Phage steering in the presence of a competing bacterial pathogen

Abstract

The rise of antibiotic-resistant bacteria has necessitated the development of alternative therapeutic strategies, such as bacteriophage therapy, where viruses infect bacteria, reducing bacterial burden. However, rapid bacterial resistance to phage treatment remains a critical challenge, potentially leading to failure. Phage steering, which leverages the evolutionary dynamics between phage and bacteria, offers a novel solution by driving bacteria to evolve away from virulence factors or resistance mechanisms. In this study, we examined whether phage steering using bacteriophage Luz19 could function in the presence of a competing pathogen, Staphylococcus aureus (SA) (USA300), while targeting Pseudomonas aeruginosa (PAO1). Through in vitro co-evolution experiments with and without the competitor, we observed that Luz19 consistently steered P. aeruginosa away from the Type IV pilus (T4P), a key virulence factor, without interference from SA. Genomic analyses revealed mutations in T4P-associated genes, including pilR and pilZ, which conferred phage resistance. Our findings suggest that phage steering remains effective even in polymicrobial environments, providing a promising avenue for enhancing bacteriophage therapy efficacy in complex infections.IMPORTANCEPhage steering-using phages that bind essential virulence or resistance-associated structures-offers a promising solution by selecting for resistance mutations that attenuate pathogenic traits. However, it remains unclear whether this strategy remains effective in polymicrobial contexts, where interspecies interactions may alter selective pressures. Here, we demonstrate that Pseudomonas aeruginosa evolves phage resistance via loss-of-function mutations in Type IV pilus (T4P) when challenged with the T4P-binding phage Luz19 and that this evolutionary trajectory is preserved even in the presence of a competing pathogen, Staphylococcus aureus. Phage resistance was phenotypically confirmed via twitching motility assays and genotypically via whole-genome sequencing. These findings support the robustness of phage steering under interspecies competition, underscoring its translational potential for managing complex infections-such as those seen in cystic fibrosis-where microbial diversity is the norm.

Keywords: bacteriophage evolution; bacteriophage therapy; bacteriophages; steering; virulence.

Fig 1

Twitching motility compared across evolution conditions. PA twitching motility decreased significantly (P < 0.05) in conditions exposed to Luz19 phage relative to the ancestral PAO1, with no significant difference in twitching motility measured between the positive-phage groups with/without SA (P > 0.05) or between non-phage conditions and the ancestral control (P > 0.05). Boxplot ranges are 1.5× interquartile ranges.

Fig 2

Growth assays and area under the curve (AUC). (A) Control treatment AUC boxplots (above) and mean growth curves (below). PA from the five conditions was grown in LB with no additional challenge—PA evolved alongside SA had significantly higher AUC than all other conditions (P < 0.05). (B) SA treatment AUC boxplots (above) and mean growth curves (below). PA grown in LB with SA added. PA evolved with SA had significantly greater AUC than the ancestral (P < 0.05) but not significantly different from the transfer control PAO1t10 (P > 0.05). (C) Luz19 treatment AUC boxplots (above) and mean growth curves (below). Conditions subject to co-evolution with Luz19 (+Luz19 and +SA+Luz19) had significantly greater AUC than those not evolved with the phage (P < 0.05) and were not significantly different from one another (P > 0.05). There was no significant difference between the AUC of conditions not previously exposed to the phage (P > 0.05). (D) SA and Luz19 treatment AUC boxplots (above) and mean growth curves (below). Conditions subject to co-evolution with Luz19 (+Luz19 and +SA+Luz19) had significantly greater AUC than those not evolved with the phage (P < 0.05) and were not significantly different from one another (P > 0.05). There was no significant difference between the AUC of conditions not previously exposed to the phage (P > 0.05). Boxplot ranges are the 1.5× interquartile ranges, and error bars on the growth curves are standard error.

Fig 3

Chromosomal single-nucleotide polymorphisms (SNPs) defined for all co-evolution condition replicates and the ancestral PAO1. SNPs were determined using breseq (version 0.31). SNPs of two variations in 5 out of 5 (5/5) +Luz19 co-evo replicates and 3/5 +SA+Luz19 co-evo replicates. In all but one of these replicate sequences, the SNP in pilR is C192Y (TGC → TAC). Replicate B of the +Luz19 condition had a different SNP in pilR, P189Q (CCG → CAG). Of the replicates in the two conditions that were subject to co-evolution with Luz19, two lacked an SNP in pilR, both in the +SA+Luz19 condition. Replicate B instead had a 197 bp deletion in pilZ. Replicate A of +SA+Luz19 was the only replicate co-evolved alongside Luz19 that sequencing did not reveal a pil gene mutation; instead, a PQS system mutation was identified in pqsR.

Fig 4

Mutations in key regulatory factors of T4P biogenesis confer phage resistance. (A) PilR serves as the response regulator of pilA transcription. The sensor kinase, PilS, phosphorylates PilR, activating it, which, in turn, activates transcription of pilA, which encodes for the major pilin subunit, PilA. (B) PilZ regulates T4P biogenesis. By binding to the extension ATPase, PilB activates shaft polymerization. PilZ and FimX are necessary for PilB activation. (C) Without PilZ binding, PilB remains inactive, and energy is not generated for shaft assembly. (D) Defective PilR does not activate transcription of pilA, no PilA subunits are synthesized, and thus, the shaft is not assembled.