Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. Jan 25, 2005; 102(4): 1205–1210.
Published online Jan 12, 2005. doi:  10.1073/pnas.0409026102
PMCID: PMC544342
From the Cover

Genetic deletion of the Nogo receptor does not reduce neurite inhibition in vitro or promote corticospinal tract regeneration in vivo


Axon regeneration failure in the adult mammalian CNS is attributed in part to the inhibitory nature of CNS myelin. Three myelin-associated, structurally distinct proteins, Nogo, myelin-associated glycoprotein, and oligodendrocyte myelin glycoprotein, have been implicated in this inhibition. Neuronal Nogo receptor (NgR) binds to each of the three inhibitors and has been proposed to mediate their inhibitory signals by complexing with a signal-transducing coreceptor, the neurotrophin receptor p75NTR. To assess the contribution of NgR to mediating myelin inhibitory signals and regeneration failure in vivo, we generated and characterized NgR-deficient mice. Nogo transcripts are up-regulated in NgR mutants, indicating that NgR regulates Nogo in vivo. However, neurite outgrowth from NgR-deficient postnatal dorsal root ganglion or cerebellar granule neurons is inhibited by myelin and by a Nogo-66 substrate to the same extent as is from wild-type neurons, whereas p75NTR-deficient neurons are less inhibited. The NgR ligand-binding domain promotes neurite outgrowth on Nogo-66, regardless of the genotype of the neurons, indicating that the NgR ligand-binding domain can act independent of NgR. Thus, NgR is not essential for mediating inhibitory signals from CNS myelin, at least in the neurons tested, whereas p75NTR plays a central role in this response. Neither NgR-nor p75NTR-deficient mice showed enhanced regeneration of corticospinal tract axons in comparison with wild-type controls after spinal dorsal hemisection. Our results thus fail to support a central role for NgR in axonal growth inhibition in vitro or in corticospinal tract regeneration block in vivo.

Keywords: p75, myelin inhibitors, spinal cord injury

Axons in the adult mammalian CNS have very limited capability to regenerate after injury. This inability to regenerate has been partly attributed to CNS myelin, which appears to contain inhibitory molecules that actively block regeneration. Three myelin-associated proteins, Nogo, myelin-associated glycoprotein (MAG) and oligodendrocyte myelin glycoprotein (OMgp), possess potent inhibitory activity for growing neurites in vitro, and have been proposed to contribute significantly to the inhibitory activity of CNS myelin in vivo (1). Nogo is a member of the Reticulon family of proteins with two transmembrane domains, MAG is a transmembrane protein of the Ig superfamily, and OMgp is a leucine-rich repeat (LRR) protein with a glycosylphosphatidylinositol (GPI) anchor. Although structurally distinct, in cell binding assays the three proteins can all bind to Nogo receptor (NgR), which is itself a GPI-anchored LRR protein. Interfering with the NgR/Nogo interaction with a peptide derived from Nogo has been shown to reduce inhibition by myelin and an inhibitory Nogo peptide in vitro, and to elicit regenerative axonal growth after spinal dorsal hemisection in vivo (2, 3). Furthermore, a dominant-negative NgR construct introduced into neurons, and a soluble NgR ectodomain that retains Nogo-binding properties have similarly been shown to reduce neurite inhibition in vitro (46). Based on these observations, it has been proposed that NgR might be required for mediating the inhibitory actions of Nogo, MAG, and OMgp. However, a direct test of NgR involvement by genetic ablation has not been performed, and it remains possible that the functional perturbations that have been performed to date all interfered with a distinct receptor.

Whether it is the functional Nogo receptor, NgR is GPI-anchored and unlikely to be a direct signal transducer. The neurotrophin receptor p75NTR binds to NgR and has been proposed to act as a coreceptor for NgR that provides the signal-transducing moiety in a receptor complex (4, 7). As such, the NgR/p75NTR receptor complex is suggested to be a potential converging point for all of the inhibitory cues from CNS myelin, leading to the prediction that disruption of this receptor complex should remove much, if not all, of the inhibitory signals from Nogo, MAG, and OMgp.

