P3-022: transcriptional regulation at the trem gene cluster in ad brains

Association of missense variants in TREM2 and TREML2 with susceptibility of late-onset Alzheimer's disease (LOAD) has been well replicated in several large case-control series including our own Mayo series (N=10,000). More recently, higher blood TREM2 mRNA and TREM2 protein levels were reported by others to correlate strongly with AD status. In this study, we evaluate the association of brain specific transcripts encoded by the TREM gene cluster on chromosome 6 (TREM1, TREML1, TREM2, TREML2, TREML4) with variants at this locus and their effect on risk of LOAD. Utilizing Illumina DASL array data from our published brain expression genome-wide association study (eGWAS), expression levels in ADs (197 cerebellum, 202 temporal cortex) were compared to levels in non-AD (177 cerebellum, 197 temporal cortex) via multivariable linear regression, adjusting for covariates (age-at-death, sex, APOE - e 4, RIN) and transcript levels of 5 CNS cell-type specific genes. RNAseq data from 96 LOAD brains has already been obtained and will be employed for characterization of the mechanism of transcriptional changes of TREM genes. For validation, TREM transcripts detected via RNAseq will be measured individually using isoform-specific TaqMan® gene expression assays in additional samples. The associations of gene isoforms with cis SNPs will be evaluated using multivariable linear regression analyses controlling for appropriate covariates. TREML1 and TREML4 were significantly higher in the ADs compared to the non-ADs (beta=0.085, p=0.048 and beta=0.048, p=0.048, respectively) in our eGWAS of temporal cortex tissue. Furthermore, in preliminary blood DASL array data from 7 ADs and 26 clinically, non-demented controls from the Mayo Clinic Study of Aging, TREML1 also has higher expression in the ADs compared to controls (beta=0.38, p=0.014). TREM family genes appear to have differential expression in brain and blood. Completion of RNAseq and isoform specific expression analysis is expected to fine tune our understanding of these gene expression changes in this gene family. eQTL study of these transcript levels may uncover novel AD risk variants at the TREM gene cluster that have an effect on transcriptional regulation.