Synapse Location during Growth
Depends on Glia Location
Zhiyong Shao,1Shigeki Watanabe,2Ryan Christensen,1Erik M. Jorgensen,2and Daniel A. Colo ´n-Ramos1,*
1Program in Cellular Neuroscience, Neurodegeneration and Repair, Department of Cell Biology, Yale University School of Medicine,
P.O. Box 9812, New Haven, CT 06536-0812, USA
2Howard Hughes Medical Institute, Department of Biology, University of Utah, Salt Lake City, UT 84112-0840, USA
Synaptic contacts are largely established during
embryogenesis and are then maintained during
growth. To identify molecules involved in this pro-
cess, we conducted a forward genetic screen in
C. elegans and identified cima-1. In cima-1 mutants,
synaptic contacts are correctly established during
embryogenesis, but ectopic synapses emerge dur-
ing postdevelopmental growth. cima-1 encodes a
solute carrier in the SLC17 family of transporters
that includes sialin, a protein that when mutated in
humans results in neurological disorders. cima-1
does not function in neurons but rather functions in
the nearby epidermal cells to correctly position glia
during postlarval growth. Our findings indicate that
CIMA-1 antagonizes the FGF receptor (FGFR), and
does so most likely by inhibiting FGFR’s role in
epidermal-glia adhesion rather than signaling. Our
data suggest that epidermal-glia crosstalk, in this
case mediated by a transporter and the FGF recep-
tor, is vital to preserve embryonically derived circuit
architecture during postdevelopmental growth.
The nervous system is largely established during embryogen-
esis, but connectivity persists throughout the lifetime of the
organism (Benardand Hobert,
C. elegans, as in other metazoans, neural circuitry is laid out
largely during embryogenesis (Sulston et al., 1983). Thereafter
the worm grows 100-fold in volume, yet axonal architecture
and synaptic contacts are maintained (Benard and Hobert,
2009; Knight et al., 2002). Genetic studies have identified mole-
cules required for the maintenance of axon positions during
growth and movement, including L1-CAM, F-spondin, and the
FGF receptor among others (Aurelio et al., 2002; Benard and
Hobert, 2009; Benard et al., 2012; Benard et al., 2006; Bu ¨low
et al., 2004; Johnson and Kramer, 2012; Pocock et al., 2008;
Sasakura et al., 2005; Woo et al., 2008). This work indicates
two important features regarding the maintenance of nervous
system architecture during development. First, the molecules
required formaintenance of axonposition aredistinct fromthose
required for circuit formation. Second, these studies underscore
the importance of regulated adhesion in maintenance of axon
Circuit architecture also requires the maintenance of synaptic
contacts. Synaptic maintenance studies have largely focused on
the stability of synapses (Lin and Koleske, 2010; Shi et al., 2012;
interactions can play key roles in the maintenance of synaptic
stability (Pfrieger, 2010). Less is known about how synaptic
distribution is maintained during postdevelopmental growth.
Synaptic distribution, which is critical for maintenance of the
embryonically derived circuit architecture, requires both mainte-
nance of correct synaptic contact and prevention of formation of
We perform a screen in C. elegans and identify cima-1. In
cima-1 mutants, inappropriate contacts between glia and axons
promote the formation of ectopic synapses. Glia inappropriately
contact axons in these mutants, likely due to increased adhesion
with epidermal cells during growth. CIMA-1 is a SLC17 family
solute transporter that modulates epidermal-glia interaction via
FGFR. We reveal a potential mechanism for the role of SLC17
transporters in maintenance of synaptic distribution. Further-
more, we suggest that reducing adhesion during growth is as
important as promoting adhesion to maintain correct synaptic
AIY Synapses Form during Embryogenesis and Are
Maintained during Growth
The AIY interneurons are a pair of bilaterally symmetric neurons
in the nematode nerve ring (Figure 1A). Although these neurons
contact many potential synaptic partners, they display remark-
able specificity for both synaptic partner and position (White
et al., 1986). In adult animals, the observed pattern of synaptic
outputs in AIY is reproducible across animals (Colo ´n-Ramos
et al., 2007). To determine when AIY synaptic outputs are estab-
lished, we examined the AIY presynaptic pattern in C. elegans
larval stages using GFP::RAB-3 (Nonet et al., 1997). We
observed that the presynaptic pattern was already established
by the time the animals hatched at larval stage 1 (L1) (Figures
1A, 1B, 1C and Colo ´n-Ramos et al., 2007). We also observed
that although the length of the neurite and synaptic zones
Cell 154, 337–350, July 18, 2013 ª2013 Elsevier Inc. 337
increase as the animal grows, the relative distribution of presyn-
aptic sites is maintained (Figures 1B, 1C, 1J, and 1K). Our find-
ings indicate that the presynaptic pattern in the AIY interneuron
is established during embryogenesis and is maintained as the
cima-1 Inhibits Ectopic Synapses during Growth
To identify the mechanisms underlying the maintenance of syn-
aptic distribution during growth, we performed a visual forward
genetic screen and isolated cima-1(wy84) (for circuit mainte-
nance). GFP::RAB-3 distribution was indistinguishable between
cima-1(wy84) and wild-type animals at larval L1 stage (Figures
1B, 1F, 1J, and 1K). cima-1(wy84) adult animals, however, dis-
played a highly penetrant ectopic GFP::RAB-3 localization in
the normally asynaptic zone 1 of AIY (Figure 1G, >90% of
animals, n > 200 animals). Quantification of the cima-1(wy84)
phenotype in adult animals revealed that although the length of
zone 3 remained similar between mutants and wild-type adult
animals, the ventral presynaptic region (zone 2, and ectopic pre-
synaptic structures in zone 1) was twice as long in cima-1(wy84)
mutants (Figures 1G and 1J). The synaptic defect in cima-
1(wy84) was confirmed using synaptic vesicle proteins SNB-1
and SNG-1, as well as active zone protein SYD-1 (Figures 1D,
1E, 1H, and 1I and data not shown). Fluorescence electron
microscopy (fEM) also demonstrated the presence of ectopic
presynaptic sites in cima-1(wy84) animals (Figure S1 available
online) (Watanabe and Jorgensen, 2012; Watanabe et al.,
2011). Together, our data indicate that cima-1(wy84) is not
required for establishing presynaptic distribution but is required
for maintaining it.
Ectopic Presynaptic Sites in cima-1 Mutants Are Not
onto Postsynaptic Partner RIA
The synapses in zone 2 of wild-type animals are formed primarily
onto postsynaptic partner RIA (Figure 2A and White et al., 1986).
To determine whether the ectopic synapses are targeted to RIA,
c i t
o i t
) l a
t / l a
L1 L4 Adult
o i g
c i t
L1 L4 Adult
Figure 1. cima-1 Is Required for Maintenance of AIY Presynaptic
Distribution during Growth
(A) Schematic diagram of bilaterally symmetric AIYs (gray) in the C. elegans
head (modified from WormAtlas with permission). Green marks presynaptic
positions. There are three distinct anatomical regions along the AIY neurite: a
segment proximal to AIY cell body that is devoid of synapses (zone 1, dashed
box), a dense presynaptic region at the dorsal turn of the AIY neurite (zone 2),
and a region with discrete presynaptic clusters at the distal part of the neurite
(zone 3)(Colo ´n-Ramos et al., 2007; White et al.,1986). Schematic illustration is
a modification with permission from the Neuron pages of WormAtlas (http://
www.wormatlas.org) by Z.F. Altun and D.H. Hall.
(B–E) The AIY presynaptic pattern in wild-type animals. Using synaptic vesicle
associated GFP::RAB-3, we observe that AIY presynaptic pattern is already
established in newly hatched larval L1 stage animals (B) and maintained in
adults (C). A similar pattern was observed when we visualized synaptic vesicle
protein SNG-1::YFP (D) and active zone protein GFP::SYD-1 (E) in adults and
L1 larva (data not shown).
