Low molecular weight thiol-dependent antioxidant and antinitrosative defenses 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 to the adaptive response afforded by inducible systems such as the hmp-encoded flavohemoprotein, 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.

Introduction

Many of the 2,500 serovars of Salmonella enterica are associated with self-limiting or severe gastroenteritis in domestic animals and humans, and some of them can cause life-threatening systemic disease. Salmonella infections are normally acquired through the ingestion of tainted food or water. Shortly after reaching the gastrointestinal lumen, Salmonella is taken up by mononuclear phagocytic cells (Vazquez-Torres et al., 1999), where it is exposed to the antimicrobial activity of a battery of innate host effectors (Fang, 2004). Reactive oxygen and nitrogen species generated by the NADPH phagocyte oxidase and inducible NO synthase (iNOS) are arguably the best characterized effectors of the anti-Salmonella arsenal of macrophages (Fang, 2004; Mastroeni et al., 2000; Vazquez-Torres et al., 2000a; Vazquez-Torres and Fang, 2001). These reactive species damage metal cofactors, thiol groups and DNA molecules (Henard and Vazquez-Torres, 2012b), thereby posing a significant demand on bacterial physiology. In turn, the coordinated actions of Salmonella-specific virulence factors and defense systems shared with many saprophytic, commensal and pathogenic microorganisms counteract the metabolic and genomic challenges imposed by reactive oxygen and nitrogen species. Salmonella-specific virulence factors are frequently encoded within the virulence plasmid or are clustered in pathogenicity islands in the chromosome. For instance, the Salmonella pathogenicity island 2 (SPI2) type III secretion system increases the intracellular fitness of Salmonella in macrophages and epithelial cells (Ochman et al., 1996; Ruiz-Albert et al., 2002). The SPI2 type III secretion system protects Salmonella against lysosomal enzymes of the degradative pathway, and lessens the exposure of this intracellular bacterium to the reactive oxygen and nitrogen species generated by NADPH phagocyte oxidase and iNOS enzymatic complexes (Berger et al., 2010; Chakravortty et al., 2002; Gallois et al., 2001; Suvarnapunya and Stein, 2005; Vazquez-Torres et al., 2000b). Salmonella also uses base excision DNA repair systems, iron storage proteins, and the LMW thiol homocysteine to detoxify reactive species or repair the resulting oxidative and nitrosative insults (De Groote et al., 1996; Richardson et al., 2009; Velayudhan et al., 2007). In addition, the Salmonella genome encodes several superoxide dismutases, catalases and peroxidases that detoxify a variety of inorganic and organic reactive oxygen species (Fang et al., 1999; Fang, 2004). Moreover, Salmonella detoxifies NO to nitrate (NO₃⁻), nitrous oxide (N₂O), or ammonia (NH₃) through the enzymatic activity of the flavohemoglobin Hmp, the flavorubredoxin NorV, or the respiratory nitrite reductase NrfA, respectively (Bang et al., 2006; Crawford and Goldberg, 1998; Gilberthorpe and Poole, 2008; Mills et al., 2005; Mills et al., 2008).

LMW thiols help maintain redox homeostasis and protect a variety of phylogenetically diverse organisms against the cytotoxicity of reactive oxygen and nitrogen species, alkylating agents, antibiotics and toxins (Fahey, 2012; Masip et al., 2006; Meister and Anderson, 1983). Some prokaryotes such as members of the Firmicutes or Actinobacteria use the LMW thiols mycothiol, or bacillithiol instead of GSH. However, as is the case for most eukaryotes and Gram-negative bacteria (Fahey, 2012; Masip et al., 2006; Meister and Anderson, 1983), GSH is the most abundant LMW thiol in Salmonella, raising the possibility that this tripeptide (i.e., γ-Glu-Cys-Gly) may be an important aspect of Salmonella pathogenesis. GSH is synthesized in two sequential enzymatic steps. The gshA-encoded γ-glutamylcysteine synthetase ligates the amino group of cysteine to the γ-carboxyl group of glutamate. In turn, the gshB-encoded glutathione synthetase condenses the resulting γ-glutamylcysteine with glycine to generate GSH. Between 90 and 99% of the GSH pool is reduced in resting cells (Henard et al., 2010). H₂O₂ and RNS generated from the reaction of NO with O₂ or superoxide anion (O₂^·−) oxidize GSH to its GSSG couple, S-nitrosoglutathione (GSNO), mixed disulfides or high order oxidation species (Keshive et al., 1996; Koppenol et al., 1992; Winterbourn and Metodiewa, 1999). GSSG can be recycled back to GSH by the gor-encoded glutathione reductase, in a reaction that draws reducing power from NADPH (Ritz and Beckwith, 2001). Both the concentration of GSH and GSSG, and the ratio of this redox couple affect the antioxidant potential of a cell (Henard et al., 2010; Schafer and Buettner, 2001). Despite being the most abundant LMW thiol in Salmonella, a role for GSH in Salmonella pathogenesis remains to be demonstrated. To examine the contribution of the GSH biosynthetic pathway to the antioxidant and antinitrosative arsenal of Salmonella, we tested the virulence of gshA, gshB and gor mutants in a variety of in vitro and in vivo experimental conditions that recapitulate the oxidative and nitrosative stress encountered by Salmonella in the course of an infection.

