FIG 4.

Hydrogen peroxide selects for the SCV phenotype. (A) Tryptic soy broth aliquots were inoculated with 10⁵ CFU tetracycline-sensitive wild-type S. aureus SH1000 and 10 CFU tetracycline-resistant SCVs in the absence (TSB) or presence of either H₂O₂ or paraquat (para), and the sizes of the total (open bars) and SCV (filled bars) populations were determined after 16 h of culture. (B) After 16 h of culture, the percentage of SCVs that were resistant to tetracycline was determined. Values significantly (P < 0.05 [Student t test]) different from those obtained with bacteria cultured in the absence of oxidants (TSB) are indicated (*). Bars represent the mean averages of 12 independent cultures. Error bars represent the standard deviations of the mean. (C, D, and E) The relative stability of SCVs isolated from cultures without (TSB) (C) or with H₂O₂ (D) or paraquat (para) (E) were determined by using a previously described assay (22). Individual SCVs were classified as stable (stab), partially stable (part), or unstable (unst) as described in Materials and Methods. The data points represent a single independent culture. The percentag of SCVs classified as unstable was significantly lower in cultures containing oxidants than those without (an asterisk [*] indicates a significant difference relative to TSB without oxidants). (F and G) SCVs that were incubated either in the absence (F) or presence (G) of a subinhibitory concentration of H₂O₂ were assessed for phenotypic stability. Each data point represents a single independent culture. SCVs that were exposed to H₂O₂ or paraquat were significantly more stable than those incubated in TSB alone. Each data point represents a single culture (10 from each condition). Significant differences between each stability category (P < 0.05 [Student t test corrected for multiple comparisons via the Bonferroni method]) are indicated (*).