state most associated with injury that is
ing CNS disease and injury, emerging evidence suggests characterized by morphological characteristics more
that these mysterious cells have important roles at CNS closely resembling peripheral macrophages (amoeboid,
synapses in the healthy, intact CNS. Groundbreaking live rod-like, etc.) as well as functional changes such as
imaging studies in the anesthetized, adult mouse demon- increased phagocytic capacity, chemotaxis, proliferation,
strated that microglia processes dynamically survey their and expression of proinflammatory molecules (Ketten-
environment and interact with other brain cells including mann et al., 2011; Ransohoff and Perry, 2009). Thus,
neurons and astrocytes. More recent imaging studies have while in vitro studies are valuable for understanding
revealed that microglia dynamically interact with synapses
where they appear to serve as ‘‘synaptic sensors,’’ respond-
some of the basic cellular biology, in vivo strategies
ing to changes in neural activity and neurotransmitter must also be utilized to understand the function of these
release. In the following review, we discuss the most recent cells in the context of the healthy, intact CNS.
work demonstrating that microglia play active roles at In 2000, a transgenic mouse was engineered that
developing and mature synapses. We first discuss the enabled visualization of microglia by EGFP (CX3CR11/
important imaging studies that have led us to better under- EGFP) (Jung et al., 2000). These mice, combined with
stand the physical relationship between microglia and syn- new technology to image the brain in a live, anesthe-
apses in the healthy brain. Following this discussion, we tized mouse by two-photon microscopy, resulted in
review known molecular mechanisms and functional conse- groundbreaking findings in 2005. Two studies used a
quences of microglia-synapse interactions in the developing
thin-skulled transcranial approach and demonstrated
and mature CNS. Our current knowledge sheds new light
on the critical functions of these mysterious cells in synapse that, in contrast to their name, ‘‘resting’’ microglia in
development and function in the healthy CNS, but has also the healthy, adult cerebral cortex were dynamic, contin-
incited several new and interesting questions that remain uously surveying their extracellular environment
to be explored. We discuss these open questions, and how (Davalos et al., 2005; Nimmerjahn et al., 2005). Time-
the most recent findings in the healthy CNS may be related lapse imaging revealed that microglia rapidly extend
to pathologies associated with abnormal and/or loss of and retract their processes on the order of minutes while
neural circuits. V 2012 Wiley Periodicals, Inc.
C
their cell bodies remain stationary. In fact within a few
hours, microglia processes have the capacity to sample
the entire brain parenchyma and physically interact
INTRODUCTION with other cortical cells including astrocytes and neu-
rons (Nimmerjahn et al., 2005). These key findings beg
Nearly a century has passed since the concept of the question, what is the function of microglia in
microglia was originally described in the healthy mam- the healthy brain? The following review will focus on
malian CNS by Pio del Rio Hortega (del Rio-Hortega, the most recent studies that suggest that microglia play
1932). Surprisingly, there is still very little known important roles at synapses in the developing and
regarding the role of these mysterious cells in the nor- mature CNS and that synaptic activity may concomi-
mal, healthy CNS. A vast majority of studies have tantly modulate microglia function.
focused on the role of microglia in the context of disease,
in which they display an extraordinary ability to
respond rapidly and perform a broad range of functions *Correspondence to: Beth Stevens, Department of Neurology, Children’s Hospital
Boston, Boston, MA 02115, USA. E-mail: [email protected]
such as shielding injury sites, phagocytosing cellular Received 6 February 2012; Accepted 21 June 2012
material, and releasing inflammatory signals to initiate DOI 10.1002/glia.22389
and/or propagate the immune response (Hanisch and Published online 24 July 2012 in Wiley Online Library (wileyonlinelibrary.
Kettenman, 2007; Kreutzberg, 1996; Ransohoff and com).

C 2012
V Wiley Periodicals, Inc.
 MICROGLIA-SYNAPSE INTERACTIONS 25
tation, data suggesting a role for ATP in regulating
IMAGING INTERACTIONS BETWEEN MICROGLIA
microglia dynamics under more basal conditions (Fontai-
AND SYNAPTIC CIRCUITS nhas et al., 2011).
Similar to purinergic and glutamatergic signaling, in-
The pioneering imaging studies conducted in 2005 laid hibitory neurotransmission has also been implicated in
a foundation for more work focused on the interactions regulating microglial dynamics. In the same ex vivo
between microglia and synapses in response to sponta- study in the retina described above the inhibitory neuro-
neous and sensory experience-driven changes in neuro- transmitter, GABA, decreased microglia process motility
nal activity. The following sections will discuss key and overall velocity, while the GABAA receptor antago-
imaging studies demonstrating that microglia have the nist, bicuculline, increased motility and velocity. These
capacity to rapidly respond to changes in neurotrans- data are similar to the in vivo work by Nimmerjahn
mitters and physically interact with synapses in an ac- et al. (2005) that also showed an increase in motility
tivity-dependent manner. upon bicuculline application; however, overall velocity
was unchanged, data that may reflect differences
between ex vivo and in vivo preparations and/or region
Imaging the Effects of Neurotransmission on analyzed (i.e., retina vs. cortex).
Microglial Dynamics Overall, the studies discussed above suggest that
increased excitatory neurotransmission and purinergic
In vitro preparations of microglia demonstrate that signaling can cause an increase in microglial process
microglia have the capacity to express receptors for and motility, while inhibitory neurotransmission may
respond to neurotransmitters such as acetylcholine, decrease this motility. In contrast, one recently pub-
gamma-aminobutyric acid (GABA), glutamate, and puri- lished study by Grinberg et al. (2011) used an in vitro
nergics including adenosine-50 -triphosphate (ATP). Such hippocampal slice preparation to assess microglial dy-
studies have demonstrated that neurotransmitters can namics in response to spreading depression and has
affect microglia in vitro in numerous ways, including results that suggest a seemingly opposite result. When
changes in membrane potential, intracellular calcium, long term potentiation (LTP) was induced by increasing
cytokine release, and overall cellular motility, reviewed cyclic adenosine monophosphate (cAMP), enhanced neu-
in (Biber et al., 2007; Kettenmann et al., 2011; Pocock ral transmission induced a decrease in microglial move-
and Kettenmann, 2007). Since these original studies, ment (Grinberg et al., 2011). Conversely, blockade of ac-
several live imaging studies have demonstrated the tivity with tetrodotoxin (TTX) resulted in increased
capacity of microglia to rapidly react and change their movement, and restoration of activity with glutamate or
dynamics in response to neurotransmitters. ATP subsequently decreased movement. However, these
Of the many neurotransmitters known to elicit an results may reflect the in vitro nature of this sample
effect, in vitro and in vivo preparations alike have dem- preparation, regional differences in neurotransmission
onstrated that ATP is one of the most potent signals to (e.g., hippocampus vs. retina or cortex), regional micro-
evoke a microglial response (Farber and Kettenmann, glial heterogeneity, and/or differences in stimulation
2006; Inoue et al., 2007). In 2005, Davalos et al. demon- paradigms (e.g., cAMP, TTX 6 glutamate or ATP vs.
strated that purinergic signaling increased basal micro- direct application of neurotransmitters or antagonists).
glia motility in vivo. In the anesthetized mouse, the Despite discrepancies, it is clear that microglia are
authors applied either apyrase, which hydrolyzes ATP responding to changes in neurotransmission and that
and ADP, or pharmacological blockers of gap junction neurotransmitters can have direct and rapid effects on
hemichannels (carbenoxolone or flufenamic acid) to the the overall dynamics of these cells. Future in vivo work
cortical surface followed by live imaging (Davalos et al., will be required to further dissect the nature of micro-
2005). In all instances, disrupting ATP-dependent sig- glia responses to excitatory and inhibitory neurotrans-
naling resulted in reduced basal velocity of microglial mission as well as identify whether these effects have
processes. More recently, using mice harboring a dele- regional specificity.
tion in the purinergic receptor P2Y12 or A2A, it has been
demonstrated in vitro and in vivo that purinergic signal-
ing through these receptors can mediate microglial Imaging Physical Interactions Between
process extension and chemotaxis (P2Y12) or retraction Microglia and Synapses
(A2A), in response to local nucleotide application or lipo-
polysaccharide (LPS), a potent activator of microglia. The imaging studies discussed above suggest that
(Haynes et al., 2006; Orr et al., 2009). However, it microglia can rapidly change their dynamics in response
remains to be determined whether purinergic signaling to neurotransmission. To better understand how micro-
can regulate microglia under more basal, physiological glia may be physically interacting with synapses in
conditions in the absence of local nucleotide or LPS paradigms known to change neural activity and induce
application. Interestingly, a more recent live imaging synaptic remodeling, recent work has taken advantage
studying using ex vivo retinal explants revealed that of two-photon live imaging in the anesthetized mouse
glutamatergic transmission increased microglia motility which express fluorescent proteins in microglia and neu-
indirectly via increased ATP release upon neuronal exci- rons (Tremblay, 2012). The first imaging study to char-

GLIA
26 SCHAFER ET AL.

