Search

The myelin-associated glycoprotein (MAG) is selectively localized in periaxonal Schwann cell and oligodendroglial membranes of myelin sheaths suggesting that it functions in glia-axon interactions in the PNS and CNS, and this is supported by much experimental evidence. In addition, MAG is now well known as one of several white matter inhibitors of neurite outgrowth in vitro and axonal regeneration in vivo, and this latter area of research has provided a substantial amount of information about neuronal receptors or receptor complexes for MAG. This article makes the hypothesis that the capacity of MAG to inhibit outgrowth of immature developing or regenerating neurites is an aberration of its normal physiological function to promote the maturation, maintenance, and survival of myelinated axons. The overview summarizes the literature on the function of MAG in PNS and CNS myelin sheaths and its role as an inhibitor of neurite outgrowth to put this hypothesis into perspective. Additional research is needed to determine if receptors and signaling systems similar to those responsible for MAG inhibition of neurite outgrowth also promote the maturation, maintenance, and survival of myelinated axons as hypothesized here, or if substantially different MAG-mediated signaling mechanisms are operative at the glia-axon junction.

Two brothers with a leukodystrophy, progressive spastic diplegia, and peripheral neuropathy were found to have proteinaceous aggregates in the peripheral nerve myelin sheath. The patients' mother had only subclinical peripheral neuropathy, but the maternal grandmother had adult-onset leukodystrophy. Sequencing of the proteolipid protein (PLP) gene showed a point mutation IVS4 + 1 G-->A within the donor splice site of intron 4. We identified one transcript with a deletion of exon 4 (Deltaex4, 169bp) encoding for PLP and DM20 proteins and lacking two transmembrane domains, and a second transcript with exon 4 + 10bp encoding three transmembrane domains. Immunohistochemistry showed abnormal aggregation in the myelin sheath of MBP and P0. Myelin-associated glycoprotein was present in the Schmidt-Lanterman clefts but significantly reduced in the periaxonal region. Using immunogold electron microscopy, we demonstrated the presence of mutated PLP/DM20 and the absence of the intact protein in the patient peripheral myelin sheath. We conclude that insertion of mutant PLP/DM20 with resulting aberrant distribution of other myelin proteins in peripheral nerve may constitute an important mechanism of dysmyelination in disorders associated with PLP mutations.

Biochemical and cellular properties of three immortalized Schwann cell lines expressing different levels of the myelin-associated glycoprotein (MAG) were compared. The S16 line generated by repetitive passaging was described previously and expresses a level of MAG comparable to that in adult sciatic nerve. The S42 line was generated independently by the same procedure, divides more slowly than the S16 line, and expresses an even higher level of MAG. The S16Y line arose spontaneously from a passage of the S16 cells, divides much more rapidly, and does not express MAG. The levels of MAG expression in the three lines are inversely related to their rates of proliferation, and MAG mRNA levels parallel the amounts of MAG. The S16 and S42 lines consist mainly of flat cells at low density and develop many processes at high density, whereas most of the S16Y cells are spindle-shaped, resembling primary Schwann cells in appearance. Surface immunostaining with the O4 antibody was positive for the S16 and S42 cells and negative for the S16Y cells, but all three lines were negative for surface staining with the O1 antibody. The overall protein compositions of the three lines are very similar, but the S16 and S42 cells express larger amounts of several glycoproteins than the S16Y cells, including the adhesion proteins, neural cell adhesion molecule, L1, and laminin. S16 and S42 cells (but not S16Y cells) also express P0 glycoprotein, galactocerebroside, and sulfatide, but, unlike MAG, these other myelin-related components were present at much lower levels than in adult nerve. Myelin basic protein and proteolipid protein were not detected in any of the lines, although all three lines contained proteolipid protein mRNA. 2',3'-Cyclic nucleotide 3'-phosphodiesterase and glial fibrillary acidic protein were present in all three lines. Conditions have not yet been found in which any of the lines will myelinate dorsal root ganglion neurons in vitro, but the S16 and S42 cells differ from the S16Y cells by clustering around neurons after 1 week in coculture. In many respects, the S16 and S42 cells biochemically resemble Schwann cells at an early stage in their preparation to myelinate and should be useful for investigating the cell biology of MAG and other myelin-related components.

Several sulfated lipids were detected in the ganglioside fraction isolated from a cell line of oligodendrocyte progenitors that had been metabolically labeled with [35S] sulfate. Separation of the ganglioside fraction by two-dimensional TLC showed that, except for galactosylceramide-sulfate, none of the sulfate-labeled lipids comigrated with those glycosphingolipids visualized by orcinol staining, indicating that these sulfolipids were quantitatively minor components. At least eight sulfate-labeled lipid bands were susceptible to desialylation by Arthrobacter ureafaciens neuraminidase, which resulted in the formation of three new bands that retained the labeled sulfate. Six of the sulfate-labeled lipid bands containing sialic acid were also susceptible to Vibrio cholerae neuraminidase, which generated two labeled bands that appeared identical to the two major products formed after treatment with A. ureafaciens neuraminidase. In vivo labeling of lipids from 14-day-old rat brain with [35S]-sulfate demonstrated that the synthesis of sulfated lipids containing sialic acid also occurred in intact brain tissue. These results show that sulfated gangliosides are synthesized in the CNS and that oligodendrocytes are one cell type that contributes to this synthesis.

Hemispheres, spinal cords, and sciatic nerves were taken from taiep, carrier, and control rats at ages ranging from 1 day to 16 months. Absolute myelin yields from CNS taiep tissues peaked at approximately 2 months and then decreased until they reached a low but stable level. Myelin yield from the affected hemispheres expressed as a percentage of age-matched controls decreased continuously from 2 weeks until it reached a stable level of approximately 10-15%. The same was true for the spinal cords, but here the myelin yield reached a plateau at a slightly higher percentage of 20-25%. In comparison with control rats, isolated CNS myelin fractions from the affected rats had a greater content of high molecular weight proteins. Western blot analyses of CNS homogenates revealed that myelin basic protein (MBP), proteolipid protein, and 2',3'-cyclic nucleotide 3'-phosphodiesterase were all present but decreased to levels generally consistent with the deficiencies of myelin. However myelin-associated glycoprotein (MAG) levels always were reduced much more than those of the other three myelin proteins, and at younger ages the apparent molecular weight for MAG was increased in the mutants. Western blot analyses of sciatic nerve homogenates showed that the levels of MBP, MAG, and P0 were not significantly different in control and mutant animals. These results suggested an early hypomyelination of the CNS, with peak levels of myelin at 2 months, followed by a prolonged period of myelin loss, until a very low but stable myelin level was reached. The consistently greater loss of MAG, in comparison with other CNS myelin proteins, is different from most other hypomyelinating mutants in which MAG is relatively preserved in comparison with the proteins of compact myelin. This might be due to microtubular abnormalities in the taiep mutant interfering with transport of myelin proteins and having the greatest effect on MAG because of its most distal location in the periaxonal oligodendroglial membranes.

