Search

The Egg Products Inspection Act of 1970 requires that egg products in the U.S. must be pasteurized prior to release into commerce. The USDA Food Safety and Inspection Service (FSIS) is responsible for regulating egg products. Salmonellae are infrequently isolated from pasteurized egg products by food manufacturers or the FSIS and may be present as a result of either pasteurization-resistant bacteria or post-processing contamination. In this study, seventeen strains of Salmonella isolated from pasteurized egg products and three heat-resistant control strains were compared for the following attributes: thermal resistance in liquid whole egg (LWE) at 60 °C, enzymatic profiles, and serotyping and phage typing, antibiotic susceptibility, fatty acid analysis and strain morphological variation evaluated by scanning electron microscopy. Isolates were serotyped as Heidelberg (4 isolates), Widemarsh, Mbandaka, Cerro, Thompson, 4,12:i:-, and Enteritidis (8 isolates). All 20 isolates were sensitive to all 14 antibiotics tested for. The D60 values in LWE ranged from 0.34 to 0.58 min. All 20 strains were recovered from LWE inoculated with 8.5 logCFU/mL of Salmonella and pasteurized at 60 °C for 3.5 min; however, some isolates were not recovered from pasteurized LWE that had been inoculated with only 4.5 logCFU/mL Salmonella and treated at 60 °C for 3.5 min. Although some strains exhibited atypical enzymatic activity (e.g., reduction of adonitol, hydrolysis of proline nitroanilide or p-n-p-beta-glucuronide, and nonreduction of melibiose), differences in biochemical reactions could not be correlated with differences in thermal resistance. Furthermore, fatty acid analysis revealed that differences insaturate/unsaturated profiles may be correlated with differences in heat resistance, in two instances. One heat resistant strain (#13, Enteritidis) had the statistically lowest unsaturated/saturate ratio at 39%. However, one heat sensitive strain (#3, serovar 4,12:i:-) had the highest unsaturated/saturate ratio at 81%, and also the lowest concentration of stearic acid. This data represents the first steps in determining whether Salmonella contamination in pasteurized egg products may be the result of either thermally-resistant isolates or post-processing contamination. Contamination of LWE by Salmonella strains with higher heat resistance, (e.g., isolate #'s 2, 6, 10 and 12) may indicate the ability of Salmonella to survive pasteurization, while contamination of LWE strains with lower heat resistance (e.g., isolate #'s 1, 3, 5, 7, 8, 11, and 15) may indicate post-processing contamination of LWE by this foodborne pathogen., (Copyright © 2015. Published by Elsevier B.V.)

Increasingly, molecular methods have become important in identification and confirmation of bacteria at the species level. Rapid molecular methods provide sensitivity and specificity while reducing cost and resources. The primary goal of this study was to develop a real-time PCR assay for identification of Escherichia coli from an agar plate. GadE (gadE) directly regulates the glutamate-dependent acid response system (GDAR) in E. coli and is responsible for survival of at pH 2. Based on gene sequence data, a real-time PCR assay targeting gadE was developed for this purpose. Seventy bacterial isolates recovered from ground beef enrichments and 714 isolates from caecal contents were identified biochemically and tested with the real-time PCR assay developed in this study. The PCR assay and the biochemical identification had 100% agreement on the tested isolates. The gadE real-time PCR assay was demonstrated in this study to be an inexpensive, reliable method for confirming E. coli colonies within 1.5 h from an agar plate, thereby saving on final identification time., (Published 2014. This article is a U.S. Government work and is in the public domain in the USA.)

