Circulating proteins and bone mineral density: A Proteome-Wide Mendelian Randomization Study

Introduction: Current osteoporosis medications often prove ineffective for various reasons. Alongside optimizing available agents, new genetic targets should be proposed for drug development. Mendelian randomization (MR) may resolve throughput and confounding issues in traditional observational studies for druggable targets. Methods: We employed two-sample MR with protein quantitative trait loci (pQTLs) and expression quantitative trait loci (eQTLs) data as exposures and six bone mineral density (BMD) sites as outcomes. By meta-analyzing pQTL evidence, validating eQTL evidence, conducting MR sensitivity tests, and assessing druggability, key druggable targets for BMD were identified. Additionally, we performed functional analysis, drug repurposing annotation, transcriptome analysis, in-house PCR, ELISA, and micro-CT validation to further investigate the functionality and expression levels of these targets across different tissues and conditions. Results: Out of 5,928 pQTLs from deCODE and UKB-PPP datasets, 16 were identified as prioritized targets with significant meta pQTL evidence. Tyrosine-protein kinase Lyn (LYN, meta beta -0.09, 95% CI -0.13 to -0.05), Chondroadherin (CHAD, meta beta -0.39, 95% CI -0.18 to -0.20), Tumor necrosis factor receptor superfamily member 19 (TNFRSF19, meta beta -0.03, 95% CI -0.05 to -0.02), and Transforming growth factor beta induced (TGFBI, meta beta -0.04, 95% CI -0.06 to -0.03) were identified as key druggable targets for BMD. R-spondin-3 (RSPO3) and SPARC-related modular calcium- binding protein 2 (SMOC2) were also suggested with consistent MR associations with previous studies. Discussion: We identified four novel BMD-related targets (CHAD, LYN, TGFBI, TNFRSF19) through pQTL meta-analysis, and validated RSPO3/SMOC2's positive effects. By integrating multi-tissue transcriptomics and OVX experiments, we further revealed elevated expression of TNFRSF19/TGFBI negatively correlated with BMD, providing new therapeutic insights. Conclusion: This large-scale Proteome-Wide MR study introduced novel targets for BMD and osteoporosis at transcriptional and translational levels, presenting new prospects for drug repurposing and development.