A possible association between CFA (I) fimbriae and K99 fimbriae on Escherichia coli and bacterial adherence to human lymphocytes

We have explored a possible association between Escherichia coli binding to human lymphocytes and plasmid coded fimbriae on the bacterial surface. E. coli with or without the plasmid coded membrane CFA(I), K99 and K88 were mixed with freshly-drawn human peripheral blood lymphocytes. When the lymphocytes were mixed with E. coli possessing the CFA(I) fimbriae, 59% of the lymphocytes bound bacteria onto the surface, whereas only 22% of the lymphocytes bound the CFA(I)- derivative. The lymphocytes bound 53% and 56% of two K9+ strains, whereas 22% and 8% of the lymphocytes adhered the same strains without the K99 fimbriae. Twelve per cent and 7% of lymphocytes bound bacteria when the strain was K88+ or K88-, respectively. Likewise a low (8%) adherence to lymphocytes was found when the E. coli did not possess fimbriae or flagella.     

Characterization of colonization factor antigen CFA/I from an enterotoxigenic Escherichia coli strain (0128:H12)

CFA/I antigen was isolated and purified from E. coli, mutant 279 B-1-14, serotype 0128:H12, and had the following biochemical and biological features: a) amino-acid content was similar to that of purified antigen prepared from strain H10407; b) latex particles sensitization with purified CFA/I antigen produced bovine and human erythrocytes group A/II hemagglutination in carbohydrates presence; c) purified anti-CFA/I specific antibodies agglutinated CFA/I-positive enterotoxigenic E. coli strains; d) 3H-leucine-labelled CFA/I antigen adhered to rabbits intestinal mucosa at significant values; e) intestinal mucosa pretreating with purified CFA/I antigen, followed by 3H-leucine labelled enterotoxigenic bacteria infection, had a least 3 local effects: 1) intestinal mucosa protection against parental enterotoxigenic bacteria; 2) inhibition of CFA/I-positive bacteria adherence to intestinal mucosa; 3) release of approximately 96% intraluminally inoculated bacteria.     

Maintenance of broad host range vector pKT230 in marine unicellular cyanobacteria

Clinical isolate of Vibrio mimicus were examined for production of cell-associated hemagglutinin (HA) and pili and for adherence to formalin-fixed human intestinal mucosa. V. mimicus grown on CFA agar for 3 h at 37 degrees C possessed HA and adhered better to the mucus layer than to the epithelial cell surface. A significant correlation was found between the HA titers and adherence ability to the epithelial cell surface of villi (P less than 0.05); adherence to the ileal lymphoid follicle-associated epithelium occurred at higher levels. In contrast, V. mimicus grown on CFA agar for 20 h at 37 degrees C exhibited lower levels of HA and reduced adherence ability. The production of pili was more pronounced after 20 h of incubation than after 3 h of incubation. In comparison with V. cholerae 01 and V. cholerae non-01 cultured under similar conditions, V. mimicus showed inferior adherence, but with similar HA production or piliation.     

The role of CFA/I in adherence and toxin delivery by ETEC expressing multiple colonization factors in the human enteroid model

Enterotoxigenic Escherichia coli (ETEC) is a primary causative agent of diarrhea in travelers and young children in low-to-middle-income countries (LMICs). ETEC adhere to intestinal epithelia via colonization factors (CFs) and secrete heat-stable toxin (ST) and/or heat-labile toxin (LT), causing dysregulated cellular ion transport and water secretion. ETEC isolates often harbor genes encoding more than one CF that are targets as vaccine antigens. CFA/I is a major CF that is associated with ETEC that causes moderate-to-severe diarrhea and plays an important role in pathogenesis. The Global Enteric Multicenter Study finding that 78% of CFA/I-expressing ETEC also encode the minor CF CS21 prompted investigation of the combined role of these two CFs. Western blots and electron microscopy demonstrated growth media-dependent and strain-dependent differences in CFA/I and CS21 expression. The critical role of CFA/I in adherence by ETEC strains expressing CFA/I and CS21 was demonstrated using the human enteroid model and a series of CFA/I- and CS21-specific mutants. Furthermore, only anti-CFA/I antibodies inhibited adherence by global ETEC isolates expressing CFA/I and CS21. Delivery of ST and resulting cGMP secretion was measured in supernatants from infected enteroid monolayers, and strain-specific ST delivery and time-dependent cGMP production was observed. Interestingly, cGMP levels were similar across wildtype and CF-deficient strains, reflecting a limitation of this static aerobic infection model. Despite adherence by ETEC and delivery of ST, the enteroid monolayer integrity was not disrupted, as shown by the lack of decrease in transepithelial electrical resistance and the lack of IL-8 cytokines produced during infection. Taken together, these data demonstrate that targeting CFA/I in global clinical CFA/I-CS21 strains is sufficient for adherence inhibition, supporting a vaccine strategy that focuses on blocking major CFs. In addition, the human enteroid model has significant utility for the study of ETEC pathogenesis and evaluation of vaccine-induced functional antibody responses.     

