Genetically predicted ZDHHC8 expression and osteoporosis risk: A two-sample Mendelian randomization and metabolite...

Protein S-palmitoylation regulates protein localization, stability, and signaling and has emerging roles in bone-cell biology. We investigated whether genetically predicted expression of palmitoylation-related genes is associated with osteoporosis risk and explored circulating metabolites as candidate mediators. We performed a 2-sample Mendelian randomization (MR) study using cis-expression quantitative trait locus instruments for 22 palmitoylation-related genes, FinnGen R10 osteoporosis data (8017 cases and 391,037 controls), and 1400 circulating metabolite traits (1091 metabolites and 309 metabolite ratios). Inverse-variance weighted was the primary estimator. Weighted-median and MR-Egger analyses, Cochran Q testing, and instrument-strength assessment were used as sensitivity analyses. Summary-data-based MR with the heterogeneity in dependent instruments test was used as an additional gene expression validation. Gene Expression Omnibus bone marrow mesenchymal stem cell datasets were used for expression comparison; GSE35959 was treated as the SuperSeries containing GSE35956 and GSE35958 rather than as an independent replication cohort. Two-step MR and the product-of-coefficients approach were used to estimate candidate mediation effects. Genetically predicted zinc finger DHHC-type palmitoyltransferase 8 (ZDHHC8) expression was associated with higher osteoporosis risk (odds ratio = 1.23, 95% confidence interval: 1.09-1.39, P = .0007), and the association met the Bonferroni threshold for the 22-gene screen. Summary-data-based MR provided additional, nominal evidence for ZDHHC8 (P = .0415). ZDHHC8 expression was higher in osteoporosis-derived bone marrow mesenchymal stem cells in the Gene Expression Omnibus comparison. Three metabolite traits: 3-(3-amino-3-carboxypropyl)uridine (acp3U), 2-hydroxyhippurate, and X-25343, showed directionally coherent 2-step MR estimates, with estimated mediated proportions of 11%, 9%, and 10%, respectively. Because these metabolite findings did not meet a Bonferroni threshold for 1400 tests, they are interpreted as exploratory candidate mediators. Genetically predicted ZDHHC8 expression is associated with osteoporosis risk across complementary genetic and expression analyses. The metabolite results suggest possible intermediate pathways but remain hypothesis-generating. These findings support further functional study of ZDHHC8 in bone cells rather than immediate clinical targeting.