Results

Oxidation of GSH in Salmonella treated with NO

H₂O₂ depletes the pool of GSH in Salmonella (Henard et al., 2010). We tested herein whether exposure of Salmonella to NO also affects the intracytoplasmic concentrations of this LMW thiol. Treatment of wild-type Salmonella with 500 μM spermine NONOate reduced by 3-fold the intracytoplasmic concentration of GSH (figure 1A) without affecting the viability of the bacteria. Exposure of Salmonella to 500 μM spermine NONOate resulted in a ~10-fold decrease in the GSH/GSSG ratio (figure 1B). It should be noted that control and NO-treated Salmonella harbored approximately 25 and 250 μM GSSG, respectively. Because NO-treated Salmonella lost about 2.5 mM of its intracellular pool of GSH, our data indicate that most of the GSH consumed in NO-treated Salmonella is oxidized to species other than GSSG.

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.

Contribution of GSH to the antioxidant and antinitrosative defenses of Salmonella

GSH is a key component of the antioxidant arsenal of organisms as evolutionarily diverse as bacteria and humans (Meister and Anderson, 1983). Salmonella exposed to NO or H₂O₂ suffer a dramatic reduction in the intracellular GSH pool (studies herein and (Henard et al., 2010), raising the possibility that this LMW thiol contributes to the antioxidant and antinitrosative defenses of Salmonella. To test this hypothesis, we measured the cytotoxicity of various reactive oxygen and nitrogen species against wild-type Salmonella as compared to gshA- and gshB-deficient controls. In addition, we also tested the susceptibility of a gor mutant that lacks the glutathione reductase enzymatic activity needed for recycling GSSG to GSH. As reported in E. coli (Bouter et al., 1988; Greenberg and Demple, 1986), Salmonella strains lacking the gshA-encoded γ-glutamylcysteine synthetase or the gshB-encoded glutathione synthetase had no GSH (figure 2A). Stationary phase, wild-type S. Typhimurium or isogenic mutants lacking gshA, gshB or gor were challenged with increasing concentrations of H₂O₂ (figure 2B). All Salmonella strains tested were killed in a dose-dependent manner 2 h after exposure to H₂O₂. The gshB or gor mutants were as susceptible to H₂O₂ as wild-type controls, an observation that is consistent with the phenotype of gshB-deficient E. coli (Imlay and Linn, 1987). In contrast, the gshA mutant was hypersusceptible (p < 0.05) to the bactericidal activity of 200 or 400 μM H₂O₂ (figures 2B and 2C). For instance, a 3 h exposure to 200 μM H₂O₂ killed about 90 and 99.5 % of wild-type and gshA-deficient Salmonella, respectively. The hypersusceptibility of the gshA-deficient Salmonella to H₂O₂ appears to be specific to a lack of γ-glutamylcysteine synthetase, because the pGSHA plasmid expressing a gshA allele complemented the survival defect of gshA-deficient Salmonella (figure 2C).

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.

As shown in figure 1 and elsewhere (Singh et al., 1996), GSH is a target of nitrosative stress. Consequently, we tested the sensitivity of the aforementioned Salmonella strains to the bacteriostatic activity of the NO donor DETA NONOate. Compared to bacteria exposed to the DETA polyamine control, 5 μM NO generated by DETA NONOate exerted bacteriostatic activity against all the strains tested (figure 2D). Wild-type and gor mutant bacteria initiated growth 5 h after challenge, whereas the gshA or gshB mutants did not resume growth until 12 or 7 h, respectively. As reported above for H₂O₂, the pGSHA plasmid complemented the growth defect noted in NO-treated, gshA-deficient Salmonella (not shown). Mutants lacking gshA or gshB were equally susceptible to GSNO (figure 2E), a RNS that can generate NO through homolytic cleavage. In addition, GSNO can heterolytically transfer nitrosonium (NO⁺) to nucleophiles such as thiolates (-S⁻) in the side chains of redox active cysteines via transnitrosation reactions (De Groote et al., 1995). Collectively, our investigations indicate that LMW thiols of the GSH biosynthetic pathway contribute to the antioxidant and antinitrosative defenses of Salmonella.

Contribution GSH and the Hmp flavohemoprotein to the antinitrosative defenses of Salmonella

The NO-detoxifying activity of Hmp is thought to be the most important antinitrosative defense of Salmonella (Bang et al., 2006). We therefore deemed it important to compare the relative contribution of gshA and hmp to the resistance of this enteropathogen to NO-dependent bacteriostasis. Wild-type, gshA, hmp and gshA hmp mutants grew with similar kinetics in LB broth supplemented with 5 mM DETA (figure 3A). Salmonella lacking hmp was highly susceptible to 5 mM DETA NONOate, and did not even initiate growth 30 h after treatment (figure 3B). As reported above, the gshA mutant was also hypersusceptible to the cytotoxicity of DETA NONOate. Compared to hmp-deficient Salmonella, the lag phase of the gshA mutant was less dramatic. A double mutant lacking both gshA and hmp was as susceptible as the hmp deficient control. The considerable pressures imposed by nitrosative stress to the gshA hmp double mutant do not appear to select for suppressors with high growth rates and increased resistance to nitrosative stress. The gshA hmp double mutant recovered after NO treatment remained highly susceptible to a subsequent exposure to DETA NONOate (not shown). Cumulatively, our investigations indicate that both Hmp and GSH contribute to the antinitrosative defenses of Salmonella, although the hmp-encoded flavohemoprotein appears to be more important than the latter.

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.

