Role of peroxiredoxin of the AhpC/TSA family in antioxidant defense mechanisms of Francisella tularensis

Francisella tularensis is a Gram-negative, facultative intracellular pathogen and the causative agent of a lethal human disease known as tularemia. Due to its extremely high virulence and potential to be used as a bioterror agent, F. tularensis is classified by the CDC as a Category A Select Agent. As an intracellular pathogen, F. tularensis during its intracellular residence encounters a number of oxidative and nitrosative stresses. The roles of the primary antioxidant enzymes SodB, SodC and KatG in oxidative stress resistance and virulence of F. tularensis live vaccine strain (LVS) have been characterized in previous studies. However, very fragmentary information is available regarding the role of peroxiredoxin of the AhpC/TSA family (annotated as AhpC) of F. tularensis SchuS4; whereas the role of AhpC of F. tularensis LVS in tularemia pathogenesis is not known. This study was undertaken to exhaustively investigate the role of AhpC in oxidative stress resistance of F. tularensis LVS and SchuS4. We report that AhpC of F. tularensis LVS confers resistance against a wide range of reactive oxygen and nitrogen species, and serves as a virulence factor. In highly virulent F. tularensis SchuS4 strain, AhpC serves as a key antioxidant enzyme and contributes to its robust oxidative and nitrosative stress resistance, and intramacrophage survival. We also demonstrate that there is functional redundancy among primary antioxidant enzymes AhpC, SodC, and KatG of F. tularensis SchuS4. Collectively, this study highlights the differences in antioxidant defense mechanisms of F. tularensis LVS and SchuS4.


Introduction
Francisella tularensis is a Gram-negative, facultative intracellular pathogen and the causative agent of a lethal human disease known as tularemia. F. tularensis has a very broad host range and can infect a wide range of ticks, arthropods, and mammals [1]. F. tularensis subsp. tularensis (Type A) cause lethal tularemia in North America. The strains belonging to F. tularensis subsp. holarctica (Type B) are less infectious than the Type A strains and are prevalent a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 peroxidase/peroxireducatse proteins and has been annotated as peroxiredoxin of the AhpC/ TSA family. It has been reported that AhpC in F. tularensis SchuS4 is required for resistance against endogenous H 2 O 2 and ONOO - [10]. However, very fragmentary information is available regarding the role of AhpC of F. tularensis SchuS4 in the pathogenesis of tularemia [10,16]; whereas the role of AhpC of F. tularensis LVS is not known. This study was undertaken to exhaustively investigate the role of AhpC in oxidative stress resistance of F. tularensis LVS and SchuS4. We report that AhpC of F. tularensis LVS confers resistance against a wide range of ROS and RNS, and serves as a virulence factor. This study also demonstrates that there is a functional redundancy among primary antioxidant enzymes AhpC, KatG and SodC of F. tularensis SchuS4. However, AhpC serves as a key antioxidant enzyme and contributes to robust oxidative and nitrosative stress resistance and intramacrophage survival of the highly virulent F. tularensis SchuS4 strain.

Ethics statement
This study was carried out in strict accordance with the recommendations and guidelines of the National Council for Research (NCR) for care and use of animals. All the animal experiments were conducted in the centralized Animal Resources Facility of New York Medical College licensed by the USDA and the NYS Department of Health, Division of Laboratories and Research and accredited by the American Association for the Accreditation of Laboratory Care. The use of animals and protocols were approved by the Institutional Animal Care and Use Committee (IACUC) of New York Medical College (Protocol Number 69-2-0914H). Mice were administered an anesthetic cocktail consisting of ketamine (5 mg/kg) and xylazine (4 mg/ kg) and underwent experimental manipulation only after they failed to exhibit a toe pinch reflex. Mice exhibiting more than 25% weight loss, anorexia, dehydration and impairment of mobility were removed from the study and euthanized by approved means. Humane endpoints were also necessary for mice which survived at the conclusion of the experiments. Mice were administered an anesthetic cocktail of ketamine and xylazine intraperitoneally and then euthanized via cervical dislocation followed by cardiac puncture, a method that is consistent with recommendations of the Panel on Euthanasia of the American Veterinary Medical Association. In all experimental procedures, efforts were made to minimize pain and suffering. All the work with Category A select agent F. tularensis SchuS4 was performed in CDC Certified Biosafety Level 3 (BSL3) laboratory of New York Medical College (Registration No. C20160722-1812) in accordance with protocols approved by Institutional Biosafety Committee (Protocol No. 01-2015-3).

