A sulfated carbohydrate epitope inhibits
axon regeneration after injury
Joshua M. Browna,1, Jiang Xiaa,1, BinQuan Zhuanga, Kin-Sang Chob, Claude J. Rogersa, Cristal I. Gamaa,
Manish Rawata, Sarah E. Tullya, Noriko Uetanic, Daniel E. Masond, Michel L. Tremblayc, Eric C. Petersd,
Osami Habuchie, Dong F. Chenb,f, and Linda C. Hsieh-Wilsona,2
aDivision of Chemistry and Chemical Engineering and Howard Hughes Medical Institute, California Institute of Technology, Pasadena, CA 91125;
bSchepens Eye Research Institute, Massachusetts Eye and Ear, Department of Ophthalmology, Harvard Medical School, Boston, MA 02114;
Cancer Center, McGill University, Montreal, Quebec H3A 1A3, Canada;
eDepartment of Chemistry, Aichi University of Education, Igaya-cho, Kariya, Aichi 448-8542, Japan; and
South Huntington Avenue, Boston, MA 02130
dGenomics Institute of the Novartis Research Foundation, San Diego, CA 92121;
fVeterans Affairs Boston Healthcare System, 150
Edited by* Peter G. Schultz, The Scripps Research Institute, La Jolla, CA, and approved January 23, 2012 (received for review December 27, 2011)
Chondroitin sulfate proteoglycans (CSPGs) represent a major bar-
rier to regenerating axons in the central nervous system (CNS), but
the structural diversity of their polysaccharides has hampered
efforts to dissect the structure-activity relationships underlying
their physiological activity. By taking advantage of our ability to
chemically synthesize specific oligosaccharides, we demonstrate
that a sugar epitope on CSPGs, chondroitin sulfate-E (CS-E),
potently inhibits axon growth. Removal of the CS-E motif signifi-
cantly attenuates the inhibitory activity of CSPGs on axon growth.
Furthermore, CS-E functions as a protein recognition element to
engage receptors including the transmembrane protein tyrosine
phosphatase PTPσ, thereby triggering downstream pathways that
inhibit axon growth. Finally, masking the CS-E motif using a CS-E-
specific antibody reversed the inhibitory activity of CSPGs and
stimulated axon regeneration in vivo. These results demonstrate
that a specific sugar epitope within chondroitin sulfate polysac-
charides can direct important physiological processes and provide
new therapeutic strategies to regenerate axons after CNS injury.
Chondroitin sulfate (CS) polysaccharides and their associated
proteoglycans (CSPGs) are the principal inhibitory components
of the glial scar, which forms after neuronal damage and acts as
a barrier to axon regeneration (1–3). It is well established that the
inhibitory activity of CSPGs is derived from their CS chains, as
chondroitinase ABC (ChABC) treatment promotes axon regen-
eration, sprouting, and functional recovery after injury in vivo
(4–6). However, the mechanisms by which CS polysaccharides
inhibit axon growth are poorly understood. Dissection of the
structural determinants and mechanisms underlying CS activity
is essential for understanding the barriers to axon regeneration
and for developing new treatments to promote regeneration
and functional recovery after spinal cord and other CNS injuries.
CS polysaccharides are composed of repeating disaccharide
units, which undergo regiochemical sulfation during development
and after injury (7–10). The CS-A (GlcA-4SGalNAc), CS-C
(GlcA-6SGalNAc), and CS-E (GlcA-4S,6SGalNAc) disacchar-
ides represent major sulfation motifs in the mammalian CNS
(Fig. S1). Although the diverse sulfation patterns of CS polysac-
charides lie at the heart of their biological activity, these complex
patterns have also hampered efforts to understand the biological
functions of CS. For example, genetic approaches are challenged
by the presence of multiple sulfotransferase isoforms with over-
lapping specificities, and deletion of a single sulfotransferase
gene can propagate global changes throughout the carbohydrate
chain. The structural complexity of CS has also thwarted bio-
chemical efforts to isolate well-defined, sulfated molecules. As
such, only heterogeneous mixtures or purified samples biased
toward abundant, readily isolable sequences have been available
for biological investigations. Although studies have suggested
major obstacle to functional recovery after CNS injury is the
inhibitory environment encountered by regenerating axons.
that the CS-A, CS-E, and CS-C motifs are upregulated after neu-
ronal injury and may play roles in axon regeneration (7, 9, 11),
only heterogeneous polysaccharides were utilized for those
studies, and there have been conflicting data, confounding the
question of whether specific sulfation sequences are important.
Indeed, because of the lack of structure-activity relationships,
relatively nonspecific mechanisms have also been proposed, such
as those brought about by steric blockage of the extracellular space
(12), arrays of negatively charged sulfate (7), or obstruction of
substrate adhesion molecules (13).
Here, we exploited chemically synthesized CS oligosaccharides
and glycopolymers to examine systematically the role of specific
sulfation sequences in axon regeneration. Our studies demon-
strate that the CS-E sulfation motif is a key structural determi-
nant responsible for the inhibitory activity of CSPGs. Moreover,
we provide mechanistic insights into how CS-E enables CSPGs
to inhibit axon growth through the identification of a specific
neuronal receptor for CS-E. Finally, we show that blocking the
inhibitory CS-E sugar motif can reverse CSPG-mediated inhibi-
tion and promote axon regeneration in vivo, providing a unique
therapeutic approach to neural regeneration.