We have shown previously that two different Nogo mutants (a Nogo-A,B mutant and a Nogo-A,B,C mutant) that we generated display no enhanced regeneration of corticospinal tract (CST) axons after spinal cord dorsal hemisection injury (8). Two other groups have reported limited or more extensive CST regeneration in two independently generated Nogo mutants. Simonen et al. (9) reported that a small subset of their Nogo-A mutants showed a higher degree of possible regeneration/sprouting past the injury site, although, on average, there did not appear to be a statistically significant difference in this measure between the two genotypes. In contrast, Kim et al. (10) reported mild to striking regeneration in half of their Nogo-A,B mutants (and no regeneration in the rest of their mutants) (10); similar to our Nogo-A,B mutation, theirs also evidently abolished expression of Nogo-A,B. It is not clear why there is high individual variability among different mice carrying the same mutation in the latter study, nor is it known why different regenerative responses were observed in the different Nogo mutants. Nevertheless, the fact that deleting Nogo does not necessarily lead to enhanced CST regeneration in at least one mouse model is, at the very least, consistent with a potential redundancy in the myelin-derived inhibitory cues, as illustrated by the presence of not just one, but at least three, inhibitors in the CNS myelin.

Because NgR, together with p75NTR, is suggested to mediate the inhibitory signals from Nogo, MAG, and OMgp, disruption of these receptors is predicted to be more effective in releasing myelin inhibition and thereby promoting axon regeneration than removal of Nogo alone. Here, we describe the generation and characterization of a NgR-null mutant, and we test the in vitro response of neurons from these mice to inhibitory substrates, as well as the ability of CST neurons to regenerate their axons after injury. For comparison, we examine in vitro and in vivo neuronal responses of p75NTR mutant mice. Our in vitro data indicate that removing NgR, in contrast to removing p75NTR, does not reduce axonal growth inhibition, at least in the neurons tested. We also observe a lack of regeneration of CST axons after spinal injury in both the NgR and the p75NTR mutant mice. Our data thus do not provide support for a key role for NgR in regenerative responses in vitro or in vivo.


Neurite Outgrowth Assays. Neurite outgrowth assays were performed as described (8, 11, 12). Recombinant Fc-Nogo66 and NgR ligand-binding domain (LBD) proteins were produced by using baculovirus expression systems (6). All experiments were performed in duplicate wells from which the average neurite length was calculated, and repeated three times, giving similar results. Four hundred to 500 neurons per condition were analyzed.

Surgery and Care of Animals. All analyzed NgR mutants and their controls were in a (C57BL/6 × 129S7) N2 background. Female mice, 6–13 weeks old (>80% were 6–8.5 weeks old), and age-matched between genotypes, were used. The p75NTR mutation analyzed was an exon 3 deletion (13) maintained in a mixed 129/BalbC background. Female mice, 7–9 weeks old, age-matched between genotypes, were used. The surgical procedures were as described (8) with minor modifications.

Histology. The spinal cords and brains were processed as described (8). The tracer was visualized by staining with the horseradish peroxidase-based Vectastain ABC system (Vector Laboratories) with diaminobenzidine as the chromogen. Quantification of the axons was performed as described in (8, 9). Seven to eight sagittal sections centered at the main dorsal medial CST were selected for quantification, thus avoiding counting axons in the vicinity of the dorsolateral CST (14). Only those fibers running outside the main thick bundle of CST axons were counted.

More detailed descriptions can be found in Supporting Methods, which is published as supporting information on the PNAS web site.


Generation of NgR Mutant Mice. The NgR gene has two exons, with the second being the primary coding exon that is separated from the first exon by 23 kb. We constructed a targeting vector to replace most of the coding sequence in exon 2 with an IRES-Tau-LacZ reporter gene (Fig. 1A). Any mutant transcript, if translated, would encode only the first 19 aa residues of the NgR protein and the mutation was therefore expected to be a null allele. The targeted allele was obtained in AB2.2 ES cells (derived from strain 129S7; Fig. 1B). The mutation in 129S7 was bred to C57BL/6 twice before being homozygosed. Homozygous mutants were obtained in expected Mendelian ratios (26 +/+, 45 +/m, and 24 m/m of 12 litters) from intercrosses between heterozygotes, and were viable, fertile, and morphologically indistinguishable from their wild-type littermates, indicating that NgR is not essential for embryonic development. Northern blot analysis of brain tissue from adult homozygous mice confirmed the absence of NgR transcripts (Fig. 1C). We conclude that the NgR mutant is a null allele. The IRES-Tau-LacZ reporter gene expression in the NgR mutants reflects the endogenous NgR expression as assessed by RNA in situ hybridization; NgR is expressed in a variety of brain regions, including the hippocampus, dorsal thalamus, cerebellum (the granule cell layer), and the olfactory bulb (the mitral cell layer; Fig. 1 D and E), which is consistent with previous reports (15, 16). In particular, NgR is strongly expressed by neurons throughout layers 2–6 of the neocortex, including layer 5 pyramidal neurons that project axons in the CST.

