Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Mol Microbiol. Author manuscript; available in PMC 2009 September 29.
Published in final edited form as:
PMCID: PMC2754204

Cholera stool bacteria repress chemotaxis to increase infectivity


Factors that enhance the transmission of pathogens are poorly understood. We show that Vibrio cholerae shed in human ‘rice-water’ stools have a 10-fold lower oral infectious dose in an animal model than in vitro grown V. cholerae, which may aid in transmission during outbreaks. Furthermore, we identify a bacterial factor contributing to this enhanced infectivity: The achievement of a transient motile but chemotaxis-defective state upon shedding from humans. Rice-water stool V. cholerae have reduced levels of CheW-1, which is essential for chemotaxis, and were consequently shown to have a chemotaxis defect when tested in capillary assays. Through mutational analyses, such a state is known to enhance the infectivity of V. cholerae. This is the first report of a pathogen altering its chemotactic state in response to human infection in order to enhance its transmission.


Vibrio cholerae causes explosive outbreaks characterized by profuse secretory diarrhoea, which results largely from the action of the ADP-ribosylating cholera toxin (Gill, 1977). Shedding of bacteria within ‘rice-water’ stools is central to disease propagation within such outbreaks. Surprisingly, a remarkably high dose of V. cholerae is required to produce disease in human volunteers: the infectious dose was determined to be between 108 and 1011 colony-forming units (cfu). However, as V. cholerae are sensitive to low pH, the infectious dose can be reduced to 104−106 cfu if bicarbonate, which neutralizes gastric acid, is administered to volunteers prior to ingestion of the V. cholerae inoculum (Cash et al., 1974). Although it is widely believed that the infectious dose of V. cholerae during cholera epidemics is likely to be considerably less than 106 cfu, transmission of V. cholerae within this natural setting is a poorly investigated area.

In order to begin to study V. cholerae transmission, a previous study demonstrated that V. cholerae shed in rice-water stools out-compete V. cholerae grown in vitro 10-to-100-fold during a mixed infection in infant mice (in vivo) (Merrell et al., 2002). This hypervirulence persisted following incubation of the human-shed V. cholerae for 5 h in pond water, but was abolished by subsequent growth in vitro. Thus, passage through the human host imparts V. cholerae with a transient competitive advantage when infecting subsequent hosts. These data suggested that rice-water stool V. cholerae might also have a lower infectious dose than V. cholerae grown in vitro, although this was not determined. Furthermore, as this phenomenon was observed with rice-water stool of the Inaba serotype of V. cholerae O1 El Tor, it was also unknown whether hypervirulence following human passage was specific to the Inaba serotype or is a general feature of O1 El Tor biotype V. cholerae.

A similar competitive advantage during infection of infant mice has also been observed for non-chemotactic mutants of V. cholerae (Freter and O'Brien, 1981a; Lee et al., 2001). Chemotaxis allows motile bacteria to sense and respond to their surrounding environment by modulating flagellar rotation, such that bacteria swim up concentration gradients of attractants and down concentration gradients of repellents (Berg and Brown, 1972). V. cholerae, which has a single polar sheathed flagellum (Sjoblad et al., 1983), appears to use chemotaxis to direct it towards the distal portion of the small intestine, thus avoiding colonizing the upper small intestine of infant mice. As a result, smooth-swimming non-chemotactic mutants of V. cholerae colonize through-out the infant mouse small intestine and out-compete the wild-type strain approximately 70-fold in vivo (Lee et al., 2001). In addition, Freter and colleagues reported that non-chemotactic V. cholerae might avoid a host antimicrobial response by virtue of their inefficient penetration into the intestinal crypts, a condition that wild-type chemotactic V. cholerae are exposed to (Freter and O'Brien, 1981a; Freter et al., 1981). Consequently, smooth-swimming non-chemotactic V. cholerae have an order of magnitude lower infectious dose compared with the wild-type strain (Butler and Camilli, 2004). Therefore, any defect in chemotaxis on the part of rice-water stool V. cholerae could explain the hypervirulence observed and could potentially result in a lower infectious dose. However, transcriptome analyses of rice-water stool V. cholerae have produced conflicting results regarding their predicted chemotactic state. Although one study showed downregulation of chemotaxis genes (Merrell et al., 2002), another one that used a different internal reference detected no such downregulation (Bina et al., 2003).

