Inhibition of axon growth by oligodendrocyte precursor cells

Two separate metalloproteinase activities are responsible for the shedding and processing of the NG2 proteoglycan in vitro

A high proportion of NG2 in the adult rat spinal cord is saline-soluble and migrates slightly faster than intact NG2 on SDS-PAGE, suggesting that it represents the shed ectodomain of NG2. In the injured cerebral cortex, much of the overall increase in NG2 is due to the saline-soluble (shed), rather than the detergent-soluble (intact), form. Hydroxamic acid metalloproteinase inhibitors, but not TIMPs, were able to prevent NG2 shedding in oligodendrocyte precursor cells (OPCs) in vitro. The generation of another truncated form of NG2 was, however, sensitive to TIMP-2 and TIMP-3. Two observations suggest that NG2 is involved in PDGF signaling in OPCs: the rate of NG2 shedding increased with cell density and NG2 expression was increased in the absence of PDGF. Ectodomain shedding converts NG2 into a diffusible entity able to interact with the growth cone, and we suggest that this release is likely to enhance its axon growth-inhibitory activity.

Matrix metalloproteases: degradation of the inhibitory environment of the transected optic nerve and the scar by regenerating axons

After injury to the central nervous system, a glial/collagen scar forms at the lesion site, which is thought to act as a physicochemical barrier to regenerating axons. We have shown that scar formation in the transected optic nerve (ON) is attenuated when robust growth of axons is stimulated. Matrix metalloproteases (MMP), modulated by tissue inhibitors of MMP (TIMP), degrade a wide variety of extracellular matrix components (ECM) and may be activated by growing axons to remodel the ECM to allow regeneration through the inhibitory environment of the glial or collagen scar. Here, we investigate whether MMP levels are modulated in a nonregenerating (scarring) versus a regenerating (nonscarring) model of ON injury in vivo. Western blotting and immunohistochemistry revealed that MMP-1, -2, and -9 levels were higher and TIMP-1 and TIMP-2 levels were lower in regenerating compared to nonregenerating ON and retinae. In situ zymography demonstrated significantly greater MMP-related gelatinase activity in the regenerating model, mainly colocalized to astrocytes in the proximal ON stump and around the lesion site. These results suggest that activation of MMP and coincident down-regulation of TIMP may act to attenuate the inhibitory scarring in the regenerating ON, thus transforming the ON into a noninhibitory pathway for axon regrowth.

Differential responses of spinal axons to transection: influence of the NG2 proteoglycan

Spinal cord transections were performed in wild type and NG2 proteoglycan null mice in order to study penetration of regenerating axons into the scar that forms in response to this type of injury. Aside from the presence or absence of NG2, the features of the transection scar did not differ between the two genotypes. In both cases, the rostral and caudal spinal cord stumps were separated by collagenous connective tissue that was continuous with the spinal cord meninges. In wild type mice, oligodendrocyte progenitors, macrophages, and microvascular pericytes contributed to up-regulation of NG2 expression in and around the scar. Substantial amounts of non-cell associated NG2 were also observed in the scar. The abilities of two classes of spinal axons to penetrate the transection scar were examined. Serotonergic efferents and calcitonin gene-related peptide-positive sensory afferents both were observed within the lesion, with calcitonin gene-related peptide-positive axons exhibiting a greater capability to penetrate deeply into the scar tissue. These observations demonstrate inherent differences in the abilities of distinct types of neurons to penetrate the scar. Significantly, growth of serotonergic axons into the transection scar was observed twice as frequently in wild type mice as in NG2 knockout mice, suggesting a stimulatory role for the proteoglycan in regeneration of these fibers. These findings run counter to in vitro evidence implicating NG2 as an inhibitor of nerve regeneration. This work therefore emphasizes the importance of including in vivo models in evaluating the responses of specific types of neurons to spinal cord injury.

