Program Nr: 18 for the 2006 ASHG Annual Meeting

A set of putative functional IDE and PLAU variants shows significant, replicable association with Alzheimers Disease. M. Carrasquillo1, S. Younkin1, M. Kashino1, S. Wilcox1, T. Dincman1, L. Younkin1, L. Ma1, F. Zou1, N. Taner1,2, M. Allen1, R. Petersen2, N. Graff-Radford1, S. Younkin1. 1) Mayo Clinic, Jacksonville, FL; 2) Mayo Clinic, Rochester, MN.
   In IDE and PLAU, which are excellent positional and functional candidate genes for late onset Alzheimers Disease (LOAD), conserved regions (>70% human vs. mouse) were screened for variants likely to be functional. Using logistic regression, conserved IDE variants were tested for association with AD in three large case/control series. Six conserved variants in non-coding regions formed 4 haplotypes with frequencies over 1%: three were protective, one was risky. In the young age group where association was strongest, these 4 haplotypes showed the same significant association with AD in the exploratory (nominal global p = 0.010) and two follow-up series (global p = 0.013 and 0.007). The three protective haplotypes (H2, H3, H4) pair to form 6 genotypes likely to be protective. These were used as a referent group and compared to the four remaining common genotypes (H1/H1, H1/H2, H1/H3, H1/H4). Three genotypes were equivalently risky, one (H1/H4) was indistinguishable from the referent group, and they showed the same significant association with AD in the exploratory (nominal global p = 0.005) and two follow up series (global p = 0.002 and 0.004). Compared to the set of 7 protective genotypes (16%), the combined set of 3 risky IDE genotypes (84%) had ORs of 4.1, 4.3, and 2.6 in the three series (p = 0.0004, 0.0005, and 0.0015). In the combined series, the OR was 3.2 (2.1-5.0) for the set of risky genotypes compared to the set of 7 protective genotypes with a p-value of 6.5x10-8 and a PAR of 65% (48% - 77%). Importantly, we observed a significant 1.5-fold elevation of IDE mRNA in the cerebella of 140 subjects with protective genotypes as compared to 22 subjects with risky genotypes (p=0.005). These results indicate that IDE is likely to be an important LOAD gene with variants that influence risk of AD by altering IDE expression. Similar analysis of variants in conserved regions of PLAU showed that multivariate PLAU haplotypes and genotypes also show significant association with LOAD.