Activation of the Vibrio cholerae SOS Response Is Not Required for Intestinal Cholera Toxin Production or Colonization

Abstract

Cholera toxin, one of the main virulence factors of Vibrio cholerae, is encoded in the genome of CTXφ, a V. cholerae-specific lysogenic filamentous bacteriophage. Although the genes encoding cholera toxin, ctxAB, are known to have their own promoter, the toxin genes can also be transcribed from an upstream CTXφ promoter, P_rstA. The V. cholerae SOS response to DNA damage induces the CTX prophage by stimulating gene expression initiating from P_rstA. Here, we investigated whether ctxA mRNA levels increase along with the levels of the transcripts for the other CTXφ genes following stimulation of the V. cholerae SOS response. Treatment of V. cholerae with the SOS-inducing agent mitomycin C increased the level of ctxA mRNA approximately sevenfold, apparently by augmenting the activity of P_rstA. However, using suckling mice as a model host, we found that intraintestinal ctxA transcription does not depend on P_rstA. In fact, the suckling mouse intestine does not appear to be a potent inducer of the V. cholerae SOS response. Furthermore, alleviation of LexA-mediated repression of the V. cholerae SOS regulon was not required for V. cholerae growth in the suckling mouse intestine. Our observations suggest that pathogenicity of V. cholerae does not depend on its SOS response.

Many key virulence factors, including diphtheria toxin, Shiga toxin, cholera toxin, and several staphylococcal toxins, are encoded by genes found in the genomes of lysogenic bacteriophages (5, 29). Phages provide a means for the horizontal transfer of genes encoding virulence factors and because of this have had a profound influence on the evolution of many pathogens. In addition, recent studies with Shiga toxin (Stx)-encoding phages in Escherichia coli O157:H7 (17, 19, 25, 28) and staphylococcal enterotoxin-encoding phages in Staphylococcus aureus (23) revealed that prophage induction can provide a mechanism to control toxin production. It is not yet known if prophage induction controls the production of other phage-encoded virulence factors.

The molecular processes controlling prophage induction have been extensively studied for lambda and closely related (lambdoid) bacteriophages. Lambda induction requires the SOS response to DNA damage, which, via activated RecA, promotes the autocleavage of the lambda repressor CI. Falling repressor levels trigger a cascade of events that ultimately leads to virion production and lysis of the host cell (20). Virtually no phage production is detectable from lambda lysogens when the SOS response is disabled.

Several studies with lambdoid Stx-encoding phages in E. coli O157:H7 have indicated that the SOS response and the resulting prophage induction also promote Stx production (26, 28). Stx prophage induction promotes Shiga toxin production via several means, including (i) augmenting stx transcription through activation of the late phage promoter p_R′ (28), (ii) increasing stx copy number through phage replication (27), and (iii) enabling toxin release through phage-mediated cell lysis (27). Deletion of p_R′ greatly reduced Stx production in the mouse intestine (28), suggesting that prophage induction makes a significant contribution to pathogenesis. This observation also suggests that the SOS response is induced in a subpopulation of E. coli O157:H7 during infection.

The E. coli SOS response to DNA damage has been the subject of considerable investigation and is one of the best-understood bacterial stress responses. DNA damage leads to an elevated level of intracellular single-stranded DNA, which is thought to be the intracellular inducer of the SOS response. Single-stranded DNA stimulates the “coprotease” activity of RecA (2, 11, 13, 16). Activated RecA promotes the autocleavage of LexA, a repressor that regulates transcription of many genes that function in DNA repair, including lexA and recA (for reviews see references 9 and 12). A decrease in LexA levels alleviates repression of the LexA (SOS) regulon, enabling expression of genes that act to repair DNA damage; afterwards, new synthesis of LexA restores repression of the SOS regulon.