Chondroitin 6-sulphate synthesis is up-regulated in injured CNS, induced by injury-related cytokines and enhanced in axon-growth inhibitory glia

Chondroitin sulphate proteoglycans (CSPGs) are up-regulated in the CNS after injury and inhibit axon regeneration mainly through their glycosaminoglycan (CS-GAG) chains. We have analysed the mRNA levels of the CS-GAG synthesizing enzymes and measured the CS-GAG disaccharide composition by chromatography and immunocytochemistry. Chondroitin 6-sulfotransferase 1 (C6ST1) is up-regulated in most glial types around cortical injuries, and its sulphated product CS-C is also selectively up-regulated. Treatment with TGFalpha and TGFbeta, which are released after brain injury, promotes the expression of C6ST1 and the synthesis of 6-sulphated CS-GAGs in primary astrocytes. Oligodendrocytes, oligodendrocyte precursors and meningeal cells are all inhibitory to axon regeneration, and all express high levels of CS-GAG, including high levels of 6-sulphated GAG. In axon growth-inhibitory Neu7 astrocytes C6ST1 and 6-sulphated GAGs are expressed at high levels, whereas in permissive A7 astrocytes they are not detectable. These results suggest that the up-regulation of CSPG after CNS injury is associated with a specific sulphation pattern on CS-GAGs, mediating the inhibitory properties of proteoglycans on axonal regeneration.

Adipocyte-derived collagen VI affects early mammary tumor progression in vivo, demonstrating a critical interaction in the tumor/stroma microenvironment

The interactions of transformed cells with the surrounding stromal cells are of importance for tumor progression and metastasis. The relevance of adipocyte-derived factors to breast cancer cell survival and growth is well established. However, it remains unknown which specific adipocyte-derived factors are most critical in this process. Collagen VI is abundantly expressed in adipocytes. Collagen(-/-) mice in the background of the mouse mammary tumor virus/polyoma virus middle T oncogene (MMTV-PyMT) mammary cancer model demonstrate dramatically reduced rates of early hyperplasia and primary tumor growth. Collagen VI promotes its growth-stimulatory and pro-survival effects in part by signaling through the NG2/chondroitin sulfate proteoglycan receptor expressed on the surface of malignant ductal epithelial cells to sequentially activate Akt and beta-catenin and stabilize cyclin D1. Levels of the carboxyterminal domain of collagen VIalpha3, a proteolytic product of the full-length molecule, are dramatically upregulated in murine and human breast cancer lesions. The same fragment exerts potent growth-stimulatory effects on MCF-7 cells in vitro. Therefore, adipocytes play a vital role in defining the ECM environment for normal and tumor-derived ductal epithelial cells and contribute significantly to tumor growth at early stages through secretion and processing of collagen VI.

An oligodendrocyte lineage-specific semaphorin, Sema5A, inhibits axon growth by retinal ganglion cells

In the mammalian CNS, glial cells repel axons during development and inhibit axon regeneration after injury. It is unknown whether the same repulsive axon guidance molecules expressed by glia and their precursors during development also play a role in inhibiting regeneration in the injured CNS. Here we investigate whether optic nerve glial cells express semaphorin family members and, if so, whether these semaphorins inhibit axon growth by retinal ganglion cells (RGCs). We show that each optic nerve glial cell type, astrocytes, oligodendrocytes, and their precursor cells, expressed a distinct complement of semaphorins. One of these, sema5A, was expressed only by purified oligodendrocytes and their precursors, but not by astrocytes, and was present in both normal and axotomized optic nerve but not in peripheral nerves. Sema5A induced collapse of RGC growth cones and inhibited RGC axon growth when presented as a substrate in vitro. To determine whether sema5A might contribute to inhibition of axon growth after injury, we studied the ability of RGCs to extend axons when cultured on postnatal day (P) 4, P8, and adult optic nerve explants and found that axon growth was strongly inhibited. Blocking sema5A using a neutralizing antibody significantly increased RGC axon growth on these optic nerve explants. These data support the hypothesis that sema5A expression by oligodendrocyte lineage cells contributes to the glial cues that inhibit CNS regeneration.