Results and Discussion
CS-E-Enriched Polysaccharides Inhibit Neurite Outgrowth and Repel
Axons.To understand the role of specific sulfation motifs, we used
CS polysaccharides enriched in particular motifs and exploited
our ability to chemically synthesize defined CS-A, CS-C, and
CS-E oligosaccharides. First, we compared the inhibitory effects
of CSPGs andCS polysaccharides enriched in CS-A, CS-C,or CS-
E disaccharide units on neurite outgrowth. Neurite outgrowth of
dissociated dorsal root ganglion (DRG) neurons was inhibited by
58% of untreated control levels when grown on CSPGs (Fig. 1A).
ChABC digestion largely abolished the effects, confirming the
importance of the CS chains. CS polysaccharides enriched in
the CS-E motif potently inhibited neurite outgrowth to approxi-
mately 50% of untreated control levels as suggested previously
(7) and in a dose-dependent manner (Fig. 1 A and B). In contrast,
polysaccharides enriched in the CS-A or CS-C motif had no
appreciable effects on neurite outgrowth at the same glucuronic
acid concentrations. The lack of inhibition observed for CS-A and
Author contributions: J.M.B., J.X., B.Z., and L.C.H.-W. designed research; J.M.B., J.X., B.Z.,
K.-S.C., C.J.R., C.I.G., S.E.T., and D.E.M. performed research; M.R., S.E.T., N.U., M.L.T., and
O.H. contributed new reagents/analytic tools; J.M.B., J.X., B.Z., K.-S.C., C.J.R., C.I.G., D.E.M.,
E.C.P., D.F.C., and L.C.H.-W. analyzed data; and J.M.B., B.Z., D.F.C., and L.C.H.-W. wrote
The authors declare no conflict of interest.
*This Direct Submission article had a prearranged editor.
1J.M.B. and J.X. contributed equally to this work.
2To whom correspondence should be addressed. E-mail: firstname.lastname@example.org.
This article contains supporting information online at www.pnas.org/lookup/suppl/
4768–4773 ∣ PNAS ∣ March 27, 2012 ∣ vol. 109 ∣ no. 13 www.pnas.org/cgi/doi/10.1073/pnas.1121318109
CS-C, even when used at 100-fold higher concentrations than
CS-E (Fig. 1B), suggests that the inhibitory activity of CS-E poly-
saccharides is not simply due to their high overall negative charge.
Similar results were obtained with cerebellar granule neurons
(CGNs; Fig. S2), whose neurite growth is inhibited by CSPGs (14).
As CSPGs in the glial scar form an inhibitory boundary to
growing axons, we examined whether polysaccharides enriched
in the CS-E sulfation motif could repel axons in a boundary assay.
Like CSPGs (15), CS-E-enriched polysaccharides formed an in-
hibitory zone that was strongly repellent to CGN axons (Fig. 1C).
In contrast, axons freely crossed into boundaries enriched in the
CS-A or CS-C motifs. CS-A-enriched polysaccharides also exhib-
ited repulsive behavior as reported (8), but much higher concen-
trations of sugar were required (Fig. S3A).
It is known that CSPGs can acutely collapse growth cones to
form dystrophic axonal structures that no longer extend, thus
leading to long-term inhibition of regrowth (16). To examine
whether CS-E is involved in the acute phase of the inhibitory re-
sponse, we performed growth cone collapse assays. Application
of CS-E-enriched polysaccharides to DRG or CGN explants
significantly increased the number of collapsed growth cones
within minutes (Fig. 1D and Fig. S3B), whereas CS-A- and
CS-C-enriched polysaccharides had no effect. Taken together,
these results indicate that CS polysaccharides are sufficient to
recapitulate the inhibitory effects of CSPGs on neurons, and this
activity depends critically on the CS sulfation pattern.
Pure CS-E Potently Inhibits Neurite Outgrowth and Collapses Growth
Cones. Although natural polysaccharides enriched in CS-E shed
light on how specific sulfation motifs function in CSPG-mediated
axon inhibition, these data should be interpreted cautiously be-
cause polysaccharides containing a single, pure sulfation sequence
have not traditionally been isolated from natural sources, and thus
the possibility that the inhibitory activity is due to minor, contam-
inating motifs cannot be eliminated. Indeed, about 40% of the CS-
E-enriched polysaccharide contains other sulfation motifs, and
rare sulfation sequences are likely to be biologically important,
as in the case of heparan sulfate glycosaminoglycans (17). As such,
the intrinsic structural complexity and heterogeneity of CS pose a
major obstacle to understanding structure-activity relationships.
To overcome this problem, we synthesized homogeneously
sulfated glycopolymers displaying only the CS-A, CS-C, or CS-E
sulfation motifs (Fig. 2A). Norbornene-linked CS-A, CS-C, or
CS-E disaccharides were polymerized using ruthenium-catalyzed
ring-opening metathesis polymerization (ROMP) chemistry. This
approach generates glycopolymers of pure, defined sulfation se-
quence with molecular weights and biological activities compar-
able to natural CS polysaccharides (18). Previously, we showed
that these molecules were powerful tools to study the roles of
specific CS motifs in promoting neurite outgrowth of developing
hippocampal neurons (19). In the context of DRG neurons,
glycopolymers containing pure CS-E inhibited neurite outgrowth,
whereas those containing pure CS-A or CS-C had minimal activ-
ity (Fig. 2B and Fig. S4). Moreover, the monovalent CS-E disac-
charide at the same uronic acid concentration did not inhibit
neurite outgrowth, confirming that the multivalent presentation
of CS-E is critical for biological activity. Similarly, we found that
glycopolymers containing pure CS-E potently induced growth
cone collapse in DRG explants (Fig. 2C), whereas CS-A or CS-C
glycopolymers had no effect. As CS polysaccharides are found as a
complex mixture of different sulfation patterns in vivo, we also ex-
amined the activity of a glycopolymer mixture. A 1∶1 mixture of
CS-A and CS-E glycopolymers had no further effects on neurite
outgrowth compared to the pure CS-E glycopolymer alone, con-
firming that sulfated mixtures do not confer additional inhibitory
properties (Fig. S5).
