Schwann cell chondroitin sulfate proteoglycan inhibits dorsal root ganglion neuron neurite outgrowth and substrate specificity via a soma and not a growth cone mechanism

Sensory axons do not regenerate into or within the spinal cord because of the presence of the axon regeneration inhibitor chondroitin sulfate proteoglycan (CSPG) on activated astrocytes. In the peripheral nervous system, CSPG associated with denervated Schwann cells retards axon regeneration, but regeneration occurs because the balance of regenerating, inhibiting, and promoting factors favors regeneration. The present experiments were aimed at determining the mechanism by which Schwann cells inhibit adult human dorsal root ganglia (H‐DRG) neuron growth cone elongation and substrate specificity, restricting the growth cones to Schwann cell membranes and inhibiting their growth onto a poly‐l‐lysine/laminin substrate. Neurites of H‐DRG neurons free of soma contact with Schwann cells, or after the Schwann cell membranes' CSPG had been digested, were 11.1‐fold longer than those of neurons in soma contact with untreated Schwann cells. Growth cones of DRG neuron somas without Schwann cell CSPG showed no outgrowth inhibition or substrate specificity. These results indicate that the Schwann cell CSPG influences act via contact with neuron somas but not growth cones. These results suggest that eliminating CSPG associated with Schwann cells within DRG in vivo will make the neurons' growth cones insensitive to the regeneration inhibitory influences of CSPG, allowing them to regenerate through the dorsal root entry zone and into and within the spinal cord, where they can establish appropriate and functional synaptic connections. © 2009 Wiley‐Liss, Inc.     

Neuronal age influences the response to neurite outgrowth inhibitory activity in the central and peripheral nervous systems.

Axonal regeneration is abortive in the central nervous system (CNS) of adult mammals, but readily occurs in the injured peripheral nervous system (PNS). Recent experiments indicate an important role for both intrinsic neuronal features and extrinsic substrate properties in determining the propensity for axonal regrowth. In particular, certain components of adult mammalian CNS myelin have been shown to exert a strong inhibitory influence on neurite outgrowth. To determine whether the potent neurite outgrowth inhibitory activity found in CNS myelin may also be present in PNS myelin and to study the influence of neuronal age on neurite outgrowth, we used a cryoculture assay in which dissociated rat dorsal root ganglion (DRG) neurons of different ages were challenged to extend neurites on fractionated myelin and cryostat sections from the PNS (sciatic nerve and myelin-free degenerated sciatic nerve) and CNS (optic nerve) of adult rats. The CNS environment of the optic nerve did not support E17 to P8 DRG neurite adhesion or outgrowth. E17 DRG neurons, unlike their older counterparts, however, were able to attach and extend neurites onto normal sciatic nerve and onto purified PNS myelin. In contrast, a vigorous neurite outgrowth response from all the ages tested was observed on the myelin-free degenerated sciatic nerve. These results indicate that PNS myelin is a potent inhibitor of neurite outgrowth and that DRG neuronal age plays an important role in determining the propensity for neurite outgrowth and regenerative response on inhibitory PNS and CNS substrata.     

Neurite outgrowth inhibition by chondroitin sulfate proteoglycan: stalling/stopping exceeds turning in human neuroblastoma growth cones

Chondroitin sulfate proteoglycan (CSPG) inhibits outgrowth from embryonic chick and rodent neurons in vivo and in vitro and is upregulated during development and following injury. The role of CSPG in outgrowth from human neurons has been largely untested, but is critical for our understanding of regeneration in humans following nervous system injury. Here we determined the effects of CSPG on platelet-derived growth factor (PDGF)-stimulated neurite outgrowth from SH-SY5Y human neuroblastoma cells, a well-accepted model of neuronal differentiation. Cells were plated on glass coverslips adsorbed with laminin (LN), CSPG, or a patterned substratum consisting of alternating stripes of the two molecules. Similar to other studies using chick or rodent neurons, SH-SY5Y cells extend neurites on LN, displaying a 15.2% increase in the total neurite length/cell as compared to cells plated on glass. Cells plated on CSPG alone exhibited reduced neurite outgrowth compared to cells plated on glass or LN. Interestingly, SH-SY5Y growth cones extending on LN and then encountering a CSPG border display more stopping/stalling (62.3%) than turning (27.9%) behaviors. Soluble CSPG inhibits neurite initiation from SH-SY5Y cells plated on glass, but not on LN. These data demonstrate that several CSPG-elicited responses of human neuron-like cells are similar to those from nonhuman neurons. However, approximately 70% of SH-SY5Y growth cones stop or stall at a CSPG border while over 80% of chick sensory neurons turn at a CSPG border. The experimental difference between these models may well indicate a functional difference between animal and human neuronal regeneration.     

