Phage Resistance Reshapes Drug Response in Pseudomonas aeruginosa
Phage resistance did not behave like a single biological switch in this Pseudomonas aeruginosa preprint. Across eight strains, the collateral effects moved in different directions depending on the bacteriophage, the antibiotic, and the genetic background of the host strain.
The work focused on a practical question with direct relevance to combination therapy: when bacteria evade phage killing, do they become easier or harder to treat with antibiotics? The answer here was mixed, but not random. Resistance to PAO1-1 produced clear collateral sensitivity to colistin and polymyxin B, and in the same PAO1-1-resistant background the response to tobramycin also shifted toward sensitivity. Resistance to the other four phages tested pushed the phenotype in the opposite direction for that aminoglycoside, with collateral resistance rather than sensitivity.
One pattern stood out immediately. Phage escape did not create a universal antibiotic weakness; it created a set of genotype-dependent liabilities.
How the Experiment Measured Collateral Shifts
Investigators generated mutants resistant to phage PAO1-1 in each of eight strains, including seven clinical isolates and the laboratory strain PAO1. Separate mutant sets were created for resistance to six antibiotics, and additional phage-resistant mutants were produced using four more phages: M32, NP1, NP3, and PNM. A double-resistant group was also created in one strain, allowing direct comparison of how combined phage and antibiotic resistance reshaped downstream susceptibility.
Bacterial burden was quantified through growth curves across antibiotic concentration ranges, then converted into area-under-the-curve values. From there, dose-response curves were built to estimate fold change in EC50 relative to the parental strains. That design matters because it captures the overall growth impact of the drug, not only survival at a single concentration. In a pathogen such as P. aeruginosa, where resistance mechanisms can change membrane structure, transport, and metabolic balance at once, a full concentration-response curve reveals far more than a binary susceptible-or-resistant label.
Colistin and Polymyxin B Show the Sharpest Sensitivity Shift
The clearest drug-class signal involved the polymyxins. PAO1-1-resistant mutants showed significant collateral sensitivity to both colistin and polymyxin B. Those findings fit the biology of phage resistance in a Gram-negative organism, where mutations that alter surface structures used by phages can also disrupt envelope properties that matter for polymyxin action.
Because polymyxins depend on interactions with the bacterial outer membrane, even modest changes in surface composition can tilt the balance back toward drug susceptibility. In clinical terms, a phage-resistance mutation may leave the organism exposed at a point where last-line antibiotics still exert pressure. That is a useful possibility in difficult infections, but it is not a blanket rule across all phages or all strains.
Tobramycin Flips Between Sensitivity and Resistance
The aminoglycoside result was more complicated. PAO1-1-resistant mutants showed collateral sensitivity to tobramycin, while mutants resistant to the other four phages were associated with collateral resistance. The direction of effect therefore depended on which phage drove the resistance phenotype in the first place.
This split is clinically important because aminoglycoside susceptibility can be influenced by membrane permeability, efflux, and broader physiological stress responses. A mutation that weakens one part of the bacterium’s defensive architecture may make drug entry easier. Another mutation, even if it also blocks a phage receptor, may trigger a different envelope adaptation and reduce intracellular antibiotic exposure. The study does not flatten those pathways into a single mechanism; it shows how phage selection can produce distinct collateral phenotypes in the same species.
Double-Resistant Mutants Still Carried Weak Points
Double phage- and antibiotic-resistant mutants from one strain remained vulnerable in specific ways. Even after accumulating resistance to both classes, these mutants still showed collateral sensitivity to colistin and tobramycin relative to the parental strain.
That finding is a reminder that resistance stacking does not necessarily close every therapeutic door. The bacterial genome can absorb a tradeoff only so far before changes in membrane organization, metabolism, virulence regulation, and DNA replication begin to create new liabilities. In a treatment setting, this kind of tradeoff could be useful if a phage is chosen not only to kill directly, but also to steer the pathogen into an antibiotic-vulnerable state.
Resistance Mutations Clustered in Core Cellular Pathways
Sequencing and annotation of resistance mutations pointed to genes involved in metabolism, virulence, and DNA replication. Those categories are broad, but they are biologically plausible in a bacterium under phage pressure. Receptor loss is only part of the story; phage escape can reverberate through cell physiology, and those reverberations may be what reshape antibiotic response.
The authors also found the reverse relationship: antibiotic-resistant mutants became more sensitive to phages, with the strongest effect in colistin-resistant mutants. In other words, the same evolutionary pressure that protects the bacterium from a drug can expose it to viral predation. That reciprocal vulnerability is the conceptual center of phage-antibiotic combination design.
Why the Strain and Phage Pairing Matters So Much
The most defensible conclusion from this preprint is not that phage resistance causes collateral sensitivity in general. It is narrower than that and more useful: collateral effects were most specific to both strain and antibiotic. The same phage did not generate identical antibiotic phenotypes across the tested panel, and the same antibiotic did not respond the same way across phage backgrounds.
For clinicians and translational microbiologists, that means any attempt to use phage to sensitize P. aeruginosa to antibiotics will need organism-specific and phage-specific validation. A therapeutic strategy built on collateral sensitivity cannot rely on class-wide assumptions. It will need pairing logic, susceptibility testing, and a clear sense of which resistance pathway is likely to emerge under selection.
This preprint does not claim clinical deployment data, and it does not solve the dosing or sequencing problem. What it does show is a mechanistic opening: phage resistance can be harnessed to alter antibiotic sensitivity, but the effect depends on the exact phage, the exact strain, and the resistance route that evolves.