OxyR-activated expression of Dps is important for Vibrio cholerae oxidative stress resistance and pathogenesis

Vibrio cholerae is the causative agent of cholera, a dehydrating diarrheal disease. This Gram-negative pathogen is able to modulate its gene expression in order to combat stresses encountered in both aquatic and host environments, including stress posed by reactive oxygen species (ROS). In order to further the understanding of V. cholerae’s transcriptional response to ROS, we performed an RNA sequencing analysis to determine the transcriptional profile of V. cholerae when exposed to hydrogen hydroperoxide. Of 135 differentially expressed genes, VC0139 was amongst the genes with the largest induction. VC0139 encodes a protein homologous to the DPS (DNA-binding protein from starved cells) protein family, which are widely conserved and are implicated in ROS resistance in other bacteria. Using a promoter reporter assay, we show that during exponential growth, dps is induced by H2O2 in a manner dependent on the ROS-sensing transcriptional regulator, OxyR. Upon entry into stationary phase, the major stationary phase regulator RpoS is required to transcribe dps. Deletion of dps impaired V. cholerae resistance to both inorganic and organic hydroperoxides. Furthermore, we show that Dps is involved in resistance to multiple environmental stresses. Finally, we found that Dps is important for V. cholerae adult mouse colonization, but becomes dispensable in the presence of antioxidants. Taken together, our results suggest that Dps plays vital roles in both V. cholerae stress resistance and pathogenesis.


Introduction
The human pathogen Vibrio cholerae, a motile gram-negative bacterium, is the causative agent of the waterborne disease, cholera [1,2] that is still a major threat to public health in developing a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 countries [3]. V. cholerae survives in various environments by sensing and responding to environmental cues. Its pathogenesis is dependent on the oral-fecal route, where it enters the human gastrointestinal tract through oral ingestion and propagates its own release into the environment through toxin production that causes choleric diarrhea [4]. Within a human host, V. cholerae senses signals such as changing oxygen tension and the presence of bile salts and bicarbonate, enabling the activation of a regulatory cascade leading to virulence gene expression [5][6][7][8]. V. cholerae also encounters oxidative stress during the later stages of infection [9,10] as well as in the aquatic environment [11]. In V. cholerae, the ROS-sensing activator OxyR is important for resistance to hydrogen peroxide [12], while OhrR, a regulator of the organic hydroperoxide resistance gene ohrA, regulates V. cholerae resistance to organic hydroperoxides [13]. Quorum sensing systems [14] and the virulence regulator AphB also play important roles in oxidative stress response [15].
Oxidative stress response regulation in bacteria has been extensively studied [16]. Many bacteria have evolved sophisticated regulatory systems to overcome ROS that are acutely toxic to bacterial cells. For example, during oxidative stress, Escherichia coli utilizes OxyR and SoxRS to sense ROS signals and subsequently coordinate the expression of a set of genes encoding ROS scavenging enzymes, such as catalases and peroxidases [17]. In addition, Dps (the DNA-binding protein from starved cells), a non-specific DNA-binding protein, has been known to be implicated in E. coli ROS resistance [18,19]. Dps is the most abundant protein in stationary phase cells, and has been shown to be regulated by OxyR during exponential phase and RpoS during stationary phase [20][21][22]. The non-specific DNA binding of Dps protects DNA against ROS through the physical association with DNA and the ability to nullify the toxic combination of Fe (II) and H 2 O 2 [23]. In addition to playing a role in oxidative stress resistance [24][25][26], Dps is also involved in E. coli resistance to acid stress [27], iron and copper toxicity [25,26,28]. Homologues of Dps are widely distributed throughout bacteria and are important for ROS resistance and other physiological functions such as pathogenesis [29][30][31]. In this study, using RNA sequencing and transcriptional reporters, we found that V. cholerae dps expression is induced by hydrogen peroxide in an OxyR-dependent manner. Deletion analysis indicates that Dps is important for V. cholerae oxidative stress resistance and pathogenesis.