In Escherichia coli, chemotaxis involves a signal transduction cascade that regulates the direction of flagellar rotation (reviewed in Manson et al., 1998; Bourret and Stock, 2002; Wadhams and Armitage, 2004). CheW serves as the linker protein between the methyl-accepting chemotaxis proteins (MCPs) and the histidine kinase CheA and is therefore required for signal transduction (Wadhams and Armitage, 2004). Another essential feature of chemotaxis is adaptation to continued changes in chemoattractant and chemorepellent concentrations through modulation of the methylation state of MCPs (Goy et al., 1978; Borkovich et al., 1992) via the methyltransferase CheR (Springer and Koshland, 1977) and methylesterase CheB (Yonekawa et al., 1983). Mutations in cheW and cheR result in a smooth-swimming non-chemotactic phenotype (Springer and Koshland, 1977; Boukhvalova et al., 2002). In contrast to E. coli, which has one gene encoding each Che protein, many motile bacteria (including V. cholerae) encode multiple paralogues of each of the chemotaxis components, which are often organized into distinct gene clusters (Bourret et al., 2002). These additional systems have been shown to regulate processes such as twitching motility and biofilm formation in Pseudomonas aeruginosa (Heydorn et al., 2002; Whitchurch et al., 2004), and adventurous and social motility in Myxococcus xanthus (Sun et al., 2000; Black and Yang, 2004; Vlamakis et al., 2004).

The transcriptional profiling study by Merrell et al. (2002) found that transcripts encoding all three CheW and CheR paralogues, as well as 17 of the 43 MCPs, were downregulated in rice-water stool V. cholerae compared with a wild-type strain grown overnight in vitro. Although these data suggested the presence of a chemotaxis defect, this hypothesis was not tested. There is some uncertainty associated with this hypothesis: first, some of these chemotactic signalling protein paralogues may not be involved in chemotaxis, as has been shown for the majority of the V. cholerae CheY and CheA paralogues (Gosink et al., 2002; Hyakutake et al., 2005). Second, Bina et al. (2003) did not report downregulation of many of these same chemotaxis gene paralogues in their transcriptional profiling study of rice-water stool V. cholerae. In the present study, we have determined the requirements of the cheW and cheR paralogues for chemotaxis and hypervirulence, and shown that rice-water stool V. cholerae have decreased levels of the required CheW. Further, we show that V. cholerae in rice-water stools are phenotypically motile but reduced for chemotaxis and, accordingly, have a reduced infectious dose in the infant mouse model.


cheW-1 and cheR-2 are required for chemotaxis in V. cholerae

Although the V. cholerae genome encodes three cheW and three cheR paralogues (Heidelberg et al., 2000), only one of these (cheW-1) is located within the genetic cluster that has previously been shown to contain genes that are required for chemotaxis (Gosink et al., 2002). However, despite its location within this cluster, the cheZ gene is not required for chemotaxis in V. cholerae (Butler and Camilli, 2004), so it was possible that cheW-1 might also be dispensable for chemotaxis. In order to determine the requirements of the cheW and cheR paralogues for chemotaxis, we constructed in frame deletions of cheW-2,3 (the start codon of cheW-2 was fused to the stop codon of the adjacent cheW-3), cheR-1, cheR-2 and cheR-3. Additionally, a non-polar G69D point mutation in cheW-1 was constructed. The homologous point mutation in E. coli disrupts the interactions of CheW with CheA and the Tar MCP and confers a smooth-swimming non-chemotactic phenotype (Boukhvalova et al., 2002). As shown in Fig. 1A and B, the cheW-1G69D and ΔcheR-2 mutants were non-chemotactic as determined using chemotaxis plates, and both mutants exhibited smooth swimming as opposed to ‘tumbly’ swimming (data not shown). The chemotaxis defect of the cheW-1G69D mutant was complemented in trans by wild-type cheW-1 on a low copy number vector (Fig. 1A). In contrast, the cheW-2, cheW-3, cheR-1 and cheR-3 genes were dispensable for chemotaxis. Thus, of the three cheW and cheR paralogues encoded by V. cholerae, all of which were predicted to be downregulated in one transcriptome analysis of stool V. cholerae (Merrell et al., 2002), only cheW-1 and cheR-2 are required for chemotaxis.

Fig. 1
Examination of the chemotactic abilities of cheW and cheR paralogue mutants using semisolid agar chemotaxis plates (1% tryptone, 0.5% NaCl, 0.3% agar; 37°C overnight).

The cheW and cheR mutants were also tested for virulence in an infant mouse competition assay within the small intestine following intragastric inoculation. Consistent with the requirements of cheW-1 and cheR-2 for chemotaxis, only the cheW-1G69D and ΔcheR-2 mutant strains out-competed the wild-type strain (Fig. 2A and B respectively). The out-competition phenotype observed in the cheW-1G69D mutant was fully complemented in trans (Fig. 2A), whereas complementation of ΔcheR-2 was not examined. The out-competition phenotype of the mutants is consistent with their smooth-swimming non-chemotactic phenotype, as had been previously reported for a ΔcheY-3 mutant (Butler and Camilli, 2004). In further support of this, the cheW-1G69D and ΔcheR-2 mutants competed equally with the ΔcheY-3 strain (Fig. 2A and B respectively). The remaining cheW and cheR mutants displayed no observable phenotype in vivo (Fig. 2A and B) and therefore repression of these paralogues is not responsible for the hypervirulence of rice-water stool V. cholerae.

Fig. 2
Examination of virulence in an infant mouse competition assay. Each data point represents the CI from an individual mouse. The horizontal bars show the geometric means from each experiment. The CI is calculated as the ratio of mutant to the test strain ...