Transcription of gshA and hmp in NO-treated Salmonella

Since both Hmp and GSH provide protection against NO, we compared the activity of the gshA::lacZY and hmp::lacZY transcriptional fusions in NO-treated Salmonella. The basal level of gshA transcription was slightly decreased in wild-type or hmp-deficient Salmonella exposed to increasing concentrations of spermine NONOate (figure 4A). These findings indicate that gshA is constitutively expressed. As expected, NO derepressed hmp transcription (figure 4B). Compared to untreated controls, Salmonella challenged with 50 μM spermine NONOate experienced a 15-fold increase in hmp::lacZY transcription, the expression of which went up 30-fold in the absence of gshA (figure 4B). Our investigations suggest that LMW thiols temper the adaptive expression of the hmp-encoded flavohemoprotein in Salmonella experiencing moderate levels of nitrosative stress. Irrespective of the gshA status of the bacteria, hmp transcription was gradually repressed as the levels of nitrosative stress increased. NsrR inhibits hmp transcription (Bodenmiller and Spiro, 2006), and thus the nitrosylation of the [2Fe-2S] cluster of NsrR derepresses hmp expression (Tucker et al., 2008). The accumulation of NADH that follows the repression of respiratory activity by high concentrations of NO promotes reduction of ferric to ferrous iron (Woodmansee and Imlay, 2003). Binding of ferrous iron to NsrR may consequently repress hmp transcription at high NO concentrations that block respiration and increase the concentration of NADH in the cell.

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.

Effects of gshA and hmp on the respiratory activity of Salmonella undergoing nitrosative stress

NO directly binds to heme and Cu⁺⁺ prosthetic groups of terminal quinol oxidases of the electron transport chain (Butler et al., 1997; Husain et al., 2008). The NO-detoxifying activity of Hmp protects the electron transfer chain from the inhibitory pressures that this diatomic radical exerts on quinol oxidases (Stevanin et al., 2000). Because gshA is a critical aspect of the antinitrosative defenses of Salmonella, we tested whether the expression of gshA shields respiration from the inhibitory activity of NO (figure 5). Wild-type and gshA mutant Salmonella were exposed to 5 μM proli NONOate at a time when the concentration of dissolved O₂ in the media reached about 130 μM. NO inhibited the respiratory activity of wild-type and gshA-deficient bacteria for about 250 sec. In contrast, NO blocked the respiration of hmp and gshA hmp mutant bacteria for about 335 and 350 sec, respectively. These findings indicate that the denitrosylase activity of the flavohemoprotein, but not LMW thiols, protects terminal quinol oxidases from the cytotoxicity exerted by NO itself. Our investigations also suggest that LMW thiols and the Hmp flavohemoprotein detoxify different RNS.

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.

LMW thiols protect SPI2 gene expression from the inhibitory effects of NO congeners

The SPI2 type III secretion system protects Salmonella from NO produced during the innate response (Chakravortty et al., 2002); however, high NO fluxes generated by IFNγ-primed macrophages inhibit SPI2 gene transcription (McCollister et al., 2005). Because LMW thiols of the GSH biosynthetic pathway antagonize the anti-Salmonella activity of NO (see above), we examined whether the presence of gshA preserves SPI2 transcription in NO-treated Salmonella. To this end, the transcription of the SPI2 gene spiC was followed in wild-type and gshA-deficient Salmonella treated with increasing concentrations of spermine NONOate. At low NO fluxes, spiC expression was unaffected in either wild-type or gshA mutant Salmonella; however, the expression of this transcriptional fusion was gradually repressed as the concentration of NO increased in the medium (figure 6A). When compared to wild-type bacteria, gshA-deficient Salmonella expressed 4-fold less spiC::lacZY 2 h after treatment with 250 μM spermine NONOate (figure 6A). The pGSHA complementing plasmid fully restored the expression of spiC::lacZY in gshA-deficient Salmonella. The differences in spiC::lacZY expression between wild-type and gshA mutant controls cannot be explained by differences in growth because the effects of NO on the expression of this SPI2 gene were tested in bacterial cells that had already reached early stationary phase. NO-treated, wild-type and gshB mutant Salmonella expressed similar levels of spiC::lacZY (figure 6B). Collectively, these findings suggest that LMW thiols protect the transcription of SPI2 genes from the inhibitory effects of NO.

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.

The Hmp flavohemoprotein plays a central role in the antinitrosative defenses of Salmonella (Bang et al., 2006), and the presence of a functional Hmp protects the transcription of SPI2 genes in Salmonella undergoing moderate nitrosative stress (McCollister et al., 2007). We therefore measured the relative expression of spiC::lacZY in hmp-deficient controls under the experimental conditions used in our investigations. The absence of hmp resulted in a marked inhibition of spiC::lacZY in NO-treated Salmonella (figure 6C). The addition of 50 μM spermine NONOate partially repressed spiC transcription in the hmp mutant. Moreover, 100 μM spermine NONOate completely repressed spiC::lacZY transcription in hmp-deficient bacteria. In contrast, wild-type bacteria fully expressed spiC::lacZY at these moderate levels of nitrosative stress. Thus, both Hmp and LMW thiols shield SPI2 transcription from the inhibitory effects of NO, although the former appears to be more efficient than the latter.

The constitutive expression of gshA suggests that LMW thiols may protect molecular targets shortly after Salmonella encounters the reactive species. Consequently, we investigated the effects of gshA gene expression on the temporal transcription of the spiC and sifA SPI2 genes, which encode translocon and effector functions (Ruiz-Albert et al., 2002; Uchiya et al., 1999; Yu et al., 2002). The expression of spiC::lacZY and sifA::lacZY transcriptional fusions was comparable in wild-type and gshA-deficient Salmonella grown in 8 μM MgCl₂ N9 medium (not shown). The expression of spiC::lacZY could be detected in wild-type bacteria 2 h after treatment with 250 μM spermine NONOate. In contrast, the gshA mutant did not initiate spiC transcription until about 3 h into NO treatment, when the expression reached about a third of wild-type controls (figure 6D). Although with slightly different kinetics, the transcription of the SPI2 effector sifA was also more susceptible to the inhibitory effects of NO in the absence of gshA (figure 6E).

