ThioredoxinA1 Controls the Oxidative Stress Response of Francisella tularensis Live Vaccine Strain (LVS)

ABSTRACT

Francisella tularensis is an intracellular, Gram-negative bacterium known for causing a disease known as tularemia in the Northern Hemisphere. F. tularensis is classified as a category A select agent by the CDC based on its possible use as a bioterror agent. F. tularensis overcomes oxidative stress encountered during its growth in the environment or host macrophages by encoding antioxidant enzymes such as superoxide dismutases, catalase, and alkylhydroperoxy reductase. These antioxidant enzymes are regulated by the oxidative stress response regulator, OxyR. In addition to these antioxidant enzymes, F. tularensis also encodes two thioredoxins, TrxA1 (FTL_0611) and TrxA2 (FTL_1224); however, their role in the oxidative stress response of F. tularensis is not known. This study investigated the role of thioredoxins of F. tularensis in the oxidative stress response and intracellular survival. Our results demonstrate that TrxA1 but not TrxA2 plays a major role in the oxidative stress response of F. tularensis. Most importantly, this study elucidates a novel mechanism through which the TrxA1 of F. tularensis controls the oxidative stress response by regulating the expression of the master regulator, oxyR. Further, TrxA1 is required for the intramacrophage survival and growth of Francisella. Overall, this study describes a novel role of thioredoxin, TrxA1, in regulating the oxidative stress response of F. tularensis.

IMPORTANCE The role of thioredoxins in the oxidative stress response of F. tularensis is not known. This study demonstrates that of the two thioredoxins, TrxA1 is vital to counter the oxidative stress in F. tularensis live vaccine strain (LVS). Furthermore, this study shows differences in the well-studied thioredoxins of Escherichia coli. First, the expression of TrxA1 of F. tularensis is independent of the oxidative stress response regulator, OxyR. Second and most importantly, TrxA1 regulates the expression of oxyR and, therefore, the OxyR-dependent oxidative stress response of F. tularensis. Overall, this study reports a novel regulatory role of TrxA1 of F. tularensis in the oxidative stress response.

INTRODUCTION

Francisella tularensis is a Gram-negative, intracellular bacterium known for causing a disease known as tularemia in the Northern Hemisphere. Tularemia occurs in three major clinical forms—ulceroglandular, oropharyngeal, and pneumonic tularemia. These clinical forms result from contact with infected animals, contaminated food or water consumption, or direct inhalation of the bacteria (1–3). Each clinical form presents its own unique set of symptoms. However, pneumonic tularemia is lethal, with mortality rates as high as 60% (4, 5). The major concern is the potential of F. tularensis to be used as a bioweapon or a bioterror agent. F. tularensis may be aerosolized and survives for extended periods in the air, and as little as ~10 CFU are sufficient to cause lethal disease (4, 6). F. tularensis subsp. tularensis, also known as type A Francisella, possesses the most virulent strains of Francisella. The F. tularensis subsp. holarctica, also known as type B Francisella, is not as virulent as F. tularensis subsp. tularensis but is still able to cause disease in humans. The live vaccine strain (LVS) is derived from F. tularensis subsp. holarctica. Very little is known about F. tularensis subsp. mediasiatica regarding its ability to infect humans or the disease associated with this Francisella subspecies (4, 7). F. novicida is avirulent to humans; however, it still can cause disease in immunocompromised individuals (8).

F. tularensis is an intracellular pathogen. F. tularensis primarily infects macrophages but can also infect other cell types (9, 10). Once F. tularensis is taken up by the macrophages, the bacteria are enclosed in a phagosome (11, 12). F. tularensis prevents the phagosome fusion with lysosomes through mechanisms that are not entirely understood (11, 13–16). Bacteria then escape the phagosome into the macrophage cytosol, where replication occurs. Replication of F. tularensis inside the cytosol may eventually lead to cell death. Francisella may also enter an endocytic compartment through host-cell-mediated autophagy (17, 18).

The respiratory burst is generated by the NADPH oxidase complex immediately after the phagocytosis. It results in the production of reactive oxygen species (ROS) and reactive nitrogen species (RNS) (19). These toxic ROS/RNS include superoxide radicals (O_2–), hydrogen peroxide (H₂O₂), and nitric oxide (NO). H₂O₂ may also combine with ferrous iron (Fe²⁺) in the Fenton reaction to produce hydroxyl radicals (OH⁻), which are extremely toxic (20). F. tularensis possesses powerful antioxidant enzymes to overcome the bactericidal effects of ROS and RNS. Francisella encodes iron and Cu-Zn-containing superoxide dismutases (SodB and SodC, respectively), a catalase (KatG), and alkyl hydroperoxide reductase (AhpC) (21, 22). The Sods work by converting O_2– into H₂O₂ by adding hydrogen and the oxidation of the metal ion contained within the enzyme (23). KatG then degrades H₂O₂ into water and oxygen (24). AhpC works by reducing H₂O₂ and other oxidized compounds through its peroxidative cysteine (25). These primary antioxidant enzymes of F. tularensis are regulated by a transcriptional regulator, OxyR (22). In addition to F. tularensis requiring OxyR to produce antioxidant enzymes, F. tularensis also depends on the major facilitator superfamily (MFS)-type efflux pump, especially the EmrA1 membrane fusion protein for secretion of SodB and KatG (26). In addition, the outer membrane component of the MFS efflux pump, SilC, is also required for oxidative stress resistance and virulence in mice (27). F. tularensis also relies on the stringent stress response genes relA and spoT to resist oxidative stress. RelA and SpoT regulate the production of the alarmone (p)ppGpp, and mutants lacking relA and spoT genes show increased susceptibility to oxidants and attenuation of virulence (28, 29). Additionally, a transcriptional regulator belonging to the AraC/XylS family also modulates the oxidative stress response of F. tularensis (30).

In addition, enzymes belonging to the thioredoxin (Trx) system are also required to overcome oxidative stress. Thioredoxins maintain the delicate redox state of the cell by reducing oxidized sulfur or thiol residues that may otherwise be damaging to the organism. The thioredoxin system consists of an electron carrier, NADP, and a thioredoxin reductase. The NADP and thioredoxin reductase contribute to thioredoxin function by recycling electrons and maintaining it in a reduced state. Reduced thioredoxin acts by reducing oxidized sulfur residues on proteins (31). Studies have shown that thioredoxins are also required for the virulence of many bacterial pathogens (32, 33). F. tularensis possesses two thioredoxin genes, trxA1 (FTL_0611) and trxA2 (FTL_1224), and a single thioredoxin reductase gene, trxR (FTL_1571). However, the role of thioredoxins in the oxidative stress response of F. tularensis is not known. This study investigated how thioredoxins of F. tularensis contribute to its oxidative stress response and intracellular survival.

RESULTS

Bioinformatic analysis and confirmation of trxA1 and trxA2 gene deletions and transcomplementation.

