A transcriptome-wide association study provides new insights into the etiology of osteoarthritis

BACKGROUND: Osteoarthritis (OA) is a common clinical disease caused by a variety of factors, including genetic variants. Although genome-wide association studies (GWAS) have been performed to elucidate the genetic basis of OA, some loci of risk located in noncoding regions of the genome have been neglected. Therefore, we integrated multiple data types to detect the genetic component of gene expression in OA patients through transcriptome-wide association studies (TWAS) and summary-data-based Mendelian randomization (SMR) analysis. METHODS: TWAS was performed by integrating the larger GWAS summary-data for OA (n=30,727 cases, n=297,191 controls) and 2 expression weight sets (muscle-skeletal tissue and whole blood). Colocalization analysis, conditional analysis, and fine-mapping analysis were also conducted. A broad description of the identified associations was obtained. In addition, a causal relationship between certain risk genes and OA was identified with SMR. RESULTS: New significant genome-wide associations were found, including on chromosome 1q36.12 (rs1555024, P=4.24E−07) near the ASAP3 and TCEA3 genes, on chromosome 17q24.2 (rs2521348, P=1.01E−06) near the ABCA9 gene, on chromosome 20q11.22 (rs224331, P=8.17E−09) near the UQCC1 and MYH7B genes, and on chromosome 21q21.3 (rs2832155, P=5.39E−08) near the RWDD2B gene. In addition, SMR results exhibited that upregulated UQCC1 and downregulated ASAP3 were associated with OA development and both had a significant causal relationship with OA. CONCLUSIONS: We revealed some novel OA-associated genes and risk loci by integrating multiple data types and analysis methods, thus providing new clues for the study of genetic mechanisms of OA.