Bacterial strains and growth conditions
F. tularensis subspecies holarctica LVS and F. tularensis subspecies tularensis SchuS4 used in this study were obtained from BEI Resources (Manassas, VA). The ahpC (FTL_1015) gene deletion (ΔahpC) mutant of F. tularensis LVS and a transcomplemented strain (ΔahpC+-pahpC) were generated and used in this study. Previously published ΔsodC mutant of F. tularensis LVS available in our laboratory was also used in this study (8). The gene deletion mutants of F. tularensis SchuS4; ΔahpC (FTT_0557), ΔkatG (FTT_0721c) and the ΔsodC (FTT_0879), and F. tularensis LVS ΔkatG (11) mutants were kindly provided by Dr. Andres Sjostedt (Umea University, Sweden). All the bacterial strains used in this study are shown in Table 1. All the experiments involving F. tularensis SchuS4 strain were conducted in the CDC certified BSL3 laboratory of New York Medical College.
All bacterial strains were grown on Mueller-Hinton (MH)-chocolate agar plates (BD Biosciences, San Jose, CA) at 37˚C with 5% CO 2 or Muller-Hinton broth (MHB) (BD Biosciences, San Jose, CA) supplemented with IsoVitaleX and ferric pyrophosphate at 37˚C with constant shaking (175 rpm). Transcomplemented ΔahpC+pahpC strain of F. tularensis LVS was grown on MH-chocolate agar plates supplemented with hygromycin (200μg/mL). Bacterial strains were grown in MHB to mid-log phase, aliquoted and stored at -80˚C until further use.

Construction of ΔahpC mutant and transcomplementation
Allelic replacement method was used to construct the ΔahpC mutant of F. tularensis LVS [18]. The entire 557-bp coding region of the ahpC gene (FTL_1015) was deleted employing an approach described previously [19]. Briefly, a 5' 1218 bp fragment upstream of the start codon and first 5 bp of the ahpC gene was amplified with primers MP241 and 243. A 3' fragment containing last 10 bp and the stop codon of the ahpC gene and 1218 bp of the downstream region was amplified with primers MP245 and 246. Both the upstream and downstream fragments were joined by overlapping extension PCR with primers MP241 and MP246 engineered with BamHI and SalI restriction sites at 5' and 3' ends, respectively. The generated single fragment with ahpC gene deletion was digested and cloned into the pJC84 vector using the BamHI and SalI sites. The resultant plasmid, pMM06, was electroporated into the wild type F. tularensis LVS as described previously [14,20]. After the primary selection of positive colonies using kanamycin and a counter selection with sucrose, the positive colonies were screened by colony PCR with primers MP260 and MP261 to identify the ΔahpC mutant.
For transcomplementation of the ΔahpC mutant of F. tularensis LVS, full-length ahpC gene sequence was amplified with primers MP274 and MP275 and cloned into a pMP822 vector at BamHI site generating a plasmid, pMM09. The pMM09 plasmid was transformed into chemically competent E. coli DH5α cells and selected on LB-hygromycin plates. The pMM09 was purified, and the orientation of the ahpC gene in the pMM09 vector was confirmed by PCR.  Table 2.

Growth curves
Growth curves were generated by resuspending bacterial cultures grown on MH-chocolate agar plates to an Optical Density at 600 nm (OD 600 ) of 0.2 (corresponds to 1×10 9 CFU/mL) in MHB. The bacterial suspensions were grown for 28 hours in the absence or presence of 750μM H 2 O 2, and the OD 600 was recorded at 4-hour intervals. ). An identical protocol was used for disc diffusion assays performed with wild type F. tularensis SchuS4 and the ΔahpC, ΔsodC, and ΔkatG mutants. However, higher concentrations of oxidants than those used for F. tularensis LVS were used. Specifically, the concentrations of menadione (6.25μg/disc), paraquat (15μg/disc) and pyrogallol (250 and 500μg/ ΔahpC, ΔsodC and ΔkatG mutants were exposed to 2-fold diluted menadione (Starting concentration 12.5 μg), TBH (875 μg), CHP (62.5 μg), SNP (15.7μg) and Sin-1 (12.5μg) to test their sensitivities towards these compounds. The plates were incubated at 37˚C in the presence of 5% CO 2 for 1 and 3 hours post-exposure, and 3μL bacterial cultures from each dilution were spotted on MH-chocolate agar plates using a multichannel pipette. The sensitivity to the compounds tested was determined on the basis of observable growth pattern on the plates after 48 hours of incubation.

