Integrative multi-omics for discovery of molecular pathways associated with diabetic retinopathy

PROJECT SUMMARY One of the areas of emphasis on the NEI strategic plan is to bridge the gap between genomics and mechanisms of disease. Diabetic retinopathy (DR) is a leading cause of blindness and disproportionately affects Hispanics/ Latinos (HL). DR-specific blood biomarkers and systemic treatments are not used clinically. Although research

into biomarkers and genetic variants has informed our understanding of DR, whether these signatures cause or are caused by DR, and how genetic variants contribute to DR etiology is still not clear. Further, few of these studies have meaningful representations of HL despite their high DR burden. Because transcript and metabolite

levels provide a window into gene function, they can help us interpret genome-wide association study (GWAS) results to better understand complex diseases like DR. Given the growing HL population in the US and their disproportionate cardiometabolic disease burden, HL representation in functional genomic studies of diabetic

sequelae is imperative. The Cameron County Hispanic Cohort (CCHC) is a cohort of HL with poor cardiometabolic health and high risk for DR. Participants range between metabolically healthy and metabolically disordered individuals without or with diabetes, some with various levels of DR. We will quantify differential

transcript (Aim 1) and metabolite (Aim 2) abundance between participants with Type 2 diabetes (T2D) with or without clinical diagnosis of DR in CCHC and replicate our findings in the Hispanic Community Heath Study/Study of Latinos (HCHS/ SOL), an independent, extant, diverse, larger HL dataset with T2D and self-

reported DR status. Using gene regulatory and metabolite network and pathway analyses we will clarify DR pathophysiology, inform novel blood biomarkers and drug targets for DR, and validate their temporal precedence with DR using both cohorts. In Aim 3, we will apply two-sample Mendelian randomization and colocalization

methods to interpret known and novel transcriptomic and metabolomic biomarkers (eQTLs and mQTLs) using results from a novel HL-specific GWAS meta-analysis and the literature. Finally, using diverse electronic health record (EHR)-linked databases, we will perform eQTL-, mQTL-, variant-, and polygenic risk score-based

phenome wide association and enrichment studies to expand our understanding of the broader clinical comorbid profile of, and pathogenic mechanisms shared with, DR. Our study will clarify mechanistic causes/consequences of DR-associated expression patterns in the blood. The aims will independently inform novel diagnostic and

therapeutic systemic DR-targets and our understanding of the causality and broader clinical implications of blood signatures dysregulated in DR. This project meets several NIH/NEI strategic goals including: 1) enhancing our understanding of ocular epidemiology in groups under-represented in vision science; 2) enriching data capturing

molecular signatures of disease; 3) using data science to integrate cross-modality data to understand disease; 4) understanding how genomic variants affect disease; 5) identifying risk factors for ocular disease in underserved populations to decrease visual impairment; and 6) promoting a diverse vision research workforce.