The U.S. Department of Agriculture, Food Safety Inspection Service has determined that six non-O157 Shiga toxin-producing Escherichia coli (STEC) serogroups (O26, O45, O103, O111, O121, and O145) are adulterants in raw beef. Isolate and phenotypic discrimination of non-O157 STEC is problematic due to the lack of suitable agar media. The lack of distinct phenotypic color variation among non-O157serogroups cultured on chromogenic agar poses a challenge in selecting colonies for confirmation. In this study, visible and near-infrared hyperspectral imaging and chemometrics were used to detect and classify non-O157 STEC serogroups grown on Rainbow agar O157. The method was first developed by building spectral libraries for each serogroup obtained from ground-truth regions of interest representing the true identity of each pixel and thus each pure culture colony in the hyperspectral agar-plate image. The spectral library for the pure-culture non-O157 STEC consisted of 2,171 colonies, with spectra derived from 124,347 of pixels. The classification models for each serogroup were developed with a k nearest-neighbor classifier. The overall classification training accuracy at the colony level was 99%. The classifier was validated with ground beef enrichments artificially inoculated with 10, 50, and 100 CFU/ml STEC. The validation ground-truth regions of interest of the STEC target colonies consisted of 606 colonies, with 3,030 pixels of spectra. The overall classification accuracy was 98%. The average specificity of the method was 98% due to the low false-positive rate of 1.2%. The sensitivity ranged from 78 to 100% due to the false-negative rates of 22, 7, and 8% for O145, O45, and O26, respectively. This study showed the potential of visible and near-infrared hyperspectral imaging for detecting and classifying colonies of the six non-O157 STEC serogroups. The technique needs to be validated with bacterial cultures directly extracted from meat products and positive identification of colonies by using confirmatory tests such as latex agglutination tests or PCR.

Non-O157 Shiga toxin-producing Escherichia coli (STEC) infections, particularly those caused by the "big six" or "top six" non-O157 serogroups (O26, O45, O103, O111, O121, and O145) can result in severe illness and complications. Because of their significant public health impact and the notable prevalence of STEC in cattle, methods for detection of the big six non-O157 STEC in ground beef have been established. Currently, the U.S. Department of Agriculture, Food Safety and Inspection Service detection methods for screening beef samples for non-O157 STEC target the stx(1), stx(2), and eae virulence genes, with the 16S rRNA gene as an internal control, in a real-time PCR multiplex assay. Further, the serogroup is determined by PCR targeting genes in the E. coli O-antigen gene clusters of the big six non-O157 serogroups. The method that we previously reported was improved so that additional stx variants, stx(1d), stx(2e), and stx(2g), are detected. Additionally, alignments of the primers targeting the eae gene were used to improve the detection assay so that eae subtypes that could potentially be of clinical significance would also be detected. Therefore, evaluation of alternative real-time PCR assay primers and probes for the stx and eae reactions was carried out in order to increase the stx and eae subtypes detected. Furthermore, a Tris-EDTA DNA extraction method was compared with a previously used procedure that was based on a commercially available reagent. The Tris-EDTA DNA extraction method significantly decreased the cycle threshold values for the stx assay (P < 0.0001) and eae assay (P < 0.0001), thereby increasing the ability to detect the targets. The use of different stx primers and probes increased the subtypes detected to include stx(1d), stx(2e), and stx(2g), and sequence data showed that modification of the eae primer should allow the known eae subtypes to be detected.

Latex agglutination assays utilizing polyclonal antibodies were developed for the top six non-O157 Shiga toxin-producing Escherichia coli (STEC) serogroups. Rabbit antisera were affinity purified through protein A/G columns, and the isolated immunoglobulins (IgGs) were covalently immobilized onto polystyrene latex particles. The resulting latex-IgG complex had a protein (IgG) load of 0.20 to 0.28 mg/ml in a 1% latex suspension. Optimum conditions for the agglutination assay consisted of utilizing 20 μm l of latex-IgG reagent containing 2.0 to 2.8 μm g IgG in a 0.5% latex suspension. Agglutination or flocculation was observed almost instantly after mixing the colonies with the latex-IgG, indicating STEC strains. More than 100 target and nontarget strains were tested in more than 3,000 test replicates. All target organisms produced positive results, but three antisera (anti-O26, anti-O103, and anti-O145) cross-reacted with some other STECs. The anti-O103 and anti-O145 latex reagents cross-reacted with O26 strains, and the anti-O26 cross-reacted with O103 strains. The latex-IgG reagents are stable for at least 1 year and are easy to prepare. These agglutination assays can be used for identification of presumptive non-O157 STEC colonies from agar media. The techniques used to prepare the latex reagents also can be utilized for testing other STEC serogroups, other E. coli serotypes, or other pathogens to ensure safe foods to consumers.