acterize physical interactions between microglia and number (Bence and Levelt, 2005; Hooks and Chen,
synaptic elements in vivo used two-photon time-lapse 2007; Majewska and Sur, 2003). During the peak of the
imaging and transgenic mice expressing EGFP in both critical period (P28), high resolution imaging demon-
microglia and neurons (Iba-1-EGFP/Thy1-EGFP M line) strated that microglia in layer II of V1 normally con-
(Feng et al., 2000; Hirasawa et al., 2005; Wake et al., tacted spines, synaptic terminals, and synaptic clefts
2009). This study revealed that microglial processes (Tremblay et al., 2010). Importantly, this landmark
briefly (
5 min) contacted synaptic elements in layers study revealed for the first time that spines often
II–III of the somatosensory and visual cortices at a rate changed size upon microglia contact, data suggesting
of
1 structure/h and that microglial processes appeared that microglia may be key regulators of structural spine
to enlarge once presynaptic terminals were contacted. plasticity. Consistent with this idea, the authors noted
Following these baseline measurements, the authors that spines that changed size upon microglial contact
investigated how changes in neural activity might affect tended to be smaller and, during later imaging sessions,
physical interactions between microglia and synaptic were often eliminated. These data raise the intriguing
elements. Using three independent methods to decrease possibility that microglia may mediate the elimination of
neural activity (enucleation, retinal TTX administration, spines in response to sensory experience and suggest
or lowered body temperature), the authors observed that one potential mechanism could be phagocytosis.
microglial processes retracted and made fewer contacts The idea that glial cells could be phagocytosing cir-
with less active presynaptic terminals within the cortex. cuits undergoing active remodeling is substantiated by
These data are in discordance with previously published earlier work in the Drosophila mushroom body and mam-
work demonstrating no change in microglial motility in malian neuromuscular junction (NMJ) (Freeman, 2006;
response to direct cortical application of TTX in vivo Mallat et al., 2005). In developing Drosophila, glial cells
(Nimmerjahn et al., 2005) or increased microglial move- engulf the axons of mushroom body g neurons undergoing
ment in response to TTX administration in an in vitro anatomical pruning (Awasaki and Ito, 2004; Watts et al.,
slice preparation (Grinberg et al., 2011). However, these 2004). In addition, in the mammalian system, Schwann
discrepancies may easily be explained by method of TTX cells phagocytose axonal arbor remnants (i.e., ‘‘axo-
administration, regional differences, and/or in vivo vs. in somes’’) undergoing pruning at the developing NMJ
vitro preparations. To further assess interactions (Bishop et al., 2004). Consistent with these data, glia
between microglia and synapses, Wake et al. (2009) have been shown to engulf destabilized synaptic boutons
induced ischemic injury by photochemical occlusion of and presynaptic debris at the developing Drosophila NMJ
the middle cerebral artery. Using this paradigm, the (Fuentes-Medel et al., 2009). However, except for early
authors demonstrated that under, microglial contact work suggesting that microglia and astrocytes could be
with presynaptic structures was prolonged and often engulfing axonal debris during large-scale cortical prun-
accompanied or followed by a disappearance of these ing (Berbel and Innocenti, 1988), very little was known
structures. These data suggest a potential function for regarding glial-mediated phagocytosis of developing
microglia in the removal of synapses undergoing, albeit mammalian CNS synapses undergoing local, small-scale
injury-induced, remodeling. remodeling. Thus, to further assess physical interactions
To address whether microglia were indeed playing a between microglia and remodeling CNS synapses, Trem-
role at remodeling synapses under more physiological blay et al. used a dark adaptation paradigm in which ju-
conditions, Tremblay et al., published a recent study venile mice were placed in the dark for 6 days during the
assessing in vivo changes in microglia-synapse interac- visual system critical period (Tremblay et al., 2010). In
tions in the developing primary mouse visual cortex (V1) some cases, mice were then re-exposed to light for 2 days,
(Tremblay et al., 2010). This study was able to attain a paradigm known to elicit synaptic remodeling in V1
better spatial resolution than previous studies by com- (Mower et al., 1983; Philpot et al., 2001; Tropea et al.,
bining high resolution 3D serial electron microscopy (3D 2010; Viegi et al., 2002). Although microglia in V1 con-
serial EM) with two-photon, transcranial imaging. In tacted synapses in a similar manner in dark-adapted
addition, in contrast to Wake et al. (2009), Tremblay mice 6 light exposure as compared with normal, light
et al. (2010) were better able to distinguish processes exposed mice, the authors provided exciting new evidence
that microglia had more phagocytic inclusions upon dark
belonging to microglia versus neurons by using a trans-
adaptation 1 light exposure, which resembled pre- and
genic mouse line expressing EGFP in microglia and YFP
postsynaptic elements. Consistent with these data, a
in neurons (CX3CR11EGFP/Thy-1 YFP). Furthermore, more recent study by the same group demonstrated an
to assess more physiologically-relevant interactions age-dependent increase in phagocytic inclusions resem-
between microglia and synapses, Tremblay et al. (2010) bling pre- and postsynaptic elements in either V1 or pri-
utilized the critical period in the mouse visual system to mary auditory cortex (A1) in association with age-depend-
assess sensory experience-induced changes in microglia- ent loss of vision or hearing, respectively (Tremblay et al.,
synapse interactions (Gordon and Stryker, 1996; Trem- 2012). Furthermore, another recent study used stimu-
blay et al., 2010). During this particular period in visual lated emission depletion microscopy and postembedding
cortex development [postnatal day 21–30 (P21–30) in EM to assess microglia-mediated phagocytosis of synap-
the mouse], several aspects of visual perception develop ses in the developing hippocampus (Paolicelli et al., 2011),
(e.g., direction selectivity, ocular dominance, etc.) and data that further suggest that microglia may engulf
are associated with changes in spine dynamics, size, and remodeling synaptic elements.

GLIA
 MICROGLIA-SYNAPSE INTERACTIONS 27
The data described above suggest that microglia are Taken together, these provocative studies demonstrate
phagocytosing synaptic elements and that this interac- that microglia, indeed, interact with and engulf synaptic
tion may underlie synaptic remodeling in response to elements and that microglia-synapse interactions are
changes in sensory experience. To more directly address regulated by neural activity and sensory experience.
whether microglia are phagocytosing synapses under- Importantly, these studies raise new and interesting
going activity-dependent synaptic remodeling, we questions including: (1) What are the molecular mecha-
recently assessed phagocytosis in a system in which ac- nisms underlying microglia-synapse interactions? and
tivity-dependent synaptic remodeling has been very well (2) What are the physiological consequences of these
characterized (Schafer et al., 2012): the postnatal retino- interactions.
geniculate system. In this system, retinal ganglion cells
(RGCs) form synapses on relay neurons residing in the
dorsal lateral geniculate nucleus (dLGN) of the thala- MICROGLIA-SYNAPSE INTERACTIONS IN THE
mus, and these synapses are known to undergo activity- DEVELOPING CNS: MOLECULAR MECHANISMS
dependent remodeling during a very small developmen- AND FUNCTIONAL CONSEQUENCES
tal window (see Section ‘‘Microglia-synapse interactions
in the mature CNS: molecular mechanisms and func- The mature nervous system is characterized by a
tional consequences’’ for details regarding retinogenicu- remarkably precise neural circuitry. However, in the
late remodeling) (Guido, 2008; Hong and Chen, 2011; developing nervous system, this circuitry is far less pre-
Huberman, 2007; Sretavan and Shatz, 1986). Using this cise as neurons form exuberant transient synapses during
system, we developed a high throughput in vivo phago- early development . In a process termed synaptic pruning,
cytic assay to demonstrate that microglia phagocytose many synapses are eliminated while the remaining synap-
presynaptic RGC inputs during a peak period in retino- ses are maintained and strengthened. While it is clear
geniculate synaptic remodeling (P5 in mouse), data that that neural activity regulates pruning as well as synapse
were validated by several different light and ultrastruc- maturation (Hua and Smith, 2004; Huberman et al., 2008;
tural assays (Schafer et al., 2012). By EM, this phagocy- Katz and Shatz, 1996; O’Leary and McLaughlin, 2005;
tosed material had characteristic features of presynaptic Sanes and Lichtman, 1999; Torborg and Feller, 2005), lit-
terminal machinery such as 40 nm vesicles; however, tle is known regarding cellular and molecular mediators.
engulfment of axonal arbors cannot be excluded. Unlike Interestingly, several key articles published in the last
previous work, this study observed little evidence sug- decade have identified a critical role for molecules tradi-
gesting engulfment of postsynaptic machinery, which tionally associated with immune function (MHC class I
may be due to regional differences, particularly given molecules and receptors, complement components and
that RGCs synapse on the cell body and proximal den- receptors, and neuronal pentraxins) as modulators of de-
drites of postnatal dLGN relay neurons (Guido, 2008). velopmental synaptic pruning (Bjartmar et al., 2006; Bou-
We further demonstrated that engulfment and synaptic langer, 2009; Corriveau et al., 1998; Datwani et al., 2009;
remodeling were temporally correlated such that as Goddard et al., 2007; Huh et al., 2000; Schafer and Ste-
remodeling was nearly complete, engulfment of RGC vens, 2010; Stevens et al., 2007; Syken et al., 2006). Taken
inputs was also dramatically reduced. together with the emerging evidence that activity regu-
In addition to assessing developmental regulation, we lates microglia at synaptic sites (Schafer et al., 2012;
developed an in vivo assay to test whether microglia Tremblay et al., 2010; Wake et al., 2009), it was hypothe-
synapse-interactions are affected by activity-dependent sized that microglia, the resident CNS immune cells, may
synaptic competition. Activity-dependent competition be cellular mediators of activity-dependent synaptic prun-
was established by injecting one eye with either TTX to ing and maturation. To determine whether microglia-syn-
silence firing or forskolin, a cAMP analog, to increase apse interactions in the developing brain have physiologi-
firing (Dunn et al., 2006; Stellwagen and Shatz, 2002; cal consequences and dissect molecular mechanisms
Stellwagen et al., 1999). The other eye was injected with underlying these interactions, recent studies have utilized
vehicle. After pharmacologically decreasing (TTX) or mice harboring deletions of genes specific to microglia in
increasing (forskolin) firing in one eye, data revealed the context of the healthy developing CNS (Paolicelli
that microglia preferentially engulfed inputs originating et al., 2011; Pascual et al., 2011; Roumier et al., 2004,
from the ‘‘weaker’’ or less active eye (Schafer et al., 2008; Schafer et al., 2012).
2012). Importantly, these particular pharmacological
manipulations are known to disrupt normal synaptic
remodeling such that inputs from the ‘‘weaker’’ or less A Role for Microglia in Developmental
active eye lose territory and inputs from the ‘‘stronger’’ Synaptic Pruning
or more active eye gain territory (Cook et al., 1999; Del
Rio and Feller, 2006; Huberman et al., 2008; Penn et al., One molecular pathway that has been proposed as a
1998; Shatz, 1990; Shatz and Stryker, 1988; Stellwagen potential mediator of microglia-synapse interactions and
and Shatz, 2002). Thus, these data suggest that micro- developmental synaptic pruning is the classical comple-
glia are dynamic sensors during activity-dependent syn- ment cascade (Stephan et al., 2012; Schafer and
aptic remodeling and may be actively engulfing synapses Stevens, 2010; Stevens et al., 2007). Molecules belonging
destined for elimination. to the classical complement cascade, C1q and C3, are

GLIA
28 SCHAFER ET AL.