Ethics statement
These studies were limited to the use of mice only. The protocol was approved by the Ethical Committee of Animal Experiments of Nanjing Agricultural University (Permit Number: SYXK (su) 2011-0036). All efforts were made to minimize animal suffering and the number of animals to be used. After infection, mice were monitored until awake and were monitored for signs of distress throughout the duration of experiments. Moribund animals, or animals that appeared to be experiencing pain or suffering, were sacrificed at earlier time points. Upon termination of experiments, the adult mice were euthanized by CO 2 inhalation followed by decapitation.

Strains, plasmids and culture conditions
All strains used in this study were derived from V. cholerae El Tor C6706 [32]. In-frame deletions of dps and rpoS mutants were constructed by cloning the regions flanking the gene of interest into suicide vector pWM91 containing a sacB counter-select marker [33]. Doublecrossover recombinant mutants were selected using sucrose plates. The construction of oxyR mutants is described in [12]. The P dps -lux transcriptional fusion reporter was constructed by cloning dps promoter sequences into pBBR-lux which contains a promoterless luxCDABE reporter [34]. The P tcpA -lux construct is described in [35]. The dps overexpression plasmid was constructed by cloning the PCR-amplified coding regions into pBAD24 [36] and the construction of the oxyR overexpression plasmid is described in [12]. Strains were propagated in LB containing appropriate antibiotics at 37˚C, unless otherwise noted.

RNA sequencing
Wild type V. cholerae were inoculated into AKI medium [37] and incubated without shaking for 4 hrs at 37˚C. One set of cultures were then exposed to 0.5 mM H 2 O 2 for 30 min. RNA was then purified using TRIzol1 (ThermoFisher Sci) and RNeasy purification kits (Qiagen). RNA sequencing was performed by PrimBio Research Institute LLC (Exton, PA, USA). Ribo-Zero rRNA Removal Kit (Bacteria) (Illumina) was used for rRNA removal. Subsequently, the rRNAdepleted RNA was used to construct a cDNA library. cDNA libraries were then constructed using the Ion Total RNA-Seq Kit (Life Technologies). The purified cDNA libraries were then amplified by PCR using Platinum PCR Super-Mix High Fidelity and Ion Xpress RNA Barcode reverse and forward primers. Approximately 10 pM of pooled barcoded libraries were then used for templating using Life Technologies Ion PI IC 200 Kit. Samples were then loaded on Ion P1 chips for Ion Torrent RNA-Seq. Following proton run, the raw sequences were aligned to the V. cholerae N16961 genome. Aligned BAM files were used for further analysis. BAM files, separated by the specific barcodes, were uploaded to the Strand NGS software (San Francisco, CA). Quality control was assessed by the Strand NGS program, which determined the pre-and post-alignment quality of the reads for each sample. The aligned reads were then filtered based on alignment score, match count, mapping quality, and average base quality. After filtering, the aligned reads were normalized and quantified using the Deseq algorithm by Strand NGS. The standard t-test was used to determine significant differentially expressed genes based on two biological replicates of each condition. Sequencing data for RNA-seq experiments are accessible at SRP095162 in the Sequence Read Archive (SRA).

Measuring dps expression using transcriptional reporters
Overnight cultures of wild type, ΔoxyR, ΔoxyR (pBAD-oxyR), and ΔrpoS, all containing P dps -luxCDABE transcriptional fusion plasmids were diluted into fresh LB containing appropriate antibiotics and shaken at 37˚C until early-log/mid-log/late-log or stationary phase. When indicated, cultures were exposed to H 2 O 2 and were incubated for 1 hr. When appropriate, culture medium was supplemented with 0.1% arabinose. Luminescence was then measured and normalized to OD 600. Three independent experiments were performed.