Both the FTL_0611 and the FTL_1224 genes of F. tularensis LVS are annotated as trxA genes belonging to the thioredoxin family in the F. tularensis LVS genome. We reannotated these two genes in the manuscript as trxA1 (FTL_0611) and trxA2 (FTL_1224) for convenience. The trxA1 and trxA2 genes of F. tularensis LVS are 324 (107 amino acids) and 327 (108 amino acids) base pairs long, respectively. The amino acid sequences of TrxA1 of F. tularensis LVS show 100% homology to TrxA1 (FTT_1445) of F. tularensis SchuS4 and Trx1 (FTN_1415) of F. novicida. TrxA2 shows 100% amino acid sequence homology to TrxA2 (FTT_0976) of F. tularensis SchuS4 and F. novicida (not shown). TrxA1 and TrxA2 of F. tularensis share 40% homology at the amino acid level (Fig. 1A), suggesting that these two are functionally different proteins. Both TrxA1 and TrxA2 have a catalytic WCxxC active site motif conserved in all thioredoxins. TrxA1 has two additional cysteines, which are absent in TrxA2. The TrxA1 protein of F. tularensis exhibits 38%, 47%, and 24% identity to the well-characterized TrxC proteins of Escherichia coli, Helicobacter pylori, and Rhodobacter capsulatus, respectively (Fig. 1B). However, TrxA1 of F. tularensis LVS lacks the two characteristic N-terminal CxxC motifs present in thioredoxins of E. coli (34). The trxA1 gene is flanked by genes encoding transcription termination factor Rho (FTL_0610), and an unannotated FTL_0612 gene and is predicted to be transcribed as an operon with FTL_610 (see Fig. S1A in the supplemental material). The trxA2 gene is flanked by genes encoding hypothetical proteins and transcribed as an operon with the neighboring genes (Fig. S1D). Deletion of the trxA1 and trxA2 genes in the ΔtrxA1 and ΔtrxA2 mutants, respectively, was confirmed by PCR. Amplification of fragments of smaller sizes (382 and 242 bp) than those of the wild-type F. tularensis LVS using the primers flanking the trxA1 and trxA2 genes confirmed the deletion of these genes in the ΔtrxA1 and ΔtrxA2 mutants, respectively (Fig. S1B and Fig. S1E). The trxA1 and trxA2 gene deletions and transcomplementations were further confirmed using gene-specific primers (Fig. S1C and Fig. S1F). The in-frame gene deletions of the trxA1 and trxA2 genes were confirmed by DNA sequencing (not shown). Our attempts to generate trxA1 trxA2 double gene-deletion mutants were unsuccessful, indicating that the deletion of both genes is detrimental to Francisella.

FIG 1.

Bioinformatic analysis. (A) Amino acid sequence alignments of TrxA1 and TrxA2 of F. tularensis LVS. (B) Amino acid sequence alignments of TrxA1 of F. tularensis LVS with TrxC of E. coli, H. pylori, and R. capsulatus. The dashed red boxes indicate the characteristic WCxxC catalytic motif and cysteine residues.

The ΔtrxA1 mutant of F. tularensis LVS is susceptible to oxidative stress.

Disc diffusion and bacterial killing assays were used to determine if the loss of either trxA1 or trxA2 genes reduces the ability of F. tularensis LVS to resist oxidative stress induced by the superoxide-generating compounds paraquat and pyrogallol. Paraquat induces superoxide radicals intracellularly via redox cycling of molecular oxygen (35), whereas pyrogallol undergoes auto-oxidation to produce superoxide extracellularly (36). The ΔtrxA1 mutant showed a higher sensitivity toward both paraquat and pyrogallol as observed by the significantly larger zone of inhibition in the disc diffusion assay (Fig. 2A and C) and enhanced killing in the bacterial killing assay compared to the wild-type F. tularensis LVS and the ΔtrxA2 mutant. Enhanced killing of the ΔtrxA2 mutant compared to the wild-type F. tularensis LVS was observed only after 3 h of exposure to paraquat and pyrogallol. However, the number of viable ΔtrxA2 mutant bacteria remained significantly higher than that of the ΔtrxA1 mutant after 3 h of exposure to both these superoxide-generating compounds (Fig. 2B and D). Collectively, these results demonstrate that TrxA1 plays an important role in overcoming the oxidative stress induced by superoxide radicals.

FIG 2.

Oxidant susceptibilities of the thioredoxin mutants of F. tularensis LVS. (A to D) The susceptibility of the wild-type F. tularensis LVS, ΔtrxA1 mutant, its transcomplemented strain (ΔtrxA1+ptrxA1), the ΔtrxA2 mutant, and its transcomplemented strain (ΔtrxA2+ptrxA2) was tested against paraquat and pyrogallol using disc diffusion (A and C) and bacterial killing assays (B and D). (E and F) Growth curves of the indicated Francisella strains in the absence (E) and presence (F) of hydrogen peroxide (H₂O₂). (G) Bacterial killing assay of the indicated Francisella strains in the presence of H₂O₂. (H to J) Disc diffusion assays using tert-butyl hydroperoxide (TBH) and cumene hydroperoxide (CHP) and diamide. (K) Bacterial killing assay using diamide. The data shown in panels A to D and G to K are cumulative of three independent experiments, each conducted with three technical replicates, and are expressed as the mean ± SEM. The data were analyzed using one-way ANOVA. *, P < 0.05; **, P < 0.01; ***, P < 0.001. The data shown in panels E and F are representative of two to three independent experiments.

The ΔtrxA1 and ΔtrxA2 mutants of F. tularensis LVS were tested further for their susceptibilities to oxidative stress induced by H₂O₂. Since culture medium on the plates impedes the diffusion of H₂O_2, growth curves, rather than the disc diffusion assay, were used to determine the sensitivity of the wild-type F. tularensis LVS, the ΔtrxA1 and ΔtrxA2 mutants, and their respective transcomplemented strains to H₂O₂. In the absence of H₂O₂, wild-type F. tularensis LVS grew to an optical density at 600 nm (OD₆₀₀) of approximately 2.0 after 28 h of growth. The ΔtrxA1 mutant grew slowly and achieved an OD₆₀₀ of 1.4 in the same period of growth. The ΔtrxA2 mutant also grew marginally slower than the wild-type F. tularensis LVS, with an OD₆₀₀ of 1.7. The ΔtrxA1+ptrxA1 transcomplemented strain grew better than its mutant counterpart, while the ΔtrxA2+ptrxA2 transcomplemented strain grew similarly to wild-type F. tularensis LVS. However, these growth characteristics altered significantly in the presence of H₂O₂. Wild-type F. tularensis LVS, the ΔtrxA2 mutant, and its transcomplemented strain, ΔtrA21+ptrxA2, grew similarly and achieved an OD₆₀₀ of approximately 1.4 to 1.7. The ΔtrxA1 mutant failed to grow beyond the starting OD₆₀₀ of 0.05 in the presence of H₂O₂, and transcomplementation restored the growth of the ΔtrxA1 mutant to wild-type F. tularensis LVS levels (Fig. 2E and F). Bacterial killing assays further confirmed the results from the growth curves. The results showed the enhanced killing of the ΔtrxA1 mutant after 1 and 3 h of exposure to H₂O₂. Transcomplementation with the trxA1 gene restored the viability of the ΔtrxA1 mutant similar to that of wild-type F. tularensis LVS. The viability of the ΔtrxA2 mutant remained identical to the that of wild-type F. tularensis LVS after 1 h of exposure to H₂O₂. Significantly reduced numbers of viable ΔtrxA2 mutant bacteria compared to wild-type F. tularensis LVS were recovered 3 h postexposure. However, these numbers were significantly higher than those observed for the ΔtrxA1 mutant (Fig. 2G). Disc diffusion assays were performed to assess the sensitivity of the ΔtrxA1 or ΔtrxA2 mutants of F. tularensis LVS to the organic peroxides cumene hydroperoxide (CHP) and tert-butyl hydroperoxide (TBH). Enhanced sensitivity toward TBH was observed for both the ΔtrxA1 and ΔtrxA2 mutants, while only the ΔtrxA1 mutant was more sensitive to CHP than wild-type F. tularensis or the ΔtrxA2 mutant. Transcomplementation of the ΔtrxA1 mutant restored the sensitivity to wild-type F. tularensis LVS levels against both TBH and CHP (Fig. 2H and I). Together, these results indicate that TxrA1 is required for resistance against the oxidative stress induced by H₂O₂ and organic peroxides. The TrxA2, on the other hand, provides oxidative stress resistance when exposed to oxidants for prolonged periods.