Fig. 1. Models of microglia-synapse interactions in the CNS. (A) In express high levels of complement receptor 3 (CR3, red squares) on
the embryonic and early postnatal brain, microglias are of an amoeboid their surface. We propose that complement component C3 (red stars) is
morphology resembling ‘‘activated" cells associated with disease and tagging synapses for removal and have demonstrated that synapses are
injury. During this stage, they are actively dividing and recruited to engulfed by phagocytic microglia in a complement-dependent manner.
regions throughout the CNS. Fractalkine (purple circles), which may be Disruptions in any of these processes results in deficits in synaptic
released by neurons, is proposed to act on fractalkine receptors (purple pruning or maturation. (C) In the adult brain, evidence suggests that
squares) expressed by microglia to regulate activation, cell number, microglia are releasing soluble factors (gray and black circles) such as
and/or recruitment to synaptic-enriched regions. DAP12 (orange BDNF, TNFa, glycine, L-serine etc., which affect basal neurotransmis-
squares) expressed on the surface of microglia is also thought to affect sion and synaptic plasticity (i.e., LTP) via direct action on neurons or
synapse function during this period, perhaps, by regulating the ‘‘acti- indirectly via astrocytes (orange). In addition, fractalkine signaling via
vation" state of microglia. (B) In the postnatal brain, during the first 3 soluble fractalkine (purple circles), most likely released by neurons,
weeks of postnatal life, microglia, which are still ‘‘activated" but now and the fractalkine receptor (purple squares), expressed by microglia,
have processes, participate in synaptic pruning. Along with the fractal- modulates microglia-synapse interactions to affect LTP and behavior in
kine receptor (purple squares) and DAP12 (orange squares), microglia the mature CNS.

localized to synaptic compartments and mediate synap- ‘‘tagging’’ synapses for removal by complement-receptor
tic pruning in the developing retinogeniculate system expressing microglia (Fig. 1, panel B). However, it
(Stevens et al., 2007). In the innate immune system, remains to be determined in vivo whether complement
C1q and/or C3 bind cellular material, inducing its re- components are regulated by activity and localized to
moval by several different mechanisms including phago- synapses destined for elimination by microglia-mediated
cytic pathways (Gasque, 2004; Lambris and Tsokos, engulfment. However, because engulfment of RGC
1986; van Lookeren Campagne et al., 2007). Thus, an in- inputs was reduced by 50% in C3 and CR3 KO mice,
triguing hypothesis has been proposed that complement data suggest that other phagocytic pathways must be
proteins bind synapses, which are subsequently elimi- involved (Schafer et al., 2012). Interesting candidates
nated by the primary CNS phagocyte, microglia. may be proteins belonging to the ‘‘find-me’’, ‘‘eat-me,’’
Recently, we demonstrated in vivo that complement and ‘‘don’t-eat-me’’ signaling pathways traditionally
component C3, enriched in synaptic compartments, and associated with engulfment of apoptotic cells, for review
its receptor, complement receptor 3 (CR3), expressed on see (Elward and Gasque, 2003; Griffiths et al., 2009;
the surface of microglia, mediated engulfment of presyn- Grimsley and Ravichandran, 2003; Nagata et al., 2010;
aptic terminals in the developing retinogeniculate sys- Ravichandran, 2011). For example, the Drosophila
tem during a peak period in synaptic remodeling (P5 Draper (CED-1 in C. elegans and MEGF10 in mammals)
mouse dLGN) (Schafer et al., 2012). Similarly, an in and CED-6 (GULP in mammals) pathways have been
vitro study suggested that sialic acid localized to neu- demonstrated to be involved in the removal of axonal de-
rites regulates binding of complement components (C1q bris during developmental axon pruning and following
and C3) and CR3-dependent engulfment of neurites by axonal injury (Awasaki et al., 2006; Fuentes-Medel
microglia. (Linnartz et al., 2012). These data are consist- et al., 2009; Logan et al., 2012; MacDonald et al., 2006;
ent with the idea that complement proteins may be Ziegenfuss et al., 2008). It remains to be determined

GLIA
 MICROGLIA-SYNAPSE INTERACTIONS 29
whether these pathways may also be involved in the tex and hippocampus has suggested a role for microglia
mammalian system. in remodeling and/or maturation of synaptic circuits
Given that C3 and CR3 KO mice had a deficit in (Paolicelli et al., 2011; Roumier et al., 2004, 2008; Trem-
engulfment of remodeling RGC inputs, we directly blay et al., 2010). For example, Paolicelli et al. (2011)
addressed the role of microglia in developmental synap- demonstrated a role for the fractalkine receptor
tic pruning by assessing C3 and CR3 KO mice for prun- (CX3CR1), expressed specifically on the surface of micro-
ing defects in the developing retinogeniculate system. glia in the developing CNS (Harrison et al., 1998), in
Early in development, RGCs in the retina project to and hippocampal synapse development. In the context of dis-
form exuberant, transient synaptic connections within ease, CX3CR1 has the capacity to modulate microglia
the early dLGN of the thalamus (Guido, 2008; Hong and number, activation, and recruitment to sites of injury by
Chen, 2011; Huberman, 2007; Sretavan and Shatz, binding its ligand fractalkine (CX3CL1), expressed by
1986). Within the first-week of rodent postnatal develop- injured neurons (Cardona et al., 2006; Jung et al.,
ment, RGC synaptic inputs compete for territory, result- 2000). In the healthy, developing hippocampus,
ing in the elimination of many transient synaptic con- Cx3cr1KO mice exhibited an increase in spine density
nections and the maintenance and strengthening of and PSD-95 immunoreactivity, enhanced hippocampal
those synapses that remain. This competition can occur long-term depression, and decreased duration of and la-
between synapses originating from the same eye (monoc- tency to pentylenetetrazol (PTZ)-induced seizure response,
ular) as well as between synapses originating from dif- characteristics associated with less mature synapses and
ferent eyes (binocular) (Chen and Regehr, 2000; Hooks possibly associated with abnormal pruning (Paolicelli
and Chen, 2006; Jaubert-Miazza et al., 2005; Ziburkus et al., 2011). However, postnatal (P15) Cx3cr1KO mice had
and Guido, 2006). One gross assessment for deficits in no deficits in engulfment of postsynaptic elements. In con-
retinogeniculate pruning is eye specific segregation, in trast, the authors demonstrated that fewer microglia were
which inputs from both eyes compete for territory, ulti- present in the postnatal Cx3cr1KO hippocampus as com-
mately resulting in the termination of ipsilateral and con- pared with age-matched wild-type (WT) controls. Impor-
tralateral synaptic inputs in distinct nonoverlapping tantly, abnormalities in microglia number and synapse
domains in the mature dLGN (Godement et al., 1984; density in postnatal Cx3cr1KO mice returned to normal lev-
Guido, 2008; Huberman et al., 2008; Jaubert-Miazza els by adulthood. Thus, in the context of the developing
et al., 2005; Sretavan and Shatz, 1986; Ziburkus and CNS, it is most likely that fractalkine signaling regulates
Guido, 2006). Using this system, pharmacological (mino- microglia activation, number, and/or recruitment to hippo-
cycline) or more specific genetic (C3 or CR3 KO) disrup- campal synaptic sites in the early postnatal brain (Fig. 1,
tions in microglia function resulted in deficits in eye spe-
panels A and B). It remains to be determined if microglia
cific segregation as well as an increased density of struc-
play any sustained functional role at these synapses. In
turally intact synapses (Schafer et al., 2012; Stevens
contrast to Paolicelli et al. (2011), another recent study
et al., 2007). Importantly, these effects were sustained
demonstrated different defects in microglia and synaptic
into adulthood and, in the case of the CR3 KO, could be
function in the mature (three-month-old) hippocampus of
specifically attributed to microglia in the context of the
healthy developing brain. One question that remains Cx3cr1KO mice (i.e., increased microglia numbers and
unanswered is whether microglia-dependent pruning reduced LTP; see Section ‘‘Microglia-synapse interactions
effects represent an active or passive process. That is, in the mature CNS: molecular mechanisms and functional
while C3/CR3-mediated engulfment and pruning data are consequences" for more details) raising the possibility that
temporally correlated, it remains to be determined fractalkine signaling may play different roles in the devel-
whether microglia mediate pruning by actively engulfing oping and mature brain (Fig. 1, panels A–C) (Rogers et al.,
intact terminals destined for elimination via C3/CR3 sig- 2011).
naling or whether C3/CR3-mediated engulfment repre- Consistent with microglia playing a role in synapse
sents a ‘‘clean-up" process independent of C3/CR3-de- maturation in the hippocampus, earlier in vitro studies
pendent pruning. In addition, future work should aim to demonstrated that maturation of hippocampal synapses
assess whether these effects are specific to the retinogeni- was altered in mice harboring a mutation in KARAP/
culate system or have broader implications throughout DAP12 (DAP12KI), a transmembrane receptor expressed
the CNS and whether other immune pathways previously by microglia from birth and a known regulator of macro-
identified to play a role in retinogeniculate pruning and phage activation in the immune system (Hamerman
plasticity throughout the CNS (e.g., MHC class I mole- et al., 2005; Roumier et al., 2004, 2008; Tomasello et al.,
cules) may interact with complement and/or microglia to 1998, 2000; Turnbull et al., 2005). Acute hippocampal
mediate synaptic pruning (Corriveau et al., 1998; Dat- slices prepared from P22 DAP12KI mutant mice had
wani et al., 2009; Goddard et al., 2007; Huh et al., 2000). increased NR2B-containing N-Methyl-D-aspartate
(NMDA) receptors as assessed by ifenprodil sensitivity
and increased 2-amino-3-(5-methyl-3-oxo-1,2-oxazol-4-yl)
A Role for Microglia in Synapse Maturation in the propanoic acid (AMPA) receptor calcium permeability,
Developing CNS characteristics indicative of less mature synapses
(Roumier et al., 2004). In a later study, acute hippocam-
Consistent with microglia having a broader role in pal slices prepared from P18–P25 DAP12KI mice had
synapse development, recent work in the developing cor- increased AMPA/NMDA ratios and, while most synapses

