Integrative genomics establishes GNL3 as a pleiotropic hub and causal gene for osteoarthritis

Osteoarthritis (OA) is a leading cause of disability, yet there are no approved disease-modifying therapies (DMOADs) capable of halting its progression. While genome-wide association studies (GWAS) have robustly linked the 3q21 locus to OA, the causal effector gene and its underlying mechanism have remained elusive, hindering translational progress. We implemented a multi-tiered, systematic integrative genomics strategy to proceed from genetic association to causal mechanism. By integrating summary statistics from large-scale OA GWAS with multi-tissue molecular quantitative trait loci (QTL) and single-cell expression QTL (sc-eQTL) data, we employed a combination of Bayesian colocalization and Mendelian randomization (MR) to establish a robust chain of causal evidence. Our analysis identifies GNL3 as the causal effector gene at the 3q21 locus. MR analysis formally demonstrated that genetically predicted higher GNL3 expression is causally protective against both knee and hip OA (e.g., Knee OA: Odds Ratio = 0.85, p = 4.4e-4). Further single-cell causal inference pinpointed this protective effect to specific cellular contexts, most prominently in activated T-cells, neural cells under cellular stress, and developmental progenitors. Moreover, we establish this locus as a pleiotropic hub, showing that its causal variants are shared with systemic OA risk factors, including Body Mass Index (BMI) and vitamin D levels. This study establishes, for the first time, a coherent chain of evidence from a shared genetic signal to the regulation of GNL3 expression in specific cell types and, ultimately, to a causal impact on OA risk. Our findings converge on a novel mechanistic model: GNL3 acts as a master regulator of systemic homeostasis, where its genetically determined expression level modulates immune and neural cell responses to stress, thereby dictating an individual’s susceptibility to OA. This work validates GNL3 as a high-confidence therapeutic target and provides a new framework for developing DMOADs aimed at reinforcing systemic homeostatic pathways.