DNA Repair Enzymes RecA and UvrD Emerge as Central Players in Antibiotic Persister Survival

Antibiotic persisters have long frustrated clinicians and researchers alike. These phenotypically tolerant bacterial cells are not resistant in the classical genetic sense; instead, they survive antibiotic exposure through physiological states that shield them from drugs designed to kill actively growing bacteria. Yet a growing body of evidence suggests that survival is not simply a matter of dormancy. For fluoroquinolones, one of the most effective antibiotic classes against non-growing bacteria, persisters appear to endure treatment by actively repairing the DNA damage the drugs inflict. A new study published in PLOS Genetics by Nashaly Soto-Echevarria, Annabel S. Lemma, and Mark P. Brynildsen of Princeton University now maps the genetic architecture underlying that repair process, revealing an intricate web of interactions between two major DNA repair pathways and pinpointing two proteins that sit at the heart of the network.

Fluoroquinolones, which include widely used drugs such as ciprofloxacin and levofloxacin, kill bacteria by trapping DNA gyrase and topoisomerase IV on the chromosome, converting these essential enzymes into agents of DNA destruction. The resulting double-strand breaks and associated lesions are lethal to most cells. But in stationary-phase cultures, where nutrients are scarce and most bacteria have stopped dividing, persisters survive fluoroquinolone treatment at high frequency. Previous work from the same research group demonstrated something surprising: transient growth inhibition after the antibiotic was removed significantly increased the recovery of persisters from stationary-phase Escherichia coli populations, but only if two specific genes were intact. One was recA, the central mediator of homologous recombination, and the other was uvrD, a DNA helicase that functions in nucleotide excision repair. That finding suggested that persister recovery depends on coordinated action between two repair systems that biologists usually study in isolation.

To dissect this relationship, the team turned to epistasis analysis, a classical genetic approach in which researchers construct double mutants and compare their behavior to that of the corresponding single mutants. If two genes operate in the same pathway, deleting both typically produces no worse a defect than deleting either one alone. If they operate in parallel, compensatory, or synergistic pathways, the double mutant reveals a different phenotype. By deleting DNA repair enzymes known to physically or functionally interact with RecA or UvrD, and then measuring how the resulting mutant strains recovered from fluoroquinolone treatment, the researchers could chart the boundaries of the persister recovery network with unprecedented resolution.

The results confirmed that the network extends well beyond the two original genes. Within nucleotide excision repair, the excision nuclease components uvrA and uvrB, along with mfd, a transcription-repair coupling factor that directs repair machinery to RNA polymerase-stalled sites, were all found to epistatically interact with recA during persister recovery. On the homologous recombination side, recB and recC, which encode the subunits of the RecBCD nuclease-helicase complex that processes DNA ends for recombination, emerged as additional epistatic partners of uvrD. These findings establish that both pathways genuinely contribute to the recovery phenomenon rather than serving as incidental bystanders, and they suggest that fluoroquinolone-induced damage in persisters generates substrates that require both excision of bulky lesions and recombinational repair of strand breaks.

Yet the data also revealed that the network is more nuanced than a simple story of one pathway compensating for the loss of the other. Loss of recA had far-reaching epistatic consequences that were not mirrored by the loss of other homologous recombination genes, indicating that RecA performs functions in persister recovery that its pathway partners do not fully duplicate. Similarly, loss of uvrD had a greater impact on the network than the loss of any other nucleotide excision repair gene tested. This asymmetry is significant because it implies that RecA and UvrD are not interchangeable cogs in a repair pipeline but rather hub proteins whose specific biochemical activities are indispensable for the recovery program. The double-mutant combinations across the two pathways produced patterns that could not be explained by straightforward redundancy, hinting at regulatory or structural roles that go beyond canonical repair chemistry.