Decreased phosphorylation of neurofilaments in mice lacking myelin-associated glycoprotein (MAG) was shown to be associated with decreased activities of extracellular-signal regulated kinases (ERK1/2) and cyclin-dependent kinase-5 (cdk5). These in vivo changes could be caused directly by the absence of a MAG-mediated signaling pathway or secondary to a general disruption of the Schwann cell-axon junction that prevents signaling by other molecules. Therefore, in vitro experimental paradigms of MAG interaction with neurons were used to determine if MAG directly influences expression and phosphorylation of cytoskeletal proteins and their associated kinases. COS-7 cells stably transfected with MAG or with empty vector were co-cultured with primary dorsal root ganglion (DRG) neurons. Total amounts of the middle molecular weight neurofilament subunit (NF-M), microtubule-associated protein 1B (MAP1B), MAP2, and tau were up-regulated significantly in DRG neurons in the presence of MAG. There was also increased expression of phosphorylated high molecular weight neurofilament subunit (NF-H), NF-M, and MAP1B. Additionally, in similar in vitro paradigms, total and phosphorylated NF-M were increased significantly in PC12 neurons co-cultured with MAG-expressing COS cells or treated with a soluble MAG Fc-chimera. The increased expression of phosphorylated cytoskeletal proteins in the presence of MAG in vitro was associated with increased activities of ERK 1/2 and cdk5. We propose that interaction of MAG with an axonal receptor(s) induces a signal transduction cascade that regulates expression of cytoskeletal proteins and their phosphorylation by these proline-directed protein kinases.

Microtubule-associated protein (MAP) 1B is a high-molecular-weight cytoskeletal protein that is abundant in developing neuronal processes and appears to be necessary for axonal growth. Various biochemical and immunocytochemical results are reported, indicating that a significant fraction of MAP1B is expressed as an integral membrane glycoprotein in vesicles and the plasma membrane of neurons. MAP1B is present in microsomal fractions isolated from developing rat brain and fractionates across a sucrose gradient in a manner similar to synaptophysin, a well-known vesicular and plasma membrane protein. MAP1B is also in axolemma-enriched fractions (AEFs) isolated from myelinated axons of rat brain. MAP1B in AEFs and membrane fractions from cultured dorsal root ganglion neurons (DRGNs) remains membrane-associated following high-salt washes and contains sialic acid. Furthermore, MAP1B in intact DRGNs is readily degraded by extracellular trypsin and is labeled by the cell surface probe sulfosuccinimidobiotin. Immunocytochemical examination of DRGNs shows that MAP1B is concentrated in vesicle-rich varicosities along the length of axons. Myelinated peripheral nerves immunostained for MAP1B show an enrichment at the axonal plasma membrane. These observations demonstrate that some of the MAP1B in developing neurons is an integral plasma membrane glycoprotein.

Treatment of cultured rat oligodendroglial progenitors with either platelet-derived growth factor (PDGF) or fibroblast growth factor-2 (FGF-2) activated extracellular signal regulated kinase 2 (ERK2). Activation was transient in response to PDGF, whereas it was greater and more prolonged in response to FGF-2. ERK2 activation by PDGF was preceded by a very rapid, robust and transient tyrosine phosphorylation of the PDGF receptor. Although there was consistently more activation of ERK2 in response to FGF-2 than to PDGF, immunostaining of FGF receptors 1 (FGFR1) and 2 (FGFR2) and their tyrosine phosphorylation in progenitors was very weak, and both receptors were up-regulated during differentiation to oligodendrocytes. Tyrosine phosphorylation of the FGF receptors was maximal from 15 to 60 min of treatment and was sustained for many hours. Binding of radioiodinated FGF-2 to FGFR1 was predominant in progenitors, whereas binding to FGFR2 was predominant in oligodendrocytes. ERK2 activation by PDGF was more sensitive to inhibition of tyrosine kinases, whereas ERK2 activation by FGF-2 was relatively more sensitive to inhibitors of protein kinase C. These differences in signal transduction pathways probably contribute to the different cellular responses of oligodendroglial lineage cells to PDGF and FGF-2, respectively.

Myelin-associated glycoprotein (MAG) is expressed in periaxonal membranes of myelinating glia where it is believed to function in glia–axon interactions by binding to a component of the axolemma. Experiments involving Western blot overlay and coimmunoprecipitation demonstrated that MAG binds to a phosphorylated neuronal isoform of microtubule-associated protein 1B (MAP1B) expressed in dorsal root ganglion neurons (DRGNs) and axolemma-enriched fractions from myelinated axons of brain, but not to the isoform of MAP1B expressed by glial cells. The expression of some MAP1B as a neuronal plasma membrane glycoprotein (Tanner, S.L., R. Franzen, H. Jaffe, and R.H. Quarles. 2000. J. Neurochem. 75:553–562.), further documented here by its immunostaining without cell permeabilization, is consistent with it being a binding partner for MAG on the axonal surface. Binding sites for a MAG-Fc chimera on DRGNs colocalized with MAP1B on neuronal varicosities, and MAG and MAP1B also colocalized in the periaxonal region of myelinated axons. In addition, expression of the phosphorylated isoform of MAP1B was increased significantly when DRGNs were cocultured with MAG-transfected COS cells. The interaction of MAG with MAP1B is relevant to the known role of MAG in affecting the cytoskeletal structure and stability of myelinated axons.