(F–I) cima-1(wy84) mutant animals fail to maintain the AIY presynaptic pattern
in adults. cima-1(wy84) mutant L1 animals display a wild-type presynaptic
pattern as visualized with GFP::RAB-3 (F). Adult animals display an abnormal
presynaptic pattern as visualized with presynaptic proteins GFP::RAB-3 (G),
SNG-1::YFP (H), and GFP::SYD-1 (I). The ectopic presynaptic structure was
confirmed with fluorescence electron microscopy (fEM) (Figure S1). In all
images, dashed box corresponds tonormallyasynapticzone1,andscale bars
(J and K) Quantification of the AIY presynaptic pattern. Note that the length of
2) is similar in cima-1(wy84) and wild-type animals at L1 stage but becomes
longeratL4or adultstages(J).Theratioof thepresynaptic length(thelengthof
ventral presynaptic region divided by total presynaptic region as shown in [K])
is a metric that reflects the general pattern of AIY, and persists in wild-type
animals during growth (black bars). Note that in cima-1(wy84) mutant animals
this ratio becomes abnormally larger, particularly in postdevelopmental
*p < 0.05, **p < 0.01, ***p < 0.001 by t test comparison.
See also Figure S1.
338 Cell 154, 337–350, July 18, 2013 ª2013 Elsevier Inc.
we simultaneously imaged RIA and the presynaptic sites in AIY.
We observed that in wild-type and cima-1(wy84) juvenile
animals, RIA projects to zone 2 correctly (Figures 2B and 2F).
However, we noted two differences in the adult mutants. First,
in adult cima-1(wy84) animals, the RIA neurite is extended
posteriorly (Figure 2G). Second, we also observed ectopic pre-
synaptic sites in zone 1 that extend beyond the area of contact
between AIY and RIA (compare Figures 2C with 2G and 2D
To further examine the relationship between AIY:RIA contact
and the ectopic presynaptic sites in zone 1, we performed GFP
reconstitution across synaptic partners (GRASP) (Feinberg
et al., 2008). GRASP is based on two GFP fragments (‘‘GFP
1-10’’ and ‘‘GFP 11’’) that can reconstitute a functional GFP
molecule only when they are in close proximity. A version of
GRASP based on the transmembrane protein CD4 allows
assessment of cell-cell contact sites (Feinberg et al., 2008).
We expressed this version of GRASP in AIY (CD4::GFP 11)
and RIA (CD4::GFP 1-10) to specifically visualize AIY:RIA con-
tact and simultaneously labeled AIY presynaptic sites with
mcherry::RAB-3. We observed that ectopic presynaptic sites
were present beyond the AIY:RIA contact region (Figures 2E
and 2I). Our studies indicate that the abnormal distribution of
presynaptic structures in cima-1(wy84) adults results from
two events: a posterior displacement of synapses between
AIY and RIA in zone 2 and the emergence of ectopic presynap-
tic sites, which are not apposed to postsynaptic cell RIA, in
cima-1 Encodes a Membrane Transporter in the SLC17
Our SNP mapping, genetic rescue and Sanger sequencing
data suggest that the cima-1(wy84) allele corresponds to a G
to A mutation in the unnamed gene F45E4.11, and that this
mutation is predicted to alter a conserved glycine at residue
388 to glutamate (Figures 3A, 3B, and 3F). A second allele of
F45E4.11(gk902665), with a nonsense mutation at R476,
We also observed a similar AIY presynaptic maintenance
defect when knocking down F45E4.11 by RNAi (Figures 3E).
Together, our genetic data indicate that cima-1(wy84) is a
missense, loss-of-function mutation in F45E4.11 (hereafter
cima-1 encodes a 12 transmembrane domain protein that is
homologous to the SLC17 family of solute carrier transporters
y t i s e t n i
c s e r o
y t i s e t n i
c s e r o
RIA AIY RAB-3
Figure 2. cima-1(wy84) Mutants Have a Posteriorly Extended Zone 2
and Ectopic Presynaptic Structures in Zone 1
(A) AIY (gray) forms synapses onto postsynaptic partner RIA (blue) in zone 2
(Colo ´n-Ramos et al., 2007; White et al., 1986); Schematic diagram modified
from WormAtlas with permission. (B and C) Simultaneous visualization of
synaptic vesicles in AIY (pseudocolor green) and postsynaptic GLR-1 sites in
(C), RIA contacts AIY in zone 2 and not in zone 1. The arrow indicates the
contacts AIY. Schematic illustration is a modification with permission from the
Neuron pages of WormAtlas (http://www.wormatlas.org) by Z.F. Altun and
(D) 3D profile of AIY presynaptic CFP::RAB-3 (pseudocolor green) fluores-
cence intensity (arbitrary units) of (C). The arrow indicates the transition
between zone 2 and zone 1.
(E) Simultaneous visualization of GRASP GFP (which indicates contact
between presynaptic neuron AIY and postsynaptic partner RIA) and
mCherry::RAB-3 in a wild-type animal. Note that the AIY:RIA contact indicated
by GFP overlaps with presynaptic mCherry::RAB-3 at zone 2 region.
(F and G) The AIY presynaptic pattern and RIA morphology are wild-type in
newly hatched (L1 stage) cima-1(wy84) mutant animals (F). However, at the
adult stage the RIA neurite is posteriorly extended (from AIY ventral turn
indicated by yellow arrow to the posterior site indicated by white arrow) and
ectopic presynapses are seen in zone 1 (bracket) (G).
(H) 3D profile of AIY presynaptic GFP::RAB-3 fluorescence intensity (arbitrary
units) of the image in (G).
(I) Simultaneous visualization of GRASP GFP and mCherry::RAB-3 in a cima-
1(wy84) animal. Note that presynaptic mCherry::RAB-3 extends beyond the
AIY:RIA contact region (indicated by GRASP GFP signal).
Ectopic presynaptic sites in AIY are bracketed. All arrows except the yellow in
(G) indicate the end of zone 2 and the beginning of zone 1. The yellow arrow in
(G) indicates where the end of zone 2 and the beginning of zone 1 should be.
Scale bars, 10mm (scale bar in [B] applies to [B] and [F]; scale bar in [C] applies
to [C], [G], [E] and [I]).
Cell 154, 337–350, July 18, 2013 ª2013 Elsevier Inc. 339
and similar to sialin, a lysosomal transporter associated with
neurodegenerative disease (Reimer, 2013; Sreedharan et al.,
2010; Verheijen et al., 1999) (Figures 3F and S2).
cima-1 Acts in Epidermal Cells
To understand the mechanism by which cima-1 suppresses
ectopic presynapses, we first examined its endogenous expres-
sion pattern. We generated a transcriptional GFP reporter and
found cima-1 expression starts during embryogenesis and per-
sists in adult animals (Figures 4A–4D and S3A–S3H). In adults,
cima-1 is primarily expressed in epidermal cells. Expression
AIY) or in glia (Figures S3E–S3H).
Consistent with the expression pattern, we observed that
expression of CIMA-1 cell specifically in epidermal cells results
in robust rescue of the AIY phenotype (Figure 4E and S3I).
Conversely, expression of CIMA-1 cell specifically in the AIY
interneurons (ttx-3 promoter), in all neurons (rab-3 promoter),
or in neurons and intestine (aex-3 promoter) did not rescue (Fig-
ure 4E and S3I). These data suggest that cima-1 acts in
epidermal cells to maintain the AIY presynaptic pattern.
We note that although cima-1 is required in epidermal cells,
cima-1(wy84) mutant animals do not exhibit obvious defects in
body morphology, epidermal cell morphology, cellular fusion
events, epidermal development, or general neuroarchitecture
in larvae or young adults (Figure S4 and data not shown).
tct gct gga cta
IV 2M 4M 6M 8M 10M 12M 14M 16M
e r p
c i p
o t o
cima-1 wy84 wy84 wy84
transgene - WRM0612bA03 F45E4.11
Figure 3. cima-1(wy84) Is an Allele of F45E4.11, which Encodes a Conserved SLC17 Family Transporter
(A) SNP mappingindicates that thegenetic lesion corresponding tothecima-1(wy84) allele isbetween7.40Mband 7.83 MbonchromosomeIV.Two overlapping
fosmids in this region (WRM0612bA03 and WRM0615cC03) rescue the cima-1(wy84) AIY presynaptic defect. Those two fosmids overlap in a genomic area that
includes just two genes: F45E4.11 and C08G9.1. Only F45E4.11 was able to rescue the AIY defect. Sequencing data indicate a missense mutation in the coding
region that alters conserved G388 to E.
