Identifying causal brain structures, genes and proteins for osteoarthritis: A large-scale genetic correlation study based on brain imaging-derived phenotypes, transcriptomes, and proteomes

Abstract

Background Recent epidemiological studies have linked the central nervous system (CNS) to osteoarthritis (OA), suggesting that targeting the CNS could offer new therapeutic strategies. This study aimed to validate the correlation between brain structures and OA risk, and to identify key causal genes by integrating brain transcriptomic and proteomic data with OA genome-wide association studies (GWAS). Methods We analyzed OA summary statistics from 826,690 participants. Using linkage disequilibrium score regression (LDSC) and Mendelian randomization (MR), we explored the genetic relationships between brain structures and OA. A transcriptome-wide association study (TWAS) was conducted with 5,138 brain transcriptomes. Additionally, a proteome-wide association study (PWAS) combined GWAS data with 152 human brain proteomes. Single-cell RNA-Seq eQTL data helped identify causal genes in brain cells linked to OA. Results LDSC and MR indicated brain structures, such as the putamen and amygdala, are strongly associated with OA. Seven genes were linked to knee OA and four to hip OA (FDR < 0.05). Proteins associated with knee, hip, and thumb OA were also identified. scRNA analysis revealed CPNE1 in excitatory neurons and EMILIN2 in oligodendrocyte progenitor cells as causally linked to knee OA. Conclusions This study enhances our understanding of the brain-joint axis in OA genetics, potentially informing new treatment strategies and therapeutic targets.