Search

Two polypeptide isoforms of myelin-associated glycoprotein (MAG) with molecular masses of 72 KDa (large MAG, L-MAG) and 67 kDa (small MAG, S-MAG) are produced by alternative splicing of the exon 12 portion of the MAG gene in rodent brain. We prepared antisera specific to the L-MAG and the S-MAG amino acid residues, respectively. Immunohistochemical staining of developing mouse brain showed that the L-MAG antiserum stained the myelin sheaths and the cytoplasm of the oligodendrocytes in the rapid myelinating stage, while the S-MAG antiserum stained only the myelin sheathes in the adult stage. The L-MAG positive cells were also present in the white matter around the experimental cerebral infarcts produced by occlusion of the left middle cerebral artery in the rat 3 days before observation. These finings suggest that L-MAG protein play an important role in myelin formation during rapid myelination in development and remyelination.

cDNA cloning of human myelin-associated glycoprotein (MAG) revealed complete amino acid sequence. In rodents, expression of the two forms of mRNA is developmentary regulated; the mRNA without exon 12 portion is expressed mainly in the active myelinating stage of development. Although the cDNA library used here was prepared from adult human brain poly (A)+ RNA all clones obtained corresponded to the mRNA without exon 12 portion. It has been demonstrated that the myelin membrane contains an endogenous calcium activated neutral protease (CANP) which degrades myelin basic protein (MBP) and MAG. The myelin endogenous CANP cleaves the 100KDa MAG molecule to a 90KDa derivative called derivative of MAG (dMAG). Since dMAG does not appear to be as tightly bound to membranes as intact MAG, the CANP is supposed (not proven) to cleave C-terminus of MAG. To confirm the proteolytic cleavage sites in MAG molecules, we raised antisera against the sequences of C-terminus of MAG. The peptides unique to large form of MAG (L-MAG) and small form of MAG (S-MAG) were synthesized using an automatic peptide synthesizer. We incubated purified myelin to produce dMAG and examined each anti-MAG antibody. Both anti-L-MAG and anti-S-MAG antibodies did not recognize dMAG. These results indicate that dMAG does not contain the sequences of C-termini in both L-MAG and S-MAG. Phosphorylation sites are thought to be important for functions of MAG as a recognition molecule and located near C-terminus of the cytoplasmic sequence. As dMAG lacks C-terminus, production of dMAG may result in the disturbance of myelin-axon interactions during demyelinating process.

Calcium-dependent protease was extracted from human brain myelin membrane and partially purified on a phenyl Sepharose CL 4B column. It was activated by calcium ion in the millimolar range, and therefore was determined to be calpain II. This enzyme fraction was electrophoresed and immunostained with anti-chicken muscle calpain antibody, resulting in staining as a single band with apparent molecular weight of 80K. This protease degraded exogenous myelin-associated glycoprotein. From the present result, it is suggested that calpain is bound to myelin membranes and involved in the turnover of myelin proteins. The vacuolar degeneration was produced by oral administration of triethyl tin （TET）. The wet weight of brain stems which seems to reflect the degree of accumulation of water increased during the administration of the toxin, whereas the activity of 2', 3'-cyclic nucleotide 3'-phosphodiesterase altered less remarkably. When TET was withdrawed from the drinking water, the rats showed a dramatic clinical improvement along with reduction in wet weight of brain stems. Treatment with acetazolamide following TET inhibited the clinical improvement and reduction in wet weight of brain stems. The present results indicates that central myelin has plasticity in recovering from the vacuolar degeneration by removing the accumulated fluid and carbonic anhydrase is possibly involved in the dehydration of myelin in such a recovery phase.

The structures of three forms of mouse myelin-associated glycoprotein(MAG) mRNAs were determined from full-length cDNA clones. Two forms of mRNAs have been reported to be different by alternate inclusion of exon 2 and 12 portions in the rat brain. One of the three forms of clones obtained here appeared to be a novel mRNA which lacked both the exon 2 and 12 portions, although others were identical splicing patterns to those of rat. Two polypeptide isoforms of MAG with molecular masses of 72 and 67 kDa are produced by alternative splicing of the exon 12 portion containing a termination codon. Evidence was presented that expression of the two MAG mRNAs was developmentally regulated in the mouse brain. In the quaking mouse, the mRNA without the exon 12 portion coding the large MAG isoform was scarcely expressed throughout development. Specific antibodies to L-MAG and S-MAG respectively were prepared, and immunoblots showed that the L-MAG band was scarcely detectable in the quaking mouse brain, whereas the S-MAG had a higher apparent molecular mass than in the normal control. Immunohistochemical study also showed that L-MAG was not stained in the quaking mouse brain. These results seemed to reflect a reduction in the L-MAG mRNA and abnormal glycosylation in the quaking mouse brain.

Three (GC-J4, MC-P2 and MC-P4) out of five murine monoclonal antibodies (mAbs) to myelin-associated glycoprotein (MAG) bound to human mononuclear cells (MNC). The results of the competitive inhibition assay suggested that MC-P2 and MC-P4 bind directly to or close to the Leu 7 epitope, and that GC-J4 binds to the epitope which is distinct from the Leu 7 epitope. The electrophoretic patterns of immunoprecipitates with GC-J4, MC-P2 and anti-Leu 7 from detergent lysates of surface labeled human MNC were very similar. The target molecules of anti-Leu 7, MC-P2 and GC-J4 have apparent molecular weights of 205, 170, 150, 135, 110, 85, 65 and 55 kDa, and they are monomeric or noncovalently associated proteins. The patterns of immunoprecipitates with MC-P4 were slightly different from those with anti-Leu 7 and MC-P2. The implications of these results are as follows: 1) The mAb with specificity to the carbohydrate part of MAG reacts with the Leu 7-reactive molecule on human MNC. 2) At least two epitopes detected by anti-MAG mAbs coexist on the surface molecules with various apparent molecular weights. 3) The fine antigenic specificity of MC-P4 is different from those of anti-Leu 7 or MC-P2.

The myelin-associated glycoprotein (MAG) is a quantitatively minor component of the myelin sheath, but it may play a regulatory role in myelinogenesis. We have isolated a cDNA clone for mouse MAG from an expression library using polyclonal antisera against mouse MAG. The clone was found to contain 1350nt cDNA, and the sequenced regions were homologous to the corresponding regions of the previousely reported rat MAG cDNA. Evidence is presented that expression of the two MAG mRNAs is developmentally regulated in mouse brain. In the quaking mouse, the mRNA without a 45nt exon portion was scarcely expressed throughout development. We conclude that the mechanism of splicing-out the 45nt exon portion is lacking in the quaking mouse.