Cell culture assays
A murine macrophage cell line Raw264.7 was used in cell culture-based assays. The macrophages were infected with the wild-type F. tularensis LVS, the ΔahpC mutant, and the ΔahpC +pahpC transcomplemented strain at a multiplicity of infection (MOI) of 10 and 100 in a volume of 1 mL bacterial suspension. In separate experiments, Raw264.7 macrophages were infected with the wild-type F. tularensis SchuS4, the ΔahpC, ΔsodC, and ΔkatG mutants at an MOI of 100 as described previously [5,14]. The infected cells were lysed after 4 and 24 hours of infection with 0.1% sodium deoxycholate, diluted 10-fold in sterile PBS and plated on MHchocolate agar plates. The plates were incubated at 37˚C in the presence of 5% of CO 2 for 48 hours, and the colonies were counted. Results were expressed as Mean ± SD of three biological replicates and presented as Log 10 colony forming units (CFU)/mL.

Mouse challenge studies
All mice studies followed the protocols approved by the Institutional Animal Care and Use Committee (IACUC) of New York Medical College. Six to eight-week-old wild type C57BL/6 and gp91phox -/mice were obtained from Jackson Laboratories. Mice were maintained in a pathogen-free environment in the Animal Facility of New York Medical College (Valhalla, NY) Mice (n = 4 mice/group) were deeply anesthetized by intraperitoneal injection of Ketamine/Xylazine cocktail. The loss of reflexes in anesthetized mice was determined by the loss of toe-pinch reflex. The deeply anesthetized mice were inoculated intranasally with 1x10 4 CFU of the wild-type F. tularensis LVS strain or the ΔahpC mutant resuspended in 20μL PBS (10μL/ nare). The infected mice were observed for morbidity and mortality for 21 days. The survival results were plotted as Kaplan-Meier survival curves, and the data were analyzed statistically by the Log-rank test.

Statistical analysis
Statistical analysis was performed by using GraphPad Prism and InStat software. The results were expressed as Mean ± S.E.M. or S.D., and statistical significance between groups was determined by one-way ANOVA followed by Bonferroni's corrections or student t-test. As detailed earlier, the survival results were expressed as Kaplan-Meier survival curves, and P values were determined by the Log-rank test.