ROS resistance assays
Overnight cultures of wild type, Δdps, and Δdps (pBAD-dps) were diluted 1:1000 into LB containing 0.1% arabinose without or with 250 μM H 2 O 2 or 80 μM cumene hydroperoxide and incubated aerobically at 37˚C. OD 600 was measured at the indicated time points. Three independent experiments were performed.

In vitro stress assays
Mid-log, stationary-phase, and starved cultures (mid-log cultures starved in artificial sea water (ASW) at 22˚C for 2 days) of wild type and dps mutants were exposed to the following stress conditions: 2 mM H 2 O 2 or 3 mM cumene hydroperxide (CHP) exposure for 15', 10 mM FeSO 4 challenge for 15', and acid shock (pH 4.5) for 30'. Survival rate was then determined by plating samples on LB plates after serial dilution and percent survival was calculated by comparing with unexposed cells. For high osmolality stress assay, different growth-staged cultures of wild type and dps mutant cells were incubated in 1 M NaCl at 37˚C for 24 hrs. Percentages of surviving cells were calculated by comparing with the number of cells surviving in 0.8% NaCl /LB (pH 7). Three independent experiments were performed.

VBNC assays
V. cholerae viable but not culturable (VBNC) assays were performed as described in [38]. Briefly, late-log LB cultures of wild type and Δdps were washed and diluted in ASW to a final concentration of 10 8 CFU/ml. The cell suspensions in artificial sea water were then incubated at 4˚C for 70 days. At the indicated time, the number of culturable cells was determined by plating the cell suspension on tryptic soy agar plates supplemented with 0.1% sodium pyruvate. To determine the number of viable cells, samples were treated with propidium monoazide (PMA), which is a DNA-binding PCR inhibitor that selectively crosses compromised cell membranes. After PMA treatment, DNA was isolated from samples and quantitative real time PCR assays were performed by using primers for VC1376. Three independent experiments were performed.

In vitro assays for tcpA expression and TCP pilin production
Overnight cultures of wild type and Δdps containing P tcpA -luxCDABE transcriptional fusion plasmids were inoculated 1:10000 into AKI medium [37] and incubated without shaking at 37˚C for 4 hrs, followed by shaking at 37˚C for an additional 3 hrs. Luminescence was then measured at the indicated time points and normalized to OD 600 . At the final time point, 10 9 cells were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblotting using anti-TcpA antiserum. Three independent experiments were performed. Representative data are shown.

In vivo competition colonization assay
The infant mouse model was used as previously prescribed [39]. Briefly, overnight cultures of wild type (lacZ + ) and Δdps (lacZ -) were mixed in a 1:1 ratio and approximately 10 5 V. cholerae cells were intragastrically inoculated into 5-day-old CD-1 suckling mice. After an 18hr infection period, the mice were sacrificed. Small intestines were homogenized and the ratio of mutants to wild type was determined by plating on LB agar containing 5-bromo-4-chloro-3-indolyl-β-Dgalacto-pyranoside (X-Gal). Each experiment consisted of a sample size of 5 mice.
The streptomycin-treated adult mouse model was used as previously described [13]. Fiveweek-old CD-1 mice were provided drinking water with or without the antioxidant, N-acetyl cysteine (NAC) [1% (wt/vol)] for one week. 0.5% (wt/vol) streptomycin and 0.4% aspartame were then added to the drinking water for the remainder of the experiment. One day after streptomycin treatment, approximately 10 8 wild type and Δdps cells were mixed in a 1:1 ratio and intragastrically administered to each mouse. Fecal pellets were collected at the indicated time points, resuspended in LB, serially diluted, and plated on plates containing X-gal. The competitive index was calculated as the ratio of mutant to wild type colonies normalized to the input ratio. Each experiment consisted of a sample size of 5 mice.

