Deciphering the shared genetic structure between hip osteoarthritis and femoral neck bone mineral density

Zongkun Jiang¹Xiaoyan Dou²Shixuan Wang¹Hongtao Li¹Jian Kang^3*

• ¹The Second Affiliated Hospital of Liaoning University of Traditional Chinese Medicine, Shenyang, Liaoning Province, China
• ²Affiliated Hospital of Liaoning University of Traditional Chinese Medicine, Shenyang, Liaoning Province, China
• ³Liaoning University of Traditional Chinese Medicine, Shenyang, China

While the association between hip osteoarthritis (HOA) and femoral neck bone mineral density (FN-BMD) is established, their shared genetic architecture remains elusive. The present study applied bidirectional Mendelian randomisation (MR) to investigate causal relationships between HOA and FN-BMD. The quantification of genetic correlations was achieved by employing linkage disequilibrium score regression (LDSC) and high-definition likelihood (HDL) inference. Multi-trait genome-wide association studies (MTAG) have been shown to enhance statistical resolution, thereby unveiling hitherto unreported genetic associations. Independent MTAG-HOA risk loci were identified through conditional joint analysis (GCTA-COJO), complemented by FUMA functional annotation. The application of both MAGMA and GCTA-fastBAT has revealed pleiotropic genes linked to MTAG-HOA susceptibility. Integration of fine-mapped genes from these loci with riskassociated candidates has enabled the identification of 13 key HOA-related genes. Functional annotation of these 13 key genes was performed using Gene Ontology (GO) and KEGG (Kyoto Encyclopedia of Genes and Genomes) pathway enrichment analyses. Multi-tissue transcriptome-wide association studies (TWAS) explored the expression of key genes across different tissues and their association with HOA. SMR analysis evaluated the causal relationship between key gene expressions in various tissues and HOA. Proteomic profiling is conducted via proteome-wide association studies (PWAS) and biomarker expression level imputation using summary-level statistics (BLISS). The application of stratified LDSC-SEG has revealed a genetic enrichment profile in cell types.Results: Bidirectional MR analysis revealed a significant negative causal effect of FN-BMD on HOA, whereas the reverse MR analysis did not identify a causal effect. LDSC and HDL analyses revealed genetic correlations between HOA and FN-BMD of rg = 0.132 and rg = 0.1697, respectively. GCTA-COJO and FUMA collectively identified 28 independent risk SNPs associated with HOA. MAGMA and GCTA-fastBAT identified 48 pleiotropic genes. Integrating independent risk loci and pleiotropic genes culminated in the identification of 13 key genes associated with HOA. Multi-tissue TWAS and SMR analyses indicated that seven genes were associated with HOA across 22 tissues, with brain tissues accounting for 28.6%. Furthermore, PWAS and BLISS methods were utilized to analyze the proteomic features of these key genes.