Soluble myelin-associated glycoprotein released from damaged white matter inhibits axonal regeneration.

The adult, mammalian CNS does not regenerate after injury largely because of a glial scar and inhibitors of regeneration in myelin. To date, two myelin inhibitors, myelin-associated glycoprotein (MAG) and Nogo, both transmembrane proteins, have been identified. No secreted inhibitors of regeneration have been described. However, a proteolytic fragment of MAG (dMAG), consisting of the entire extracellular domain, is readily released from myelin and is found in vivo. Here, we show, first, that a soluble, chimeric form of MAG (MAG-Fc), when secreted from CHO cells in a collagen gel and hence in the absence of a fixed substrate, inhibits/deflects neurite outgrowth from P6 dorsal root ganglion (DRG) neurons. This inhibition was blocked when a MAG monoclonal antibody was included in the gel and a control chimera sialoadhesin-Fc (Sn-Fc), which, like MAG, binds neurons in a sialic acid-dependent manner but does not inhibit axonal growth, had no effect. Using the same assay system we showed that factors secreted from damaged white matter inhibited/deflected neurite outgrowth. This inhibition was neutralized when a MAG monoclonal antibody was included in the gel and there was no inhibition when white matter from a MAG knockout mouse was used. Factors secreted from damaged white matter from wild-type mice had no effect on neurite outgrowth from E18 DRG neurons. These results show that factors secreted from damaged white matter inhibit axonal regeneration and that the majority of inhibitory activity can be accounted for by dMAG. Thus, released dMAG is likely to play an important role in preventing regeneration, immediately after injury before the glial scar forms.

Neuronal cyclic AMP controls the developmental loss in ability of axons to regenerate.

Unlike neonatal axons, mammalian adult axons do not regenerate after injury. Likewise, myelin, a major factor in preventing regeneration in the adult, inhibits regeneration from older but not younger neurons. Identification of the molecular events responsible for this developmental loss of regenerative capacity is believed key to devising strategies to encourage regeneration in adults after injury. Here, we report that the endogenous levels of the cyclic nucleotide, cAMP, are dramatically higher in young neurons in which axonal growth is promoted both by myelin in general and by a specific myelin component, myelin-associated glycoprotein (MAG), than in the same types of neurons that, when older, are inhibited by myelin-MAG. Inhibiting a downstream effector of cAMP [protein kinase A (PKA)] prevents myelin-MAG promotion from young neurons, and elevating cAMP blocks myelin-MAG inhibition of neurite outgrowth in older neurons. Importantly, developmental plasticity of spinal tract axons in neonatal rat pups in vivo is dramatically reduced by inhibition of PKA. Thus, the switch from promotion to inhibition by myelin-MAG, which marks the developmental loss of regenerative capacity, is mediated by a developmentally regulated decrease in endogenous neuronal cAMP levels.

Epitope-tagged P(0) glycoprotein causes Charcot-Marie-Tooth-like neuropathy in transgenic mice.

In peripheral nerve myelin, the intraperiod line results from compaction of the extracellular space due to homophilic adhesion between extracellular domains (ECD) of the protein zero (P(0)) glycoprotein. Point mutations in this region of P(0) cause human hereditary demyelinating neuropathies such as Charcot-Marie-Tooth. We describe transgenic mice expressing a full-length P(0) modified in the ECD with a myc epitope tag. The presence of the myc sequence caused a dysmyelinating peripheral neuropathy similar to two distinct subtypes of Charcot-Marie-Tooth, with hypomyelination, altered intraperiod lines, and tomacula (thickened myelin). The tagged protein was incorporated into myelin and was associated with the morphological abnormalities. In vivo and in vitro experiments showed that P(0)myc retained partial adhesive function, and suggested that the transgene inhibits P(0)-mediated adhesion in a dominant-negative fashion. These mice suggest new mechanisms underlying both the pathogenesis of P(0) ECD mutants and the normal interactions of P(0) in the myelin sheath.

Acylation of myelin Po protein is required for adhesion.