Fig. 1.
Generation of NgR mutant mice. (A) Targeting strategy for the NgR mutant. IRES, internal ribosomal entry site; LacZ, β-gal gene; Neo, neomycin resistance gene. The black bars indicate exons. (B) Southern analysis of ES cell DNA to identify the ...

Neurite Outgrowth of NgR- or p75NTR-Deficient Neurons on Inhibitory Substrates. We first addressed whether NgR-deficient neurons lose response to inhibitory substrates such as CNS myelin and Nogo-66 in neurite outgrowth assays. Dissociated P7 cerebellar neurons (Fig. 2A) or P10 dorsal root ganglion (DRG) neurons (Fig. 2C) were cultured from NgR heterozygous mice and their NgR-null littermates. Neurons were plated either on control poly-d-lysine (PDL) substrate or on myelin, incubated overnight, and then neurite outgrowth was quantified. As expected, a dose-dependent inhibition of neurite outgrowth was observed when NgR-heterozygous neurons were plated on increasing concentrations of myelin. This inhibition was identical to that seen with wild-type neurons (data not shown). Surprisingly, myelin also inhibited neurite outgrowth from both cerebellar and DRG neurons cultured from NgR-null mice to a similar extent as seen in control neurons. As a control, we also examined neurons isolated from p75NTR mutant animals, because these neurons have previously been reported to lose their sensitivity to myelin (4). We observed that cerebellar neurons from p75NTR mutant mice did not appear to be any less resistant to myelin inhibition compared with controls (Fig. 2B). In contrast, inhibition by myelin of neurite outgrowth from p75NTR-deficient DRG neurons was dramatically reduced (Fig. 2D), as previously reported (4). Myelin inhibition of P7 cerebellar neurons was previously reported to be reversed in p75NTR mutant mice (4); one possible explanation for the difference may be that we use whole-myelin extract in our culture experiments, whereas Wang et al. (4) used detergent-solubilized myelin, which may be a less inhibitory substrate to begin with. These results indicate that neurons lacking NgR are still fully inhibited by myelin, at least as we prepare it, whereas at least one population of p75NTR mutant neurons are much less sensitive to myelin.

Fig. 2.
Myelin-dependent inhibition of axon outgrowth by cultured NgR-deficient neurons. Mouse P7 cerebellar neurons (A and B) and P10 DRG neurons (C and D) were dissociated and plated on PDL or myelin for 22 h, stained with an anti-tubulin antibody, and neurite ...

Because the precise constituents of myelin can vary, depending on the method of its preparation, we wished to use a more defined inhibitory substrate, and, therefore, turned to the Nogo-66 peptide. P7 cerebellar neurons cultured on recombinant Fc-Nogo66 fusion protein were inhibited in a dose-dependent fashion (Fig. 3), but we again failed to observe any release from this inhibition in neurons cultured from NgR mutant mice compared with control neurons (Fig. 3A). In contrast, P7 cerebellar neurons cultured from p75NTR mutant mice were dramatically less inhibited by Fc-Nogo66 (Fig. 3B), in agreement with previously reported data (4). We were also able to rescue neurons from inhibition by Fc-Nogo66 by precoating with an excess of the NgR LBD, which is consistent with previous data (5, 6). Importantly, this rescue worked efficiently, regardless of neuronal genotype (Fig. 3), suggesting that the NgR LBD functions by a mechanism independent of NgR, presumably by simply masking the inhibitory epitope of the Nogo-66 peptide and thereby eliminating its activity. Thus, loss of p75NTR, but not NgR, reduces inhibition by Nogo-66.