(B) Quantification of the percentage of animals displaying the AIY presynaptic patterning defect in cima-1(wy84) mutants transformed with fosmid
WRM0612bA03 (which includes gene F45E4.11) or with just gene F45E4.11. n R 86 for each category. Error bars are SEM, ***p < 0.001 by t test.
(C) A different F45E4.11 allele, gk902655, contains a nonsense mutation in the coding region (R476 to opal stop codon) and phenocopies the wy84 allele.
(D and E) Knockdown of F45E4.11 by RNAi phenocopies the cima-1(wy84) presynaptic phenotype in AIY. Animals fed with bacteria transfected with control
vector L4440 show normal AIY presynaptic distribution (D), whereas animals fed with bacteria expressing F45E4.11 dsRNA phenocopy the cima-1(wy84) AIY
presynaptic phenotype (E).
Scale bar, 10mm (C–E). In (C–E), zone 1 region is highlighted with a dashed box.
(F) A schematic diagram of the predicted cima-1 topology. The mutated G388 (asterisk) in cima-1(wy84) is located in ninth transmembrane domain. CIMA-1 is a
member of SLC17 transporter family (See also Figure S2).
340 Cell 154, 337–350, July 18, 2013 ª2013 Elsevier Inc.
Next, we wanted to determine the temporal requirement of
cima-1. As AIY presynaptic defects are only observed in adults,
we hypothesized that expression of cima-1 after larval develop-
ment would be sufficient to rescue the AIY presynaptic defect.
To test this, we expressed cima-1 using a col-19 promoter that
expresses just in epidermal cells and after larval development
s e s p
n y s e r p
Y I A
c i p
o t c
) s l a
m i n
Figure 4. cima-1 Is Expressed and Required in Epidermal Cells for Maintenance of the AIY Presynaptic Pattern
(A) A larval animal displaying the endogenous cima-1 expression pattern as determined by rescuing construct CIMA-1(genomic)::SL2::GFP. The dashed box
(B–D) Simultaneous visualization of cima-1(genomic)::SL2::GFP (green) and body wall muscle reporter Pmyo-3::mcherry (red). Note the nonoverlapping ex-
pressing pattern of cima-1 in epidermal cells and Pmyo-3::mcherry in muscles, both in the sagittal cross-section (B) and in the transverse cross-section (C).
Dashed line in (B) indicates site of transverse cross-section image in (C). And (D) is the schematic drawing of (C), with muscle quadrants (‘‘M’’), epidermal cells
(green) and pharynx (P) labeled.
(E) Quantification of tissue-specific rescue. Expression of cima-1 cDNA in AIY (ttx-3 promoter) or pan-neuronally (rab-3 promoter) does not rescue the AIY
presynaptic defect in cima-1(wy84) mutant animals. However, expression of cima-1 cDNA in epidermal cells (dpy-4 promoter) robustly rescues the AIY pre-
synaptic defect (see also rescue with epidermal promoters rol-6, dpy-7 and col-19 in Figure S3I). n R 50 for each group. Error bars are SEM, n.s.: not significant,
***p < 0.001 by t test. Scale bars, 10 mm.
See also Figures S3, S4 and S5.
Cell 154, 337–350, July 18, 2013 ª2013 Elsevier Inc. 341
AIY: RAB-3 Glia
AIY: RAB-3 Glia
AIY: RAB-3 Glia
AIY: RAB-3 Glia
AIY: RAB-3 Glia
AIY: RAB-3 Glia
AIY: RAB-3AIY: RAB-3AIY: RAB-3
cima-1 rescued adult
AIY: RAB-3 Glia
AIY: RAB-3 Glia
cima-1 rescued adult
cima-1 rescued adult
cima-1 rescued adult
Figure 5. cima-1 Is Required for Maintenance of VCSC Glial Morphology during Growth
(A) Relativeposition of epidermal cells, VCSC glia and AIY in C. elegans. A cross section of EM image of a wild-type animal in the zone 2 region of AIY (from http://
www.WormAtlas.org and http://www.WormImage.org) (White et al., 1986). VCSC glia (pseudocolored red in the micrograph) lie between the epidermal cells
(purple) and AIY at zone 2 synapses (green, note vesicles and dense projections in the two AIY-neurite cross sections). The dashed box in the schematic of
the worm head represents the region where images (and schematic illustrations) in ([B)–(Q) were obtained. The electron micrograph is from animals ‘‘JSH’’ and
(legend continued on next page)
342 Cell 154, 337–350, July 18, 2013 ª2013 Elsevier Inc.
(Cox and Hirsh, 1985; Liu and Ambros, 1991). We observed sig-
nificant rescue of the AIY presynaptic defect, consistent with
cima-1 playing a postdevelopmental role in epidermal cells to
maintain synaptic distribution (Figure S3I).
The cima-1 Phenotype Is Affected by Animal Size
Animals increase their size after concluding larval development.
To examine if cima-1 is required to maintain synaptic positions
during postdevelopmental growth, we visualized mutants that
have abnormally short (dumpy or Dpy) or long (Lon) body length
(Figures S5A–S5C; Page and Johnstone, 2007). The presynaptic
pattern in Dpy animals retained a wild-type distribution in AIY
synaptic region in AIY, but their synaptic distribution does not
phenocopy cima-1 (Figures S5C, S5F, and S5J). However,
enhanced AIY presynaptic maintenance defect (Figures S5I and
S5J). Importantly, the enhancement of the presynaptic distribu-
tion phenotype in cima-1(wy84) lon-3(e2175) double mutants
was observed after late L4 stage (data not shown). Moreover,
cima-1(wy84) dpy-4(e1166) double mutants suppressed the
cima-1(wy84) AIY presynaptic defect (Figure S5E). This suppres-
sion is likely due to the shorter size of the animal, as a number of
Dpy alleles with different genetic identities gave us the same
result (dpy-7(e88), dpy-9(e12), and dpy-6(e14)) (Figure S5H and
data not shown).Together, our findings support a role for cima-
1 in maintaining presynaptic distribution during postdevelop-
cima-1 Is Required for Maintenance of Glial Morphology
The ventral cephalic sheath cells (VCSC) are nonneuronal cells
the cima-1-expressing hyp7 epidermal cell and the AIY interneu-
these glial cells, we expressed cytoplasmic mCherry using the
glia-specific hlh-17 promoter (McMiller and Johnson, 2005).
In wild-type animals, VCSC glia are in close proximity to AIY
5B–5E). In larval stages, cima-1(wy84) VCSC morphology is
indistinguishable from that of wild-type animals (Figures 5F–5I).
However, in adult stages, the glial processes (endfeet) were
abnormally distended posteriorly onto zone 1 of AIY (Figures
5J–5M). Cell-specific expression of cima-1 in epidermal cells
was sufficient to rescue the glial morphology (Figures 5N–5Q).
Thus, cima-1 is required in epidermal cells for maintenance of
glial morphology during postlarval growth.
Doespositionoftheglial endfeetcorrelatewith theemergence
of ectopic synaptic sites? To address this question, we conduct-
edlongitudinal studiesand observed a tight temporal and spatial
correlation between the position of the glial endfeet and the po-
sition of AIY presynaptic sites (Figure S6). Our findings indicate
that cima-1 acts in epidermal cells to maintain VCSC glial
tween the defective maintenance of glial morphology and the
emergence of ectopic presynaptic sites in AIY.
