Low-molecular-weight thiol-dependent antioxidant and antinitrosative defences in Salmonella pathogenesis

Abstract

We found herein that the intracytoplasmic pool of the low-molecular-weight (LMW) thiol glutathione (GSH) is readily oxidized in Salmonella exposed to nitric oxide (NO). The hypersusceptibility of gshA and gshB mutants lacking γ-glutamylcysteine and glutathione synthetases to NO and S-nitrosoglutathione indicates that GSH antagonizes the bacteriostatic activity of reactive nitrogen species. Metabolites of the GSH biosynthetic pathway do not affect the enzymatic activity of classical NO targets such as quinol oxidases. In contrast, LMW thiols diminish the nitrosative stress experienced by enzymes, such as glutamine oxoglutarate amidotransferase, that contain redox active cysteines. LMW thiols also preserve the transcription of Salmonella pathogenicity island 2 gene targets from the inhibitory activity of nitrogen oxides. These findings are consistent with the idea that GSH scavenges reactive nitrogen species (RNS) other than NO. Compared with the adaptive response afforded by inducible systems such as the hmp-encoded flavohaemoprotein, gshA, encoding the first step of GSH biosynthesis, is constitutively expressed in Salmonella. An acute model of salmonellosis has revealed that the antioxidant and antinitrosative properties associated with the GSH biosynthetic pathway represent a first line of Salmonella resistance against reactive oxygen and nitrogen species engendered in the context of a functional NRAMP1(R) divalent metal transporter.

Figure 1. Consumption of glutathione in NO-treated Salmonella

The concentration of glutathione (GSH) was quantified in stationary phase, wild-type (WT) Salmonella 15 min after exposure to 500 μM spermine NONOate (sNO) in LB broth. The GSH/GSSG (reduced/oxidized) ratio in untreated and sNO-treated Salmonella is shown in B. The data are the mean +/− SD of 2–5 independent observations.

Figure 2. Susceptibility of gshA-deficient Salmonella to reactive oxygen and nitrogen species

The concentration of glutathione (GSH) in wild-type (WT) and gshA-or gshB-deficient Salmonella grown in EG medium, pH 7.0, is shown in A. Survival of 10⁵ CFU/well of WT, ΔgshA, ΔgshB and Δgor mutant Salmonella 2 h after exposure to increasing concentrations of H₂O₂ (B). Panel C shows the effect of the pGSHA complementing plasmid on the susceptibility of ΔgshA mutant Salmonella to 200 μM H₂O₂. Growth of Salmonella in LB broth containing 5 mM DETA or 5 mM of the DETA NONOate NO donor (DETANO) (D). The anti-Salmonella activity of GSNO was estimated by measuring the zone of inhibition of bacteria grown on M9 glucose agar plates (E). Data in panel D represent the mean +/− SEM of 10 independent observations obtained from 3 separate experiments. The data in panel E gathered from 7 independent observations in 2 separate experiments are represented in box-and-whiskers plots as median, intraquartile and total ranges. p < 0.01; ***, p < 0.001 compared to WT controls.

Figure 3. Contribution of gshA and hmp to the antinitrosative defenses of Salmonella

Effects of 5 mM DETA (A) or 5 mM of the NO donor DETA NONOate (DETANO) (B) on the growth of wild-type (WT) Salmonella, and gshA- or hmp-deficient controls grown in LB broth. Data represent the mean +/− SD of 10–15 independent observations obtained from 3–4 separate experiments.

Figure 4. Expression of gshA and hmp in NO-treated Salmonella

The expression of gshA::lacZY (A) and hmp::lacZY (B) was recorded in wild-type (WT) Salmonella, or gshA and hmp mutant controls grown in 8 μM Mg²⁺ N salts medium 2 h after exposure to increasing concentrations of the NO donor spermine NONOate. Data represent the mean +/− SEM of 4 independent observations obtained from 2 separate experiments.

Figure 5. Respiratory activity of Salmonella undergoing nitrosative stress

Respiratory activity of gshA, hmp or gshA hmp mutant Salmonella after the addition of 5 μM of the NO donor proli NONOate (pNO) to bacterial cultures containing about 130 μM O₂. Wild-type (WT) Salmonella were used as controls. The data are representative of 3 independent observations obtained on 3 separate days.

Figure 6. Transcription of SPI2 genes in NO-treated Salmonella

Expression of spiC::lacZY in gshA (A), gshB (B) or hmp (C) mutant Salmonella grown in 8 μM Mg²⁺ N salts medium in the presence of increasing concentrations of the NO donor spermine NONOate. The wild-type strain AV0207 (WT) and the ghsA strain AV09315 expressing the pGSHA complementing plasmid were used as controls. Data represent the mean of 4 independent observations obtained from 2 separate experiments. Panels D and E show the expression of spiC::lacZY and sifA::lacZY transcriptional fusions in WT or gshA-deficient Salmonella at the indicated times after exposure to 250 μM spermine NONOate. The data on D and E are representative 3 independent experiments.

Figure 7. Virulence of gshA-deficient Salmonella in an acute model of infection

NRAMP1^S C57BL/6 (A) and NRAMP1^R C3H/HeNCrl (B) mice were challenged i.p. with about 100 and 500 CFU/mouse, respectively, of the indicated Salmonella strains. To inhibit the function of the NADPH oxidase, 100 μg/ml of acetovanillone were added drinking water (C). The water of the mice in panel D was treated with 500 μg/ml of the iNOS specific inhibitor L-NIL. The survival of Salmonella-infected mice was scored over time. The data are from five mice per group.

Figure 8. Glutamine oxoglutarate amidotransferase enzymatic activity in gshA-deficient Salmonella undergoing nitrosative stress

GOGAT enzymatic activity was measured by following the consumption of NADPH at A₃₄₀ (A). A ΔgltBD mutant strain lacking GOGAT was used as a control. The specific GOGAT enzymatic activity shown in parenthesis is expressed as μmoles of NADPH/min/mg of protein. % GOGAT activity remaining in wild-type (WT) and gshA-deficient Salmonella 30 min after exposure to 100 μM spermine NONOate (sNO) (B). The bacteria were grown in media A containing 124 μM or 248 μM FeCl₃. The data represent the mean +/− SEM of 8 independent observations obtained from 2 separate experiments. ***, p < 0.001 compared to sNO-treated wild-type controls grown in 124 μM FeCl₃ media A.