Genetically Predicted Gene Expression and Circulating Metabolites Associated with Cervical High-Grade Squamous Intraepithelial Lesion: A Mendelian Randomization Study
Source: PubMed Central Open Access, NCBI / U.S. National Library of Medicine
Background High-grade squamous intraepithelial lesion (HSIL) is a precancerous condition of the cervix. Identifying risk factors associated with HSIL and understanding their potential mechanisms may inform prevention strategies. This study aimed to investigate the associations of genetically predicted gene expression and circulating metabolites with HSIL risk using Mendelian randomization (MR). Methods We performed two-sample MR analysis to evaluate the associations of genetically predicted gene expression (eQTLGen consortium, N=31,684) and circulating metabolites (genome-wide association study [GWAS], N=8,299) with HSIL risk (FinnGen R12, N=293,218; 8,291 cases). Mediation analysis was conducted to explore whether metabolites might mediate the associations between genes and HSIL. Sensitivity analyses, including Mendelian randomization pleiotropy residual sum and outlier (MR-PRESSO), leave-one-out, and colocalization, were performed to assess the robustness of the findings. All GWAS data used in this study were derived from European-ancestry populations. Results Eleven genes showed significant associations with HSIL after false discovery rate (FDR) correction (q 0.98), while COL11A2 showed weak evidence of colocalization (PP.H4=1.58×10). Functional enrichment analysis indicated that COL11A2-related genes were enriched in extracellular matrix (ECM)-receptor interaction and PI3K-Akt signaling pathways. Conclusion This MR study identified 11 genes and 11 circulating metabolites
Abstract
Background High-grade squamous intraepithelial lesion (HSIL) is a precancerous condition of the cervix. Identifying risk factors associated with HSIL and understanding their potential mechanisms may inform prevention strategies. This study aimed to investigate the associations of genetically predicted gene expression and circulating metabolites with HSIL risk using Mendelian randomization (MR). Methods We performed two-sample MR analysis to evaluate the associations of genetically predicted gene expression (eQTLGen consortium, N=31,684) and circulating metabolites (genome-wide association study [GWAS], N=8,299) with HSIL risk (FinnGen R12, N=293,218; 8,291 cases). Mediation analysis was conducted to explore whether metabolites might mediate the associations between genes and HSIL. Sensitivity analyses, including Mendelian randomization pleiotropy residual sum and outlier (MR-PRESSO), leave-one-out, and colocalization, were performed to assess the robustness of the findings. All GWAS data used in this study were derived from European-ancestry populations. Results Eleven genes showed significant associations with HSIL after false discovery rate (FDR) correction (q 0.98), while COL11A2 showed weak evidence of colocalization (PP.H4=1.58×10). Functional enrichment analysis indicated that COL11A2-related genes were enriched in extracellular matrix (ECM)-receptor interaction and PI3K-Akt signaling pathways. Conclusion This MR study identified 11 genes and 11 circulating metabolites associated with HSIL risk. Among these, COL11A2 showed a protective association that appeared to be largely independent of circulating phospholipid metabolites, suggesting potential local mechanisms. These findings provide genetic and metabolic clues for future studies on HSIL etiology. Graphical Abstract The flowchart illustrates the process of analyzing multi-omics data through Mendelian randomization. It starts with ′Multi-Omics Data′ including ′Gene expression (eQTLGen, N=31,684)′ and ′Metabolites (GWAS, N=8,299)′, leading to ′HSIL outcome (FinnGen, 8,291 cases)′. This data is processed using ′Two-sample MR′ methods: IVW, MR-Egger, Weighted median and MR-PRESSO, with additional analyses like Mediation analysis, Colocalization and Leave-one-out. The ′Key Findings′ section highlights ′Causal Genes′ such as VWA7, PAX8 and others and ′Circulating Metabolites′ including Phospholipids and Sphingomyelin. A ′Novel Finding′ is noted: COL11A2 with an odds ratio of 0.60, indicating a protective effect linked to ECM-receptor interaction and PI3K-Akt signaling. A flowchart of multi-omics data analysis and key findings in Mendelian randomization. http://www.w3.org/1999/xlink print-only float portrait IJWH-18-632165-g0001.webp anchor uf0001 portrait graphical