INTEGRATING METABOLOMICS AND NETWORK BIOLOGY TO PRIORITIZE REPURPOSABLE DRUG FOR GALLSTONE DISEASE

Abstract

Gallstone disease (cholelithiasis) is one of the most common hepatobiliary metabolic disorders, which involve the development of solid masses in the gallbladder or biliary tract caused by disorders of cholesterol metabolism, bile acid homeostasis, gallbladder motility, and inflammatory regulation. Obesity, dyslipidemia, insulin resistance, metabolic syndrome and gut microbiota dysbiosis are all highly associated with the disease and they all are involved in the promotion of cholesterol supersaturation and lithogenic bile formation. The symptomatic treatment with ursodeoxycholic acid and cholecystectomy is associated with a risk of recurrence, the risk of surgical complications, but also with the lack of pharmacological alternatives that target the mechanisms involved. The aim of the present study was to identify and prioritize therapeutically relevant drug candidates for gallstone disease in an integrated framework of metabolomics-guided computational workflow. The metabolomics data from gallstone patients and healthy controls were retrieved from the Metabolomics Workbench database (ST004186) and underwent extensive preprocessing to ensure accuracy: feature filtering using quality control, probabilistic quotient normalization, missing value imputation, and log2 transformation. The findings of differential metabolite analysis and metabolite set enrichment analysis revealed major dysregulation of a multi-pathway disease architecture including those involved in bile acid biosynthesis and transport, fatty acid metabolism, amino acid metabolism, the glutathione-mediated antioxidant pathways and PPAR-associated lipid signalling. The altered metabolites were identified and assigned to genes associated with the disease, such as NR1H4, ABCB11, SLC10A2, SLCO1B1, CYP3A4, CYP2C9, ABCB1, PTGS2, and LIPC, etc., by the gene–drug interaction analysis. An integrated metabolite–gene–drug interaction network was then built in Cytoscape, which created a highly interconnected structure, with hub nodes focused on bile acid transport, cholesterol metabolism, hepatic drug processing and inflammatory signalling pathways. The network was broken down into 15 subnetworks for each drug, allowing for the granular analysis of each drug–target–metabolite relationship for the 15 candidate drugs. A multi-criteria network centrality scoring method, combining degree, betweenness, closeness, and eigenvector centrality, was also used to prioritize a list of candidate molecules, which was subsequently evaluated according to a systematic literature-based evaluation of biological relevance, pathway specificity, clinical safety profile and therapeutic novelty. This multi-layered evaluation resulted in the exclusion of statins on account of inconsistent evidence on gallstones and absence of pathway specificity to bile acid transport, anti-inflammatory agents such as Aspirin, Indomethacin, and Celecoxib due to toxicities, and Fenofibrate Micronized, Rifampin, and dietary flavonoids (Quercetin, Hesperetin, and Apigenin) due to known lithogenic effects on the bile, absence of pathway specificity to bile acid transport, and lack of regulatory approval and clinical evidence respectively. Curcumin was excluded for poor v bioavailability, Pioglitazone was excluded for serious adverse effects, and Colchicine was excluded due to absence of bile acid transport pathway engagement. Mechanistically, Odevixibat inhibited ileal bile acid transporter, which is the mechanism by which it targets the enterohepatic bile acid recirculation identified in the metabolomics results of this study, and was found to have the highest closeness centrality of all prioritized candidates in the drug network to ABCB11, the bile salt export pump, and SLC10A2, the apical sodium-dependent bile acid transporter. The approved clinical status, the alignment with the disease pathway and the therapeutic novelty in the context of gallstones met all the criteria for the multi-criteria prioritization framework. Overall, these results highlight the potential of using a combined approach of metabolomics, pathway enrichment analysis and network pharmacology as a computational tool for hypothesis-driven drug repurposing across complex metabolic diseases. These results support a biologically motivated computational rationale for the use of Odevixibat as a repurposable candidate in the treatment of gallstone disease and highlight the bile acid transport and enterohepatic circulation as the most therapeutically actionable axes in the pathogenesis of gallstones. These computational predictions need to be further validated in vitro, in vivo and clinically. In future, the integration of multi-omics, microbiome–metabolomics co-analysis, machine-learning assisted therapeutic prioritization, and patient-level metabolomics will be pursued to further develop systems medicine strategies in hepatobiliary diseases.