P-Orn CSPG CSPG+ChABC
CS-A CS-C CS-E
0 20 4060 80 100
CS coated (µg/ml)
(% of control)
(% of control)
Axon crossing (%)
Growth cone collapse (%)
stratum of poly-DL-ornithine (P-Orn control), CSPGs, chondroitinase ABC-treated CSPGs, or CS polysaccharides enriched in the CS-A, CS-C, or CS-E sulfation motifs.
Representative images and quantitation of average neurite length (?SEM, error bars) from three experiments (n ¼ 50–200 cells per experiment). (B) Polysac-
charides enriched in the CS-E sulfation motif, but not the CS-A or CS-C motifs, inhibit DRG neurite outgrowth in a dose-dependent manner. (C) CS-E-enriched
polysaccharides repel axon crossing in a boundary assay. Polysaccharides (1 mg∕mL) or PBS control were mixed with Texas Red and spotted on P-Orn coated
coverslips. Dissociated rat P5-9 CGN neurons were immunostained with an anti-βIII-tubulin antibody. Representative images and quantitation of percentage
of axon crossing (?SEM, error bars) from two experiments (n ¼ 30–50 axons per experiment). (D) CS-E-enriched polysaccharides induce growth cone collapse.
DRG explants from chick E7-9 embryos were grown on a P-Orn/laminin substratum, treated with medium (control) or the indicated polysaccharides, and stained
with rhodamine-phalloidin. Representativeimages and quantitation of growth conecollapse(?SEM, error bars)from five experiments (n ¼ 50–100 growth cones
per experiment). Arrowsindicate collapsed growth cones.Allstatistical analyses were performed usingthe one-wayANOVA test(*P < 0.0001, relativetocontrol).
CS-E-enriched polysaccharides inhibit DRG neurite outgrowth and induce growth cone collapse. (A) Dissociated chick E7 DRGs were cultured on a sub-
Brown et al. PNAS
March 27, 2012
To complement our chemical approaches, we also investigated
the contribution of the CS-E motif using genetic methods. We
isolated CSPGs from mice containing a targeted gene disruption
of N-acetylgalactosamine 4-sulfate 6-O sulfotransferase 15
(Chst15), the enzyme that generates CS-E via addition of a sul-
fate group to the 6-O position of GalNAc on CS-A (20). Consis-
tent with potent inhibitory activity for CS-E, removal of CS-E
from CSPGs resulted in significant loss of inhibitory activity on
DRG neurite outgrowth (Fig. 2D). The remaining inhibitory
effect of CSPGs from Chst15−∕−mice is likely due to the proteo-
glycan core protein or other proteins in the mixture, as treatment
with ChABC to remove CS chains did not reduce the inhibitory
effects any further. Taken together, our chemical and genetic stu-
dies demonstrate conclusively that the CS-E motif is a potent in-
hibitor ofaxongrowthanda criticalinhibitory structure onCSPGs.
the molecular mechanisms by which CS-E inhibits axon growth,
we examined the ability of CS-E to activate signaling pathways
associated with inhibition of axon regeneration. CSPGs and
myelin inhibitors have been shown to activate Rho/Rho-kinase
(ROCK) and epidermal growth factor receptor (EGFR) path-
ways (8, 14, 15). Pharmacological inhibition of these signaling
pathways effectively reversed the inhibitory effects of CSPGs on
CGNs (Fig. 3A and Fig. S6). Specifically, the EGFR competitive
inhibitor AG1478 and the ROCK inhibitor Y27632 restored
neurite outgrowth to within 79–88% of untreated control levels,
in agreement with previous studies (14, 15). Importantly, we
found that the EGFR and ROCK inhibitors also neutralized the
inhibitory activity of CS-E polysaccharides and rescued neurite
outgrowth to a similar extent. In contrast, inhibition of c-Jun
N-terminal kinase (JNK) pathways using JNK inhibitor II showed
no effect on either CS-E- or CSPG-mediated neurite inhibition,
as expected (15). Moreover, treatment of COS-7 cells with CS-E
or CSPGs led to activation of RhoA (Fig. S7). Thus, CS-E acti-
vates intracellular signaling pathways involved in CSPG-mediated
inhibition of axon regeneration, further supporting the notion
that this sugar epitope is a major inhibitory component of CSPGs.
The CS-E Motif Inhibits Neurite Outgrowth via PTPσ. The ability of
CS-E to trigger downstream signaling pathways suggests that
CS-E may directly engage protein receptors at the cell surface,
thereby initiating intracellular signaling. Recently, CSPGs were
shown to interact with protein tyrosine phosphatase PTPσ, a
transmembrane receptor known to bind heparan sulfate proteo-
glycans (21, 22). PTPσ gene disruption reduced axon inhibition
by CSPGs in culture (22) and enhanced regeneration in sciatic,
facial, optic, and spinal cord nerves in vivo (22–25). However, it
remains unknown whether (and which) specific sulfation motifs
on CS mediate the interactions of CSPGs with PTPσ.