GLIA
30 SCHAFER ET AL.

appeared normal by EM, there was an observed increase Microglia-Dependent Synaptic Plasticity in the
in perforated synapses in the stratum radiatum of CA1 Mature CNS
(Roumier et al., 2008). To address whether these effects
originated prenatally the authors prepared near-pure While studies in the developing brain have demon-
dissociated hippocampal neuron cultures from P0 strated that physical interactions between microglia and
DAP12KI mice and mice subjected to in utero inflamma- synaptic elements can regulate plasticity, most of the
tion with LPS. After 14 days in culture (DIV 14), func- work in the mature CNS has suggested that microglia
tional calcium imaging revealed significantly reduced can function to modulate the plasticity of neural circuits
calcium fluctuations in the presence of the AMPAR by paracrine signaling. In vitro studies have demon-
blocker 6-cyano-7 nitroquinoxline-2 (CNQX) in cultures strated that when either cultured microglia or condi-
prepared from P0 mutant and LPS-treated mice as com- tioned media from cultured microglia was added to corti-
pared with WT or vehicle-treated mice. In addition, cal slice cultures, NMDA-receptor mediated excitatory
there was an increase in immunohistochemical colocali- postsynaptic currents (EPSCs) were larger in amplitude
zation of pre- and postsynaptic markers in cultures pre- and longer in duration (Moriguchi et al., 2003). Work in
pared from DAP12KI mice. Because DAP12 expression is hippocampal neuron cultures confirmed this finding and
specific to prenatal and early postnatal microglia and demonstrated that microglia conditioned media also
cultures were presumably void of microglia, these data enhanced LTP induction (Hayashi et al., 2006). The
suggest that the physiological and structural deficits are same study identified glycine and L-serine secreted by
the result of a developmental defect in synapse matura- microglia as molecular mediators of this effect. Consist-
tion that originates prenatally (Fig. 1, panel A) ent with a role for microglia in hippocampal LTP, a
(Roumier et al., 2008). However, because the oldest age recent study found that LTP induction was reduced in
at which synapse function was assessed was P25, it organotypic hippocampal slices prepared from adult (3
remains to be determined whether deficits in DAP12KI month) Cx3cr1KO mice as compared with WT littermates
hippocampal synapse maturation represent a develop- (Rogers et al., 2011). Importantly, these in vitro deficits
mental delay and are later recovered. were coincident with in vivo impairments in learning
These new insights offer significant advancements in and memory as assessed by Morris water maze and con-
our understanding of molecular mechanisms and func- textual and cued fear conditioning. Because these mice
tional consequences of microglia-synapse interactions in also had impaired adult neurogenesis in vivo, it remains
the developing brain and have provoked interest in iden- to be determined whether CX3CR1 signaling in micro-
tifying more microglia-related molecular pathways and glia has a direct or indirect effect on synapse function.
functions throughout the CNS. Indeed, identification of Interestingly, unlike postnatal hippocampus, which was
such mechanisms and functions will have tremendous characterized by a transient decrease in microglia den-
impact on the way we think about synapse development sity (Paolicelli et al., 2011), microglia density was
and plasticity and may also greatly increase our under- increased in adult Cx3cr1KO hippocampus (Rogers et al.,
standing of diseases associated with abnormal brain wir- 2011). These data raise the intriguing possibility that
ing (e.g., autism, schizophrenia, etc.) and/or synaptic fractalkine signaling could have differential effects on
degeneration (e.g., Alzheimer’s disease, Parkinson’s, microglia depending on the context (Fig. 1, panels A–C).
etc.) (See Section ‘‘Microglia-synapse interactions in the In addition to synaptic plasticity associated with LTP,
healthy CNS: disease relevance" for discussion of disease microglia also have been suggested to play a role in syn-
relevance). aptic scaling, a homeostatic mechanism that promotes
long-term stability of neural networks (Stellwagen and
Malenka, 2006; Turrigiano, 2008; Turrigiano and Nel-
son, 2004). Using cultured hippocampal neurons, glial-
MICROGLIA-SYNAPSE INTERACTIONS IN THE derived TNFa was shown to be necessary for synaptic
MATURE CNS: MOLECULAR MECHANISMS AND scaling in excitatory and inhibitory neurons. Because
FUNCTIONAL CONSEQUENCES both astrocytes and microglia produce TNFa, the rela-
tive contribution of these two cell types to in vitro and
The previously discussed studies suggest that interac- in vivo synaptic scaling is an interesting and important
tions between microglia and synapses play an important area for future investigation.
role in the pruning and/or maturation of synaptic con-
nectivity in the developing CNS. Could microglia play a
role in the plasticity and function of synaptic circuits in Microglia-Mediated Effects on Basal Synaptic
the mature CNS? In addition to roles at developing syn- Transmission
apses, several studies have now described potential
functions for microglia at synapses in the healthy, adult Besides synaptic plasticity, it has been suggested that
CNS, such as regulating LTP, synaptic scaling, and ba- microglia could regulate basal glutamatergic and
sal glutamatergic and GABAergic transmission (Fig. 1, GABAergic synaptic transmission (Coull et al., 2005;
panel C) (Ben Achour and Pascual, 2010; Bessis et al., Pascual et al., 2011; Tsuda et al., 2003). For example,
2007; Kettenmann et al., 2011). The following section when acute organotypic hippocampal slices were exposed
will review several key findings. to the proinflammatory molecule LPS, microglia became

GLIA
 MICROGLIA-SYNAPSE INTERACTIONS 31
more ‘‘activated" and AMPA receptor-mediated sponta- their role in the reversal of GABA currents resulting in
neous EPSC frequency was increased in CA1 neurons pain hypersensitivity (discussed in previous section). In
(Pascual et al., 2011). Pascual et al. (2011) then demon- neurodegenerative diseases of the CNS (e.g., Multiple
strated that this effect was attenuated in slices prepared Sclerosis, Alzheimer’s disease, Huntington’s disease,
from PU.1 null mice that lack several cells types belong- glaucoma, etc.), abnormal synapse function and loss are
ing to the lymphoid and myeloid lineages (McKercher increasingly being recognized as hallmarks of early
et al., 1996; Scott et al., 1994), data suggesting that the stages of disease progression (Stephan et al., 2012;
effect was specific to microglia in the context of the Coleman et al., 2004; Mandolesi et al., 2010; Milnerwood
CNS. The authors further demonstrated that the micro- and Raymond, 2010). While there are instances in which
glia-specific effect was, most likely, indirect and sug- synapse loss associated with neurodegeneration (i.e.,
gested that LPS induces microglia to release ATP that Prion disease) is a neuron-autonomous event (Perry and
binds P2Y1 receptors on astrocytes. Astrocytes subse- O’Connor, 2010; Siskova et al., 2009), data in other con-
quently mediate an increase in excitatory transmission texts suggest that microglia are contributing to early
via a metabotropic glutamate receptor 5 (mGluR5)-de- synapse loss and/or dysfunction. (Alexander et al., 2008,
pendent mechanism. While this study was performed in 2012; Beggs and Salter, 2010; Rosen and Stevens, 2010;
vitro and may be more disease relevant as LPS is a Schafer and Stevens, 2010). For example, in a mouse
proinflammatory agent thought to more closely mimic model of tauopathy (P301S), in vivo hippocampal syn-
disease states, it raises the intriguing question as to apse loss and microglial activation were coincident and
whether microglia have the capacity to regulate basal occurred as early as 3 months of age, while significant
glutamatergic transmission under less pathological con- atrophy is not observed until 9–12 months of age (Yosh-
ditions in vivo. iyama et al., 2007). Thus, data place microglia at the
In addition to affecting basal glutamatergic signaling, right time and place to contribute to early synapse loss
microglia have the capacity to modulate GABAergic syn- and/or dysfunction. In addition, microglia are thought to
aptic transmission; however, this effect has only been be involved in synaptic stripping following axotomy
demonstrated in the context of injury (Coull et al., 2005; (Cullheim and Thams, 2007; Kreutzberg, 1993; Perry
Tsuda et al., 2003). Briefly, upon injury, ATP is released, and O’Connor, 2010; Trapp et al., 2007). These experi-
which in turn, stimulates microglia to release brain- ments were first performed in the context of the injured
derived neurotrophic factor (BDNF). Microglia-derived facial nerve where microglial processes were shown to
BDNF induces a depolarizing shift in the anion reversal intercalate between pre- and postsynaptic elements
potential that results in an inversion of current polarity within the facial nerve nucleus following injury, which
activated by GABA. Excitatory transmission via GABA effectively ‘‘stripped’’ away presynaptic terminals from
receptor activation results in hyperactivity and, on a be- their postsynaptic targets (Blinzinger and Kreutzberg,
havioral level, increased allodynia. It remains to be 1968). Interestingly, a recent study in the aging brain
determined whether similar microglia-mediated mecha- demonstrated an increase in phagocytic inclusions
nisms may be involved in modulating GABAergic neuro- within microglia resembling pre- and postsynaptic ele-
transmission in the healthy CNS. ments that was concommittant with age-dependent loss
of vision and hearing independent of any significant
neuronal cell loss (Tremblay et al., 2012). Future work
MICROGLIA-SYNAPSE INTERACTIONS IN THE will be necessary to better define the role of microglia at
HEALTHY CNS: DISEASE RELEVANCE dysfunctional synapses associated with neurodegenera-
tive disease and aging and to determine whether these
The studies discussed in previous sections demon- cells, indeed, contribute to early synapse loss.
strate roles for microglia in the remodeling and matura- In addition to synaptic dysfunction and loss associated
tion of synaptic circuits in the developing CNS as well with neurodegenerative disorders and acute neuronal
as in contributing to basal transmission and plasticity in injury, particularly timely and relevant is the potential
the adult. In addition to elucidating mechanisms of syn- contribution of microglia to synaptic abnormalities asso-
apse development and function in the healthy brain, ciated with psychiatric disorders such as autism spec-
these studies have important implications for under- trum disorder (ASD), obsessive compulsive disorder
standing mechanisms underlying synapse pathology in (OCD), and schizophrenia (Chen et al., 2010; Hashimoto,
disease. In the following section, we will briefly touch on 2008; Havik et al., 2011; Monji et al., 2009a; Morgan
some key studies suggesting a function for microglia at et al., 2010; Pardo et al., 2005; Vargas et al., 2005). Sev-
diseased synapses. eral groups have published interesting studies demon-
Microglial pathological functions have been tradition- strating that early infection, a risk factor for many psy-
ally associated with diseases in which they are known to chiatric disorders, can result in abnormalities in synap-
perform several functions, ranging from removal of de- ses, microglia, and/or behavior, particularly after a
bris and shielding injury sites to initiating and propa- second immune or stress challenge (Bilbo, 2010; Bilbo
gating immune responses (Hanisch and Kettenmann, et al., 2006; Bitanihirwe et al., 2010; Ito et al., 2010; Shi
2007; Kreutzberg, 1996; Prinz et al., 2011; Ransohoff et al., 2003). One of the most relevant and exciting new
and Perry, 2009). In the context of models of neuropathic studies that implicates microglia in the abnormal brain
pain, one of the most intriguing synaptic functions is wiring associated with ASD is a very recent study by

