From Pathway Tracing to Actionable Targets: Integrative Mendelian Randomization and Experimental Triangulation Map Metabolic Pathways Across Ovarian Cancer Histotypes

Ovarian cancer (OC) comprises multiple histotypes with distinct mechanisms, molecular features, and clinical behavior. We used Mendelian randomization (MR) to map histotype-stratified metabolic pathways and connect them to drug targets, establishing a translatable target–metabolic node–histotype risk chain. We built a multi-stage MR framework using Integrative Epidemiology Unit (IEU) OpenGWAS summary statistics. After screening 1400 plasma metabolites against overall ovarian cancer in UK Biobank and Ovarian Cancer Association Consortium (OCAC) with KEGG enrichment, we traced a prespecified amino acid/energy–nitrogen axis using histotype-stratified univariable MR and pathway-restricted multivariable MR. We then performed cis drug-target MR for PPARG, DPP4, ABCC8/KCNJ11, and SLC5A2, integrated triangulation, colocalization, and mediation analyses, and experimentally interrogated the prioritized PPARG/ABCC8-KCNJ11–lactate–invasive mucinous ovarian cancer (IMOC) triangle. Screening nominated 55 and 72 metabolites in UK Biobank and OCAC, respectively (IVW p < 0.05), highlighting amino-acid nitrogen and central-carbon metabolism. Univariable Mendelian randomization (UVMR) showed marked heterogeneity: alanine increased low-grade serous ovarian cancer (LGSOC) risk, glutamate was protective for endometrioid OC, and lactate-related traits most consistently implicated the low-grade/borderline serous lineage. In multivariable Mendelian randomization (MVMR), tryptophan and lactate levels emerged as independent risk nodes for serous low-grade plus low malignant potential (LG + LMP). Drug-target MR prioritized PPARG as protective (OR = 0.18) and ABCC8/KCNJ11 as risk-increasing (OR = 7.50) for IMOC, with opposite target → lactate effects supporting a directionally symmetric target–lactate–IMOC triangle. Experimental perturbation in mucinous ovarian cancer models produced concordant reciprocal changes in lactate and malignant phenotypes, extending this triangle biologically. This integrative MR framework delineates histotype-specific metabolic drivers and links them to actionable targets, providing a roadmap from genetic prioritization to mechanistic and translational validation.