Ultrastructural, biochemical, and electrophysiological analyses were done on 12–14-month-old mice deficient for myelin-associated glycoprotein (MAG) to further characterize the neuropathy that develops as they age. Electron microscopy demonstrated normal myelin compaction and axonal degeneration in a large number of myelinated nerve fibers. Western blots showed that the proteins of compact myelin, P0 glycoprotein, and myelin basic protein were not significantly altered in the mutants; however, the Schwann cell protein, 2′,3′-cyclic nucleotide 3′-phosphodiesterase, was reduced to less than half the control level. Also, both total and phosphorylated high-molecular-weight neurofilament proteins (TNFH and PNFH, respectively) were significantly decreased, as was the PNFH:TNFH ratio. Electrophysiological evaluation revealed a mild, but statistically significant, reduction of conduction velocity and a nonsignificant mild decrease in compound muscle action potential amplitudes. This constellation of findings in aging MAG-null mice is consistent with an axonopathy that resembles axonal Charcot–Marie–Tooth (CMT2) disease in many respects. Thus, mutation of a myelin-associated gene expressed by Schwann cells can induce axonal degeneration and cause a neuropathy with minimal signs of demyelination. © 2001 John Wiley & Sons, Inc. Muscle Nerve 24: 1380–1387, 2001

Although MAG-null mice myelinate relatively normally except for subtle structural abnormalities in the periaxonal region of myelin sheaths, they develop more severe pathological changes as they age. The purpose of this study was to further define the biochemical aspects of CNS pathology caused by an absence of MAG. Proteins associated with myelin and oligodendrocytes were quantified by densitometry of western blots in whole brain homogenates, as well as in isolated myelinated axons and myelin. Neither myelin yields, nor levels of myelin basic protein and proteolipid protein, were decreased in comparison to control levels in 14-month-old MAG-null mice. On the other hand, 2',3'-cyclic nucleotide 3'-phosphodiesterase (CNPase) and the 120 kD neural cell adhesion molecule (N-CAM) were substantially reduced in whole brain, myelinated axons, and myelin. Tubulin, Na(+)K(+)ATPase and Fyn tyrosine kinase were also reduced significantly in myelin-related fractions, but not in whole brain homogenate. The decreased levels of these proteins suggest pathological abnormalities in oligodendrocytes. Furthermore, significant reductions of CNPase and 120 kD NCAM were also present at 2 months, indicating that the oligodendroglial abnormalities begin at a relatively early age. Neither TUNEL assays nor multiplex RT-PCR for mRNAs of apoptosis-related proteins in the aging MAG-null mice provided evidence for apoptotic oligodendrocytes. These biochemical findings suggest oligodendroglial damage in MAG-null mice and support the morphological observations pointing to a progressive "dying-back oligodendrogliopathy" as a consequence of MAG deficiency.

Among the developmentally regulated antigens expressed on the surface of bipotential glial precursors isolated from neonatal rat brain are the gangliosides recognized by the monoclonal antibody A2B5. Immunostaining of thin layer chromatograms showed that oligodendrocyte-type 2 astrocyte (O2A) progenitors in culture express two ganglioside antigens that react strongly with the A2B5 antibody. Both ganglioside antigens are down-regulated as the cells differentiate to oligodendrocytes, corresponding to the disappearance of cell surface immunostaining by A2B5 in mature oligodendrocytes. By contrast, both gangliosides continue to be expressed when the cells differentiate to type-2 astrocytes. These two A2B5-reactive gangliosides in O2A lineage cells were identified as GT3 and O-acetyl GT3 by using the monoclonal antibody 18B8 that recognizes GT3 and an influenza C virus esterase that specifically removes O-acetyl moieties from sialic acids. Thin-layer chromatographic immunostaining with the JONES monoclonal antibody demonstrated that the progenitor cells in culture also express O-acetyl GD3, which is similarly down-regulated in oligodendrocytes, but retained in type-2 astrocytes. The 18B8 and JONES antibodies also immunostained the surface of O2A progenitors. Therefore, expression of GT3 and O-acetylation of GT3 and GD3 occur during the proliferative and migratory stages of glial cell development. Published 1999 Wiley-Liss, Inc.

High titers of serum antibodies to neural antigens occur in several forms of neuropathy. These include neuropathies associated with monoclonal gammopathy, inflammatory polyneuropathies, and paraneoplastic neuropathies. The antibodies frequently react with glycosylated cell surface molecules, including glycolipids, glycoproteins, and glycosaminoglycans, but antibodies to intracellular proteins have also been described. There are several correlations between antibody specificity and clinical symptoms, such as anti-MAG antibodies with demyelinating sensory or sensorimotor neuropathy, anti-GM1 ganglioside antibodies with motor nerve disorders, antibodies to gangliosides containing disialosyl moieties with sensory ataxic neuropathy and Miller-Fisher syndrome, and antibodies to the neuronal nuclear Hu antigens with paraneoplastic sensory neuronopathy. These correlations suggest that the neuropathies may be caused by the antibodies, but evidence for a causal relationship is stronger in some examples than others. In this review, we discuss the origins of the antibodies, evidence for and against their involvement in pathogenic mechanisms, and the implications of these findings for therapy.

Polyclonal immunoglobulin M antibodies to the monosialoganglioside GM2, sulfoglucuronyl glycolipids, and sulfatide were detected by thin-layer chromatography and enzyme-linked immunosorbent assay in the serum of a patient with melanoma and chronic inflammatory demyelinating polyneuropathy. Both the patient's serum and polyclonal antibodies against GM2 reacted strongly with a biopsy of melanomatous tissue from the patient, suggesting a process of molecular mimicry.

Antibodies reacting with carbohydrate epitopes on neural glycoconjugates are present in several forms of neuropathy. These include monoclonal antibodies to the myelin-associated glycoprotein (MAG) and to gan gliosides in patients with neuropathy in association with IgM gammopathy, as well as polyclonal antibodies to gangliosides in inflammatory polyneuropathies, such as Guillain-Barré syndrome and multifocal motor neuropathy. There are several correlations between antibody specificity and clinical symptoms, including anti-MAG antibodies with demyelinating sensory or sensorimotor neuropathy, anti-GM1 ganglioside anti bodies with motor nerve disorders, anti-GQ1b ganglioside antibodies with Miller-Fisher syndrome, and antibodies to gangliosides containing disialosyl moieties with sensory ataxic neuropathy. This review will emphasize recent developments concerning the origins of the anti-glycoconjugate antibodies in patients, pathogenic mechanisms by which the antibodies may cause the neuropathies, and the implications of these findings for therapy. NEUROSCIENTIST 3:195-204, 1997

A growing number of glycoproteins have been identified and characterized in myelin and myelin-forming cells. In addition to the major P0 glycoprotein of compact PNS myelin and the myelin-associated glycoprotein (MAG) in the periaxonal membranes of myelin-forming oligodendrocytes and Schwann cells, the list now includes peripheral myelin protein-22 (PMP-22), a 170 kDa glycoprotein associated with PNS myelin and Schwann cells (P170k/SAG), Schwann cell myelin protein (SMP), myelin/oligodendrocyte glycoprotein (MOG), and oligodendrocyte-myelin glycoprotein (OMgp). Many of these glycoproteins are members of the immunoglobulin superfamily and express the adhesion-related HNK-1 carbohydrate epitope. This review summarizes recent findings concerning the structure and function of these glycoproteins of myelin sheaths with emphasis on the physiological roles of oligosaccharide moieties.