Cholera toxin, one of the principal virulence factors of the agent of cholera, Vibrio cholerae, is encoded in the genome of a lysogenic filamentous bacteriophage (30). Cholera toxin is an AB₅ subunit toxin, and its activity is thought to account for the secretory diarrhea that is characteristic of cholera. The genes encoding cholera toxin, ctxAB, are found at the 3′ end of the CTX prophage, ∼5 kb downstream of P_rstA, which is the only CTXφ promoter required for CTXφ virion production (Fig. 1). Several cellular transcription factors are known to regulate ctxAB expression from a single promoter found immediately upstream of ctxAB (14, 22). We previously found that transcription initiating at P_rstA could extend into ctxAB (6), raising the possibility that factors that influence the activity of P_rstA could influence the production of cholera toxin.

FIG. 1.

SOS-induced increase in ctxA transcript abundance is dependent on P_rstA activity. pIII^CTX and ctxA transcripts levels were measured by quantitative RT-PCR. Fold increases were calculated by dividing the amount of mRNA measured in MMC-treated cells by the amount of mRNA measured in untreated cells. In the schematic diagram of the CTX prophage, the lollipop symbol above rstA indicates the site of the terminator in BD899. N16961 (WT) and BD899 were treated with 20 ng/ml of mitomycin C for 3 h. The data are data from one representative experiment.

We recently reported that the V. cholerae SOS response induces the CTX prophage (21). Exposure of V. cholerae to SOS-inducing, DNA-damaging agents, such as UV light and mitomycin C (MMC), increases production of CTXφ virions. However, the consequences of CTX prophage induction differ considerably from those of λ induction. CTX prophage induction does not result in prophage excision from the chromosome or in cell death, which occurs with induction of the lambda prophage (20). In addition, the molecular interactions that lead to CTX prophage induction are not identical to the interactions that lead to λ induction. V. cholerae LexA binds to and represses P_rstA, along with the CTXφ-encoded repressor RstR. Cleavage of LexA is essential for CTXφ induction; no MMC or UV-stimulated CTX prophage induction was detectable in a CTXφ lysogen that produces a noncleavable form of LexA (LexA A91D) (21). Apparently, the decrease in LexA levels that results from the V. cholerae SOS response alleviates repression of P_rstA, enabling greater transcription from this promoter and elevated production of CTXφ virions.

Here we investigated whether ctxA transcription increases during the V. cholerae SOS response. We found that the SOS response increased the level of ctxA mRNA, probably by increasing transcription initiating from P_rstA. However, unlike what happens with Stx-encoded phages, the SOS response does not play a significant role in the dramatic elevation of transcription of ctxA that occurs within the murine intestine. In fact, growth in the suckling mouse intestine is not a potent stimulus for the V. cholerae SOS response. Furthermore, cleavage of LexA and activation of the SOS regulon were not required for V. cholerae growth in the intestine.

MATERIALS AND METHODS

Bacterial strains and plasmids.

All strains used in this study are derivatives of N16961, an El Tor O1 V. cholerae clinical isolate (10). The construction of BD899, N16961 harboring a terminator in the CTX prophage gene rstA, and MQ164, N16961 containing lexA_A91D, which encodes a noncleavable form of LexA, have been described previously (6, 21). Antibiotics were used at the following concentrations: streptomycin, 200 μg/ml; sulfamethoxazole, 160 μg/ml; trimethoprim, 32 μg/ml; and mitomycin C 20 ng/ml.

Real-time RT-PCR assays.

Quantitative reverse transcription-PCR (RT-PCR) assays were performed basically as described previously (7, 21). V. cholerae RNA was purified using an RNeasy mini kit (QIAGEN), and contaminating DNA was removed by DNase I digestion using a DNA-free kit (Ambion). RNA was isolated from intestinal homogenates of infected mice by use of TRIzol LS (Invitrogen). Specific oligonucleotide DNA primers were used for the RT reactions; these primers have been described previously (21). Each RT reaction mixture contained 2.0 μg of RNA purified from V. cholerae grown in LB medium or 12 μg of RNA purified from intestinal homogenates from suckling mice infected with ∼10⁶ V. cholerae cells ∼20 h earlier. In vitro-synthesized RNA for each transcript of interest was used to generate standard curves for the PCR assays. Ultimately, the amount of cDNA measured in each PCR assay was normalized to the amount of rpoB cDNA, which served as a control for the total amount of cDNA in each reaction. rpoB was chosen as the control gene because we found that there was very little variation in its expression in the culture conditions used in this study (unpublished data).