GLIA
32 SCHAFER ET AL.

TABLE 1. Overview of Synapse-Related Microglia Functions

Molecules mediation
Microglia function Age of analysis function Method of study Region studied References
Activity-dependent First postnatal C3, CR3, C1q in vivo: Engulfment Retinogeniculate Schafer et al. 2012
synaptic pruning week assay, EM, system
anatomical tracing,
array tomography,
IHC
Phagocytic removal of Cultures prepared C3, CR3, C1q, Sialic in vitro: Sialidase Hippocampus Linnartz et al. 2012
neurites from embryonic acids treatment IHC, co-
mice culture of dissociated
neurons and
microglia
Regulation of synapse Second and third CX3CL1, CX3CXR1 in vivo: IHC, EM, and Hippocampus Paolicelli et al. 2011
maturation postnatal weeks chemical seizure
induction
ex vivo: Physiology in
acute slice
Regulation of synapse Third postnatal KARAP/DAP12 in vivo: EM ex vivo: Hippocampus Roumier et al. 2004,
maturation week Physiology in acute Roumier et al. 2008
slice
in vitro: Physiology in
dissociated neuron
cultures, calcium
imaging, IHC
Activity-dependent Adulthood Unknown in vivo: 2-photon Somatosensory and Wake et al. 2009
structural remodeling transcranial imaging, visual cortex
EM, calcium
imaging,
pharmacology,
enucleation,
ischemic injury
Sensory experience Third and fourth Unknown in vivo: 2-photon Visual cortex Tremblay et al. 2010
driven structural postnatal weeks transcranial imaging,
remodeling serial 3D EM, IHC,
alternation of sensory
experience
Age-dependent Adulthood Unknown in vivo: EM, behavioral Visual and auditory Tremblay et al. 2012
structural remodeling studies, IHC cortex
Extracellular Adulthood Unknown in vivo: 2-photon Cerebral cortex Davalos et al. 2005,
surveillance transcranial imaging, Nimmerjahn et al.
pharmacology, 2005
targeted injury
Enhancement of Adulthood Glycine, L-serine in vitro: Physiology in Cortex, Hippocampus Morguchi et al. 2003,
NMDA receptor organotypic slice and Hayashi et al. 2006
mediated responses, dissociated neuron
LTP cultures
Regulation of LTP, Adulthood CX3CL1, CX3CR1 in vivo: Behavioral Hippocampus Rogers et al. 2011
learning and memory assays, pharmocology
ex vivo: Physiology in
acute slice
Synaptic scaling Adulthood TNFa ex vivo: Physiology in Hippocampus Stewllwagen &
acute slice Malenka 2006
in vitro: Physiology in
dissociated neuron
cultures
Modulation of Adulthood BDNF, ATP in vivo: Pharmacology, Spinal cord Tsuda et al. 2003, Coull
GABAergic behavioral assays for et al. 2005
transmission allodynia
ex vivo: Physiology in
acute slice
Modulation of basal Adulthood ATP ex vivo: Physiology in Hippocampus Pascual et al. 2011
glutamatergic acute slice,
signaling pharmacology
in vitro: Physiology in
organotypic slice and
dissociated neuron
cultures
Unknown Third postnatl week ATP, Glutamate, ex vivo: Time lapse Retina Fontainhas et al. 2011
to adulthood GABA imaging in explants,
pharmocology,
physiology
Unknown Cultures prepared ATP, Glutamate, in vitro: Live imaging Hippocampus Grinberg et al. 2011
from postnated cAMP in slice cultures,
mice pharmacology,
physiology
Summary of studies discussed in this review. IHC, Immunohistochemistry; EM, Electron Microscopy; CR3, Complement receptor 3; CX3CL1-Fractalkine, CX3CR1-Frac-
talkine receptor; TNFa, Tumor necrosis factor-alpha; BDNF, Brain derived neurotrophic factor; ATP, Adenosine triphosphate; GABA, Gamma-aminobutyric acid; cAMP,
Cyclic adenosine monophosphate.

GLIA
 MICROGLIA-SYNAPSE INTERACTIONS 33
Derecki et al. (2012). Using a genetic model of Rett syn- active process by which microglia selectively engulf
drome, the Mecp2-null, that has behavioral and synaptic intact synapses destined for elimination. Furthermore,
phenotypes resembling those associated with ASD (Chen while current imaging data in the cortex and hippocam-
et al., 2001; Guy et al., 2001), Derecki et al. demon- pus suggest a role for microglia in activity-dependent
strated that replenishing an Mecp2-null with WT micro- synaptic remodeling or maturation (Paolicelli et al.,
glia resulted in the attenuation of several behavioral 2011; Tremblay et al., 2010), it is still unclear whether
and physiological deficits (e.g., body weight, breathing microglia are necessary for bona fide synaptic pruning
rate, locomotion, etc.). Furthermore, the beneficial in these other brain regions, and if so, what are the
effects of WT microglia on Mecp2-null mice were dimin- underlying mechanisms? In the context of the mature
ished when phagocytic activity was blocked pharmaco- adult CNS, while several molecular pathways have been
logically with Annexin-V. However, it remains unclear identified to contribute to synaptic plasticity and basal
precisely how abnormal phagocytic activity may be con- transmission in slice and cultured cell preparations,
tributing to the phenotype and whether and how micro- in vivo contributions are still relatively unclear with
glia may be contributing to synaptic abnormalities asso- only a few behavioral correlates. In addition, in the
ciated with ASD. Interestingly, one in vitro study using mature CNS, it is unknown whether and how microglia-
dissociated hippocampal neuron cultures demonstrated specific pathways may interact with one another to
that cells treated with conditioned media isolated from affect neurotransmission and plasticity and whether
Mecp2-null mice had delayed and abnormal dendritic many of these molecular pathways have similar or dif-
morphology, signs of microtubule disruption, and dam- ferent functions in the context of microglia in the devel-
age to postsynaptic glutamatergic components due to oping brain (Fig. 1, panels A–C).
toxic levels of glutamate released by the microglia (Mae- Imperative to our understanding of these mysterious
zawa and Jin, 2010). Given that deficits in synaptic cir- cells is the use of in vivo strategies. Microglia are highly
cuit development are emerging as important underlying reactive cells that respond within minutes to manipula-
correlates of behavioral outcomes (Belmonte et al., 2004; tion (Davalos et al., 2005; Hanisch and Kettenmann,
LeBlanc and Fagiolini, 2011; Melom and Littleton, 2011; 2007; Kreutzberg, 1996; Nimmerjahn et al., 2005; Ran-
Rubenstein and Merzenich, 2003; Waites and Garner, sohoff and Perry, 2009). Thus, while in vitro prepara-
2011) and the new data that microglia are participating tions (e.g., isolated cells, acute slice, slice culture, etc.)
in synaptic pruning via phagocytic immune pathways are important and necessary strategies for dissecting
(Schafer et al., 2012), it is a highly speculative yet pro- mechanism and function, the physiological relevance
vocative hypothesis that microglia are contributing to and molecular mechanisms identified must be confirmed
ASD symptoms, in part, through aberrant neural- in vivo. Future in vivo studies using a combinatorial
immune signaling at developing synapses. approach of live imaging and molecular biology, includ-
ing the use of newly derived cre lines (Parkhurst et al.,
2011), should significantly advance our understanding of
SUMMARY AND REMAINING QUESTIONS the function of microglia at synapses and the molecular
mechanisms underlying these interactions. Given that
In summary, the recent attention focused on the role several prevalent psychiatric and developmental disor-
of microglia in the healthy brain has elicited several ders (e.g., schizophrenia, OCD, autism, etc.) have
exciting new findings suggesting that microglia play recently been linked to deficits in synapse development
dynamic roles at developing and mature synapses (see and/or function, as well as a growing body of evidence
Table 1). The interactions between microglia and synap- suggesting abnormalities in microglia (Derecki et al.,
ses are dependent upon direct, physical contact as well 2012; Maezawa and Jin, 2010; Monji et al., 2009b; Mor-
as signaling via soluble factors. On a functional level, it gan et al., 2010; Steiner et al., 2008; Vargas et al., 2005;
is now clear that microglia interact with and/or engulf Yang and Lu, 2011), understanding functions and molec-
synaptic elements in a manner dependent upon neural ular pathways underlying microglia-synapse interactions
activity, mediate synaptic pruning in at least one region in the healthy brain becomes imperative. This knowl-
of the developing CNS, regulate synapse maturation, edge of molecules and function is important for advanc-
and modulate plasticity (LTP and synaptic scaling) and ing our understanding of basic biological mechanisms as
basal transmission in the mature CNS. As a result of well as for the promise of developing novel diagnostic
these important first studies, several questions have and therapeutic strategies associated with CNS disease
arisen and remain unanswered in the field. First, in the and injury.
context of the developing CNS, while it is clear that
complement-dependent phagocytic signaling is one path-
way underlying physical interactions between microglia
and remodeling synapses, other, yet to be identified, ACKNOWLEDGMENTS
pathways must also be involved. Second, at this point,
data demonstrate that phagocytosis of synaptic elements Work was supported by grants from the Smith Family
and pruning are temporally correlated. It remains to be Foundation (BS), Dana Foundation (BS), John Merck
determined whether engulfment of synaptic elements is, Scholars Program (BS), NINDS (RO1NS07100801-A1;
indeed, an underlying mechanism of plasticity and an BS), and NRSA (F32NS066698-01;D.P.S.).