Growth cone collapse (%)
CS-A polymer (n=108)
CS-C polymer (n=98)
CS-E polymer (n=97)
Inhibition of neurite
(% of control)
synthetic glycopolymers displaying pure CS-A, CS-C, and CS-E disaccharides.
(B) The synthetic CS-E glycopolymer inhibits neurite outgrowth of chick E7
DRGs, whereas the CS-A glycopolymer, CS-C glycopolymer, and monovalent
CS-E disaccharide have little effect. (C) The synthetic CS-E glycopolymer
induces DRG growth cone collapse. (D) CSPGs from CS-E-deficient mice show
significant loss of inhibitory activity on DRG neurite outgrowth. Mouse P8
DRGs were cultured on CSPGs purified from Chst15 knockout or wild-type
mice. Statistical analyses were performed using the one-way ANOVA test
(*P < 0.0001, relative to control).
The CS-E motif is a potent inhibitor of axon growth. (A) Structures of
CS-C capture CS-E capture
Inhibition of neurite
CS polysaccharide (nM)
Relative PTP binding
(% of control)
CS-A CS-C CS-D
Relative PTP binding
against EGFR (AG1478, 15 nM) and ROCK (Y27632, 5 μM) rescued CS-E-
and CSPG-mediated inhibition of neurite outgrowth in dissociated rat P5-9
CGN cultures, whereas JNK inhibitor II (10 μM) had no effect. Quantitation
of neurite outgrowth from three experiments is reported. (One-way ANOVA,
*P < 0.0001, relative to CS-E control without inhibitors, **P < 0.0001, relative
to CSPG control without inhibitors; n ¼ 50–200 cells per experiment). (B) PTPσ
binds selectively to CS-E-enriched polysaccharides on glycosaminoglycan
microarrays. Microarrays were incubated with PTPσ-Fc, followed by a Cy3-
conjugated antihuman IgG secondary antibody, and analyzed using a Gene-
Pix 5000a scanner. Graphs show quantification from three experiments
(n ¼ 10 per condition). (C) Coprecipitation of CS-E and PTPσ. Full-length
PTPσ-mycHis was expressed in COS-7 cells and incubated with biotinylated
CS-E or CS-C polysaccharides bound to streptavidin beads. PTPσ binding was
detected by immunoblotting with an anti-myc antibody. (D) Specific, high
affinity binding of CS-E polysaccharides to PTPσ. (E) PTPσ -/- neurons show
significantly less inhibition by CS-E than wild-type control neurons. For each
genotype, the percentage inhibition of neurite outgrowth is plotted relative
to neurons treated with only P-Orn. Quantification from three experiments
is shown. (One-way ANOVA, *P < 0.005, relative to control; n ¼> 200 cells
The CS-E sulfation motif inhibits axon growth via PTPσ. (A) Inhibitors
www.pnas.org/cgi/doi/10.1073/pnas.1121318109Brown et al.
In light of our results showing that CS-E is a major inhibitory
motif on CSPGs, we examined the potential interaction between
CS-E and PTPσ using carbohydrate microarrays (26). A soluble
PTPσ-Fc fusion protein, but not other receptors such as EphA2-
Fc or Fc alone, bound efficiently to CS-E polysaccharides arrayed
on poly-lysine-coated glass slides (Fig. 3B and Fig. S8). PTPσ
showed strong binding to heparin and CS-E polysaccharides, with
weaker binding to chondroitin sulfate and dermatan sulfate (both
of which contain some CS-E) and heparan sulfate. Little or no
binding to CS-A, CS-C, or CS-D polysaccharides was observed,
highlighting the specificity of PTPσ for the CS-E sulfation motif.
To confirm further the PTPσ-CS-E interaction, biotinylated
CS-E or CS-C polysaccharides were conjugated to streptavidin
beads and incubated with COS-7 cell lysates expressing full-
length PTPσ. We found that CS-E polysaccharides were capable
of pulling down PTPσ, whereas CS-C polysaccharides showed no
interaction (Fig. 3C). In addition to this heterologous cell system,
we captured PTPσ from a rat brain membrane protein-enriched
fraction and identified the protein by mass spectrometry analysis
(Fig. S9). Lastly, we showed that biotinylated CS-E, but not CS-A
or CS-C, polysaccharides bind immobilized PTPσ with high
affinity according to a Langmuir binding model (Fig. 3D). The
apparent dissociation constant (KD;app) of approximately 1 nM
is similar to values reported for the association of PTPσ with
the CSPGs neurocan and aggrecan (22).
Having demonstrated that CS-E interacts specifically with
PTPσ, we next tested whether CS-E and PTPσ form a functional
association. Deletion of PTPσ significantly attenuated CS-E-in-
duced inhibition of neurite outgrowth in DRG neurons (Fig. 3E),
indicating that PTPσ is required for CS-E to inhibit neurite out-
growth. Interestingly, residual inhibition by CS-E (approximately
22%) remained in PTPσ-deficient neurons, consistent with pre-
vious observations with CSPGs (22). These results suggest that
CS-E may also engage other receptors, possibly leukocyte com-
mon antigen-related phosphatase (LAR) (27) and as-yet-undis-
covered receptors, although we cannot rule out additional
receptor-independent mechanisms, such as charge repulsion or
reduced cell adhesion. Together, these studies demonstrate that
the fine structure of CS chains mediates interactions with recep-
tors involved in axon regeneration, and they identify PTPσ as a
critical functional receptor for CS-E.