Results and discussion
Global transcriptomic responses of V. cholerae to hydrogen peroxide In order to study how V. cholerae manipulate their genetic reservoirs to resist oxidative stress, we analyzed the transcriptome of V. cholerae grown in the presence and absence of hydrogen peroxide using RNA-seq. We grew V. cholerae in AKI medium [37], in which virulence genes are highly induced, until mid-log phase. Cultures were then exposed to 0.5 mM H 2 O 2 and further incubated at 37˚C for 30 min. Total RNA was then harvested and subjected to subsequent RNA-seq. Read mapping against the V. cholerae N16961 genome was performed and allowed for the identification of differentially expressed genes. The analysis identified the expression of 3689 coding DNA sequence (CDS) tags. Biological replicates were tightly clustered, indicating consistency between replicates. As shown in S1 Table, we identified a total of 135 genes that displayed at least 2-fold differential expression upon H 2 O 2 exposure. These genes are scattered along the two chromosomes of the V. cholerae genome (Fig 1A). Among those differentially expressed genes, expression of 99 genes was repressed in the presence of H 2 O 2 . Many of the downregulated genes are related to primary metabolism and cellular transport systems, suggesting that hydrogen peroxide attenuates cellular metabolism and transport through the cell cholerae were grown in AKI medium [37] without shaking for 4 hrs. One set of cultures were then exposed to 0.5 mM H 2 O 2 for 30 min. RNA was extracted and then subjected to RNA sequencing analysis. A. Chromosomal location of differentially expressed genes. The map was constructed using the Circos program [43]. membrane. Similar phenotypes were also observed in Pseudomonas aeruginosa when exposed to sublethal doses of H 2 O 2 [40]. Interestingly, transcription of the key virulence regulator ToxR, which is essential for V. cholerae pathogenesis [41], was repressed 2.1-fold by H 2 O 2 , indicative of a relationship between ROS resistance and V. cholerae pathogenesis. How oxidative stress influences virulence is subject to another study.
Many of the genes induced by H 2 O 2 , are known to be involved in cellular protective mechanisms (Fig 1B). For example, expression of peroxiredoxin PrxA (VC2637) showed an over 10-fold increase in response to hydrogen peroxide. PrxA in V. cholerae is important for ROS resistance and is induced by H 2 O 2 [12,42]. Both catalase genes, VC1560 (katG) and VC1585 (katB) were also induced. We previously showed [12] that both of these catalases are critical for V. cholerae survival upon exposure to H 2 O 2 . Amongst these strongly induced genes was VC0139, induced over 25-fold, which putatively encodes a Dps family protein. As Dps family proteins have been shown to be involved in ROS resistance in many other bacteria, we chose to further investigate Dps.

dps expression is controlled by OxyR and RpoS at different growth phases
To verify RNA-seq results, we constructed a P dps -luxCDABE transcriptional fusion plasmid to monitor dps expression. We found that in the absence of H 2 O 2 , dps expression was relatively low throughout the growth curve (Fig 2A). The addition of hydrogen peroxide induced dps differentially based on growth phase (Fig 2A), confirming our RNA-seq data that dps transcription was significantly induced by H 2 O 2 . dps induction by H 2 O 2 was increased dramatically during exponential growth, similar to that in E. coli [44]. In stationary phase cultures, however, dps expression was induced less prominently (Fig 2A).
It has been reported that dps expression requires the redox sensor, OxyR in E. coli. [44]. To test whether dps is also regulated by OxyR in V. cholerae, we examined dps expression in oxyR deletion mutants. During exponential growth, compared to wild type, dps induction as a result of H 2 O 2 exposure was abolished in ΔoxyR (Fig 2B, left panel). Complementation of oxyR on a plasmid restores dps induction when exposed to H 2 O 2 (Fig 2B, left panel). These data suggest that OxyR induces dps expression upon exposure to hydrogen peroxide in exponential growth phases in V. cholerae, similar to that in E. coli. Upon entry into stationary phase, however, dps expression in ΔoxyR mutants was similar to that of wild type (Fig 2B, right panel), indicating that OxyR does not regulate dps at this growth phase. It has been reported that in E. coli, the major stationary phase regulator RpoS (σ 38 ) is required for dps expression in stationary phase [44]. To test whether RpoS also regulates dps in V. cholerae, we compared dps expression between wild type and an rpoS in-frame deletion mutant. We found that in mid-log phase, RpoS did not affect dps expression (Fig 2B, left panel), whereas in stationary phase, dps expression was decreased in ΔrpoS mutants (Fig 2B, right panel). These data suggest that RpoS is the key regulator for dps in stationary phase.