GLIA
34 SCHAFER ET AL.

REFERENCES inogeniculate refinement and limit ocular dominance plasticity. Neu-

ron 64:463–470.
Davalos D, Grutzendler J, Yang G, Kim JV, Zuo Y, Jung S, Littman
Alexander JJ, Anderson AJ, Barnum SR, Stevens B, Tenner AJ. 2008. DR, Dustin ML, Gan WB. 2005. ATP mediates rapid microglial
The complement cascade: Yin-Yang in neuroinflammation—neuro- response to local brain injury in vivo. Nat Neurosci 8:752–758.
protection and -degeneration. J Neurochem 107:1169–1187. del Rio-Hortega P. 1932. Microglia. In: Penfield W, editor. Cytology
Awasaki T, Tatsumi R, Takahashi K, Arai K, Nakanishi Y, Ueda R, Ito and cellular pathology of the nervous system. New York: Hoeber.
K. 2006. Essential role of the apoptotic cell engulfment genes draper Del Rio T, Feller MB. 2006. Early retinal activity and visual circuit de-
and ced-6 in programmed axon pruning during Drosophila metamor- velopment. Neuron 52:221–222.
phosis. Neuron 50:855–867. Derecki NC, Cronk JC, Lu Z, Xu E, Abbott SB, Guyenet PG, Kipnis J.
Awasaki T, Ito, K. 2004. Engulfing action of glial cells is required for 2012. Wild-type microglia arrest pathology in a mouse model of Rett
programmed axon pruning during Drosophila metamorphosis. Curr syndrome. Nature 484:105–109.
Biol 14:668–677. Dunn TA, Wang CT, Colicos MA, Zaccolo M, DiPilato LM, Zhang J,
Belmonte MK, Allen G, Beckel-Mitchener A, Boulanger LM, Carper Tsien RY, Feller MB. 2006. Imaging of cAMP levels and protein ki-
RA, Webb SJ. 2004. Autism and abnormal development of brain con- nase A activity reveals that retinal waves drive oscillations in sec-
nectivity. J Neurosci 24:9228–9231. ond-messenger cascades. J Neurosci 26:12807–12815.
Ben Achour S, Pascual O. 2010. Glia: The many ways to modulate syn- Elward K, Gasque P. 2003. ‘‘Eat me’’ and ‘‘don’t eat me’’ signals govern
aptic plasticity. Neurochem Int 57:440–445. the innate immune response and tissue repair in the CNS: Emphasis
Bence M, Levelt CN. 2005. Structural plasticity in the developing vis- on the critical role of the complement system. Mol Immunol 40:85–94.
ual system. Prog Brain Res 147:125–139. Farber K, Kettenmann H. 2006. Purinergic signaling and microglia.
Berbel P, Innocenti GM. 1988. The development of the corpus callosum Pflugers Arch 452:615–621.
in cats: A light- and electron-microscopic study. J Comp Neurol Feng G, Mellor RH, Bernstein M, Keller-Peck C, Nguyen QT, Wallace
276:132–156. M, Nerbonne JM, Lichtman JW, Sanes JR. 2000. Imaging neuronal
Bessis A, Bechade C, Bernard D, Roumier A. 2007. Microglial control of subsets in transgenic mice expressing multiple spectral variants of
neuronal death and synaptic properties. Glia 55:233–238. GFP. Neuron 28:41–51.
Biber K, Neumann H, Inoue K, Boddeke HW. 2007. Neuronal ÔOnÕ and Fontainhas AM, Wang M, Liang KJ, Chen S, Mettu P, Damani M, Far-
ÔOffÕ signals control microglia. Trends Neurosci 30:596–602. iss RN, Li W, Wong WT. 2011. Microglial morphology and dynamic
Bilbo SD. 2010. Early-life infection is a vulnerability factor for aging- behavior is regulated by ionotropic glutamatergic and GABAergic
related glial alterations and cognitive decline. Neurobiol Learn Mem neurotransmission. PLoS One 6:e15973.
94:57–64. Freeman MR. 2006. Sculpting the nervous system: glial control of neu-
Bilbo SD, Rudy JW, Watkins LR, Maier SF. 2006. A behavioural charac- ronal development. Curr Opin Neurobiol 16:119–125.
terization of neonatal infection-facilitated memory impairment in Fuentes-Medel Y, Logan MA, Ashley J, Ataman B, Budnik V, Freeman
adult rats. Behav Brain Res 169:39–47. MR. 2009. Glia and muscle sculpt neuromuscular arbors by engulfing
Bishop DL, Misgeld T, Walsh MK, Gan WB, Lichtman JW. 2004. Axon destabilized synaptic boutons and shed presynaptic debris. PLoS Biol
branch removal at developing synapses by axosome shedding. Neuron 7:e1000184.
44:651–661. Gasque P. 2004. Complement: A unique innate immune sensor for dan-
Bitanihirwe BK, Peleg-Raibstein D, Mouttet F, Feldon J, Meyer U. ger signals. Mol Immunol 41:1089–1098.
2010. Late prenatal immune activation in mice leads to behavioral Goddard CA, Butts DA, Shatz CJ. 2007. Regulation of CNS synapses
and neurochemical abnormalities relevant to the negative symptoms by neuronal MHC class I. Proc Natl Acad Sci U S A 104:6828–6833.
of schizophrenia. Neuropsychopharmacology 35:2462–2478. Godement P, Salaun J, Imbert M. 1984. Prenatal and postnatal devel-
Bjartmar L, Huberman AD, Ullian EM, Renteria RC, Liu X, Xu W, Pre- opment of retinogeniculate and retinocollicular projections in the
zioso J, Susman MW, Stellwagen D, Stokes CC, Cho R, Worley P, mouse. J Comp Neurol 230:552–575.
Malenka RC, Ball S, Peachey NS, Copenhagen D, Chapman B, Naka- Gordon JA, Stryker MP. 1996. Experience-dependent plasticity of binoc-
moto M, Barres BA, Perin MS. et al. 2006. Neuronal pentraxins ular responses in the primary visual cortex of the mouse. J Neurosci
mediate synaptic refinement in the developing visual system. J Neu- 16:3274–3286.
rosci 26:6269–6281. Griffiths MR, Gasque P, Neal JW. 2009. The multiple roles of the innate
Blinzinger K, Kreutzberg GW. 1968. Displacement of synaptic terminals immune system in the regulation of apoptosis and inflammation in
from regenerating motoneurons by microglial cells. Z Zellforsch Mik- the brain. J Neuropathol Exp Neurol 68:217–226.
rosk Anat 85:145–157. Grimsley C, Ravichandran KS. 2003. Cues for apoptotic cell engulf-
Boulanger LM. 2009. Immune proteins in brain development and syn- ment: eat-me, don’t eat-me and come-get-me signals. Trends Cell Biol
aptic plasticity. Neuron 64:93–109. 13:648–656.
Cardona AE, Pioro EP, Sasse ME, Kostenko V, Cardona SM, Dijkstra Grinberg YY, Milton JG, Kraig RP. 2011. Spreading depression sends
IM, Huang D, Kidd G, Dombrowski S, Dutta R, Lee JC, Cook DN, microglia on Levy flights. PLoS One 6:e19294.
Jung S, Lira SA, Littman DR, Ransohoff RM. et al. 2006. Control of Guido W. 2008. Refinement of the retinogeniculate pathway. J Physiol
microglial neurotoxicity by the fractalkine receptor. Nat Neurosci 586:4357–4362.
9:917–924. Guy J, Hendrich B, Holmes M, Martin JE, Bird A. 2001. A mouse
Chen C, Regehr WG. 2000. Developmental remodeling of the retinoge- Mecp2-null mutation causes neurological symptoms that mimic Rett
niculate synapse. Neuron 28:955–966. syndrome. Nat Genet 27:322–326.
Chen RZ, Akbarian S, Tudor M, Jaenish R. 2001. Deficiency of methyl- Hamerman JA, Tchao NK, Lowell CA, Lanier LL. 2005. Enhanced toll-
CpG binding protein-2 in CNS neurons results in a Rett-like pheno- like receptor responses in the absence of signaling adaptor DAP12.
type in mice. Nat Genet 27:327–331. Nat Immunol 6:579–586.
Chen SK, Tvrdik P, Peden E, Cho S, Wu S, Spangrude G, Capecchi MR. Hanisch UK, Kettenmann H. 2007. Microglia: Active sensor and versa-
2010. Hematopoietic origin of pathological grooming in Hoxb8 mutant tile effector cells in the normal and pathologic brain. Nat Neurosci
mice. Cell 141:775–785. 10:1387–1394.
Coleman P, Federoff H, Kurlan R. 2004. A focus on the synapse for neu- Harrison JK, Jiang Y, Chen S, Xia Y, Maciejewski D, Mcnamara RK,
roprotection in Alzheimer disease and other dementias. Neurology Streiti WJ, Salafranca MN, Adhikari S, Thompson DA, Botti P, Bacon
63:1155–1162. KB, Feng L. et al. 1998. Role for neuronally derived fractalkine in
Cook PM, Prusky G, Ramoa AS. 1999. The role of spontaneous retinal ac- mediating interactions between neurons and CX3CR1-expressing
tivity before eye opening in the maturation of form and function in the microglia. Proc Natl Acad Sci U S A 95:10896–10901.
retinogeniculate pathway of the ferret. Visual Neurosci 16:491–501. Hashimoto K. 2008. Microglial activation in schizophrenia and minocy-
Corriveau RA, Huh GS, Shatz CJ. 1998. Regulation of class I MHC cline treatment. Prog Neuropsychopharmacol Biol Psychiatry
gene expression in the developing and mature CNS by neural activ- 32:1758–1759; author reply 1760.
ity. Neuron 21:505–520. Håvik B, Le Hellard S, Rietschel M, Lybæk H, Djurovic S, Mattheisen
Coull JA, Beggs S, Boudreau D, Boivin D, Tsuda M, Inoue K, Gravel C, M, M€ uhleisen TW, Degenhardt F, Priebe L, Maier W, Breuer R,
Salter MW, De Koninck Y. 2005. BDNF from microglia causes the Schulze TG, Agartz I, Melle I, Hansen T, Bramham CR, N€ othen MM,
shift in neuronal anion gradient underlying neuropathic pain. Nature Stevens B, Werge T, Andreassen OA, Cichon S, Steen VM. 2011. The
438:1017–1021. complement control-related genes CSMD1 and CSMD2 associate to
Cullheim S, Thams S. 2007. The microglial networks of the brain and Schizophrenia. Biol Psychiatry 70:35–42.
their role in neuronal network plasticity after lesion. Brain Res Rev Hayashi Y, Ishibashi H, Hashimoto K, Nakanishi H. 2006. Potentiation
55:89–96. of the NMDA receptor-mediated responses through the activation of
Datwani A, McConnell MJ, Kanold PO, Micheva KD, Busse B, Shamloo the glycine site by microglia secreting soluble factors. Glia 53:
M, Smith SJ, Shatz CJ. 2009. Classical MHCI molecules regulate ret- 660–668.

