Bidirectional Expression of the SCA8 Expansion Mutation: One Mutation, Two Genes

Spinocerebellar ataxia type 8 (SCA8) is a dominantly inherited, slowly progressive neurodegenerative disorder caused by a CTG.CAG repeat expansion located on chromosome 13q21. The expansion mutation was isolated directly from the DNA of a single patient using RAPID cloning and subsequently shown to co-segregate with disease in additional ataxia families including a seven-generation kindred (the MN-A family). The size-dependent penetrance of the repeat found in the large MN-A kindred makes it appear as though some parts of the family have a dominant disorder while other parts of this same family have recessive or sporadic forms of ataxia. While the linkage and size-dependent penetrance of the SCA8 CTG.CAG expansion in the MN-A family argue that the SCA8 expansion causes ataxia, the reduced penetrance in other SCA8 families and the discovery of expansions in the general population have led to a controversy surrounding whether or not the SCA8 expansion is pathogenic. A recently reported mouse model in which SCA8 BAC-expansion but not BAC-control lines develop a progressive neurological phenotype now demonstrates the pathogenicity of the (CTG.CAG)(n) expansion. These mice show a loss of cerebellar GABAergic inhibition and, similar to human patients, have 1C2-positive intranuclear inclusions in Purkinje cells and other neurons. Additional studies demonstrate that the SCA8 expansion is expressed in both directions (CUG and CAG) and that a novel gene expressed in the CAG direction encodes a pure polyglutamine expansion protein (ataxin 8, ATXN8). Moreover, the expression of non-coding (CUG)(n) expansion transcripts (ataxin 8 opposite strand, ATXN8OS) and the discovery of intranuclear polyglutamine inclusions suggest SCA8 pathogenesis may involve toxic gain-of-function mechanisms at both the protein and RNA levels. Our data, combined with the recently reported antisense transcripts spanning the DM1 repeat expansion in the CAG direction and the growing number of reports of antisense transcripts expressed throughout the mammalian genome, raises the possibility that bidirectional expression across pathogenic microsatellite expansions may occur in other expansion disorders, and that potential pathogenic effects of mutations expressed from both strands should be considered.