Six Shiga toxin-producing Escherichia coli (STEC) serogroups, which include O26, O45, O103, O111, O121, and O145, are responsible for the majority of non-O157 STEC infections in the United States, representing a growing public health concern. Cattle and other ruminants are reservoirs for these pathogens; thus, food of bovine origin may be a vehicle for infection with non-O157 STEC. Methods for detection of these pathogens in animal reservoirs and in food are needed to determine their prevalence and to develop intervention strategies. This study describes a method for detection of non-O157 STEC in ground beef, consisting of enrichment in modified tryptic soy broth at 42°C, followed by real-time multiplex polymerase chain reaction (PCR) assays targeting stx(1), stx(2), and genes in the O-antigen gene clusters of the six serogroups, [corrected] and then immunomagnetic separation (IMS) followed by plating onto Rainbow® Agar O157 and PCR assays for confirmation of isolates. All ground beef samples artificially inoculated with 1-2 and 10-20 CFU/25 g of ground beef consistently gave positive results for all of the target genes, including the internal amplification control using the multiplex real-time PCR assays after enrichment in modified tryptic soy broth for a total of 24 h (6 h at 37°C and 18 h at 42°C). The detection limit of the real-time multiplex PCR assays was ∼50 CFU per PCR. IMS for O26, O103, O111, and O145 was performed with commercially available magnetic beads, and the IMS beads for O45 and O121 were prepared using polyclonal antiserum against these serogroups. A large percentage of the presumptive colonies of each serogroup picked from Rainbow Agar O157 were confirmed as the respective serogroups; however, the percent recovery of STEC O111 was somewhat lower than that of the other serogroups. This work provides a method for detection and isolation in ground beef and potentially other foods of non-O157 STEC of major public health concern.

Escherichia coli O157:H7 is a foodborne pathogen that causes hemorrhagic colitis and hemolytic uremic syndrome. Positive identification of E. coli O157:H7 is made using biochemical tests and specific antisera or latex agglutination reagents for the O157 and H7 antigens. However, under certain conditions, some E. coli O157:H7 isolates can appear to be nonreactive with H7 antisera and may require multiple passages on motility medium to restore H7 antigenicity. In this study, we compared the performance of a real-time PCR test with that of a method using latex agglutination reagents to detect the presence of the fliC(h7) gene or the H7 antigen, respectively, in E. coli O157:H7 isolates. One hundred twenty-six E. coli strains were tested including reference strains and strains isolated from meat. Lyophilized E. coli O157:H7 isolates were rehydrated and were plated on sheep blood agar without passage on motility medium. All strains were analyzed in parallel by a real-time PCR test targeting the fliC(h7) gene and by a latex agglutination test that detects the H7 antigen. The real-time PCR assay showed 100% agreement with the H7 status reported for reference strains and E. coli O157:H7 meat isolates. The latex agglutination test results agreed with the H7 status reported for the E. coli O157:H7 reference strains and non-O157:H7 strains, except for one, E. coli O117:H7; however, 42% (42 of 100) of the E. coli O157:H7 meat isolates tested negative for the H7 antigen by latex agglutination. The real-time fliC(h7) PCR test can be used to confirm E. coli O157:H7 strains that are not expressing the immunoreactive H7 antigen.

In eight separate trials, four groups of raw ground beef samples were inoculated with 0.04 to 0.3 CFU/g of Salmonella Typhimurium (DT 104). Each group consisted of four 25-g samples (three inoculated and one uninoculated). After inoculation, these samples were shipped by overnight courier in Shipping containers with ice packs from the U.S. Department of Agriculture (USDA), Eastern Regional Research Center, in Wyndmoor, Pa., to the U.S. Food Safety and Inspection Service (FSIS), Eastern Laboratory, in Athens, Ga. A total of 128 samples (32 in each of four groups) were shipped. A temperature data logger was placed inside each shipping container to record the temperature during shipping and storage. The first group of ground beef samples was analyzed within approximately 1 h of arrival. The second group of samples was left in the original containers, with a gel ice pack, for 24 h before processing. The third and fourth groups of samples were removed from the original shipping containers and stored at room temperature (21 +/- 2 degrees C) for 6 h and then in a refrigerator at 4 +/- 2 degrees C for 24 and 48 h, respectively, before analysis. The samples were analyzed for the presence of Salmonella according to the USDA/FSIS Microbiological Laboratory Guidebook, chapter 4.02. There was no significant difference in the presence and levels of Salmonella in ground beef among the four test groups. These data show that it is acceptable to process the late-arriving ground beef samples for the detection of Salmonella if they are kept in a refrigerator (4 +/- 2 degrees C) for 24 to 48 h or when the shipments arrive late (24 h in the container with ice pack).