Both ΔtrxA1 and ΔtrxA2 mutants exhibit enhanced susceptibility to the thiol-oxidizing agent diamide.

To determine if the enhanced sensitivity of thioredoxin mutants toward oxidants is due to a failure to counter thiol-related oxidative stress, we next tested the ability of ΔtrxA1 and ΔtrxA2 mutants to resist thiol-oxidizing stress. Diamide, a compound that specifically induces thiol oxidation, was used in disc diffusion assays. Both the ΔtrxA1 and ΔtrxA2 mutants showed enhanced sensitivity toward diamide as indicated by larger zones of inhibition (14.7 ± 0.5 and 14.5 ± 0.5 mm, respectively) than wild-type F. tularensis LVS (13 ± 0.5 mm) (Fig. 2J). Similar results were replicated in the bacterial killing assay, with significantly lower numbers of viable ΔtrxA1 mutant bacteria after 1 h, and ΔtrxA1 and ΔtrxA2 mutant bacteria were recovered after 3 h of exposure to diamide (Fig. 2K). Transcomplemented strains of both ΔtrxA1 and ΔtrxA2 mutants exhibited zones of inhibition in disc diffusion assays and bacterial viability in bacterial killing assays similar to those of wild-type F. tularensis LVS. These results indicated that loss of thioredoxin genes, especially trxA1, increases the sensitivity of F. tularensis LVS toward the thiol-oxidizing agent diamide.

Enhanced expression of glutaredoxins is observed in the ΔtrxA1 mutant of F. tularensis.

Our preceding results demonstrated an enhanced susceptibility of the ΔtrxA1 mutant to oxidative stress compared to F. tularensis LVS or the ΔtrxA2 mutant. We next investigated the mechanisms responsible for this enhanced susceptibility. In addition to Sods, katG, ahpC, and trx, Francisella also encodes glutathione (GSH)-dependent systems. The GSH-dependent antioxidant system, also known as glutaredoxin (Grx), comprises Grx, NADP, GSH, and GSH reductase. The Grx system, together with the Trx system, controls cellular redox homeostasis. Glutathione peroxidase (gpx) encoded by the FTT_0733 gene provides oxidative stress resistance to the F. tularensis SchuS4 strain (37). In addition to gpx, F. tularensis also encodes glutaredoxin grx2 (FTL_0923). The expression of trxA2, gpx, and grx2 genes was determined in the wild-type F. tularensis LVS, the ΔtrxA1 mutant, and the transcomplemented strain in the absence and presence of oxidative stress induced by H₂O₂ by quantitative reverse transcription PCR (qRT-PCR). The results revealed that expression of trxA2, gpx, and grx2 genes remained similar in the wild-type F. tularensis LVS, the ΔtrxA1 mutant, and the transcomplemented strain in both the absence and presence of oxidative stress (Fig. 3A). These results indicated that the trxA2, gpx, and grx2 genes are not overexpressed in the absence of the trxA1 gene.

FIG 3.

Transcription profile of the glutaredoxin genes. Wild-type F. tularensis LVS, the ΔtxrA1 mutant, and ΔtxrA1+ptxrA1 transcomplemented strains were either left untreated or exposed to 1 mM H₂O₂ for 2 h to induce oxidative stress. (A and B) The RNA was isolated and quantitated for the expression of (A) trxA2, gpx, and grx2 genes and (B) rrse, grx1, and grx3 by qRT-PCR. The data are represented as the relative fold change compared to F. tularensis LVS. The data shown are cumulative of three independent experiments, each conducted with three technical replicates, expressed as the mean ± SEM and analyzed using one-way ANOVA. *, P < 0.05; **, P < 0.01.

In addition to gpx and grx2, the glutaredoxin 1 (grx1) and glutaredoxin 3 (grx3) genes are cotranscribed with ribonucleoside diphosphate reductase subunit alpha (rrse) (Fig. 3B). We next determined if genes of the Grx system are upregulated in the absence of the trxA1 gene. Significantly upregulated expression of all these genes of the Grx system was observed in the ΔtrxA1 mutant compared to that of F. tularensis LVS. Transcomplementation of the ΔtrxA1 mutant restored the expression of rrse, grx1, and grx3 genes to wild-type levels in both the absence and the presence of the oxidative stress induced by H₂O₂ (Fig. 3B). These results demonstrated compensatory changes in the Grx system in response to trxA1 deficiency.

The expression of oxyR is dependent on TrxA1 under oxidative stress conditions.

The expression of trxA1 and trxA2 in the absence or the presence of oxidative stress induced by H₂O₂ was next determined in the wild-type F. tularensis LVS by qRT-PCR. The expression of the trxA1 gene was significantly upregulated upon exposure to oxidative stress induced by H₂O_2. However, the expression of trxA2 remained unaltered, indicating an association of trxA1 with the oxidative stress response (Fig. 4A). We determined the expression of the oxyR gene in the absence or the presence of oxidative stress. Similar to the txrA1 gene, the expression of the oxyR gene was significantly upregulated in the wild-type F. tularensis LVS upon exposure to oxidative stress (Fig. 4B). These results indicated that oxyR and txrA1 are upregulated under oxidative stress conditions.

FIG 4.

TrxA1-dependent expression of the oxyR gene. Wild-type F. tularensis LVS was either left untreated or exposed to 1 mM H₂O₂ for 2 h to induce oxidative stress. (A and B) The RNA was isolated and quantitated for the expression of (A) trxA1 and trxA2 and (B) oxyR genes. (C) Wild-type F. tularensis LVS, the ΔoxyR mutant, and ΔoxyR+poxyR transcomplemented strains were either left untreated or exposed to 1 mM H₂O₂ for 2 h to induce oxidative stress. The RNA was isolated and quantitated for the expression of trxA1 and trxA2 genes. (D) Wild-type F. tularensis LVS, the ΔtxrA1 mutant, and ΔtxrA1+ptxrA1transcomplemented strains were either left untreated or exposed to 1 mM H₂O₂ for 2 h to induce oxidative stress. The RNA was isolated and quantitated for the expression of oxyR by qRT-PCR. The data are represented as the relative fold change compared to F. tularensis LVS. The data shown are cumulative of three independent experiments, each conducted with three technical replicates, expressed as the mean ± SEM and analyzed using one-way ANOVA. *, P < 0.05; **, P < 0.01.

