The Nogo-66 Receptor NgR1 Is Required Only for the Acute Growth Cone-Collapsing But Not the Chronic Growth-Inhibitory Actions of Myelin Inhibitors

Discussion

The Nogo-66 receptor NgR1 supports high-affinity binding of three potent myelin-associated inhibitors of neurite outgrowth, (namely, Nogo-A, OMgp, and MAG) and neuronal NgR1 has been proposed to serve as an important convergence point for inhibitory responses toward each of its ligands. Recently, two independent groups used a mouse genetic approach to more directly test the hypothesis that NgR1 is essential for inhibitory responses toward its ligands (Kim et al., 2004; Zheng et al., 2005). At face value, the two studies seemed to come to opposite conclusions, but on closer examination there were important differences in the way the two studies were performed. Specifically, Kim et al. (2004) studied the modest growth cone-collapsing activity of NgR1 ligands when applied acutely to DRG neurons, and found that this required NgR1 function because it was not observed on DRG neurons from NgR1 knock-out mice. In contrast, Zheng et al. (2005) examined the axon growth-inhibitory activity of Nogo-66 and CNS myelin when presented as substrates for CGN and DRG neurons in chronic culture experiments, and found that NgR1 was dispensable for this activity, thereby questioning the importance of NgR1 in myelin inhibition. Here, we reconcile these two studies by showing that NgR1 is required for acute collapsing effects of MAG and OMgp, but not for their ability to inhibit neurite outgrowth when presented as substrates.

Recent evidence suggests that different neuronal cell types use different mechanisms for MAG inhibition (Venkatesh et al., 2007); we therefore examined the contribution of NgR1 in MAG-mediated outgrowth inhibition in CGNs, DRGs, and cortical neurons. These are cell types commonly used to assay for CNS myelin inhibition; they all express NgR1 and include myelinated and nonmyelinated neurons. Each of these neuronal cell types isolated from wild-type mice is strongly inhibited by membrane-bound MAG. Furthermore, we found that neurite outgrowth of NgR1-deficient neurons is strongly inhibited on MAG and to a similar extent as for wild-type neurons. Thus, in three different neuronal cell types, loss of NgR1 does not reduce inhibition by membrane-bound MAG. In a similar vein, substrate-bound Nogo-66 and OMgp strongly inhibit neurite outgrowth of NgR1^−/− DRG neurons.

Although neither p75^NTR nor NgR2 expression was altered in NgR1-null brain tissue, it remains possible that germline ablation of NgR1 leads to compensatory changes in gene expression that may influence signaling pathways regulating neuronal growth behavior. We therefore wanted to independently replicate our findings with NgR1-null neurons by acute shRNAi-mediated gene knock-down. Consistent with our studies with different populations of primary neurons from NgR1-null mice, we found no disinhibition of neurite outgrowth after shRNAi-mediated knock-down of NgR1 in cortical neurons cultured on CHO-MAG feeder cells. Thus, two different strategies used to functionally ablate NgR1, one chronic and one acute, failed to demonstrate any significant role for NgR1 in MAG-mediated neurite outgrowth inhibition. Together with the previous finding that NgR1^−/− neurons are not disinhibited on Nogo-66 substrate (Zheng et al., 2005), these studies show that NgR1 is not important for Nogo-66-, OMgp-, or MAG-mediated neurite outgrowth inhibition in vitro.

An independent study by Ahmed et al. (2005), using an RNAi-based approach to functionally deplete NgR1 in adult rat DRG neurons, found a significant release of neurite outgrowth inhibition in the presence of crude CNS myelin. We did not see a similar effect after acute shRNAi-mediated knock-down of NgR1 in primary cortical neurons cultured on MAG (Fig. 3). Although different neuronal populations may use different mechanisms to respond to CNS myelin, we also noticed several important differences between the two experiments. First, Ahmed et al. cultured DRG neurons in the presence of FGF2 (10 ng/ml) and the study did not address whether loss of NgR1 modulates FGF2 signaling rather than responsiveness to CNS myelin. Second, Ahmed et al. used soluble rat CNS myelin extracts of unknown composition, directly added to the culture medium to inhibit neurite growth. In the present study, we used membrane-bound MAG or substrate-bound recombinant OMgp or Nogo-66 to inhibit neurite outgrowth. Third, to monitor NgR1 knock-down, Ahmed et al. show Western blot analyses of DRG lysates 3 d after RNAi transfection and found a significant decrease of an ∼80 kDa immunoreactive band. Western blot analysis of brain extracts from NgR1 mutant mice results in complete loss of an anti-NgR1 immunoreactive band with an apparent molecular weight of 65 kDa (Fig. 1C). Consistent with this, we previously reported that endogenously expressed rat NgR1 has a molecular weight of 65 kDa (Venkatesh et al., 2005). Thus, it is not clear whether in the study by Ahmed et al. (2005) proteins other than NgR1 were knocked-down by the RNAi constructs used (Ahmed et al., 2005). For a more complete assessment of whether acute loss of NgR1 alters myelin-mediated neurite growth inhibition of different cell types, additional studies, including conditional ablation of NgR1, will be needed.

