Rational Design of Resistance-Suppressing Phage-Antibiotic Cocktails via Receptor Tradeoffs

Abstract

Bacteriophages (phages) are promising antibiotic alternatives, yet their utility is limited by rapid resistance evolution. Current phage-antibiotic combination strategies rely on empirical screening for efficacy in killing, which often fails to prevent resistance. Here, we establish a mechanistic framework to rationally design phage-antibiotic cotreatments that specifically suppress resistance by leveraging evolutionary tradeoffs. We show that traditional killing efficacy metrics do not predict resistance suppression. Instead, we identify predictable evolutionary trade-offs, including species-specific targets like TolC (Escherichia coli) and Type IV pili (Pseudomonas aeruginosa). We also identify a universal vulnerability for Gram-negative bacteria, lipopolysaccharide (LPS). We find that LPS-targeting phages are ubiquitous and that phage-steered LPS perturbation creates a predictable hypersensitivity to lipophilic antibiotics. This mechanistic trade-off, governed by physicochemical rules rather than drug-specific targets, directly correlates with resistance suppression. This framework shifts combination therapy from empirical screening to the a priori design of evolutionarily stable cocktails.