Multi-Target Gene Therapy for Osteoarthritis: Dual-Axis Modeling and In Silico Validation

Background: Osteoarthritis (OA) remains therapeutically intractable despite advances in regenerative medicine. Single-target interventions—including mesenchymal stromal cells, platelet-rich plasma, and gene therapies—consistently yield transient symptomatic relief without durable structural modification. The recent Phase 2 failure of GLPG1972, a potent ADAMTS-5 inhibitor (ROCCELLA trial, n=932), exemplifies this persistent challenge. Objective: This study proposes a multi-target gene therapy strategy integrating computational structural analysis, network perturbation modeling, and clinical failure analysis to address the dual-axis nature of OA pathogenesis. Methods: Crystal structures of key therapeutic proteins (IL-1Ra, SOX9, IGF-1) were analyzed using PyMOL. Molecular docking was performed using AutoDock Vina to evaluate ADAMTS-5 inhibitor binding. Network perturbation analysis (n=1,000 Monte Carlo iterations) quantified multi-target synergy using an ECM Recovery Score. shRNA sequences were designed using the Reynolds algorithm with BLAST off-target validation. Human–canine sequence homology was assessed across all therapeutic targets. Results: Structural analysis confirms therapeutic transgenes preserve critical functional domains. Network perturbation demonstrates significantly higher ECM Recovery Scores for multi-target intervention (76.2 ± 8.3) versus single-target control (43.6 ± 12.1; p < 0.001). Analysis of GLPG1972 failure reveals an "Exosite Bypass" mechanism supporting complete enzyme elimination via shRNA rather than catalytic inhibition. Validated shRNA sequences for ADAMTS-5 and MMP-13 achieved Reynolds scores of 8–9/9. Human–canine homology analysis (mean 90.5% identity) supports canine models for translational studies. Conclusions: Computational analysis, clinical failure mechanisms, and network modeling converge on the necessity of simultaneous multi-axis targeting. The proposed dual-vector AAV system addressing inflammation (IL-1Ra), anabolism (SOX9, IGF-1), and catabolism (ADAMTS-5/MMP-13 shRNA) provides a framework for experimental validation. This work remains hypothesis-generating and requires experimental confirmation in cell-based and animal models.