GSH increases Salmonella fitness in a murine model of systemic infection

As previously noted in BALB/c mice (Bjur et al., 2006), the gshA mutant was found to be slightly attenuated when inoculated intraperitoneally into C57BL/6 mice (figure 7A). Given that LMW thiols protect SPI2 expression from the inhibitory activity exerted by nitrosative stress, while lessening the bactericidal and bacteriostatic activities of H₂O₂ and NO, we found it surprising that gshA-deficient Salmonella were as virulent as wild-type controls. BALB/c and C57BL/6 mice express a mutant allele of the natural resistance-associated macrophage protein 1 (NRAMP1, also known as Slc11a1) that has been linked to low NO synthesis and a reduced anti-Salmonella activity of macrophages (Ables et al., 2001; Fritsche et al., 2003). We reasoned that, as described for hmp, metL, ssrB cys²⁰³ variants and sitA (Bang et al., 2006; De Groote et al., 1996; Husain et al., 2010; Zaharik et al., 2004), the superior anti-Salmonella activity of wild-type mice expressing the wild-type NRAMP1^R allele could bring to light a possible contribution of gshA to Salmonella pathogenesis. To test this hypothesis, NRAMP1^R C3H/HeNCrl mice were challenged intraperitoneally with wild-type Salmonella or isogenic gshA or gshB mutants. Salmonella lacking gshA, and to a lesser extent a gshB mutant, were attenuated in this acute model of salmonellosis (figure 7B). These findings indicate that, as it has recently been demonstrated in pneumococci (Potter et al., 2012), GSH is important for Salmonella virulence. Given its contribution to the resistance of Salmonella to H₂O₂ and NO, we tested whether gshA antagonizes the antimicrobial activity associated with host NADPH oxidase and iNOS hemoproteins. Salmonella lacking gshA gained virulence in C3H/HeNCrl mice treated with the NADPH phagocyte oxidase inhibitor acetovanillone (figure 7C). In contrast, acetovanillone did not increase the virulence of wild-type Salmonella (P = 0.61; Mantel-Cox survival test). Because gshA-deficient Salmonella are hypersusceptible to the bacteriostatic effects of NO (figures 2 and 3), we also tested the virulence of the gshA mutant in C3H/HeNCrl mice treated with the iNOS inhibitor N6-(1-iminoethyl)-L-lysine, dihydrochloride (L-NIL). The addition of L-NIL to drinking water enhanced the virulence of gshA-deficient Salmonella, but did not affect the virulence of wild-type bacteria (figure 7D). Cumulatively, these findings indicate that LMW thiols protect Salmonella against the innate antimicrobial activity of reactive oxygen and nitrogen species generated by host NADPH phagocyte oxidase and iNOS hemoproteins.

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.

LMW thiols and iron decrease the nitrosative stress endured by glutamine oxoglutarate amidotransferase

By pumping cations from the phagosomal lumen into the cytosol, the NRAMP1 cationic transporter limits the access of intracellular Salmonella to iron (Nairz et al., 2009). Because a gshA mutant strain is attenuated in NRAMP1^R mice, we explored a possible relation between iron, LMW thiols and nitric oxide. To this end, we followed the enzymatic activity of glutamine oxoglutarate amidotransferase (GOGAT). This enzyme was chosen because the cysteines coordinating the [4Fe-4S] clusters of GOGAT are targets of nitrosative stress (Brandes et al., 2007). Wild-type Salmonella grown to OD₆₀₀ of 0.8 in media A containing 124 μM FeCl₃ harbored a specific GOGAT enzymatic activity of 0.2 μmoles/min (figure 8A). In comparison, log phase, gshA mutant Salmonella only contained 0.120 μmoles/min of specific GOGAT enzymatic activity, perhaps reflecting the importance of GSH in maintaining thiols coordinating the [4Fe-4S] clusters in a reduced state (Tran et al., 2000). The addition of 100 μM spermine NONOate for 30 min inhibited 40% and 70% of GOGAT enzymatic activity in wild-type and gshA-deficient Salmonella (figure 8B). Supplementation of media A with 248 μM FeCl₃ shielded most GOGAT enzymatic activity from NO-mediated toxicity, regardless of the expression of gshA. These findings indicate that both the LMW thiol content of the bacteria and the availability of extracellular iron influence the ability of RNS to damage biological targets.

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.

Discussion

Salmonella uses a type III secretion system, storage proteins, scavengers, and enzymes to prevent, detoxify and repair the damage incurred by reactive species generated by NADPH phagocyte oxidase and iNOS hemoproteins (Fang, 2004). The investigations presented herein add LMW thiols of the GSH pathway to the antioxidant and antinitrosative defenses of Salmonella. The contribution of GSH to Salmonella pathogenesis was exposed in NRAMP1^R mice. It is possible that LMW thiols assume a key role in Salmonella pathogenesis in the context of the increased oxidative and nitrosative stress engendered by a functional NRAMP1 transporter (Ables et al., 2001; Barton et al., 1995; Fritsche et al., 2003). Furthermore, the increased availability of iron in NRAMP1^S macrophages (Nairz et al., 2009) may decrease the NO-mediated anti-Salmonella activity by two independent, but mutually nonexclusive mechanisms. First, iron in the phagosome could scavenge NO, thereby reducing the nitrosative stress endured by Salmonella. This model could explain why RNS were not as efficient at damaging GOGAT enzymatic activity in Salmonella cultures grown with increased iron concentrations. Second, the availability of iron in NRAMP1^S macrophages could help intracellular Salmonella repair oxidized [4Fe-4S] clusters.