GLIA
 MICROGLIA-SYNAPSE INTERACTIONS 35
Haynes SE, Hollopeter G, Yang G, Kurpius D, Dailey ME, Gan WB, Ju- Morgan JT, Chana G, Pardo CA, Achim C, Semendeferi K, Buckwalter
lius D. 2006. The P2Y12 receptor regulates microglial activation by J, Courchesne E, Everall IP. 2010. Microglial activation and
extracellular nucleotides. Nat Neurosci 9:1512–1519. increased microglial density observed in the dorsolateral prefrontal
Hirasawa T, Ohsawa K, Imai Y, Ondo Y, Akazawa C, Uchino S, Koh- cortex in autism. Biol Psychiatry 68:368–376.
saka S. 2005. Visualization of microglia in living tissues using Iba1- Moriguchi S, Mizoguchi Y, Tomimatsua Y, Hayashia Y, Kadowaki T,
EGFP transgenic mice. J Neurosci Res 81:357–362. Kagamiishi Y, Katsube N, Yamamoto K, Inoue K, Watanabe S, Nabe-
Hong YK, Chen C. 2011. Wiring and rewiring of the retinogeniculate kura J, Nakanishi H. et al. 2003. Potentiation of NMDA receptor-medi-
synapse. Curr Opin Neurobiol 21:228–237. ated synaptic responses by microglia. Mol Brain Res 119:160–169.
Hooks BM, Chen C. 2006. Distinct roles for spontaneous and visual activ- Mower GD, Christen WG, Caplan CJ. 1983. Very brief visual experience
ity in remodeling of the retinogeniculate synapse. Neuron 52:281–291. eliminates plasticity in the cat visual cortex. Science 221:178–180.
Hooks BM, Chen C. 2007. Critical periods in the visual system: Changing Nagata S, Hanayama R, Kawane K. 2010. Autoimmunity and the clear-
views for a model of experience-dependent plasticity. Neuron 56:312–326. ance of dead cells. Cell 140:619–630.
Hua JY, Smith SJ. 2004. Neural activity and the dynamics of central Nimmerjahn A, Kirchhoff F, Helmchen F. 2005. Resting microglial cells
nervous system development. Nat Neurosci 7:327–332. are highly dynamic surveillants of brain parenchyma in vivo. Science
Huberman AD. 2007. Mechanisms of eye-specific visual circuit develop- 308:1314–1318.
ment. Curr Opin Neurobiol 17:73–80. O’Leary DD, McLaughlin T. 2005. Mechanisms of retinotopic map de-
Huberman AD, Feller MB, Chapman B. 2008. Mechanisms underlying velopment: Ephs, ephrins, and spontaneous correlated retinal activ-
development of visual maps and receptive fields. Annu Rev Neurosci ity. 147:43–65.
31:479–509. Orr AG, Orr AL, Li XJ, Gross RE, Traynelis SF. 2009. Adenosine A(2A) re-
Huh GS, Boulanger LM, Du H, Riquelme PA, Brotz TM, Shatz CJ. ceptor mediates microglial process retraction. Nat Neurosci 12:872–878.
2000. Functional requirement for class I MHC in CNS development Paolicelli RC, Bolasco G, Pagani F, Maggi L, Scianni M, Panzanelli P,
and plasticity. Science 290:2155–2159. Giustetto M, Ferreira TA, Guiducci E, Dumas L, Ragozzino D, Gross
Inoue K, Koizumi S, Tsuda M. 2007. The role of nucleotides in the neu- CT. et al. 2011. Synaptic pruning by microglia is necessary for nor-
ron—Glia communication responsible for the brain functions. J Neu- mal brain development. Science 333:1456–1458.
rochem 102:1447–1458. Pardo CA, Vargas DL, Zimmerman AW. 2005. Immunity, neuroglia and
Ito HT, Smith SE, Hsiao E, Patterson PH. 2010. Maternal immune acti- neuroinflammation in autism. Int Rev Psychiatry 17:485–495.
vation alters nonspatial information processing in the hippocampus Parkhurst CN, Littman DN, Gan W. 2011. Generation of characteriza-
of the adult offspring. Brain Behav Immun 24:930–941. tion of a CX3CR1-CreER for the study of microglia function in the
Jaubert-Miazza L, Green E, Lo FS, Bui K, Mills J, Guido W. 2005. CNS. Soc Neurosci 2011 Abstract 664.16/I10.
Structural and functional composition of the developing retinogenicu- Pascual O, Ben Achour S, Rostaing P, Triller A, Bessis A. 2011. Micro-
late pathway in the mouse. Vis Neurosci 22:661–676. glia activation triggers astrocyte-mediated modulation of excitatory
Jung S, Aliberti J, Graemmel P, Sunshine MJ, Kreutzberg GW, Sher A, neurotransmission. Proc Natl Acad Sci U S A 109:E197–E205.
Littman DR. 2000. Analysis of fractalkine receptor CX(3)CR1 func- Penn A, Riquelme P, Feller MB, Shatz C. 1998. Competition in retino-
tion by targeted deletion and green fluorescent protein reporter gene geniculate patterning driven by spontaneous activity. Science
insertion. Mol Cell Biol 20:4106–4114. 279:2108–2112.
Katz LC, Shatz CJ. 1996. Synaptic activity and the construction of cort- Perry VH, O’Connor V. 2010. The role of microglia in synaptic stripping
ical circuits. Science 274:1133–1138. and synaptic degeneration: a revised perspective. ASN Neuro
Kettenmann H, Hanisch UK, Noda M, Verkhratsky A. 2011. Physiology 2:e00047.
of microglia. Physiol Rev 91:461–553. Philpot BD, Sekhar AK, Shouval HZ, Bear MF. 2001. Visual experience
Kreutzberg GW. 1993. Dynamic changes in motoneurons during regen- and deprivation bidirectionally modify the composition and function
eration. Restor Neurol Neurosci 5:59–60. of NMDA receptors in visual cortex. Neuron 29:157–169.
Kreutzberg GW. 1996. Microglia: A sensor for pathological events in the Pocock JM, Kettenmann H. 2007. Neurotransmitter receptors on micro-
CNS. Trends Neurosci 19:312–318. glia. Trends Neurosci 30:527–535.
Lambris JD, Tsokos GC. 1986. The biology and pathophysiology of com- Prinz M, Priller J, Sisodia SS, Ransohoff RM. 2011. Heterogeneity of
plement receptors. Anticancer Res 6:515–523. CNS myeloid cells and their roles in neurodegeneration. Nat Neuro-
LeBlanc JJ, Fagiolini M. 2011. Autism: A ‘‘critical period’’ disorder? sci 14:1227–1235.
Neural Plast 2011:921680. Ransohoff RM, Perry VH. 2009. Microglial physiology: Unique stimuli,
Linnartz B, Kopatz J, Tenner AJ, Neumann H. 2012. Sialic acid on the specialized responses. Annu Rev Immunol 27:119–145.
neuronal glycocalyx prevents complement c1 binding and complement Ravichandran KS. 2011. Beginnings of a good apoptotic meal: The find-
receptor-3-mediated removal by microglia. J Neurosci 32:946–952. me and eat-me signaling pathways. Immunity 35:445–455.
Logan MA, Hackett R, Doherty J, Sheehan A, Speese SD, Freeman Rogers JT, Morganti JM, Bachstetter AD, Hudson CE, Peters MM,
MR. 2012. Negative regulation of glial engulfment activity by Draper Grimmig BA, Weeber EJ, Bickford PC, Gemma C. 2011. CX3CR1
terminates glial responses to axon injury. Nat Neurosci 15:722–730. deficiency leads to impairment of hippocampal cognitive function and
MacDonald JM, Beach MG, Porpiglia E, Sheehan AE, Watts RJ, Free- synaptic plasticity. J Neurosci 31:16241–16250.
man MR. 2006. The Drosophila cell corpse engulfment receptor Draper Rosen AM, Stevens B. 2010. The role of the classical complement cas-
mediates glial clearance of severed axons. Neuron 50:869–881. cade in synapse loss during development and glaucoma. Adv Exp
Maezawa I, Jin LW. 2010. Rett syndrome microglia damage dendrites Med Biol 703:75–93.
and synapses by the elevated release of glutamate. J Neurosci Roumier A, Bechade C, Poncer JC, Smalla KH, Tomasello E, Vivier E,
30:5346–5356. Gundelfinger ED, Triller A, Bessis A. 2004. Impaired synaptic func-
Majewska A, Sur M. 2003. Motility of dendritic spines in visual cortex tion in the microglial KARAP/DAP12-deficient mouse. J Neurosci
in vivo: Changes during the critical period and effects of visual depri- 24:11421–11428.
vation. Proc Natl Acad Sci U S A 100:16024–16029. Roumier A, Pascual O, Bechade C, Wakselman S, Poncer JC, Real E,
Mallat M, Marin-Teva JL, Cheret C. 2005. Phagocytosis in the develop- Triller A, Bessis A. 2008. Prenatal activation of microglia induces
ing CNS: More than clearing the corpses. Curr Opin Neurobiol delayed impairment of glutamatergic synaptic function. PLoS One
15:101–107. 3:e2595.
Mandolesi G, Grasselli G, Musumeci G, Centonze D. 2010. Cognitive Rubenstein JL, Merzenich MM. 2003. Model of autism: increased ratio
deficits in experimental autoimmune encephalomyelitis: neuroinflam- of excitation/inhibition in key neural systems. Genes Brain Behav
mation and synaptic degeneration. Neurol Sci 31:S255–S259. 2:255–267.
McKercher SR, Torbett BE, Anderson KL, Henkel GW, Vestal DJ, Bari- Sanes JR, Lichtman JW. 1999. Development of the vertebrate neuro-
bault H, Klemsz M, Feeney AJ, Wu GE, Paige CJ, Maki RA. et al. muscular junction. Annu Rev Neurosci 22:389–442.
1996. Targeted disruption of the PU. 1 gene results in multiple hema- Schafer DP, Lehrman EK, Kautzman AG, Koyama R, Mardinly AR,
topoietic abnormalities. EMBO J 15:5647–5658. Yamasaki R, Ransohoff RM, Greenberg ME, Barres BA, Stevens B.
Melom JE, Littleton JT. 2011. Synapse development in health and dis- 2012. Microglia sculpt postnatal neural circuits in an activity and
ease. Curr Opin Genet Dev 21:256–261. complement-dependent manner. Neuron 74:691–705.
Milnerwood AJ, Raymond LA. 2010. Early synaptic pathophysiology in Schafer DP, Stevens B. 2010. Synapse elimination during development
neurodegeneration: Insights from Huntington’s disease. Trends Neu- and disease: immune molecules take centre stage. Biochem Soc Trans
rosci 33:513–523. 38:476–481.
Monji A, Kato T, Kanba S. 2009a. Cytokines and schizophrenia: Micro- Scott EW, Simon MC, Anastasi J, Singh H. 1994. Requirement of tran-
glia hypothesis of schizophrenia. Psychiatry Clin Neurosci 63: scription factor PU.1 in the development of multiple hematopoietic
257–265. lineages. Science 265:1573–1577.
Monji A, Kato T, Kanba S. 2009b. Cytokines and schizophrenia: Micro- Shatz CJ. 1990. Competitive interactions between retinal ganglion cells
glia hypothesis of schizophrenia. Psychiatry Clin Neurosci 63:257–265. during prenatal development. J Neurobiol 21:197–211.