Results
The peroxiredoxin of the AhpC/TSA family (ahpC) gene in F. tularensis LVS and SchuS4 is transcribed divergently from the LysR family oxidative stress transcriptional regulator gene, oxyR. A similar genomic organization of ahpC gene is also present in Mycobacterium tuberculosis. However, in other bacterial pathogens including Yersinia pestis, the ahpC gene is not transcribed divergently from the oxyR gene ( Fig 1A). To characterize the functional role of the peroxiredoxin AhpC of F. tularensis LVS, we generated a gene deletion mutant of ahpC (ΔahpC). The deletion of the ahpC gene was confirmed by PCR followed by DNA sequencing to determine that ahpC gene deletion did not alter reading frames of the downstream genes. A transcomplement of the ΔahpC mutant was generated by providing a copy of ahpC gene intrans. Transcomplementation was confirmed by PCR using ahpC gene-specific primers. The ΔahpC mutant was tested for any growth defect under aerobic growth conditions. It was observed that growth pattern of the ΔahpC mutant was identical to that of the wild type F. tularensis LVS or the transcomplemented strain when grown aerobically indicating that the loss of ahpC is not associated with any growth defect in the ΔahpC mutant ( Fig 1B). The ΔahpC mutant of F. tularensis LVS exhibits enhanced sensitivities towards superoxide-generating compounds The contribution of AhpC of F. tularensis LVS in conferring resistance to superoxide-generating compounds menadione, pyrogallol and paraquat were determined by disc diffusion and spot assays. The ΔahpC mutant of F. tularensis LVS revealed enhanced sensitivities towards superoxide-generating compounds as indicated by significantly larger zones of inhibition around the discs impregnated with menadione (21.6±1.4 mm), pyrogallol (11.6±0.6 mm) and paraquat (26.0±1.0 mm) as compared to those observed for wild type F. tularensis LVS (18.0 ±0.2, 10.3±0.6, 23.0±1.0 mm, respectively) and the transcomplemented strain (17.1±0.6, 10.0 ±0.0, 25.6±1.5 mm, respectively) (Fig 2A, 2B and 2C). Similar to the ΔahpC mutant, the ΔsodC and ΔkatG mutants of F. tularensis LVS also exhibited increased susceptibility towards menadione (S1A and S1B Fig).
We next confirmed the results obtained with the disc diffusion assays by performing spot assays that determine the bacterial viability. Wild type F. tularensis LVS, the ΔahpC mutant or the transcomplemented strains were exposed to varying concentrations of two-fold serial dilutions of menadione, pyrogallol and paraquat for 1 and 3 hours, and plated to determine the bacterial viability. Reduced viability of the ΔahpC mutant was observed after 1 and 3 hours of exposure to menadione (31.25 and 15.62μg, respectively), pyrogallol (38.8 and 19.4μg, respectively) and paraquat (0.30 and 0.15μg, respectively) as compared to the wild type F. tularensis LVS corroborating the results observed with the disc diffusion assays. The transcomplementation either restored the wild type phenotype or exhibited an intermediate phenotype (Fig 2D,  2E and 2F). Collectively, these results demonstrate that loss of AhpC in F. tularensis LVS is associated with enhanced sensitivities towards the superoxide-generating compounds.
We confirmed the results obtained with the disc diffusion assays by performing spot assays and by generating growth curves in the presence of H 2 O 2 . The wild type F. tularensis LVS, the ΔahpC mutant or the transcomplemented strains were exposed to varying concentrations of serially diluted TBH, CHP, and H 2 O 2 for 1 and 3 hrs and plated to determine the bacterial viability. The viability of the ΔahpC mutant of F. tularensis LVS was markedly reduced after 1 and 3 hours of exposure to TBH (1.1μg), CHP (13.7 and 0.9μg, respectively) and H 2 O 2 (0.2 and 0.05mM, respectively) as compared to the wild-type F. tularensis LVS. Transcomplementation of the ΔahpC mutant restored the wild type phenotype (Fig 3D, 3E and 3F). The ΔahpC mutant grew very slowly as compared to the wild type or the transcomplemented counterparts when grown in the presence of 750μM of H 2 O 2 ( Fig 3G). Collectively, these results demonstrate that AhpC of F. tularensis LVS plays an important role in providing resistance against organic peroxides and H 2 O 2 .

The ΔahpC mutant of F. tularensis LVS exhibits enhanced sensitivity towards RNS
Our preceding results demonstrated that AhpC of F. tularensis LVS provides resistance against superoxide-generating compounds and peroxides. We further tested the role of AhpC in providing resistance against RNS by using nitric oxide (NO) donor sodium nitroprusside (SNP) and Sin-1. Wild type F. tularensis LVS, the ΔahpC mutant or the transcomplemented strains were exposed to varying concentrations of serially diluted SNP and Sin-1 for 1 and 3 hours and plated to determine the bacterial viability. The ΔahpC mutant of F. tularensis LVS was found to be highly sensitive to both SNP (93.8 μg) and Sin-1 (0.1 and 0.05μg, respectively) as evidenced by marked reduction in viability after 1 and 3 hours of exposure to these compounds as compared to the wild type or the transcomplemented strain (Fig 4A and 4B). These results demonstrate that AhpC of F. tularensis LVS also plays an important role in providing resistance against RNS. These results indicate that ahpC is not required for intramacrophage survival of F. tularensis LVS (Fig 5A).
We next examined the contribution of AhpC of F. tularensis LVS in virulence in mice. Since our preceding results indicated that the ΔahpC mutant is highly sensitive to ROS, we also determined the contribution of NADPH oxidase-dependent ROS in clearance of ΔahpC mutant of F. tularensis LVS by infecting Phox -/mice. These mice are defective in ROS generation. Wild-type C57BL/6 and Phox -/mice were infected intranasally with 1×10 4 CFUs of to serially diluted TBH, CHP, and H 2 O 2 for 1 and 3 hours and spotted on MH-chocolate agar plates to determine the bacterial killing. The red arrows indicate enhanced killing of the ΔahpC mutant at the indicated concentrations of the compounds. (G) Growth curves of F. tularensis LVS, the ΔahpC mutant and the transcomplemented strain (ΔahpC + pahpC) in the absence or presence of 750μM H 2 O 2 . Equal numbers of bacteria were suspended in Mueller-Hinton broth and the optical density (OD 600 ) was recorded every 4 hours. All the results shown are representative of 3 independent experiments conducted with identical results. The p values were determined by oneway ANOVA and a p value of <0.05 is considered statistically significant. � p<0.05; ��� p<0.001.
https://doi.org/10.1371/journal.pone.0213699.g003 either the wild-type F. tularensis LVS or the ΔahpC mutant and observed for mortality for 21 days. 100% of wild type C57BL/6 mice infected with the ΔahpC mutant survived the infection; while mice infected with similar doses of the wild type F. tularensis LVS succumbed to infection by day 8 post-infection, indicating that AhpC is required for virulence. On the other hand, 100% of Phox -/mice infected either with F. tularensis LVS, or the ΔahpC mutant succumbed to infection indicating that NADPH-oxidase induced ROS is required for clearance of the ΔahpC mutant (Fig 5B).