Regulating axon growth within the postnatal central nervous system

As neuronal development enters its final stages, axonal growth is restricted. Recent work indicates that several myelin-derived proteins, Nogo, MAG and OMgp, play a critical role in restricting axonal growth in the mature central nervous system (CNS). These proteins function by binding to an axonal NgR protein that limits axonal growth by activating Rho in neurons. Hypoxic conditions during the later stages of neuronal development have a prominent effect on oligodendrocytes and hence on the expression of these axon growth inhibitors. Reduced expression of these proteins caused by the developmental hypoxia, or direct blockade of the myelin inhibitor pathways in the adult CNS leads to axonal sprouting and the formation of new neuronal connections. The regulation of axonal growth, sprouting and connections in the postnatal brain by myelin proteins is an area of important investigation and potential therapeutic intervention.

Effect of Vimentin on Reactive Gliosis:In VitroandIn VivoAnalysis

Reactive gliosis is a prominent result of many types of insult to the central nervous system (CNS) and leads to the formation of glial scar that impedes the regeneration of axons. The intermediate filament protein vimentin is found in pathology of the CNS, mainly in the vicinity of injuries to the CNS. In the present study we investigated the role of vimentin in the formation of glial scars in vitro and in vivo by using immunohistochemistry, Western blot analysis, and in situ hybridization. In vitro experiments showed that the intensity of immunofluorescent labeling for vimentin and glial fibrillary acidic protein (GFAP) was consistently decreased in astrocytes after transfection with a retrovirus carrying antisense complementary DNA (cDNA) for vimentin. Transfection also inhibited the growth of astrocytes and decreased the expression of vimentin mRNA. In vivo studies demonstrated that transfection with the retrovirus carrying the antisense cDNA vimentin inhibited the upregulation of vimentin and GFAP in stab wounds in rat cerebrum. These results suggest that vimentin may play a key role in the formation of glial scars in the CNS.

Modulation of Rho GTPase activity alleviates chondroitin sulfate proteoglycan-dependent inhibition of neurite extension

The central nervous system (CNS) fails to regenerate after injury. A glial scar forms at the injury site, contributing to regenerative failure partly resulting from the chondroitin sulfate proteoglycans (CSPGs) in the glial scar. The family of Rho GTPases, which includes Cdc42, Rac1, and RhoA, is involved in growth cone dynamics. Although the response of neural cells to the inactivation of Rho when contacting myelin-related substrates, or CSPG, has been investigated, Rac1's and Cdc42's abilities to modulate CSPG-dependent inhibition have yet to be explored. In this study, a stripe assay was utilized to examine the effects of modulating all three Rho GTPases on neurite extension across inhibitory CSPG lanes. Alternating laminin (LN) and CSPG lanes were created and NG108-15 cells and E9 chick dorsal root ganglia (DRGs) were cultured on the lanes. By using the protein delivery agent Chariot, the neuronal response to exposure of constitutively active (CA) and dominant negative (DN) mutants of the Rho GTPases, along with the bacterial toxin C3, was determined by quantifying the percentage ratio of neurites crossing the CSPG lanes. CA-Cdc42, CA-Rac1, and C3 transferase significantly increased the number of neurites crossing into the CSPG lanes compared with the negative controls for both the NG108-15 cells and the E9 chick DRGs. We also show that these mutant proteins require the delivery vehicle, Chariot, to enter the neurons and affect neurite extension. Therefore, activation of Cdc42 and Rac, as well as inhibition of Rho, helps overcome the CSPG-dependent inhibition of neurite extension.

