PirB is a Functional Receptor for Myelin Inhibitors of Axonal Regeneration

See allHide authors and affiliations

Science  07 Nov 2008:
Vol. 322, Issue 5903, pp. 967-970
DOI: 10.1126/science.1161151


A major barrier to regenerating axons after injury in the mammalian central nervous system is an unfavorable milieu. Three proteins found in myelin—Nogo, MAG, and OMgp—inhibit axon regeneration in vitro and bind to the glycosylphosphatidylinositol-anchored Nogo receptor (NgR). However, genetic deletion of NgR has only a modest disinhibitory effect, suggesting that other binding receptors for these molecules probably exist. With the use of expression cloning, we have found that paired immunoglobulin-like receptor B (PirB), which has been implicated in nervous system plasticity, is a high-affinity receptor for Nogo, MAG, and OMgp. Interfering with PirB activity, either with antibodies or genetically, partially rescues neurite inhibition by Nogo66, MAG, OMgp, and myelin in cultured neurons. Blocking both PirB and NgR activities leads to near-complete release from myelin inhibition. Our results implicate PirB in mediating regeneration block, identify PirB as a potential target for axon regeneration therapies, and provide an explanation for the similar enhancements of visual system plasticity in PirB and NgR knockout mice.

Myelin, an insulating layer surrounding axons, is thought to pose an obstacle to axon regeneration, inhibiting neurite outgrowth in vitro and contributing to regeneration failure in vivo. The NgR, a candidate receptor for the myelin-derived inhibitors Nogo, MAG, and OMgp (15), appears to be required for the acute inhibitory activity of these proteins, because genetic removal of NgR blocks acute growth-cone collapse in response to these factors when added in solution (6, 7). However, genetic deletion of NgR does not relieve the chronic inhibition of neurite outgrowth by myelin inhibitors presented as substrates (7, 8). Furthermore, genetic deletion of NgR does not enhance regeneration of corticospinal tract (CST) axons after dorsal hemisection (6, 8), although some regeneration of raphespinal and rubrospinal tracts after spinal cord injury has been reported (6). These data suggest that NgR is important for mediating some of the inhibitory activity of myelin inhibitors but that other binding receptors for these factors remain to be identified. Such putative receptors could work either independently or in concert with NgR.

To identify candidate receptors for myelin inhibitors, we used expression cloning to screen an arrayed library of human cDNA pools (9). Our screen identified only two receptors for Nogo66: (i) NgR and (ii) the human leukocyte immunoglobulin (Ig)–like receptor B2 (LILRB2). LILRB2 is part of the B type subfamily of LILR receptors, which consists of five highly homologous family members in humans (10) (fig. S1). In mice, however, there is a single ortholog, paired immunoglobulin-like receptor B (PirB) (11). PirB shares only ∼50% amino acid similarity with LILRB2, and it contains six Ig-like repeats, as opposed to four Ig-like repeats in LILRB proteins (Fig. 1A). Despite this low level of homology, we found that alkaline phosphatase (AP)–Nogo66 can bind PirB (Fig. 1B). The affinity of the AP-Nogo66-PirB interaction was similar to that of the AP-Nogo66 interaction with NgR (fig. S2). Given the unusual promiscuity of NgR binding to Nogo66, MAG, and OMgp, it is possible that LILRB2 and PirB also bind other myelin inhibitors. Indeed, we found that MAG-Fc and AP-OMgp bind PirB (Fig. 1B and fig. S2). The MAG-Fc-PirB interaction is high-affinity (half-maximal saturation of the interaction between purified MAG-Fc and purified PirB ectodomain: 13.8 ± 6 nM) (Fig. 1C).

Fig. 1.

LILRB2 and PirB can bind Nogo66 and MAG. (A) Schematic diagram of LILRB, PirB, and NgR receptors. ITIM, immunoreceptor tyrosine-based inhibitory motif; LRR, leucine-rich repeat. (B) COS cells were transfected with empty vector (control) or PirB or LILRB2 cDNA. After 48 hours, cells were incubated with AP-Nogo66 or MAG-Fc, and bound ligand was detected. NgR and NgR2 were used as positive controls for AP-Nogo66 and MAG-Fc binding, respectively. Scale bars, 200 μm. (C) The affinity of the MAG-Fc-PirB interaction is shown by one representative enzyme-linked immunosorbent assay binding curve. The experiment was repeated twice with similar results.