Generation of a Selective CS-E Blocking Antibody. An important im-
plication of these results is that blocking CS-E interactions may
prevent the inhibition caused by CSPGs and promote axon regen-
eration. To generate a CS-E blocking agent, we raised a mono-
clonal antibody against a pure synthetic CS-E tetrasaccharide
(28). Although antibodies have been generated previously using
CS polysaccharides as antigens (29, 30), their specificity has been
limited by the structural heterogeneity of natural polysaccharides.
Synthetic chemistry has the advantage of providing defined
molecules of precise sulfation sequence, which can be used as
antigens, for screening antibodies, and for characterizing binding
specificities. An antibody generated in this manner was highly
selective for the CS-E sulfation motif, as measured by dot blot,
ELISA, carbohydrate microarrays, and surface plasmon resonance
(Fig. 4A and Figs. S10 and S11). Strong binding to pure CS-E tet-
rasaccharides and natural CS-E polysaccharides was observed,
with minimal binding to CS-A or CS-C tetrasaccharides and other
glycosaminoglycan classes. Notably, this antibody also bound a
mixture of CSPGs derived from chick brain (Fig. 4B), confirming
the presence of the CS-E epitope on CSPGs, and blocked the in-
teraction of CS-E polysaccharides with PTPσ (Fig. S12).
CS-E Blocking Antibody Promotes Axon Regeneration. To test
whether blocking the CS-E epitope reverses the inhibitory effects
of CSPGs, we added the CS-E antibody to DRG neurons grown
on a substratum of CSPGs. Neurite inhibition by CSPGs was sig-
nificantly decreased by addition of the CS-E antibody, with neur-
ite outgrowth returning to 79% of control levels (Fig. 4C). In
contrast, neither a CS-A monoclonal antibody nor an IgG control
antibody had any effect on CSPG-mediated neurite outgrowth.
Having demonstrated specific blocking of CSPG activity in vi-
tro, we next examined whether the CS-E antibody could promote
axon regeneration in vivo. We performed an optic nerve crush
injury in mice (31), which causes focal damage and glial scarring
in the optic nerve and thus presents an ideal model for evaluating
the effects of local application of the CS-E antibody on axon
regeneration. Supporting the notion that CS-E is a prominent in-
hibitory component associated with CSPGs, pronounced upregu-
lation of CS-E was rapidly observed around the lesion site within
1 d after the injury (Fig. 5A). To examine the effects of the CS-E
antibody on axon regeneration, gelfoam soaked in a solution con-
taining the CS-E or control IgG antibody was placed around the
crush site of the nerve immediately after the injury and replaced
twice at day three and six. The extent of axonal regrowth was
assessed 2 weeks after injury by anterograde axon tracing with
choleratoxin-B subunit (CTB), which was injected intravitreally
3 d before mice were killed. Little axon regeneration was ob-
served in the control antibody-treated group. In contrast, the
CS-E antibody treatment resulted in substantial axonal regrowth,
with a sixfold increase in the number of regenerating axons when
counted at 0.25 mm beyond the injury site, as compared with
control antibody-treated mice (Fig. 5B). Notably, the extent of
axon regeneration observed after CS-E antibody treatment was
comparable to that seen in mice treated with ChABC alone
(50 U∕mL) or with ChABC and CS-E antibody applied simulta-
neously (50 U∕mL and 1.7 mg∕mL, respectively). Thus, blockade
of CS-E activity induced a similar extent of axon regeneration as
Relative Ab binding
OD (450 nm)
0 200 400600
(% of control)
mediated neurite inhibition. (A) Binding of the CS-E antibody to carbohy-
drate microarrays. Little binding to other sulfated CS polysaccharides or
glycosaminoglycan classes was detected. Experiments were performed in tri-
plicate (n ¼ 10 per condition). (B) Dose-dependent binding of the anti-CS-E
antibody to CSPGs, as shown by an enzyme-linked immunosorbent assay.
The experiment was performed in triplicate, and average values (?SD, error
bars) are shown for one representative experiment. (C) The CS-E antibody
blocks CSPG-mediated inhibition of neurite outgrowth. Dissociated chick E7
DRGs were cultured on a substratum of P-Orn (control) or CSPGs (0.5 μg∕mL)
in the presence of the indicated antibodies (0.1 mg∕mL) for 12 h. Quantita-
tion from three experiments is shown (One-way ANOVA, *P < 0.0001,
relative to CSPG without antibody treatment control; n ¼ 50–200 cells per
A monoclonal antibody binds specifically to CS-E and blocks CSPG-
Brown et al.PNAS
March 27, 2012
removal of CS chains from the CSPGs, underscoring the inhibitory
axon regrowth after CS-E antibody treatment was simply due to
improved cell survival, we stained retinal sections with an anti-
βIII-tubulin antibody to image retinal ganglion cells and counted
the number of surviving cells. No detectable increase in retinal
ganglion cell survival was found in the CS-E antibody-treated mice,
as compared with control antibody-treated mice (Fig. S13).