The role played by LMW thiols of the GSH biosynthetic pathway in an acute model of Salmonella infection appears to rely on the direct protection of essential biomolecules and the indirect potentiation of Salmonella-specific virulence factors. DNA molecules, metal cofactors and redox active cysteines in enzymes of central metabolism are the preferred targets of H₂O₂ and NO congeners (Henard and Vazquez-Torres, 2012b). Therefore, LMW thiols are likely to diminish the genotoxicity and metabolic stress that Salmonella undergoes upon exposure to reactive oxygen and nitrogen species. In support of this model, GOGAT enzymatic activity was more vulnerable to the inhibitory effects of RNS in gshA-deficient Salmonella. However, the role of LMW thiols in Salmonella pathogenesis might not be limited to protecting enzymes of central metabolism from the noxious effects of oxidative and nitrosative stress. Because the SPI2 type III secretion system minimizes the contact of Salmonella phagosomes with vesicles harboring NADPH oxidase or iNOS enzymatic complexes (Berger et al., 2010; Chakravortty et al., 2002; Gallois et al., 2001; Suvarnapunya and Stein, 2005; Vazquez-Torres et al., 2000b), the capacity of LMW thiols to preserve SPI2 transcription in Salmonella enduring moderate levels of nitrosative stress could indirectly boost the antioxidant and antinitrosative arsenal of this intracellular pathogen. Our previous work showed that nitrosative stress represses SsrB- and PhoP-dependent SPI2 transcription in Salmonella (Bourret et al., 2009; McCollister et al., 2005). Accordingly, LMW thiols could protect SsrB and PhoP signaling cascades from the inhibitory effects of RNS. Protection of DNA and central metabolic pathways, together with the preservation of transcription of Salmonella specific virulence factors such as the SPI2 type III secretion system, provide a rational framework for the role played by LMW thiols in the antioxidant and antinitrosative defenses that contribute to Salmonella pathogenesis.

Our genetic analysis indicates that gshA and gshB mediate the resistance of Salmonella to various RNS. In particular, the susceptibility of gshB-deficient Salmonella to the nitrosating agent GSNO suggests that GSH is a central piece of the antinitrosative arsenal of this facultative pathogen. The considerable drop in the concentration of GSH in NO-treated Salmonella provides further support for this notion. It is then surprising that, despite the marked oxidation of GSH in wild-type cells undergoing oxidative stress (Henard et al., 2010), gshB mutant E. coli and Salmonella, which are deficient in GSH, show similar resistance to H₂O₂ (studies herein and Imlay and Linn, 1987). It is possible that the γ-glutamylcysteine accumulated in gshB mutant cells (Cameron and Pakrasi, 2010; Helbig et al., 2008) assumes the antioxidant functions normally played by GSH in wild-type bacteria. This model could explain why gshA-deficient Salmonella are susceptible to H₂O₂ (studies herein and Bjur et al., 2006). Nonetheless, gshA-deficient E. coli are resistant to H₂O₂ (Greenberg and Demple, 1986). It is not likely that γ-glutamylcysteine per se plays different roles in E. coli and Salmonella. However, the relative concentrations of γ-glutamylcysteine could explain the seemingly contradictory contributions of gshA to the antioxidant defenses of in E. coli and Salmonella. The intracellular pool of γ-glutamylcysteine is quite small in E. coli (Helbig et al., 2008); the concentration of γ-glutamylcysteine in Salmonella could be different. Compared to gshB-deficient controls, gshA mutant Salmonella were hypersusceptible to NO and suffered a greater degree of NO-dependent inhibition of SPI2 transcription. These findings indicate that γ-glutamylcysteine may have a unique role as a scavenger of reactive species. Independent of whether or not γ-glutamylcysteine is a relevant scavenger of reactive species, the enzymatic activity of γ-glutamylcysteine synthetase could ameliorate oxidative stress by diminishing the intracytoplasmic cysteine pool, thereby limiting the cysteine-mediated reduction of ferric to ferrous iron (Park and Imlay, 2003). Regardless of the mechanism, this is not the first occasion that closely related members of the Enterobacteriaceae have been shown to deal with oxidative stress in different ways. A case in point, respiratory arrest promotes oxidative stress in E. coli, whereas it boosts the antioxidant potential of Salmonella (Husain et al., 2008; Woodmansee and Imlay, 2003). Our investigations indicate that if GSH biosynthesis inhibition were to be considered as a potential target for new chemotherapies, then GshA rather than GshB would perhaps be a more suitable target.

Our transcriptional analysis identifies LMW thiols as a substantial component of the innate antioxidant and antinitrosative arsenal of Salmonella. The defense provided by LMW thiols takes place before adaptive mechanisms such as Hmp are fully engaged (Stevanin et al., 2000). In analogy to the immune system, the innate antinitrosative defenses afforded by LMW thiols are not only less specific but also less potent than adaptive mechanisms. For instance, GSH scavenges a variety of reactive oxygen and nitrogen species, whereas the inducible Hmp flavohemoprotein exclusively detoxifies NO (Hausladen et al., 2001). In spite of its remarkable specificity, the detoxification of NO by the denitrosylase enzymatic activity of Hmp limits the production of all RNS, thereby not only safeguarding direct NO targets such as quinol oxidases but also protecting a cadre of biomolecules with redox active thiolates or metal centers from nitroxidative stress. On the other hand, GSH preferentially detoxifies oxidative products of this diatomic radical (Keszler et al., 2010). Consequently, in comparison with the NO-detoxifying activity of Hmp, GSH is expected to protect a more limited set of biomolecules. Following this line of reasoning, LMW thiols are quantitatively less important to the antinitrosative defenses of Salmonella than Hmp. Having said that, the overexpression of hmp (studies herein), trxA and grxA (Miranda-Vizuete et al., 1996) in gshA-deficient bacteria may have understimated the real contribution of LMW thiols of the GSH biosynthetic pathway to the antioxidant and antinitrosative potential of wild-type Salmonella.

