Prioritising Candidate Therapeutic Targets in Degenerative Spinal Diseases Through Immune-Cell-Specific Single-Cell eQTL-Guided Cis-Mendelian Randomization and Colocalisation

Intervertebral disc disorders (IVDD), spinal stenosis and spondylosis are the commonest degenerative spinal phenotypes and a leading source of pain and disability worldwide, yet no disease-modifying drugs exist; immune regulation is implicated, but causal evidence at immune-cell resolution is scarce. We performed cis-Mendelian randomization (MR) using single-cell cis-eQTL instruments from OneK1K (982 donors; 14 immune cell types; 8,733 independent eGenes) against genome-wide association data for the three phenotypes from FinnGen R13 and the Million Veteran Program (MVP), analysed independently, with Bayesian colocalisation and cross-cohort evidence tiering, followed by network, druggability, novelty and phenome-wide association (PheWAS) annotation. Significant MR exposures numbered 395, 291 and 134 (FinnGen) and 96, 99 and 51 (MVP) for stenosis, IVDD and spondylosis, with 96.8 - 100% cross-cohort direction concordance. Tiering nominated 145 gene - cell-type pairs covering 72 genes. GFPT1 was the only gene with top-tier evidence in all three phenotypes (consistent risk-increasing direction; its lead variant's 13 PheWAS associations were all musculoskeletal); METTL21B showed cell-type-dependent bidirectional effects; 51 genes (71%) had no prior spinal-disease association. Immune-cell-resolved, cross-cohort-tiered evidence nominates GFPT1 and METTL21B as candidate therapeutic targets for degenerative spinal diseases, delineating candidates for functional follow-up rather than constituting validation.