Thiol Peroxidase Protects Salmonella enterica from Hydrogen Peroxide Stress In Vitro and Facilitates Intracellular Growth

Deletion of tpx leads to higher susceptibility to exogenous H₂O₂in vitro.

The contribution of Tpx to the oxidative stress response of S. Typhimurium was assessed in vitro using exogenous H₂O₂. Deletion of tpx (STM1682) was performed by the method of Datsenko and Wanner (9). Overnight cultures of the wild type and the tpx mutant were adjusted to an optical density at 600 nm (OD₆₀₀) of 0.2 and diluted 1:10 in LB medium, and 100 μl of the diluted suspension was challenged with 2 mM H₂O₂ at 37°C. Killing of S. Typhimurium was enumerated by plating for CFU on LB agar after the indicated time points (Fig. 1A). After 6 h, the difference in survival between the wild type and the mutant became significant (P < 0.005 using a two-sided t test). While the wild type was reduced to ≈2.1 × 10⁵ CFU/ml after 9 h, the tpx mutant was reduced to ≈6 × 10³ CFU/ml. Expression of wild-type tpx on the pBAD30 plasmid (11) in the tpx mutant restored the wild-type phenotype upon challenge with H₂O₂ (Fig. 1A). Taking into account that there are three catalases (KatE, KatG, and KatN) and two other peroxiredoxins (AhpC and TsaA) present in S. Typhimurium, the clear phenotype of the tpx mutant under exogenous H₂O₂ was surprising. Apparently the remaining scavenging capacity in the tpx mutant was not sufficient to protect S. Typhimurium from exogenous H₂O₂.

• Open in new tab
• Download powerpoint

FIG. 1.

(A) Susceptibility of S. Typhimurium to exogenous H₂O₂. Wild-type (closed squares), Δtpx (circles), and complemented (open squares) strains were exposed to 2 mM H₂O₂ in LB broth and plated for CFU at the indicated time points. The results are representative of three independent experiments. Data are presented as means ± standard deviations (SD) of duplicates. Symbols for the WT and complemented strains overlie each other. P values were <0.005 (two-sided t test). (B) Intracellular proliferation of S. Typhimurium in IFN-γ-activated RAW 264.7 cells. RAW macrophages were infected with the wild-type, Δtpx, and complemented strains. Treatment with 10 μM DPI is indicated (white bars). Intracellular proliferation was calculated as the ratio of intracellular proliferation at 16 h versus 2 h postinfection (p.i.). The inset represents the ratio of intracellular proliferation at 6 h versus 2 h postinfection. All results shown are the means ± standard errors of the means (SEM) for three independent experiments, each in triplicate. P values of <0.05 (two-sided t test) were considered to be significant. ns, not significant.

The tpx mutant is affected in proliferation within activated RAW macrophages.

To address the role of Tpx during intracellular replication in gamma interferon (IFN-γ)-activated RAW 264.7 macrophages (10 ng/ml), cells were infected with S. Typhimurium as described previously (6). Intracellular proliferation was calculated by dividing the bacterial load of the macrophages at 6 h or 16 h postinfection by the number of intracellular bacteria at 2 h after infection. In activated macrophages, intracellular proliferation of the tpx mutant was decreased significantly after 16 h (Fig. 1B) compared to that of the wild type (≈20.1-fold and ≈34.2-fold, respectively; P < 0.05). To test whether the attenuation of the tpx mutant in activated macrophages was due to an oxidative burst, RAW 264.7 cells were treated at 2 h postinfection with 10 μM diphenyleneiodonium (DPI), a known inhibitor of the phagocyte oxidase. In the presence of DPI, the tpx mutant showed the same proliferation rate as the WT and complemented strains (Fig. 1B). Finally, the intracellular proliferation was investigated at 6 h after infection, as an oxidative burst in macrophages starts very early after infection. The tpx mutant showed a slightly decreased proliferation compared to the WT (≈2.9-fold and ≈3.4-fold, respectively; P > 0.05) which could be restored in the complemented strain (≈3.6-fold; P < 0.05).

Hébrard et al. reported that the catalases of S. Typhimurium are dispensable during infection of IFN-γ stimulated RAW cells, whereas a mutant lacking ahpCF/tsaA shows a reduced proliferation, similar to the result described here for the tpx mutant (14). These results emphasize the role of peroxiredoxins in intracellular growth of Salmonella in activated macrophages.

The tpx mutant has a reduced capacity to degrade H₂O₂ in cell extracts.