Reactive species modify biomolecules with remarkable specificity. For example, NO directly nitrosylates metal cofactors of quinol oxidases with a second-rate constant of 10⁻⁶ M⁻¹ sec⁻¹ (Butler et al., 1997), but modifies thiol groups quite inefficiently (Herold and Rock, 2005). Conversely, autoxidation products of NO and O₂ readily react with thiolates but fail to bind to metals in quinol oxidases of the electron transport chain (Herold and Rock, 2005). We noticed that NO more easily inhibited SPI2 gene transcription and GOGAT enzymatic activity in the absence of gshA. On the other hand, the absence of gshA did not affect the kinetics by which NO repressed the respiratory activity of Salmonella. These observations lend support to the notion discussed above that GSH does not scavenge NO itself, but rather NO congeners such as ONOO⁻, N₂O₃, and NO₂. LMW thiols could also undergo electrophilic attack by the NO⁺ cation released from dinitrosyl-iron complexes or S-nitrosothiols such as GSNO (Vazquez-Torres, 2012). Collectively, our investigations are consistent with the hypothesis that the anti-Salmonella activity emanating from iNOS is mediated by a rich chemistry involving NO, S-nitrosothiols, dinitrosyl-iron complexes, as well as nitrosating and nitroxidating species generated in the reaction of NO with O₂ and O₂^·−.

GSH is modified by reactive oxygen and nitrogen species to GSSG, mixed disulfides and high oxidation products such as sulfinic and sulfonic acids. Our biochemical studies indicate that about 5% of the GSH pool is oxidized to GSSG in NO-treated Salmonella. The small fraction of GSH that ends up as GSSG suggests that this is not a significant pathway by which LMW thiols add to the antinitrosative defenses of Salmonella. The dispensability of the gor-encoded glutathione reductase in the resistance of Salmonella to H₂O₂ and NO lends further support to this notion. LMW thiols could boost antioxidant and antinitrosative defenses by at least two independent mechanisms. First, terminal oxidation of GSH may be an important mechanism for defense against the attack of reactive species, in which case its de novo biosynthesis becomes an important aspect of antioxidant and antinitrosative defense. Second, formation of mixed disulfides between GSH and cysteines in proteins could not only explain the drop in GSH noted in NO-treated Salmonella, but it may represent a cytoprotective mechanism as well. Mixed disulfides can function as redox-sensitive switches that mediate signal transduction, and can protect redox-active cysteines from becoming over oxidized to frequently irreversible sulfinic and sulfonic acids (Chi et al., 2011). Moreover, mixed disulfides can be reduced by glutaredoxins (Peltoniemi et al., 2006). Further investigations will be needed to determine whether and to what extent mixed disulfides contribute to the antioxidant and antinitrosative defenses of Salmonella.

In summary, our investigations demonstrate that LMW thiols of the GSH biosynthetic pathway represent a first line of antioxidant and antinitrosative defense against the reactive oxygen and nitrogen species generated by the NADPH oxidase and iNOS in an acute model of NRAMP1^R Salmonella infection. The role played by LMW thiols in Salmonella pathogenesis likely represents the combined protection of essential biomolecules and the potentiation of Salmonella-specific virulence factors that lessen the cytotoxicity associated with host cell NADPH oxidase and iNOS hemoproteins.

Experimental procedures

Bacterial Strains

Table 1 lists the plasmids and the strains of S. enterica serovar Typhimurium derived from ATCC 14028s used in the course of these investigations. Mutations in the Salmonella chromosome were generated by following the λ Red-mediated gene replacement method (Datsenko and Wanner, 2000). Primers encoding 40 nucleotides homologous to a target gene followed by 20 nucleotides homologous to the pKD13 template plasmid were used in PCR reactions to amplify the Flp recombinant target (FRT)-flanked kanamycin resistance cassette (Table 2). The amplicons were digested with DpnI and the specimens were electroporated into S. Typhimurium strain TT22236 carrying the pTP2223 plasmid that expresses the λ Red recombinase under P_tac control. Mutations were moved into S. Typhimurium strain 14028s by P22HT int-mediated transduction. In-frame deletions were generated by recombining the two FRT sites flanking the kanamycin resistance cassette with the Flp recombinase encoded in the pCP20 plasmid (Cherepanov and Wackernagel, 1995; Datsenko and Wanner, 2000). The mutations were confirmed by PCR analysis. Transcriptional lacZY fusions were constructed by pCP20-mediated integration of the pCE36 plasmid carrying a promoterless lacZY operon into unique FRT scars (Ellermeier et al., 2002). A wild-type gshA allele driven by its native promoter was cloned into the backbone of the low-copy vector pWSK29 (Wang and Kushner, 1991), generating the complementing plasmid pGSHA.

Table 1.

Strains of Salmonella enterica serovar Typhimurium and plasmids

Strain	Description	Source
14028s	Wild type	ATCC
AV0203	sifA::lacZY::km	(McCollister et al., 2005)
AV0207	spiC::lacZY::km	(McCollister et al., 2005)
AV0305	hmp::lacZY::km	This study
AV0468	Δhmp	(McCollister et al., 2005)
AV0539	Δhmp spiC::lacZY::km	This study
AV06119	Δgor	This study
AV09094	ΔgltBD	This study
AV09286	ΔgshA	This study
AV09290	Δhmp ΔgshA	This study
AV09291	gshA::lacZY::km	This study
AV09292	ΔgshA sifA::lacZY::km	This study
AV09296	Δhmp gshA::lacZY::km	This study
AV09301	ΔgshA pGSHA	This study
AV09302	ΔgshA spiC::lacZY::km	This study
AV09307	ΔgshA hmp::lacZY::km	This study
AV09315	ΔgshA spiC::lacZY::km pGSHA	This study
AV09328	ΔgshB	This study
AV09337	ΔgshB spiC::lacZY::km	This study
Plasmids
pKD13	bla FRT ahp FRT PS1 PS4 oriR6K	(Datsenko and Wanner, 2000)
pCP20	bla cat cI857 λPR flp pSC101 oriTS	(Datsenko and Wanner, 2000)
pCE36	ahp FRT lacZY+ this oriR6K	(Ellermeier et al., 2002)
pWSK29	bla lacZ oripSC101	(Wang and Kushner, 1991)
pGSHA	bla lacZ oripSC101::gshA	This study

Table 2.

