Proteomic insights into chronic pancreatitis: Proteome-wide mendelian randomization to identify potential therapeutic...

Chronic pancreatitis (CP) is characterized by progressive fibrosis and irreversible pancreatic dysfunction, lacking disease-modifying therapies. We aimed to identify druggable targets through proteome-wide mendelian randomization (PW-MR).

Protein quantitative trait loci (pQTLs) from UK Biobank Pharma Proteomics Project (2923 proteins) and deCODE study (4719 proteins) were analyzed. Genome-wide association study (GWAS) data for CP were sourced from the FinnGen (3875 cases and 361,641 controls). PW-MR identified proteins meeting the 5 % false discovery rate (FDR) threshold (FDR-corrected P < 0.05), followed by sensitivity analyses (Egger's intercept, Cochran's Q, contamination mixture, the Pleiotropy RESidual Sum and Outlier test). Prioritized proteins underwent colocalization, reverse MR, summary-data-based mendelian randomization (SMR), and external validation. Functional annotation included tissue specificity, Gene Ontology (GO)/Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment, and protein-protein interaction analyses.

Eight proteins met FDR <0.05 in both cohorts: four tier 1 targets (CST3, CTSB, MMP1, PI3) passed at least one sensitivity analysis, and the remaining four (AMY2A, GSTT2B, BCHE, MMP12) were classified as tier 2 targets. Eighteen additional tier 3 targets (e.g., AMY2B, IL6R, GRN) met FDR <0.05 in one cohort and nominal significance (P < 0.05) in the other. For tiers 1-3 targets, functional annotation revealed enrichment in digestive/metabolic pathways and extracellular matrix regulation. Drug target analysis identified eight proteins with available therapeutics, while the remaining targets currently lack clinical agents.

This study prioritizes 26 proteins (tiers 1-3 targets) as potential therapeutic targets for CP, bridging proteomic insights to clinical translation.