Glial Cell Ablation Suppresses Ectopic Synaptic Sites in
Our findings are consistent with a model in which distended glial
endfeet in cima-1(wy84) mutants result in ectopic contact
between glia and AIY, which in turn induces the formation of
ectopic synaptic sites. To examine this hypothesis, we per-
formed GFP reconstitution across synaptic partners (GRASP)
between glia and AIY (Feinberg et al., 2008). We expressed
CD4::GFP 11 in AIY and CD4::GFP 1-10 in glia to examine sites
of AIY:glia contact. Consistent with published EM data (Figures
S1A–S1B and White et al., 1986), we observed that in wild-
type animals the GRASP signal was restricted to zone 2 and
colocalized with the synaptic marker mCherry::RAB-3 (Figures
6A–6C). cima-1(wy84) mutant animals displayed an ectopic
GRASP signal in zone 1 that colocalized with the ectopic
mCherry::RAB-3 (Figures 6D–6F). Our findings suggest that
ectopic AIY:glia contact in cima-1(wy84) mutant animals results
in ectopic presynaptic specializations (Figures 6E and 6F).
lated the glia by expressing caspases in these cells (Chelur and
Chalfie, 2007). The cell death was confirmed by the absence of
VCSC GFP. We observed that cell-specific ablation of the glia in
tic sites in zone 1 (Figures 6H and 6I). Second, we examined the
AIY presynaptic pattern in cima-1(wy84) mls-2(ns156) double
ment, and in mls-2(ns156) mutant animals the glia do not properly
results, we also observed that ectopic presynaptic sites in zone 1
are suppressed in cima-1(wy84) mls-2(ns156) double mutant
animals (Figures 6I). It should be noted, however, that loss of the
type. Incomplete suppression could result from two possibilities:
(1) caspase and genetic ablations do not completely eliminate
the glia, or (2) other tissues besides VCSC glia also influence
the persistence of ectopic presynaptic sites in AIY.
‘‘N2U’’ shown with permission, each was prepared by J. White, E. Southgate, N. Thomson, and S. Brenner at the LMB/MRC labs in Cambridge, England (White
org. Schematic illustration is a modification with permission from the Neuron pages of WormAtlas (http://www.wormatlas.org) by Z.F. Altun and D.H. Hall.
(B–Q) Simultaneous visualization of AIY presynaptic sites (green) and VCSC glia (red) in wild-type adult animal (B–D), cima-1(wy84) L4 animal (F)–(H),
cima-1(wy84) adult animal (J)–(L), or cima-1(wy84) adult animal rescued with Pdpy-4::cima-1 (N)–(P). Images (D), (H), (L), and (P) are as (C), (G), (K), and (O), but
overlaid with DIC. Note that in cima-1(wy84) adult animals, VCSC glia abnormally distend to the zone 1 region and overlap with AIY ectopic presynapses (dashed
box; [J]–[M]). Both glia and AIY presynaptic defects are rescued by expressing cima-1 cDNA in epidermal cells (N)–(Q). In all images, white and black asterisk
represent the location of the AIY cell body, and orange asterisks mark pharyngeal grinder. Note that in K and L, synapses are formed in zone 1, past the orange
asterisk. (E), (I), (M), and (Q) are cartoons of (D), (H), (L), and (P) with pharynx in gray (see schematic in [A]). The hlh-17 promoter labels both dorsal and ventral glial
cells. The dorsal glia not labeled in (K) and (L) is due to the mosaic retention of the transgenic marker.
Scale bar in (B) corresponds to 10 mm and applies to all images. See also Figure S6.
Cell 154, 337–350, July 18, 2013 ª2013 Elsevier Inc. 343
Together, these data suggest that epidermally expressed
cima-1 regulates glial morphology during postlarval growth and
that maintenance of correct glial morphology during growth is
required for preventing the emergence of ectopic presynaptic
sites in AIY.
Ectopic Synapses Do Not Require Embryonic
During embryogenesis, AIY synapse formation requires UNC-6/
Netrin, which is transiently expressed by VCSC glia (Colo ´n-
Ramos et al., 2007; Wadsworth et al., 1996). UNC-6 instructs
presynaptic assembly by signaling through its receptor UNC-
40/DCC, expressed in AIY. We used the Netrin receptor, UNC-
40, to test whether Netrin signaling is required for the cima-1
mutant phenotype. To achieve this, we generated cima-
1(wy84) unc-40(e271) double mutants and visualized the AIY
presynaptic pattern. In juvenile worms, we observed a general
reduction of presynaptic vesicle clusters in cima-1(wy84)
unc-40(e271) compared to wild-type animals, as expected (Fig-
ure S7H and Colo ´n-Ramos et al., 2007). In adults, cima-1(wy84)
s e s p
n y s e r p
Y I A
c i p
o t c
) s l a
m i n
AIY: RAB-3 RIA
AIY: RAB-3 RIA
mCherry::RAB-3 GRASP AIY:VCSCMerge
Figure 6. VCSC Glia Abnormally Contact AIY and Are Required for Formation of Ectopic Presynaptic Sites in Zone 1 in cima-1(wy84).
(A–F) Simultaneous visualization of GRASP GFP signal (which indicates contact between AIY and VCSC glia) (A, D) and mCherry::RAB-3 ([B] and [E]) in wild-type
([A]–[C]) and cima-1(wy84) ([D]–[F]) adult animals. Note that in cima-1(wy84) mutants VCSC glia abnormally contact AIY in zone 1 (indicated by GRASP GFP in
dashed box), and that these sites correlate with sites of ectopic mCherry::RAB-3 in zone 1.
(G) In cima-1(wy84) adult animals, AIY forms ectopic presynaptic sites in zone 1 (dashed box) beyond zone 2 (determined by the position of postsynaptic partner
RIA in blue). Arrow indicates the end of zone 2 and the beginning of zone 1.
(H) As in (G), but with VCSC genetically ablated through the cell-specific expression of caspases. Note suppression of ectopic presynaptic sites in zone 1
(I) Quantification of the percentage of animals displaying ectopic presynaptic sites in zone 1 in cima-1(wy84) adult mutants; in cima-1(wy84) adult mutants
expressing caspases cell specifically in VCSC glia, or in cima-1(wy84) mls-2(ns156) double mutants.
The scale bar in (A) is 10mm and applies to (A)–(H). n R 37 for each genotype. Error bars represent 95% confidence interval. ***p < 0.001 between groups as
determined by Fisher’s exact test. See also Figure S1.
344 Cell 154, 337–350, July 18, 2013 ª2013 Elsevier Inc.
unc-40(e271) animals still displayed ectopic presynaptic sites in
zone 1 similar to cima-1(wy84) single mutants (Figures S7A–
S7G). These findings suggest that VCSC glia use different mo-
lecular mechanisms for presynaptic assembly during embryonic
during postdevelopmental growth.
EGL-15(5A)/FGFR Is Required for Ectopic Synapse
To understand the molecular mechanisms by which cima-1
affects AIY presynaptic maintenance during postlarval growth,
we conducted candidate suppressor screens using RNAi and
available mutants. These genetic approaches revealed that
mutations of egl-15, the only FGFR in C. elegans, suppress the
cima-1 AIY presynaptic defect (Figures 7A, 7B, and 7C, and
S7I; n > 300 animals).
egl-15/FGFR is required during C. elegans development for
sex myoblast migration, axon outgrowth, and fluid homeostasis,
and null alleles of egl-15 are larval lethal (Bu ¨low et al., 2004;
Goodman et al., 2003). Splicing isoform egl-15(5A), but not iso-
form egl-15(5B), is also required postembryonically to maintain
axon position during growth and movement (Bu ¨low et al.,
2004). We observed that egl-15(n484), a nonsense allele specific
to the egl-15(5A) isoform, suppressed cima-1(wy84) ectopic
presynaptic sites (Figures 7A, 7B, and S7I). Similarly, egl-
15(ay1), an egl-15(5A) splicing acceptor mutation (Goodman
et al., 2003), also suppressed cima-1(wy84) AIY ectopic synap-
ses (Figures 7C and S7I). These data demonstrate that the for-
mation of ectopic presynaptic sites in cima-1 mutants requires
We next examined VCSC glial morphology in cima-1(wy84)
egl-15(n484) double mutants. We observed that egl-15(n484)
also suppresses the abnormal glial morphology observed in
cima-1(wy84) mutants (Figures 7D and 7E). Consistent with this
observation, glia ablation did not enhance suppression of the
cima-1(wy84) egl-15(n484) double mutants (Figure S7I). Our
data suggest that egl-15(n484) suppresses the AIY presynaptic
defect by suppressing abnormal VCSC glial morphology in
EGL-15(5A)/FGFR Acts in the Epidermis Independent of
Expression of egl-15(5A) cDNA using a pan-neuronal promoter
did not restore ectopic presynapses to egl-15(n484) cima-
1(wy84) double mutants. However, expression of egl-15(5A)
in epidermal cells restored ectopic presynapses to egl-
15(n484) cima-1(wy84) double mutants (Figures 7F and S7I),
suggesting that FGFR isoform 5A, like cima-1, acts in the
EGL-15 isa receptor tyrosine kinase activated byFGFligands,
whichactivates downstreamRas pathways (Borlandetal.,2001;
Sundaram, 2006); kinase activity is required by EGL-15(5B) to
regulate axon outgrowth and fluid homeostasis (Bu ¨low et al.,
2004; Goodman et al., 2003; Huang and Stern, 2004). However,
EGL-15(5A) does not require the intracellular kinase domain to
maintain axon positions (Bu ¨low et al., 2004). Similarly, we found
that neither egl-17(ay6) nor let-756(s2613) suppress the presyn-
aptic phenotype in cima-1(wy84) mutants (data not shown),
suggesting thatFGFligands, egl-17andlet-756, arenotrequired
for ectopic presynapse formation in cima-1 mutants.