In E. coli, the expression of the thioredoxin 2 (Trx2)-encoding gene trxC is OxyR-dependent (38). To determine if OxyR similarly regulates the txrA1 gene of F. tularensis, the expression of txrA1 was determined in the ΔoxyR mutant of F. tularensis LVS in the absence or presence of oxidative stress induced by H₂O₂. The expression of the trxA1 gene remained unaltered in the ΔoxyR mutant compared to wild-type F. tularensis LVS or the transcomplemented strain in the absence or the presence of oxidative stress. Similar results were also observed for the trxA2 gene (Fig. 4C). These results indicated that expression of trxA1 and trxA2 is independent of OxyR in F. tularensis. We next determined if the expression of the oxyR gene is dependent on trxA1. The expression of the oxyR gene remained unaltered in the absence of oxidative stress in all three bacterial strains tested. However, the expression of the oxyR gene was significantly downregulated in the ΔtrxA1 mutant when exposed to oxidative stress. Transcomplementation of the trxA1 gene restored the expression of oxyR to a level similar to that observed in wild-type F. tularensis LVS (Fig. 4D). These results indicated that the expression of the oxyR gene is dependent on TrxA1 when F. tularensis LVS is exposed to oxidative stress.

OxyR-dependent antioxidant enzyme genes ahpC, katG, and sodB are downregulated in the ΔtrxA1 mutant under oxidative stress conditions.

The expression of antioxidant enzyme genes ahpC, katG, and sodB of F. tularensis are dependent on OxyR (22). The expression of these antioxidant enzyme genes was determined in the absence or the presence of oxidative stress induced by H₂O₂ in the wild-type F. tularensis LVS, the Δtrx mutant, and the transcomplemented strain by qRT-PCR. In the absence of oxidative stress, all three antioxidant enzyme genes showed little to no difference in their expression profiles compared to the wild-type F. tularensis LVS. However, upon exposure to H₂O₂, the expression of the ahpC, katG, and sodB genes was significantly downregulated in the ΔtrxA1 mutant compared to wild-type F. tularensis LVS. Transcomplementation of the ΔtrxA1 mutant restored the expression of all three genes to levels similar to those observed in the wild-type F. tularensis LVS (Fig. 5A and B). Our preceding results demonstrated that the ΔtrxA1 mutant is susceptible to H₂O₂, resulting in reduced viability when exposed to 1 mM H₂O₂ (Fig. 2F and G). To validate that the downregulated expression of ahpC, katG, and sodB genes is not due to the enhanced killing of the ΔtrxA1 mutant at 1 mM H₂O₂, we performed gene expression analysis using the lower concentrations (250 and 500 μM) of H₂O₂ that are sufficient to induce oxidative stress without significant alteration of the bacterial viability. The expression profiles of the ahpC, katG, and sodB genes at 250 and 500 μM H₂O₂ mirrored those observed for the higher concentration of H₂O₂ (Fig. 5C and D).

FIG 5.

Expression of antioxidant enzymes in the ΔtrxA1 mutant of F. tularensis. (A to D) Wild-type F. tularensis LVS, the ΔtxrA1 mutant, and ΔtxrA1+ptxrA1 transcomplemented strains were either left untreated (A) or exposed to 1 mM (B), 250 μM (C), or 500 μM (D) of H₂O₂ for 2 h to induce oxidative stress. The RNA was isolated and quantitated for the expression of ahpC, katG, and sodB genes by qRT-PCR. The data are represented as the relative fold change compared to F. tularensis LVS. The data shown are cumulative of three independent experiments, each conducted with three technical replicates, expressed as the mean ± SEM and analyzed using one-way ANOVA. *, P < 0.05; **, P < 0.01; ***, P < 0.001. (E) The Western blot analysis of the lysates of the indicated Francisella strains in the absence or the presence of 1 mM H₂O₂ probed with anti-Francisella KatG antibodies. The blots were stripped and reprobed with anti-Francisella sodB antibodies.

We also confirmed the results of qRT-PCR by detecting the protein levels of KatG and SodB in the wild type F. tularensis LVS and the ΔtrxA1 mutant in the absence or the presence of oxidative stress by Western blotting. In addition, the ΔtrxA2 mutant and the transcomplemented strains were included as additional controls. The results showed equivalent levels of KatG and SodB in lysates of all the strains tested in the absence of oxidative stress. However, the levels of both KatG and SodB were much lower in the ΔtrxA1 mutant when exposed to H₂O₂ and were restored to the wild-type levels in the transcomplement strain (Fig. 5E). Collectively, these results indicate that trxA1 regulates the expression of oxyR, which in turn regulates the expression of primary antioxidant enzyme genes to resist oxidative stress. This notion was investigated further.

TrxA1 of F. tularensis LVS regulates the expression of oxyR under the conditions of oxidative stress.

To validate the results from gene expression studies showing that TrxA1 indeed regulates the expression of oxyR, we performed an OxyR reporter assay. We generated lacZ-control and lacZ-oxyR reporter strains of the wild-type F. tularensis LVS and the ΔtrxA1 mutant (Fig. 6A). The expression of the oxyR gene in LacZ reporter strains of F. tularensis LVS and the ΔtrxA1 mutant was determined by a β-galactosidase activity assay. The lacZ-control reporter (pCZ) elicited virtually no β-galactosidase activity in wild-type F. tularensis LVS and the ΔtrxA1 mutant. The β-galactosidase activity detected in the wild-type F. tularensis LVS lacZ-oxyR reporter (F. tularensis pOZ) in the absence of oxidative stress indicated the base-level expression of oxyR. However, the β-galactosidase activity in the wild-type F. tularensis LVS lacZ-oxyR reporter strain was nearly three times that observed when the bacteria were exposed to oxidative stress compared to the untreated controls, indicating upregulated expression of oxyR. In the ΔtrxA1 mutant, however, the β-galactosidase activity was significantly lower than that observed for wild-type F. tularensis LVS. Furthermore, it remained unaltered in the absence or the presence of oxidative stress (Fig. 6B). These results demonstrate that TrxA1 of F. tularensis regulates the expression of oxyR in the presence of oxidative stress.

FIG 6.

TrxA1 of F. tularensis LVS regulates the expression of oxyR under the conditions of oxidative stress. (A) Schematic for generation of lacZ-control (pCZ) and lacZ-oxyR reporter (pOZ) strains of F. tularensis LVS and theΔtrxA1 mutant. (B) F. tularensis LVS and the ΔtrxA1 mutant and their reporter strains were grown in MHB for 3 h with or without 1 mM H₂O₂, and the β-galactosidase activity was determined. The data shown are cumulative of three independent experiments, each conducted with three technical replicates, expressed as the mean ± SEM and analyzed using one-way ANOVA. **, P < 0.01; ***, P < 0.001.

