This area of investigation began with the demonstration of monoclonal anti~MAG antibodies in patients with mixed sensory~motor polyneuropathies occurring in associated with IgM gammopathy (parapro~teinemia). It was subsequently demonstrated that these anti~MAG antibodies were all directed toward carbohydrate epitopes in MAG and cross~reacted with 19 to 28 kD glycoproteins of PNS myelin and a sphingoglycolipid, sulfate~3~glucuronyl paragloboside (SGPG). This year we have established that one of the principal 19 to 28 kD glycoprotein antigens for these human anti~MAG~SGPG antibodies is PMP~22, a glycoprotein of peripheral nerve myelin that has only recently been cloned and characterized. This suggests that PMP~22, which has been implicated in the genetic abnormalities causing Charcot~Marie Tooth disease, may also be a target of the anti~MAG~SGPG antibodies causing autoimmune neuropathy associated with gammopathy. Monoclonal antibodies that are MAG~SGPG~negative in patients with gammopathy and neuropathy frequently react with ganglioside antigens in nerve. In the current year, we have completed the characterization of a monoclonal IgA reacting with the major LM1 ganglioside of peripheral nerve myelin. Although monoclonal IgM antibodies reacting with gangliosides in patients with neuropathy are common, this is the first example of a patient with IgA gammopathy in which the antigen has been identified as a ganglioside. Little is known about the molecular mechanisms by which antibodies to acidic glycolipids (SGPG or gangliosides) in patients with demyelinating neuropathy exert their pathogenic effects. In order to probe the function of acidic glycolipids in myelination and to understand mechanisms by which the human anti~glycolipid antibodies may perturb function, we are investigating gangliosides in differentiating Schwann cells. In comparison to Schwann cells cultured in the absence of neurons in serum~containing medium on polylysine, culturing them on a basement membrane substratum (Matrigel) resulted in increased synthesis of gangliosides (predominantly GM3), but had no effect on synthesis of the galactosphingoglycolipids characteristic of myelin. This effect of basement membrane required the presence of a serum factor and was specific for ganglioside synthesis, since the synthesis of proteins, glycoproteins and phospholipids was unaffected. Although basement membrane is well known to be required for myelination, its presence in the absence of neurons signals Schwann cells to synthesize gangliosides rather than the galactosphinglipids that are enriched in myelin.