The extracellular domains of myelin Po protein interact homophilically and hence hold myelin compact at the intraperiod line. The cytoplasmic domain of Po, however, can also affect the interactions of its extracellular sequences. Po is acylated, mostly with palmitic acid, at Cys 153, just at the transmembrane:cytoplasmic domain interface. Here we show that Po mutated at Cys 153 to alanine (C153A), is not acylated and is not adhesive. Like wild-type Po, C153A Po clusters within the membrane and seems to interact with the cytoskeleton. On the other hand, the rate of turnover of C153A Po in transfected Chinese hamster ovary cells is almost 4 times faster than wild-type Po. The increased instability of C153A Po compared to wild-type Po may account for its loss of adhesion.

Axon regeneration: Vaccinating against spinal cord injury.

Myelin is a potent inhibitor of axon regeneration, but has been viewed as just one of many factors that prevent regeneration after injury. So it comes as a surprise that immunization against myelin has been found to allow extensive axon regeneration after injury, without apparent autoimmune-induced demyelination.

Glial inhibition of nerve regeneration in the mature mammalian CNS.

The lack of axonal regeneration in the adult mammalian CNS is due to both unfavorable environmental glial factors and the intrinsic neuronal state. Inhibitors associated with myelin and the glial scar have been extensively studies and it has been shown that neutralizing at least some of the inhibitors can lead to improved growth. Meanwhile, important advances have also been made towards our understanding of the neuronal intrinsic state, particularly the intracellular levels of cyclic nucleotide, that influence the capacity of mature CNS neurons to initiate and maintain a regrowth response. It is well recognized that successful regeneration may only be achieved by application of a combination of strategies that both block glial inhibitors and enhance the intrinsic neuronal growth capacity.

Characterization of the effect on adhesion of different mutations in myelin P0 protein.

Table 1 summarizes the results obtained from expressing in CHO cells C21A P0 and N93A P0 alone, and each with wild-type P0. As can be seen, each of these mutated proteins reach the cell surface when expressed alone, but neither is adhesive. Finally, when each of these mutated P0 molecules is expressed with the wild-type P0, wild-type P0 is no longer adhesive. Therefore, both C21A Po and N93A P0 each have a dominant negative effect on adhesion of wild-type P0. This approach to address the functional consequences of mutations in P0 can now be used to assess the effects of different mutations associated with CMT1B.

Myelin-associated glycoprotein, MAG, selectively binds several neuronal proteins.

Myelin-associated glycoprotein (MAG) is a potent inhibitor of axonal regeneration and also, depending on the age and type of neuron, can promote axonal growth. In addition, MAG influences stability of both myelin and the axon in the intact, mature nervous system. The identity of the neuron/axonal MAG-binding receptor responsible for effecting these responses is not known. Here we show that a soluble, chimeric form of MAG, MAG-Fc, can bind to the neuronal cell body and neurites equally well, in a sialic acid-dependent manner. Importantly, MAG-Fc specifically precipitates a number of surface proteins from post-natal cerebellar, dorsal root ganglion (DRG) and PC12 neurons. These proteins are not precipitated by a control Fc-containing chimera and are not apparent when a MAG antibody is included in the precipitation mix as a competitive inhibitor. Based on molecular weight, two prominent proteins of 190 and 250 kD are precipitated from all three neuron types. The 190 kD protein is a sialoglycoprotein, since it is not apparent in the precipitate from neurons which have been desialylated. Other proteins are precipitated but are less abundant and are different for each type of neuron. One or more of these proteins is/are likely to be the functional MAG receptor.

Prior exposure to neurotrophins blocks inhibition of axonal regeneration by MAG and myelin via a cAMP-dependent mechanism.

MAG is a potent inhibitor of axonal regeneration. Here, inhibition by MAG, and myelin in general, is blocked if neurons are exposed to neurotrophins before encountering the inhibitor; priming cerebellar neurons with BDNF or GDNF, but not NGF, or priming DRG neurons with any of these neurotrophins blocks inhibition by MAG/myelin. Dibutyryl cAMP also overcomes inhibition by MAG/myelin, and cAMP is elevated by neurotrophins. A PKA inhibitor present during priming abrogates the block of inhibition. Finally, if neurons are exposed to MAG/myelin and neurotrophins simultaneously, but with the Gi protein inhibitor, inhibition is blocked. We suggest that priming neurons with particular neurotrophins elevates cAMP and activates PKA, which blocks subsequent inhibition of regeneration and that priming is required because MAG/myelin activates a Gi protein, which blocks increases in cAMP. This is important for encouraging axons to regrow in vivo.