Stool V. cholerae have decreased CheW-1 levels

Due to the differing published microarray results concerning cheW-1 expression in stool V. cholerae, we measured CheW-1 protein levels by quantitative Western blot analysis. As V. cholerae in dark-field positive rice-water stools are motile and each bacterium has a single polar sheathed flagellum (Sjoblad et al., 1983), we normalized CheW-1 to flagellin using polyclonal antisera to each and radiolabelled Protein A. Western blot analysis of five cholera patient stool samples, which included both O1 Inaba and O1 Ogawa stool samples, showed greatly decreased levels of CheW-1 compared with V. cholerae grown in vitro either to stationary phase or on plates (Fig. 3A). As cheW-1 is essential for chemotaxis in V. cholerae, this decrease in CheW-1 levels is predicted to negatively affect the chemotactic ability of the rice-water stool V. cholerae.

Fig. 3
Analysis of CheW-1 content and chemotactic ability of rice-water stool V. cholerae.

Rice-water stool V. cholerae have a chemotaxis defect

As mentioned in the Introduction, the out-competition phenotype of human-passaged V. cholerae is a transient phenomenon that is abolished by growth of the bacteria in vitro. We were therefore unable to use chemotaxis plates to determine the chemotactic state of stool V. cholerae because this method is not only qualitative, but requires growth in vitro. Instead, we used a modified version of the Adler capillary assay (Adler, 1973). These assays were done in 96 well plates for rapid analysis of a large number of capillaries. Because Freter and colleagues found greater reproducibility when capillary assays with V. cholerae were performed as a competition experiment (Freter and O'Brien, 1981b), these experiments were done as a competition between rice-water stool and wild-type V. cholerae grown in vitro on Luria–Bertani (LB) agar plates at 37°C. The latter growth condition results in consistently motile and chemotactic V. cholerae cells. In addition, because this condition produces a mixture of growth phases, it may be more representative of a possible mixed growth phase population that was proposed for V. cholerae in rice-water stools (Bina et al., 2003).

Nine cholera patient stool samples were tested using eight PBS-filled and eight chemoattractant-filled capillary tubes. As shown in Fig. 3B, stool V. cholerae enter into capillary tubes containing PBS to the same extent as a chemotactic wild-type strain grown in vitro (che+). However, in the presence of a mixture of known chemoattractants (Freter and O'Brien, 1981b) the ratio of stool to che+ V. cholerae in capillaries decreases approximately threefold (P < 0.001, two-tailed Student's t-test). This defect in chemotaxis was observed with rice-water stool V. cholerae of both the Ogawa and Inaba serotypes and is therefore a general property of O1 El Tor V. cholerae. In control capillary assays the wild-type out-competed a ΔcheY-3 non-chemotactic mutant (AC-V1736) by 5.9 and 6.5-fold in two independent experiments. Both Ogawa and Inaba serotype strains isolated from rice-water stools by single-colony purification exhibited a fully chemotactic phenotype in chemotaxis plates (data not shown). Thus, V. cholerae in human stool are impaired in their ability to chemotax, and this impediment is predicted to increase both the virulence and infectivity of these bacteria.

Stool V. cholerae of the O1 El Tor serotype Ogawa exhibit enhanced virulence

As the hypervirulence phenotype of rice-water stool V. cholerae had been documented with O1 El Tor serotype Inaba V. cholerae, we determined whether the same phenomenon would occur with V. cholerae of the Ogawa serotype. As shown in Fig. 4, rice-water stool V. cholerae O1 El Tor serotype Ogawa also out-compete V. cholerae grown overnight in vitro. This also is therefore a general feature of V. cholerae O1 El Tor strains and not a serotype-specific phenomenon. In addition, as rice-water stool V. cholerae of both serotypes have a chemotaxis defect, we competed rice-water stool V. cholerae against the ΔcheY-3 mutant in an infant mouse competition experiment. The rice-water stool V. cholerae and ΔcheY-3 mutant competed equally well (Fig. 4), consistent with the defect in chemotaxis observed.

Fig. 4
Competition experiments with rice-water stool V. cholerae O1 El Tor serotype Ogawa. Infant mice competition experiments were performed with V. cholerae O1 El Tor serotype Ogawa from human stool (LacZ+) competed against either the wild-type strain AC-V984 ...

