Two small (p)ppGpp synthases in Staphylococcus aureus mediate tolerance against cell envelope stress conditions

Abstract

The stringent response is a conserved global regulatory mechanism that is related to the synthesis of (p)ppGpp nucleotides. Gram-positive bacteria, such as Staphylococcus aureus, possess three (p)ppGpp synthases: the bifunctional RSH (RelA/SpoT homolog) protein, which consists of a (p)ppGpp synthase and a (p)ppGpp hydrolase domain, and two truncated (p)ppGpp synthases, designated RelP and RelQ. Here, we characterized these two small (p)ppGpp synthases. Biochemical analyses of purified proteins and in vivo studies revealed a stronger synthetic activity for RelP than for RelQ. However, both enzymes prefer GDP over GTP as the pyrophosphate recipient to synthesize ppGpp. Each of the enzymes was shown to be responsible for the essentiality of the (p)ppGpp hydrolase domain of the RSH protein. The staphylococcal RSH-hydrolase is an efficient enzyme that prevents the toxic accumulation of (p)ppGpp. Expression of (p)ppGpp synthases in a hydrolase-negative background leads not only to growth arrest but also to cell death. Transcriptional analyses showed that relP and relQ are strongly induced upon vancomycin and ampicillin treatments. Accordingly, mutants lacking relP and relQ showed a significantly reduced survival rate upon treatments with cell wall-active antibiotics. Thus, RelP and RelQ are active (p)ppGpp synthases in S. aureus that are induced under cell envelope stress to mediate tolerance against these conditions.

FIG 1

The essentiality of the RSH-hydrolase to prevent toxic accumulation of (p)ppGpp. (A) Growth of wild-type strain HG001, the rsh_syn mutant, the conditional rsh_cond mutant, the rsh_cond relP double mutant, and the rsh_cond relQ double mutant in CYPG medium without the inducer IPTG. Growth curves represent the means of four independent experiments. (B) Detection of (p)ppGpp accumulation in the conditional rsh_cond mutant due to RelP and/or RelQ by TLC. Strains were grown with IPTG to the early exponential phase and shifted to different conditions: Extended incubation for 2 h without mupirocin and without IPTG resulted in accumulation of ppGpp and pppGpp in the conditional rsh_cond strain but not in the triple mutant (rsh_cond relP relQ) strain or strains with active hydrolase (wild type or rsh_Syn mutant). As a control, the stringent response was induced for 20 min with mupirocin (0.5 μg/ml) in the conditional rsh_cond mutant grown with and without IPTG. (C) Overexpression of (p)ppGpp synthases from an ATc-inducible promoter in a (p)ppGpp⁰ strain showed decreased abilities to form colonies on TSA plates. Wild-type, (p)ppGpp⁰, (p)ppGpp⁰ plus pRelP, (p)ppGpp⁰ plus pRelQ, and (p)ppGpp⁰ plus pRSH_Hyd strains were grown to the exponential growth phase (OD₆₀₀ of 0.5, time 0), followed by a further incubation with 0.2 μg/ml ATc. At indicated time points, percent survival was calculated by counting the number of cells able to form colonies, normalized to CFU before ATc treatment (time 0). CFU counts represent the means of three independent experiments.

FIG 2

The staphylococcal RSH-hydrolase is very efficient to degrade (p)ppGpp molecules. Shown are growth of the WT, WT plus pRelP, WT plus pRelQ, and WT plus pRSH_Hyd strains (A) and growth of the (p)ppGpp⁰, (p)ppGpp⁰ plus pRelP, (p)ppGpp⁰ plus pRelQ, and (p)ppGpp⁰ plus pRSH_Hyd strains (B) to the exponential growth phase, followed by further incubation with (filled symbols) or without (open symbols) 0.2 μg/ml ATc. Growth curves represent the means of four independent experiments.

FIG 3

In vivo and in vitro evidence for (p)ppGpp synthetic activities of RelP and RelQ. (A) Detection of (p)ppGpp synthesis after amino acid deprivation in vivo. ³²P-labeled nucleotides of formic acid extracts of S. aureus were detected by thin-layer chromatography. Strain HG001 (WT), the rsh_syn mutant, and the rsh_syn relP relQ [(p)ppGpp⁰] triple mutant were grown to the exponential phase, followed by the addition of mupirocin (0.5 μg/ml) for 20 min. (B) Characterization of the (p)ppGpp synthase activities of RelP and RelQ in vitro. Purified RelP or RelQ proteins (0.5 μg) were assayed for (p)ppGpp synthase activity in the presence of [γ-³²P]ATP, 2 mM ATP, and either 2 mM GTP or 2 mM GDP. Reaction mixtures were analyzed by 1D thin-layer chromatography. The positions of the origin and signals corresponding to pppGpp, ppGpp, and ATP are indicated.

FIG 4

Induction of a stringent response by overexpression of relP and relQ. Shown is (p)ppGpp synthase-dependent repression or induction of target genes. Wild-type (WT) strain HG001, the (p)ppGpp⁰ mutant, and the (p)ppGpp⁰ mutant complemented with pRelP, pRelQ, or pRSH_Hyd synthases were grown in CYPG to the exponential growth phase followed by further incubation with (+) or without (−) ATc (0.2 μg/ml) for 1 h. RNA was hybridized with digoxigenin-labeled PCR fragments. The 16S rRNA detected in the ethidium bromide-stained gels is indicated as loading control in the bottom lane.

FIG 5

RelP and RelQ synthases mediate tolerance to cell wall-active antibiotics. (A) Upregulation of genes upon vancomycin and ampicillin treatments. Wild-type strain HG001 and its isogenic vraR mutant were grown in CYPG medium to the exponential growth phase (OD₆₀₀ of 0.5), followed by the addition of vancomycin (3.12 μg/ml, corresponding to 5× the MIC) and ampicillin (9.4 μg/ml, corresponding to 6× the MIC) for 30 min. RNA was hybridized with digoxigenin-labeled PCR fragments. The 16S rRNA detected in the ethidium bromide-stained gels as a loading control is indicated at the bottom. (B) Growth behavior of strains upon treatments with vancomycin and ampicillin. Strains were grown to the exponential phase in Mueller-Hinton (MH) broth and subsequently treated with 3.12 μg/ml vancomycin or 9.4 μg/ml ampicillin for 24 h. (C) Time-kill curves of strains in the presence of vancomycin, ampicillin, and ciprofloxacin. Strains grown to the exponential growth phase (OD₆₀₀ of 0.5, time 0) were treated with 3.12 μg/ml (5× the MIC) vancomycin, 9.4 μg/ml (6× the MIC) ampicillin, and 2.5 μg/ml (10× the MIC) ciprofloxacin for 24 h. Viable counts were determined by plating known dilutions of the samples on tryptic soy agar (TSA) plates before antibiotic treatments (time 0) and after 24 h. Time-kill curves represent the means of four independent experiments. The levels of significance after 24 h were determined by the two-tailed Student's t test. After vancomycin treatment, the significance level was P < 0.005 for the comparisons of the WT versus the relP relQ mutant and the WT versus the (p)ppGpp⁰ mutant. After ampicillin treatment, the significance level was P < 0.05 for the comparisons of the WT versus the relP relQ mutant and the WT versus the (p)ppGpp⁰ mutant. There were no significant differences upon ciprofloxacin treatment.