Myelin-associated glycoprotein in myelin and expressed by Schwann cells inhibits axonal regeneration and branching.

The mammalian CNS does not regenerate after injury due largely to myelin-specific inhibitors of axonal growth. The PNS, however, does regenerate once myelin is cleared and myelin proteins are down-regulated by Schwann cells. Myelin-associated glycoprotein (MAG), a sialic acid binding protein, is a potent inhibitor of neurite outgrowth when presented to neurons in culture. Here, we present additional evidence that strongly supports the suggestion that MAG contributes to the overall inhibitory properties of myelin. When myelin from MAG-/- mice is used as a substrate, axonal length is 100 and 60% longer for neonatal cerebellar and older DRG neurons, respectively, compared to MAG+/+ myelin. The converse is true for neurites from neonatal DRG neurons, which are twice as long on MAG+/+ relative to MAG-/- myelin, consistent with MAG's dual role of promoting axonal growth from neonatal DRG neurons but inhibiting growth in older DRG and all other postnatal neurons examined. Furthermore, desialylating neurons reverses inhibition by CNS myelin by 45%. Contrary to previous reports, under these conditions PNS myelin is also inhibitory for axonal regeneration. Importantly, results using PNS MAG-/- myelin as a substrate suggest that MAG contributes to this inhibition. Finally, when Schwann cells not expressing MAG and permissive for axonal growth are induced to express MAG by retroviral infection, not only is axonal outgrowth greatly inhibited by these cells but so also is neurite branching. This suggests for the first time that MAG not only affects axonal regeneration but may also play a role in the control of axonal sprouting.

Myelin Po protein mutated at Cys21 has a dominant-negative effect on adhesion of wild type Po.

The homophilic adhesion of the peripheral nervous system myelin protein, Po, holds myelin compact at the extracellular leaflets. Po carries a single immunoglobulin (Ig)-like domain that is stabilized by a disulfide bond between Cys21 and Cys98. We showed previously that Po mutated at Cys21 to Ala (C21A Po), unlike wild type Po, when it was expressed in Chinese hamster ovary (CHO) cells after transfection, was not adhesive. To determine whether C21A Po could influence the adhesion of wild type Po, cells that already expressed wild type Po and that were shown to be adhesive were retransfected with plasmids containing C21A Po. After selection in hygromycin, clones that coexpressed wild type Po and Cys21-mutated Po were identified by polyacrylamide gel electrophoresis in the presence and absence of beta-mercaptoethanol, because only in coexpressors will there be two forms of Po of different apparent molecular weights under nonreducing conditions. Two clones that coexpressed wild type Po and C21A Po and a third clone, which, although it was resistant to hygromycin, expressed only wild type Po, were chosen for further study. An enzyme-linked immunosorbent assay of fixed, unpermeabilized cells showed that all of these clones expressed approximately equivalent amounts of Po at the cell surface. However, in an aggregation adhesion assay, only the cells that expressed wild type Po alone formed large aggregates and dropped in total particle number with time. The cells that coexpressed wild type and C21A Po did not form aggregates, and their adhesive properties were indistinguishable from the control transfected cells, which did not express Po. These results suggest that C21A Po has a dominant-negative effect on adhesion of wild type Po.

Myelin-associated glycoprotein interacts with neurons via a sialic acid binding site at ARG118 and a distinct neurite inhibition site.

Inhibitory components in myelin are largely responsible for the lack of regeneration in the mammalian CNS. Myelin-associated glycoprotein (MAG), a sialic acid binding protein and a component of myelin, is a potent inhibitor of neurite outgrowth from a variety of neurons both in vitro and in vivo. Here, we show that MAG's sialic acid binding site is distinct from its neurite inhibitory activity. Alone, sialic acid-dependent binding of MAG to neurons is insufficient to effect inhibition of axonal growth. Thus, while soluble MAG-Fc (MAG extracellular domain fused to Fc), a truncated form of MAG-Fc missing Ig-domains 4 and 5, MAG(d1-3)-Fc, and another sialic acid binding protein, sialoadhesin, each bind to neurons in a sialic acid- dependent manner, only full-length MAG-Fc inhibits neurite outgrowth. These results suggest that a second site must exist on MAG which elicits this response. Consistent with this model, mutation of arginine 118 (R118) in MAG to either alanine or aspartate abolishes its sialic acid-dependent binding. However, when expressed at the surface of either CHO or Schwann cells, R118-mutated MAG retains the ability to inhibit axonal outgrowth. Hence, MAG has two recognition sites for neurons, the sialic acid binding site at R118 and a distinct inhibition site which is absent from the first three Ig domains.