Retinal axon regeneration in the lizard Gallotia galloti in the presence of CNS myelin and oligodendrocytes

Retinal ganglion cell (RGC) axons in lizards (reptiles) were found to regenerate after optic nerve injury. To determine whether regeneration occurs because the visual pathway has growth-supporting glia cells or whether RGC axons regrow despite the presence of neurite growth-inhibitory components, the substrate properties of lizard optic nerve myelin and of oligodendrocytes were analyzed in vitro, using rat dorsal root ganglion (DRG) neurons. In addition, the response of lizard RGC axons upon contact with rat and reptilian oligodendrocytes or with myelin proteins from the mammalian central nervous system (CNS) was monitored. Lizard optic nerve myelin inhibited extension of rat DRG neurites, and lizard oligodendrocytes elicited DRG growth cone collapse. Both effects were partially reversed by antibody IN-1 against mammalian 35/250 kD neurite growth inhibitors, and IN-1 stained myelinated fiber tracts in the lizard CNS. However, lizard RGC growth cones grew freely across oligodendrocytes from the rat and the reptilian CNS. Mammalian CNS myelin proteins reconstituted into liposomes and added to elongating lizard RGC axons caused at most a transient collapse reaction. Growth cones always recovered within an hour and regrew. Thus, lizard CNS myelin and oligodendrocytes possess nonpermissive substrate properties for DRG neurons—like corresponding structures and cells in the mammalian CNS, including mammalian-like neurite growth inhibitors. Lizard RGC axons, however, appear to be far less sensitive to these inhibitory substrate components and therefore may be able to regenerate through the visual pathway despite the presence of myelin and oligodendrocytes that block growth of DRG neurites. GLIA 22:61–74, 1998. © 1998 Wiley-Liss, Inc.     

Neurite guidance by non-neuronal cells in culture: preferential outgrowth of peripheral neurites on glial as compared to nonglial cell surfaces

Growing axons in the peripheral nervous system (PNS) encounter a variety of cellular and extracellular substrates. Since it is difficult to sort out the possible contributions of these diverse components of the extracellular environment to axonal guidance in vivo, I have developed an in vitro system to study neurite outgrowth on two classes of cells which may provide as substrates for growing axons during development or regeneration: glial cells, e.g., astrocytes and Schwann cells, and nonglial cells, e.g., fibroblasts. Although neurites from sympathetic and spinal sensory ganglia explants grew onto preformed monolayers of both glial and nonglial cells, glial cells were a markedly better substrate. On the glial cells the neurites extended at a rate of 25 to 30 micron/hr and traveled singly or in fine fascicles; their growth cones displayed long filopodia and migrated on the upper surface of the monolayer cells. Conditioned media experiments suggested that neurite outgrowth on glial cell monolayers was not mediated by soluble secreted factors. These results indicate that the glial cell surface is an attractive substrate for neurite outgrowth. In contrast, on nonglial cells the rate of outgrowth was only 10 to 15 micron/hr, large neurite fascicles were common, and the growth cones migrated beneath the monolayer cells in contact with the underlying artificial substrate. This location of the growth cone, coupled with the observation that conditioned medium from these cells promoted neurite outgrowth only when bound to artificial substrates, suggests that secreted substrate-associated components may be an important determinant of neurite outgrowth on nonglial cell monolayers.(ABSTRACT TRUNCATED AT 250 WORDS)     

Characterization of a Schwann cell neurite-promoting activity that directs motoneuron axon outgrowth

Schwann cells support and facilitate axonal growth during development and successful regeneration in the peripheral nerve. In the regenerating rat sciaticnerve, Schwann cells provide a trophic milieu for primary sensory, sympathetic, and motoneurons. We have characterized a neurotrophic activity produced by adult rat sciatic nerve Schwann cells and a spontaneously immortal Schwann cell clone (iSC). This activity elicits neurite outgrowth from chick embryo explants of both CNS and PNS. The iSC activity has been concentrated by cation-exchange chromatography and compared to known neurotrophins in bioassay. Pooled bound fractions elicit neurite outgrowth from sympathetic, ciliary and motoneurons. In collagen matrix cocultures of iSC and E4 ventral horn(before motor axon extension to muscle targets), the iSC activity can direct the initial axonal extension from motoneurons. The data presented suggest that Schwann cell-produced activity may mediate motoneuron axonal extension before contact with their peripheral source of neurotrophin. © 1994 Wiley-Liss, Inc.     

