Training a phage to expand its host range: directed evolution against Staphylococcus aureus from diabetic foot infections
TL;DR
Both the potential and the limitations of phage training to expand therapeutic coverage are demonstrated and the need to integrate evolutionary approaches with physiologically relevant models to optimize phage therapy for chronic infections, such as DFIs is emphasized.
Abstract
ABSTRACT Diabetic foot infections (DFIs) are a major complication of diabetes frequently involving multidrug-resistant Staphylococcus aureus. Their persistence and therapeutic complexity underscore the urgent need for alternatives to conventional antibiotics. Phage therapy offers a promising solution, though its clinical application is often limited by the narrow host range of individual phages. Here, SAVM02, a recently characterized staphylococcal Kayvirus, was trained using directed evolution against DFI isolates. After 20 iterative passages on a mixed panel of susceptible and non-susceptible S. aureus strains, the trained phage pool (SAVM02-P20) displayed markedly enhanced activity, infecting approximately 77% of tested S. aureus isolates and showing partial cross-species activity against coagulase-negative staphylococci. Improved infectivity was linked to increased replication efficiency rather than adsorption. Genomic and phylogenetic analyses further suggested recurrent adaptive changes and preferential host-range expansion within genetically related lineages. In vivo testing in a zebrafish embryo model confirmed therapeutic efficacy against a fully susceptible strain but revealed no survival benefit against an initially poorly permissive isolate, highlighting the gap between in vitro adaptation and in vivo efficacy. These findings demonstrate both the potential and the limitations of phage training to expand therapeutic coverage and emphasize the need to integrate evolutionary approaches with physiologically relevant models to optimize phage therapy for chronic infections, such as DFIs.