TrxA1 of F. tularensis LVS facilitates intramacrophage survival by overcoming oxidative stress.

Our preceding results demonstrated that TrxA1 is important for oxidative stress under acellular growth conditions. We next determined the contribution of TrxA1 to intramacrophage survival and growth. We employed cell culture assays using bone marrow-derived macrophages (BMDMs) derived from wild-type C57BL/6 and gp91phox^−/− mice that cannot generate NADPH oxidase-dependent ROS. Significantly lower numbers of the ΔtrxA1 mutant bacteria were recovered from the BMDMs derived from wild-type mice compared to the wild-type F. tularensis LVS at 4 and 24 h postinfection. Transcomplementation restored the intramacrophage survival of the ΔtrxA1 mutant, similar to that of wild-type F. tularensis LVS. Further, the survival and replication of the ΔtrxA1 mutant were also restored, similar to those of the wild-type F. tularensis LVS in gp91phox^−/− BMDMs (Fig. 7). These results indicate that TrxA1 of F. tularensis contributes to intramacrophage survival by overcoming ROS-induced oxidative stress.

FIG 7.

The ΔtrxA1 mutant is attenuated for intramacrophage growth. Intramacrophage growth of the ΔtrxA1mutant compared to F. tularensis LVS and the transcomplemented ΔtrxA1+ptrxA1 strain at 100 MOI in wild-type and gp91phox^−/− bone marrow-derived macrophages at 4 and 24 h postinfection. The data, represented as the mean ± SEM, are cumulative of three independent experiments, each conducted with four technical replicates and analyzed using one-way ANOVA. *, P < 0.05; ***, P < 0.001.

DISCUSSION

Bacterial thioredoxins are required for several cellular functions. Thioredoxins are important for oxidative stress response and function by reducing H₂O₂ and quenching and scavenging superoxide and hydroxyl radicals, respectively (39). They are required for the reduction of important enzymes such as ribonucleotide reductase and methionine sulfoxide reductase involved in DNA synthesis and protein repair mechanisms (40, 41). Thioredoxins also play a role in cell division, transcriptional regulation, energy transduction, protein folding and degradation, and biosynthetic pathways (42). In addition to these critical cellular processes, thioredoxins are vital for the intracellular survival and virulence of several bacterial pathogens. This study reports how thioredoxin of F. tularensis controls the oxidative stress response by regulating a critical component of the oxidative stress response machinery and contributes to intracellular survival.

The thioredoxin system of Francisella comprises two thioredoxins encoded by FTL_0611 and the FTL_1224 genes. Since both these genes are annotated as trxA in the F. tularensis LVS genome, we designated FTL_0611 as trxA1 and FTL_1224 as trxA2. The TrxA1 exhibits structural and regulatory features characteristic of bacterial thioredoxins, including the CxxC motif, but lacks two additional N-terminal CxxC motifs present in Trx2 of E. coli. Instead, the TxrA1 has two cysteine residues. The amino acid sequences of both TrxA1 and TrxA2 are 100% conserved in different subspecies of Francisella. Both TrxA1 and TrxA2 of Francisella exhibit only 38% homology at the amino acid level with the thioredoxins Trx1 and Trx2 of E. coli. Furthermore, recovery of the gene deletion mutants of both trxA1 and trxA2 indicates that these genes individually are not essential for viability, especially when bacteria are not exposed to exogenous oxidative stress. However, we could not recover the trxA1 trxA2 double gene deletion mutant, indicating that deletion of both genes is detrimental to Francisella.

The results demonstrated that the ΔtrxA1 mutant is more susceptible to superoxide radicals, organic peroxides, H₂O₂, and diamide than the wild-type F. tularensis LVS and the ΔtrxA2 mutant. These observations indicate that the TrxA1 protein of F. tularensis is similar in functionality to Trx2 encoded by the trxC gene of E. coli and the photosynthetic bacterium Rhodobacter capsulatus and TrxA encoded by the trxA gene of Helicobacter pylori (43–46). Furthermore, similar to that of the latter bacterial species, the second thioredoxin, TrxA2 of Francisella, appears to play a secondary and less important role in oxidative stress response. In addition to the Trx system, the GSH-dependent Grx antioxidant system participates in oxidative stress resistance mechanisms in parallel. Also, a cross talk between the Trx and Grx systems has been established. Studies have shown that the GSH-Grx system compensates for the Trx deficiency (47, 48). Accordingly, we observed overexpression of rrse, grx1, and grx3 genes but not of gpx and grx2 genes in the ΔtrxA1 mutant, in both the absence and the presence of oxidative stress. However, these compensatory changes are not sufficient to alter the oxidant-sensitive phenotype of the ΔtrxA1 mutant. On the other hand, the E. coli Grx system can functionally complement the thioredoxin system, and therefore the mutants carrying trxA and trxC gene deletions are viable (47). However, we could not recover a viable trxA1 trxA2 double gene mutant, which further supports the observation that the Grx system does not complement the loss of the Trx system of F. tularensis.

The results from this study strongly suggest that TrxA1 is vital to counter the oxidative stress in F. tularensis LVS. Our results demonstrate enhanced expression of trxA1 and oxyR but not the trxA2 gene when Francisella is subjected to oxidative stress. The results also showed that trxA2 does not compensate for the loss of trxA1. In E. coli and Rhodobacter, OxyR regulates trxC and other antioxidant enzyme genes such as katG, ahpC, and grxA (38, 43). Our data showed that the trxA1 gene of F. tularensis appeared to be functionally similar to the trxC gene of E. coli. Further, an identical expression profile of trxA1 and oxyR in the presence of oxidative stress indicated that OxyR similarly regulates trxA1 gene expression in Francisella. However, the trxA1 and trxA2 gene expression levels in the absence or presence of oxidative stress remained similar to those of the wild-type F. tularensis in the ΔoxyR mutant, indicating that OxyR does not control the expression of Francisella thioredoxins. These results corroborated an earlier report that OxyR of F. tularensis controls the expression of thioredoxin reductase (FTL_1571) under oxidative stress but not of either the trxA1 or trxA2 genes (22). Further, the results from this study show that oxyR gene expression is significantly downregulated in the ΔtrxA1 mutant when exposed to oxidative stress. These findings were corroborated by significantly downregulated expression of OxyR-regulated ahpC, katG, and sodB genes in the ΔtrxA1 mutant exposed to oxidative stress. The LacZ-oxyR reporter assay further established that, indeed, TrxA1 regulates the expression of OxyR in F. tularensis LVS. Collectively, these results demonstrate a novel finding that contrary to what has been reported for E. coli, TrxA1 controls the oxyR gene expression in F. tularensis.

The role of thioredoxins as transcriptional regulators is not fully established; however, several studies indicate their role in transcriptional regulation. In Rhodobacter, the thioredoxin, TrxC, regulates the expression of photosynthetic genes and ROS generation. The TrxC acts by sensing the oxygen tension. A lower oxygen tension keeps the TrxC in its reduced form, which, upon interaction with the gyrase B subunit of topoisomerase, decreases its supercoiling activity and increases the transcription of genes involved in photosynthesis (44). Further, although a direct connection of thioredoxin A to the transcriptional regulation of type IV pili genes is not established, the loss of trxA is associated with downregulated expression of type IV genes in Acinetobacter baumannii (49, 50). Similarly, downregulation of virulence-associated genes in the thioredoxin mutant in the trxA gene of Listeria has been reported (33). The results from the present study represent a novel and unique role of TrxA1 of F. tularensis LVS in the regulation of oxyR gene expression. The exact molecular mechanisms through which the TrxA1 of F. tularensis regulates the expression of oxyR would require further in-depth investigations.