The formation of basement membrane around Schwann cells that are in contact with axons is necessary for Schwann cell differentiation and myelin formation in the peripheral nervous system. However, primary Schwann cells grown on basement membrane in the absence of neuronal influence show increased proliferation rather than differentiation, which implies that the signals generated by Schwann cell-basement membrane interactions are multipotential. We examined the effect of matrigel, an exogenous basement membrane preparation, and other extracellular matrix growth surfaces on primary Schwann cells to determine if the resulting interactions play a role in the control of glycosphingolipid synthesis. Isolated primary Schwann cells grown on a thin layer of matrigel rapidly adhered to the surface and exhibited a greater degree of cell spreading when compared to cells grown on the nonspecific substrate polylysine. Labeling of the cells with [3H]galactose between 24 and 48 hr after plating revealed that the incorporation of [3H]galactose into glucosylceramide-based glycosphingolipids increased from 1.5-3-fold on matrigel in comparison to cells grown on polylysine. The major labeled glycolipids under both conditions were GM3 ganglioside and two neutral glycolipids that comigrated with GbOse4Cer (GalNAc beta 1-3Gal alpha 1-4Gal beta 1-1Cer) and GbOse5Cer (GalNAc alpha 1-3Gal-NAc beta 1-3Gal alpha 1-4Gal beta 1-4Glc beta 1-1Cer) standards. There was little or no increase in the incorporation of [3H]leucine, [3H]galactose, or [3H]glucosamine into proteins or [3H]palmitic acid into phospholipids, free ceramides, or sphingomyelin, suggesting that the matrigel-induced increase in the synthesis of the glycolipids was selective. In the absence of serum, there was little or no difference in the levels of glycolipid labeling between cells grown on the two substrata, demonstrating that serum factors were required for matrigel to have this effect. When cells were grown on surfaces coated with individual extracellular matrix components, those cells grown on laminin and collagen IV showed an increase in glycolipid labeling similar to that produced by matrigel, while labeling increased to a lesser degree for the other components tested. Thus, the signals generated by interactions between Schwann cells and basement membrane, particularly the laminin and collagen IV constituents, contribute to the regulation of glycolipid synthesis which in turn may affect cell morphology and proliferation.

The myelin-associated glycoprotein (MAG) exhibits an abnormally high apparent molecular weight in sciatic nerve, but not in brain, of dysmyelinating trembler mutants (Inuzuka et al.: J Neurochem 44:793-797, 1985). Antibodies to the large and small isoforms of MAG (L- and S-MAG) and probes for oligosaccharide structure were used to determine if this was due to overexpression of L-MAG or increased glycosylation. Nerves from both control and trembler 36-day-old mice contained primarily S-MAG with only traces of L-MAG. The distribution of the two isoforms appeared normal in trembler mice, and both isoforms exhibited the higher apparent molecular weight. Lectin binding showed that, in contrast to brain in which most glycoproteins contain primarily alpha 2-3 linked sialic acid, most glycoproteins of both control and trembler nerve contained primarily alpha 2-6 linked sialic acid. Lectin binding and glycosidase treatments demonstrated that the higher molecular weight of MAG in trembler nerves was due to an increased content of alpha 2-3 linked sialic acid and galactose. The abnormal glycosylation of MAG in trembler mutants may contribute to the myelin pathology.

Quaking is an autosomal recessive hypo/dysmyelinating mutant mouse which has a 1-Mbp deletion on chromosome 17. The mutation exhibits pleiotrophy and does not include genes encoding characterized myelin proteins. The levels of the 67-kD isoform of the myelin-associated glycoprotein (S-MAG) relative to those of the 72-kD isoform (L-MAG) are increased in the quaking CNS, but not in other dysmyelinating mutants. Abnormal expression of MAG isoforms in quaking may result from altered transcription of the MAG gene or from abnormal sorting, transport, or targeting of L-MAG or S-MAG. To test these hypotheses, we have determined the distribution of L-MAG and S-MAG in cervical spinal cord of 7-, 14-, 21-, 28-, and 35-d-old quaking mice. In 7-d-old quaking and control spinal cord, L- and S-MAG was detectable in periaxonal regions of myelinated fibers and in the perinuclear cytoplasm of oligodendrocytes. Between 7 and 35 d, L-MAG was removed from the periaxonal membrane of quaking but not control mice. Compared to control mice, a significant increase in MAG labeling of endosomes occurred within oligodendrocyte cytoplasm of 35-d-old quaking mice. S-MAG remained in periaxonal membranes of both quaking and control mice. Analysis of the cytoplasmic domain of L-MAG identifies amino acid motifs at tyrosine 35 and tyrosine 65 which meet the criteria for "tyrosine internalization signals" that direct transmembrane glycoproteins into the endocytic pathway. These results establish that L-MAG is selectively removed from the periaxonal membrane of CNS-myelinated fibers by receptor-mediated endocytosis. The loss of L-MAG from quaking periaxonal membranes results from increased endocytosis of L-MAG and possibly a decrease in L-MAG production.

The relative expression of large (L) and small (S) iso-forms of the myelin-associated glycoprotein (MAG) and their glycosylation were compared in developing spinal cord of quaking and control mice. Using anti-sera specific for L- and S-MAG, respectively, it was shown that S-MAG is the principal isoform in quaking mice at all ages between 13 and 72 days, although L-MAG was just detectable by western blotting at the early ages. Both L- and S-MAG have higher apparent molecular weights in quaking mice than in controls. Experiments involving lectin binding and glycosidase treatment demonstrated that the higher molecular weight of MAG in the quaking mutant was due to a higher content of N-acetyineuraminic acid residues linked α2-3 to galactose as well as to more branching of oligosaccharide moieties indicated by a higher content of subterminal galactose residues. The total sialic acid measured by HPAE-chromatography in purified quaking MAG was 40% higher than in control MAG. By contrast, quaking MAG contained less of the adhesion-related, HNK-1 carbohydrate epitope. Another difference was that a lower molecular weight form of MAG with predominantly high mannose oli-gosaccharides was prominent in young quaking mice, but not in controls. The abnormalities of MAG expression related to splicing of its mRNA and glycosylation may contribute to the myelin pathology in quaking mutants. © 1995 Wiley-Liss, Inc.