Here, we report that NgR1 is not necessary for Nogo-66 (Zheng et al., 2005), OMgp, and MAG-mediated neurite outgrowth inhibition. The apparent disparity between these results and the results reported by Kim et al. (2004), using an independently generated NgR1 knock-out mouse, appeared striking, but now appear to be related to differences in methods used to assay inhibition. Indeed, the conclusions made by Kim et al. are based on a DRG growth cone collapse assay, whereas our results are based on neurite outgrowth assays on substrate-bound inhibitor. Specifically, Kim et al. found that each of the three myelin inhibitors when applied acutely caused a modest increase in the collapse of wild-type DRG growth cones from ∼40 to ∼60%, an increase that was not seen with their NgR1-deficient neurons (Kim et al., 2004). Using the same assay and DRG neurons from our independently generated NgR1-null mice, we found that, in the presence of MAG-Fc or OMgp-Fc, a significantly greater number of growth cones is collapsed in wild-type neurons compared with NgR1 mutant neurons. Moreover, in DRG neurons, NgR1 is important for MAG-Fc- or OMgp-Fc-induced growth cone collapse. Thus, neurons from two independently generated NgR1-null mouse lines behave very similarly when analyzed in the same assay, which strengthens our conclusion that NgR1 is important for growth cone responses after acute application of inhibitor, but is dispensable for neurite outgrowth inhibition on Nogo-66 (Zheng et al., 2005), MAG, and OMgp substrate.

Two homologs of NgR1 have been identified in the mouse genome, NgR2 and NgR3, raising the question of whether these proteins function as alternate receptors for myelin inhibitors. With respect to Nogo-66 or OMgp signaling, NgR2 and NgR3 are likely not to serve as functional substitutes for NgR1. This is primarily based on previous observations that Nogo-66 and OMgp selectively bind to NgR1 but do not bind to NgR2 or NgR3 (Barton et al., 2003). Although NgR2 has been found to support high-affinity MAG binding, its CNS distribution is overlapping, yet distinct from NgR1 (Venkatesh et al., 2005; Barrette et al., 2007). For example, we have previously shown that NgR1 but not NgR2 is expressed in postnatal cerebellar granule neurons (Venkatesh et al., 2005). Furthermore, no upregulation of NgR2 expression is observed in NgR1-null mice (Fig. 1C), and thus, MAG inhibition of NgR1-deficient CGNs is likely to be independent of NgR1 and NgR2. Adult DRG neurons show a heterogeneous expression of NgR2 (Venkatesh et al., 2005), but loss of NgR1 results in a significant attenuation of MAG-Fc-induced growth cone collapse, suggesting that NgR2 does not contribute significantly to the collapse-inducing response. To what extent NgR2 expressed in other neuronal populations participates in MAG inhibition is not known, and future studies with primary neurons from NgR2 mutant or NgR1 and NgR2 double knock-out mice will be needed.

Our results imply that some other receptor(s) must exist that can bind Nogo-66, MAG, and OMgp to mediate the actions of these factors in substrate inhibition. If so, how does NgR1 fit in with this alternate receptor(s)? A first possibility is that NgR1 pairs with this receptor(s) to increase its affinity for the inhibitory ligands. In this model, loss of NgR1 would reduce the sensitivity of the receptor, which could lead to loss of collapse but not substrate inhibition if the concentration of each ligand was limiting in the collapse assay but saturating in the substrate assay. Arguing strongly against this possibility, however, is our finding from a substrate dilution series that the degree of inhibition by different concentrations of MAG is the same for wild-type and NgR1 knock-out neurons (supplemental Fig. 2, available at www.jneurosci.org as supplemental material). Also, neurons from p75 knockout mice showed enhanced neurite outgrowth in similar experimental procedures (Wang et al., 2002b; Yamashita et al., 2002; Zheng et al., 2005). Instead, our results appear to imply that different mechanisms are required for collapse and for growth inhibition. This conclusion is unexpected, because it is widely assumed in the field that growth cone collapse-inducing activities simply reflect the effect of acute activation of an inhibitory receptor. Instead, the finding that myelin inhibitors can inhibit growth of NgR1^−/− neurons without causing collapse of their growth cones shows that the two effects can be dissociated. How this occurs mechanistically is not clear, although one could imagine several ways in which differential effects could be generated; for example, one could speculate that the alternate receptor(s) impinges on mechanisms regulating protrusive activity of actin filaments or stabilization of microtubules, and that NgR1, either in collaboration with this receptor(s) or through some other means, is poised to trigger actin depolymerization. Elucidating the precise mechanism will require additional studies.

