A novel high-throughput assay identifies small molecules with activity against persister cells

Persister cells are a subpopulation of transiently antibiotic tolerant bacteria, which are believed to be the main cause of relapsing bacterial infections. Due to the importance of persister cells in human infections, there is a need for new antibiotics that kill bacteria independently of their activity. However, high-throughput assays to screen for drugs with such activity are missing. This is partly due to the transient nature of the phenotype, which makes it is difficult to prepare a concentrated population of persister cells that remain inactive during incubation with antibiotics in standard growth media. The purpose of this study was to develop a simple and high-throughput assay to identify compounds with antimicrobial activity against persister cells during a 24 h incubation period. Subsequently, this assay was used to screen a selection of small molecules with hypothesized antimicrobial activity. The fraction of S. aureus that tolerate bactericidal concentrations of ciprofloxacin were defined as persister cells. We first quantified how the cell concentration, growth phase, antibiotic concentration, duration of antibiotic exposure, and presence/absence of nutrients during antibiotic exposure affected the fraction of persister cells in a population. After optimizing these parameters, we compared our approach to generate persister cells, to a process that generated persister cells by a short exposure to rifampicin. Finally, we used the optimised protocol to identify molecular structures that have anti-persister activity by performing screening on initially compound fragments and then selecting compounds that incorporated the fragments that displayed activity. We show that exponential- and stationary-phase cultures transferred to nutrient-rich media only contain a small fraction (0.001 to 0.07 %) of persister cells that tolerated 10, 50 and 100 x MIC ciprofloxacin. Exponential-phase cultures displayed a bi-phasic time-kill curve, which plateaued after 5 h exposure, while stationary phase cultures displayed a low, but constant death rate at 50 and 100 x MIC ciprofloxacin. Inducing the persister phenotype with a short rifampicin treatment resulted in 100% persister cells when evaluated after [≤]5 h exposure to ciprofloxacin. However, after longer incubation times, cells resumed activity and lost their tolerance to ciprofloxacin. Tolerance was only maintained in the majority of the population for the full 24 h incubation period if cells were transferred to a carbon-free minimal medium before exposure to ciprofloxacin. We conclude that keeping cells starved in a carbon-free medium enables generation of high concentrations of S. aureus cells that tolerate 50 x MIC ciprofloxacin, and we find this protocol easily applicable for rapid screening of anti-persister drugs that act on dormant or non-dividing cells.