Thioredoxins, in addition to their role in oxidative stress response, are also required for immune evasion, colonization, and virulence. The TrxA protein of A. baumannii targets and reduces the disulfide bonds of the secretory IgA. The neutralization of secretory IgA facilitates the colonization of A. baumannii on the mucosal surfaces, especially the gastrointestinal tract (51). Similarly, TrxA and TrxC of H. pylori facilitate its colonization by reducing mucin and decreasing the viscosity of the mucus (52, 53). Further, mutants of Francisella novicida, Acinetobacter baumannii, Listeria monocytogenes, and Salmonella in thioredoxin genes are attenuated for virulence (32, 33, 50, 54). In this study, attenuation of intramacrophage growth of the ΔtrxA1 mutant in the wild-type BMDMs and restoration of its growth in gp91phox^−/− BMDMs, demonstrate that TrxA1 contributes to intramacrophage survival by overcoming the intracellular oxidative stress. A majority of the mutants of F. tularensis with attenuated intramacrophage growth also exhibit attenuated virulence in mice (22, 26, 27, 29). The results showing attenuated intramacrophage growth indicate that TrxA1 may also be involved in the virulence of F. tularensis. Further, the conserved sequence of TrxA1 between F. tularensis LVS and F. tularensis SchuS4 also indicates that TrxA1 may have a similar role in oxidative stress resistance and intramacrophage survival of F. tularensis SchuS4.

To conclude, this study describes the role of thioredoxins in F. tularensis LVS. Specifically, the results from this study demonstrate the role of TrxA1 of F. tularensis in oxidative stress resistance and intramacrophage survival. Furthermore, this study reports that TrxA1 of F. tularensis plays a novel regulatory role in the oxidative stress response by controlling the expression of the oxidative stress response regulator, the oxyR gene.

MATERIALS AND METHODS

Bacterial strains and culture conditions.

F. tularensis live vaccine strain (LVS) was obtained from BEI Resources, Manassas, VA. The thioredoxin gene deletion mutants ΔtrxA1 (FTL_0611) and ΔtrxA2 (FTL_1224) of F. tularensis LVS and their respective transcomplemented strains were generated in this study. The F. tularensis ΔoxyR mutant and its transcomplemented strain previously developed in our laboratory were also used in this study (22). F. tularensis LVS and the ΔtrxA1, ΔtrxA2, and ΔoxyR mutants were streaked on Mueller-Hinton (MH)-chocolate agar plates (Hardy Diagnostics). A single colony was picked and grown in modified MH broth (MHB) supplemented with 2% IsoVitaleX (BD), 1% glucose, and 0.25% ferric pyrophosphate until an OD₆₀₀ of 0.5 was achieved. The cultures were aliquoted, snap-frozen in liquid nitrogen, and stored at −80°C until further use. The transcomplemented strains were grown on MH-chocolate agar plates or MHB containing 200 μg/mL hygromycin to retain the plasmid containing the transcomplemented gene. All bacterial strains used in the present study are summarized in Table 1.

TABLE 1.

List of bacterial strains and plasmid vectors used in this study

Strain	Genotype	Source or reference
Francisella strains
F. tularensis LVS	Wild-type strain	BEI Resources
ΔtrxA2	Deletion mutant of F. tularensis LVS trxA2 gene	This study
ΔtrxA1	Deletion mutant of F. tularensis LVS trxA1 gene	This study
ΔtrxA1+ptrxA1	Transcomplement, ΔtrxA1, pMM022(pMP822+trxA1), Hygror	This study
ΔtrxA2+ptrxA2	Transcomplement, ΔtrxA2,pMM021(pMP822+trxA2), Hygror	This study
F. tularensis LVS pOZ	oxyR promoter-lacZ fusion, Kanr reporter	This study
ΔtrxA1 pOZ	ΔtrxA1mutant oxyR promoter-lacZ fusion, Kanr reporter	This study
F. tularensis LVS pCZ	Control fragment-lacZ fusion, Kanr reporter	This study
ΔtrxA1 pCZ	ΔtrxA1 mutant control fragment-lacZ fusion, Kanr reporter	This study
E. coli strain
DH5α	F– Φ80lacZΔM15 Δ(lacZYA-argF) U169 recA1 endA1 hsdR17 (rK–, mK+) phoA supE44 λ– thi-1 gyrA96 relA1	Invitrogen
Plasmids
pMP822	E. coli-Francisella shuttle vector, Hygror	56
pJC84	E. coli-Francisella suicide vector, Kanr	55
pBSK(iglA-lacZ)	Integration vector, Kanr	Simon Dove
pMM020	pJC84 + fused flanking fragment of trx1 gene, Kanr	This study
pMM022	pJC84 + fused flanking fragment of trx gene, Kanr	This study
pMM021	pMP822 + trx1, Hygror	This study
pMM023	pMP822 + trx, Hygror	This study
pMM026	pBSK+ oxyR-lacZ (pOZ), Kanr	This study
pMM027	pBSK+ control fragment-lacZ (pCZ), Kanr	This study

Generation and screening of thioredoxin gene deletion mutants and transcomplemented strains.

Thioredoxin gene deletion mutants of the FTL_0611 (ΔtrxA1) and FTL_1224 (ΔtrxA2) genes were generated by an allelic-replacement method. As previously described, the pJC84 suicide plasmid, which allows for SacB-assisted allelic replacement in F. tularensis LVS, was used (55). Briefly, fragments upstream and downstream of the thioredoxin genes were generated by PCR amplification using the Platinum Pfx DNA polymerase. These fragments were then fused via overlapping extension PCR and cloned into the pJC84 plasmid at the BamHI and SalI sites. Following confirmation of cloning via PCR, the newly constructed plasmids were transformed into F. tularensis by electroporation and plated on MH-chocolate agar plates containing 10 μg/mL kanamycin for selection. For counterselection with sucrose, kanamycin-resistant clones were grown to an OD₆₀₀ of 0.5 in MHB with 5% sucrose. The cultures were incubated for an additional 2 h and serially diluted onto MH-chocolate agar plates containing 8% sucrose and incubated at 37°C with 5% CO₂ for 48 to 72 h. Sucrose-resistant clones were replated onto kanamycin-containing MH-chocolate agar plates to confirm the loss of kanamycin resistance. The sucrose-resistant and kanamycin-sensitive clones were screened for gene deletion by using either gene-specific primers or primers flanking the thioredoxin genes. DNA sequencing was used to confirm in-frame deletions of trxA1 and trxA2 genes.

