New Insights on the Therapeutic Potential of Runt-Related Transcription Factor 2 for Osteoarthritis: Evidence from Mendelian Randomization

Introduction: Research has highlighted the role of runt-related transcription factor 2 (Runx2) in the development of osteoarthritis (OA); however, its causal association remains unclear. This study aimed to explore whether Runx2 expression is causally associated with OA and assess its therapeutic potential for OA. Methods: Genetic proxy instruments for Runx2 expression were obtained from gene expression quantitative trait locus (eQTLs) study of eQTLGen Consortium (n = 31,684). Aggregated genome-wide association study (GWAS) data for OA (including all OA [177,517 cases and 649,173 controls], knee OA (KOA) [62,497 cases and 333,557 controls], and hip OA (HOA) [36,445 cases and 316,943 controls]) were extracted from the Genetics of Osteoarthritis Consortium. We integrated eQTLs data with OA GWAS data to estimate their causal association and to estimate the potential of Runx2 as a drug target in the treatment of OA using summary data-based Mendelian randomization (SMR) analysis. Furthermore, different OA GWAS data (including all OA [77,052 cases and 378,169 controls], KOA [24,955 cases and 378,169 controls], and HOA [15,704 cases and 378,169 controls]) derived from the GWAS Catalog database were used for replication study. Results: SMR analysis showed that high expression levels of Runx2 were associated with an increased risk of all OA [odds ratio (OR) 1.044, 95% confidence interval (CI) 1.023–1.067; P = 5.03 × 10⁻⁵], KOA (OR 1.040, 95% CI 1.006–1.075; P = 0.021), and HOA (OR 1.067, 95% CI 1.022–1.113; P = 0.003). This suggests that Runx2 inhibitors may have promising potential for the treatment of OA. Notably, the causal effects of Runx2 with all OA (OR 1.053, 95% CI 1.027–1.079; P = 3.95 × 10⁻⁵) and KOA (OR 1.043, 95% CI 1.001–1.087; P = 0.045) were repeated in the replication study, but limited evidence supported the association of Runx2 expression levels with HOA (OR 1.045, 95% CI 0.993–1.101; P = 0.094). Conclusions: Our analyses indicate a positive correlation between Runx2 expression and OA risk across all three phenotypes, suggesting the potential of Runx2 inhibitors in the treatment of OA and providing evidence from a genetic perspective.