Two groups of calves were subjected to dietary stress by withholding of food beginning 1 or 14 days after inoculation with 10(10) CFU of Escherichia coli O157:H7. Following treatment, neither group had a significant increase in fecal shedding of E. coli O157:H7. A third group of calves had food withheld for 48 h prior to inoculation with 10(7) CFU of E. coli O157:H7. These calves were more susceptible to infection and shed significantly more E. coli O157:H7 organisms than calves maintained on a normal diet.

Cattle are an important reservoir of Shiga toxin-producing enterohemorrhagic Escherichia coli (EHEC) O157:H7 strains, foodborne pathogens that cause hemorrhagic colitis and hemolytic uremic syndrome in humans. EHEC O157:H7 strains are not pathogenic in calves >3 weeks old. Our objective was to determine if EHEC O157:H7 strains are pathogenic in neonatal calves. Calves <36 h old inoculated with EHEC O157:H7 developed diarrhea and enterocolitis with attaching and effacing (A/E) lesions in both the large and small intestines by 18 h postinoculation. The severity of diarrhea and inflammation, and also the frequency and extent of A/E lesions, increased by 3 days postinoculation. We conclude that EHEC O157:H7 strains are pathogenic in neonatal calves. The neonatal calf model is relevant for studying the pathogenesis of EHEC O157:H7 infections in cattle. It should also be useful for identifying ways to reduce EHEC O157:H7 infections in cattle and thus reduce the risk of EHEC O157:H7 disease in humans.

Escherichia coli isolates from 1,305 (of 6,894) fecal samples collected during the 1991-1992 USDA, Animal and Plant Health Inspection Service, National Health Monitoring System, Diary Heifer Evaluation Project were tested for virulence attributes associated with human enterohaemorrhagic E. coli (EHEC) and the enterotoxin commonly associated with diarrhoea in newborn calves. Single, random isolates from each heifer were hybridized to probes derived from the 60 mDa EHEC plasmid (CVD 419), E. coli attaching and effacing gene (eae), Shiga-like toxin (slt) genes I and II, and E. coli heat-stable enterotoxin a (STaP). Seventy-seven of the 1305 isolates (5.9%) were slt-positive. Most (81.8%) slt-positive E. coli were also CVD 419 and eae-positive. Only 2 of the slt-positive E. coli isolates were STaP-positive.

The antibody-direct epifluorescent filter technique (Ab-DEFT) was adapted for direct detection of Escherichia coli O157:H7 in bovine feces. The method involved suspension of bovine feces in buffer, centrifugation for 30 s, treatment of the supernatant with trypsin and Triton X-100 at 50 degrees C for 10 min, pre-filtration through 5 and 1.2 microns pore filters, and filtration through a 0.4 micron pore filter. The final filter was stained with fluorescein-labeled polyclonal antibody specific for the O157 antigen and examined by epifluorescence microscopy. The Ab-DEFT was correlated with viable plate counts for enumeration of the pathogen in artificially inoculated bovine feces (r = 0.96). The limit of detection was approximately 10(4)-10(5) CFU/g feces. The procedure provided a clean background for microscopic visualization of cells; however, cell loss and inaccurate quantitation sometimes resulted. E. coli O157:H7 was detected in feces of an inoculated calf for more than 3 weeks post-inoculation. The Ab-DEFT may be useful for rapid screening of cattle for the presence of the E. coli O157:H7 and as an analytical method in ecological studies of the pathogen.

Oral inoculation of calves and steers with 10(10) CFU of Escherichia coli O157:H7 induced prompt and sustained increases in serum antibodies to O157 lipopolysaccharide. Neutralizing antibodies to verotoxin 1 also increased rapidly in most steers but more gradually in calves. None of the animals developed neutralizing antibodies to verotoxin 2. These serological responses were not correlated with elimination of infection in calves or steers or protection of calves against reinfection with the same strain.