Cultures consisting primarily of O-2A progenitor cells and immature oligodendrocytes with a few microglia and astrocytes were obtained by shaking primary cultures from neonatal rat brain after 12-14 days in vitro. Addition of 50 micrograms/ml exogenous Neu-NAc alpha 2-3Gal beta 1-4Glc beta 1-1'ceramide (GM3 ganglioside) to the cultures resulted in an increase in the number and thickness of cell processes that stained intensely for sulfatide and galactocerebroside (galC) in comparison to control cultures without added GM3. The treated cultures also contained fewer astrocytes than control cultures as revealed by immunostaining for glial fibrillary acidic protein (GFAP). Cells that immunostained for both GFAP and sulfatide/galC were very rare in control cultures but were frequently seen in the GM3-treated cultures, suggesting that these may represent cells changing their direction of differentiation away from type II astrocytes toward oligodendrocytes under the influence of GM3. These effects on the developing rat oligodendrocytes were specific for GM3 ganglioside and were not produced by adding GM1, GM2, GD3, or GD1a to the cultures. Lactosyl ceramide and neuraminyl lactose were also ineffective. When control cultures were initially plated on polylysine and incubated with [14C]galactose, GD3 was the principal labeled ganglioside. However, as the control cells differentiated over time in culture without the addition of exogenous GM3 and produced increasing amounts of myelin-related components, the incorporation of [14C]galactose into endogenous GM3 increased to become the predominant labeled ganglioside by 6 days after plating. Metabolic labeling of the GM3-treated oligodendrocytes with [14C]galactose revealed increased incorporation into galC and sulfatide in comparison to control cultures, but a decreased labeling of endogenous GM3. Similarly, incorporation of an amino acid precursor into the myelin-associated glycoprotein (MAG) was increased by GM3 treatment, but incorporation into myelin basic protein (MBP) was not affected. Although the overall effect of added GM3 was to decrease the phosphorylation of most proteins in the oligodendrocytes, including MBP, GM3 enhanced the phosphorylation of MAG. These findings indicate that GM3 ganglioside has an important role in the differentiation of cells of the O-2A lineage toward myelin production, since differentiation is associated with increased metabolic labeling of endogenous GM3 in control cultures and is enhanced by the addition of exogenous GM3.

An autoantibody occurring in the serum of an apparently normal rabbit that immunocytochemically stains myelin sheaths and oligodendrocytes in rat brain was shown to react specifically with the 46-kDa isoform of 2′,3′-cyclic nucleotide 3′-phosphodiesterase (CNP) (EC 3.1.4.37) in a number of species. Identification of the shorter isoform of the enzyme (CNP1) as the antigen was achieved by comparing the immunostaining of Western blots by the autoantibody with that of a well-characterized anti-CNP antiserum. The 46-kDa antigen reacting with the autoantibody exhibited the same Mr and pI as the small isoform of CNP on two-dimensional gels and showed a similar enrichment in purified CNS myelin. The autoantibody has very high affinity for CNP1 and is capable of detecting the very low amounts of this enzyme in peripheral nerve, spleen, adrenal gland, pancreas, testis, and intestine. Testing the reactivity of the autoantibody with synthetic peptides by enzyme-linked immunosorbent assay revealed that it reacted with the N-acetylated decapeptide corresponding to the N-terminus of CNP1, but did not react if the peptide was not acetylated or if the acetyl group was replaced with a palmityl group. The lack of reactivity with CNP2, which differs from CNP1 by a 20-amino acid extension at the N-terminus of the protein as a result of alternative splicing, may be due to the absence of the N-acetyl moiety that is part of the epitope and/or blocking of antibody binding to the decapeptide by extension of the polypeptide chain. This highly specific antibody for the shorter isoform of CNP may be useful in studies to establish possible differences in the location or function of the two isoforms of this enzyme.

Isolated adult bovine oligodendrocytes maintained in vitro for 10 days were treated for 1 day with 50 micrograms/ml of GM3 ganglioside (NeuNac alpha 2-3Gal beta 1-4Glc beta 1-1'ceramide) in serum-free culture medium. The treated oligodendrocytes had significantly longer processes with more branching than control cells in the same medium without GM3. The treatment also stimulated the release of a series of 22-100-kDa, [3H]glucosamine-labeled glycoproteins into the culture medium. Treatment of oligodendrocytes maintained in vitro for 50 days with GM3 for 1 day resulted in a thickening of the processes and the appearance of many fine branches on existing processes as well as a similar stimulation of glycoprotein release into the medium.

The major 24- and 28-kDa glycoproteins in shark PNS and CNS myelin express high levels of the adhesion-associated HNK-1/L2 carbohydrate epitope. The 28-kDa protein, but not the 24-kDa protein, cross-reacts strongly with one of two anti-bovine P0 antisera not previously tested against fish myelin proteins. Shark PNS and CNS myelin also contains smaller amounts of high-molecular-weight HNK-1-positive proteins, including a prominent broad band in the 65-85-kDa range. Although myelin-associated glycoprotein (MAG) is well known to react with HNK-1 in some mammals, monoclonal and polyclonal anti-MAG antibodies did not react with the high-molecular-weight HNK-1-positive material in shark myelin, a result suggesting that it is not a MAG-like protein. The high expression of the HNK-1/L2 epitope in glycoproteins of shark myelin, including the major P0-related ones, suggests that this adhesion-related carbohydrate structure may have had an important role in the molecular evolution of the myelinating process.

Phosphorylation of the myelin-associated glycoprotein (MAG) in the peripheral nervous system is demonstrated by immunoprecipitation from myelin proteins radiolabeled in vivo , in nerve slices and in a cell-free system. Phosphoamino acid analysis of immuno-precipitated MAG revealed the presence of radioactivity in phosphoserine, but not in phosphothreonine or phosphotyrosine. Only the shorter isoform of MAG (S-MAG) was detected by immunostaining of nitrocellulose sheets with anti-MAG anti-serum after enzymatic deglycosylation of immunoprecipitated MAG labeled in nerve slices. Autoradiography of the same Western blots revealed that most of the radioactive phosphate was in S-MAG, demonstrating that the polypeptide backbone of S-MAG is phosphorylated in the PNS.

IgM paraproteins in about 50% of the patients with neuropathy associated with IgM gammopathy react with carbohydrate moieties in myelin-associated glycoprotein (MAG) and in sulfated glucuronic glycolipids (SGGLs) in human peripheral nerves. However, the role of anti-MAG/SGGL antibodies in the pathogenesis of neuropathy remains unclear. In order to induce an animal model of neuropathy associated with anti-MAG/SGGL antibodies, cats were immunized with sulfoglucuronyl paragloboside (SGPG). All four cats immunized with SGPG developed clinical signs of sensory neuronopathy within 11 months after initial immunization, characterized by unsteadiness, falling, hind limb weakness and ataxia. In two cats the ataxia and hind limb paralysis were so severe that the animals had to be euthanized. Pathological examination revealed sensory ganglionitis with inflammatory infiltrates in the dorsal root ganglia. No overt signs of pathology were noted in the examined roots or nerves. High titer anti-SGPG/MAG antibodies were detected in all 4 cats immunized with SGPG but not in 3 control cats. Our data demonstrate that immunization of cats with SGPG induced anti-SGPG antibodies and sensory neuronopathy clinically resembling the sensory ataxia of patients with monoclonal IgM anti-MAG/SGPG antibodies. This study suggests that these anti-MAG/SGPG antibodies play a role in the pathogenesis of this neuropathy.