Fig. 3.
Fc-Nogo66-dependent inhibition of axon growth by cultured NgR-deficient neurons. (A and B) Mouse P7 cerebellar neurons were dissociated and plated on PDL, Fc-Nogo66, or Fc-Nogo66 combined with a 5-fold excess of NgR LBD for 22 h, stained with an anti-tubulin ...

Lack of Detectable Regeneration in the CST of NgR-Deficient Mice. Because NgR is highly expressed in layer 5 pyramidal neurons in the adult cerebral cortex (see above), the temporal and spatial expression of NgR is compatible with a role in mediating responses of corticospinal neurons to inhibitory cues after injury. To address the in vivo role of NgR in regeneration failure, we therefore examined CST axon regeneration in the NgR mutant by using a dorsal hemisection model of spinal cord injury, as described (8). Animals were subjected to T8 dorsal hemisection, their CST axons were simultaneously traced by biotinylated dextran amine injection into the right sensorimotor cortex, and the animals were allowed to survive for 2–3 weeks (see Methods for details). An 0.8-cm segment of the spinal cord centered around the injury site was sectioned in the sagittal plane to follow severed axons around the injury site; a 2- to 3-mm segment immediately rostral or caudal to this 0.8-cm block was sectioned in the transverse plane to detect any traced CST axons above and below the injury. In total, we examined and compared the regenerative response of CST fibers in 12 NgR mutants and 10 wild-type controls after dorsal hemisection.

No obvious differences in the regenerative responses of the CST of wild-type and NgR mutant mice were observed. Representative sections from a wild-type animal and from a NgR mutant animal are shown in Fig. 4A. On transverse sections ≈0.5 cm above injury, we detected labeling of the main dorsal CST (thousands of axons), the dorsolateral CST (one to a few dozen axons), and extensive collaterals in the gray matter in both genotypes. There was no obvious defect in the CST of the NgR mutants. On transverse sections ≈0.5 cm below injury, there were either no axons or few axons in the gray matter. In several cases, including both wild-type and NgR mutants, a small number of dorsolateral CST fibers were found on transverse sections below injury, indicating some sparing of the dorsolateral CST (except for one mutant that showed a more-than-usual degree of sparing, see below). Such sparing has also been observed by others in a similar injury model (9), and the degree of sparing does not appear to differ significantly between the genotypes. Overall, the pattern of traced CST axons both above and below the injury site is comparable between wild-type and NgR mutant animals. These results indicate that deleting NgR did not elicit a striking regenerative/sprouting response above or below injury, as assessed on transverse sections.

Fig. 4.
Lack of detectable CST regeneration in NgR mutants. (A) Representative spinal cord sections from wild-type and NgR mutant mice. (Top and Middle) Transverse sections, ≈0.5 cm rostral and ≈0.5 cm caudal to the injury site are shown and the ...

To follow the behavior of severed axons more closely and to detect any milder regenerative response in the mutants, we analyzed sagittal sections surrounding the injury site and quantified the traced axons caudal to the lesion compared with those rostral to the lesion. Representative images of traced axons on sagittal sections at the plane of the main dorsal CST are shown in Fig. 4A Bottom.In both wild-type and NgR mutant mice, the main CST axons stopped just before the injury site, and either none or few axons passed the lesion site. As described previously, we manually counted axons at five fixed locations (0.25, 0.5, 1, 2, and 3 mm or beyond) caudal to the injury on seven to eight sagittal sections surrounding the main CST, summed these numbers up, and normalized to the number of axons outside of the main CST 0.5 mm rostral to the end of the severed CST (which ends in retraction bulbs) from the same sagittal sections to obtain the caudal/rostral axon ratio. As shown in Fig. 4B, there is no statistically significant difference between the two genotypes in regenerative response caudal to the injury, as assessed by this measure. One NgR mutant animal showed an unusually high ratio of caudal/rostral axon counts. Further examination of this mutant indicated that there was a more-than-usual degree of sparing of the dorsolateral CST, and the plane of section was significantly off the sagittal plane; both of these effects might have contributed to the unusually high ratio. We thus do not believe that this animal represents a true regenerative response. Instead, it appears that, overall, the severed CST axons of NgR mutants behave similarly to those of wild-type animals, with no detectable enhancement of regeneration in the CST.