Phenotypic and genetic features of Escherichia coli strains showing simultaneous expression of localized and diffuse adherence

We have previously shown that some Escherichia coli strains isolated from children with diarrhea present the so-called 'localized and diffused adherence (LA/DA) pattern' in which both localized adherence (LA) and diffused adherence (DA) are expressed simultaneously. In the present study, we show that the LA adherence of these strains is genetically and phenotypically similar to that so far described for enteropathogenic E. coli (EPEC) as determined by DNA hybridization and electron microscopy. On the other hand, the DA is encoded by genes not homologous to the DAEC or AIDA-I DNA probes. In addition, the LA/DA strains are able to invade eukaryotic cells both in vitro and in vivo. In the rabbit ileal loop assay their invasion capacity goes beyond the enterocyte and reaches the muscularis mucosae as determined by transmission electron microscopy. These findings suggest that the LA/DA adherence pattern may be linked to a new E. coli virulence category which in the case of the strains studied may be associated to other virulence traits that enable them to more deeply invade the intestinal mucosa.     

The binding of colonization factor antigens of enterotoxigenic Escherichia coli to intestinal cell membrane proteins

We examined the binding of colonization factor antigens (CFAs) of enterotoxigenic Escherichia coli to electrophoretically separated membrane components of rabbit intestinal brush borders or human intestinal (and non-intestinal) cell lines using an immunoblotting technique. Both CFA/I and CFA/II bound to distinct membrane components which seemed to be identical in rabbit brush borders and in a human intestinal cell line; these binding structures were mainly missing in membranes from epithelial cell lines of non-intestinal origin. Both shared and specific binding components were identified for CFA/I and the different subcomponents of CFA/II (CS1, CS2 and CS3), respectively. Chloroform-methanol extraction of lipids from the cell membranes did not change the binding pattern for either CFA/I or CFA/II suggesting that the binding occurred to (glyco)proteins rather than to (glyco)lipids.     

Detection of attachment of enterotoxigenic Escherichia coli (ETEC) to human small intestinal cells by enzyme immunoassay

Abstract Simple immunoassays were developed to study the binding between enterocytes of the small intestine and other cell types, and enterotoxigenic Escherichia coli (ETEC). CFA/I or CFA/II pilus protein or CFA-positive E. coli bacteria were wells of microtitre plates and incubated with vesicles or crude mucus prepared from human brush border enterocytes. Binding of the cell preparations was detected by adding specific rabbit anti-brush border IgG followed by urease-labelled goat anti-rabbit IgG and urea substrate. The binding of purified CFA/I to human or rabbit small intestine, human oral epithelial cells or Caco-2 cells was detected with specific anti-CFA/I IgG. Both human brush border and mucus-derived preparations were able to attach to ETEC. The binding was CFA-specific and strong enough to withstand several washings. In contrast, CFA/I did not bind to small intestinal cells of non-human small intestinal origin, indicating that there may be important differences in affinity between receptors present on human small intestinal cells and cells of non-human small intestinal origin. Antibodies directed against human small intestinal and non-small intestinal cells did not cross-react with either preparation, indicating that receptors between these different cell sources are different. The EIA proved useful during the identification of a newly-recognised 15 kDa bacterial surface component of ETEC strain H10407P, which may function as a putative attachment factor. The EIAs developed in this study were easy to perform and multiple tests could be performed on small samples, including biopsy samples obtained during endoscopy.     