In vitro evidence for a neurite growth-promoting activity in Trembler mouse serum

Basal lamina components, such as heparan sulfate proteoglycan (HSPG) and laminin play an important role in neuritic outgrowth for CNS and PNS neurons in culture. The mutant mouse 'Trembler' is characterized by hypomyelinization and production of an excess of basal lamina layers around Schwann cells in peripheral nerves. In order to analyse whether or not the serum of the mutant animals contains neurite outgrowth-promoting factors, we cultured rat spinal cord neurons in the presence of Trembler serum. Under these conditions, the outgrowth of neurites was increased approx. 2 times as compared to control serum. Trembler serum induces neuritic outgrowth characterized both by an increase in number of primary neurites emerging from the nerve cell body as well as by an increase in peripheral branching of neurites. To characterize the factors implicated in this increase we added antibodies directed against HSPG or laminin to the mutant serum. As a result, the increase in neuritic outgrowth was reduced or abolished in both cases. Trembler effect on neurite growth disappeared when the number of the non-neuronal cells was reduced, suggesting that the mutant serum did not act directly on neurons but by the intermediary action of non-neuronal cells.     

Epac mediates cyclic AMP-dependent axon growth, guidance and regeneration

A decline in developing neuronal cAMP levels appears to render mammalian axons susceptible to growth inhibitory factors in the damaged CNS. cAMP elevation enhances axon regeneration, but the cellular mechanisms involved have yet to be fully elucidated. Epac has been identified as a signaling protein that can be activated by cAMP independently of PKA, but little is known of its expression or role in the nervous system. We report that Epac expression is developmentally regulated in the rat nervous system, and that activation of Epac promotes DRG neurite outgrowth and is as effective as cAMP elevation in promoting neurite regeneration on spinal cord tissue. Additionally, siRNA mediated knockdown of Epac reduces DRG neurite outgrowth, prevents the increased growth promoted by cAMP elevation and also diminishes the ability of embryonic neurons to grow processes on spinal cord tissue. Furthermore, we show that asymmetric activation of Epac promotes attractive growth cone turning in a similar manner to cAMP activation. We propose that Epac plays a role in mediating cAMP-dependent axon growth and guidance, and may provide an important target for inducing axon regeneration in vivo.     

GROWTH CONE BEHAVIOR IN THE PRESENCE OF SOLUBLE CHONDROITIN SULFATE PROTEOGLYCAN (CSPG), COMPARED TO BEHAVIOR ON CSPG BOUND TO LAMININ OR FIBRONECTIN

Proteoglycans (PGs) are complex macromolecules of the extracellular matrix (ECM) that have a wide variety of effects on developing and regenerating neurons in vivo and in vitro. One hypothesis regarding the mechanisms of PG regulation of neuronal behavior states that the conformation of PGs may be critical, and thus that ECM- or cell surface-bound PGs may operate differently than secreted (soluble) PGs. Therefore, this study examined differences between the effects of soluble chondroitin sulfate proteoglycan (CSPG) and substratum-bound CSPG on neuronal growth cone behavior. Dissociated chicken dorsal root ganglion (DRG) neurons were cultured on either laminin (LN) or fibronectin (FN), both sensory neurite outgrowth-promoting glycoproteins. CSPG (or chondroitin sulfate alone) was either bound to FN or LN, or was added to the culture media. Subsequently, using time lapse video microscopy and image analysis, this study measured: (1) neuronal attachment, (2) neurite outgrowth, (3) rate of neurite elongation, and (4) filopodial length and lifespan. To determine the site of CSPG action, DRG neurons were grown on either: CS-1, a FN peptide [Humphries M. J. et al. (1987) J. biol. Chem. 262, 6886–6892], or a recombinant FN protein, RFN_IIICS (Maejne, submitted), both of which permit DRG attachment and outgrowth but do not have recognized CSPG binding sites, and the resulting neuronal behavior was compared to that of DRG neurons grown on intact FN.The results of these studies confirm that the effect of CSPG on DRG neurons is concentration-, conformation- and substratum-dependent. On LN, soluble CSPG had little to no effect on neurite initiation or outgrowth, while substratum-bound CSPG inhibited neurite outgrowth. In contrast, on FN, soluble CSPG inhibited neurite outgrowth and decreased the rate of neurite elongation. Soluble CSPG did not affect the length of sensory growth cone filopodia or filopodial lifespan on either LN or FN. From the FN fragment experiments, we found that: (1) soluble CSPG reduces neurite outgrowth on FN or FN fragments, but not on LN, up to 80%, and reduces elongation rate on FN up to 50%, and (2) soluble CSPG regulates neuronal behavior by binding directly to growth cones elongating on FN.Given that substratum-bound CSPG from a variety of sources is inhibitory to neurite outgrowth and to the rate of neurite elongation, while soluble CSPG often has different effects on growth cone behavior, the regulation of growth cone behavior by CSPGs may be dependent upon CSPG conformation. Further, CSPG may affect growth cone behavior by either binding to the substratum or by binding directly to growth cones. Copyright © 1996 ISDN. Published by Elsevier Science Ltd.     