Intestinal colonization assay.

Competition assays were performed with 5-day-old CD-1 mice as previously described (1, 24). N16961 lacZ and MQ169 cells were mixed at a ratio of 1:1 to generate the inocula (10⁵ to 10⁶ CFU) used for competition assays. For in vitro competition assays, the same cell mixtures were diluted 1:100 in LB broth and grown for ∼20 h at 37°C with rotation.

RESULTS AND DISCUSSION

Intraintestinal CTX prophage induction does not contribute to expression of ctxA during infection.

We recently reported that activation of the V. cholerae N16961 SOS response induces CTX prophage gene expression and virion production (21). For example, stimulation of the V. cholerae N16961 SOS response with MMC increases transcript abundance for the CTXφ genes pIII^CTX and rstB more than 50-fold (Fig. 1 and data not shown). In addition, we detected an approximately sevenfold increase in the amount of ctxA mRNA following MMC treatment (Fig. 1). As the ctxA promoter is not known to be influenced by the SOS response, we hypothesized that this increase in ctxA mRNA levels depended upon the upstream SOS-regulated promoter P_rstA (Fig. 1). Consistent with this hypothesis, the ctxA transcript abundance did not increase following MMC treatment of BD899, an N16961 derivative with a transcription terminator inserted between P_rstA and P_ctxA (6).

Since cholera toxin production is known to be induced during V. cholerae growth in the intestine (15, 18), we wondered if production of at least some of the intraintestinal ctxA transcripts is attributable to transcription initiation from P_rstA as a consequence of intraintestinal CTX prophage induction. To explore this possibility, we used quantitative RT-PCR to compare the levels of ctxA and rstB transcripts in V. cholerae grown in the infant mouse small intestine and in vitro in LB medium. RNA extracted from intestinal homogenates contained ∼1,000-fold more ctxA mRNA than was present in the in vitro culture (Fig. 2). In contrast, there was only a minimal increase (approximately twofold) in the rstB transcript levels in intestinal homogenates compared to LB medium cultures. These data suggest that the suckling mouse intestine does not provide a potent stimulus for CTX prophage induction and that P_rstA does not significantly contribute to expression of ctxA in this environment. The limited role of P_rstA in intraintestinal ctxA expression was verified using BD899. The intraintestinal ctxA transcript levels in this strain did not differ significantly from those in wild-type V. cholerae (Fig. 2), again suggesting that the phage promoter P_rstA makes little or no contribution to ctxA expression in the suckling mouse intestine.

FIG. 2.

Transcript abundance in V. cholerae from intestinal homogenates of infected infant mice compared to in vitro cultures. Transcript levels were measured by quantitative RT-PCR. Each symbol indicates the amount of mRNA found in an intestinal homogenate from a single mouse relative to the amount of mRNA measured in late-exponential-phase cells cultured in LB medium. All comparisons are comparisons for strain N16961 unless indicated otherwise.

V. cholerae SOS response is not induced during intraintestinal growth.

The limited increase in rstB transcript abundance in the suckling mouse small intestine suggests that the SOS pathway and its effectors (e.g., the LexA regulon) are not induced in this environment. To further investigate this idea, we compared the levels of lexA transcripts in cells from intestinal homogenates and from in vitro cultures (lexA belongs to the LexA regulon). As observed with rstB transcripts, there was only a minimal increase (∼1.6-fold) in the amount of lexA mRNA in intestinal homogenates from infected mice compared to the amount in overnight LB medium cultures. In contrast, we previously found that UV treatment of N16961 resulted in a fivefold increase in lexA transcripts (21). These observations suggest that the suckling mouse small bowel does not provide potent stimuli for induction of the V. cholerae SOS response.

LexA cleavage is not required for V. cholerae growth in the suckling mouse small intestine.