Stool V. cholerae are an order of magnitude more infectious

Although rice-water stool V. cholerae were shown to be hypervirulent compared with V. cholerae grown in vitro, the effects of passage through humans on infectivity per se has not been determined. Because rice-water stool V. cholerae have a chemotaxis defect, we therefore predicted an increase in infectivity. We determined the infectious dose at which 50% of infant mice become infected (ID50) by human stool V. cholerae and find that stool V. cholerae are an order of magnitude more infectious than V. cholerae grown in vitro: the ID50 of rice-water stool V. cholerae was approximately 15 cfu compared with 140 cfu for wild-type AC-V999 grown overnight in vitro, which are the conditions used for the in vivo competition experiments (Fig. 5). This magnitude decrease in ID50 is identical to that observed for non-chemotactic mutants of V. cholerae (Butler and Camilli, 2004), consistent with the chemotaxis defect we observe in the human stool V. cholerae. Similar results were obtained in an independent experiment using a stool sample from another patient (data not shown). Both stool samples used for the ID50 determinations were of the Ogawa serotype; however, the wild-type competitor strain AC-V999 is an Inaba clinical isolate from 2001. In order to confirm that the increase in infectivity observed did not result from differences between the V. cholerae O1 El Tor Ogawa and Inaba strains used, the rice-water stool V. cholerae used for the ID50 determination in Fig. 5 was grown overnight in vitro and the ID50 was determined along with that of AC-V999. The ID50 for both strains grown in vitro was identical and thus the increased infectivity of stool V. cholerae is only observed upon exit from the human host (data not shown).

Fig. 5
Determination of the ID50 for stool and in vitro grown V. cholerae. Groups of five mice were infected with either stool V. cholerae O1 El Tor serotype Ogawa (filled circles) or strain AC-V999 grown overnight in vitro (empty circles). The doses used are ...


In this study, we show that human-shed V. cholerae have a transiently reduced chemotactic state. This is, to our knowledge, the first report demonstrating that a pathogen alters its chemotactic state in response to human infection. In addition, these findings provide the first explanation for the enhanced virulence previously observed for stool V. cholerae. Consistent with the presence of defective chemotaxis, rice-water stool V. cholerae are an order of magnitude more infectious than when grown in vitro. If infection of infant mice mimics that in humans, then it appears that pathogenic V. cholerae have evolved to repress chemotaxis upon exiting the human host in order to enhance transmissibility. This lower infectious dose would be predicted to aid in the spread of V. cholerae during localized outbreaks of cholera.

Although all three cheW and three cheR paralogues appeared to be transcriptionally downregulated in rice-water stool V. cholerae compared with V. cholerae grown in vitro, we have determined that only the cheW-1 and cheR-2 paralogues are required for chemotaxis. We also showed that the level of CheW-1 protein is decreased, which would suggest the presence of smooth swimming by stool V. cholerae. We have not confirmed the downregulation of CheR-2 by Western blot analysis; however, decreased levels of CheW-1 would be sufficient to cause a chemotaxis defect. If the remaining cheW and cheR paralogues are also downregulated at the protein level, the significance of that finding would currently be unclear because no function has yet been ascribed to the chemotaxis genes located outside of the dominant chemotaxis cluster (with the exception of cheR-2). As mentioned previously, alternative chemotaxis clusters often regulate processes like biofilm formation and flagellar-independent motility such as twitching motility. However, V. cholerae do not appear to exhibit flagellar-independent surface motility (Watnick et al., 1999), and we have been thus far unable to detect any defects in biofilm formation for any of the cheW and cheR paralogue mutants (S. Butler and A. Camilli, unpubl. data). Therefore the significance of any potential downregulation of these proteins may not become evident until such time as their function becomes known.

No functional analysis of the chemotactic state of rice-water stool V. cholerae has been reported previously. We therefore examined the chemotactic ability of rice-water stool V. cholerae using a capillary assay and found that these bacteria have an approximately threefold defect in chemotaxis compared with V. cholerae that are wild-type for chemotaxis. Furthermore this chemotaxis defect was observed with two independent O1 El Tor serotypes (Ogawa and Inaba). These data suggest that this defect in chemotaxis is a genuine characteristic V. cholerae O1 El Tor exiting the human host.

We have previously shown that smooth-swimming non-chemotactic mutants of V. cholerae have a lower infectious dose than the wild-type strain. This increase in infectivity is likely a combination of the increased area of the small intestine colonized by these mutants and the potential avoidance of an antimicrobial response as proposed earlier. We hypothesize that by temporarily downregulating chemotaxis, rice-water stool V. cholerae can initially colonize the upper small intestine. If so, this property would reduce the infectious dose required for infection because more of these bacteria will initially colonize the intestine than if they exhibited functional chemotaxis. Presumably, after a period of time these bacteria may reinitiate and use chemotaxis to penetrate deeper in the intervillous spaces and to reach the distal small intestine. This is likely to be important because it has been previously shown that spontaneous non-chemotactic mutants exhibit an approximately threefold decrease in penetration of the infant mouse intestinal mucosa (Allweiss et al., 1977), which is remarkably similar in magnitude to the threefold decrease in chemotaxis we observe for stool V. cholerae. An additional role for the downmodulation of chemotaxis may be to facilitate dissemination from the host. If the V. cholerae were fully chemotactic they might respond to chemoattractants present within the host and be more likely to remain in the intestinal crypts, where they are known to colonize (Freter et al., 1981). A temporary decrease in chemotaxis may therefore facilitate improved entry into the lumen and subsequent exit from the host.