Lack of Detectable Regeneration in the CST of p75NTR-Deficient Mice. Our in vitro data indicate that it is possible to dissociate p75NTR from NgR in mediating neuronal response to neurite outgrowth inhibitors, such that p75NTR appears to play a central role in mediating the inhibitory response of postnatal cerebellar and DRG neurons, whereas NgR does not. These data would suggest that disrupting p75NTR may elicit a more extensive regenerative response in vivo. We therefore analyzed CST regeneration in p75NTR-deficient mice in parallel to our study on the NgR-deficient mice. In total, we examined and compared the regenerative response of CST fibers in eight p75NTR mutants and seven wild-type controls after dorsal hemisection. Representative images of transverse sections rostral and caudal to the injury and sagittal sections around the injury site are shown in Fig. 5A. The pattern of traced axons on both transverse sections and sagittal sections are comparable between p75NTR-deficient mice and wild-type controls. Quantification of CST fibers indicated that the ratio of traced axons below the lesion to non-main CST axons rostral to the lesion (as described for analysis of NgR mutants) is not significantly different between the two genotypes. These data indicate that, despite its importance in mediating the inhibitory signals for multiple neuronal types in vitro, disrupting p75NTR is insufficient to elicit detectable enhancement of regeneration of the CST axons. Similar results were reported by Song et al. (17).

Fig. 5.
Lack of detectable CST regeneration in p75NTR mutants. (A) Representative spinal cord sections from wild-type and p75NTR mutant mice presented as in Fig. 4A.(B) Ratio of caudal axon counts over rostral axon counts as in Fig. 4B. (Error bar, SEM.) No statistically ...


The identification of NgR as a candidate receptor for three myelin-associated proteins (Nogo, MAG, and OMgp) and p75NTR as a coreceptor has suggested that a NgR/p75NTR receptor complex may represent a potential converging point for multiple inhibitory signals from the CNS myelin to severed axons attempting to regenerate. This hypothesis predicts that disrupting the NgR/p75NTR receptor complex should lead to an enhanced regenerative response by CNS axons. Here. we applied in vitro neurite outgrowth assays and an in vivo CST axon regeneration model to test this hypothesis.

We present two major findings that contradict this hypothesis. First, we found that although p75NTR-deficient neurons exhibit a reduced propensity to be inhibited by CNS myelin and Nogo-66 (in two of three assays), NgR-deficient neurons are still strongly inhibited by either CNS myelin or Nogo-66, at least for the neuronal populations tested here (postnatal cerebellar and DRG neurons). Second, we found no enhanced regeneration of CST axons after a dorsal hemisection injury in either NgR- or p75NTR-deficient mice. These results challenge the view that NgR plays a central role in mediating the inhibitory signals from myelin, and indicate that blocking additional inhibitory influences and/or providing growth-stimulating signals is required to effectively promote axon regeneration in the CNS. Here, we discuss the significance of these results in the context of published reports.

We first analyzed the ability of NgR- or p75NTR-deficient neurons to grow on inhibitory substrates. If NgR binds and mediates inhibitory signals from Nogo, MAG, and OMgp, deleting NgR from neurons is expected to release part or all of the inhibition by CNS myelin and by Nogo-66. To our surprise, NgR-deficient postnatal cerebellar and DRG neurons are still inhibited by CNS myelin and Nogo-66 to the same extent as are the wild-type controls; no reduction in inhibition was observed under any of the circumstances tested. In contrast, p75NTR-deficient cerebellar neurons exhibit a diminished response to Nogo-66, which is consistent with a previous report (4), and DRG but not cerebellar neurons also showed reduced inhibition by myelin. As discussed, we assume that our failure to observe reduced inhibition of cerebellar neurons by myelin, contrary to what Wang et al. (4) observed, may reflect differences in the manner of preparation of the myelin. Whatever the explanation, the fact that strikingly different results are obtained for p75NTR versus NgR mutant neurons indicates that the contributions of these two putative receptors to mediating the response to inhibitory signals can be dissociated: p75NTR appears to play a central role in mediating the response to inhibitory signals in the neurons tested, whereas NgR does not.