AhpC of F. tularensis SchuS4 is a major antioxidant enzyme that protects against oxidative stress induced by superoxide-generating compounds
Previous studies conducted with mutants of F. tularensis LVS deficient in SodB, SodC, or KatG have reported that loss of only one antioxidant enzyme results in an enhanced sensitivity of F. tularensis LVS to oxidative stress, attenuated intramacrophage growth and virulence in mice [7,8,11]. The results obtained in this study with the ΔahpC mutant of F. tularensis LVS also support this notion. On the contrary, the reported phenotype of the SchuS4 ΔkatG mutant is quite different from that reported for the corresponding mutant of F. tularensis LVS [11]. Moreover, unlike F. tularensis LVS mutants, the ΔkatG, ΔsodC and ΔahpC mutants of F. tularensis SchuS4 retain their virulence in mice [10,11,16]. We next investigated to establish if AhpC is one of the major antioxidant enzymes of F. tularensis SchuS4 by determining the sensitivities of the ΔahpC, ΔsodC and ΔkatG mutants of F. tularensis SchuS4 to oxidants and RNS.
Exposure of F. tularensis SchuS4, the ΔahpC, ΔsodC and the ΔkatG mutants to the superoxide-generating compound menadione revealed that the ΔahpC mutant was extremely sensitive to menadione as evident by significantly enlarged zone of inhibition (25.3±1.1 mm) as compared to the wild type F. tularensis SchuS4, ΔsodC and ΔkatG mutants (6.0±0.0 mm for all the three strains, respectively). No differences in sensitivity towards menadione were observed between the wild type F. tularensis SchuS4 or the ΔsodC and the ΔkatG mutants (Fig 6A). We further confirmed these findings by performing spot-and bacterial killing assays. Results from the spot assays ( Fig 6B) demonstrated that exposure to increasing concentrations of menadione resulted in reduced viability of the ΔahpC mutant as compared to wild type F. tularensis SchuS4, or the ΔkatG mutant. In another approach, equal numbers of wild type F. tularensis SchuS4 and the ΔahpC mutant were exposed to menadione (6.25μg/mL) for 1 and 4 hours, diluted 10-fold, and the bacterial killing was determined. The results demonstrated that after 1-hour post-treatment with menadione, significantly lower numbers of the ΔahpC mutant bacteria (4.7±0.1 Log 10 CFU/mL) survived as compared to the wild type F. tularensis SchuS4 strain (6.7±0.3 Log 10 CFU/mL). After 4 hours of treatment, no colonies of the ΔahpC mutant were recovered, while the viability of the wild type F. tularensis SchuS4 was only reduced by 10-fold (5.8±0.1 Log 10 CFU/mL). The viability of both F. tularensis SchuS4 and the ΔahpC mutant were not affected in the PBS control or exposure to the volume of ethanol that was used to resuspend menadione (Fig 6C).
Exposure to paraquat resulted in a significantly larger zone of inhibition for the ΔahpC mutant (31.67 ± 1.53 mm) as compared to the wild type F. tularensis SchuS4 (26.3±0.5 mm). However, treatment of ΔsodC (26.0±1.0 mm) and ΔkatG (28.0±1.0 mm) mutant strains with paraquat did not show any enhanced sensitivity as compared with the wild type F. tularensis SchuS4 (Fig 6D). Disc diffusion assays using pyrogallol (250 and 500μg/disc) displayed similar results, with ΔahpC mutant strain showing a significantly enlarged zone of inhibition (21.6 ±1.5 and 27.3±1.5 mm, respectively) as compared to the wild type F. tularensis SchuS4 strain (17.6±1.1 and 21.3±0.5 mm, respectively). Further, similar to paraquat, the ΔsodC (18.6±0.5 and 23.0±2.0 mm, respectively) and ΔkatG (16.3±1.5 and 21.3±0.5 mm, respectively) mutant strains did not show any increased sensitivity to pyrogallol when compared with the wild type F. tularensis SchuS4 (Fig 6E). Collectively, these results indicate that AhpC of F. tularensis SchuS4 is primarily responsible for providing resistance against oxidative stress induced by superoxide radicals. These results also demonstrate that both the SodC and KatG are dispensable, as the loss of these antioxidant enzymes do not alter the sensitivities of the ΔsodC and ΔkatG mutants to superoxide-generating compounds and remain similar to the wild type F. tularensis SchuS4 strain.