Remarkably, these results indicate that the complex process of
CSPG-mediated neuronal inhibition can be broken down into
discrete, active components, which when blocked are sufficient to
promote axonal regeneration in vivo.
Combining the CS-E Blocking Antibody with Other Treatments. The
failure of axons to regenerate has been attributed to inhibitory
molecules in the extrinsic environment and a reduced intrinsic
regenerative capacity of mature CNS neurons (1, 2). We there-
fore examined the ability of the CS-E antibody to enhance axon
regrowth in vivo when used in combination with 8-(4-chlorophe-
nylthio)-cyclic AMP (CPT-cAMP), a cAMP analog known to
penetrate the cell membrane and activate the intrinsic growth
state of neurons (32). In agreement with stimulation of the
growth potential of retinal ganglion cell axons, treatment of
CPT-cAMP increased the number of regenerating axons by 12-
fold compared to the control treated group (Fig. S13). Combined
delivery of the CS-E antibody and CPT-cAMP stimulated longer
axonal regrowth than either drug treatment alone, increasing the
distance of regeneration by more than 3-fold (Fig. 5C). The num-
ber of regenerating axons compared to the CPT-cAMP treatment
alone was not affected (Fig. S13), further supporting the notion
that CS-E contributes tothe environmental inhibition, but not the
intrinsic growth status, of retinal ganglion cell axons. These re-
sults demonstrate the potential of combining the CS-E antibody
to block inhibitory CSPGs in the extracellular matrix with growth-
promoting treatments to enhance the regenerative outcome.
It has long been recognized that CSPGs are one of the major in-
hibitors of axon regeneration, but until recently, the structural
determinants and mechanisms underlying their activity have been
poorly understood. In particular, the precise role of the CS sugars
and the importance of specific sulfation motifs have been unclear,
limiting the development of molecular approaches to counteract
CSPGs. Our studies identify a sugar epitope on CSPGs that is
primarily responsible for the inhibitory effects of CSPGs. We show
that the CS-E motif interacts directly with the PTPσ receptor and
activates signaling pathways involved in inhibiting axon growth.
These findings defy the conventional view that CSPGs function
primarily as a mechanical barrier to axon regrowth and that chon-
droitin sulfate sugars play nonspecific, passive roles. The ability to
upregulate particular sulfated epitopes on the sugar side chains
may be essential for regulating CSPG activity by allowing for more
precise control beyond mere expression of the core protein.
Further, the concerted expression of diverse sulfated epitopes
on different CSPGs could provide an elegant mechanism to coor-
dinate the activities of various proteoglycan core proteins.
These studies also provide a potential strategy for promoting
axon regeneration and neural plasticity after injury. We show that
CS-E blocking strategies can increase axon regeneration in vivo
and can be combined effectively with other treatments, such as
stimulation of neuronal growth, to further improve the regenera-
tive outcome. Previous studies have demonstrated that antibodies
Number of regenerating
125 250 375 500625750 875
ChABC+ CS-E Ab
Distance of axon
1 sham-operation (a), optic nerve crush injury (b), or optic nerve crush injury plus ChABC treatment (c). Note upregulation of CS-E around the injury site (b) that
was removed by ChABC treatment (c). d–g: Representative epifluorescence photomicrographs of optic nerve sections taken from mice treated with control IgG
(d),CS-E antibody(e), ChABC(f)or ChABCplus CS-E antibody (g). Asteriskindicates thecrush site. Retinal ganglioncell axons (red) are labeledby an anterograde
axon tracer, CTB, which was injected into the vitreous 3 d prior to scarify, followed by immunostaining with goat-anti-CTB antibody. In control antibody-treated
mice (d), few regenerating axons are evident. In contrast, numerous regenerating axons were seen extending pass the crush site in CS-E antibody, ChABC or the
combine-treated groups (e–g). Scale bars: 75 μm (a–g); 25 μm (d′–f′). Arrowheads indicate regenerating RGC axons. (B) Quantification of the numbers of re-
generating axons at different distances from the injury site. Nerve fibers were counted at 125-μm intervals from the crush site from three nonconsecutive
sections, and the number of fibers at a given distance was calculated (?SEM, error bars). Both the anti-CS-E-and ChABC-treated groups showed significantly
more regenerating axons as compared with the control IgG antibody-treated group (ANOVA with Bonferroni posttests at each distance, *P < 0.001 as com-
pared to controls; n ¼ 6 for each group). (C) Quantification of the distances of axon regeneration. Longest distance of axon regeneration was measured from at
least four nonconsecutive optic nerve sections from each mouse (?SEM, error bars). Combined treatment of CS-E mAb and CPT-cAMP more than tripled the
distance of axon regeneration but did not affect the number of regenerating axons compared to the anti-CS-E or CPT-cAMP treatment alone (Fig. S13).
CS-E neutralizing antibodies promote optic nerve regeneration. (A) a–c: Immunofluorescence labeling of CS-E expression in optic nerve sections at day
www.pnas.org/cgi/doi/10.1073/pnas.1121318109Brown et al.
delivered to the spinal cord can improve function after spinal Download full-text
cord injury (33, 34), and new techniques may even allow anti-
bodies into the brain for the treatment of neurodegenerative dis-
eases (35, 36). Additionally, the development of small-molecule
antagonists of CS-E function should also be feasible by inhibiting
the sulfotransferase Chst15. Targeting specific CS sugar epitopes
using antibodies, small molecules, or other approaches may offer
fewer undesirable side effects and a more stable, selective, and
less immunogenic alternative to chondroitinase ABC, which is
currently being evaluated as a therapeutic treatment for spinal
cord injury. Given that CS-E appears to interact with multiple
protein receptors and activate multiple signaling pathways, stra-
tegies that block the sulfated CS-E epitope may also prove more
effective at neutralizing CSPGs than targeting individual CSPG
receptors or pathways.