GLIA
36 SCHAFER ET AL.

Shatz CJ, Stryker MP. 1988. Prenatal tetrodotoxin infusion blocks Tremblay ME, Zettel ML, Ison JR, Allen PD, Majewska AK. 2012.
segregation of retinogeniculate afferents. Science 242:87–89. Effects of aging and sensory loss on glial cells in mouse visual and
Shi L, Fatemi SH, Sidwell RW, Patterson PH. 2003. Maternal influenza auditory cortices. Glia 60:541–558.
infection causes marked behavioral and pharmacological changes in Tropea D, Majewska AK, Garcia R, Sur M. 2010. Structural dynamics
the offspring. J Neurosci 23:297–302. of synapses in vivo correlate with functional changes during experi-
Siskova Z, Page A, O’Connor V, Perry VH. 2009. Degenerating synaptic ence-dependent plasticity in visual cortex. J Neurosci 30:
boutons in prion disease: microglia activation without synaptic strip- 11086–11095.
ping. Am J Pathol 175:1610–1621. Tsuda M, Shigemoto-Mogami Y, Koizumi S, Mizokoshi A, Kohsaka S,
Sretavan DW, Shatz CJ. 1986. Prenatal development of retinal ganglion Salter MW, Inoue K. 2003. P2X4 receptors induced in spinal micro-
cell axons: Segregation into eye-specific layers within the cat’s lateral glia gate tactile allodynia after nerve injury. Nature 424:778–783.
geniculate nucleus. J Neurosci 6:234–251. Turnbull IR, McDunn JE, Takai T, Townsend RR, Cobb JP, Colonna M.
Steiner J, Bielau H, Brisch R, Danos P, Ullrich O, Mawrin C, Bernstein 2005. DAP12 (KARAP) amplifies inflammation and increases mortal-
HG, Bogerts B. 2008. Immunological aspects in the neurobiology of ity from endotoxemia and septic peritonitis. J Exp Med 202:363–369.
suicide: Elevated microglial density in schizophrenia and depression Turrigiano GG. 2008. The self-tuning neuron: Synaptic scaling of exci-
is associated with suicide. J Psychiatr Res 42:151–157. tatory synapses. Cell 135:422–435.
Stellwagen D, Malenka RC. 2006. Synaptic scaling mediated by glial Turrigiano GG, Nelson SB. 2004. Homeostatic plasticity in the develop-
TNF-alpha. Nature 440:1054–1059. ing nervous system. Nat Rev Neurosci 5:97–107.
Stellwagen D, Shatz CJ. 2002. An instructive role for retinal waves in the van Lookeren Campagne M, Wiesmann C, Brown EJ. 2007. Macro-
development of retinogeniculate connectivity. Neuron 33:357–367. phage complement receptors and pathogen clearance. Cell Microbiol
Stellwagen D, Shatz CJ, Feller MB. 1999. Dynamics of retinal waves 9:2095–2102.
are controlled by cyclic AMP. Neuron 24:673–685.
Vargas DL, Nascimbene C, Krishnan C, Zimmerman AW, Pardo CA.
Stephan AH, Barres BA, Stevens B. 2012. The complement system: an 2005. Neuroglial activation and neuroinflammation in the brain of
unexpected role in synaptic pruning during development and disease.
patients with autism. Ann Neurol 57:67–81.
Annu Rev Neurosci 35:369–389.
Viegi A, Cotrufo T, Berardi N, Mascia L, Maffei L. 2002. Effects of dark
Stevens B, Allen NJ, Vazquez LE, Howell GR, Christopherson KS,
rearing on phosphorylation of neurotrophin Trk receptors. Eur J Neu-
Nouri N, Micheva KD, Mehalow AK, Huberman AD, Stafford B, Sher
rosci 16:1925–1930.
A, Litke AM, Lambris JD, Smith SJ, John SW, Barres BA. et al.
2007. The classical complement cascade mediates CNS synapse elimi- Waites CL, Garner CC. 2011. Presynaptic function in health and dis-
nation. Cell 131:1164–1178. ease. Trends Neurosci 34:326–337.
Syken J, Grandpre T, Kanold PO, Shatz CJ. 2006. PirB restricts ocular- Wake H, Moorhouse AJ, Jinno S, Kohsaka S, Nabekura J. 2009. Rest-
dominance plasticity in visual cortex. Science 313:1795–1800. ing microglia directly monitor the functional state of synapses in vivo
Tomasello E, Desmoulins PO, Chemin K, Guia S, Cremer H, Ortaldo J, and determine the fate of ischemic terminals. J Neurosci 29:
Love P, Kaiserlian D, Vivier E. 2000. Combined natural killer cell 3974–3980.
and dendritic cell functional deficiency in KARAP/DAP12 loss-of-func- Yang XW, Lu XH. 2011. Molecular and cellular basis of obsessive-com-
tion mutant mice. Immunity 13:355–364. pulsive disorder-like behaviors: Emerging view from mouse models.
Tomasello E, Olcese L, Vely F, Geourgeon C, Blery M, Moqrich A, Gau- Curr Opin Neurol 24:114–118.
theret D, Djabali M, Matteii M, Vivier E. 1998. Gene structure, Watts RJ, Schuldiner O, Perrino J, Larsen C, Luo L. 2004. Glia engulf
expression pattern, and biological activity of mouse killer cell activat- degenerating axons during developmental axon pruning. Curr Biol
ing receptor-associated protein (KARAP)/DAP-12. J Biol Chem 14:678–684.
273:34115–34119. Yoshiyama Y, Higuchi M, Zhang B, Huang SM, Iwata N, Saido TC,
Torborg CL, Feller MB. 2005. Spontaneous patterned retinal activity Maeda J, Suhara T, Trojanowski JQ, Lee VM. 2007. Synapse loss and
and the refinement of retinal projections. Prog Neurobiol 76:213–235. microglial activation precede tangles in a P301S tauopathy mouse
Trapp BD, Wujek JR, Criste GA, Jalabi W, Yin X, Kidd GJ, Stohlman model. Neuron 53:337–351.
S, Ransohoff R. 2007. Evidence for synaptic stripping by cortical Ziburkus J, Guido W. 2006. Loss of binocular responses and reduced
microglia. Glia 55:360–368. retinal convergence during the period of retinogeniculate axon segre-
Tremblay MÈ. 2012. The role of microglia at synapses in the healthy CNS: gation. J Neurophysiol 96:2775–2784.
Novel insights from recent imaging studies. Neuron Glia Biol 15:1–10. Ziegenfuss JS, Biswas R, Avery MA, Hong K, Sheehan AE, Yeung YG,
Tremblay ME, Lowery RL, Majewska AK. 2010. Microglial interactions Stanley ER, Freeman MR. 2008. Draper-dependent glial phagocytic
with synapses are modulated by visual experience. PLoS Biol activity is mediated by Src and Syk family kinase signalling. Nature
8:e1000527. 453:935–939.

GLIA