For the generation of transcomplemented strains, full-length FTL_0611 (trxA1) and FTL_1224 (trxA2) gene sequences were amplified with gene-specific primers using the Pfx polymerase enzyme. The amplified fragments were digested with BamHI and XhoI restriction enzymes and cloned into the E. coli-Francisella shuttle vector pMP822 (56). The resulting plasmids were validated by PCR and DNA sequencing. The plasmids were electroporated into the corresponding gene-deletion mutants and selected on MH-chocolate agar containing 200 μg/mL hygromycin. The transcomplemented strains ΔtrxA1+ptrxA1 and ΔtrxA2+ptrxA2 were confirmed by PCR. The primers and vectors used to generate gene deletion mutants and the transcomplemented strains are summarized in Tables 1 and 2.

TABLE 2.

List of primers used in this study

Primer	Sequence	Purposea
Deletion constructs
trxA2 upstream fragment
MP584	5′-CAAGGATCCCTCTCAGCAAAGCCAAGAAATC-3′	F-primer with a BamHI site
MP585	5′-CAATGTCATAATCTATTTCTCCAT-3′	R-primer
trxA2 downstream fragment
MP586	5′-ATGGAGAAATAGATTATGACATTGGAGTAAATCTAAAGAAACCGATTC-3′	F-primer
MP588	5′-TGATGTCGAC CGATATCATCCCTCTATTAACTCTTATG-3′	R-primer with a SalI site
trxA1 upstream fragment
MP611	5′-CAAGGATCCTGGTCCCAGAAGAACCAAGCACAG-3′	F-primer with a BamHI site
MP612	5′-TGACATGCTTTGACTTTCCT-3′	R-primer
trxA1 downstream fragment
MP613	5′-AGGAAAGTCAAAGCATGTCACTATAATTAGTTATATCAACAT-3′	F-primer
MP614	5′-TGATGTCGACCCACCAGATAGTACACGACC-3′	R-primer with a SalI site
trxA2 mutant screening and verification
MP037	5′-CCGGATCCATGAAATTTGAATTACCAAAAC-3′	F-primer for sodB as control
MP038	5′-CGCTGCAGCTAATCAGCGAATTGCTCAGAAAC-3′	R-primer for sodB as control
MP597	5′-GATACTAAAGATTACCTTGCTAAC-3′	F-primer
MP598	5′-CCATTGATTTG ATATACCCAT-3′	R-primer
MP621	5′-TATGCTGATTGGTGTGGTCC-3′	F-primer
MP622	5′-TTGAGAAGCAGTATGAACGC-3′	R-primer
trxA1 mutant screening and verification
MP615	5′-CTTATTCCTTAACTACCCCC-3′	F-primer
MP616	5′-CCTTCACCATAAATGTCTTCACC-3′	R-primer
MP623	5′-GTTAGATTTTTGGGCACCATG-3′	F-primer
MP624	5′-CGCTGTCTCTTCATTATCATC-3′	R-primer
Complementation plasmids
ThioredoxinA2 (trxA2)
MP601	5′-CAAGGATCCATGACATTGAGTAATGTTATAA-3′	F-primer with a BamHI site
MP602	5′-CGACCTGCAGTTACTCATATGCTTTTAGTTTCTG-3′	R-primer with a PstI site
ThioredoxinA1 (trxA1)
MP619	5′-CAAGGATCCATGTCAAAATGTATCGATATATCAG-3′	F-primer with a BamHI site
MP620	5′-CGACCTGCAGTTATAGATATTTGTCTACGATTGA-3′	R-primer with a PstI site
LacZ fusion constructs
poxyR-lacZ
MP643	5′-TAGAGCGGCCGCTGAGAGTCTTGGATATCTACAGCA-3′	F-primer with a NotI site
MP644	5′-CGATTAATTAATTTTGTCTCATACACAGAAATG-3′	R-primer with a PacI site
pcontrol-lacZ
MP645	5′-TAGAGCGGCCGCAAATGGCGCATATTTATCAAGAGC-3′	F-primer with a NotI site
MP646	5′-CGATTAATTAAtCAAAGAGCTAGGTATAGATGAGA-3′	R-primer with a PacI site
Transcriptional analysis
tul4 (internal control)
MP029	5′-TCGCAGGTTTAGCGAGCTGTTCTA-3′	F-primer
MP030	5′-ACAGCAGCAGCTTGCTCAGTAGTA-3′	R-primer
Peroxidase/catalase (FTL_1504, katG)
MP077	5′-CCTGCCAAATAAAGTTTTGCTC-3′	F-primer
MP078	5′-AGCTCACCAATGGACTCCTAC-3′	R-primer
Superoxide dismutase [Fe](FTL_1791, sodB)
MP101	5′-GGCGGAATATTTAATAACGCTGC-3′	F-primer
MP102	5′-GTGCTCCCAAACATCAAAAG-3′	R-primer
AhpC/TSA family protein (FTL_1015, ahpC)
MP258	5′-TTGTATTCTCATTACCAGGAGCA-3′	F-primer
MP259	5′-ACAATCATTGCATAGCGCCA-3′	R-primer
Oxidative stress transcriptional regulator (FTL_1014, oxyR)
MP528	5′-ATGCCTAAAATTCTCCCTGC-3′	F-primer
MP529	5′-GACCTTCATCAAGAAGCATAAGA-3′	R-primer
Thioredoxin 1 (FTL_1224, trx1)
MP286	5′-CAGACGAAGCTAATTTTGACAAAC-3′	F-primer
MP287	5′-AGTCTCAAGTTGCTTACGATTTTT-3′	R-primer
Glutaredoxin 2 (FTL_092, grx2)
MP534	5′-TAGAATTGGCTCAAAACGAGTACC-3′	F-primer
MP535	5′-CTAAGTAAAGCATCGAAATCACCA-3′	R-primer
Glutaredoxin 1 (FTL_0985, grx1)
MP538	5′-GTTGTCCATATTGTGTTTGGGC-3′	F-primer
MP539	5′-CCTTAAGCTCTGTAAAGCCACCTA-3′	R-primer
Glutathione peroxidase (FTL_1383, gpx)
MP544	5′-GTAAGTGCGGTTTTACTAAGCAG-3′	F-primer
MP545	5′-TGGCTCAGCATCTTTACCAT-3′	R-primer
Glutaredoxin 3/ribonucleoside-diphosphate reductase, beta subunit (FTL_0984, grx3)
MP633	5′-TGCTTGGTTCATTTGCAGCTCG-3′	F-primer
MP634	5′-CGACGCGTAGTAGGGTCTGC-3′	F-primer
Ribonucleoside-diphosphate reductase, alpha subunit (FTL_0986)
MP635	5′-CGTTACCGTACAGGTGAACCAT-3′	F-primer
MP636	5′-TTGAACCTTTGATTGTCAAGCCCA-3′	R-primer

^aF, forward; R, reverse.

Disc diffusion assays and growth curves.