To further test whether the kinase domain of EGL-15(5A) is
required for the formation of ectopic synapses, we expressed
the previous described egl-15(5A) extracellular domain (‘‘egl-
15(5A)ecto’’; Bu ¨low et al., 2004) in the epidermal cells in
egl-15(n484) cima-1(wy84) double mutant animals, and found
that expression of the ectodomain reverted egl-15(n484) sup-
pression of cima-1(wy84) (Figure S7I). Together, these findings
indicate that egl-15(5A) is required in epidermal cells in a
kinase-independent manner to distort glial morphology in
CIMA-1 Negatively Regulates EGL-15(5A)/FGFR Protein
To determine howcima-1 acts in epidermal cells to maintain pre-
synaptic positions, we first examined its subcellular localization.
We observed that CIMA-1::RFP is largely colocalized with lyso-
somal marker GFP::CUP-5 (Figures S3V–S3Y, Pearsons cor-
relation = 0.53), suggesting that CIMA-1 localizes primarily to
lysosomes. However, alleles that disrupt lysosomal function
(such as glo-1(zu391), glo-4(ok623), rab-7(ok511), and cup-5
(ar465)) did not phenocopy cima-1(wy84) (data not shown). Our
findings suggest that the cima-1 phenotype does not result
from general lysosomal dysfunction.
CIMA-1 localization is reminiscent of that seen for sialin, a
vertebrate homolog of CIMA-1 that also localizes to lysosomes
and vesicles and that in humans is associated with neurodegen-
erative disease (Verheijen et al., 1999). Sialin regulates trans-
membrane and extracellular adhesion molecules (Galuska
et al., 2010; Hildebrandt et al., 2009; Morin et al., 2004; Myall
et al., 2007; Prolo et al., 2009; Verheijen et al., 1999). Based on
CIMA-1 localization, and its genetic interaction with EGL-
some to regulate EGL-15(5A) in epidermal cells. To examine this
hypothesis, we generated transgenic animals expressing C
terminus HA-tagged EGL-15(5A) just in epidermal cells and
probed protein levels by western blots in both cima-1(wy84)
and wild-type animals. We consistently found that EGL-15(5A)
levels are 5-fold higher in cima-1(wy84) animals than in wild-
typeanimals (Figures 7Gand7H,p =0.007). Thisresult suggests
that cima-1 is required for the regulation of EGL-15(5A) protein
levels in epidermal cells.
To further test whether the synaptic defect in cima-1 is due
to the increase of EGL-15(5A) in epidermal cells, we generated
transgenic lines overexpressing EGL-15(5A) in epidermal
cells and observed that overexpression of EGL-15(5A) in
wild-type animals resulted in abnormal glial morphology and
ectopic presynaptic specializations in zone 1, similar to the
phenotypes observed for cima-1(wy84) mutants (Figures 7I–
7K). Together, our findings support a model whereby the trans-
porter cima-1 maintains glial morphology and presynaptic
distribution by negatively regulating EGL-15(5A)/FGFR during
growth. We hypothesize that reduction of EGL-15(5A) levels
would result in reduced adhesion between glia and epidermal
cells, which would be critical in maintaining glia location
and, therefore, synapse location during postdevelopmental
Cell 154, 337–350, July 18, 2013 ª2013 Elsevier Inc. 345
D RAB-3 Glia
E RAB-3 Glia
J RAB-3 Glia
cima-1(lof) or egl-15 (OE)
anti- EGL-15 anti-HA
n i e t o r p
v i t a l a
d l o f (2.0
c i p
o t c
) s l a
Figure 7. egl-15/fgfr Is Required for Ectopic Synapse Formation in cima-1(wy84) Animals
(A–C) GFP::RAB-3 in cima-1(wy84) mutant (A), cima-1(wy84) egl-15(n484) double mutant (B) and cima-1(wy84) egl-15(ay1) double mutant (C) adult animals. Note
that alleles egl-15(n484) and egl-15(ay1), which specifically disrupt egl-15(5A) isoform, suppress cima-1(wy84) presynaptic distribution defect.
(legend continued on next page)
346 Cell 154, 337–350, July 18, 2013 ª2013 Elsevier Inc.
maintains presynaptic distribution during C. elegans postdeve-
lopmental growth. We uncovered the requirement of a SLC17
family transporter CIMA-1. CIMA-1 antagonizes a specific
FGFR isoform, EGL-15(5A), in epidermal cells to modulate glial
morphology, in turn modulating AIY synaptic distribution.
CIMA-1 acts postdevelopmentally to regulate synaptic posi-
tions during growth. Between the L1 to L4 stage, animals un-
dergo a 4-fold change in length. However, during these stages,
cima-1 mutants resemble wild-type animals. We hypothesize
that cima-1-independent mechanisms scale synaptic growth
during development and that cima-1 is required to regulate syn-
aptic positions during postdevelopmental growth. Two lines of
evidence support this hypothesis. First, cima-1 expression after
larval stages using the col-19 promoter rescues the presynaptic
distribution phenotype in AIY. Second, in the cima-1 mutant
background, Lon mutants display an enhanced presynaptic
maintenance phenotype and Dpy mutants suppress the cima-1
phenotype. These findings demonstrate a relationship between
animal size and the postdevelopmental expressivity of the
cima-1 phenotype and indicate a role for cima-1 in maintenance
of the synaptic pattern during postdevelopmental growth.
cima-1 negatively regulates an FGFR isoform, egl-15(5A), to
modulate glial morphology during growth. Four lines of evidence
support this model. First, like cima-1, egl-15(5A) is also ex-
epidermal-glia interaction. Second, in the epidermal cells of
cima-1 mutant animals, EGL-15(5A) protein levels are signifi-
cantly increased. Third, loss of egl-15(5A) in cima-1 mutants
mostly restores glial morphology and correct presynaptic
patterning in AIY. Fourth, overexpression of egl-15(5A) in wild-
type animals results in abnormally extended glia and in ectopic
presynaptic sites in AIY. EGL-15(5A) was previously implicated
in the maintenance of axon position in the ventral nerve cord
(Bu ¨low et al., 2004). This function of EGL-15(5A) is also mediated
byits ecto-domain, which was hypothesized to provide an adhe-
sive substratum as a part of a larger adhesive complex (Benard
and Hobert, 2009; Bu ¨low et al., 2004). Our findings now demon-
strate that CIMA-1 can negatively regulate EGL-15(5A) to main-
tain glial morphology during growth.
CIMA-1 could act as a sugar transporter to regulate EGL-
15(5A) protein levels. CIMA-1 is a member of the solute carrier
SLC17 family that includes sialin. Although sialin is capable of
transporting a variety of cargos depending on its biological
context (Miyaji et al., 2010, 2011; Qin et al., 2012), its role as a
lysosomal transporter of acidic monosaccharides has been
implicated in neurological diseases (Verheijen et al., 1999;
Wreden et al., 2005). Importantly, several studies suggest that
sialin regulates intercellular adhesion via the modulation of
glycoconjugate export from the lysosome (Galuska et al., 2010;
Hildebrandt et al., 2009; Morin et al., 2004; Myall et al., 2007;
Prolo et al., 2009). CIMA-1 also localizes to the lysosome.