AhpC of F. tularensis SchuS4 protects against oxidative stress induced by peroxides
Disc diffusion assays using peroxides TBH, CHP and H 2 O 2 exhibited results similar to those observed following treatment with superoxide-generating compounds. Exposure of ΔahpC mutant to 3.5mg/disc of TBH demonstrated a significantly larger zone of inhibition (28.00 ± 2.0 mm) as compared to the wild type F. tularensis SchuS4 strain (9.3±1.1mm) ( Fig  7A). However, the ΔsodC mutant strain (6.0±0.0mm) was observed to be more resistant to TBH than the wild type SchuS4 strain. The susceptibility of the ΔkatG mutant (10.0±0.0 mm) to TBH treatment remained similar to that observed for the wild type F. tularensis SchuS4. The spot assay demonstrated similar results as observed for the disc diffusion assays; the ΔahpC mutant was more sensitive to increasing concentrations of TBH than the wild type F. tularensis SchuS4 strain. However, the sensitivity of the ΔkatG mutant remained similar to that observed for the wild type F. tularensis SchuS4 strain (Fig 7B). AhpC of F. tularensis SchuS4 is a major antioxidant enzyme that protects against oxidative stress induced by superoxide generating compounds. The sensitivities of the wild type F. tularensis (Ft) SchuS4, the ΔahpC, ΔsodC and the ΔkatG mutants of SchuS4 as determined by disc diffusion (A), spot assay (B) and bacterial killing assay (C) against superoxide-generating compound, menadione. The sensitivity of the indicated strains against paraquat (D) and pyrogallol (E) was determined using the indicated concentration of the compounds by disc diffusion assay. For the disc diffusion assays, the results are expressed as a zone of inhibition in millimeters and are expressed as Mean ± S.D. The red arrows in (B) indicate enhanced killing of the SchuS4 ΔahpC mutant at the indicated concentrations of menadione. For bacterial killing assay (C) indicated bacterial strains were exposed to menadione (6.25μg/mL) and the bacterial numbers were enumerated after 1 and 4 hours of exposure. PBS, and ethanol required for suspension of menadione were used as controls. The data shown are representative of 2 independent experiments each conducted with 3 biological replicates and were analyzed by one-way ANOVA. �� P<0.01; ��� P<0.001. https://doi.org/10.1371/journal.pone.0213699.g006 Exposure of wild type F. tularensis SchuS4 and the ΔahpC, ΔsodC and the ΔkatG mutants to 500μg/disc of CHP demonstrated that the ΔahpC mutant was significantly more sensitive to the compound (29.0 ± 1.0 mm) as compared to the wild type F. tularensis SchuS4 (19.3 ± 1.1 mm). The sensitivity of the ΔkatG mutant to CHP (16.6±1.5 mm) remained similar to that observed for the wild type F. tularensis SchuS4. On the other hand, similar to that observed for TBH, the ΔsodC mutant was also more resistant to CHP (14.3±0.6mm) as compared to the wild type F. tularensis SchuS4 strain (Fig 7C) as determined by the disc diffusion assays as well as by spot assay (Fig 7D). The ΔahpC mutant demonstrated higher sensitivity to 50mM/disc of H 2 O 2 as indicated by a greater zone of inhibition (28.00±1.4 mm) compared to the wild type F. tularensis SchuS4 (20.0±0.0 mm) (Fig 7E), whereas, the sensitivities of the ΔsodC and ΔkatG mutants remained similar to those observed for the wild type SchuS4 strain (20.5±0.7 and 21.0 ±1.1 mm, respectively). Collectively, these results indicate that the requirement of AhpC for resistance against oxidative stress induced by superoxide radicals and peroxides.