A number of questions remain unanswered regarding the temporary defect in chemotaxis that is observed. First, we do not know whether the chemotaxis defect observed is due to an equivalent defect among the population as a whole, or rather results from a subpopulation of V. cholerae that are impaired for chemotaxis, as the capillary assays do not distinguish between these two possibilities. Regardless, a defect in chemotaxis in any proportion of the population may have important implications for transmission of V. cholerae following dissemination from a host. Second, it is unknown whether it is the environment of the rice-water stool itself that causes the decrease in CheW-1 levels and the resulting increase in infectivity, or whether they have undergone these changes prior to entering into the intestinal lumen. Finally, it is unclear how long the chemotaxis defect and subsequent increase in infectivity persists in the environment after exiting the human host. We know from previous studies that the hypervirulent phenotype survives incubation in pond water for 5 h; however, the chemotactic ability and infectivity of such bacteria has not been tested.

A recent paper has described the use of mouse-passaged V. cholerae as an equivalent model for the increase in infectivity that is observed with human-passaged V. cholerae (Alam et al., 2005). Although V. cholerae in mid-exponential phase have a competitive advantage over stationary-phase V. cholerae during infection, the authors showed that the increase in infectivity observed when V. cholerae are first passaged through mice is not solely due to differences in growth phase or priming for virulence gene expression. It is therefore also likely that the increase in infectivity observed with rice-water stool V. cholerae is not due to differences in growth phase alone. Unfortunately, this model is not suitable for an analysis of the chemotactic ability of these mouse-passaged V. cholerae because it would be extremely difficult to remove the bacteria from the small intestine homogenate material. Therefore, the use of rice-water V. cholerae is crucial in this instance for examining the chemotactic ability of stool V. cholerae.

We do not know whether non-chemotactic smooth-swimming mutants or rice-water stool V. cholerae are more infectious in humans. However, the infant mouse model has been a good predictor of factors important for V. cholerae colonization and virulence in past studies (Klose, 2000). In addition, although we have determined the infectivity of human-passaged V. cholerae in infant mice, the infectivity of pathogenic V. cholerae in the environment prior to human passage is unknown. Determination of this infectivity would be difficult because V. cholerae cannot be directly cultured in large numbers from the environment, but require an enrichment step to enable culturing (Faruque et al., 2004). The increased infectivities of natural forms of pathogens compared with experimental infection using in vitro grown bacteria have been demonstrated in pathogens such as Citrobacter rodentium and entero-haemorrhagic E. coli (Cornick and Helgerson, 2004; Wiles et al., 2005). Interestingly, all toxigenic V. cholerae isolated from local water sources during a 2004 outbreak in Dhaka, Bangladesh were of a single clone, which was the same clone present in clinical isolates from that outbreak (Faruque et al., 2005). It is conceivable that this widespread outbreak was caused by a single clone that was initially passaged by one or a few susceptible individuals, and with the subsequent increased numbers of shed bacteria combined with a lower infectious dose, was able to be amplified within the human population and thus surrounding bodies of water.

We have shown in this study that V. cholerae shed from humans are an order of magnitude more infectious than V. cholerae grown in vitro for infection of infant mice. Furthermore, these human-shed V. cholerae have a chemotaxis defect compared with in vitro grown V. cholerae. This downmodulation of chemotactic ability by V. cholerae shed from humans is likely to be one reason for the hyperinfectivity observed and provides us with an insight into the multiple strategies that V. cholerae may employ during its life cycle to take advantage of its human host.

Experimental procedures

Bacterial strains and growth conditions

The strains and plasmids used in this study are shown in Table 1. Strains were grown to stationary phase in LB broth at 37°C with aeration or on LB plates at 37°C. Chemotaxis plates (1% tryptone, 0.5% NaCl, 0.3% agar) were incubated at 37°C overnight. L agar (without added NaCl) supplemented with 10% sucrose was used for counter selection of pCVD442-lac derivatives. Antibiotics were used at the following concentrations; ampicillin (Ap), 50 μg ml−1; streptomycin (Sm), 100 μg ml−1; and tetracycline (Tc), 1 μg ml−1. 5-bromo-4-chloro-3-indolyl β-D-galactopyranoside (Xgal) was used at a concentration of 40 μg ml−1.

Table 1
Bacterial strains and plasmids used in this study.

Strain constructions

The alleles for construction of the deletion and point mutations were generated by splicing overlap extension polymerase chain reaction (PCR) using PFU polymerase (Stratagene). All gene deletions were in frame, and the sequences deleted are listed in Table 1. The resulting PCR products were initially cloned into the pCR-Script Amp SK vector (Stratagene), followed by subcloning into the pCVD442-lac plasmid using restriction sites that had been incorporated into the outer primers used for PCR amplification. The cheW-1G69D and ΔcheR-1 constructs were subcloned using SacI and SphI restriction sites; the ΔcheW-2,3 and ΔcheR-2 constructs were subcloned using SphI and XbaI restriction sites; the ΔcheR-3 construct was subcloned using SacI and XbaI restriction sites. The pCVD442-lac derivatives were electroporated into E. coli Sm10λpir and transformants were then mated with V. cholerae. Mutant alleles were introduced into the V. cholerae genome by allelic exchange as described previously (Donnenberg and Kaper, 1991). PCR and DNA sequencing were used to confirm the presence of the correct mutations.