Comparative study of inhibition of mannose-resistant haemagglutination caused by CFA/I, CFA/II, K88 and K99-positive Escherichia coli strains

Abstract We have studied the inhibition of mannose-resistant haemagglutination (MRHA) caused by Escherichia coli strains with CFA/I, CFA/II, K88, K99 and by other faecal E. coli lacking these colonisation antigens, by means of 30 sugar compounds and by enzymatic treatment of erythrocytes with neuraminidase, α-mannosidase, β-galactosidase, trypsin and pronase, and with formaldehyde. Inhibition of MRHA by sugars was effective only in K88-positive strains with d (+)glucosamine, mucic acid and bovine submaxillary mucin. Enzymatic treatment and the formolisation of erythrocytes gave different results on MRHA activity in strains possessing each colonisation antigen type. Results suggest that the erythrocyte receptor for CFA/I and CFA/II may possibly be sialoglycoprotein in which N -acetylneuraminic acid (NANA) plays an important role, because MRHA activity in these strains was inhibited by treatment of erythrocytes with neuraminidase and pronase. On the other hand, erythrocyte receptors for K88 and K99, like receptors for haemagglutinins of faecal E. coli lacking these colonization antigens, may have other glycoconjugate structures in which proteins and NANA are not essential. Our observations also suggest that the nature (or structure) of the receptor for a specific colonisation antigen on diverse erythrocyte types may be different.     

Non-replicating oral whole cell vaccine protective against enterotoxigenic Escherichia coli (ETEC) diarrhea: stimulation of anti-CFA (CFA/I) and anti-enterotoxin (anti-LT) intestinal IgA and protection against challenge with ETEC belonging to heterologous serotypes

An oral killed (non-replicating) whole-cell anti-ETEC vaccine was prepared by treating enterotoxigenic Escherichia coli strain H-10407 (ST + LT +; 078: H11: CFA/I) with a 100%-lethal amount of colicin E2. Colicin E2 is a potent DNA endonuclease which enters the target bacterial cells without disrupting cellular integrity. Thus the vaccine consists of intact cells lacking chromosomal and plasmid DNA but possessing a normal complement of antigens, including CFA/I and enterotoxin(s), unaltered by chemical- or heat-treatment. Young healthy volunteers were administered two oral doses, one month apart, of approximately 3 x 10(10) vaccine cells. Of 22 vaccinees, 17 (77.3%) showed an intestinal anti-CFA/I IgA response and 19 (86.4%) showed an increase in intestinal anti-LT IgA. Twenty of 22 (90.9%) vaccinees had antibody responses to either CFA/I, LT, or both antigens, demonstrating that colicin E2-treated CFA-positive E. coli cells are an efficient vehicle in terms of delivery of antigens to the gut immune system. We previously demonstrated protection of vaccinees against challenge with the living homologous ETEC (strain H-10407). In this study, two groups of 8 vaccinees were challenged with a diarrheagenic dose of virulent ST + LT + ETEC of heterologous serotype; one group was challenged with a CFA/I-positive 063: H- strain and the other group was challenged with a CFA/II-positive 06: H16 strain. Approximately 75% efficacy was achieved in both challenge groups. None of the 16 vaccinees who had responded to both CFA/I and LT became ill upon challenge while both of the vaccinees who had not responded to either antigen did.(ABSTRACT TRUNCATED AT 250 WORDS)     

Competitive exclusion of diarrheagenic Escherichia coli (ETEC) from human enterocyte-like Caco-2 cells by heat-killed Lactobacillus

Diarrheagenic Escherichia coli (ETEC) bearing CFA/I or CFA/II adhesive factors specifically adhere onto the brush border of the polarized epithelial human intestinal Caco-2 cells in culture. Heat-killed Lactobacillus acidophilus strain LB, that adheres onto Caco-2 cells, inhibits diarrheagenic Escherichia coli adhesion in a concentration-dependent manner. Since the L. acidophilus does not express ETEC-CFA adhesive factors, it can be postulated that the heat-killed L. acidophilus LB cells inhibit diarrheagenic E. coli attachment by steric hindrance of the human enterocytic ETEC receptors.     