Exposure to NO-generating compounds results in the enhanced killing of ΔahpC mutant of F. tularensis SchuS4
We next investigated the role of F. tularensis SchuS4 antioxidants in providing resistance to RNS. Results of this assay demonstrated that the ΔahpC mutant was highly sensitive to increasing concentrations of SNP and SIN-1 (Fig 8A and 8B) as compared to the wild type F. tularensis SchuS4 or the ΔkatG mutant. However, the ΔsodC mutant showed enhanced resistance to SNP as compared to the wild type F. tularensis SchuS4 strain (Fig 8A). These results demonstrate that AhpC in addition to ROS also protects F. tularensis SchuS4 against RNS.

The ΔahpC mutant of F. tularensis SchuS4 is attenuated for intramacrophage growth
To determine the role of F. tularensis SchuS4 antioxidants in intramacrophage survival, we infected Raw264.7 macrophages with the wild type CFU/mL) were recovered from Raw264.7 cells at 24 hours post-infection as compared to the wild type F. tularensis SchuS4 strain (6.9±0.1 Log 10 CFU/mL). Higher numbers of ΔkatG mutant bacteria were taken up by the macrophages as compared to the wild type F. tularensis SchuS4, the ΔahpC, and the ΔsodC mutants at 4 hours post-infection. However, both ΔkatG and ΔsodC mutants survived and replicated similarly to the wild type F. tularensis SchuS4 strain and equal numbers of bacteria (7.0±0.1 and 7.0±0.2 Log 10 CFU/mL, respectively) were recovered at 24 hours post-infection. The wild type F. tularensis SchuS4 and the ΔsodC bacteria showed a 25-fold increase at 24 hours post-infection than those recovered from macrophages after 4 hours of infection. The ΔkatG mutants exhibited 20-fold increase; while the ΔahpC mutants increased by 17-fold at 24 hours post-infection. These results demonstrate that AhpC contributes to intramacrophage growth of F. tularensis SchuS4 (Fig 9).