To test whether the canonical enzymatic functions of these two hub proteins were actually required, the researchers deployed separation-of-function mutants. RecA(N304D) is a variant that retains some activities of the protein but is specifically defective in its recombination function, while UvrD(R284A) carries a mutation that abolishes its ATP-dependent helicase activity. If persister recovery depended merely on the physical presence of these proteins, for example as DNA-binding scaffolds, these mutants should have supported recovery. They did not. The recombination-proficient form of RecA and the helicase-active form of UvrD were both strictly required for their participation in fluoroquinolone persister recovery. This result anchors the phenomenon in biochemistry: the actual motor and strand-exchange activities of these enzymes, not just their occupancy of the DNA, drive the repair that allows persisters to resume growth after antibiotic treatment ends.

The findings carry weight for a long-standing debate in the microbiology of antibiotic tolerance. For years, persisters were assumed to survive because their sluggish metabolism rendered them invisible to antibiotics. Fluoroquinolones complicate that picture, because these drugs can form lesions even in metabolically quiescent cells, and the new work reinforces the view that survival after fluoroquinolone exposure is an active, repair-dependent process. In other words, persisters are not merely hiding from the drug; they are fighting back with the same DNA repair arsenal that dividing cells use, albeit under conditions of transient growth arrest that appear to give the repair machinery time to work. The post-treatment growth inhibition that boosts recovery in the laboratory may mimic conditions persisters encounter in infected tissues, where nutrient limitation and immune pressures can stall bacterial growth after antibiotic levels decline.

There are also therapeutic implications worth considering. If RecA and UvrD sit at the center of the persister recovery network, they become attractive targets for adjunctive drugs designed to sensitize persisters. Inhibitors that compromise RecA-mediated recombination or UvrD helicase activity could, in principle, be paired with fluoroquinolones to prevent the repair that underpins survival, thereby reducing the reservoir of tolerant cells that can seed recurrent infections. The epistasis map constructed in this study provides a rational starting point for such efforts, because it identifies which combinations of repair functions must be blocked to collapse the network. The asymmetry in the data suggests that targeting the hubs, RecA and UvrD, might be more effective than targeting peripheral components such as UvrA, UvrB, Mfd, RecB, or RecC.

Caveats remain, as they always do in laboratory genetics. The study was conducted in Escherichia coli grown to stationary phase in culture, and persister physiology in vivo, within the context of an actual infection, may involve additional stress responses, host factors, and heterogeneous metabolic states that alter the repair network’s composition. Moreover, epistasis analysis reveals functional relationships but does not by itself identify the molecular substrates being repaired or the sequence in which the pathways act. Future work combining the genetic framework established here with direct measurements of DNA damage, repair kinetics, and single-cell recovery dynamics will be needed to convert the network map into a mechanistic model.

Even so, the study represents a substantial advance in understanding how bacteria survive one of medicine’s most important antibiotic classes. By systematically expanding the persister recovery network and demonstrating that the core enzymatic activities of RecA and UvrD are non-negotiable requirements, Soto-Echevarria, Lemma, and Brynildsen have transformed a curious observation about post-treatment growth arrest into a genetically grounded framework. The work deepens appreciation of the interplay between homologous recombination and nucleotide excision repair in non-dividing cells, and it sharpenens the rationale for attacking persister survival at its enzymatic roots. As antibiotic resistance continues to erode treatment options worldwide, strategies that target tolerance rather than resistance are gaining urgency, and this study offers a detailed molecular blueprint for where such strategies might strike.

Subject of Research: Epistatic interactions between nucleotide excision repair and homologous recombination in fluoroquinolone persister recovery in Escherichia coli

Article Title: Genetic analysis of epistasis between nucleotide excision repair and homologous recombination in the recovery of persisters after fluoroquinolone treatment

Article References: Soto-Echevarria, N., Lemma, A. S., & Brynildsen, M. P. (2026). Genetic analysis of epistasis between nucleotide excision repair and homologous recombination in the recovery of persisters after fluoroquinolone treatment. PLOS Genetics, 22(9), e1012305. https://doi.org/10.1371/journal.pgen.1012305

Image Credits: AI Generated

DOI: 10.1371/journal.pgen.1012305

Keywords: antibiotic persisters, fluoroquinolones, DNA repair, RecA, UvrD, homologous recombination, nucleotide excision repair, epistasis, Escherichia coli, antibiotic tolerance, stationary phase, PLOS Genetics