Assessment of TREM2 R47H in a population-based sample: The Cache County Study

Recent studies by Guerreiro et al 2012, and Jonsson et al 2012, identified the R47H variant in TREM2 as a risk factor for developing Alzheimer's disease suggesting that other rare variants in this gene may also affect risk. Here we present an association analysis of the R47H variant in a large population based sample, the population attributable fraction for the R47H variant, and the results of sequencing exon 2 of TREM2 in several hundred AD cases. We genotyped rs75932628 (R47H) in 3832 samples from the population based Cache County Memory Study using a custom Taqman Assay. This includes 387 clinically ascertained Alzheimer's cases and 635 screened or ascertained controls. We also performed Sanger Sequencing of exon 2 of TREM2 in our AD cases to search for unknown significant genetic variants. The R47H variant had a minor allele frequency of 0.0029, with population attributable fraction for R47H in this large, population-based sample of 0.004. We identified 3 R47H heterozygotes among 635 clinically ascertained controls and 3 R47H heterozyotes among 384 clinically ascertained Alzheimer's disease cases. The chi-squared test of the contingency table yielded a one-tailed p-value of 0.27 and an Odds Ratio of 1.65. We found one novel amino acid substitution in our Sanger sequencing, A49S. A49S is not present in the 1000 genomes or exome variant databases. This variant is predicted to be “possibly damaging” in POLYPHEN with a score of 0.90. As expected, the low frequency of the R47H variant gives it a minimal population attributable fraction. However, the results of our association analysis support those found by Guerreiro et al, and Johnsson et al, for the R47H variant, which correlates to a higher risk of developing Alzheimer's. The novel variant A49S is near R47H and is predicted to have a functional effect by POLYPHEN, providing further evidence for a role of the TREM2 gene in Alzheimer's pathology.