Sialic acid specificity of myelin-associated glycoprotein binding.

Myelin-associated glycoprotein (MAG), a nervous system cell adhesion molecule, is an I-type lectin that binds to sialylated glycoconjugates, including gangliosides bearing characteristic structural determinants (Yang, L. J.-S., Zeller, C. B., Shaper, N. L., Kiso, M., Hasegawa, A., Shapiro, R. E., and Schnaar, R. L. (1996) Proc. Natl. Acad. Sci. U.S.A. 93, 814-818). Two cell adhesion systems, COS-1 monkey kidney fibroblasts transiently transfected to express MAG and Chinese hamster ovary (CHO) cells stably transfected to express MAG, were used to probe the structural specificity of MAG-ganglioside binding. Both cell types bound to the same gangliosides: GQ1b alpha (IV3NeuAc,III6NeuAc,II3(NeuAc)2Gg4Cer) > GT1b = GD1a > GM3 > GM1, GD1b, and GQ1b (the latter do not support adhesion). Binding was enhanced by pretreatment of MAG-expressing cells with neuraminidase. MAG-expressing Chinese hamster ovary cells bound directly to gangliosides resolved on thin layer chromatograms, allowing detection of MAG binding species in a mixture. The simplest ganglioside ligand for MAG was GM3 bearing N-acetylneuraminic acid, whereas GM3 bearing N-glycolylneuraminic acid did not support adhesion. Chemical modifications of N-acetylneuraminic acid residues (on GD1a) abrogated MAG binding. Mild periodate oxidation of sialic acids to their corresponding seven-carbon (or eight-carbon) sialic acid aldehydes abolished MAG binding, as did further conversion to the corresponding primary alcohols. Eliminating the anionic charge by ethyl esterification, amidation, or reduction also abolished MAG-mediated cell adhesion. These data demonstrate that MAG-ganglioside binding is highly specific and defines key carbohydrate structural determinants for MAG-mediated cell adhesion to gangliosides.

Soluble myelin-associated glycoprotein (MAG) found in vivo inhibits axonal regeneration.

Myelin-associated glycoprotein (MAG) is a potent inhibitor of axonal regeneration when used as a substrate for growth. However, to be characterized definitively as inhibitory rather than nonpermissive, MAG must also inhibit axonal regeneration when presented in solution. Here, we show that soluble dMAG (extracellular domain only), released in abundance from myelin and found in vivo and chimeric MAG-Fc, can potently inhibit axonal regeneration. For both dMAG and MAG-Fc, inhibition is dose-dependent. If myelin-conditioned medium is immunodepleted of dMAG, or if a MAG antibody is included with MAG-Fc, inhibition is completely neutralized. Together with MAG's ability to induce growth cone collapse, these results demonstrate that MAG is an inhibitory molecule and not merely nonpermissive. The results also suggest that MAG binds to a specific receptor and initiates a signal transduction cascade to effect inhibition. Importantly, these results indicate that soluble dMAG detected in vivo could contribute to the lack of regeneration in the mammalian CNS after injury.

Dominant-negative effect on adhesion by myelin Po protein truncated in its cytoplasmic domain.