Although we have not yet identified the specific cargo for
CIMA-1, our phenotypic characterization uncovers an in vivo
function for this transporter in maintaining presynaptic dis-
tribution by regulating specific isoform of the FGFR, EGL-
15(5A). EGL-15 is N-glycosylated (Polanska et al., 2009). We
hypothesize that CIMA-1 could act like sialin in transporting
acidic monosaccharides that could modify and regulate protein
levels of EGL-15(5A). Although our findings are consistent with
this hypothesis and demonstrate a genetic interaction between
cima-1 and EGL-15(5A) in vivo, it is likely that cima-1 also regu-
lates other molecules, as mutations of egl-15(5A) do not
completely suppress the cima-1 phenotype.
The role of glia in synapse formation and function is well-
established. In both vertebrates and invertebrates, glia-derived
molecules promote synapse formation during development
(Allen et al., 2012; Colo ´n-Ramos et al., 2007; Fuentes-Medel
etal., 2012;Pfrieger,2010; Stevens,2008). Glia arealso required
in vivo for maintenance of synaptic function in adult animals. For
aptic function at neuromuscular junctions (Reddy et al., 2003).
Although glia are known to play roles in maintenance of synaptic
function, their role in maintenance of synaptic positions is less
(D–E) Simultaneous visualization of presynaptic sites in AIY (green) and glia (red) in cima-1(wy84) (D) and cima-1(wy84) egl-15(n484) double mutant adult animals
(E). Note that egl-15(n484) allele suppresses both the AIY presynaptic defect and the glia morphology defect in cima-1(wy84).
(F) Expression of egl-15(5A) cDNA in the epidermal cells (using the dpy-7 promoter) reverts the suppression of the cima-1(wy84) AIY presynaptic phenotype by
(G) EGL-15 crackle antibody (a gift from M Stern), detects endogenous EGL-15, the 141 kD band in wild-type (lane 1), but not in egl-15(n1477) mutants (lane 2),
which produced the C terminus truncated EGL-15 (M Stern, personal communication). HA antibody specifically recognizes HA-tagged EGL-15(5A) expressed in
epidermal cells (lane 4 and 5), but not in wild-type animals without the transgene (lane 3). For comparison in lanes 4 and 5, the same HA-tagged EGL-15(5A)-
expressing transgenic line were used. Note that EGL-15(5A) protein levels are higher in cima-1(wy84) mutant animals (lane 5) as compared to wild-type animals
(lane 4). Actin and coinjection marker Punc-122::GFP were used as loading control.
(H) Quantification of the EGL-15(5A)::HA protein levels from four independent blots. Error bars are SEM, **p < 0.01 by Student’s t test.
(I–J) AIY presynaptic GFP::RAB-3 (I) is mislocalized to zone 1 region abnormally ensheathed by VCSC (J) upon overexpression of egl-15(5A) in epidermal cells by
using the dpy-7 promoter (compare to Figure 5C).
(K) Quantification of the percentage of animals with the phenotype shown in (I). n R 150 for each group. Error bars represent 95% confidence interval. **p < 0.01
groups as determined by Fisher’s exact test.
(L) A model for cima-1 and egl-15(5A) in epidermal cells (purple) regulating VCSC glia (red) morphology and AIY presynaptic distribution (green). In wild-type
animals, cima-1 negatively regulates egl-15(5A), thereby reducing epidermal-glia adhesion and preventing glia extension during growth. This interaction
contributes to maintaining wild-type VCSC morphology, which in turn specifies correct synaptic distribution during growth (left cartoon). In animals with cima-1
loss-of-function, or animals in which egl-15(5A) is overexpressed, the interaction between the epidermal cells and VCSC is misregulated, resulting in VCSC glia
distension posteriorly, ectopic contacts between glia and axons, and ectopic presynaptic sites. Schematic illustration is a modification with permission from the
Neuron pages of WormAtlas (http://www.wormatlas.org) by Z.F. Altun and D.H. Hall.
See also Figure S7.
Cell 154, 337–350, July 18, 2013 ª2013 Elsevier Inc. 347
clear. A recent study demonstrates that growth of glial pro-
(Brink et al., 2012). We show a postdevelopmental role for glia in
maintaining synaptic distribution and circuit architecture during
The transduction of growth information from epithelial cell, to
glia, to neuron may be common. The epidermal epithelium in
C. elegans coordinates growth in the organism. For example,
genes expressed in epidermal cells regulate molting, body
morphogenesis and animal size (Chisholm and Hardin, 2005;
Chisholm and Hsiao, 2012; Chisholm and Xu, 2012). Our work
shows that the interaction of epidermal cells with glia translates
growth information to the neurons to limit synapse distribution.
The epithelial-glia interaction we uncovered here is reminiscent
of the neurovascular unit in vertebrates. Astrocytes play a funda-
mental role mediating communication between epithelial cells
and neurons in the vasculature of the brain (Abbott et al., 2006;
Banerjee and Bhat, 2007; Giaume et al., 2010; Kim et al., 2006;
Tam and Watts, 2010; Wolburg et al., 2009). For example, astro-
cytes indirectly control blood flow to neurons by coupling
neuronal activity to the epithelial cells of the vasculature (Allan,
2006; Koehler et al., 2009). Astrocytes also developmentally
couple epithelial cells of the vasculature with neurons during
embryogenesis (Tam and Watts, 2010). It is not yet known if
astrocytes translate growth information from epithelial cells in
the vasculature to position synapses as the animal grows. How-
ever, given existing roles for astrocytes in coupling functional
and developmental information between epithelial cells andneu-
rons, we speculate that analogous, glia-dependent mechanisms
like those in C. elegans could maintain synaptic position during
animal growth in metazoans.
For further experimental details on strains (Table S1), culture conditions, and
statistics, please see the Extended Experimental Information.
EMS Screen and RNAi
cima-1(wy84) was isolated from a visual forward EMS mutagenesis screen
aimed at identifying mutants with defects in GFP::RAB-3 distribution. cima-1
was mapped to an interval between 7.40 Mb and 7.83 Mb on chromosome
IV. The 18 fosmids that cover this region were injected into cima-1(wy84) mu-
tants and examined for rescue of AIY presynaptic defects. The F45E4.11
region in cima-1(wy84) was sequenced with Sanger sequencing technique.
Bacteria-mediated RNAi was performed as described in (Kamath et al., 2001).
Constructs and Nematode Transformation
Expression clones were made in the pSM vector, a derivative of pPD49.26 (A.
Fire) with extra cloning sites. Constructs are listed in Table S2, and detailed
cloning information will be provided upon request. Transgenic strains
(1–30 ng/ml) were generated using standard techniques (Mello and Fire,
1995) and listed in Table S2.
Microscopy and Imaging
electron microscopy as previously described (Watanabe et al., 2011). AIY
neurons in two cima-1(wy84) animals were identified based on fluorescence.
Fluorescence images and electron micrographs were correlated based on
the fiducial markers. Epidermal and VCSC glial cells were identified based on
theirmorphology. A totallength of6mmof anAIY andaVCSCglial cell werere-
constructed from serial electron micrographs as shown in Figure S1F. Recon-
prepared by Dr. John White (White et al., 1986), downloaded from http://www.
wormimage.org and reconstructed with TrakEM2 EM. For confocal micro-
scopy, images of fluorescently tagged fusion proteins were captured in live
C. elegans using an UltraView VoX spinning disc confocal microscope
(PerkinElmer) as described and analyzed using Volocity software (Improvision)
or ImageJ (Schneider et al., 2012; Stavoe and Colo ´n-Ramos, 2012).
We quantified EGL-15::HA in wild-type and cima-1(wy84) mutant animals
using olaEx1288 (Pdpy-7::egl-15(5A)::HA::MYC). 40-60 transgenic L4 animals
wereusedfortheblots.Reagents and detailedproceduresareintheExtended
Experimental Procedures section. This experiment was repeated four times
with similar results. We also performed an identical experiment with a second,
independent transgenic strain (olaEx1411) and three replicated western blots
with similar results.
figures,andtwotablesand canbefoundwiththisarticleonline athttp://dx.doi.
Some illustrations were created by Z.F. Altun and D.H. Hall from WormAtlas.