More broadly, our results demonstrate the importance of the
fine structure of CS chains in modulating the activity of CSPGs in
vivo. In contrast to heparan sulfate, where a handful of important
sequences have been identified (17), much less is known about
the roles of CS sulfation. We provide in vivo evidence that a spe-
cific CS-E sulfation motif within CS polysaccharides signals
through protein receptors so as to direct important physiological
responses. Our studies underscore the power of synthetic chem-
istry to deliver sulfated sequences with precise spacing and
orientation to assess an underappreciated component of the
mechanism. Given the importance of glycosaminoglycans in pro-
cesses ranging from development to viral invasion and spinal cord
injury, an expanded view of these sulfated sugars may provide
new insights into many critical biological processes.
For a detailed description of the materials and methods used, see SI Methods.
Neurite Outgrowth Assays. E7 chick DRGs or P5-9 rat CGNs were grown on
coverslips coated with poly-DL-ornithine, followed by CS-A, -C, -E polysacchar-
ides or glycopolymers, CSPGs, ChABC-digested CSPGs, or PBS control. For
the signaling pathway inhibitor studies, inhibitors against EGFR (AG1478,
15 nM), ROCK (Y27632, 5 μM), and JNK (JNK Inhibitor II, 10 μM) were added
in solution at the start of culturing. For inhibition studies using neurons from
PTPσ−∕−mice, DRGs from adult KO mice or WT controls were grown on la-
minin-coated plates. Neurons were immunostained with an anti-βIII tubulin
antibody, and neurite outgrowth was quantified using NIH software Image J
or MetaMorph software. Statistical analysis was performed using the one-
way ANOVA test; n ¼ 50–500 cells per experiment, and results from at least
three independent experiments were reported. Further details can be found
in SI Methods.
Optic Nerve Regeneration Assay. Immediately after crush injury in the optic
nerve of adult mice, gelfoam soaked in a solution containing control IgG,
CS-E antibody, or ChABC plus CS-E antibody was placed around the crush site.
Other groups of mice received an intravitreal injection of CPT-cAMP alone or
CPT-cAMP plus CS-E antibody. To label retinal ganglion cell axons, a solution
containing the anterograde axon tracer CTB was injected intravitreally 3 d
before mice were killed. The extent of axonal regrowth was assessed 2 weeks
after injury. For immunofluorescence labeling, sectioned optic nerve and ret-
inal tissue was immunostained with anti-CS-E, anti-CTB, or anti-βIII-tubulin
antibodies. To quantify the number of CTB-positive regenerating axons,
the number of regenerating axons was counted at 125 μm stepwise from
the crush site of the optic nerve. Further details can be found in SI Methods.
ACKNOWLEDGMENTS. We thank J. Vielmetter, Director of the Caltech Protein
Expression Center, for assistance with the surface plasmon resonance analy-
sis, I. Antoshechkin and the Caltech Millard and Muriel Jacobs Genetics and
Genomics Laboratory for printing of the polysaccharide microarrays, and P.
Clark for helpful comments on the manuscript. This research was supported
by National Institutes of Health grants R01 GM093627-03 (L.H.W.) and 5T32
GM07616-30S1 (J.M.B.), the Roman Reed Spinal Cord Injury Research Fund of
California (L.H.W.), and a Christopher & Dana Reeve Foundation Individual
Research Grant (B.Z.).
1. Yiu G, He Z (2006) Glial inhibition of CNS axon regeneration. Nat Rev Neurosci
2. Busch SA, SilverJ (2007) Therole ofextracellular matrixin CNS regeneration.Curr Opin
3. Carulli D, Laabs T, Geller HM, Fawcett JW (2005) Chondroitin sulfate proteoglycans in
neural development and regeneration. Curr Opin Neurobiol 15:116–120.
4. Pizzorusso T, et al. (2006) Structural and functional recovery from early monocular
deprivation in adult rats. Proc Natl Acad Sci USA 103:8517–8522.
5. Cafferty WB, et al. (2008) Chondroitinase ABC-mediated plasticity of spinal sensory
function. J Neurosci 28:11998–12009.
6. Bradbury EJ, et al. (2002) Chondroitinase ABC promotes functional recovery after
spinal cord injury. Nature 416:636–640.
7. Gilbert RJ, et al. (2005) CS-4,6 is differentially upregulated in glial scar and is a potent
inhibitor of neurite extension. Mol Cell Neurosci 29:545–558.
8. Wang H, et al. (2008) Chondroitin-4-sulfation negatively regulates axonal guidance
and growth. J Cell Sci 121:3083–3091.
9. Properzi F, et al. (2005) Chondroitin 6-sulphate synthesis is up-regulated in injured
CNS, induced by injury-related cytokines and enhanced in axon-growth inhibitory glia.
Eur J Neurosci 21:378–390.