Prevalence of Colonization Factor Antigens (CFAs) and Adherence to HeLa Cells in Enterotoxigenic Escherichia coli Isolated from Feces of Children in São Paulo

Fifty-eight enterotoxigenic Escherichia coli (ETEC) strains, isolated from children with and without diarrhea in Sao Paulo, were examined for the presence of colonization factor antigens (CFAs) and their ability to adhere to HeLa cells. Antisera to CFA/I, the coli surface (CS) antigens CS1CS3, CS2CS3, and CS2 of CFA/II, CFA/III, and CS5CS6 and CS6 of CFA/IV were used. CFAs were identified in 43% of the ETEC strains: 40% of the strains with CFAs harbored CFA/I, 24% carried CFA/II (CS1CS3), 24% carried CFA/IV (CS6), and 12% carried CFA/IV (CS5CS6). CFAs occurred mainly among ETEC strains producing only heat-stable (ST-I) enterotoxin and in strains also producing heat-labile toxin (LT-I). No ETEC strains tested expressed CFA/III. A marked change in serotypes of ST-I-producing strains was found in Sao Paulo between 1979 and 1990. Adherence to HeLa cells was detected in 14% of the ETEC strains. All of them had a diffuse adherence pattern and produced only ST-I, and 88% carried CS6 antigen.     

Identification and characterization of a gene cluster mediating enteroaggregative Escherichia coli aggregative adherence fimbria I biogenesis.

The aggregative pattern of adherence (AA) exhibited by enteroaggregative Escherichia coli upon HEp-2 cells is a plasmid-associated property which correlates with aggregative adherence fimbria I (AAF/I) expression and human erythrocyte hemagglutination. By using cloning and mutagenesis strategies, two noncontiguous plasmid segments (designated regions 1 and 2) required for AA expression have previously been identified in enteroaggregative E. coli 17-2. TnphoA mutagenesis was performed on clones containing region 1, and 16 TnphoA mutants which were negative for the AA phenotype were analyzed. The TnphoA insertion site for each mutant was determined by junctional DNA sequencing. All 16 mutations occurred within a 4.6-kb span in region 1. Nucleotide sequence analysis of the region revealed four contiguous open reading frames, designated aggDCBA, in the same span. AA-negative TnphoA insertions into all open reading frames except aggB were obtained. On the basis of mutational analysis and protein homology data, it is inferred that aggA, aggC, and aggD are involved in biogenesis of AAF/I, encoding a major fimbrial subunit, outer membrane usher, and periplasmic fimbrial chaperone, respectively. By immunogold electron microscopy, polyclonal antiserum raised against the aggA gene product decorated AAF/I fimbriae, affirming that AggA encodes an AAF/I subunit.     

Adrenergic modulation of Escherichia coli O157:H7 adherence to the colonic mucosa

Enteric neurotransmitters can modulate the biodefensive functions of the intestinal mucosa, but their role in mucosal interactions with enteropathogens is not well defined. Here we tested the hypothesis that norepinephrine (NE) modulates interactions between enterohemorrhagic Escherichia coli O157:H7 (EHEC) and the colonic epithelium. Mucosal sheets from porcine distal colon were mounted in Ussing chambers. Drugs and an inoculum of either Shiga toxin-negative or -positive EHEC were added to the contraluminal and luminal bathing medium, respectively. After 90 min, adherent bacteria were quantified by an adherence assay and by immunohistochemical methods; short-circuit current (I(sc)) was measured continuously to assess changes in active ion transport. NE-treated tissues exhibited concentration-dependent increases in I(sc) and EHEC adherence. NE did not alter adherence of a rodent-adapted, noninfectious E. coli strain or two porcine-adapted non-O157 E. coli strains. The actions of NE on EHEC adherence but not I(sc) were prevented by the alpha-adrenergic antagonist yohimbine and the PKA activator Sp-8-bromoadenosine-3',5'-cyclic monophosphorothioate. Like NE, the PKA inhibitor Rp-8-bromoadenosine-3',5'-cyclic monophosphorothioate or indirectly acting sympathomimetic agents increased EHEC adherence. Nerve fibers immunoreactive for the NE-synthesizing enzymes tyrosine hydroxylase and dopamine beta-hydroxylase appeared to innervate the colonic epithelium. EHEC-like immunoreactivity on the colonic surface had the appearance of bacterial microcolonies and increased after NE treatment by a phentolamine-sensitive mechanism. Through interactions with alpha(2)-adrenergic receptors, NE appears to increase EHEC adherence to the colonic mucosa. Changes in sympathetic neural outflow may alter intestinal susceptibility to infection.     