Preweaned calves and adult cattle were inoculated with 10(10) CFU of Escherichia coli O157:H7 strain 3081, a calf isolate which produces Shiga-like toxin, to define the magnitude and duration of fecal shedding of E. coli O157:H7 for each age group. Fecal samples of eight of eight, eight of eight, three of eight, and two of eight calves were positive at 2, 7, 14, and 20 weeks, respectively. In contrast, nine of nine, two of nine, and one of nine steers were positive at 2, 7, and 14 weeks, respectively. The magnitude of shedding (CFU per gram) by individual animals at any one time postinoculation varied widely within each age group but was greater for calves as a group. The differences in shedding patterns between adults and calves were statistically significant. After inoculation, 25 of 29 animals remained healthy and 4 of 17 calves had transient diarrhea. Histologic sections of the brain, kidney, jejunum, ileum, cecum, and colon taken at necropsy from nine calves either 3, 14, or 18 days postinoculation or three adults either 2, 3, or 4 days postinoculation were normal. E. coli O157:H7 was recovered from the alimentary tracts of all of the animals necropsied, and there was no evidence of spread to the liver, spleen, or kidneys. Four calves that had ceased shedding were reinfected when inoculated again with the same strain. E. coli O157:H7 was recovered from none of five and two of five adults inoculated with 10(4) and 10(7) CFU, respectively.(ABSTRACT TRUNCATED AT 250 WORDS)

Symptoms of SLT E coli-induced enteric disease in human beings include watery diarrhea, hemorrhagic diarrhea, and, in some cases, HUS. The most frequent serotype associated with HUS is O157:H7, although several other serotypes have also been implicated. These organisms produce SLT-I, SLT-II, or both toxins. Factors other than SLT are implicated as virulence attributes, such as adhesins and enterohemolysins, but roles for these factors in the pathogenicity of these organisms have not been defined. Colonization mechanisms for enterohemorrhagic E coli have not been defined, nor is there a defined set of characteristics by which enterohemorrhagic E coli pathogenic for human beings can be identified. Because virulence attributes are ill-defined, experimental animal models are useful in studies of pathogenicity. Gnotobiotic pigs, infant rabbits, streptomycin-treated mice, and one-day-old chickens have been used. Although the epidemiologic evidence implicating cattle as a source of zoonotic SLT E coli is strong, there is a paucity of direct evidence documenting this relationship. Until we have a better set of criteria with which to identify SLT E coli that are human pathogens, we are probably limited to epidemiologic criteria. Cattle excrete a variety of SLT E coli that includes many serotypes, in addition to O157:H7, that have been associated with disease in human beings. Surveys of the incidence of O157:H7 indicate a low incidence of these organisms in healthy cattle. However, much of these data have been derived from surveys of clinically normal cattle in daries.(ABSTRACT TRUNCATED AT 250 WORDS)

We investigated the role of the rumen fermentation as a barrier to the foodborne pathogen, Escherichia coli O157:H7. Strains of E. coli, including several isolates of O157:H7, grew poorly in media which simulated the ruminal environment of a well-fed animal. Strains of E. coli O157:H7 did not display a superior tolerance to ruminal conditions which may facilitate their colonization of the bovine digestive tract. Unrestricted growth of E. coli was observed in rumen fluid collected from fasted cattle. Growth was inhibited by rumen fluid collected from well-fed animals. Well-fed animals appear less likely to become reservoirs for pathogenic E. coli. These results have implications for cattle slaughter practices and epidemiological studies of E. coli O157:H7.

The mosquitocidal glycoprotein endotoxin of Bacillus thuringiensis subsp. israelensis was digested with chymotrypsin to yield protease-resistant domains which were then separated from smaller protease digestion products by high-performance liquid chromatography. Once purified, the domains no longer bound wheat germ agglutinin, a lectin which binds N-acetylglucosamine (GlcNAc) and GlcNAc oligomers. Purified protease-resistant domains were as toxic for Culex quinquefasciatus larvae as intact solubilized toxin. In separate experiments, the toxicity of chymotrypsin-digested endotoxin for Aedes aegypti larvae was reduced fivefold or more. A model is presented in which GlcNAc-containing oligosaccharides are required for toxicity for A. aegypti larvae but not C. quinquefasciatus larvae.