To test how much Tpx, AhpC, and TsaA contribute to the detoxification of H₂O₂ in S. Typhimurium, the peroxidase activities of tpx, ahpCF, and tsaA mutants were compared in cell extracts using H₂O₂ as a substrate. Deletion of ahpCF (STM0608, STM0609) and tsaA (STM0402) was performed using the method of Datsenko and Wanner (9). Overnight cultures of a tpx, ahpCF, or tsaA mutant were washed in 20 mM Tris-1 mM EDTA (pH 7.4), and extracts were prepared by mechanical disruption of cells. The degradation of H₂O₂ was measured using a xylenol orange assay quantifying H₂O₂ based on the oxidation of Fe²⁺ to Fe³⁺ in the presence of xylenol orange (18). A reaction mixture (100 μl) containing 1 mM H₂O₂ and 10 μg total protein of the cell extracts in 50 mM HEPES-1 mM EDTA (pH 7.4) was incubated for 15 min at room temperature. One milliliter of the xylenol assay reagent, consisting of 250 μM ammonium iron(II) sulfate, 125 μM xylenol orange, 25 mM H₂SO₄, and 100 mM sorbitol, was added and incubated for a further 15 min at room temperature. Absorbance was measured at 560 nm and compared to that of a known standard of H₂O₂. A concentration of 84.4 μM H₂O₂ was measured in extracts from the tpx mutant, compared to 18 μM, 38.9 μM, and 20.9 μM in extracts from the wild type and the tsaA and ahpCF mutants, respectively (Fig. 2), indicating that the tpx mutant possesses less reductive capacity toward H₂O₂ than the wild type, the ahpCF mutant, and the tsaA mutant.

• Open in new tab
• Download powerpoint

FIG. 2.

Biochemical analysis of peroxidase activities in S. Typhimurium WT, ΔahpC, ΔtsaA, Δtpx, and Δtpx::tpx complemented strains with H₂O₂ as a substrate. Bacterial cells were mechanically disrupted, and the capacity to reduce H₂O₂ in the cell extracts was determined using the xylenol orange assay. The color change was quantified by measuring the optical density at 560 nm and compared to a known H₂O₂ standard. Data shown are the means ± SEM for three independent experiments, each in triplicate. P values were <0.0001 using a two-sided t test.

The tpx mutant is not attenuated in mice.

The fitness of the mutant strain compared to the wild type was tested in a competition experiment in mice. Groups of five female BALB/cJ mice at the age of 6 to 8 weeks were infected intraperitoneally with 1 × 10⁵ cells consisting of the wild-type and mutant strains at a 1:1 ratio. Mice were sacrificed, and the livers and spleens were removed and homogenized in phosphate-buffered saline (PBS) on day 1 and day 3 after infection. Bacteria were plated on LB agar and on LB agar containing chloramphenicol. Competitive indices (CIs) were calculated as described previously (2). The tpx mutant was not attenuated in the livers and spleens of infected mice compared to the wild type (CIs were 0.93 ± 0.13 and 1.27 ± 0.06 for the mutant versus the WT in livers and spleens, respectively).

In this study we identified a third peroxiredoxin, Tpx, in S. Typhimurium by testing three aspects of the oxidative defense system of Salmonella: in vitro H₂O₂ challenge, growth inside activated phagocytes, and virulence in mice. The attenuated phenotype of the tpx mutant in response to exogenous H₂O₂ and in activated macrophages seems to be surprising considering the redundant capacities to counteract oxidative stress in Salmonella. However, similar observations have been reported for other pathogens, such as Escherichia coli (4), Mycobacterium tuberculosis (15), Helicobacter pylori (8), and Enterococcus faecalis (19). A single deletion in tpx rendered all four pathogens more susceptible to exogenous H₂O_2. In vitro challenge with H₂O₂ in this study was done in whole-cell assays and in cell extracts. In both assays the tpx mutant was significantly affected either in survival or in its ability to degrade H₂O_2. We directly compared tpx, ahpCF, and tsaA mutants in cell extracts challenged with 1 mM H₂O₂. This comparison suggested that Tpx might be superior to AhpC and TsaA in its ability to protect against H₂O₂ even at millimolar concentrations. Biochemical studies indicate that Tpx has broad substrate specificity. E. coli Tpx has been shown to reduce H₂O₂ and organic peroxides (1, 5, 7). M. tuberculosis Tpx has a similar substrate range and, in addition, is known to reduce peroxynitrite (16, 23). Studies with purified Tpx and AhpC from E. coli revealed that Tpx is the most potent peroxidase. Inactivation of tpx in E. coli resulted in increased sensitivity to H₂O₂ and various organic hydroperoxides. The level of sensitivity was always higher than that of an ahpC mutant (3, 4). In E. faecalis the comparison of mutants lacking either an NADH peroxidase (Npr), an alkyl hydroperoxide reductase (AhpC), or a thiol peroxidase (Tpx) showed that Tpx mediates the most effective antioxidant activity for survival inside macrophages (19). In this study we did not compare growth of the ahpCF, tsaA, and tpx single mutants in macrophages. However, the presently available results suggest that all three peroxiredoxins play an essential role during growth in activated macrophages. In M. tuberculosis and in E. faecalis, a single tpx mutant was attenuated in mice. We could not detect reduced virulence of a single tpx mutant of Salmonella in mice. Hébrard et al. found that neither a double mutant with deletions in ahpCF and tsaA nor a triple mutant with deletions in katG, katE, and katN was attenuated in mice. Only a mutant with deletions in all five genes was highly attenuated in vivo (14). Thus, Salmonella has evolved multiple routes to survive oxidative stress in vivo. In summary, this study shows that Salmonella Tpx is efficient in scavenging H₂O₂ in vitro and facilitates intracellular growth in activated macrophages.