ABSTRACT: Type 2 diabetes (T2D) is among the leading causes of death in the U.S. Ethnic differences in T2D prevalence are evident, including especially for Native Hawaiians (NHs) who remain disproportionately affected by it. This difference in T2D susceptibility involves an interplay between genetic and environmental factors, of which epigenetic mechanisms, including DNA methylation (DNAm) provide a novel approach to investigating gene-environment interactions of health and disease. Monocytes, an innate immune cell intrinsic to the inflammatory response, are a fundamental immune cell component that likely underlies T2D pathogenesis, given their involvement in inflammation and inflammation-associated insulin resistance and metabolic dysfunction. From participants enrolled into the Multiethnic Cohort Study (MEC), who self-identified as NH (n=152), Japanese American (JA; n=119), or White (n=121), we investigated monocyte-specific DNAm patterns in participants at a baseline visit, when free of T2D, using the HumanMethylation850K (850K) to determine whether monocytes harbor an ethnic-specific epigenetic signature of T2D risk that precedes T2D diagnosis. Using an epigenome-wide association study (EWAS), we found 904 significantly (q < 0.01) differentially methylated loci (DML) at a 5% difference in DNAm between participants who remained T2D free at a 15-year follow-up (i.e., controls) and those that would be diagnosed with T2D by follow-up (i.e., incident T2D) after adjusted for age, sex, and education level. These methylation differences were enriched at regulatory regions of the genome, including intergenic and intragenic regions. Notably, these DML were able to distinctly stratify NHs by T2D risk groups, however, this signature was inapparent in Whites and JAs. Likewise, NHs in the incident T2D group displayed a higher degree of DNAm variability. Sensitivity analysis with traditional risk factors (fasting glucose and body mass index [BMI]) and neighborhood socioeconomic status (nSES) found these risk factors had minimal effect on T2D risk-associated DML. Next, using a similar approach we found ethnic-specific DML in monocytes that was able to uniquely stratify NHs, JAs, and Whites, however, in NHs this epigenetic landscape displayed a higher degree of DNAm variability. Similarly, these DML were enriched at regulatory regions of the genome. In both cases, we found genes associated with biological functions and pathways relevant to monocyte functionality, including immune activation and cellular metabolism. Due to the higher degree of epigenetic variability in NHs, including especially NHs with T2D risk, we investigated differentially variable CpGs associated with T2D risk and ethnic-differences. We found differentially variable CpGs between T2D risk groups were able to stratify NHs with and without T2D risk, whereas it was unresolvable in JAs and Whites. Ethnic-specific DNAm variability clustered each ethnic population distinctly, and NH controls were further partitioned from NHs with T2D risk. Our findings suggest monocytes harbor a unique DNAm signature associated with T2D risk, which may be related to ethnic differences that underlie DNAm variability in monocytes. The increased epigenetic variability in monocytes from NHs may underlie epigenetic plasticity that could allow these cells to readily respond to adverse environmental conditions throughout the life course that may underlie long-term disease risk, whereas in other less susceptible populations with similar environmental exposure, this epigenetic plasticity may not be apparent in monocytes.