The objective of this chapter is to provide an overview of the lipids and proteins in myelin, which will serve as a foundation for other chapters in this section that cover neuroimmunogical aspects of disorders affecting myelin. The structure and functions of lipids and proteins of myelin are summarized. Other proteins localized in associated glial membranes, which have important roles in the formation and maintenance of myelinated axons, are also described. Finally, the roles of lipids and proteins as target antigens in the immunological aspects of demyelinating diseases are discussed. Keywords: multiple sclerosis; myelin; myelin-associated glycoprotein; myelin basic protein; myelin-oligodendrocyte glycoprotein; neuropathy; oligodendrocyte; protein zero; proteolipid protein; Schwann cell

This brief overview summarizes some of the ways that the structure, location and function of proteins in myelin sheaths affect the pathological abnormalities brought on by immunological, viral, genetic or other insults occurring in myelin disorders. The composition of compact myelin membranes is novel in comparison to other biological membranes, because of its high lipid content and the presence of several major proteins that are relatively specific for myelin [1,2]. The overall morphological structures of compact central nervous system (CNS) and peripheral nervous system (PNS) myelin are similar, even though they are formed by different cell types, that is, oligodendrocytes and Schwann cells, respectively. On the other hand, there are morphological differences in their periodicity and relationship to the myelin-forming glial cells, quantitative differences in lipid composition, and important qualitative differences in protein composition. These differences in protein composition are likely to be especially important for determining the pathologies that characterize disorders of myelin formation and degeneration in the CNS and PNS. Proteolipid protein (PLP) is a very hydrophobic tetraspan protein that accounts for over half the total protein in CNS myelin [1,3], whereas the amount of PLP in the PNS is minimal. By contrast, the type 1 transmembrane P0 glycoprotein accounts for over half the protein in PNS myelin [1,4,5]. Compact PNS myelin also contains a 22kDa tetraspan glycoprotein, peripheral myelin protein-22 (PMP-22), which accounts for less than 5% of the total and is absent in CNS myelin [1,4,6]. Myelin basic protein (MBP) is a prominent, positively charged, extrinsic membrane protein of both CNS and PNS myelin, which accounts for about 30% of the total in CNS myelin, but only 5–15%

Members of the immunoglobulin superfamily include a large number of cell adhesion molecules and receptors that are important for mediating interactions between cells in the nervous system. Keywords: cell adhesion molecule (CAM); glia; neural cell adhesion molecule (N-CAM); neuron; signal transduction

Primary and immortalized cultured Schwann cells are commonly utilized in analyses of myelin gene promoters, but few lines are well-characterized in terms of their endogenous expression of myelin genes. This is particularly significant in that cultured Schwann cells typically do not express myelin genes at levels comparable to those observed in vivo. In this study, the steady-state levels of mRNA and protein for five Schwann cell markers (PMP22, P0, MBP, MAG, and LNGF-R) were assessed in primary Schwann cells and six representative Schwann cell lines (RT4-D6P2T, JS-1, RSC96, R3, S16, and S16Y). RT4-D6P2T and S16 cells were the most similar to myelinating Schwann cells based on their comparatively high expression of PMP22 and P0 mRNA. Both RT4-D6P2T and S16 also expressed P0 protein. In addition, the previously reported P1-A positive regulatory region from the myelination-specific PMP22 promoter demonstrated significant activity in both these cell lines. However, nuclear proteins that interacted with P1-A were different in extracts prepared from RT4-D6P2T and S16 cells. Primary Schwann cells expressed myelin proteins at levels that were equal or less than those observed with the RT4-D6P2T and S16 lines, indicating that primary Schwann cells are not necessarily better than immortalized Schwann cells as model systems for the study of myelin gene regulation. The data presented here demonstrate that cultured Schwann cells used to study myelin gene promoters have to be carefully selected on the basis of the endogenous level of expression of the myelin gene under study.

Decreased phosphorylation of neurofilaments in mice lacking myelin-associated glycoprotein (MAG) was shown to be associated with decreased activities of extracellular-signal regulated kinases (ERK1/2) and cyclin-dependent kinase-5 (cdk5). These in vivo changes could be caused directly by the absence of a MAG-mediated signaling pathway or secondary to a general disruption of the Schwann cell-axon junction that prevents signaling by other molecules. Therefore, in vitro experimental paradigms of MAG interaction with neurons were used to determine if MAG directly influences expression and phosphorylation of cytoskeletal proteins and their associated kinases. COS-7 cells stably transfected with MAG or with empty vector were co-cultured with primary dorsal root ganglion (DRG) neurons. Total amounts of the middle molecular weight neurofilament subunit (NF-M), microtubule-associated protein 1B (MAP1B), MAP2, and tau were up-regulated significantly in DRG neurons in the presence of MAG. There was also increased expression of phosphorylated high molecular weight neurofilament subunit (NF-H), NF-M, and MAP1B. Additionally, in similar in vitro paradigms, total and phosphorylated NF-M were increased significantly in PC12 neurons co-cultured with MAG-expressing COS cells or treated with a soluble MAG Fc-chimera. The increased expression of phosphorylated cytoskeletal proteins in the presence of MAG in vitro was associated with increased activities of ERK 1/2 and cdk5. We propose that interaction of MAG with an axonal receptor(s) induces a signal transduction cascade that regulates expression of cytoskeletal proteins and their phosphorylation by these proline-directed protein kinases.

The spontaneously immortalized S16 Schwann cell line expresses higher levels of myelin-associated glycoprotein (MAG) and PO glycoprotein and their messenger RNAs when grown at high density than at low density [Sasagasako et al: J Neurochem 1996;66:1432–1439]. This up-regulation of myelin protein expression at high density is not associated with decreased cellular proliferation and may be caused by direct cell-to-cell contact. To investigate the hypothesis that increased mRNA levels for myelin proteins are caused by contact between Schwann cells, sparse S16 cell cultures were treated for 48 h with plasma-membrane-enriched fractions isolated from dense S16 cells. The treatment had no effect on the proliferation of the cells, but MAG and PO mRNAs were elevated 2- and 1.3-fold, respectively, in comparison to untreated cells. These effects on levels of myelin protein mRNAs were eliminated by pretreatment of the membrane fraction with heat or trypsin and were not caused by plasma membrane fractions from NIH 3T3 cells. These data support the hypothesis that homotypic contact between Schwann cells up-regulates expression of myelin proteins and suggest the possibility of autotypic contact-mediated regulation of myelinogenesis by adjacent spiraled membranes of individual Schwann cells.

