Abstract 5705: Identifying DNA methylation quantitative trait loci across multi-ethnic populations

Abstract

Abstract Objective: To date, genome-wide association studies (GWAS), which are primarily conducted in Europeans, have identified a large number of validated DNA methylation quantitative trait loci (meQTLs) acting in both cis and trans. However, it is recognized that there are racial differences in genetic architectures such as allele frequencies and linkage equilibrium patterns, which may differentially influence DNA methylation levels. This points to a critical need to conduct a methylome wide association study (MWAS) to identify meQTLs across multiethnic populations including understudied African Americans (AAs), Latinos, Japanese Americans (JAs), and Native Hawaiians (NHs). Methods: We are performing a GWAS to identify meQTLs in blood leukocytes in a multiethnic population using data from the Multiethnic Cohort Study (MEC). In MEC, “genome-wide” germline genetic variants and DNA methylation levels of blood leukocytes have been measured using the Illumina 1M and MethEPIC array, respectively, for 372 AAs, 408 Latinos, 531 JAs, 319 NHs, and 406 European Americans who were current smokers at time of blood draw. Genotyped data has been imputed with 1000 Genomes phase 3 dataset. For DNA methylation data, standardized quality control (QC) and normalization have been performed. We adjusted for the following potential confounders at time of blood draw: age, sex, body mass index, estimated cell type composition variables, genetic principal components, smoking pack-years, and urinary total nicotine equivalents (a biomarker for internal smoking dose). A stringent Bonferroni-corrected threshold (P<5×10-8/(850,355×6) = 9.80×10-15) was used to define statistical significance. An independent dataset of MEC participants including 114 AAs, 142 Latinos, 286 JAs, 108 NHs, and 270 European Americans will be used to replicate the identified meQTLs. Results: Our preliminary analyses for identifying cis-meQTL SNPs (residing within 1 Mb upstream or downstream of a CpG site of interest) found 169,823 CpG sites (20.0%) are associated with at least one nearby genetic variant in AAs. Similarly, 284,371 CpG sites (33.4%) in JAs, 168,580 CpG sites (19.8%) in Latinos, 123,032 CpG sites (14.5%) in NHs, and 234,822 CpG sites (27.6%) in European Americans are significantly associated with at least one nearby genetic variant. In the pan-ethnic analyses including all five ethnic/racial groups, 193,743 CpG sites (23.0%) are associated with at least one nearby genetic variant. Conclusion: Our preliminary findings identified ethnic-specific and pan-ethnic cis-meQTL SNPs across a multiethnic population that includes understudied AAs, Latinos, JAs, NHs, along with European Americans. If replicated in independent datasets, these findings will significantly improve our understanding of genetic regulation of DNA methylation levels in these populations. Citation Format: Lang Wu, Xiequn Xu, Dalia Ghoneim, Kayla Kim, Alexandra Binder, Yesha Patel, Daniel O. Stram, Loïc Le Marchand, Sungshim L. Park. Identifying DNA methylation quantitative trait loci across multi-ethnic populations [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 5705.