Collapse of growth cone structure on contact with specific neurites in culture

We have studied the morphology of embryonic chick retinal and sympathetic growth cones as they meet retinal and sympathetic neurites grown in culture. Growth cones preserve their normal morphology and ability to locomote when retinal growth cones contact retinal neurites or when sympathetic growth cones contact sympathetic neurites. Growth cones collapse and their motility ceases when retinal growth cones contact sympathetic neurites or when sympathetic growth cones contact retinal neurites. Collapse was never observed before a growth cone touched a neurite. As a growth cone collapses, the neurite it leads retracts. After a brief pause, a new growth cone is organized and extension recommences. These results suggest that contact-mediated inhibition of locomotion could play a role in growth cone guidance.     

Growth of embryonic retinal neurites elicited by contact with Schwann cell surfaces is blocked by antibodies to L1

Explants from embryonic rat retina plated on Schwann cell monolayers were used to examine the mechanisms by which these central neurons interact with Schwann cell surfaces. Embryonic retinal explants extend neurites reliably on Schwann cell surfaces (Kleitman et al., 1988, J. Neurosci. 8: 653). Antibodies to molecules thought to be present on Schwann cell surfaces (laminin and the 217C antigen), on retinal neurite surfaces (Thy-1.1), or on both surfaces (L1) were tested for their ability to influence this neurite growth. Of these, only antibodies to L1 were effective in blocking retinal neurite extension on Schwann cells. Inhibition of neurite growth by anti-L1 was shown to be specific to growth on Schwann cell surfaces because neurite growth on air-dried collagen (a substratum known to support retinal neurite outgrowth) was not affected. This blockage was dose-dependent. At a low titer of anti-L1 Fab fragments defasciculation of neurites was prominent; at high titers 95% of neurite outgrowth was inhibited. This virtual elimination of the ability of Schwann cell surfaces to support embryonic retinal neurite growth in the presence of antibodies to L1 indicates that binding of the L1 molecule is a critical component of the mechanism by which Schwann cells foster the growth of these neurites. The present experiments concur with the growing body of evidence that L1 plays an important role in supporting neurite growth on cell surfaces and raise the possibility that L1 may also mediate the striking ability of adult retinal axons to regenerate in a peripheral nerve environment.     

ROCK inhibition promotes adult retinal ganglion cell neurite outgrowth only in the presence of growth promoting factors

Lesioned central nervous system (CNS) axons fail to regenerate because of limited availability of neurotrophic factors (NTF) to promote neuron survival and drive axon regeneration through an environment rich in multiple myelin- and non myelin-derived axon growth inhibitory ligands that initiate growth cone collapse through the Rho/Rho kinase (ROCK) signalling pathway. However, pharmacological inhibition of Rho and ROCK promotes neurite outgrowth in PC12, Ntera-2 cells and embryonic/early postnatal neurons in culture. We have used our well-characterised CNS myelin-inhibited adult rat retinal culture model to show that Y27632 only promotes disinhibited neurite outgrowth if RGC are co-stimulated with ciliary neurotrophic factor (CNTF). Y27632 in CNTF-stimulated retinal cultures promotes optimal RGC neurite outgrowth at 10 μM concentrations, while higher concentrations negatively correlate with RGC neurite outgrowth and survival. Raising the levels of cAMP in Y27632-treated retinal cultures also promotes significant RGC neurite outgrowth, an effect that is potentiated by the further inclusion of CNTF. Our results suggest that Y27632-induced ROCK inhibition promotes robust disinhibited axon regeneration of adult neurons only when growth promoting factors are added and/or cAMP levels are raised.     

