The PhoU mutant identified in our previous transposon mutant screen has the transposon inserted near the C terminus of the phoU gene and has a more obvious persister phenotype than the phoU deletion mutant (Y. Li & Y. Zhang, unpublished data). Thus the finding that the PhoU deletion mutant selleck compound did not come up in our screen may be due to compensatory changes or mutations, which may indicate a limitation of the deletion mutant library approach. Like the PhoU mutant (Li & Zhang, 2007), the sucB and ubiF mutants have increased susceptibility to various stresses and different antibiotics with a two- to fourfold decrease in MIC and MBC (Table 1). It is generally assumed that mutations in genes involved

in persistence should not affect the MIC (Hansen et al., 2008). However, this may not necessarily be true. It is possible that mutation in a persister Lumacaftor molecular weight gene can affect antibiotic susceptibility not only in persisters but also in growing bacteria. As the current MIC and MBC testing is performed with a standard inoculum of 105–106 organisms of log phase cultures that may contain some persister bacteria already, it is likely that persisters may contribute to the MIC and MBC under normal MIC/MBC testing conditions. When the standard inoculum is inoculated into the culture medium containing antibiotics for MIC/MBC

testing, the mutants with defective persister formation are killed more rapidly than the wild-type bacteria at a given antibiotic concentration in the medium and therefore have lower MIC/MBC. In fact, all our persister-defective mutants, including phoU identified in the previous study (Li & Zhang, 2007) and ubiF and sucB identified in this study, have about two- to fourfold lower MIC/MBC than Coproporphyrinogen III oxidase the wild-type strain. A recent study, using the E. coli Keio mutant library screen to identify persistence genes with a short ofloxacin exposure of 6 h, found primarily stress response genes dnaJ and

dnaK (chaperones), apaH (diadenosine tetraphosphatase involved in stress resistance), surA (peptidyl-prolyl cis–trans isomerase involved in stationary phase survival), fis and hns (global regulators), hnr (response regulator of RpoS), dksA (RNA polymerase-binding transcription factor, a positive regulator of RpoS), ygfA (5-formyl-tetrahydrofolate cyclo-ligase) and yigB (FMN phosphatase) (Hansen et al., 2008). As we indicated previously (Li & Zhang, 2007), persisters are highly variable and have to be defined by specific conditions. Persisters may consist of different subpopulations of varying hierarchy in continuum (Zhang, 2007), and different times of antibiotic exposure may lead to different persister populations, with longer exposure causing increasingly fewer persisters, which can be called ‘deep persisters’ (with lower metabolism), which are not killed by antibiotics even with long antibiotic exposure.