Complementation analysis

cheW-1 (bases −35 to +478, relative to start codon) was amplified by PCR, cloned into pCR-Script Amp SK, and was subcloned into pMMB67EH (low-copy number vector) using the SacI and SphI restriction sites that were incorporated into the primers used for amplification. The pcheW-1 complementing plasmid and the pMMB67EH empty vector were introduced into V. cholerae by electroporation.

Chemotaxis assays

Rice-water stool samples were examined by dark-field microscopy and inhibition of motility using monoclonal antibodies to confirm the presence of V. cholerae. Stool samples were clarified by centrifugation for 4 min at 4000 rpm prior to subsequent study. Approval of human studies was obtained from the Research and Ethical Review Committees of the ICDDR,B in Bangladesh and from the Institutional Review Board of the Massachusetts General Hospital in Boston MA.

For chemotaxis assays, AC-V984 colonies were resuspended in 1 ml of PBS and a portion was added to stool at a ~1:1 cell ratio. This mixture was washed twice with PBS and 150 μl was aliquoted into wells of a 96 well plate. One microlitre capillaries (55 mm, Drummond) filled with PBS or PBS containing L-Serine, L-Arginine and D-Glucose (10 mM final concentration of each) were inserted into each well. After 6 min capillaries were removed and the capillary contents were expelled into 99 μl LB. The ratio of stool V. cholerae to AC-V984 from each was determined by plating dilutions onto LB agar supplemented with Sm and Xgal. The ratio in each capillary was adjusted for the input ratio.

Western blot analysis

Five hundred microlitres of an overnight culture of AC-V999 was pelleted and the cell pellet resuspended in 150 μl protein sample buffer. Four millilitres of each clarified stool sample was pelleted and the pellet resuspended in 200 μl sample buffer. Samples were subjected to SDS-PAGE followed by transfer to nitrocellulose membranes (Amersham). The following dilutions of antisera in PBS-T (PBS + 0.1% Tween) plus 5% dried skimmed milk were used: 1:5000 rabbit anti-CheW-1; 1:200 rabbit anti-V. anguillarum flagellin (from Debra Milton, Umeå University) (O'Toole et al., 1997). The latter antiserum cross-reacts with the V. cholerae flagellin subunits, which share 82−90% identity at the amino acid level with the four V. anguillarum flagellin subunits. The V. cholerae flagellins have similar molecular weights and are not separated in this experiment. A 1:2000 dilution of 125I-Protein A (PerkinElmer) was used for detection followed by exposure to a phosphor screen (Kodak). Band intensities were quantified on a phosphorimager (Molecular Dynamics) and analysed with ImageQuant (Molecular Dynamics). CheW-1 levels were normalized to flagellin.

Competition experiments and ID50 determination

The competitive index (CI) was determined following a 24 h infection of 5-day-old CD-1 or Swiss Webster mice as described previously (Taylor et al., 1987). Each mouse was inoculated intragastrically with approximately 105 cfu of a 1:1 mixture of the strains to be tested. In vitro grown competing strains were grown overnight in LB at 37°C. Mice were euthanized at 24 h post inoculation and the small intestines were removed. Following homogenization in LB + 20% glycerol, serial dilutions of intestinal homogenates were plated on LB agar supplemented with Sm and Xgal. The CI is the ratio of LacZ to LacZ+ bacteria for the cheW and cheR paralogue mutant competition experiments, or LacZ+ to LacZ for stool competition experiments corrected for the input ratio.

For ID50 determinations, 5-day-old Swiss Webster mice in groups of five mice were inoculated intragastrically with doses ranging from ~1 to 104 cfu of stool V. cholerae or an overnight culture of AC-V999. Mice were euthanized at 24 h and the small intestines were homogenized and plated on LB agar supplemented with Sm to determine the number of cfu. Values ≥10 cfu (limit of detection) were recorded as positive for infection. In these experiments, 90% of mice scored as positive had ≥104 cfu per small intestine. The percentage of infected mice was plotted against the input dose to determine the ID50 as described by Reed and Muench (Reed and Muench, 1938).


This research was supported by NIH Grant AI55058 to A.C. and the Center for Gastroenterology Research on Absorptive and Secretory Processes, NEMC (P30 DK34928). A.C. is an investigator of the Howard Hughes Medical Institute.


