Phages exploit dormancy to defeat stubborn bacterial infections

One reason infections come back after treatment is that some bacteria slip into a dormant, non-replicating state when stressed by lack of nutrients, acid, or antibiotics. These sleepy cells are often invisible to antibiotics and can fuel relapse and the emergence of resistance. To better understand whether viruses that infect bacteria—bacteriophages—can reach and kill these hard-to-reach cells, a team led by Rachit Agarwal tested interactions between several phage species and different bacteria. The researchers worked with Mycobacterium smegmatis, Mycobacterium tuberculosis and Pseudomonas aeruginosa, putting the bacteria under stresses that force them out of active growth. Rather than simply failing to infect, many lytic phages were able to enter these non-replicating cells and persist in a dormant association called pseudolysogeny. In that state the phage genome remains extrachromosomal and the virus does not immediately kill the host. Importantly, when the bacteria began to regrow, the phages could resume their lytic cycle and destroy the bacterial cells. These observations suggest phages can track and lie in wait inside dormant bacteria instead of being wasted when targets are not actively dividing.

The study used quantitative experiments across multiple host–phage pairs to map how this behavior depends on both virus and host. Under conditions such as nutrient starvation, acidic pH or antibiotic pressure, lytic phages established pseudolysogeny rather than immediate lysis. The length of this pseudolysogeny window varied with the phage and the bacterial species, and the researchers traced loss of the dormant state to degradation of extrachromosomal phage DNA. A notable finding concerned bacterial immune defenses: Pseudomonas CRISPR systems actively degraded phage DNA even when the host was non-replicating, showing that CRISPR remains relevant under dormancy and can end pseudolysogeny. To test real-world relevance, the team showed that pseudolysogeny occurs in living systems and that CRISPR-resistant bacteriophages can eliminate implant-associated and antibiotic-persistent Pseudomonas infections in mice. Together these results reveal specific molecular and ecological steps—pseudolysogeny, phage DNA persistence or degradation, and CRISPR activity—that determine whether phage therapy will reach non-replicating bacteria.

These findings shift how we should think about using phages as therapeutics against persistent infections. First, the ability of lytic phages to enter a pseudolysogenic state means some phages can sit dormant with non-replicating bacteria and spring into action when conditions favor bacterial regrowth, offering a timed approach to killing cells that evade antibiotics. Second, bacterial defenses such as Pseudomonas CRISPR systems can actively remove phage genomes during dormancy, meaning that successful therapy must consider and potentially overcome these systems. Third, the demonstration that CRISPR-resistant bacteriophages can clear implant-associated and antibiotic-persistent infections in mice suggests tailored phage selection or engineering may be necessary for hard-to-treat infections. Overall, the work highlights that phage–host dynamics and bacterial immune mechanisms matter for designing effective phage-based treatments aimed at the non-replicating reservoirs that drive relapse and chronic infection.