To address whether PirB is a functional receptor for Nogo66, we focused on juvenile (P7) cerebellar granule neurons (CGNs), whose neurite outgrowth is inhibited when grown on AP-Nogo66 (12). Adult CGNs have been shown to express PirB (13), and we found that is also the case for juvenile CGNs, as assessed by reverse transcription polymerase chain reaction and in situ hybridization (fig. S3). We first tested the ability of a soluble ectodomain of PirB (PirB-His) to interfere with AP-Nogo66 inhibition in vitro. AP-Nogo66 inhibits neurite outgrowth of P7 CGNs to ∼66% of untreated control levels (Fig. 2A). Inclusion of PirB-His in this assay reversed AP-Nogo66 inhibition, with neurite outgrowth returning essentially to control levels. These results are similar to those reported using the ectodomain of NgR to block inhibition by Nogo66 (8, 14, 15), and they indicate that PirB can bind the functionally inhibitory domain of Nogo66 but do not address whether endogenous PirB in CGNs mediates inhibition by AP-Nogo66. Therefore, antibodies to PirB that are capable of interfering with the PirB-Nogo66 interaction were generated. Using a phage display platform (16) directed against the extracellular domain of PirB, we screened multiple clones for their ability to block binding of AP-Nogo66 to PirB (fig. S4). Clone YW259.2 (hereafter referred to as anti-PirB.1), which interfered best with AP-Nogo66-PirB binding, had a dissociation constant Kd of 5 nM for PirB. Anti-PirB.1 had no effect on the baseline axon growth of CGNs. However, it significantly reduced inhibition by AP-Nogo66 or myelin in cultured CGNs (Fig. 2B), rescuing neurite outgrowth to 59% from 41% on AP-Nogo66 and to 62% from 47% on myelin. Similar results were seen with MAG as an inhibitory substrate or with a different neuronal cell type [dorsal root ganglion (DRG) neurons] (fig. S5). These results suggest that PirB is a functional receptor mediating long-term inhibition of neurite outgrowth.

Fig. 2.

Blocking PirB reverses inhibition of CGN outgrowth on AP-Nogo66 or myelin. Dissociated mouse P7 CGNs were plated on PDL/laminin (control), AP-Nogo66, or myelin to test inhibition by these substrates after various manipulations. In each panel, representative photomicrographs are shown on the left, and a graph measuring average neurite length (±SE, error bars) from one representative experiment is shown on the right. (A) Neurons were grown on PDL/laminin or AP-Nogo66, either alone or mixed with a fivefold excess of PirB extracellular domain (PirB-His). Neuronal inhibition by AP-Nogo66 was largely rescued by the presence of PirB-His. (B) Neurons grown on PDL/laminin, AP-Nogo66, or myelin were cultured in the presence or absence of anti-PirB.1 (50 μg/ml), which significantly reduced inhibition by either substrate. (C) Neurons cultured from either WT control mice or PirBTM mice were grown on PDL/laminin, AP-Nogo66, or myelin. PirBTM neurons were significantly less inhibited on either substrate (Student's t test,*P <0.01; n = 6 wells per condition). Scale bars, 50 μm.

To confirm this result, we made use of mice carrying a loss-of-function PirB allele, the PirBTM mice, in which four exons encoding the transmembrane domain and part of the PirB intracellular domain have been removed (13). CGNs were cultured from PirBTM mice or wild-type (WT) littermates on a control substrate, AP-Nogo66, or myelin. On the control substrate (PDL/laminin), PirBTM neurons behaved similarly to WT neurons (Fig. 2C). However, neurite outgrowth from PirBTM neurons was markedly less inhibited than that from WT neurons on either AP-Nogo66 or myelin. On AP-Nogo66, outgrowth from WT neurons was inhibited to 50% of control levels, whereas PirBTM neurons were inhibited to 66%. Similarly, on myelin, WT neurons were inhibited to 52% of control levels, whereas PirBTM neurons were inhibited to 70%. Again, we observed similar partial disinhibition of PirBTM DRG neurons on both myelin and AP-Nogo66 (fig. S5). We also saw disinhibition of PirBTM CGNs on MAG and OMgp (fig. S5). These findings indicate that PirB is indeed a functional receptor for AP-Nogo66, MAG, OMgp, and myelin-mediated inhibition of neurite growth. However, loss of PirB activity does not fully rescue outgrowth.

It is possible that PirB and NgR function together to mediate inhibition of neurite outgrowth. To address this concept, both PirB and NgR function were blocked together in CGNs by culturing neurons from NgR-null mice in the presence of anti-PirB.1. As we have reported previously (8), NgR–/– CGN neurite outgrowth is inhibited by AP-Nogo66 or myelin to the same extent as that in WT neurons (50% and 49%) (Fig. 3). Anti-PirB.1 antibody treatment of NgR+/– neurons partially reversed inhibition by either AP-Nogo66 or myelin, as discussed above with WT neurons. Anti-PirB.1 treatment of NgR–/– neurons resulted in a similar partial disinhibition on AP-Nogo66 but did not provide any further rescue, suggesting that NgR is not involved in AP-Nogo66 inhibition. In contrast, anti-PirB.1 treatment of NgR–/– neurons restored neurite outgrowth on myelin to nearly control levels. Thus, it appears that PirB, but not NgR, is required for substrate inhibition by AP-Nogo66 in CGNs, but only accounts for it partly. In contrast, PirB and NgR both contribute to the substrate inhibition imparted by myelin.

