Prevalence of Virulence Genes and Cytolethal Distending Toxin Production in Campylobacter jejuni Isolates from Diarrheal Patients in Bangladesh

Campylobacter is a major cause of human bacterial gastroenteritis, and may be responsible for as many as 400 to 500 million cases worldwide each year (9). Although the genus Campylobacter is composed of 17 described species, human illness is associated primarily with Campylobacter jejuni and Campylobacter coli (17). Previous studies in Bangladesh conducted in the 1990s showed that C. jejuni is typically associated with predominantly watery diarrhea and isolated in frequencies varying from 17 to 26% (2, 12). Specific properties involved in adhesion, colonization, invasion, and toxin production appear necessary in the process of infection (15). In this study, flaA, cadF, racR, and dnaJ were selected as pathogenic genes responsible for the expression of adherence and colonization; virB11, ciaB, and pldA were selected as pathogenic genes responsible for the expression of invasion; cdtA, cdtB, and cdtC were selected as pathogenic genes responsible for the expression of toxin production; and wlaN was selected as a gene that is presumably involved in the expression of ganglioside mimics in Guillian-Barré syndrome (8). Another putative virulence gene, ceuE, encoding a lipoprotein (a component of a protein-binding-dependent transport system for the siderophore enterochelin) of C. jejuni was also selected in this study (23). In this study, the prevalence of Campylobacter species and their biochemical properties were determined. The prevalence of different pathogenic genes among C. jejuni strains was detected by PCR, and cytolethal distending toxin (CDT) activity was tested on HeLa cells.

Three hundred stool specimens collected from 2% of patients attending the ICDDR,B diarrhea treatment center under the systematic surveillance system (26) between January and December 2002 were examined. From these samples, 58 strains were isolated and presumptively identified as Campylobacter species by standard microbiological and biochemical methods (1, 7). Species identifications of C. jejuni and C. coli strains were confirmed by hippurate hydrolysis test and PCR (19).

Template DNAs for PCR were extracted by the conventional boiling method (8). Four sets of primers for cadF (14), cdtA (11), ceuE (10), and wlaN (18) were described earlier. The remaining eight sets of primers were designed by Datta et al. (8). All primers were purchased from Integrated DNA Technologies, Inc., Coralville, IA. All PCR assays were performed according to the procedure described earlier (27), but the cycling was as follows: denaturation at 94°C for 1 min, annealing at a temperature specific for the primer pairs (listed in Table 1) for 1 min, and extension at 72°C for 1 min.

Bacterial cell lysates were prepared and assayed for CDT activity according to the method described earlier (13, 24, 25), with minor modifications. Briefly, assays were performed with HeLa cells grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and an antibiotic supplement (Gibco, Invitrogen Corporation) containing penicillin, streptomycin, and glutamine. Ninety-six-well microtiter plates were seeded with 1 × 10⁴ to 2 × 10⁴ HeLa cells per well. After 18 h of incubation, 20 μl of the toxin preparation in 200 μl DMEM was applied to these wells. The plates were incubated at 37°C in a humidified 5% CO₂ incubator. Morphological changes of HeLa cells were studied at every 24 h for 4 days. Toxin titers were determined by performing twofold serial dilution of the sonic lysate in HeLa cell culture medium. Toxin titers were expressed as the reciprocal of the highest dilution that caused over 50% of the cells to be distended in a well. After 3 days, cellular distension and morphological changes were assessed by staining actin filaments with Alexa Fluor 488-conjugated phalloidin (Molecular Probes, Invitrogen) according to the method described earlier (5). The average cell size was determined by using the NIH Image J package (W. Rasband, NIH, Bethesda, MD; http://rsb.info.nih.gov/ij/download.html ).

All 58 strains showed morphological properties and biochemical reactions typical of Campylobacter species. These strains were oxidase and catalase positive and urease negative, did not ferment glucose, and were resistant to cephalothin. Of 58 isolates, 40 strains were identified as C. jejuni and 5 as C. coli by hippurate hydrolysis test and species-specific PCR for the presence of hipO and ask genes. The remaining 13 strains were classified as Campylobacter species. The amplicons of five pathogenic genes (flaA, cadF, racR, dnaJ, and pldA) were detected in all of the C. jejuni isolates tested. All C. jejuni strains except one (KC-1375) were positive for cdtA, cdtB, and cdtC genes. The prevalences of the ciaB and ceuE genes were 95% and 82.5%, respectively. The virB11 gene, which is found on the pVir plasmid (6), was not detected in any of the strains. A similar observation was obtained by Louwen et al. and Muller et al. (20, 21). This result is more likely as the majority of the strains contained no plasmid (K. Talukder, M. Aslam, Z. Islam, I. J. Azmi, M. A. Islam, D. K. Dutta, A. S. G. Faruque, Y. Kabir, A. Cravioto, and H. P. Endtz, unpublished data). The wlaN gene was detected in three (7.5%) strains only.

C. jejuni strains isolated in Bangladesh from patients with gastroenteritis caused morphological changes in cells of a tissue culture monolayer. In the HeLa cell assay, all but one (KC-1375) of the C. jejuni (97.5%) isolates produced CDT. When HeLa cells were treated with the sonic lysates, the cells did not show any characteristic changes in morphology within 24 h. However, after 48 h, the cells were enlarged, whereas control cells treated with phosphate-buffered saline continued to divide and grow to confluence. Cellular distension and morphological changes were more evident after staining with Alexa Fluor 488-conjugated phalloidin (Fig. 1). Measurement of cell size of 100 cells from each group was determined. The average diameter of the control cells was 475 ± 39 μm², whereas the average diameter of CDT-treated cells was 1,504 ± 119 μm². The CDT titer of the toxin positive strains ranged from 10 to 40.

CDT activity requires the function of three genes: cdtA, cdtB, and cdtC. CdtA, CdtB, and CdtC are all necessary for toxin production and form an active tripartite holotoxin that exhibits full cellular toxicity (16). Despite numerous reports on the toxic effects of CDT on different cultured mammalian cells, its role in pathogenesis is still unclear. Recently, Purdy et al. (25) have shown that C. jejuni isogenic cdt mutants have reduced invasiveness in a mouse model of enteric infection. An epidemiological study in Bangladesh showed a trend toward increased numbers of CDT-positive Escherichia coli cells in diarrheal patients compared to asymptomatic controls, but the difference did not reach statistical significance (3). In another report, a CDT-producing E. coli strain (O55:K59:H4) was isolated from the stool of a child suffering from gastroenteritis and encephalopathy (4). No other known toxins or virulence factors were detected from this strain, and no other viral or bacterial pathogens were isolated. Recent evidence from the use of a suckling mouse model suggests that the toxin causes secretory diarrhea and necrosis of colonic epithelium (22). CDT-mediated intestinal secretion in the mouse model was rapid, with diarrhea being evident within 4 h of toxin administration (22). In our study, all samples examined were obtained from diarrheal patients. Our observations suggest that CDT is a putative virulence factor in diarrhea caused by Campylobacter in Bangladesh and is responsible for a toxin-induced increase in intestinal fluid secretion.

Further studies in developing countries are necessary to further clarify the pathogenesis of the campylobacteriosis and the role of CDT.