The EM in Fig 5A, Fig S1C and S1D were prepared by J. White, E. Southgate,
N. Thomson and S. Brenner at the LMB/MRC labs in Cambridge, England
(White et al., 1986). The adult image in the graphical abstract was created
by Dr. Maria Gallegos of Cal State University. We thank Dr. H. Bu ¨low, Dr. M.
Stern, Dr. S. Shaham, Dr. T. Kinnunen, Dr. J. Fares, Dr. K. Shen, and CGC
for strains and reagents. We also thank A. Pe ´rez and A. Roque for technical
assistance, Dr. X. Song for helping with the protein blots, and D. Hall,
S. Margolis, P. De Camilli, S. Strittmatter, M. Hammarlund, and members of
the Colo ´n-Ramos lab for helpful discussions and sharing of advice. This
work was funded by the following grants to D.C.-R. (R00 NS057931, R01
NS076558, a fellowship from the Klingenstein Foundation and the Alfred P.
Sloan Foundation and a March of Dimes Research Grant) and to E.M.J (NIH
Medical Institute. Z.S. and D.A.C-R. designed experiments. S.W. and E.M.J.
performed the fEM experiments. R. C. performed wild type EM reconstruction
and AIY Zone 1 length quantification. Z.S. performed all other experiments.
Z.S. and D.A.C-R. analyzed and interpreted the data. Z.S., S.W., E.M.J., and
D.A.C.-R. wrote the paper.
Received: November 3, 2012
Revised: April 9, 2013
Accepted: June 19, 2013
Published: July 18, 2013
Abbott, N.J., Ronnback, L., and Hansson, E. (2006). Astrocyte-endothelial
interactions at the blood-brain barrier. Nat. Rev. Neurosci. 7, 41–53.
Allan, S. (2006). The neurovascular unit and the key role of astrocytes in the
regulation of cerebral blood flow. Cerebrovasc. Dis. 21, 137–138.
Allen, N.J., Bennett, M.L., Foo, L.C., Wang, G.X., Chakraborty, C., Smith, S.J.,
and Barres, B.A. (2012). Astrocyte glypicans 4 and 6 promote formation of
excitatory synapses via GluA1 AMPA receptors. Nature 486, 410–414.
required for maintenance of ventral nerve cord organization. Science 295,
Banerjee, S., and Bhat, M.A. (2007). Neuron-glial interactions in blood-brain
barrier formation. Annu. Rev. Neurosci. 30, 235–258.
Benard, C., and Hobert, O. (2009). Looking beyond development: maintaining
nervous system architecture. Curr. Top. Dev. Biol. 87, 175–194.
348 Cell 154, 337–350, July 18, 2013 ª2013 Elsevier Inc.
Benard, C.Y., Boyanov, A., Hall, D.H., and Hobert, O. (2006). DIG-1, a novel gi-
ant protein, non-autonomously mediates maintenance of nervous system ar-
chitecture. Development 133, 3329–3340.
Benard, C.Y., Blanchette, C., Recio, J., and Hobert, O. (2012). The Secreted
Immunoglobulin Domain Proteins ZIG-5 and ZIG-8 Cooperate with L1CAM/
SAX-7 to Maintain Nervous System Integrity. PLoS Genet. 8, e1002819.
Borland, C.Z., Schutzman, J.L., and Stern, M.J. (2001). Fibroblast growth
factor signaling in Caenorhabditis elegans. Bioessays 23, 1120–1130.
Brenner, S. (1974). The genetics of Caenorhabditis elegans. Genetics 77,
Brink, D.L., Gilbert, M., Xie, X., Petley-Ragan, L., and Auld, V.J. (2012). Glial
processes at the Drosophila larval neuromuscular junction match synaptic
growth. PLoS ONE 7, e37876.
Bu ¨low, H.E., Boulin, T., and Hobert, O. (2004). Differential functions of the C.
elegans FGF receptor in axon outgrowth and maintenance of axon position.
Neuron 42, 367–374.
Chelur, D.S., and Chalfie, M. (2007). Targeted cell killing by reconstituted cas-
pases. Proc. Natl. Acad. Sci. USA 104, 2283–2288.
Chisholm, A.D., and Hardin, J. (2005). Epidermal morphogenesis. In Worm-
Book, The C. elegans Research Community, ed. http://dx.doi.org/10.1895/
Chisholm, A.D., and Hsiao, T.I. (2012). The Caenorhabditis elegans epidermis
as a model skin.I: development, patterning, and growth. Wiley Interdiscip Rev
Dev Biol. 1, 861–878.
Chisholm, A.D., and Xu, S. (2012). The Caenorhabditis elegans epidermis as a
model skin.II:differentiation and physiological roles. Wiley InterdiscipRev Dev
Biol. 1, 879–902.
Colo ´n-Ramos, D.A., Margeta, M.A., and Shen, K. (2007). Glia promote local
synaptogenesis through UNC-6 (netrin) signaling in C. elegans. Science 318,
Cox, G.N., and Hirsh, D. (1985). Stage-specific patterns of collagen gene
expression during development of Caenorhabditis elegans. Mol. Cell. Biol. 5,
Feinberg, E.H., Vanhoven, M.K., Bendesky, A., Wang, G., Fetter, R.D., Shen,
K., and Bargmann, C.I. (2008). GFP Reconstitution Across Synaptic Partners
(GRASP)defines cellcontacts andsynapsesinlivingnervoussystems.Neuron
Fuentes-Medel, Y., Ashley, J., Barria, R., Maloney, R., Freeman, M., and
Budnik, V. (2012). Integration of a Retrograde Signal during Synapse Forma-
tion by Glia-Secreted TGF-beta Ligand. Curr. Biol. 22, 1831–1838.
Galuska, S.P., Rollenhagen, M., Kaup, M., Eggers, K., Oltmann-Norden, I.,
Schiff, M., Hartmann, M., Weinhold, B., Hildebrandt, H., Geyer, R., et al.
(2010). Synaptic cell adhesion molecule SynCAM 1 is a target for polysialyla-
tion in postnatal mouse brain. Proc. Natl. Acad. Sci. USA 107, 10250–10255.
Giaume, C., Koulakoff, A., Roux, L., Holcman, D., and Rouach, N. (2010).
Astroglial networks: a step further in neuroglial and gliovascular interactions.
Nat. Rev. Neurosci. 11, 87–99.
Goodman, S.J., Branda, C.S., Robinson, M.K., Burdine, R.D., and Stern, M.J.
(2003). Alternative splicing affecting a novel domain in the C. elegans EGL-15
FGF receptor confers functional specificity. Development 130, 3757–3766.
Hildebrandt, H., Muhlenhoff, M., Oltmann-Norden, I., Rockle, I., Burkhardt, H.,
Weinhold, B., and Gerardy-Schahn, R. (2009). Imbalance of neural cell adhe-
sion molecule and polysialyltransferase alleles causes defective brain con-
nectivity. Brain 132, 2831–2838.
regulate fluid balance in C. elegans. Development 131, 2595–2604.
Johnson, R.P., and Kramer, J.M. (2012). Neural maintenance roles for the
matrix receptor dystroglycan and the nuclear anchorage complex in Caeno-
rhabditis elegans. Genetics 190, 1365–1377.
Kamath, R.S., Martinez-Campos, M., Zipperlen,P.,Fraser, A.G., and Ahringer,
J. (2001). Effectiveness of specific RNA-mediated interference through
ingested double-stranded RNA in Caenorhabditis elegans. Genome Biol 2,
Kim, J.H., Park, J.A., Lee, S.W., Kim, W.J., Yu, Y.S., and Kim, K.W. (2006).
Blood-neural barrier: intercellular communication at glio-vascular interface.
J. Biochem. Mol. Biol. 39, 339–345.
Knight, C.G., Patel, M.N., Azevedo, R.B., and Leroi, A.M. (2002). A novel mode
of ecdysozoan growth in Caenorhabditis elegans. Evol. Dev. 4, 16–27.
Koehler, R.C., Roman, R.J., and Harder, D.R. (2009). Astrocytes and the regu-
lation of cerebral blood flow. Trends Neurosci. 32, 160–169.