What are the implications of our findings for CNS regeneration? The role of the Nogo/NgR1 system in limiting neuronal growth and axonal regeneration after spinal cord injury is controversial (Cafferty et al., 2007; Steward et al., 2007). Two independent studies showed that germline ablation of NgR1 does not result in a significant regeneration of corticospinal tract axons after dorsal hemisection at T6 or complete transection at T8 (Kim et al., 2004) or dorsal hemisection at T8 (Zheng et al., 2005). Kim et al. (2004) reported regeneration of some raphespinal and rubrospinal fibers in NgR1 mutants after dorsal hemisection. In a more recent study, it was found that pyramidotomy in NgR1 mutants does not lead to long-distance axonal regeneration but induced growth of corticospinal tract axons into denervated cervical gray matter proximal to the injury site (Cafferty and Strittmatter, 2006). Consistent with the lack of long-distance axonal growth in spinal cord-injured NgR1 mutants in vivo, we report here that neurite outgrowth of NgR1-deficient neurons is strongly inhibited by substrate-bound Nogo-66, OMgp, or MAG in vitro.

Several studies have reported positive behavioral effects in injured rodents after antagonism of NgR1 with peptide inhibitors of Nogo (GrandPre et al., 2002; Li and Strittmatter, 2003), soluble receptor bodies [NgR(310)Ecto-Fc] (Li et al., 2004; Wang et al., 2006), NgR1 DNA vaccination (Yu et al., 2007), or genetic ablation (Lee et al., 2004). Some of these manipulations had the potential to interfere with more than just NgR1, so caution must be exercised in interpreting these results. To the extent that behavioral improvement in some of these studies can be attributed specifically to antagonism of NgR1, the question arises whether the improvement is related to more subtle structural changes of the neuronal architecture, rather than long-distance axonal sprouting. Short-range morphological changes, including collateral sprouting, altered shape or number of dendritic spines, and number of synaptic contacts may be achieved primarily through remodeling of the actin cytoskeleton. Spines are small dendritic perturbations that represent postsynaptic sites of excitatory synaptic input and changes in spine number and morphology have been proposed to contribute to alterations in excitatory synaptic transmission (Lamprecht and LeDoux, 2004). Functional recovery, although often very limited, is observed after CNS traumatic injury or stroke and is believed, at least in part, to be a reflection of natural plasticity of uninjured neurons adjacent to denervated areas (Belhaj-Saif and Cheney, 2000; Raineteau and Schwab, 2001). Functional recovery can be enhanced by physiotherapy (exercise) as a natural and noninvasive paradigm to promote neuronal activity, which leads to increased expression of neurotrophic factors, enhanced plasticity, and rehabilitative gains (de Leon et al., 1998; Toni et al., 1999; Vaynman and Gomez-Pinilla, 2005). Consistent with the idea that NgR1 limits neuronal plasticity in the more mature nervous system, loss of physiological NgR1 signaling has been found to disrupt consolidation of synaptic connectivity in the visual system beyond the critical period (McGee et al., 2005). The underlying molecular mechanisms of enhanced anatomical plasticity after functional depletion of NgR1 in vivo, however, are mostly unknown. Finally, functional improvement in spinal cord-injured rodents treated with NgR1 antagonists is observed 1–2 weeks after injury (GrandPre et al., 2002; Li et al., 2005), which is likely to be too rapid to reflect true long-distance axonal regeneration (Steward et al., 2003). In future studies, it will be interesting to explore whether NgR1 participates in regulating short-range structural plasticity, such as dendritic spine architecture, thereby influencing synaptic connectivity and neuronal transmission.

Together, our data and those of Zheng et al. (2005) indicate that NgR1 is not necessary for neurite outgrowth inhibition elicited by any of its high-affinity myelin ligands, including the Nogo-A inhibitory peptide Nogo-66, OMgp, and MAG. Interestingly, we find that NgR1 is important to bring about growth cone collapse in the presence of soluble MAG or OMgp ectodomains. As discussed above, our findings do not rule out a physiological role for the known myelin inhibitor–NgR1 interactions; however, they fail to support the idea of NgR1 as an important receptor for myelin-mediated inhibition of longitudinal neurite extension. Finally, given the growing interest in NgR1 as a target for drug development in nervous system regeneration (Lee et al., 2003, 2004; Hu and Strittmatter, 2004; Satoh et al., 2005), it is our hope that the data presented here will help to focus ongoing efforts to further elucidate the full complement of receptors and molecular mediators of CNS myelin-mediated outgrowth inhibition.