From metabolomics to therapeutics: identifying causal metabolites and potential drugs for the treatment of osteoarthritis.

Abstract

Osteoarthritis (OA) is a common age-related disease that causes pain and impaired mobility. Various blood metabolites are reportedly associated with bone health; however, their impact on OA remains unclear. Therefore, we conducted a metabolome-wide Mendelian randomization (MR) study to identify causal metabolites and therapeutic targets in OA. Genetic associations of metabolites were derived from the largest genome-wide association study (GWAS) of the blood metabolome, which provided summary-level data on 1091 blood metabolites. Genetic associations with OA were obtained from four large-scale GWAS: McDonald's study (140,025 cases, 344,349 controls), Zengini's study (12,658 cases, 50,898 controls), Dönertaş's study (39,515 cases, 445,083 controls), and Tachmazidou's study (39,427 cases, 378,169 controls). MR and colocalization analyses were performed to validate the causal roles of the candidate metabolites. Further analyses were conducted using expression quantitative trait locus-based MR, single-cell sequencing data, protein-protein interaction networks, and druggability assessments. These analyses aimed to identify the differentially expressed genes and prioritize them as potential therapeutic targets. The genetically predicted levels of 10 metabolites were associated with OA. Elevated levels of five metabolites and reduced levels of another five metabolites were associated with an increased OA risk. Among these, five metabolites were prioritized based on the most compelling evidence. Seven genes were identified as potentially involved and could serve as novel therapeutic targets for OA. Several blood metabolites were associated with OA, providing new insights into the etiology of OA and highlighting promising therapeutic targets.