INTRODUCTION

Top Abstract Introduction Materials and Methods Results and Discussion References

The curved, highly motile gram-negative rod Vibrio cholerae is the causative agent of the severe and sometimes lethal diarrheal disease cholera. Human biopsy studies (17) as well as animal models (15, 16) indicate that V. cholerae is a noninvasive pathogen. During intestinal colonization, V. cholerae secretes cholera toxin, an A-B subunit type toxin which catalyzes the transfer of an ADP-ribose from NAD to a GTP-binding regulatory component of adenylate cyclase in enterocytes (18). The signs and symptoms of cholera can largely be reproduced by the administration of cholera toxin to human volunteers (25), suggesting that the activity of this enterotoxin primarily accounts for the clinical manifestations of V. cholerae infection. The bacterial properties which facilitate this organism's capacity to survive and multiply in the small intestine, the principal locus of V. cholerae intestinal colonization, are incompletely understood.

A variety of animal models (31) and bacterial genetic screens have been used to study V. cholerae intestinal colonization factors. The most commonly used animal model is the suckling mouse. Unlike V. cholerae isolates in adult mice, isolates orally administered to 3- to 5-day-old mice colonize the intestine and cause fluid accumulation. Adult animals can be experimentally infected by V. cholerae either through the use of antibiotics to clear most of the normal flora prior to infection or through the use of surgical ties on the small bowel (e.g., the adult rabbit RITARD model) (30). Taylor et al. demonstrated that toxin-coregulated pili (TCP), a type 4 pilus whose expression is coregulated with cholera toxin, are required for colonization of the suckling mouse small bowel (33). Subsequent studies in human volunteers have established the requirement for TCP for V. cholerae human intestinal colonization (21). Although there may be substantive differences between the gastrointestinal (GI) tract of a suckling mouse and that of an adult human, the above-described finding lends validity to the use of the suckling mouse as a model for the study of V. cholerae intestinal colonization.

The suckling mouse model has been used to identify several other V. cholerae colonization factors. Prior to the recombinant DNA era, Baselski and colleagues determined that spontaneous lipopolysaccharide (LPS) rough V. cholerae strains are severely defective in colonization of the suckling mouse small intestine (4, 6). More recently, V. cholerae strains harboring mutations in wbf genes encoding the V. cholerae O1 O antigen (OA) (10, 23, 34) have been constructed and found to be severely defective in intestinal colonization, although the mechanism explaining this finding is unknown. Other V. cholerae gene products which have been shown to be important for colonization of the infant mouse include accessory colonization factors (13, 22, 29), a cell-associated mannose-fucose-resistant hemagglutinin (14), and metabolic factors, such as iron (20) and magnesium (10) transport proteins and arginine, purine, and biotin biosynthesis enzymes (6, 8, 10).

More than 20 years ago, Baselski and Parker studied the distribution of V. cholerae after oral infection of infant mice by using a radioactive tracer (5). They found that the capacity of different strains to colonize the infant mouse upper bowel differed significantly among strains and that the capacity to colonize the upper bowel is essential for the establishment of an infection. Although the suckling mouse model is currently commonly used to study V. cholerae pathogenesis, there has not been much attention paid to further characterization of the suckling mouse V. cholerae colonization model since the pioneering studies of Baselski and Parker. For example, the location of cells within different segments of the small bowel has not been addressed, nor has the intestinal distribution of El Tor biotype V. cholerae strains, the principal cause of cholera in the world at present, been studied. In this study, we have investigated the details of the population dynamics of a wild-type El Tor strain and a classical V. cholerae strain along the entire GI tract at multiple time points during the course of suckling mouse infection. In addition, the population dynamics of tcpA and wbf mutants of the El Tor and classical strains along the entire GI tract were determined.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results and Discussion References

Bacterial strains. The strains used in this study are listed in Table 1. The N16961 strain was a spontaneous streptomycin-resistant (Sm^r) mutant derived from the clinical isolate N16961 from Bangladesh (24). Competition experiments showed that the Sm^r N16961 strain was as fit for growth in vitro and as virulent in the suckling mouse model of cholera as the parental strain (data not shown). Strain NTCP was constructed by transduction of the tcpA::mTn5 allele from SC338 (10) into N16961 with a temperature-sensitive mutant of the generalized transducing phage CP-T1 (19, 27). The mTn5 insertion in tcpA in one transductant, designated NTCP, was confirmed by Southern blot analysis (data not shown). The N16961 wbfB::pSC95 strain, MA393, was made by plasmid integration of pSC95 into the wbfB locus. This plasmid contains an internal fragment of wbfB inserted into the suicide vector pGP704 (10, 26). To construct MA393, SM10pir(pSC95) was mated on Luria-Bertani (LB) (12) agar plates with N16961. Exconjugants were then selected as streptomycin- and ampicillin-resistant colonies. Southern blot analysis was used to confirm the integration of pSC95 into wbfB in MA393 (data not shown).