Determinants of the capacity of live Vibrio cholerae O1 isolates to evoke specific immune responses in intestinal mucosa were studied in rabbits, using mucosal immunoglobulin A (IgA) antitoxin as the measured immune response. Antitoxin responses were evoked mostly by the primary inoculation and were dose dependent; secondary-type responses were modest and occurred only when the booster inoculum was large, i.e., 10(10) CFU. The efficiency of mucosal immunization correlated closely with the mucosal colonizing capacity of the infecting strain and was otherwise independent of toxin genotype (A+ B+ or A- B+) or whether the strain was motile or nonmotile. Live bacteria evoked mucosal antitoxin more efficiently than did purified cholera toxin. Prior immunization with a nontoxinogenic (A- B-) V. cholera strain interfered significantly with the induction of mucosal antitoxin by subsequent immunization with its fully toxinogenic (A+ B+) parent. These results demonstrate the marked efficiency with which live V. cholerae stimulate a specific enteric mucosal secretory IgA response. They support the view that mucosal colonization aids efficient delivery of bacterial antigens to responsive subepithelial lymphoid tissue. This might occur by transfer of colonizing bacteria through M cells into Peyer patches or by efficient delivery of secreted toxin to M cells by mucosa-associated organisms. Preexisting antibacterial immunity interferes with colonization, which may prevent efficient antigenic stimulation and which may explain the relatively weak response to booster immunization. The same process may also limit the efficacy of hybrid enteric bacterial vaccines when there is preexisting mucosal immunity to the carrier organism due to either natural exposure or prior immunization with another vaccine that uses the same carrier.

The origin and fate of specific IgA plasma cells in intestinal lamina propria were studied in rats immunized enterically with cholera toxin (CT). Our major goal was to define how an anti-CT response is focused and sustained at the site of antigen challenge. To distinguish antigen-dependent from antigen-independent mechanisms, CT exposure was restricted to defined portions of intestine and, in some studies, the distribution of antitoxin-containing plasma cells (ACC) was examined in nonimmune adoptive recipients of post-challenge thoracic duct lymphocytes. After enteric priming and challenge, ACC appeared throughout the gut, but were most numerous at the challenged site. About 25% of ACC appearing at the site of jejunal challenge were due to antigen-driven proliferation of memory cells within the lamina propria; the remainder arose elsewhere, apparently in mucosal follicles or mesenteric lymph nodes, and migrated systemically as antitoxin-containing plasmablasts before homing to the lamina propria. The homing of these migrating ACC precursors was not affected by mucosal exposure to CT, nor did they undergo appreciable antigen-driven division after arrival in gut lamina propria. However, homing was specific for the organ from which they arose, i.e., precursors arising from duodenal challenge homed selectively to jejunum, whereas those from colonic challenge homed to the colon. The organ specificity of homing was determined during the challenge response and was independent of the origin of memory cells participating in the response. The survival of migrating ACC precursors did not differ in segments of gut exposed or nonexposed to CT. However, CT exposure at the time of their migration evoked another secondary-type response, due to stimulation of comigrating memory cells, thus sustaining the secondary response at a high level. These results and those in a previous report identify important mechanisms that affect the localization, magnitude, and duration of a specific IgA response, at least in the intestine. These include: 1) organ-specific homing of migrating IgA plasmablasts, 2) antigen-driven generation of IgA plasma cells from memory cells within the lamina propria, 3) enhanced memory at the site of mucosal priming compared to that a distant mucosae, and 4) regeneration of memory cells during the secondary response.

Four strains of Vibrio cholerae O1 were compared for their ability to colonize and immunize adult rabbit intestine. Three were virulent, toxinogenic (A+ B+) isolates, and one, an A- B+ mutant (Texas Star-SR), was derived by mutagenesis with nitrosoguanidine. When given orally to nonimmune rabbits, virulent strains colonized the small bowel with similar efficiency, whereas Texas Star-SR colonized poorly. Rabbits fed less than 50 CFU of an A+ B+ strain developed marked serotype-specific resistance to recolonization. In contrast, Texas Star-SR evoked resistance to reinfection less efficiently, with a minimum immunizing dose of 10(5) CFU when given once or 10(3) CFU when given twice. Oral inoculation with an A+ B+ strain also evoked vigorous, dose-dependent mucosal antitoxin responses; comparable inocula of Texas Star-SR were much less effective, causing antitoxin responses that were 90 to 95% smaller. Finally, rabbits inoculated once with 10(4) CFU of an A+ B+ strain were markedly protected against experimental cholera or fecal shedding of V. cholerae when challenged with 10,000 times the 50% effective dose of a virulent strain by the RITARD technique. In contrast, an inoculum of 10(4) CFU of Texas Star-SR was nonprotective, and 10(10) CFU was only partially protective. These studies reveal the remarkable efficiency with which virulent V. cholerae evokes intestinal immunity to recolonization or experimental cholera and show that the A- B+ mutant, Texas Star-SR, is substantially less effective.