Dps is critical for V. cholerae resistance of inorganic and organic hydroperoxides
To investigate the role of Dps in V. cholerae ROS resistance, we constructed a dps in-frame deletion mutant to test the effect of dps and ROS on growth. We found that the growth of Δdps was comparable to that of wild type (Fig 3A). However, in the presence of H 2 O 2 , Δdps showed significantly reduced growth (Fig 3B, triangles). Expression of dps in trans largely restored the growth of Δdps to wild type levels after 2 hours (Fig 3B, diamonds). Similarly, in the presence of organic hydroperoxide such as cumene hydroperoxide (CHP), Δdps displayed significantly delayed growth compared to wild type cells (Fig 3C, triangles) and this defect was partially compensated when dps was expressed in trans (Fig 3C, diamonds). Taken together, these Wild type, ΔoxyR, ΔoxyR (pBAD-oxyR), ΔrpoS containing P dps -luxCDABE reporter plasmids were grown to the indicated growth phase, 100 μM (early-log) or 500 μM H 2 O 2 (mid-log and stationary) were added and incubated for an additional hour. Luminescence was measured and normalized to OD 600 . When indicated, 0.1% arabinose was added in the medium to induce the P BAD promoter. Results are means and s.d. of three independent experiments. *: Student t-test, P<0.05; ns: no significance.
doi:10.1371/journal.pone.0171201.g002 results suggest that Dps is important for V. cholerae growth in the presence of both organic and inorganic hydroperoxides.

Dps is involved in resistance to multiple environmental stresses
In addition to the well-studied role Dps plays in resistance to oxidative stress [24][25][26], it has also been shown to be important for resistance to other stresses, such as starvation [18,45], osmotic stress [46], iron toxicity [25,26,28], and acid stress [27] in many bacteria. To test whether Dps is important for resistance to these stresses in V. cholerae, we compared the survival rate of wild type and Δdps when exposed to different stress signals at different growth stages: exponential phase, stationary phase, and starvation. We found that Δdps mutants were more susceptible to H 2 O 2 during exponential growth, but not at stationary phase ( Fig 4A). The viability of Δdps mutants under the starvation condition was similar to that of wild type (data not shown). However, upon exposure to H 2 O 2 , the number of dps mutants was significantly reduced (Fig 4A), suggesting that in starved cells Dps is critical for protecting V. cholerae against ROS. When cultures were exposed to organic hydroperoxide CHP, Δdps mutant cells were more susceptible than wild type cells at all tested growth phases (Fig 4B). We also exposed wild type and Δdps mutants to high osmolality, acid shock, and high concentrations of iron. We found that Δdps showed lower viability when exposed to high iron concentrations during starvation (Fig 4C). However, Δdps displayed similar survival rates when exposed to high osmolality and low pH (data not shown). These data suggest that Dps is important for V. cholerae survival in starved cells as a response to ROS and for tolerating iron toxicity.
To examine additional roles of Dps with respect to V. cholerae stress resistance, we induced the production of "viable but non culturable (VBNC)" cells. It has been reported that upon exposure to unfavorable environments, V. cholerae can survive by entering a VBNC state [47], in which bacteria fail to grow on routine bacteriological media, but are alive and capable of resuscitation under favorable conditions such as in vivo. To test whether Dps is involved in V. cholerae survival in VBNC, we inoculated mid-log wild type and Δdps into artificial sea water and incubated at 4˚C. Culturable cells were determined by plating cell suspensions on rich  Wild type and Δdps were grown in LB to mid-log and stationary phases. To induce starvation, wildtype and Δdps were grown in LB until mid-log phase. The cells were then resuspended in ASW and incubated at 22˚C for 2 days. A set of cultures was then exposed to medium agar plates, while viable cells were determined by using the PCR method described in [38]. Fig 5A shows that the percentage of culturable wild type cells declined rapidly, while a statistically significantly more culturable Δdps cells were detected. Similarly, we found that there were more viable Δdps cells than wild type cells after 70 days of incubation in ASW at 4˚C ( Fig  5B). These results suggest that Dps has a negative effect on cell viability under the VBNC condition tested. Interestingly, a previous transcriptome study shows that in the VBNC condition, dps expression is over 2-fold higher than that in vegetative cells [48]. This may imply that Dps  wild type (lacZ -) and dps mutants (lacZ + ) were intragastrically inoculated into 5-day-old CD-1 mice in a 1:1 ratio. After 18-hr incubation, CFU from small intestines were determined by serial dilution and plated on LB agar. The data shown are from three independent experiments and each symbol represents CFU recovered from one mouse intestine. Horizontal lines represent the average number of cells recovered. B. Virulence factor expression and production: wild type and Δdps containing P tcpA -luxCDABE were grown under the virulence inducing AKI condition [37]. Luminescence was measured and normalized to OD 600 (left panel), and is involved with the adaptation of cells to stresses such as cold temperatures and poor nutrients.

