Alternative polyadenylation as a novel mechanism for diabetes

PROJECT SUMMARY Obesity and the associated insulin resistance are the established risk factors for type 2 diabetes (T2D). Although more than 90% of T2D patients are overweight or obese, only about 30% of obese people develop T2D. One major determining factor for the development of T2D in obese patients is islet β−cell decomposition or failure, resulting in relative insulin deficiency. It is well documented that islet failure has a strong genetic predisposition. Genome-wide association study (GWAS) provides a powerful tool to associate genetic variants (SNPs) with T2D. However, the vast majority of the diabetes risk SNPs from GWAS are found in the non-coding regions, posing significant challenges to identifying the SNP-associated genes for T2D. 3’untranslated regions (3’UTR) are non- coding sequences containing cis-regulatory elements (CRE), such as binding sites for miRNAs and RNA-binding proteins (RBPs) that regulate mRNA fate and protein expression. Alternative polyadenylation (APA) at the 3’UTR is an RNA-processing mechanism that generates mRNA isoforms with significantly different 3’UTR lengths with distinct CREs. SNPs that regulate APA may lengthen or shorten 3’UTR, thereby altering gene expression and function. We recently developed a novel 3’UTR APA quantitative trait loci (3’aQTLs) analysis tool to colocalize disease-associated SNPs with APA. Using the 3’aQTLs tool, we found that the lengthening of 3’UTR of zinc finger CCCH domain-containing protein 13 (ZC3H13) in the pancreas was highly associated with T2D. ZC3H13 is a key epitranscriptomic factor that forms an N6-methyladenosine (m6A) RNA modification writer complex with methyltransferase-like 3 (METTL3) and methyltransferase-like 14 (METTL14). Both METTL3 and METTL14- mediated m6A have been shown to be essential for β-cell function. Our preliminary data show that longer ZC3H13 3’UTR reduced ZC3H13 protein expression without changes to mRNA levels. Knocking down ZC3H13 suppressed insulin production in the cultured β-cells. Furthermore, ZC3H13 heterozygous knockout mice exhibited impaired glucose tolerance with reduced insulin levels when challenged with a high-fat diet. We, therefore, hypothesize that the reduced ZC3H13 protein expression from 3’aQTLs-associated 3’UTR lengthening contributes to the genetic predisposition of islet failure in T2D. We propose three aims to study what SNPs cause ZC3H13 3’UTR lengthening, why ZC3H13 3’UTR lengthening reduces protein expression, and how the reduced ZC3H13 protein impairs islet function. Aim 1: To identify the causal SNPs impacting APA of ZC3H13 using functionally informed fine-mapping of 3'aQTLs. Aim 2: To determine the mechanisms by which 3’UTR lengthening reduces ZC3H13 protein expression. Aim 3: To investigate the mechanisms by which reduced ZC3H13 expression impairs insulin production and islet function. Our studies will advance the field by uncovering ZC3H13 APA as a novel genetic risk factor and elucidating ZC3H13-mediated epitranscriptomic regulation as a novel mechanism for the impaired β-cell function in T2D.