Fig. 3.

PirB and NgR together contribute to myelin inhibition. Neurons were cultured on PDL/laminin, AP-Nogo66, or myelin substrates, and the PirB and NgR receptors were functionally blocked, either independently or in combination, to assess the contribution of these two pathways to inhibition. Representative photomicrographs are shown (top), and a graph measuring average neurite length (±SE, error bars) from one representative experiment is shown (bottom). NgR functional blockade was achieved by culturing neurons from NgR-null mice. PirB blockade was achieved by culturing neurons in the presence of anti-PirB.1 (50 μg/ml) (Student's t test,*P <0.01; n = 6 per condition). Scale bars, 50 μm.

Because NgR is required for growth-cone collapse in response to various myelin inhibitors (6, 7), it is possible that PirB is also involved in this more acute response. For this experiment, we used sensory neurons from the DRG of 3-week-old mice, confirmed to express PirB (fig. S3). Similar to what has been shown by others (6, 7), we found that growth cones in this culture system have a high baseline level of collapse (∼30%), which is further increased (to ∼75%) by incubation with AP-Nogo66 or myelin (Fig. 4). As reported previously (6, 7), this collapse was largely abolished in NgR–/– neurons. In addition, blocking PirB function with anti-PirB.1 was also sufficient to reverse growth-cone collapse by these inhibitors.

Fig. 4.

Both PirB and NgR are required to mediate growth-cone collapse by myelin inhibitors. Growth cones of postnatal DRG axons were treated with medium alone (control), myelin (3 μg/ml), or AP-Nogo66 (100 nM) for 30 min to stimulate collapse and were stained with rhodamine-phalloidin to visualize growth cones. Representative photomicrographs are shown on the left, and a graph measuring percent of growth-cone collapse (±SE, error bars) from cumulative experiments is shown on the right (n ≥ 4 per condition). Scale bars, 50 μm.

Together, these results support a previously unknown role for PirB as a necessary receptor for neurite inhibition by myelin extracts and, more specifically, by the myelin-associated inhibitors Nogo66, MAG, and OMgp. PirB appears to be a more substantial mediator of substrate inhibition than NgR, as removal of PirB function alone (either genetically or with antibodies) partially disinhibits growth on both myelin extracts and inhibitors, whereas genetic removal of NgR alone does not disinhibit on any of these substrates. However, NgR appears to play an adjunct role in mediating inhibition by myelin extracts (but not Nogo66), because genetic removal of NgR can augment the disinhibition caused by anti-PirB antibodies on myelin (but not on Nogo66). Other co-receptors or modulators may also contribute in parallel, such as p75, TROY, LINGO, and gangliosides (17, 18). Conversely, other PirB ligands, including major histocompatibility complex class I proteins, may contribute to the inhibitory action of myelin (13, 19, 20). Our finding of collaboration between PirB and NgR may help to explain the surprising lack of enhanced CST regeneration after dorsal spinal cord hemisection in NgR knockout mice (6, 8), despite the reported regeneration or sprouting seen in rodents infused with the NgR ectodomain (21). Thus, it might be necessary to remove both PirB and NgR to achieve extensive regeneration in vivo. In addition, because on Nogo66 substrate the genetic removal of NgR does not further augment the partial disinhibitory effect of PirB removal, it is likely that there are additional binding receptors for Nogo66.

Although PirB appears to be a more important receptor for substrate inhibition than NgR, inactivation of either PirB or NgR alone is sufficient to block the acute growth-cone collapse caused by the addition of myelin inhibitors. This observation suggests that collapse is a more demanding process, requiring both PirB and NgR activities, acting either in parallel or together. In this context, it is of interest that PirB and NgR receptors have recently been shown to play similar roles in limiting plasticity of synaptic connections in the visual cortex. In mice lacking either receptor, eye closure during a critical developmental period results in excessive strengthening of connections via the open eye (13, 22). The mechanisms responsible for the effect of both receptors in mediating growth-cone collapse could also underlie the commonality of their role in ocular dominance plasticity.

The mechanism by which PirB signals to inhibit axon growth in response to myelin inhibitors is not clear. However, PirB has been shown to antagonize the function of integrin receptors (23) and to recruit both src homology 2–containing protein tyrosine phosphatase (SHP)–1 and SHP-2 phosphatases (13, 24); either or both of these events could attenuate normal neurite outgrowth. In humans, one or more members of the LILRB gene family might also play a role in regeneration. The blockade of PirB/LILRB activity, either with antibodies or by other means, provides an important target for therapeutic interventions to stimulate axonal regeneration.

Supporting Online Material

Materials and Methods

Figs. S1 to S5


References and Notes

View Abstract

Navigate This Article