Primers

Gene	Primers
gshA	F 5′-ttgatcccggacgtatcacaggctctggcctggctggaaagtgtaggctggagctgcttc-3′
	R 5′-tcaggcgtgttttgcaagccacgcggcaaacggctcggtaattccggggatccgtcgacc-3′
gshB	F 5′-acttaggatgaagcgttaacggagaaaatagctggagctgcttcgaagtt-3′
	R 5′-aagcgcacgccaccgggcaataaacctttattccggggatccgtcgacct-3′
gor	F 5′-cagaatctatccgtccgacaacgaacataaaggagcggtcgctggagctgcttcgaagtt-3′
	R 5′-cttgtgcgcccacccttgaagatgggcgcacaagatattattccggggatccgtcgacct-3′
gltB	F 5′ gaaggcgaacctagccacaaggtagtgcgtaccgccatacgtgtaggctggagctgcttc-3′
	R 5′-caccaccagatcggaaccgcgcacgatgtcgccaccggcgattccggggatccgtcgacc-3′

^*The sequences for pKD13 are underlined.

Quantification of GSH and GSSG

Total GSH was quantified using the GSH recycling method described by (Baker et al., 1990) with modifications by (Henard et al., 2010). Briefly, bacterial cultures grown for 2 h in Luria-Bertani (LB) broth to A₆₀₀ of 1.0 were exposed to 500 μM of spermine NONOate for 5 or 30 min. Control and NO-treated cells were centrifuged at 13,000 RPM for 1 min, and the bacterial pellets were resuspended in 250 μl of 20 mM EDTA, pH 8.0. Samples used for the determination of GSSG were resuspended in 20 mM EDTA containing 2 mM N-ethylmaleimide. The bacteria were lysed by sonication, and the cytoplasmic proteins in the specimens were precipitated upon the addition of one volume of 10% HClO₄. The samples were neutralized with 93.5 μl 5 M KOH, freeze/thawed, and centrifuged to obtain cleared lysates. Reaction buffer was freshly prepared by mixing 2.8 ml of 1 mM 5,5′-dithiobis (2-nitro-benzoic acid), 3.75 ml of 1 mM NADPH, 5.85 ml of 100 mM NaH₂PO₄-5 mM EDTA phosphate-EDTA buffer, and 20 U of glutathione reductase (Sigma-Aldrich). 50 μl of the specimens were mixed with 100 μl of the reaction buffer in a 96-well microtiter plate. After a 5-min incubation, the reactions were read at A₄₁₂ in a spectrophotometer (Versamax Micrioplate Reader, Molecular Devices, USA). The concentration of GSH in the samples was estimated by regression analysis of known standards. The intracellular concentrations of GSH and GSSG were calculated taking into account the number of colony forming units and a bacterial cell volume of 10⁻¹⁵ liters.

In vitro Susceptibility to ROS

Salmonella cultures grown overnight in LB broth were diluted in PBS to a cell density of 5 × 10⁵ CFU/ml. The bacteria were challenged with increasing concentrations of H₂O₂ at 37°C in 96-well plates. After the indicated times of H₂O₂ challenge, the bacterial cultures were serially diluted in PBS and the specimens were spotted on LB agar plates. The number of bacteria capable of forming a colony was estimated after overnight culture on LB agar plates. The data are represented as the fraction of bacteria that survives H₂O₂ treatment.

Effect of NO on Bacterial Growth

Bacterial cultures grown overnight in LB broth were inoculated into fresh LB broth to a final concentration of 5 × 10⁶ cfu/ml. The bacteria were challenged with 5 mM of the polyamine diethylenetriamine control (DETA; Sigma-Aldrich, St. Louis, MO) or the NO donor DETA NONOate (Cayman Chemical, Ann Arbor, MI) dissolved in 10 mM Tris-HCl, pH 8.5. A mole of DETA NONOate liberates 2 moles of NO with a half-life of 20 h. Experimental estimates indicate that 5 mM DETA NONOate generate a brief burst of ~7.5 μM NO, followed by a steady flux of 5 μM NO for the duration of the experiment (Henard and Vazquez-Torres, 2012a). Bacterial growth measured as A₆₀₀ in a Bioscreen-C Growth analyzer (Oy Growth Curves AB Ltd, Helsinki, Finland) was recorded every 15 min for 20–30 h. The anti-Salmonella activity of GSNO was measured in disk-diffusion assays as previously described (De Groote et al., 1995).