  • Adler J. A method for measuring chemotaxis and use of the method to determine optimum conditions for chemotaxis by Escherichia coli. J Gen Microbiol. 1973;74:77–91. [PubMed]
  • Alam A, Larocque RC, Harris JB, Vanderspurt C, Ryan ET, Qadri F, Calderwood SB. Hyper-infectivity of human-passaged Vibrio cholerae can be modeled by growth in the infant mouse. Infect Immun. 2005;73:6674–6679. [PMC free article] [PubMed]
  • Allweiss B, Dostal J, Carey KE, Edwards TF, Freter R. The role of chemotaxis in the ecology of bacterial pathogens of mucosal surfaces. Nature. 1977;266:448–450. [PubMed]
  • Berg HC, Brown DA. Chemotaxis in Escherichia coli analysed by three-dimensional tracking. Nature. 1972;239:500–504. [PubMed]
  • Bina J, Zhu J, Dziejman M, Faruque S, Calderwood S, Mekalanos J. ToxR regulon of Vibrio cholerae and its expression in vibrios shed by cholera patients. Proc Natl Acad Sci USA. 2003;100:2801–2806. [PubMed]
  • Black WP, Yang Z. Myxococcus xanthus chemotaxis homologs DifD and DifG negatively regulate fibril polysaccharide production. J Bacteriol. 2004;186:1001–1008. [PMC free article] [PubMed]
  • Borkovich KA, Alex LA, Simon MI. Attenuation of sensory receptor signaling by covalent modification. Proc Natl Acad Sci USA. 1992;89:6756–6760. [PubMed]
  • Boukhvalova MS, Dahlquist FW, Stewart RC. CheW binding interactions with CheA and Tar. Importance for chemotaxis signaling in Escherichia coli. J Biol Chem. 2002;277:22251–22259. [PubMed]
  • Bourret RB, Stock AM. Molecular information processing: lessons from bacterial chemotaxis. J Biol Chem. 2002;277:9625–9628. [PubMed]
  • Bourret RB, Charon NW, Stock AM, West AH. Bright lights, abundant operons – fluorescence and genomic technologies advance studies of bacterial locomotion and signal transduction: review of the BLAST meeting, Cuernavaca, Mexico, 14–19 January 2001. J Bacteriol. 2002;184:1–17. [PMC free article] [PubMed]
  • Butler SM, Camilli A. Both chemotaxis and net motility greatly influence the infectivity of Vibrio cholerae. Proc Natl Acad Sci USA. 2004;101:5018–5023. [PubMed]
  • Camilli A, Mekalanos JJ. Use of recombinase gene fusions to identify Vibrio cholerae genes induced during infection. Mol Microbiol. 1995;18:671–683. [PMC free article] [PubMed]
  • Camilli A, Beattie D, Mekalanos J. Use of genetic recombination as a reporter of gene expression. Proc Natl Acad Sci USA. 1994;91:2634–2638. [PubMed]
  • Cash RA, Music SI, Libonati JP, Snyder MJJ, Wenzel RP, Hornick RB. Response of man to infection with Vibrio cholerae. I. Clinical, serologic, and bacteriologic responses to a known inoculum. J Infect Dis. 1974;129:45–52. [PubMed]
  • Cornick NA, Helgerson AF. Transmission and infectious dose of Escherichia coli O157:H7 in swine. Appl Environ Microbiol. 2004;70:5331–5335. [PMC free article] [PubMed]
  • Donnenberg MS, Kaper JB. Construction of an eae deletion mutant of enteropathogenic Escherichia coli by using a positive-selection suicide vector. Infect Immun. 1991;59:4310–4317. [PMC free article] [PubMed]
  • Faruque SM, Chowdhury N, Kamruzzaman M, Dziejman M, Rahman MH, Sack DA, et al. Genetic diversity and virulence potential of environmental Vibrio cholerae population in a cholera-endemic area. Proc Natl Acad Sci USA. 2004;101:2123–2128. [PubMed]
  • Faruque SM, Naser IB, Islam MJ, Faruque AS, Ghosh AN, Nair GB, et al. Seasonal epidemics of cholera inversely correlate with the prevalence of environmental cholera phages. Proc Natl Acad Sci USA. 2005;102:1702–1707. [PubMed]
  • Freter R, O'Brien PC. Role of chemotaxis in the association of motile bacteria with intestinal mucosa: fitness and virulence of nonchemotactic Vibrio cholerae mutants in infant mice. Infect Immun. 1981a;34:222–233. [PMC free article] [PubMed]
  • Freter R, O'Brien PC. Role of chemotaxis in the association of motile bacteria with intestinal mucosa: chemotactic responses of Vibrio cholerae and description of motile nonchemotactic mutants. Infect Immun. 1981b;34:215–221. [PMC free article] [PubMed]
  • Freter R, Allweiss B, O'Brien PC, Halstead SA, Macsai MS. Role of chemotaxis in the association of motile bacteria with intestinal mucosa: in vitro studies. Infect Immun. 1981;34:241–249. [PMC free article] [PubMed]
