Topological Complementarity Predicts Synergistic Drug Pairs in Branched Metabolic Networks.

Abstract

Drug discovery typically ranks targets by individual sensitivity. Branched metabolic networks, in contrast, have redundant flux routes that buffer single-enzyme inhibition possibly deviating the best target from the most sensitive node. We developed a topology-driven framework for identifying optimal dual-target pairs and applied it to the E. coli folate pathway. A kinetic model of the folate pathway and its extended one-carbon pool showed that single-node inhibition is compensated by flux redistribution. Pairwise synergy scoring, a metric to measure drug-pair potency as compared to single drug, identified DHPS (folP) and DHFR (folA) as a top-ranked complementary pair. Experimental validation in E. coli MG1655 using sulfamethoxazole and trimethoprim confirmed strong dose-dependent synergy at intermediate inhibition and attenuation at saturation. Our results support topological complementarity as a key predictor of combination efficacy and provide a rationale for drug-pair design in branched biosynthetic pathways.