The glycoprotein component in rat brain reacting most strongly with Galanthus nivalis agglutinin (GNA) on western blots migrates as an 85-kDa band. GNA identifies mannose-rich oligosaccharides because it is highly specific for terminal alpha-mannose residues. After purification of this 85-kDa glycoprotein band by chromatography on GNA-agarose and preparative gel electrophoresis, binding of other lectins demonstrated the presence of fucose and a trace of galactose, but no sialic acid. Treatment with N-Glycanase or endoglycosidase H produced a 65-kDa band, indicating that it consisted of about one-fourth N-linked oligomannosidic carbohydrate moieties. High-performance anion-exchange chromatography and fluorescence-assisted carbohydrate electrophoresis indicated that the major carbohydrate moiety is a heptasaccharide with the structure Manalpha1-6(Manalpha1-3)Manalpha1-6(Manalpha1-3) Manbeta1-4Glc-NAcbeta1-4GlcNAc (Man5GlcNAc2). Determination of amino acid sequences of peptides produced by endoproteinase digestion demonstrated that this 85-kDa mannose-rich glycoprotein component contained the SHP substrate-1 for phosphotyrosine phosphatases and at least one other member of the signal-regulatory protein (SIRP) family. The unusually high content of oligomannosidic carbohydrate moieties on these receptor-like members of the immunoglobulin superfamily in neural tissue could be of functional significance for intercellular adhesion or signaling.

Eleven patients with demyelinating polyneuropathy associated with monoclonal IgM antibodies were randomized to receive IVIg or placebo, monthly, for 3 months in a double-blind study. After a washout period, they crossed over to the alternate therapy. Response was gauged by evaluating muscle strength, sensation, and neuromuscular symptoms at baseline, after 3 months, and at treatment's end. After IVIg therapy, the strength improved in only 2 of 11 patients, by 28 and 38.5 points from baseline, and declined after placebo. In 1 other patient, the sensory score improved by 13 points. Antibody titers to MAG/SGPG or gangliosides did not appreciably change. We conclude that IVIg has only a modest benefit to not more than 18% of patients with IgM paraproteinemic demyelinating neuropathy.

Myelin gene expression was investigated in the immortalized S16 Schwann cell line grown in the presence and absence of serum and at different densities. Protein expression was monitored by western blotting, and message levels were determined by RNase protection assays. To study cell proliferation rates at different cell densities and serum conditions, [3H]thymidine uptake assays and cell counts were performed. Although serum deprivation decreased cell proliferation as expected, the proliferation of S16 cells was unchanged or slightly increased at high density under the conditions of our experiments in either serum-containing or serum-free medium. This increased cell division at high density appeared to be due to greater release of an autocrine growth factor to the medium by dense cell populations. For both sparse and dense cells, substantially more P0 glycoprotein (P0) and myelin-associated glycoprotein (MAG) per milligram of total cellular protein were expressed when the cells were proliferating slowly in defined medium in comparison with more rapidly proliferating cells in serum-containing medium. Furthermore, in both serum-containing and defined media, dense cell populations expressed more MAG and PO than sparse ones. PO mRNa and MAG mRNA levels generally paralleled protein levels. The level of mRNA for peripheral myelin protein-22 (PMP-22) was also increased at high cell density but did not change much when proliferation was decreased by serum deprivation. PMP-22 protein was not detected under any of the growth conditions. The changes in expression of these genes with growth conditions may be specific for myelin proteins, because the expression of a nonmyelin glycoprotein, L1, remained constant. The level of cyclic AMP in the cells did not change with the different growth conditions tested. The results indicate that the S16 Schwann cell line mimics primary or secondary Schwann cells by down-regulating myelin gene expression when it proliferates more rapidly in the presence of serum. Furthermore, in both the presence and absence of serum, there was greater expression of myelin genes at high cell density that was not associated with a decreased proliferative rate. Because evidence for a role of secretory factors in affecting myelin gene expression was not obtained by treating sparse S16 cells with medium conditioned by dense S16 cells, the results suggest that the higher expression of myelin genes at high density may be mediated by cell-to-cell contact.

A significant number of patients with progressive leukodystrophy do not have a definitive diagnosis. This report describes the clinical, morphological, and biochemical characteristics of 4 unrelated girls with progressive ataxic diplegia of unknown etiology. These patients had normal development until the ages of 1.5 to 5 years. A diffuse confluent abnormality of the white matter of the central nervous system was present on computed tomography and magnetic resonance scans obtained early in the course of the illness. Dementia was not present and peripheral nerves were normal. All patients were evaluated for known metabolic and degenerative diseases and no abnormalities were observed. Light and electron microscopy of open-brain biopsy specimens from 2 girls showed selective white matter abnormalities including hypomyelination, demyelination, and gliosis. Myelin-specific proteins in the subcortical white matter were examined immunocytochemically and by Western blot analysis. They were of normal molecular size but were markedly reduced in quantity in both patients compared to control subjects. Lipid analysis revealed decreased levels of characteristic myelin lipids. When examined by magnetic resonance spectroscopic imaging, all patients showed a marked decrease of N-acetylaspartic acid, choline, and creatine in white matter only. The magnetic resonance spectroscopic imaging profile is a unique diagnostic feature of this clinicopathological syndrome.

Mammalian peripheral nervous system (PNS) myelin contains several glycoproteins with molecular weights of 19 to 28 kDa, including the major 28 kDa P0 glycoprotein and a recently cloned protein called PMP22. Some glycoproteins in this Mr range in humans, cats and some other mammals react with HNK1, a mouse monoclonal antibody that identifies a carbohydrate epitope shared between the immune system and a number of adhesion proteins in the nervous system. A variety of antibodies to PO, PMP-22, and the carbohydrate determinants reacting with HNK1 were used to characterize immunochemically these 19 to 28 kDa glycoproteins of cat PNS myelin. The HNK1-reactive components include P0 and two slightly smaller 23 to 26 kDa proteins that are immunologically related to PO. However, HNK1 reacts most strongly with a lower molecular weight glycoprotein that does not react with the antibodies to P0 and was identified as PMP-22. Since the carbohydrate structure reacting with HNK1 is generally expressed on adhesion molecules, this result suggests that PMP-22 may function in cell-cell or membranemembrane interactions. Furthermore, the related human anti-MAG monoclonal lgM antibodies from patients with neuropathy also react strongly with PMP-22, suggesting that it may be a target antigen in the pathogenesis of this disease. Purified PNS and CNS myelin from bony fish (toadfish and trout) were also shown to contain major glycoproteins, in the same 19 to 28 kDa Mr range, that react very strongly with HNK1. It is known that fish myelin has major proteins of this size that are immunologically and structurally related to mammalian P0, and it is demonstrated here that one of the strongly HNK1-positive proteins reacted well with an antiserum raised to bovine P0. The presence of high levels of the adhesion-related HNK1 epitope on these major myelin proteins of fish suggests that this carbohydrate structure may have played a role in the molecular evolution of myelin. © 1993 Wiley-Liss, Inc.1