Myelin-, reactive glia-, and scar-derived CNS axon growth inhibitors: expression, receptor signaling, and correlation with axon regeneration

Axon regeneration is arrested in the injured central nervous system (CNS) by axon growth-inhibitory ligands expressed in oligodendrocytes/myelin, NG2-glia, and reactive astrocytes in the lesion and degenerating tracts, and by fibroblasts in scar tissue. Growth cone receptors (Rc) bind inhibitory ligands, activating a Rho-family GTPase intracellular signaling pathway that disrupts the actin cytoskeleton inducing growth cone collapse/repulsion. The known inhibitory ligands include the chondroitin sulfate proteoglycans (CSPG) Neurocan, Brevican, Phosphacan, Tenascin, and NG2, as either membrane-bound or secreted molecules; Ephrins expressed on astrocyte/fibroblast membranes; the myelin/oligodendrocyte-derived growth inhibitors Nogo, MAG, and OMgp; and membrane-bound semaphorins (Sema) produced by meningeal fibroblasts invading the scar. No definitive CSPG Rc have been identified, although intracellular signaling through the Rho family of G-proteins is probably common to all the inhibitory ligands. Ephrins bind to signalling Ephs. The ligand-binding Rc for all the myelin inhibitors is NgR and requires p75(NTR) for transmembrane signaling. The neuropilin (NP)/plexin (Plex) Rc complex binds Sema. Strategies for promoting axon growth after CNS injury are thwarted by the plethora of inhibitory ligands and the ligand promiscuity of some of their Rc. There is also paradoxical reciprocal expression of many of the inhibitory ligands/Rc in normal and damaged neurons, and NgR expression is restricted to a limited number of neuronal populations. All these factors, together with an incomplete understanding of the normal functions of many of these molecules in the intact CNS, presently confound interpretive acumen in regenerative studies.

Selective differentiation of neural progenitor cells by high-epitope density nanofibers

Neural progenitor cells were encapsulated in vitro within a three-dimensional network of nanofibers formed by self-assembly of peptide amphiphile molecules. The self-assembly is triggered by mixing cell suspensions in media with dilute aqueous solutions of the molecules, and cells survive the growth of the nanofibers around them. These nanofibers were designed to present to cells the neurite-promoting laminin epitope IKVAV at nearly van der Waals density. Relative to laminin or soluble peptide, the artificial nanofiber scaffold induced very rapid differentiation of cells into neurons, while discouraging the development of astrocytes. This rapid selective differentiation is linked to the amplification of bioactive epitope presentation to cells by the nanofibers.

Prevention of gliotic scar formation by NeuroGel? allows partial endogenous repair of transected cat spinal cord

Spinal cords of adult cats were transected and subsequently reconnected with the biocompatible porous poly (N-[2-hydroxypropyl] methacrylamide) hydrogel, NeuroGel. Tissue repair was examined at various time points from 6-21 months post reconstructive surgery. We examined two typical phenomena, astrogliosis and scar formation, in spines reconstructed with the gel and compared them to those from transected non-reconstructed spines. Confocal examination with double immunostaining for glial fibrillary acidic protein (GFAP) and myelin basic protein (MBP) showed that the interface formed between the hydrogel and the spine stumps did prevent scar formation and only a moderate gliosis was observed. The gel implant provided an adequate environment for growth of myelinated fibers and we saw angiogenesis within the gel. Electron microscopy showed that regenerating axons were myelinated by Schwann cells rather than oligodendrocytes. Moreover, the presence of the gel implant lead to a considerable reduction in damage to distal caudal portions of the spine as assessed by the presence of more intact myelinated fibers and a reduction of myelin degradation. Neurologic assessments of hindlimb movement at various times confirmed that spinal cord reconstruction was not only structural but also functional. We conclude that NeuroGel lead to functional recovery by providing a favorable substrate for regeneration of transected spinal cord, reducing glial scar formation and allowing angiogenesis.

