LAUSR

Expanded alleles accounted for 74.3% of the observed variability in the age at onset, whereas normal alleles explained a discrete additional 0.5% of the total age at onset variance, and 1.95% of the variance of the age at onset was unexplained by the CAG repeats at the expanded alleles. A comparison of age at onset residues after adjusting for the CAG repeat number at expanded alleles across classes for normal alleles showed highly significant differences (F 5 6.39; P < .001; Supporting Information Figure S5A). Moreover, geometric means for the age at onset showed a trend to increase with the elongation of the CAG repeat tract across classes of expanded and normal alleles (Supporting Information Figure S5B). There is evidence of wildtype ataxin-2 being recruited in protein aggregates generated by the expanded ataxin-2. It is expected that wildtype ataxin-2 with longer polyQ stretch to interact more efficiently with mutant ataxin-2, increasing the rate of aggregate formation so the propensity of mutant ataxin-2 to aberrantly interact with other proteins will be decreased, thus slowing down the neurodegenerative processes. On the contrary, the recruitment of wildtype ataxin-2 in protein aggregates could lead to a partial loss of its proposed neuroprotective roles linked to the modulation of apoptosis or to metabolic disturbances. Even when the interaction of wildtype and mutant ataxin2 could be having a double-edge effect, the findings from this study are consistent with the hypothesis that an interaction between wildtype and mutant ataxin-2 proteins may lead to a mitigation of mutant protein toxicity. Alternatively, the results are consistent with the possible existence of a linked modifier on the normal chromosome.