Regulatory variation that affects splicing of the Alzheimer's disease risk gene TREM2

University Of Washington, Seattle WA

Investigators

Abstract

Alzheimer's disease (AD), the most frequent cause of age-related dementia, constitutes a serious burden for the US health system and economy. In the absence substantial preventative or disease-modifying treatments, AD frequency is increasing as the lifespan increases. Age and genetic predisposition are the largest contributors to AD risk. A substantial fraction of genetic variation in AD are noncoding variants that regulate gene expression, but unlike changes in protein code, impact of regulatory variants is not easily predictable. Currently known genes and risk alleles that change protein coding sequence explain at most 30% of AD susceptibility, emphasizing the need to investigate variants that regulate gene expression and splicing. According to recent estimates, up to 50% of all variants associated with human diseases are variants that have some effect on gene splicing. Triggering receptor expressed on myeloid cells 2 (TREM2) is a microglia-specific immune molecule whose dysfunction causes a continuum of neurodegenerative conditions. Some missense TREM2 variants, exemplified by R47H, are transmitted as autosomal dominant traits, conferring a significant late onset AD risk. The R47H allele is one of the strongest known contributors to the risk, with effect size similar to the APOE e4 allele. On the other hand, biallelic loss-of-function variants that inactivate the receptor cause recessive early onset dementia, such as Nasu-Hakola disorder (NHD). TREM2 is subject to alternative splicing, which is species-specific, but the extent and effect of this process has yet to be comprehensively characterized. We recently identified a novel splice isoform of TREM2 (delta e2) with altered activities due to lack of an important protein domain. In preliminary study, we found that multiple variants that cause NHD or increase risk for AD affect TREM2 splicing and reduce the dosage of functional transcript via competition with the abnormally spliced delta e2 isoform. We hypothesize that a precise balance of isoforms is important for TREM2 function and that cis-regulatory variants, in combination with protein factors variably expressed in AD microglia, reduce gene dosage and increase AD risk. The goal of this proposal is to comprehensively characterize the TREM2 transcript repertoire in human brain in health and disease (Aim 1), assess TREM2 variants that are present in patients with dementia for their effect on RNA splicing and post-splicing processing (Aim 2) and analyze the interplay between AD-associated cis- regulatory TREM2 variants and trans-acting splicing factors that are perturbed in AD brain (Aim 3). To achieve these goals, we will use state-of-the-art long-read RNA sequencing, computational variant effect predictions, high-throughput combinatorial library screening, and unbiased quantitative proteomics approaches. This will generate a cis-regulatory TREM2 map at the nucleotide resolution level that will be used to annotate the regulatory potential of hundreds of variants of unknown significance found in AD patients. Successful completion of the proposal will spearhead application of similar strategies to characterize variants that affect post- transcriptional regulation of other disease-related genes.