The effects of treatment with antibodies to transforming growth factor β1 and β2 following spinal cord damage in the adult rat

We recently showed axonal ingrowth into fibronectin (FN) mats implanted into the spinal cord. However, little axonal growth was found from FN mats into intact spinal cord. Previous research has shown that this is due in part to astrocytosis around an area of CNS damage. Antibodies to transforming growth factor beta (TGFbeta) can diminish this astrocytosis. TGFbeta also has effects on macrophages and Schwann cells, both of which infiltrate the spinal cord following damage. We examined the axonal, Schwann cell, and macrophage infiltration into FN mats as well as the level of astrocytosis and chondroitin sulfate proteoglycan NG2 around FN implants incubated in TGFbeta antibodies and implanted into a lesion cavity in the spinal cord. We also examined the effects of applying TGFbeta antibodies to a spinal cord hemisection site. Anti-TGFbeta1 within FN mats resulted in extensive cavitation, with the area of damage being larger than the original lesion. Cavitation was also seen following application of anti-TGFbeta1 to a spinal cord hemisection site. No cavitation was seen following saline, non-immune IgG or anti-TGFbeta2 treatment. However, anti-TGFbeta2 treatment did result in diminished axonal growth and Schwann cell and macrophage infiltration. Around the implant site, anti-TGFbeta2 treatment resulted in a reduction in the level of astrocytosis but had not effect on levels of NG2. Similar effects were seen following anti-TGFbeta2 application to spinal cord hemisection sites. The results suggest that anti-TGFbeta1 exacerbates secondary damage by preventing the anti-inflammatory effect of endogenous TGFbeta1. Anti-TGFbeta2 did not enhance axonal regeneration in this model but did slightly reduce astrocytosis.

Chondroitin sulphate proteoglycans: preventing plasticity or protecting the CNS?

It is well established that axonal regeneration in the adult CNS is largely unsuccessful. Numerous axon-inhibitory molecules are now known to be present in the injured CNS, and various strategies for overcoming these obstacles and enhancing CNS regeneration have been experimentally developed. Recently, the use of chondroitinase-ABC to treat models of CNS injury in vivo has proven to be highly beneficial towards regenerating axons, by degrading the axon-inhibitory chondroitin sulphate glycosaminoglycan chains found on many proteoglycans in the astroglial scar. This enzyme has now been shown to restore synaptic plasticity in the visual cortex of adult rats by disrupting perineuronal nets, which contain high levels of chondroitin sulphate proteoglycans (CS-PGs) and are expressed postnatally around groups of certain neurons in the normal CNS. The findings suggest exciting prospects for enhancing growth and plasticity in the adult CNS; however, some protective roles of CS-PGs in the CNS have also been demonstrated. Clearly many questions concerning the mechanisms regulating expression of extracellular matrix molecules in CNS pathology remain to be answered.

The astrocyte/meningeal cell interface is a barrier to neurite outgrowth which can be overcome by manipulation of inhibitory molecules or axonal signalling pathways

Invading meningeal cells form a barrier to axon regeneration after damage to the spinal cord and other parts of the CNS, axons stopping at the interface between meningeal cells and astrocytes. Axon behavior was examined using an in vitro model of astrocyte/meningeal cell interfaces, created by plating aggregates of astrocytes and meningeal cells onto coverslips. At these interfaces growth of dorsal root ganglion axons attempting to grow from astrocytes to meningeal cells was blocked, but axons grew rapidly from meningeal cells onto astrocytes. Meningeal cells were examined for expression of axon growth inhibitory molecules, and found to express NG2, versican, and semaphorins 3A and 3C. Astrocytes express growth promoting molecules, including N-Cadherin, laminin, fibronectin, and tenascin-C. We treated cultures in various ways to attempt to promote axon growth across the inhibitory boundaries. Blockade of NG2 with antibody and blockade of neuropilin 2 but not neuropilin 1 both promoted axon growth from astrocytes to meningeal cells. Blockade of permissive molecules on astrocytes with N-Cadherin blocking peptide or anti beta-1 integrin had no effect. Manipulation of axonal signalling pathways also increased axon growth from astrocytes to meningeal cells. Increasing cAMP levels and inactivation of rho were both effective when the cultures were fixed in paraformaldehyde, demonstrating that their effect is on axons and not via effects on the glial cells.