Intestinal colonization assay. A modified version of the method of Baselski and Parker (5) was used for infection and recovery of V. cholerae from the suckling mouse intestine. Briefly, 4- to 5-day-old suckling CD-1 mice were separated from their mothers 1 h prior to inoculation with V. cholerae. Then, the mice were intragastrically inoculated with cultures of V. cholerae that had been grown overnight and then diluted in LB broth. The cells for the inocula were grown at 30°C in LB broth supplemented with 50 µg of streptomycin per ml (for all strains), 50 µg of ampicillin per ml (for MA393), and 30 µg of kanamycin per ml (for NTCP). Each overnight culture was diluted 1:1,000 in LB containing 8 µl of blue food-coloring dye per ml. Each mouse was then inoculated with 50 µl of the diluted culture as previously described (7). The bacterial titers in each inoculum were determined by plating serial dilutions of the inocula on the appropriate plates. Infected mice were kept at 26°C in the absence of their mothers. Mice were sacrificed at the designated time points, and the small and large bowels as well as the ceca were removed. The small bowel was cut into three segments of equal length, designated the proximal, middle, and distal small intestines. The cecum and large bowel, extending all the way to the rectum, were also separated. Each of these five segments was then mechanically homogenized in 4.5 ml of LB containing 20% (vol/vol) glycerol with a Tissue Tearor (Biospec Products, Bartlesville, Okla.), and serial dilutions were plated onto LB agar supplemented with 100 µg of streptomycin per ml to enumerate V. cholerae CFU per segment. The detection limit for this assay is 150 CFU/segment.

	RESULTS AND DISCUSSION

Top Abstract Introduction Materials and Methods Results and Discussion References

Wild-type V. cholerae infection profiles. Determination of the population dynamics of wild-type El Tor and classical V. cholerae strains along the entire length of the GI tracts of infected suckling mice describes an important parameter of the host-pathogen interaction and establishes a baseline for comparison of the colonization properties of mutant V. cholerae strains. For the studies of wild-type V. cholerae intestinal population dynamics, we chose El Tor strain N16961 because it is the subject of the ongoing V. cholerae genome project and classical strain O395 because it has been the subject of many previous investigations. Either N16961 or O395 was intragastrically inoculated into suckling mice. Then, at various times after inoculation, the GI tract was dissected and the total number of V. cholerae CFU in each segment was determined. At each time point tested, the total number of recoverable CFU represents the sum of the cells present in the input and their progeny minus the number of cells that were killed or excreted in the stool. Figures 1 and 2 depict the infection profiles for N16961 and O395, respectively. Panel A in each figure shows the percent recovery of input from the entire GI tract, excluding the stomach. In preliminary experiments, we found that the stomach contained no detectable CFU at the time points tested. Panel B in each figure shows the percent recovery of input from the small bowel only. Finally, panel C in each figure shows the number of V. cholerae CFU in various segments of the GI tract, including the proximal, middle, and distal small bowels, the cecum, and the large bowel. Panels A and B show the average percent recovery from multiple experiments. However, due to the variability in absolute numbers of CFU recovered from intestinal segments in different experiments, the data shown in panels C are from single experiments which are representative of multiple experiments performed.

View larger version (44K): [in this window] [in a new window]   FIG. 1.   Wild-type V. cholerae N16961 (El Tor biotype) recovery from GI tracts of suckling mice over time. Five-day-old CD-1 mice were intragastrically inoculated with between 105 and 107 CFU as described in Material and Methods. Intestinal segments were aseptically removed and processed for bacterial quantification. (A) Percent CFU recovery from entire GI tract. For each time point, CFU from all segments were added and expressed as a percentage of the inoculum given to the mice at time zero. The averages and standard deviations of CFU recovered from four mice are shown. (B) Percent CFU recovery from small bowel. (C) Number of V. cholerae CFU recovered from indicated segments at various times postinoculation. The data are the averages from two mice from a single experiment in which the inoculum was 3 × 105 CFU. Columns with asterisks above them were below the limit of detection for this assay (150 CFU).

View larger version (40K): [in this window] [in a new window]   FIG. 2.   Wild-type V. cholerae O395 (classical biotype) recovery from GI tracts of suckling mice. (A) CFU recovery as a percentage of input (between 2 × 105 and 3 × 106) from entire GI tracts of four mice as described in the legend for Fig. 1A. (B) CFU recovery as a percentage of input from small bowel. (C) CFU per segment from a pair of mice inoculated with 2 × 105 CFU, as described in the legend for Fig. 1C.

