Cis-acting modifiers of trinucleotide repeat instability

The dynamic expansion of CAG.CTG repeats in otherwise unrelated genes is responsible for a growing number of late-onset progressive disorders, including Huntington disease, myotonic dystrophy type 1 (DM1) and the spinocerebellar ataxias. As toxicity increases with repeat length, the intergenerational expansion of unstable CAG.CTG repeats leads to anticipation, an earlier age-at-onset in successive generations in these disorders. Crucially, disease associated alleles are also somatically unstable and continue to expand throughout the lifetime of the individual. In addition, evidence suggests that c/s-acting elements may be major modifiers of instability. Here it was found that the toxicity of expanded polyQ-encoding CAG.CTG tracts correlates with both the expandability of the underlying CAG.CTG repeat and the GC content of the genomic DNA flanking sequences. PolyQ toxicity does not correlate with properties of mRNA or protein sequences, or with polyQ location within the gene or protein. These data thus strongly suggest that the observed inter-locus differences in polyQ. toxicity are not mediated by protein context effects, but that the rate at which somatic expansion of the DNA delivers proteins to their cytotoxic state is a critical factor in expanded polyQ-disease age-at-onset. Using human and mouse cell lines transgenic for an expanded human DM1 locus, it was found that an expanded CTGhz repeat alone is not sufficient for instability. Moreover, by generating mouse cell lines stably transfected with both a stable and unstable expanded CTG[142] repeat, it was possible to assay the effect of cis-elements on these two loci in the same cell line over time. The sequences flanking the unstable repeat were hypermethylated, whereas the sequences flanking the stable transgenic repeat were unmethylated, suggesting an association between CpG methylation and repeat instability. However, methylation of the stable transgenic repeat failed to induce instability. In addition, it was revealed that transcription of an expanded repeat was not sufficient to induce instability. Analysis of genome-wide CAG.CTG microsatellite instability revealed a significant correlation between flanking sequence GC content and microsatellite mutability. This association was most significant for short (< 7 repeats) microsatellites and for those microsatellites located within exons. However, comparison of microsatellite lengths in the human and chimpanzee genomes revealed a complex association between flanking GC content and misalignment mutations at microsatellite loci, suggesting that the modifying effect of flanking GC content on expanded repeat instability may be specific to the expanded repeat disease loci In conclusion, this work suggests that the rate of somatic repeat expansion is a major modifier of disease progression, and that cis-acting elements in turn, modify repeat instability