Integrating urinary and plasma omics to identify markers and therapeutic targets for cardiac disease

Abstract

AbstractIntroductionUrinary metabolites, representing kidney regulated filtration of metabolism end products, contain cardiac disease biomarkers such as NT-proBNP. We set out to integrate plasma proteins with urinary metabolites to identify potentially druggable metabolic pathways for cardiac disease.MethodsData was leveraged from a genome-wide association study (GWAS) on 954 urinary metabolites. Mendelian randomisation was used to identify urinary metabolites associating with atrial fibrillation (AF), heart failure (HF), dilated cardiomyopathy (DCM), or hypertrophic cardiomyopathy (HCM). By interrogating eight independent plasma protein GWAS, jointly including 92,277 participants and 1,562 unique proteins, we identified druggable plasma proteins with a directionally concordant effect on urinary metabolites and cardiac outcomes.ResultsIn total, 38 unique urinary metabolites associated with cardiac disease, predominantly covering breakdown products from amino acid metabolism (n=12), xenobiotic metabolism (n=5), and unclassified metabolism origins (n=16). Subsequently, we identified 32 druggable proteins expressed in cardiac tissue, which had a directionally concordant association with the identified urinary metabolites and cardiac outcomes. This included positive control findings, for example higher values of AT1B2 (targeted by digoxin) decreased the risk of HCM, which we were able to link to a novel unclassified urinary metabolite (X-15497). Additionally, we showed that increased plasma RET values, a mediator of GDF-15 signalling, reduced the risk of HF, and linked this to the novel unclassified urinary breakdown product X-23776.ConclusionWe were able to identify 32 druggable proteins affecting cardiac disease, and link these to urinary measurements of metabolite breakdown processes identifying potentially novel disease pathways.