Antibiotic-persistent bacterial cells exhibiting low-level ROS are eradicated by ROS-independent membrane disruption

Bacterial persistence increases therapy duration, disease relapse, and antibiotic resistance. Mechanisms underlying persistence and feasible ways to rapidly eliminate persister cells are largely unknown. The present work examined genetic and environmental perturbations to identify anti-death events occurring in Escherichia coli persister and phenotypically tolerant cells. The quiescent status of hipA7 and metG2 persister cells, which were protected from killing by multiple antibiotics, was insensitive to the presence/absence of exogenous nutrients. In contrast, stationary-phase and nutrient-starved wild-type cultures, which displayed tolerance rather than the subpopulation status of persistence, were readily killed by ciprofloxacin upon restoration of nutrients, thereby indicating that tolerance was phenotypic. Both persistent and tolerant cells suppressed accumulation of reactive oxygen species (ROS), DNA breakage, and global metabolic activity. Restoration of nutrients to stationary-phase cultures restored these three processes for phenotypically tolerant cells but not for persister cells. Cultures of high-frequency-persistent hipA7 and metG2 mutants and low-frequency-persistent wild-type cells were rapidly sterilized by ROS-independent, synergistic membrane disruption using aminoglycoside-polymyxin combinations; rapid eradication occurred at clinically achievable concentrations for both antibiotics. The aminoglycoside-polymyxin combination also killed environmentally tolerant cells but more slowly. The combination killed both laboratory and clinical isolates of gram-negative bacteria, an E. coli ptsI mutant that is pan-tolerant to diverse antibiotics and disinfectants, and E. coli cells in a biofilm model. Moderate lethality was observed with the gram-positive bacterium Staphylococcus aureus. The work indicates that suppression of ROS accumulation is a common feature of persistence and phenotypic tolerance, and it emphasizes ROS-independent strategies for controlling quiescent bacterial populations.

The report generalizes the concept that persistence and tolerance involve suppression of toxic metabolites (reactive oxygen species [ROS]). The work also shows that an environmental perturbation (nutrient deprivation) leads to antibiotic tolerance rather than persistence, thereby raising questions about the classification of other environmental perturbations. The synergistic action of multiple aminoglycoside species with polymyxins opens many treatment options. Lethality with biofilms and with S. aureus may extend polymyxin-based therapies beyond planktonic, gram-negative bacteria, and the ROS independence of the combination may allow antioxidant mitigation of drug toxicity. Overall, the work advances our knowledge of persistent and tolerant bacterial pathogens and our efforts to eradicate them.