Our test of NgR involvement is more stringent than that provided by previous experiments that had been used to argue for its involvement. On the loss-of-function side, enzymatic cleavage of GPI-linked proteins from cell surfaces was shown to render otherwise sensitive neurons insensitive to Nogo-66 inhibition (18, 19), but this manipulation would be expected to remove all GPI-linked proteins, not just NgR, and therefore does not prove the specific involvement of NgR. Similarly, two antagonists of the Nogo66/NgR interaction, a soluble NgR ectodomain and a peptide, were shown to reduce inhibition by both Nogo66 and myelin (2, 5), and we have found a similar effect of the LBD of NgR in our experiments. However, these manipulations do not directly test the involvement of NgR in mediating the response, because the antagonists could interfere with the substrates' ability to bind a variety of ligands and/or receptors. This possibility is underscored by our finding that a NgR LBD reverses the inhibitory effect of Nogo-66, even on neurons lacking NgR. In contrast, using neurons from NgR-null mutant mice provides a definitive test of the involvement of NgR, and we can therefore conclude that it is not required for responses of postnatal DRG and cerebellar granule neurons to Nogo-66, or to myelin (at least as we prepare it).

One possible reason for the lack of release of inhibitory response by NgR-deficient neurons could be that NgR may not be normally expressed in the postnatal neurons tested here. Indeed, when we examined the expression of NgR transcripts by RNA in situ hybridization in P7 cerebellum (the stage when the cells were harvested and used for the neurite outgrowth assay), we detected only background levels of NgR mRNA, whereas in the adult, we observed high-level expression in cerebellar granule neurons (Fig. 6A, which is published as supporting information on the PNAS web site), which is consistent with previous reports (15, 16). We did detect a low level of NgR expression in dissociated postnatal cerebellar neurons by RT-PCR, but this expression remained at a low level after 1 day in culture, and the level is much lower than that in adult brain (Fig. 6B). Nevertheless, because P7 cerebellar neurons are sensitive to CNS myelin and Nogo-66 inhibition, and because p75NTR is not known to directly interact with the inhibitory ligands, the inhibitory response presumably must be mediated by receptors other than NgR. Such a role might in principle be fulfilled by a homologue of NgR, of which there are two (NgR2 and NgR3) with significant sequence homology. We observed that NgR2, but not NgR3, is highly expressed in both cerebellar granule and Purkinje neurons at P7 and in adult mice (Fig. 6A). However, when we used an alkaline phosphatase linked Nogo-66 peptide (AP-Nogo-66) in a binding assay with COS cells expressing NgR or its homologues, we found that neither NgR2 nor NgR3 binds to AP-Nogo-66 (Fig. 7, which is published as supporting information on the PNAS web site), whereas NgR does, which is consistent with a previous report (20). Furthermore, it has been reported that NgR2 and NgR3 do not bind MAG or OMgp, either (20). Taken together, these results argue against, but do not rule out completely, the possibility that NgR2 and NgR3 are involved in mediating the inhibitory signals from Nogo, MAG, or OMgp. The identity of the necessary binding receptor(s) on P7 cerebellar granule neurons and DRG neurons that mediates responses to myelin and Fc-Nogo-66 thus remains to be determined, but our results specifically rule out an essential role for NgR itself.

We also examined whether deletion of either NgR or p75NTR leads to enhanced regeneration of the CST by using a mouse model of spinal cord injury. Although p75NTR appears to play a central role in mediating myelin inhibitory signals (see above), loss of p75NTR was not necessarily expected to enhance regeneration, because p75NTR is reportedly not expressed by CST axons (17). Indeed, we did not detect any enhanced regeneration of the CST after a dorsal hemisection injury to the thoracic spinal cord in p75NTR mutant mice, which is consistent with recent results by Song et al. (17). In contrast, it might have been expected that loss of NgR would enhance CST regeneration, because NgR is expressed by CST neurons. However, we failed to observe enhanced CST regeneration in the NgR mutant mice. Consistent with the lack of CST regeneration in NgR mutant mice was a lack of enhanced functional recovery as assessed by the Basso–Beattie–Bresnahan open field scale (data not shown).