View larger version (37K): [in this window] [in a new window]   FIG. 3.   Wild-type V. cholerae O395 (classical biotype) recovery from segments of the small bowel of a suckling mouse. The mouse was intragastrically inoculated with 3 × 106 CFU, and CFU were recovered from 1-cm segments proceeding from the pylorus (x-axis origin) toward the cecum 10 h after infection. The numbers of CFU recovered from three of the segments are shown above their bars in parentheses.

TcpA mutant strain infection profiles. TcpA is the repeated polypeptide subunit of TCP and is an essential V. cholerae intestinal colonization factor. In competition experiments, when a tcpA strain is coinoculated with a wild-type strain, the competitive index observed is typically 10³ to 10⁴ in the suckling mouse small bowel; i.e., the tcpA strain is outcompeted approximately 1,000- to 10,000-fold by the wild type over a 24-h period (1, 30, 33). It is not known if tcpA strains survive and replicate within the small bowel when present as the sole inoculum or in the large intestine (which is usually not assayed in competition experiments). To begin to assess this, we infected suckling mice with tcpA derivatives of N16961 and O395 and observed the population dynamics of these cells at the time points we used to study the population dynamics of the wild-type strains. Figures 4 and 5 represent infection profiles of tcpA derivatives of N16961 and O395, respectively.

View larger version (44K): [in this window] [in a new window]   FIG. 4.   V. cholerae N16961 tcpA recovery from GI tracts of suckling mice. (A) CFU recovery as a percentage of input (between 3 × 105 and 4 × 106) from entire GI tracts of four mice as described in the legend for Fig. 1A. (B) CFU recovery as a percentage of input from small bowel. (C) CFU per segment from a pair of mice inoculated with 4 × 106 CFU, as described in the legend for Fig. 1C.

View larger version (42K): [in this window] [in a new window]   FIG. 5.   V. cholerae O395 tcpA recovery from GI tracts of suckling mice. (A) CFU recovery as a percentage of input (between 5 × 105 and 5 × 106) from entire GI tracts of five mice, as described in the legend for Fig. 1A. (B) CFU recovery as a percentage of input from small bowel. (C) CFU per segment from a pair of mice inoculated with 5 × 105 CFU, as described in the legend for Fig. 1C.

3

LPS mutant strain infection profiles. The V. cholerae wbfB and wbfD (formerly rfbB and rfbD) genes are part of an operon that is required for biosynthesis of perosamine, the repeated carbohydrate moiety of the LPS O1 OA (32, 35). Strains harboring mutations within this operon lack OA and have been shown to be severely attenuated for colonization in suckling mice (23). Figures 6 and 7 depict the infection profiles for N16961 wbfB (strain MA393) and O395 wbfD (strain O395R-1), respectively. As with the tcpA strains, there is no net growth of either wbf strain after 24 h (Fig. 6A and 7A); however, during the 10- to 24-h interval, unlike the tcpA strains, these wbf strains did grow to a limited extent in the small intestine and perhaps in the large bowel (Fig. 6B and C and 7B and C). Since the data in Fig. 4 and 5 indicate that TCP is required for small intestinal colonization, the observation that there is some small bowel colonization by these two OA strains suggests that there is at least some functional TCP production by the latter strains. This finding contradicts a recent report (23) suggesting that V. cholerae strains lacking OA are defective in assembling TCP on the cell surface.

View larger version (42K): [in this window] [in a new window]   FIG. 6.   V. cholerae N16961 wbfB recovery from GI tracts of suckling mice. (A) CFU recovery as a percentage of input (between 2 × 105 and 6 × 106) from entire GI tracts of two mice, as described in the legend for Fig. 1A. (B) CFU recovery as a percentage of input from small bowel. (C) CFU per segment from a pair of mice inoculated with 3 × 106 CFU, as described in the legend for Fig. 1C.

View larger version (38K): [in this window] [in a new window]   FIG. 7.   V. cholerae O395 wbfD recovery from GI tracts of suckling mice. (A) CFU recovery as a percentage of input (between 2 × 105 and 1 × 106) from entire GI tracts of four mice, as described in the legend for Fig. 1A. (B) CFU recovery as a percentage of input from small bowel. (C) CFU per segment from a pair of mice inoculated with 2 × 105 CFU, as described in the legend for Fig. 1C.

Conclusions. Our studies characterizing the infection profiles of wild-type El Tor and classical biotype V. cholerae strains should prove useful as a baseline for future studies of V. cholerae strains harboring mutations in virulence genes, as well as for studies of host factors which influence colonization. For example, it is interesting that in suckling mice, like in humans, the middle small bowel is the main site of colonization by V. cholerae (2). Is this reflective of the distribution of the hypothesized but hitherto unidentified TCP receptor? An alternative explanation which warrants investigation is that maximal colonization of the middle small bowel may simply reflect the requirement for a short period of adaptation within the host GI tract, in this case occurring during passage through the stomach and proximal small bowel, for proper induction of V. cholerae colonization factors.