CFA/I-ST plasmids: comparison of enterotoxigenic Escherichia coli (ETEC) of serogroups O25, O63, O78, and O128 and mobilization from an R factor-containing epidemic ETEC isolate

Colonization factor antigen I (CFA/I) plays an important role in the pathogenesis of diarrhea due to enterotoxigenic Escherichia coli. In this study, we examined 11 CFA/I+ enterotoxigenic E. coli from serogroups O25, O63, O78, and O128 and found that with all strains, spontaneous loss of CFA/I was associated with the loss of heat-stable toxin (ST) and with the loss of a single plasmid ranging in size from 54 to 60 megadaltons; when heat-labile toxin was lost, this was associated with the loss of another plasmid. The R factor of one strain, TX432 (O78:H12:CFA/I+; ST+), was found to mobilize the CFA/I-ST plasmid into E. coli K-12 at a frequency of 20%. These studies provide further evidence that CFA/I production is plasmid mediated in enterotoxigenic E. coli belonging to serogroups O25, O63, O78, and O128.     

Expression of cloned plasmid regions encoding colonization factor antigen I (CFA/I) in Escherichia coli

Molecular cloning from a plasmid encoding colonization factor antigen I (CFA/I) and heat-stable enterotoxin isolated two regions, 1 and 2, that are required for the production of CFA/I fimbriae. The level of CFA/I synthesis measured by ELISA was similar in an Escherichia coli K12 strain carrying regions 1 and 2 cloned separately on compatible plasmid vectors to that in the same strain containing the parental plasmid. The structural gene for the CFA/I fimbrial subunit was within region 1. This region directed production in E. coli minicells of at least six independent polypeptides, of which the fimbrial subunit and at least three others appeared to be synthesized as precursor molecules that underwent processing. Cloned DNA containing CFA/I region 2 specified three polypeptides in minicells. Attempts to reduce the size of the cloned region 1 resulted in a derivative plasmid that carried the CFA/I structural gene but did not complement a region-2 recombinant plasmid to restore production of CFA/I fimbriae.     

Cloning of regulator genes controlling fimbrial production by enterotoxigenic Escherichia coli

Sequences regulating production of fimbriae were cloned from two enterotoxigenic Escherichia coli strains. One cloned region, from E. coli 0.25.H42, controlled expression of coli surface-associated (CS) antigen 4, whereas the function of the other, from E. coli 0167.H5, was unclear. Both regulators were related to the cfaD gene that controls expression of colonization factor antigen I (CFA/I) although low stringency conditions were required to show significant hybridization between cfaD and the regulatory fragment from E. coli 0167. The cloned regulatory genes promoted expression of CFA/I, CS1, CS2 and CS4 antigens but the levels of production in the presence of the 0167 regulator were lower than those promoted by the CS4 regulator or cfaD.     

Correlation between the presence of aggregative fibria and a type of Escherichia coli adhesion

Enteroaggregative strains of E. coli (EAEC) are an important agents possessing among many virulence factors, aggregative fimbria AAF hemagglutinating in the presence of mannose human group A or/and rats erythrocytes. The aim of the study was to determine the correlation between the presence of AAF fimbria and pattern of adherence in vitro. Tested strains of E. coli were obtained from children with diarrhea (133 strains) and healthy children (105 strains). Among strains of E. coli from children with diarrhea 81 (61%) showed the presence of AAF fimbria and 19 (23%) were adhering in aggregative pattern. In the group of strains of E. coli isolated from healthy children 31 (30%) were AAF positive and 8 (25.8%) of them presented aggregative adherence. Examination of AAF fimbria only dose not allow to distinguish EAEC strains. The data showed the participation of EAEC strains in diarrhea of children below 3 years old.     

Precise and rapid assessment of Escherichia coli adherence to vaginal epithelial cells by flow cytometry