Discussion
Francisella tularensis during its intracellular residence encounters a number of oxidative and nitrosative stresses. To overcome these, F. tularensis has evolved a multitude of mechanisms. Francisella counters the phagocyte induced oxidative stress by relying on two divergent approaches; neutralize the ROS/RNS produced by the phagocytic cells and inhibit the assembly of NADPH oxidase [21]. The roles of the primary antioxidant enzymes SodB, SodC and KatG of F. tularensis LVS have been characterized in previous studies [7,11,22]. It has been reported that these antioxidant enzymes are required for resistance of F. tularensis LVS against oxidative stress, survival in IFN-γ stimulated macrophages and virulence in mice [7,8,11]. On the contrary, it has been reported that KatG of F. tularensis SchuS4 although provides some degree of resistance against H 2 O 2 , is neither required for intramacrophage survival nor virulence in mice [11]. Similarly, both SodC and AhpC of F. tularensis SchuS4 are not required for virulence in mice [16]. Very fragmentary information is available regarding the role of both SodC and AhpC [16] of F. tularensis SchuS4, and none related to AhpC of F. tularensis LVS. This study investigated the role of AhpC in oxidative and nitrosative stress resistance of F. tularensis LVS and SchuS4.
Results from this study demonstrate that AhpC plays a role in the protection of Francisella against the oxidative and nitrosative stresses. Furthermore, it was observed that loss of ahpC in F. tularensis LVS is not associated with any intramacrophage growth defect in unstimulated naïve macrophages, but the ΔahpC mutant is attenuated for virulence in mice; a phenotype consistent with the ΔsodC and ΔkatG mutants of F. tularensis LVS (8,11). These findings indicate that antioxidant enzymes of F. tularensis LVS act independently and that loss of one enzyme is not compensated by other antioxidant enzymes in response to oxidative or nitrosative stresses. The SchuS4 ΔahpC mutant showed higher sensitivities towards superoxide-generating compounds and peroxides. However, unlike F. tularensis LVS mutants, the sensitivities of both the SchuS4 ΔsodC and ΔkatG mutants towards superoxide-generating compounds as well as peroxides remained similar to the wild type F. tularensis SchuS4. These observations indicate that F. tularensis SchuS4 AhpC serves as a major antioxidant enzyme in providing resistance against oxidative stresses. Furthermore, the SchuS4 ΔahpC mutant was found to be attenuated for intramacrophage growth indicating that AhpC in F. tularensis SchuS4 unlike LVS, play a role in overcoming the oxidative stress intracellularly.
The ΔahpC mutants of both F. tularensis LVS and SchuS4 exhibited an unusually high sensitivity towards RNS generating compounds SNP and Sin-1. SNP exerts its bactericidal effect by releasing NO which can either be oxidized or reduced to generate highly reactive and microbicidal RNS [23,24]. RNS reacts with cellular thiols, lipids and metals to inhibit metabolism, damage cell membranes and DNA [25]. NO also reacts with superoxide anion to produce highly reactive peroxynitrite anion (ONOO -) [26] which is subsequently decomposed into potent microbicidal reactive nitrogen intermediates [26,27]. Sin-1 generates ONOOby producing both NO and superoxide anions. ONOOis required for macrophage-dependent killing of F. tularensis [28]. However, our previous studies have shown that neither the ΔsodC nor the sodBΔsodC mutants of F. tularensis LVS exhibit enhanced sensitivities towards NO or preformed ONOOunder cell-free growth conditions [22]. Similarly, the viability of the F. tularensis LVS ΔkatG mutant is only partially affected; while the viability of the SchuS4 ΔkatG mutant remains similar to its parental wild type strain upon exposure to Sin-1 [11]. These observations indicate that superoxide dismutases and catalase of F. tularensis LVS and SchuS4 are primarily involved in scavenging ROS, but do not effect RNS. On the other hand, enhanced sensitivities of ΔahpC mutants of both F. tularensis LVS and SchuS4 towards SNP and Sin-1 observed in this study demonstrate that AhpC contributes to resistance against nitrosative stresses.
Majority of Gram-negative bacteria encode AhpC belonging to 2-Cys peroxiredoxins to protect bacteria from ROS and RNS induced cell damage [29]. A conserved peroxidatic cysteine in AhpC reacts with H 2 O 2 or organic peroxides to form sulfenic acid and then subsequently release water or the corresponding alcohols. The oxidized AhpC is reduced and regenerated by an NADH-dependent oxidoreductase AhpF [29]. The AhpC of F. tularensis differs from other members of the peroxiredoxin family of proteins. F. tularensis AhpC is a 1-Cys peroxiredoxin containing a conserved peroxidatic cysteine; however, it lacks the resolving cysteine as well as the reducing partner AhpF. Similar to F. tularensis, AhpC in mycobacteria protects against RNS and hydroperoxides [30]. Mycobacterial AhpC catalyzes the conversion of ONOOto nitrite very rapidly and prevents its spontaneous decomposition into highly microbicidal nitrogen dioxide and hydroxyl radicals [31]. However, unlike F. tularensis, the M. tuberculosis AhpC is a 3-Cys peroxiredoxin containing the peroxidatic cysteine, the putative resolving cysteine and the third cysteine with unknown catalytic role [32]. The peroxidatic cysteine of the mycobacterial AhpC attacks ONOOand gets oxidized to cysteine sulfenic acid residues; while the resolving cysteine completes the catalytic cycle. A thioredoxin-like protein known as AhpD reduces the oxidized AhpC in mycobacteria [31]. The mechanisms through which the AhpC of F. tularensis neutralizes ONOOin the absence of a resolving cysteine and how AhpC is regenerated in F. tularensis in the absence of AhpD/AhpF homologs is yet to be elucidated.
Collectively, this study highlights differences in antioxidant defense mechanisms of F. tularensis LVS and SchuS4 and their abilities to counter oxidative and nitrosative stresses. Nearly 4-5 times the concentration of oxidants and RNS generating compounds used for F. tularensis LVS were required to get tangible results with F. tularensis SchuS4 mutants. One hundred percent of the wild type F. tularensis LVS bacteria were killed when the concentrations of the compounds used in assays with F. tularensis SchuS4 were applied. However, these concentrations either did not affect or only moderately affected the viability of F. tularensis SchuS4. To conclude, our results demonstrate that AhpC of F. tularensis LVS confers resistance against a wide range of ROS and RNS, and serves as a virulence factor. In highly virulent F. tularensis SchuS4 strain, AhpC serves as a key antioxidant enzyme and contributes to its robust oxidative and nitrosative stress resistance, and intramacrophage survival. It also becomes evident from these results that F. tularensis SchuS4 can compensate for the loss of KatG and SodC with other antioxidant enzymes, but may not do so when AhpC is absent. The results from this study further indicate that differences in virulence attributes of F. tularensis LVS and SchuS4 may be due to the inherent differences in their antioxidant defense mechanisms.