The failure of CST regeneration in the NgR mutant contrasts with the CST regeneration that is observed in wild-type animals treated with a peptide antagonist of the Nogo66/NgR interaction (2, 3). There are two possible explanations for this discrepancy. First, the peptide might be acting by interfering with NgR function but be more effective than deletion of NgR, because it is provided acutely, whereas in the NgR mutant, there could be compensatory changes in gene expression that could block regeneration. Lending credence to the possibility of compensatory changes is our observation that all of the three Nogo transcripts are up-regulated in the adult brain of the NgR mutant; however, we also find that this does not lead to a detectable change in the levels of Nogo-A protein, as assessed by Western blotting, nor does it lead to an elevated expression of MAG or OMgp (Fig. 8, which is published as supporting information on the PNAS web site). Second, the peptide might produce its effect by interfering with something other than NgR, and specific interference with NgR might be insufficient to permit regeneration, just as it is insufficient to permit postnatal DRG and cerebellar regeneration on inhibitory substrates in vitro. Fischer et al. (21) have recently shown that expression of a dominant-negative fragment of NgR by itself does not lead to enhanced regeneration of optic nerve axons, but it does lead to enhanced regeneration of optic nerve axons in combination with a conditional injury. Although this result is consistent with NgR contributing to regeneration failure, as in the peptide experiment, this experiment does not test directly the involvement of NgR, because the dominant-negative has the potential to interfere with other inhibitory ligands/receptors.

In summary, our data fail to provide any evidence of NgR involvement in mediating growth inhibition by myelin-derived inhibitors in vitro, or in blocking CST regeneration in vivo. As such, they are consistent with our previous finding that disrupting Nogo itself is insufficient to enhance CST regeneration. Taken together, these studies indicate that any role of Nogo/NgR in regeneration failure is more limited than previously envisioned. Future studies will help clarify whether disrupting the NgR/Nogo pathway is sufficient to enhance regeneration of other axonal tracts, and whether combining Nogo/NgR disruption with other manipulations will lead to enhanced regeneration of the CST.

Supplementary Material

Supporting Information:


We thank Kelly Yee, Han Lin, Nona Velarde, and the Roman Reed Core Laboratory for expert technical assistance. This work was supported by a grant from the International Spinal Research Trust (to M.T.-L.), a postdoctoral fellowship from the Helen Hay Whitney Foundation (to B.Z.), a Human Frontier Science Program postdoctoral fellowship (to J.A.), the Roman Reed Spinal Cord Injury Research Fund of California, Research for Cure (to O.S.), individual donations to the Reeve–Irvine Research Center, and the Howard Hughes Medical Institute (to M.T.-L.).


Author contributions: B.Z., J.A., and M.T.-L. designed research; B.Z., J.A., C.H., L.C., and O.S. performed research; B.Z., J.A., X.-L.H., and K.C.G. contributed new reagents/analytic tools; B.Z., J.A., C.H., L.C., O.S., and M.T.-L. analyzed data; and B.Z., J.A., and M.T.-L. wrote the paper.

Abbreviations: NgR, Nogo receptor; MAG, myelin-associated glycoprotein; OMgp, oligodendrocyte myelin glycoprotein; GPI, glycosylphosphatidylinositol; DRG, dorsal root ganglion; PDL, poly-d-lysine; LBD, ligand-binding domain.

Note. While this work was under review, Strittmatter and colleagues (22) published results on the analysis of a NgR mutant mouse they generated. Consistent with our results, they did not detect any enhanced regeneration of the CST in their mutants (22). They did report, however, enhanced regeneration of other tracts, most notably, the raphe spinal fibers; whether this result reflects a reduction in inhibition of such fibers by myelin inhibitors, or some other change in these mice, remains to be determined. Considering that NgR is thought to mediate the inhibitory signals not only from Nogo-66 but also MAG and OMgp, it is not clear why CST axons fail to regenerate in NgR mutants, although they appear to do so in at least some Nogo-deficient mutant lines (10). Kim et al. (22) also reported that NgR-deficient DRG growth cones do not collapse in response to Nogo-66 and other myelin inhibitors. In these experiments, the inhibitors caused a modest increase in the collapse of wild-type growth cones from ≈40% (seen in control cultures) to ≈60%, an increase that was not seen with their NgR knockout neurons. However, it is unclear whether this collapse was sustained, that is, whether these neurons are inhibited in longer-term cultures by an immobilized myelin or Nogo-66 substrate, which is the assay used previously by the same group (2, 5) to suggest a role for NgR in mediating inhibition, and also the assay used here that showed no change in inhibition after NgR deletion.