Transcriptional Analysis

S. Typhimurium strains harboring gshA::lacZY, hmp::lacZY, sifA::lacZY or spiC::lacZY transcriptional fusions were grown overnight in LB broth. The cells were subcultured 1:100 in 10 mM Mg²⁺ N salts medium [5 mM KCl, 7.5 mM (NH₄)SO₄, 0.5 mM K₂SO₄, 1 mM KH₂PO₄, 38 mM glycerol, 0.1% casamino acids supplemented with 10 mM MgCl₂ and 100 mM Tris-HCl, pH 7.6]. The specimens were grown at 37°C in a shaking incubator until the bacterial cultures reached an A₆₀₀ of ~0.5. The bacterial cells were washed 3 times in 8 μM MgCl₂ N salts medium, pH 6.9, and the optical density adjusted to A₆₀₀ of 0.5. Spermine NONOate (Cayman Chemical) dissolved in 10 mM Tris-HCl buffer, pH 8.5, was added to bacteria that had been grown for 1 h in 8 μM MgCl₂ N salts medium, pH 6.9. Spermine NONOate has a half-life of 39 min at 37°C and releases 2 moles of NO per mole of the parent compound. The expression of the lacZY transcriptional fusions was quantified spectrophotometrically at the indicated times as β-galactosidase enzymatic activity using the substrate o-nitrophenyl-β-D-galactopyranoside. β-galactosidase activity is expressed in Miller units calculated according to the equation $1,000\times [({\text{A}}_{420}-1.75\times {\text{A}}_{550})/(T_{(min)}\times V_{(\text{ml})}\times {\text{A}}_{600})]$.

Measurement of Respiratory Activity

Salmonella grown overnight in LB broth were diluted in pre-warmed LB broth and the cultures incubated at 37°C with shaking until the bacterial density reached an A₆₀₀ of ~0.2. The bacterial optical density was adjusted to A₆₀₀ of 0.2 with pre-warmed LB broth and the cultures equilibrated for 3 min at 37°C in a shaking incubator. The specimens were then transferred into a multi-port measurement chamber equipped with an ISO-OXY-2 O₂ probe attached to an APOLLO 4000 free radical analyzer (World Precision Instruments, Sarasota, FL). The NO donor proli NONOate (Cayman Chemical), which has a half-life of 1.8 sec at 37°C, was added to Salmonella in the air-sealed chamber when the concentration of dissolved O₂ in the media reached about 130 μM. The respiratory activity of control and NO-treated Salmonella is expressed as μM O₂ over time.

Glutamine oxoglutarate amidotransferase (GOGAT) enzymatic assay

GOGAT enzymatic activity was estimated as previously described (Miller and Stadtman, 1972). S. Typhimurium strains 14028s, AV09286 (ΔgshA::FRT) and AV09094 (ΔgltBD:FRT) were grown to stationary phase in 20 mM D-glucose, 1.7 mM MgSO₄, 14 mM K₂SO₄, 43 mM NaCl, 108 mM potassium phosphate minimum medium A, pH 7.0, containing 0.108 mM Na₂-EDTA, 9 μM CaCl₂, 0.63 μM ZnCl₂, 0.64 μM CuSO₄, and 20 mM NH₄Cl at 37°C in a shaker incubator. Stationary phase cultures diluted 1:100 in medium A containing 0.124 or 0.248 mM FeCl₃ were grown to OD₆₀₀ of 0.8. Where indicated, late log phase cultures were treated for 30 min with 100 μM spermine NONOate in 2 ml aliquots in 14 ml polystyrene tubes at 37°C in a shaker incubator. Control and spermine NONOate-treated bacteria were harvested by centrifugation, and the bacterial pellets were washed and resuspended in ice-cold, 50 mM Tris-HCl buffer, pH 7.6. Bacterial lysates were prepared by sonication and clear supernatants containing the cytoplasmic fractions were obtained by centrifugation. The concentration of protein in the cytoplasmic fractions was estimated using the Pierce 660 nm Protein Assay Reagent (Thermo Scientific, Rockford, IL). GOGAT activity was measured by the rate of oxidation of NADPH (Research Products International Corp. Mt. Prospect, IL). Briefly, 45 μg of protein were added to 300 μl of prewarmed reaction cocktail containing 50 mM Tris-HCl, 5 mM α-ketogluterate and 5 mM glutamine, pH 7.6. The reaction was started by adding 5 μl from a 7.5 mM NADPH stock. The consumption of NADPH was recorded for 1.5 min at 340 nm in a Cary 50 spectrophotometer. Specific GOGAT enzymatic activity is expressed as μmoles of NADPH oxidized per min per mg of protein using ε₃₄₀ of 6.22 × 10³.

Salmonella Virulence in a Murine Model Acute of Infection

Eight- to ten-week old NRAMP1^S C57BL/6 or NRAMP1^R C3H/HeNcrl/Br mice bred in our animal facility according to Institutional Animal Care and Use Committee guidelines were used in the course of these investigations. C57BL/6 and C3H/HeNCrl mice were challenged intraperitoneally with about 100 and 500 CFU/mouse, respectively. Where indicated, 100 μg/ml acetovanillone (Sigma-Aldrich) or 500 μg/ml L-NIL (Calbiochem, Billerica, MA), which are specific inhibitors NADPH oxidase or iNOS, respectively, were added into drinking water. The data are expressed as % of mice surviving over time.

Statistical Analysis

The data were analyzed using a Student’s paired t test. Determination of statistical significance between multiple comparisons was achieved using one-way analysis of variance (ANOVA) followed by a Bonferroni post-test. Data were considered statistically significant when p<0.05.

Abbreviations

GSNO

S-nitrosoglutathione

GSH

glutathione

GSSG

oxidized GSH

H₂O₂

hydrogen peroxide

iNOS

inducible nitric oxide synthase

LB

Luria-Bertani

LMW

low-molecular weight

L-NIL

N6-(1-iminoethyl)-L-lysine, dihydrochloride

NO₂

nitrogen dioxide

NO

nitric oxide

NO₃⁻

nitrate

N₂O

nitrous oxide

N₂O₃

dinitrogen trioxide

O₂^·−

superoxide anion

OH^·

hydroxyl radical

RNS

reactive nitrogen species

SPI2

Salmonella pathogenicity island 2