Molecular approaches to spinal cord repair

Axon growth inhibitors associated with myelin and the glial scar contribute to the failure of axon regeneration in the injured adult mammalian central nervous system (CNS). A number of these inhibitors, their receptors, and signaling pathways have been identified. These inhibitors can now be neutralized by a variety of approaches that point to the possibility of developing new therapeutic strategies to stimulate regeneration after spinal cord injury.

Expression and glycanation of the NG2 proteoglycan in developing, adult, and damaged peripheral nerve

We have investigated expression of the axon growth-inhibitory proteoglycan NG2 in peripheral nerve. In the adult, NG2 was present on endoneurial and perineurial fibroblasts, but not on Schwann cells. At birth, peripheral nerve NG2 was heavily glycanated, but was much less so in the adult. In vitro, sciatic nerve fibroblasts also produced heavily glycanated NG2. After peripheral nerve injury in rats and humans, an accumulation of NG2-positive cells was observed at the injury site. In the rat, there was an increase in NG2 glycanation for at least 2 weeks following injury. In mixed cultures of Schwann cells and peripheral nerve fibroblasts, the axons preferred to grow on the Schwann cells and seldom crossed onto the fibroblasts. Three-dimensional cultures of sciatic nerve fibroblasts were inhibitory to the growth of dorsal root ganglion axons. Inhibition of proteoglycan synthesis made the cells more permissive. NG2 may play a part in blocking axon regeneration through scar tissue in injured human peripheral nerve.

Inhibiting cell proliferation during formation of the glial scar: effects on axon regeneration in the CNS

Following a CNS lesion many glial cell types proliferate and/or migrate to the lesion site, forming the glial scar. The majority of these cells express chondroitin sulphate proteoglycans (CS-PGs), previously shown to inhibit axonal growth. In this study, in an attempt to diminish glial scar formation and improve axonal regeneration, proliferating cells were eliminated from the lesion site. Adult rats received a continuous infusion of 2% cytosine-D-arabinofuranoside (araC) or saline for 7 days over the lesion site, immediately following a unilateral transection of the right medial forebrain bundle. Additional groups of rats that received subdural infusions prior to the lesion, and lesioned rats which received no infusion, were also compared in the analyses. Animals were killed at 4, 7, 12 or 18 days post-lesion (dpl) and immunohistochemistry was used to determine the effects of these treatments on tyrosine hydroxylase (TH)-lesioned axons, and on the injury response of glial cells. Almost complete elimination of NG2 oligodendrocyte progenitor cells from the lesion site was seen up to 7 dpl in araC-infused animals; reduced numbers of reactive CD11b microglia were also seen but no effects were seen on the injury response of GFAP astrocytes. Significantly more TH axons were seen distal to the lesion in araC-treated brains, but these numbers dwindled by 18 dpl.

Expression of 2',3'-cyclic nucleotide 3'-phosphodiesterase (CNPase) in the developing olfactory bulb and subventricular zone rostral extension

