C01 Glutamine codon usage and somatic mosaicism of the HTT cag repeat are modifiers of huntington disease severity

Background Huntington disease (HD) is caused by the expansion of a polyglutamine encoding CAG repeat in exon 1 of the HTT gene. Affected individuals inherit ≥40 repeats and longer alleles are associated with earlier onset and higher HD severity. The HTT CAG repeat is genetically unstable in both the germline and soma. Somatic mosaicism is dependent on the number of CAG repeats and age, and is expansion biased and cell-type specific. Recent genome-wide association studies (GWAS) have identified components of the DNA repair system as trans¬-acting modifiers of HD severity, some of which are known to modify somatic instability of the HTT CAG repeat in HD mouse models. We thus hypothesise that the trans¬-acting modifiers identified by GWAS affect HD severity by their direct effect on HTT CAG somatic instability. Aims Determine the exact trinucleotide structure of the HTT exon 1 repeat and quantify its somatic mosaicism to investigate their association with HD severity. Methods/techniques Using genotyping-by-sequencing, we determined the exact genotype of the polyglutamine and polyproline encoding repeats in HTT exon 1 and quantified the somatic mosaicism associated with the CAG repeat in blood DNA from 807 HD expansion carriers. Results/outcome The sequence encoding the HTT polyglutamine and polyproline tract has an atypical structure in ˜8% of the non-HD-causing alleles and ˜3% of the HD-causing alleles, differing from the typical structure by the number of glutamine encoding CAA codons and/or the number of proline encoding CCA and CCT codons. Multiple linear regression analysis revealed that the number of CAA codons is negatively correlated with HD severity and that the number of CAG repeats is a better predictor of HD severity (r2=0.559) than the number of glutamines (r2=0.537). Moreover, somatic mosaicism in blood correlates with HD severity (r2 ≥0.014, p≤2.5 × 10–3) and some of the polymorphisms associated with HD severity (p=1.5 × 10–5 for FAN1 rs3512, p=1.8 × 10–4 for MLH3 rs175080, p=3.6 × 10–3 for MLH1 rs1799977 and p=0.016 for MSH3 rs1382539). Conclusion Our data show that atypical HD-causing alleles have major implications for genetic diagnosis and counselling and confirm the correlation of somatic expansion with HD severity. The latter further supports the therapeutic potential of targeting expansion causing components of the DNA repair system.