10. Kitagawa H, Tsutsumi K, Tone Y, Sugahara K (1997) Developmental regulation of the
sulfation profile of chondroitin sulfate chains in the chicken embryo brain. J Biol Chem
11. Lin R, Rosahl TW, Whiting PJ, Fawcett JW, Kwok JCF (2011) 6-Sulphated chondroitins
have a positive influence on axonal regeneration. Plos One 6:e21499.
12. Olson L (2002) Medicine: clearing a path for nerve growth. Nature 416:589–590.
13. Mckeon RJ, Hoke A, Silver J (1995) Injury-induced proteoglycans inhibit the potential
for laminin-mediated axon growth on astrocytic scars. Exp Neurol 136:32–43.
14. Sivasankaran R, et al. (2004) PKC mediates inhibitory effects of myelin and chondroitin
sulfate proteoglycans on axonal regeneration. Nat Neurosci 7:261–268.
15. Kaneko M, Kubo T, Hata K, Yamaguchi A, Yamashita T (2007) Repulsion of cerebellar
granule neurons by chondroitin sulfate proteoglycans is mediated by MAPK pathway.
Neurosci Lett 423:62–67.
16. Tom VJ, Steinmetz MP, Miller JH, Doller CM, Silver J (2004) Studies on the development
and behavior of the dystrophic growth cone, the hallmark of regeneration failure, in
an in vitro model of the glial scar and after spinal cord injury. J Neurosci 24:6531–6539.
17. Bishop JR, Schuksz M, Esko JD (2007) Heparan sulphate proteoglycans fine-tune mam-
malian physiology. Nature 446:1030–1037.
18. Rawat M, Gama CI, Matson JB, Hsieh-Wilson LC (2008) Neuroactive chondroitin sulfate
glycomimetics. J Am Chem Soc 130:2959–2961.
19. Gama CI, et al. (2006) Sulfation patterns of glycosaminoglycans encode molecular re-
cognition and activity. Nat Chem Biol 2:467–473.
20. Ohtake-Niimi S, et al. (2010) Mice deficient in N-acetylgalactosamine 4-sulfate 6-o-sul-
fotransferase are unable to synthesize chondroitin/dermatan sulfate containing N-
acetylgalactosamine 4,6-bissulfate residues and exhibit decreased protease activity
in bone marrow-derived mast cells. J Biol Chem 285:20793–20805.
21. Aricescu AR, McKinnell IW, Halfter W, Stoker AW (2002) Heparan sulfate proteo-
glycans are ligands for receptor protein tyrosine phosphatase sigma. Mol Cell Biol
22. Shen YJ, et al. (2009) PTP sigma Is a Receptor for Chondroitin Sulfate Proteoglycan, an
Inhibitor of Neural Regeneration. Science 326:592–596.
23. Sapieha PS, et al. (2005) Receptor protein tyrosine phosphatase sigma inhibits axon
regrowth in the adult injured CNS. Mol Cell Neurosci 28:625–635.
24. Thompson KM, et al. (2003) Receptor protein tyrosine phosphatase sigma inhibits
axonal regeneration and the rate of axon extension. Mol Cell Neurosci 23:681–692.
25. McLean J, Batt J, Doering LC, Rotin D, Bain JR (2002) Enhanced rate of nerve regen-
eration and directional errors after sciatic nerve injury in receptor protein tyrosine
phosphatase sigma knock-out mice. J Neurosci 22:5481–5491.
26. Rogers CJ, et al. (2011) Elucidating glycosaminoglycan-protein-protein interactions
using carbohydrate microarray and computational approaches. Proc Natl Acad Sci
27. Fisher D, et al. (2011) Leukocyte common antigen-related phosphatase is a functional
receptor for chondroitin sulfate proteoglycan axon growth inhibitors. J Neurosci
28. Tully SE, Rawat M, Hsieh-Wilson LC (2006) Discovery of a TNF-alpha antagonist using
chondroitin sulfate microarrays. J Am Chem Soc 128:7740–7741.
29. Clement AM, et al. (1998) The DSD-1 carbohydrate epitope depends on sulfation,
correlates with chondroitin sulfate D motifs, and is sufficient to promote neurite out-
growth. J Biol Chem 273:28444–28453.
30. Pothacharoen P, et al. (2007) Two related but distinct chondroitin sulfate mimetope
octasaccharide sequences recognized by monoclonal antibody WF6. J Biol Chem
31. Park KK, et al. (2008) Promoting axon regeneration in the adult CNS by modulation of
the PTEN/mTOR pathway. Science 322:963–966.
32. Cui Q, Yip HK, Zhao RC, So KF, Harvey AR (2003) Intraocular elevation of cyclic AMP
potentiates ciliary neurotrophic factor-induced regeneration of adult rat retinal gang-
lion cell axons. Mol Cell Neurosci 22:49–61.
33. Geremia NM, et al. (2012) CD11d antibody treatment improves recovery in spinal
cord-injured mice. J Neurotrauma 29:539–550.
34. Liebscher T, et al. (2005) Nogo-A antibody improves regeneration and locomotion of
spinal cord-injured rats. Ann Neurol 58:706–719.
35. Yu YJ, et al. (2011) Boosting brain uptake of a therapeutic antibody by reducing its
affinity for a transcytosis target. Sci Transl Med 3:84ra44.
36. Atwal JK, et al. (2011) A therapeutic antibody targeting BACE1 inhibits amyloid-beta
production in vivo. Sci Transl Med 3:84ra43.
Brown et al. PNAS
March 27, 2012