1. McKerracher, L. & Winton, M. J. (2002) Neuron 36, 345–348. [PubMed]
2. GrandPre, T., Li, S. & Strittmatter, S. M. (2002) Nature 417, 547–551. [PubMed]
3. Li, S. & Strittmatter, S. M. (2003) J. Neurosci. 23, 4219–4227. [PubMed]
4. Wang, K. C., Kim, J. A., Sivasankaran, R., Segal, R. & He, Z. (2002) Nature 420, 74–78. [PubMed]
5. Fournier, A. E., Gould, G. C., Liu, B. P. & Strittmatter, S. M. (2002) J. Neurosci. 22, 8876–8883. [PubMed]
6. He, X. L., Bazan, J. F., McDermott, G., Park, J. B., Wang, K., Tessier-Lavigne, M., He, Z. & Garcia, K. C. (2003) Neuron 38, 177–185. [PubMed]
7. Wong, S. T., Henley, J. R., Kanning, K. C., Huang, K. H., Bothwell, M. & Poo, M. M. (2002) Nat. Neurosci. 5, 1302–1308. [PubMed]
8. Zheng, B., Ho, C., Li, S., Keirstead, H., Steward, O. & Tessier-Lavigne, M. (2003) Neuron 38, 213–224. [PubMed]
9. Simonen, M., Pedersen, V., Weinmann, O., Schnell, L., Buss, A., Ledermann, B., Christ, F., Sansig, G., van der Putten, H. & Schwab, M. E. (2003) Neuron 38, 201–211. [PubMed]
10. Kim, J. E., Li, S., GrandPre, T., Qiu, D. & Strittmatter, S. M. (2003) Neuron 38, 187–199. [PubMed]
11. Wang, K. C., Koprivica, V., Kim, J. A., Sivasankaran, R., Guo, Y., Neve, R. L. & He, Z. (2002) Nature 417, 941–944. [PubMed]
12. Neumann, S., Bradke, F., Tessier-Lavigne, M. & Basbaum, A. I. (2002) Neuron 34, 885–893. [PubMed]
13. Lee, K. F., Li, E., Huber, L. J., Landis, S. C., Sharpe, A. H., Chao, M. V. & Jaenisch, R. (1992) Cell 69, 737–749. [PubMed]
14. Steward, O., Zheng, B., Ho, C., Anderson, K. & Tessier-Lavigne, M. (2004) J. Comp. Neurol. 472, 463–477. [PubMed]
15. Hunt, D., Mason, M. R., Campbell, G., Coffin, R. & Anderson, P. N. (2002) Mol. Cell. Neurosci. 20, 537–552. [PubMed]
16. Josephson, A., Trifunovski, A., Widmer, H. R., Widenfalk, J., Olson, L. & Spenger, C. (2002) J. Comp. Neurol. 453, 292–304. [PubMed]
17. Song, X. Y., Zhong, J. H., Wang, X. & Zhou, X. F. (2004) J. Neurosci. 24, 542–546. [PubMed]
18. Niederost, B., Oertle, T., Fritsche, J., McKinney, R. A. & Bandtlow, C. E. (2002) J. Neurosci. 22, 10368–10376. [PubMed]
19. Fournier, A. E., GrandPre, T. & Strittmatter, S. M. (2001) Nature 409, 341–346. [PubMed]
20. Barton, W. A., Liu, B. P., Tzvetkova, D., Jeffrey, P. D., Fournier, A. E., Sah, D., Cate, R., Strittmatter, S. M. & Nikolov, D. B. (2003) EMBO J. 22, 3291–3302. [PMC free article] [PubMed]
21. Fischer, D., He, Z. & Benowitz, L. I. (2004) J. Neurosci. 24, 1646–1651. [PubMed]
22. Kim, J. E., Liu, B. P., Park, J. H. & Strittmatter, S. M. (2004) Neuron 44, 439–451. [PubMed]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...