The myelin Po protein is believed to hold myelin together via interactions of both its extracellular and cytoplasmic domains. We have already shown that the extracellular domains of Po can interact in a homophilic manner (Filbin, M.T., F.S. Walsh, B.D. Trapp, J.A. Pizzey, and G.I. Tennekoon. 1990. Nature (Lond.). 344:871-872). In addition, we have shown that for this homophilic adhesion to take place, the cytoplasmic domain of Po must be intact and most likely interacting with the cytoskeleton; Po proteins truncated in their cytoplasmic domains are not adhesive (Wong, M.H., and M.T. Filbin, 1994. J. Cell Biol. 126:1089-1097). To determine if the presence of these truncated forms of Po could have an effect on the functioning of the full-length Po, we coexpressed both molecules in CHO cells. The adhesiveness of CHO cells expressing both full-length Po and truncated Po was then compared to cells expressing only full-length Po. In these coexpressors, both the full-length and the truncated Po proteins were glycosylated. They reached the surface of the cell in approximately equal amounts as shown by an ELISA and surface labeling, followed by immunoprecipitation. Furthermore, the amount of full-length Po at the cell surface was equivalent to other cell lines expressing only full-length Po that we had already shown to be adhesive. Therefore, there should be sufficient levels of full-length Po at the surface of these coexpressors to measure adhesion of Po. However, as assessed by an aggregation assay, the coexpressors were not adhesive. By 60 min they had not formed large aggregates and were indistinguishable from the control transfected cells not expressing Po. In contrast, in the same time, the cells expressing only the full-length Po had formed large aggregates. This indicates that the truncated forms of Po have a dominant-negative effect on the adhesiveness of the full-length Po. Furthermore, from cross-linking studies, full-length Po, when expressed alone but not when coexpressed with truncated Po, appears to cluster in the membrane. We suggest that truncated Po exerts its dominant-negative effect by preventing clustering of full-length Po. We also show that colchicine, which disrupts microtubules, prevents adhesion of cells expressing only the full-length Po. This strengthens our suggestion that an interaction of Po with the cytoskeleton, either directly or indirectly, is required for adhesion to take place.

Mapping the adhesive domains of the myelin Po protein.

The Po protein holds PNS myelin compact at the intraperiod line by homophilic interactions of its single immunoglobulin (Ig)-like domain. Using transfected Chinese hamster ovary (CHO) cells expressing Po we can monitor this adhesion in vitro and have shown that the cells expressing Po when incubated as a single-cell suspension form large aggregates, whereas control-transfected cells do not. To precisely map the domains of Po responsible for Po:Po-mediated membrane adhesion, the ability of a number of antibodies raised to peptides corresponding to segments of the Ig-domain of Po, and the ability of the Po-peptides themselves, to inhibit aggregation was assessed. Both antibodies to Po-peptide, SDNGT, corresponding to amino acids Po 91-95, and the peptide itself, were able to block adhesion completely. Furthermore, within this Po sequence, amino acids Asp 92 and Gly 94 are conserved in a large number of V-like Ig-domains. To determine if these two amino acids are important for Po-mediated adhesion, the nucleotides coding for Asp 92 and Gly 94 were mutated to encode glutamate and alanine, respectively. Although the mutated Po reached the surface in transfected CHO cells and was glycosylated, the cells did not aggregate. These results suggest that the sequence SDNGT in the extracellular domain of Po is important for adhesion. In addition, antibodies to a second sequence, Po 74-82, and the peptide itself, also partially inhibited Po: Po-mediated adhesion indicating that there is more than one adhesive domain on Po-protein.

Myelin-associated glycoprotein inhibits axonal regeneration from a variety of neurons via interaction with a sialoglycoprotein.

Myelin-associated glycoprotein (MAG) is a potent inhibitor of axonal regeneration from both cerebellar neurons and adult dorsal root ganglion (DRG) neurons. In contrast, MAG promotes axonal growth from newborn DRG neurons. Here, we show that the switch in response to MAG from promotion to inhibition of neurite outgrowth by DRg neurons occurs sharply at Postnatal Day 3. To date, of all the neurons tested a postnatal switch in response is only observed for DRG neurons; MAG inhibits axonal growth from retinal, superior cervical ganglion, spinal, and hippocampal neurons of all postnatal ages. Furthermore, MAG binds to neurons from which it promotes and from which it inhibits outgrowth, in a sialic-acid-dependent manner. Now we show this binding is also trypsin-sensitive. Hence, the interaction is via a sialoglycoprotein. Binding of MAG to all the neurons tested here was also sialic-acid-dependent. Importantly, both inhibition and promotion of neurite outgrowth by MAG are reduced, or abolished completely, either by desialyation of the neurons prior to the outgrowth assay or by including small sialic-acid-bearing sugars in the cultures. These results suggest that MAG is likely to contribute to the lack of regeneration observed throughout the nervous system. Also, it is likely that MAG is exerting its effect, either directly or indirectly, on both promotion and inhibition of neurite outgrowth via a neuronal sialoglycoprotein.