Oligodendrocytes arrest neurite growth by contact inhibition

We have used video time-lapse microscopy to analyze in vitro the interactions of growth cones of newborn rat dorsal root ganglion cells with dissociated young rat CNS glial cells present in the cultures at low density. To provide optimal conditions for neurite extension, cells were grown on laminin and in NGF-supplemented medium. Our initial observation showed that there are 2 subpopulations of growth cones differing in their growth rate on laminin (averages of 12 and 45 microns/hr). When these growth cones encountered astrocytes, they maintained their normal configuration and growth velocity. They subsequently grew along or on top of astrocytes. In some cases, however, fast-moving growth cones showed a slight reduction in their growth rate. When growth cones countered oligodendrocytes, however, firm filopodial contact was sufficient to induce a rapid and long-lasting arrest of the growth cone motility, often followed by a collapse of the growth cone structure. One third of the paralyzed growth cones were observed to retract. Growth arrest and growth cone collapse were strictly dependent on membrane contact between neurons and oligodendrocytes. This contact inhibition phenomenon was exclusively found with differentiated oligodendrocytes and could be prevented by the monoclonal antibody IN-1 directed against neurite growth inhibitors NI-35 and NI-250 (Caroni and Schwab, 1988b). These results confirm previous findings that the neurite growth inhibitor proteins are important in axon outgrowth. Further, the inhibition of neurite growth exerted by oligodendrocytes is a contact-mediated phenomenon that can be triggered by the tip of growth cone filopodia.     

Effect of chondroitin sulfate proteoglycans on neuronal cell adhesion,spreading and neurite growth in culture

As one major component of extracellular matrix(ECM) in the central nervous system, chondroitin sulfate proteoglycans(CSPGs) have long been known as inhibitors enriched in the glial scar that prevent axon regeneration after injury. Although many studies have shown that CSPGs inhibited neurite outgrowth in vitro using different types of neurons, the mechanism by which CSPGs inhibit axonal growth remains poorly understood. Using cerebellar granule neuron(CGN) culture, in this study, we evaluated the effects of different concentrations of both immobilized and soluble CSPGs on neuronal growth, including cell adhesion, spreading and neurite growth. Neurite length decreased while CSPGs concentration arised, meanwhile, a decrease in cell density accompanied by an increase in cell aggregates formation was observed. Soluble CSPGs also showed an inhibition on neurite outgrowth, but it required a higher concentration to induce cell aggregates formation than coated CSPGs. We also found that growth cone size was significantly reduced on CSPGs and neuronal cell spreading was restrained by CSPGs, attributing to an inhibition on lamellipodial extension. The effect of CSPGs on neuron adhesion was further evidenced by interference reflection microscopy(IRM) which directly demonstrated that both CGNs and cerebral cortical neurons were more loosely adherent to a CSPG substrate. These data demonstrate that CSPGs have an effect on cell adhesion and spreading in addition to neurite outgrowth.     

Influence of janusin and tenascin on growth cone behavior in vitro

Janusin and tenascin are glia-derived, structurally related, extracellular matrix glycoproteins of the J1 family that are expressed in vivo at times and in locations where active neurite outgrowth occurs, but also when the formation or stabilization of cytoarchitectonic boundaries appears to be in operation. To resolve this apparent functional dichotomy, we have studied the behavioral response of growth cones, growing in culture on the permissive substrate laminin, to janusin and tenascin, by video time lapse microscopy. When janusin and tenascin were offered as sharp substrate boundaries, dorsal root ganglion (DRG) and retinal ganglion neuron growth cones avoided growing on these molecules, but were not induced to collapse. On the other hand, when janusin and tenascin were offered, in a mixture with laminin, as uniform substrates, DRG growth cones displayed a collapsed morphology and were able to advance at a faster rate than on laminin alone. In contrast, the outgrowth of retinal ganglion neuron growth cones was completely inhibited under these conditions, underscoring a cell type specificity in the response of growth cones to these molecules. Using several monoclonal antibodies binding to distinct epitopes on the tenascin molecule, we have identified two domains responsible for growth cone repulsion, on epidermal growth factor (EGF)-like repeats 3-5 and fibronectin type III homologous repeats 4 and 5. These domains are different from the one previously recognized to be involved in neurite outgrowth on a uniform tenascin substrate. We conclude that both molecules may promote or retard growth cone advance, depending on the spatial expression pattern and the neuronal cell type. © 1993 Wiley-Liss, Inc.     