Dps is important for V. cholerae colonization of inflammatory intestines
To investigate whether Dps plays a role in V. cholerae pathogenesis, we first used an infant mouse colonization model [49] to test for dps mutant colonization. We found that dps mutants could colonize in the small intestine of 5-day-old infant mice as well as wild type (Fig 6A). We also examined whether Dps affects the expression and production of TcpA, the major virulence factor in V. cholerae [50]. In vitro tcpA induction (Fig 6B, left panel) and TcpA production ( Fig 6B, right panel), were similar between wild type and Δdps, suggesting that Dps does not affect virulence gene expression. To examine the role of Dps in ROS resistance during infection, we performed an in vivo colonization competition assay using the streptomycintreated adult mouse model in which bacteria experience host-generated oxidative and nitrosative stress [42,51]. Fig 6C shows that Δdps was outcompeted by wild type in this model. However, treatment with N-acetyl cysteine (NAC), an antioxidant widely used in human and animal studies to lower ROS levels [52], restores Δdps colonization (Fig 6C, squares). These results suggest that Dps is important for ROS resistance in vivo [53,54].
In this study, we show that similar to E. coli and other bacteria, the expression of dps is activated by OxyR and H 2 O 2 during exponential growth in V. cholerae. At stationary phase, RpoS is important for dps expression. Like in other bacteria, Dps is critical for V. cholerae resistance to both inorganic and organic hydroperoxides as well as resistance to iron toxicity during specific growth phases. Interestingly, Dps also has an effect on the production of VBNCs, which may be important for V. cholerae as they reside in aquatic environments between infections. In addition, Dps plays a role in V. cholerae colonization and is critical for V. cholerae in vivo ROS resistance. Our study adds Dps as an additional factor to V. cholerae's arsenal of tools used for survival in both aquatic and host environments. As Dps is well conserved in many bacteria, including pathogens, our study contributes to the knowledge of pathogenic mechanisms required to achieve successful infection.
Supporting information S1 Table. Genes that differentially expressed more than 2-fold upon H 2 O 2 exposure. Wild type V. cholerae were inoculated into virulence-inducing AKI medium and incubated at 37˚C for 4 hrs. One set of cultures were then exposed to 0.5 mM H 2 O 2 for 30 min. RNA was purified and RNA sequencing was performed by PrimBio Research Institute (Exton, PA, USA).