BACKGROUND: In the pathogenesis of Escherichia coli urinary tract infections (UTIs) in women, infecting bacteria adhere to vaginal and periurethral epithelial cells prior to ascending to the bladder and causing infection. Complex interactions among specific bacterial adhesins and various host factors appear to influence adherence of E. coli to mucosal surfaces such as the urogenital epithelium. To conduct population-based studies assessing host epithelial cell determinants that influence bacterial attachment, a method of measuring bacterial adherence utilizing clinically derived epithelial cell samples is needed. METHODS: We developed and standardized an efficient, accurate, high-throughput method for analyzing the adherence of uropathogenic E. coli to clinical samples containing a large number of exfoliated vaginal epithelial cells (VEC). Three wild-type E. coli strains isolated from women with UTI (IA2 expressing pap-encoded, class II fimbriae only; F24 expressing pap-encoded, class II and type 1 fimbriae; and F20, without pap-encoded or type I fimbriae) were transformed with gfpmut3, encoding green fluorescent protein, incubated with VECs, and analyzed by flow cytometry. RESULTS: Enumeration of the binding of each E. coli strain to 10,000 VECs showed reproducible, highly significant strain-dependent differences in adherence to VECs. Differential analysis of the relative contributions of type 1 pili and P fimbrial-mediated binding to the adherence phenotype was performed. It demonstrated that IA2 binding was dependent entirely on P fimbriae, whereas F24 binding was dependent on both P and type 1 fimbriae. CONCLUSIONS: This method has great potential for use in high-throughput analyses of clinically derived epithelial cell samples and will be valuable in population-based investigations of host-parasite interactions in UTI utilizing VECs collected from specific patient groups.     

Antibody response in mice immunized with a plasmid DNA encoding the colonization factor antigen I of enterotoxigenic Escherichia coli

The colonization factor antigen I (CFA/I) is one of the most epidemiologically relevant enterotoxigenic Escherichia coli (ETEC) fimbrial adhesins, which mediates the binding to human small intestine epithelium. A recombinant eukaryotic expression plasmid, pRECFA, encoding the CFA/I protein fused to the glycoprotein D of herpes simplex type 1 virus, was used to generate an antibody response in a murine model following intramuscular inoculation of purified DNA. Eukaryotic cells (BHK-21) transfected with pRECFA expressed the CFA/I protein in vitro, as revealed by Western blot and immunofluorescence microscopy. Administration of a single pRECFA 100-microg dose induced a long-term CFA/I-specific antibody response in BALB/c mice composed mainly of IgG and, to a lesser extent, IgA isotypes. The major CFA/I-specific IgG subclass was IgG2a, suggesting a Th-1-type immune response. A second dose with the same amount of purified DNA, given 2 weeks later, caused a booster effect on the immunoglobulin levels, but did not qualitatively alter the isotypes and subclasses of the induced antibody response. Immunization with different amounts of purified DNA and/or number of doses showed that maximal transient CFA/I-specific antibody levels could be obtained after two 100-microg doses of pRECFA given 2 weeks apart, but long-term antibody levels were similar.     

Stimulation of mucosal immune response following oral administration of enterotoxigenic Escherichia coli fimbriae (CFA/I) entrapped in liposomes in conjunction with inactivated whole-cell Vibrio cholerae vaccine

In this study, we have searched for an effective mucosal vaccine. An oral enterotoxigenic E. coli vaccine containing colonization factor antigen (CFA/I) associated with inactivated whole-cell V. cholerae vaccine (WCV) has been tested for safety and immunogenicity in animals. Five groups of animals were used. The results showed the following: (a) vaccine containing CFA/I antigen entrapped in liposomes and associated with WCV (batch C) had increased titers of specific antibodies to CFA/I antigen in 15 to 18 (83.3%) animals; (b) specific Peyer's patches (PP), lymph nodes (LN) and spleen (SPL) lymphocytes proliferation was detected following in vitro restimulation with CFA/I antigen or WCV. This response gradually increased to the highest value by the 35th postimmunization day. Moreover, lower PP, LN and spleen (SPL) proliferation was observed in rabbits receiving soluble CFA/I antigen (S-CFA/I) or free liposomes (F-L) alone; (c) adhesion of E. coli H10407 strain labelled with 3H-leucine in immunized and control animals revealed the following local effects: (i) protection of rabbit intestinal mucosa against virulent E. coli cells; (ii) inhibition of adhesion of ETEC bacteria to intestinal mucosa and (iii) significantly faster release of E. coli H 10407 strain labelled with 3H-leucine from the intestinal tract of immunized animals. The histopathological and electron microscope findings confirmed the above results. The experimental results point out an efficient protection against infection with E. coli strains (ETEC), after mucosal vaccination with CFA/I antigen entrapped in liposomes associated with inactivated whole-cell Vibrio cholerae as immunological adjuvant.