For disc diffusion assays, wild-type F. tularensis LVS, ΔtrxA1, ΔtrxA1+ptrxA1, ΔtrxA2, and ΔtrxA2+ptrxA2 strains were grown in MHB until an OD₆₀₀ of ~0.2 was attained for each strain. Then, 200 μL of each culture was spread onto MH-chocolate agar plates (Hardy Diagnostics), and sterile paper discs were placed. Each disc was impregnated with 10 μL of paraquat (50 μg/disc), pyrogallol (62.5 μg/disc), cumene hydroperoxide (125 μg/disc), t-butyl-hydroperoxide (1.4 μg/disc), and diamide (50 mM/disc). The plates were incubated at 37°C in the presence of 5% CO₂ for 48 to 72 h or until bacterial lawns were fully grown. Zones of inhibition were measured using a digital caliper. The average zone of inhibition ± the standard deviation was recorded.

The growth curves were generated in the absence and presence of H₂O₂ to determine the ability of the wild-type bacteria, the mutants, and the transcomplemented strains to withstand oxidative stress induced by H₂O_2. Bacterial cultures grown on MH-chocolate agar plates were adjusted to an OD₆₀₀ of 0.05 in MHB, and 1 mM H₂O₂ was added. Untreated cultures were kept as controls. The cultures were incubated at 37°C with constant shaking. Growth was determined by recording OD₆₀₀ values at regular intervals over 28 h.

Bacterial killing assay.

The sensitivity of the wild-type F. tularensis LVS, ΔtrxA1, ΔtrxA1+ptrxA1, ΔtrxA2, and ΔtrxA2+ptrxA2 strains was determined by bacterial killing assays in the presence or the absence of oxidants as described previously (22, 27, 57, 58). F. tularensis LVS, ΔtrxA1, ΔtrxA1+ptrxA1, ΔtrxA2, and ΔtrxA2+ptrxA2 strains were adjusted to 0.2 OD₆₀₀ (equivalent to 1 × 10⁹ CFU/mL). The bacterial suspensions were exposed to 1 mM H₂O₂, paraquat, pyrogallol, and diamide (5 μM) for 1 and 3 h at 37°C and 5% CO₂ with constant shaking at 180 rpm. Bacteria were diluted 10-fold, plated on MH-chocolate agar plates and incubated for 48 h at 37°C and 5% CO₂. Bacterial colonies were counted and plotted as log₁₀ CFU/mL.

Western blot analysis.

The Western blot analysis was performed as described earlier (22). F. tularensis LVS, ΔtrxA1, ΔtrxA1+ptrxA1, ΔtrxA2, and ΔtrxA2+ptrxA2 strains were grown to 0.5 OD₆₀₀. The bacterial suspensions were exposed to 1 mM, 250 uM or 500 uM H₂O₂ for 2 h and centrifuged at 4,000 × g for 10 min. The bacterial cell pellets were resuspended in 200 μL lysis buffer [200 mM Tris-HCl, pH 8.0; 320 mM (NH₄)₂SO₄; 5 mM MgCl₂; 10 mM EDTA; 10 mM EGTA; 20% glycerol; 1 mM dithiothreitol (DTT); protease and phosphatase inhibitors]. The protein concentrations of the cell lysates were determined with a bicinchoninic acid (BCA) kit; 5 μg of protein from each sample was run on a 10% SDS-PAGE gel, transferred to a polyvinylidene difluoride membrane (Millipore), and probed with anti-KatG antibodies and secondary anti-rabbit IgG antibodies conjugated to horseradish peroxidase. The protein bands on the membrane were visualized using Supersignal West Pico chemiluminescent substrate (Thermo Scientific) on a Chemidoc XRS system (Bio-Rad). The blots were stripped and reprobed with anti-SodB antibodies, and the protein bands were imaged as described above.

Macrophage cell culture assay.

Bone marrow-derived macrophages (BMDMs) were derived from the wild-type C57BL/6 and gp91phox^−/− mice (Jackson Laboratories) using the standard protocol (59). The BMDMs were seeded at a concentration of 1 × 10⁶ cells/mL in a 12-well plate. Bacterial strains were added to the cells at a multiplicity of infection (MOI) of 100 and incubated at 37°C in the presence of 5% CO₂. After 2 h, 100 μg/mL gentamicin was added to the wells to kill extracellular bacteria. Cells were then incubated for a total of 4 or 24 h in an antibiotic-free medium. Macrophages were lysed using 0.01% sodium deoxycholate, and bacteria were diluted 10-fold in sterile phosphate-buffered saline (PBS) and plated on MH-chocolate agar plates. After a 48-h incubation, bacteria were enumerated by counting the colonies. The results were expressed as log₁₀ CFU/mL.

Quantitative reverse transcription-PCR (qRT-PCR).

qRT-PCR was used to assess gene expression. Bacteria were grown to an OD₆₀₀ of 0.5 in MHB and were either left untreated or subjected to 1 mM H₂O₂ treatment for 2 h. In another experiment, bacterial strains were also treated with 250 and 500 μM H₂O₂. RNA was isolated from the bacteria using the Thermo Fisher PureLink RNA minikit. cDNA was synthesized using the Bio-Rad iScript cDNA synthesis kit. qPCR was then performed using the iQ SYBR green supermix from Bio-Rad, using primers for antioxidant genes shown in Table 2. The fold change in gene expression was calculated using the ΔΔCT method (60). Results were normalized to the tul4 gene of F. tularensis LVS as described previously (29, 30). The primer sequences are summarized in Table 2.

Generation of the oxyR-lacZ reporter strains of F. tularensis LVS and the ΔtrxA1 mutant.

The promoter region of F. tularensis LVS oxyR (FTL_1014) and a control fragment located inside the ahpC open reading frame (ORF) (FTL_1015) were amplified from F. tularensis LVS genomic DNA by PCR using the pfx polymerase enzyme. The resulting PCR products were cloned into the pBSK iglA-lacZ plasmid (kindly provided by Simon Dove, Boston Children’s Hospital) (61) at PacI and NotI sites to replace the iglA promoter, resulting in plasmids pCZ (control-lacZ) and pOZ (oxyR-lacZ). Both plasmids were introduced into F. tularensis LVS and the Δtrx mutant by electroporation. The strains with the plasmids integrated into their genomes were selected and designated LVS pCZ (control-lacZ), Δtrx pCZ, LVS pOZ (oxyR-lacZ), and Δtrx pOZ, respectively. The primer sequences are summarized in Table 2.

β-galactosidase activity assay.

A β-galactosidase activity assay was used to determine the transcription of oxyR using a previously described protocol (62). Briefly, the control-lacZ and oxyR-lacZ integrated strains of F. tularensis LVS and the Δtrx mutant were grown in 5 mL MHB for 3 h at 37°C in the absence or presence of 1 mM H₂O₂ at a starting OD₆₀₀ of 0.2. Bacterial cells were added to the reaction buffer and permeabilized with 0.1% sodium dodecyl sulfate (SDS) and chloroform for 10 min. Following the permeabilization, ortho-nitrophenyl-β-galactoside (ONPG) was added, and tubes were incubated for 2 h. Stop solution of 1 molar (M) sodium carbonate (Na₂CO₃) was added, and the OD₄₂₀ and OD₅₅₀ were measured. The results were expressed as β-galactosidase Miller units.

Statistical analysis.

All data are presented as the mean ± standard deviation (SD) or mean ± standard error of the mean (SEM). All statistical analyses were performed using a one-way analysis of variance (ANOVA) using GraphPad Prism version 7.