How Persisters Patch Their DNA Back Together
Review written by: Megan Mauriello
Persisters are bacterial cells which are phenotypically tolerant to antibiotics. They are important contributors to chronic and recurrent infection and the development of antibiotic-resistant mutants. Fluoroquinolones (FQs) are effective antibiotics for treating non-growing bacteria, but persisters can survive. They target type II topoisomerases (including DNA gyrase and topoisomerase IV) and can produce single- and/or double-strand breaks. A growing body of evidence suggests that persisters survive FQ treatment by taking the DNA damage and repairing it after the treatment is over.
Three of the most prominent pathways for DNA repair are homologous recombination (HR), nucleotide excision repair (NER), and mismatch repair (MMR). HR machinery includes RecA, RecB, RecC, RecD, RecF, RecJ, RecN, and RecR. NER machinery includes UvrA, UvrB, UvrC, UvrD, Mfd, and Cho. MMR machinery includes MutH, MutL, and MutS. The RecFOR pathway repairs single-strand breaks and the RecBCD pathway repairs double-strand breaks, and RecA is involved in both pathways. NER includes global genomic repair (GGR) and transcription coupled repair (TCR). UvrABCD are involved in both types, but Mfd is only involved in some types of TCR. Cho, an excision nuclease, has homology to UvrC. UvrD is also involved in MMR.
Previous work has shown that, in stationary-phase Escherichia coli, recA or uvrD presence significantly increased persister levels - therefore, HR and/or NER can repair DNA damage to recover from FQ treatment. Additionally, inhibition of transcription or translation after FQ treatment increased persistor levels, a phenomenon the authors call inhibition-assisted recovery, that requires either recA or uvrD to be functional.
To better understand the roles of HR and NER in FQ persister survival, the authors tested deletions of DNA repair enzymes known to interact with RecA (HR) and/or UvrD (NER). None of the tested genes specific to MMR, mutH, mutL, and mutS, contributed to FQ persister recovery.
uvrA, uvrB, and mfd (all part of NER) epistatically interact with recA (HR) in FQ persister recovery. recA was the most important HR contributor to FQ persister recovery, and the recombination function of RecA was required. ΔuvrC ΔrecA, Δcho, and Δcho ΔuvrC ΔrecA did not phenotypically resemble ΔuvrD ΔrecA mutants, although Δcho ΔuvrC ΔrecA did show attenuated recovery from FQ treatment. In time-course experiments, ΔrecA mutants took 8 hours to recover from FQ treatment, ΔuvrD mutants took only 4 hours, compared to the 2 hour-recovery time of the wild-type. ΔuvrA, ΔuvrB, and Δmfd mutants had increased survival when exposed to chloramphenicol after FQ treatment, demonstrating that inactivity of each gene individually does not eliminate inhibition-assisted recovery. However, ΔuvrA, ΔuvrB, and Δmfd mutations in combination with ΔrecA eliminated inhibition-assisted recovery.
Similarly, recB and recC, and to an extent recN, (of HR) epistatically interact with uvrD (NER). uvrD was the most important NER contributor, and the helicase function of UvrD was required. ΔrecD ΔuvrD, ΔrecF ΔuvrD, ΔrecJ ΔuvrD, ΔrecO ΔuvrD, and ΔrecJ ΔuvrD mutants, and ΔuvrD mutants with an uninducible lexA allele for SOS response did not phenotypically resemble ΔuvrD ΔrecA mutants. However, ΔrecB ΔuvrD and ΔrecC ΔuvrD mutants exhibited no inhibition-assisted recovery, and ΔrecN ΔuvrD mutants had attenuated recovery. In time-course experiments, ΔrecB and ΔrecC mutants had increased survival when exposed to chloramphenicol after FQ treatment, demonstrating that inactivity of each gene individually does not eliminate inhibition-assisted recovery. However, ΔrecB and ΔrecC mutations in combination with ΔuvrD eliminated inhibition-assisted recovery.
These results allow for better mapping of the epistatic interactions of the E. coli persister recovery network, which potentially apply to other pathogens, especially those in the Enterobacteriaceae. Interactions tested showed complex patterns, suggesting that the epistatic network is more complicated and nuanced than this study can capture. It is important to understand the limits of FQ antibiotics as treatments for non-growing bacterial infections to maximize their efficacy alone or in combination with other treatments.
---
READ MORE: Soto-Echevarria N, Lemma AS, Brynildsen MP. Genetic analysis of epistasis between nucleotide excision repair and homologous recombination in the recovery of persisters after fluoroquinolone treatment. PLoS Genet. 2026 Sep 21;22(9):e1012305. doi: 10.1371/journal.pgen.1012305. PMID: 42766644; PMCID: PMC13592689.