Lin, Y.C., and Koleske, A.J. (2010). Mechanisms of synapse and dendrite
maintenance and their disruption in psychiatric and neurodegenerative disor-
ders. Annu. Rev. Neurosci. 33, 349–378.
Liu, Z., and Ambros, V. (1991). alternative temperal control systems for hypo-
dermal cell differentiation in Carnorhabditis elegans. Nature 350, 162–165.
McMiller, T.L., and Johnson, C.M. (2005). Molecular characterization of HLH-
17, a C. elegans bHLH protein required for normal larval development. Gene
Mello, C., and Fire, A. (1995). DNA transformation. Methods Cell Biol. 48,
Miyaji, T., Omote, H., and Moriyama, Y. (2010). A vesicular transporter that
mediates aspartate and glutamate neurotransmission. Biol. Pharm. Bull. 33,
Miyaji, T., Omote, H., and Moriyama, Y. (2011). Functional characterization
of vesicular excitatory amino acid transport by human sialin. J. Neurochem.
Morin, P., Sagne, C., and Gasnier, B. (2004). Functional characterization of
wild-type and mutant human sialin. EMBO J. 23, 4560–4570.
Myall, N.J., Wreden, C.C., Wlizla, M., and Reimer, R.J. (2007). G328E and
G409E sialin missense mutations similarly impair transport activity, but differ-
entially affect trafficking. Mol. Genet. Metab. 92, 371–374.
Nonet, M.L., Staunton, J.E., Kilgard, M.P., Fergestad, T., Hartwieg, E., Horvitz,
H.R., Jorgensen, E.M., and Meyer, B.J. (1997). Caenorhabditis elegans rab-3
mutant synapses exhibit impaired function and are partially depleted of vesi-
cles. J. Neurosci. 17, 8061–8073.
Page, A.P., and Johnstone, I.L. (2007). The cuticle. In WormBook, The
C. elegans Research Community, ed. http://dx.doi.org/10.1895/wormbook.
Pfrieger, F.W. (2010). Role of glial cells in the formation and maintenance of
synapses. Brain Res. Brain Res. Rev. 63, 39–46.
Pocock, R., Benard, C.Y., Shapiro, L., and Hobert, O. (2008). Functional
dissection of the C. elegans cell adhesion molecule SAX-7, a homologue of
human L1. Mol. Cell. Neurosci. 37, 56–68.
Polanska, U.M., Duchesne, L., Harries, J.C., Fernig, D.G., and Kinnunen, T.K.
(2009). N-Glycosylation regulates fibroblast growth factor receptor/EGL-15
activity in Caenorhabditis elegans in vivo. J. Biol. Chem. 284, 33030–33039.
Prolo, L.M., Vogel, H., and Reimer,R.J. (2009).The lysosomal sialic acid trans-
porter sialin is required for normal CNS myelination. J. Neurosci. 29, 15355–
Qin, L., Liu, X., Sun, Q., Fan, Z., Xia, D., Ding, G., Ong, H.L., Adams, D., Gahl,
W.A., Zheng, C., et al. (2012). Sialin (SLC17A5) functions as a nitrate trans-
Reddy, L.V., Koirala, S., Sugiura, Y., Herrera, A.A., and Ko, C.P. (2003). Glial
cells maintain synaptic structure and function and promote development of
the neuromuscular junction in vivo. Neuron 40, 563–580.
Reimer, R.J. (2013). SLC17: A functionally diverse family of organic anion
transporters. Mol. Aspects Med. 34, 350–359.
Sasakura, H., Inada, H., Kuhara, A., Fusaoka, E., Takemoto, D., Takeuchi, K.,
and Mori, I. (2005). Maintenance of neuronal positions in organized ganglia by
SAX-7, a Caenorhabditis elegans homologue of L1. EMBO J. 24, 1477–1488.
Schneider, C.A., Rasband, W.S., and Eliceiri, K.W. (2012). NIH Image to
ImageJ: 25 years of image analysis. Nat. Methods 9, 671–675.
Cell 154, 337–350, July 18, 2013 ª2013 Elsevier Inc. 349
Shaham, S. (2006). Glia-neuron interactions in the nervous system of Caeno- Download full-text
rhabditis elegans. Curr. Opin. Neurobiol. 16, 522–528.
Shi, L., Fu, A.K., and Ip, N.Y. (2012). Molecular mechanisms underlying matu-
rationandmaintenanceofthevertebrate neuromuscularjunction. TrendsNeu-
rosci. 35, 441–453.
Sreedharan, S., Shaik, J.H., Olszewski, P.K., Levine, A.S., Schioth, H.B., and
Fredriksson, R. (2010). Glutamate, aspartate and nucleotide transporters in
the SLC17 family form four main phylogenetic clusters: evolution and tissue
expression. BMC Genomics 11, 17.
Stavoe, A.K., and Colo ´n-Ramos, D.A. (2012). Netrin instructs synaptic vesicle
clustering through Rac GTPase, MIG-10, and the actin cytoskeleton. J. Cell
Biol. 197, 75–88.
Stevens, B. (2008). Neuron-astrocyte signaling in the development and plas-
ticity of neural circuits. Neurosignals 16, 278–288.
Sulston, J.E., Schierenberg, E., White, J.G., and Thomson, J.N. (1983). The
embryonic cell lineage of the nematode Caenorhabditis elegans. Dev. Biol.
Sundaram, M.V. (2006). RTK/Ras/MAPK signaling. In WormBook, The
C. elegans Research Community, ed. http://dx.doi.org/10.1895/wormbook.
Tam, S.J., and Watts, R.J. (2010). Connecting vascular and nervous system
development: angiogenesis and the blood-brain barrier. Annu. Rev. Neurosci.
Verheijen, F.W., Verbeek, E., Aula, N., Beerens, C.E., Havelaar, A.C., Joosse,
M., Peltonen, L., Aula, P., Galjaard, H., van der Spek, P.J., et al. (1999). A new
gene, encoding an anion transporter, is mutated in sialic acid storage dis-
eases. Nat. Genet. 23, 462–465.
Wadsworth, W.G., Bhatt, H., and Hedgecock, E.M. (1996). Neuroglia and
pioneer neurons express UNC-6 to provide global and local netrin cues for
guiding migrations in C. elegans. Neuron 16, 35–46.
Watanabe, S., and Jorgensen, E.M. (2012). Visualizing proteins in electron
micrographs at nanometer resolution. Methods Cell Biol. 111, 283–306.
Watanabe, S., Punge, A., Hollopeter, G., Willig, K.I., Hobson, R.J., Davis,
M.W., Hell, S.W., and Jorgensen, E.M. (2011). Protein localization in electron
micrographs using fluorescence nanoscopy. Nat. Methods 8, 80–84.
White, J.G., Southgate, E., Thomson, J.N., and Brenner, S. (1986). The Struc-
ture of the Nervous System of the Nematode Caenorhabditis elegans. Philos.
Trans. R. Soc. Lond. B Biol. Sci. 314, 1–340.
Wilcox, K.C., Lacor, P.N., Pitt, J., and Klein, W.L. (2011). Abeta oligomer-
induced synapse degeneration in Alzheimer’s disease. Cell. Mol. Neurobiol.
Wolburg, H., Noell, S., Mack, A., Wolburg-Buchholz, K., and Fallier-Becker, P.
(2009). Brain endothelial cells and the glio-vascular complex. Cell Tissue Res.
Woo, W.M., Berry, E.C., Hudson, M.L., Swale, R.E., Goncharov, A., and
Chisholm, A.D. (2008). The C. elegans F-spondin family protein SPON-1 main-
tains cell adhesion in neural and non-neural tissues. Development 135, 2747–
Wreden,C.C., Wlizla,M.,andReimer,R.J.(2005).Varied mechanisms underlie
the free sialic acid storage disorders. J. Biol. Chem. 280, 1408–1416.
Yoshimura, S., Murray, J.I., Lu, Y., Waterston, R.H., and Shaham, S. (2008).
mls-2 and vab-3 Control glia development, hlh-17/Olig expression and glia-
dependent neurite extension in C. elegans. Development 135, 2263–2275.
350 Cell 154, 337–350, July 18, 2013 ª2013 Elsevier Inc.