The olfactory bulb (OB) presents a unique pattern of permanent acquisition of primary afferents and interneurons, but not much detail is known about the differentiation of its oligodendroglial cells. We studied the expression of 2',3'-cyclic nucleotide 3'-phosphodiesterase (CNPase), a protein related to axonal ensheathment by myelinating cells. Expression of CNPase in OB follows a general caudorostral gradient, with the exception of the glomerular layer (GL). At postnatal day 5-6 (P5-P6), the first CNPase(+) profiles appeared in the dorsal lateral olfactory tract adjacent to the accessory OB (AOB), followed by rare cell bodies and processes in AOB internal plexiform layer at P7. At P9, the main OB (MOB) granular cell layer (GrCL) already showed intensely stained CNPase(+) processes. From P5 to P12, small numbers of CNPase(+) cells were found in the subventricular zone (SVZ), throughout its rostral extension (SVZ-RE), and in the intrabulbar subependymal layer. The appearance of CNPase(+) profiles delimiting glomeruli started in the GL rostralmost region at P12, extending to all GL levels, but glomeruli remained open caudally at P15. At P18, oligodendroglial glomeruli were evident throughout OB, but the adult pattern was established only after P30. There was no age-related loss of CNPase immunoreactivity in glial cell bodies, possibly indicating de novo ensheathment of neurites. Our results show an earlier onset of oligodendroglial differentiation in OB than previously reported and a rostrocaudal gradient of formation of oligodendroglial glomeruli. They also raise the possibility that a minor fraction of OB oligodendrocytes might derive from the SVZ-RE.

Accelerated glial reactivity to stroke in aged rats correlates with reduced functional recovery

Following cerebral ischemia, perilesional astrocytes and activated microglia form a glial scar that hinders the genesis of new axons and blood vessels in the infarcted region. Since glial reactivity is chronically augmented in the normal aging brain, the authors hypothesized that postischemic gliosis would be temporally abnormal in aged rats compared to young rats. Focal cerebral ischemia was produced by reversible occlusion of the right middle cerebral artery in 3- and 20-month-old male Sprague Dawley rats. The functional outcome was assessed in neurobehavioral tests at 3, 7, 14, and 28 days after surgery. Brain tissue was immunostained for microglia, astrocytes, oligodendrocytes, and endothelial cells. Behaviorally, aged rats were more severely impaired by stroke and showed diminished functional recovery compared with young rats. Histologically, a gradual activation of both microglia and astrocytes that peaked by days 14 to 28 with the formation of a glial scar was observed in young rats, whereas aged rats showed an accelerated astrocytic and microglial reaction that peaked during the first week after stroke. Oligodendrocytes were strongly activated at early stages of infarct development in all rats, but this activation persisted in aged rats. Therefore, the development of the glial scar was abnormally accelerated in aged rats and coincided with the stagnation of recovery in these animals. These results suggest that a temporally anomalous gliotic reaction to cerebral ischemia in aged rats leads to the premature formation of scar tissue that impedes functional recovery after stroke.

Multiple regions of the NG2 proteoglycan inhibit neurite growth and induce growth cone collapse

The NG2 chondroitin sulfate proteoglycan, an integral membrane proteoglycan, inhibits axon growth from cerebellar granule neurons and dorsal root ganglia (DRG) neurons in vitro. The extracellular domain of the NG2 core protein contains three subdomains: an N-terminal globular domain (domain 1), a central extended domain that has the sites for glycosaminoglycan (GAG) attachment (domain 2), and a juxtamembrane domain (domain 3). Here, we used domain-specific fusion proteins and antibodies to map the inhibitory activity within the NG2 core protein. Fusion proteins encoding domain 1 (D1-Fc) or domain 3 (D3-Fc) of NG2 inhibited axon growth from cerebellar granule neurons when the proteins were substrate-bound. These proteins also induced growth cone collapse from newborn DRG neurons when added to the culture medium. Domain 2 only inhibited axon growth when the GAG chains were present. Neutralizing antibodies directed against domain 1 or 3 blocked completely the inhibition from substrates coated with D1-Fc or D3-Fc. When the entire extracellular domain of NG2 was used as a substrate, however, both neutralizing antibodies were needed to reverse completely the inhibition. When NG2 was expressed on the surface of HEK293 cells, the neutralizing anti-D1 antibody was sufficient to block the inhibition, whereas the anti-D3 antibody had no effect. These results suggest that domains 1 and 3 of NG2 can inhibit neurite growth independently. These inhibitory domains may be differentially exposed depending on whether NG2 is presented as an integral membrane protein or as a secreted protein associated with the extracellular matrix.