Abstract: BACKGROUND: With the acceleration of China's aging population, the number of osteoporosis patients has been increasing significantly. Recent advancements in genome-wide association studies and single-cell transcriptomic sequencing have empowered researchers to identify novel osteoporosis-associated genes through integrative multi-omics analyses.
OBJECTIVE: To identify potential therapeutic targets for osteoporosis using summary data-based Mendelian randomization approaches that integrate genome-wide association studies and transcriptomic data from Asian and European populations.
METHODS: By integrating cis-expression quantitative trait loci (cis-eQTL) and protein quantitative trait loci (pQTL) datasets from multiple tissues (blood and muscle-bone)(cis-expression quantitative trait loci datasets for human blood and bone-muscle tissues were obtained from the Genotype-Tissue Expression (GTEx) Project v.8, which is a large-scale international collaborative project focusing on gene expression variation across different tissues/organs and its association with genetic regulation) with osteoporosis genome-wide association study data (the 2021 European population osteoporosis genome-wide association studies data were obtained from the FinnGen database that is a large-scale genomic research project in Finland. The 2020 large-scale East Asian population genome-wide association study was obtained from Biobank Japan that is a large-scale population-based cohort study led by Japan; we employed a summary data-based Mendelian randomization approach to identify osteoporosis-associated genes. These identified genes were further analyzed using colocalization analysis, single-cell sequencing, and enrichment analysis. All data were derived from published studies or publicly accessible sources, with ethical approval and informed consent obtained for the original studies.
RESULTS AND CONCLUSION: (1) By leveraging summary data-based Mendelian randomization analysis, a total of 64 genes (following the exclusion of duplicates) significantly associated with osteoporosis were identified. Among these, the human leukocyte antigen (HLA) alleles HLA-DQA1, HLA-DQA2, HLA-DQB1, HLA-DQB2, and HLA-DRB5 were robustly validated across two independent outcome datasets. (2) Colocalization analysis further prioritized HLA-DQA2 and HLA-DQB1 as causal candidates with strong evidence of colocalization (posterior probability PPH4 > 0.8). (3) Protein quantitative trait locus (pQTL) analysis revealed that elevated plasma levels of HLA-DQA2 were associated with a reduced risk of osteoporosis. (4) Single-cell sequencing analysis indicated that, within the immune microenvironment of osteoporosis, the abundance of dendritic cells, B cells, macrophages, and neutrophils was significantly higher than that of other cell populations. (5) Enrichment analysis results showed that the identified genes were enriched in the antigen presentation pathway of major histocompatibility complex class II molecules. (6) This study preliminarily identified several previously unreported osteoporosis-associated genes through bioinformatics analyses integrating Asian and European population genome-wide association data. These genes hold potential for further exploration as novel therapeutic targets for osteoporosis in clinical trials.

Key words: genome wide association studies, single cell sequencing, osteoporosis, summary data-based Mendelian randomization