Phage resistance accompanies reduced fitness of uropathogenic E. coli in the urinary environment

Urinary tract infections (UTIs) are among the most common infections treated worldwide each year and are primarily caused by uropathogenic E. coli (UPEC). Rising rates of antibiotic resistance among uropathogens have spurred consideration of alternative strategies such as bacteriophage (phage) therapy; however, phage-bacterial interactions within the urinary environment are poorly defined. Here, we assess the activity of two phages, HP3 and ES17, against clinical UPEC isolates using in vitro and in vivo models of UTI. In both bacteriologic medium and pooled human urine, we identified phage resistance arising within the first 6-8 hours of coincubation. Whole genome sequencing revealed that UPEC resistant to HP3 and ES17 harbored mutations in genes involved in lipopolysaccharide (LPS) biosynthesis. These mutations coincided with several in vitro phenotypes, including alterations to adherence to and invasion of human bladder epithelial HTB-9 cells, and increased biofilm formation. Interestingly, these phage-resistant UPEC demonstrated reduced growth in pooled human urine, which could be partially rescued by nutrient supplementation, and were more sensitive to several outer membrane targeting antibiotics than parental strains. Additionally, these phage-resistant UPEC were attenuated in a murine UTI model. In total, our findings suggest that while resistance to phages, such as LPS-targeted HP3 and ES17, may readily arise in the urinary environment, phage resistance is accompanied by fitness costs rendering UPEC more susceptible to host immunity or antibiotics. IMPORTANCEUTIs are one of the most common causes of outpatient antibiotic use, and rising antibiotic resistance threatens the ability to control these infections unless alternative treatments are developed. Bacteriophage (phage) therapy is gaining renewed interest, however, much like antibiotics, bacteria can readily become resistant to phage. For successful UTI treatment, we must predict how bacteria will evade killing by phage and identify the downstream consequences of phage-resistant bacterial infections. In our current study, we found that while phage-resistant mutant bacteria quickly emerged, these mutations left bacteria less capable of growing in human urine and colonizing the murine bladder. These results suggest that phage therapy poses a viable UTI treatment if phage resistance confers fitness costs for the uropathogen. These results have implications for developing cocktails of phage with multiple different bacterial targets, each of which is only evaded at the cost of bacterial fitness.