Sensory neurite growth cone guidance by substrate adsorbed nerve growth factor

The response of growth cones from embryonic chick dorsal root ganglia to a patterned substrate of adsorbed nerve growth factor (NGF) was studied. The patterned substrate presented growth cones with an adsorbed NGF pattern and NGF-free substrate. NGF-responsive growth cones from 7 and 9 day ganglia could not proceed onto NGF-free substrate, reproducing the adsorbed NGF pattern. NGF-unresponsive growth cones from 17 day ganglia did not display any preference for adsorbed NGF or NGF-free substrata, which resulted in neurites not reproducing the adsorbed NGF pattern. Neurite outgrowth from NGF responsive 7-day ganglia onto a patterned NGF substrate, in NGF-containing medium, was radially symmetrical, exhibiting no growth cone response to the patterned NGF substrate. The lack of NGF-responsive growth cone extension onto NGF-free substrate indicates that NGF is a requirement for neurite elongation. If NGF is withdrawn from growth cones by microperfusion, neurite elongation ceases. Thus, an adsorbed pattern of NGF may be duplicated because growth cones are not able to extend onto NGF-free substrate, since NGF is a requirement for neurite elongation. These results indicate that substrate adsorbed NGF can support neurite formation and elongation as well as guide the direction of neurite elongation.     

Interactions between growth cones and neurites growing from different neural tissues in culture

We have previously used retinal and sympathetic explants to show that growth cones recognize and retract from specific neurites in culture (Kapfhammer et al., 1986). In an effort to determine the generality of this phenomenon and to see how many different neurite labels can be detected by it, we have studied interactions between individual growth cones and neurites extending from a variety of neural sources in vitro. Using most of the possible pairings between sympathetic, ciliary, dorsal root ganglion (DRG), retinal, and diencephalic neurons, we have found that in most instances: (1) Growth cones do not retract from neurites originating from the same tissue; (2) retinal growth cones do not retract from diencephalic neurites; (3) sympathetic, ciliary, and DRG growth cones, with one possible exception, do not retract from sympathetic, ciliary, or DRG neurites; (4) retinal growth cones retract from sympathetic, ciliary, and DRG neurites; (5) sympathetic, ciliary, and DRG growth cones retract from retinal neurites; and (6) sympathetic growth cones retract from diencephalic neurites. A simple hypothesis consistent with these results is that 2 labels exist--one associated with central neurites and another associated with peripheral neurites--and that peripheral growth cones are programmed to retract from the central label and central growth cones are programmed to retract from sympathetic, ciliary, and DRG neurites; (5) symevant to the separation of the CNS and PNS during development.     

PC12 neurite regeneration and long-term maintenance in the absence of exogenous nerve growth factor in response to contact with Schwann cells

We have investigated mouse and rat ganglionic Schwann cells as possible sources of neurite outgrowth-promoting factors by co-culturing Schwann cells with nerve growth factor (NGF)-responsive PC12 pheochromocytoma cells primed by pretreatment with NGF. NGF-primed PC12 cells are capable of neurite regeneration when provided with an appropriate neurite promoting factor such as NGF. When primed PC12 cells were co-cultured with Schwann cells in the absence of exogenous NGF, PC12 cells that directly contacted Schwann cells became enlarged and flattened, attaining a neuron-like morphology within one day. When contact with Schwann cells was established, PC12 cells regenerated neurites by the first day of co-culture and these were maintained throughout the experiments (7 weeks). Most PC12 cells cultured in the same collagen-coated dishes with Schwann cells, but not directly in contact with them, failed to regenerate neurites. Instead, they began to proliferate, forming cell clusters. Neurite regeneration by PC12 cells in contact with Schwann cells was not blocked by antibody to NGF. These results demonstrate the presence of a neurite-promoting activity localized to the vicinity of the Schwann cell surface which is capable of eliciting regeneration and long-term maintenance of PC12 neurites in the absence of exogenous NGF. This activity does not appear to be due to NGF.