Although the myelin-associated glycoprotein (MAG) cannot be detected in primary cultures of rat Schwann cells in the absence of neurons, MAG expression was demonstrated in some lines of cultured Schwann cells that had been immortalized by repetitive passaging. Radioimmunoassay of one such Schwann cell line, S-16, showed a remarkably high MAG concentration of about 1 ng/microgram of total protein, a level that is comparable to the MAG concentration in adult sciatic nerve. The S-16 cells divide very rapidly, are rounder than normal Schwann cells, and elaborate many processes after reaching high density. The cells are galactocerebroside positive, but express little or no P0 glycoprotein or myelin basic protein. As in nerve, the MAG synthesized by the cultured cells is primarily the shorter isoform (S-MAG). Furthermore, the posttranslational processing resembles that occurring in vivo including a similar degree of glycosylation, sulfation of oligosaccharides, and phosphorylation of the polypeptide. The sensitivity of MAG to treatment of the intact cells with trypsin or neuraminidase, as well as surface labeling with [3H]borohydride reduction after periodate oxidation, demonstrated that most of the MAG expressed by the S-16 cells is located on the cell surface. This line of immortalized Schwann cells expressing a remarkably high level of MAG should be useful for investigating the cell biology and function of this glycoprotein.

A high proportion of patients with neuropathy have immunoglobulin M (IgM) paraproteins that react with carbohydrate determinants on the myelin-associated glycoprotein (MAG) and two sphingoglycolipids, 3-sulfoglucuronyl paragloboside (SGPG) and 3-sulfoglucuronyl lactosaminyl paragloboside. In order to characterize the fine specificities of these human antibodies further, the binding of 10 anti-MAG paraproteins to several chemically modified derivatives of SGPG was compared with the binding to intact SGPG by both TLC-overlay and enzyme-linked immunosorbent assay. The following derivatives were tested: the desulfated lipid, glucuronyl paragloboside (GPG); the methyl ester of GPG (MeGPG); the methyl ester of SGPG, 3-sulfomethylglucuronyl paragloboside (SMeGPG); and 3-sulfoglucosyl paragloboside (SGlcPG) produced by reduction of the carboxyl group of the glucuronic acid with sodium borohydride. All 10 IgM paraproteins and the related mouse IgM antibody, HNK-1, reacted most strongly with intact SGPG, but variations in the reactivity with the derivatives revealed striking differences in the structural requirements for binding between the antibodies. Five distinct patterns of reactivity were observed: (a) three of the human antibodies and HNK-1 exhibited partial reactivity with the sulfated derivatives, SMeGPG and SGlcPG, but not with GPG or MeGPG, indicating an absolute requirement for the sulfate group; (b) two of the human antibodies reacted only with GPG, demonstrating a requirement for the free carboxyl group on the glucuronic acid; (c) three of the antibodies bound to SMeGPG, SGlcPG, and GPG, but not to MeGPG, suggesting that at least one negative charge was needed for binding; (d) one antibody bound to SMeGPG and GPG, but not to SGlcPG or MeGPG, suggesting that both the carbonyl group and at least one negative charge were required; and (e) another antibody bound to MeGPG, SMeGPG, and SGlcPG, but not to GPG, a pattern that is difficult to explain simply based on the chemical structures. Interestingly, only those antibodies that exhibited reactivity with GPG bound to a third, minor, unidentified glycolipid of normal peripheral nerve that chromatographs faster than SGPG. The results clearly demonstrate heterogeneity in the fine specificities of the anti-MAG antibodies that may affect their pathogenic properties.

An earlier investigation from our laboratory (Ilyas AA, Wilson HJ, Quarles RH, et al. Serum antibodies to gangliosides in Guillain-Barre syndrome. Ann Neurol 1988;23:440-447) demonstrating the presence of high levels of antiganglioside ganglioside antibodies in the sera of 5 of 26 patients with Guillain-Barre Syndrome (GBS) but not in control sera is summarized. The ganglioside antigens varied among the 5 patients with positive findings, and the antiganglioside antibodies decreased concurrently with clinical improvement in those patients for whom longitudinal samples were available for analysis. The results are discussed in the context of antibodies to acidic glycolipids in other types of neuropathy and other studies on antiglycolipid antibodies in GBS. Data showing the occurrence of lower levels of antibodies to P2 protein, P0 glycoprotein, and myelin-associated glycoprotein in some of the GBS patients are also summarized. The findings from our laboratory combined with the results of others make it unlikely that antiganglioside antibodies have a notable pathogenic effect in most patients with GBS, but the possibility remains that they are of pathogenic importance in some patients with the highest antibody titers.

The myelin-associated glycoprotein (MAG) is a type I transmembrane glycoprotein localized in periaxonal Schwann cell and oligodendroglial membranes of myelin sheaths where it functions in glia-axon interactions. It contains five immunoglobulin (Ig)-like domains and is in the sialic acid-binding subgroup of the Ig superfamily. It appears to function both as a ligand for an axonal receptor that is needed for the maintenance of myelinated axons and as a receptor for an axonal signal that promotes the differentiation, maintenance and survival of oligodendrocytes. Its function in the maintenance of myelinated axons may be related to its role as one of the white matter inhibitors of neurite outgrowth acting through a receptor complex involving the Nogo receptor and/or gangliosides containing 2,3-linked sialic acid. MAG is expressed as two developmentally regulated isoforms with different cytoplasmic domains that may activate different signal transduction pathways in myelin-forming cells. MAG contains a carbohydrate epitope shared with other glycoconjugates that is a target antigen in autoimmune peripheral neuropathy associated with IgM gammopathy and has been implicated in a dying back oligodendrogliopathy in multiple sclerosis.

The monoclonal HNK-1 antibody, a marker for human natural killer cells, strongly reacted with human peripheral nerve gangliosides in the enzyme-linked immunosorbent assay. Autoradiography after the binding of HNK-1 to thin-layer chromatograms of peripheral nerve gangliosides followed by radioiodinated goat anti-mouse IgM revealed that HNK-1 was reacting with a minor ganglioside that chromatographed between GM1 and GD1a. The antigen was insensitive to digestion with neuraminidase and pronase.