While proton pump inhibitors (PPIs) show epidemiological associations with gastrointestinal disorders, their mechanistic basis remains unclear. The pharmacological effects of a drug are determined not only by its direct protein targets but also by genes involved in its metabolism and transport. Using two-sample Mendelian randomization and multi-ancestry cohorts (UKB/FinnGen), we systematically assessed genes encoding both direct targets of PPIs and key proteins involved in PPI pharmacokinetics across 24 gastrointestinal diseases. Validation included colocalization/SMR analyses, disease stratification, anatomical stratification, and HPA-based tissue expression profiling. Single-cell sequencing further elucidated target distribution. Finally, meta-analysis validates the correlation between PPI and gastrointestinal disorders. CYP2D6 emerged as the most pleiotropic locus (nine disease endpoints), demonstrating dose-dependent carcinogenic effects in hepatocellular/pancreatic cancers (OR = 1.36-1.58, p = 8.45 × 10-24, OR = 1.43-1.56, p = 2.31 × 10-72) yet protective against acute/chronic pancreatitis (OR = 0.72-0.75, p = 1.26 × 10-196, OR = 0.62-0.67, p = 1.39 × 10-111). ABCB1/SLC22A1 exhibited pan-gastrointestinal risk modulation. Anatomical stratification revealed context-dependent gene effects: SLC22A1 (esophageal), CYP2D6 (hepatic), and CYP2C19 (colorectal) showed opposing roles in carcinogenesis versus precancerous states. HPA profiling identified SLC22A3 as a pan-cancer candidate (moderate-high expression in 4 tumor types). Single-cell tumor data indicated AHR was enriched in tumor cells and significantly higher than in normal cells in all five tumor tissues. The finding suggests that AHR may be related to the potential gastrointestinal carcinogenicity of PPIs. The meta-analysis confirms a significant link between PPI use and increased gastrointestinal cancer risk. Further research is needed on PPI and benign gastrointestinal diseases. Our multi-omics integration reveals tissue-specific PPI pharmacodynamics, identifying targets with dual therapeutic/carcinogenic potential that may explain epidemiological discordances.
Keywords: drug targets; gastrointestinal diseases; mendelian randomization; meta‐analysis; proton pump inhibitor.
© 2025 Federation of American Societies for Experimental Biology.