  • Gill DM. Mechanism of action of cholera toxin. Adv Cyclic Nucleotide Res. 1977;8:85–118. [PubMed]
  • Gosink KK, Kobayashi R, Kawagishi I, Häse CC. Analyses of the roles of the three cheA homologs in chemotaxis of Vibrio cholerae. J Bacteriol. 2002;184:1767–1771. [PMC free article] [PubMed]
  • Goy MF, Springer MS, Adler J. Failure of sensory adaptation in bacterial mutants that are defective in a protein methylation reaction. Cell. 1978;15:1231–1240. [PubMed]
  • Hanahan D. Studies on transformation of Escherichia coli with plasmids. J Mol Biol. 1983;166:557–580. [PubMed]
  • Heidelberg JF, Eisen JA, Nelson WC, Clayton RA, Gwinn ML, Dodson RJ, et al. DNA sequence of both chromosomes of the cholera pathogen Vibrio cholerae. Nature. 2000;406:477–483. [PubMed]
  • Heydorn A, Ersboll B, Kato J, Hentzer M, Parsek MR, Tolker-Nielsen T, et al. Statistical analysis of Pseudomonas aeruginosa biofilm development: impact of mutations in genes involved in twitching motility, cell-to-cell signaling, and stationary-phase sigma factor expression. Appl Environ Microbiol. 2002;68:2008–2017. [PMC free article] [PubMed]
  • Hyakutake A, Homma M, Austin MJ, Boin MA, Hase CC, Kawagishi I. Only one of the five CheY homologs in Vibrio cholerae directly switches flagellar rotation. J Bacteriol. 2005;187:8403–8410. [PMC free article] [PubMed]
  • Klose KE. The suckling mouse model of cholera. Trends Microbiol. 2000;8:189–191. [PubMed]
  • Lee SH, Butler SM, Camilli A. Selection for in vivo regulators of bacterial virulence. Proc Natl Acad Sci USA. 2001;98:6889–6894. [PubMed]
  • Manson MD, Armitage JP, Hoch JA, Macnab RM. Bacterial locomotion and signal transduction. J Bacteriol. 1998;180:1009–1022. [PMC free article] [PubMed]
  • Merrell DS, Butler SM, Qadri F, Dolganov NA, Alam A, Cohen MB, et al. Host-induced epidemic spread of the cholera bacterium. Nature. 2002;417:642–645. [PMC free article] [PubMed]
  • O'Toole R, Milton DL, Horstedt P, Wolf-Watz H. RpoN of the fish pathogen Vibrio (Listonella) anguillarum is essential for flagellum production and virulence by the water-borne but not intraperitoneal route of inoculation. Microbiology. 1997;143:3849–3859. [PubMed]
  • Reed LJ, Muench H. A simple method of estimating fifty per cent endpoints. Am J Hyg. 1938;27:493–497.
  • Sjoblad RD, Emala CW, Doetsch RN. Invited review: bacterial flagellar sheaths: structures in search of a function. Cell Motil. 1983;3:93–103. [PubMed]
  • Springer WR, Koshland DE., Jr Identification of a protein methyltransferase as the cheR gene product in the bacterial sensing system. Proc Natl Acad Sci USA. 1977;74:533–537. [PubMed]
  • Sun H, Zusman DR, Shi W. Type IV pilus of Myxococcus xanthus is a motility apparatus controlled by the frz chemosensory system. Curr Biol. 2000;10:1143–1146. [PubMed]
  • Taylor RK, Miller VL, Furlong DB, Mekalanos JJ. Use of phoA gene fusions to identify a pilus colonization factor coordinately regulated with cholera toxin. Proc Natl Acad Sci USA. 1987;84:2833–2837. [PubMed]
  • Vlamakis HC, Kirby JR, Zusman DR. The Che4 pathway of Myxococcus xanthus regulates type IV pilus-mediated motility. Mol Microbiol. 2004;52:1799–1811. [PubMed]
  • Wadhams GH, Armitage JP. Making sense of it all: bacterial chemotaxis. Nat Rev Mol Cell Biol. 2004;5:1024–1037. [PubMed]
  • Watnick PI, Fullner KJ, Kolter R. A role for the mannose-sensitive hemagglutinin in biofilm formation by Vibrio cholerae El Tor. J Bacteriol. 1999;181:3606–3609. [PMC free article] [PubMed]
  • Whitchurch CB, Leech AJ, Young MD, Kennedy D, Sargent JL, Bertrand JJ, et al. Characterization of a complex chemosensory signal transduction system which controls twitching motility in Pseudomonas aeruginosa. Mol Microbiol. 2004;52:873–893. [PubMed]
  • Wiles S, Dougan G, Frankel G. Emergence of a ‘hyperinfectious’ bacterial state after passage of Citrobacter rodentium through the host gastrointestinal tract. Cell Microbiol. 2005;7:1163–1172. [PubMed]
  • Yildiz FH, Schoolnik GK. Role of rpoS in stress survival and virulence of Vibrio cholerae. J Bacteriol. 1998;180:773–784. [PMC free article] [PubMed]
  • Yonekawa H, Hayashi H, Parkinson JS. Requirement of the cheB function for sensory adaptation in Escherichia coli. J Bacteriol. 1983;156:1228–1235. [PMC free article] [PubMed]