To explore if the V. cholerae SOS response is required for V. cholerae intestinal colonization, we tested whether MQ164, an N16961 lexA_A91D mutant, colonizes the suckling mouse small bowel as well as the wild-type strain. This mutant produces a noncleavable LexA and therefore cannot initiate an SOS response. Nearly equal numbers of N16961 lacZ and MQ164 cells were coinoculated intragastrically into suckling mice. Twenty hours after inoculation, equal numbers of MQ164 and N16961 lacZ cells were found in homogenates of the small bowels of infected mice (Fig. 3). This observation indicates that LexA cleavage and elevated levels of LexA-repressed genes are not required for V. cholerae survival and growth in the suckling mouse intestine. Induction of LexA-repressed genes is also not important for V. cholerae growth in overnight cultures in LB medium at 37°C, since we also recovered equal numbers of N16961 lacZ and MQ164 cells in in vitro competition experiments (Fig. 3). However, it is important to note that the products of many SOS-induced genes are thought to be present at relatively high levels in uninduced cells (31). Therefore, our findings do not imply that LexA-regulated genes are not required for V. cholerae growth either in vivo or in vitro.

FIG. 3.

Competitive indices obtained from in vivo and in vitro competition assays with N16961 and MQ164. Competitive index = (MQ164 output/N16961 output)/(MQ164 input/N16961 input). MQ164 is N16961 lexA_A91D. For in vivo competition, each symbol represents an individual mouse, and for in vitro competition, each symbol represents a different experiment.

Conclusions.

We found that induction of the V. cholerae SOS response increases the amount of ctxA mRNA, a previously unrecognized mode of regulation of cholera toxin production. Most likely, SOS increases ctxA mRNA levels by augmenting the activity of P_rstA, a CTXφ promoter located ∼5 kb upstream of ctxA, since introduction of a terminator into rstA abrogated SOS-induced increases in ctxA mRNA. However, based on our findings using suckling mice as a model host, intraintestinal ctxA transcription does not depend on P_rstA. In fact, the suckling mouse small intestine does not appear to be a potent inducer of the V. cholerae SOS response, and derepression of the LexA regulon is not required for V. cholerae growth in the suckling mouse intestine.

It is possible that infant mice do not accurately reproduce conditions in the human intestine. Although it is not yet clear whether the mammalian adult gastrointestinal tract induces the SOS response in V. cholerae or other enteric pathogens, the V. cholerae gene expression profile obtained for cells in fresh cholera stools suggests that there is some induction of the SOS response and the CTX prophage during V. cholerae growth in humans (3). It is important to note, however, that the gene expression profile of V. cholerae in stools may differ considerably from the expression profile of organisms growing in the small bowel.

Even though we did not observe a role for the V. cholerae SOS response in the suckling mouse model of cholera, the SOS response appears to be important in controlling virulence in several other pathogens. For example, our previous work with Stx-encoding phages in E. coli O157:H7 (enterohemorrhagic E. coli [EHEC]) strongly suggests that EHEC growth in the adult mouse gastrointestinal tract leads to induction of the EHEC SOS response in a subpopulation of cells (28). Gene expression profiles for Salmonella enterica serovar Typhimurium growing in macrophage-like cells also suggest that the SOS response of this enteric pathogen is moderately induced during intracellular growth (8). The bacterial SOS response may be important for regulating virulence gene expression in extracellular pathogens outside the gastrointestinal tract as well (4). For example, the SOS response of S. aureus may also be induced during infection, as expression of fnbB, which encodes a virulence factor that mediates fibronectin binding, has been found to be regulated by LexA.

There is a striking contrast in the importance of prophage induction for cholera toxin production and Shiga toxin production. Cholera toxin production by V. cholerae does not appear to be dependent on CTX prophage induction, whereas Shiga toxin production (especially production of Shiga toxin 2) by EHEC relies on prophage induction for stx transcription and toxin release (27). Neither stx nor ctx appears to have any role in the biology of the bacteriophages in which they reside; instead, these genes have a dramatic influence on the biology of EHEC and V. cholerae, respectively. Thus, it seems as if stx has not been assimilated into host cell regulatory pathways to the same extent as ctx. Additional studies are required to learn whether the stx model of phage control of virulence factor production or the ctx model of cellular control of virulence factor production is the correct model for the many other phage-borne virulence factors.