The colonizing capacities of 16 Vibrio cholerae strains, including nine genetically and/or phenotypically defined parent-mutant pairs, were determined in unobstructed adult rabbit small bowel. There were marked interstrain differences in colonizing capacity, which was enhanced by bacterial motility and the production of cholera holotoxin but was unrelated to production of cholera toxin B-subunit or hemolysin or to bacterial serotype or biotype. The role of colonizing capacity and other bacterial features in determining the immunizing efficiency of live V. cholerae was studied by determining the efficiency with which graded inocula of each strain immunized against attempted recolonization of the ileum or induction of choleralike diarrhea by the RITARD (removable intestinal tie-adult rabbit diarrhea) challenge technique using standard inocula of virulent V. cholerae. Mucosal colonizing capacity was the only quantitative predictor of bacterial immunizing capacity; none of the other bacterial features cited above influenced bacterial immunogenicity against either type of challenge, except as they affected colonizing capacity. Live V. cholerae immunized much more efficiently than Formalin-killed bacteria; the former caused marked protection after a single inoculum of 10(2) CFU, whereas the latter gave only partial protection after three inoculations of 10(11) killed organisms. Protection induced by live bacteria was due largely to resistance to colonization and included marked inhibition of bacterial growth within the bowel lumen. These findings strongly suggest that an optimally efficient oral cholera vaccine would be composed of avirulent live V. cholerae selected for their capacity to colonize the small-bowel mucosa.

Using CT as the test antigen, we sought 1) to learn whether primary immunization at 1 mucosal site caused priming of distant nonstimulated mucosae, 2) to study the role of migrating memory cells in the dissemination of mucosal priming, and 3) to compare disseminated priming with priming that occurs at the site of initial immunization. CT given i.c. or i.d. caused priming in tracheal and nonexposed enteric mucosae; i.t. immunization, however, did not cause detectable enteric priming. Adoptive transfer of immune TDLs showed that priming was conveyed by migrating memory cells. These appeared to be of 2 types: those that recirculated briefly before settling in MALT, and those that continued to recirculate until recruited by antigen to the site of mucosal challenge. Both types were required for secondary responses at mucosae distant from the site of priming. The time-course of disseminated mucosal priming resembled that of priming at the site of initial CT exposure, both lasting at least 16 wk. Disseminated priming persisted better in jejunal than tracheal mucosa, suggesting that the subgroup of memory cells that did not continue to recirculate settled preferentially in jejunal MALT. Disseminated priming supported smaller challenge responses than priming at the site of initial CT exposure did, suggesting that sessile memory cells also contributed to the latter process. These observations extend the concept of a "common mucosal immune system" to include cellular dissemination of mucosal priming, but also show quantitative differences between local and disseminated priming that probably reflect the patterns of distribution of migrating and sessile memory cells.

Viable bacteria were found in six bottles of dimethyl sulfoxide (DMSO) at a concentration of approximately one bacterium per 4.4 mL. The 18 bacterial isolates appeared to be tolerating the DMSO rather than metabolizing it. No fungi were detected. DMSO must be assumed to be nonsterile unless it has been previously sterilized.

The induction of a jejunal antitoxin response was studied in dogs immunized with cholera toxin or toxoid. Single doses of toxoid given subcutaneously (s.c.) or of toxin given intraluminally (i.l.) were each effective in priming the mucosal immune system, whereas toxoid given i.l. was much less effective. In contrast, toxin and toxoid given i.l. were each effective as booster antigens. The local secondary response was rapid and brief, the peak occurring at about 7 days after i.l. boosting and declining by 90% after 2 more weeks. After s.c. priming and i.l. boosting with toxoid, antitoxin-containing plasma cells appeared predominantly in the portion of jejunum exposed to the i.l. booster. The appearance of antitoxin-containing plasma cells in jejunal lamina propria correlated with the amount of antitoxin recovered in jejunal washings which, in turn, correlated with protection against challenge with cholera toxin. Thus, lamina propria antitoxin-containing plasma cells appeared to be the source of protective antitoxin. However, after sequential s.c.-oral immunization with toxoid, protection against challenge with Vibrio cholerae far outlasted the major systemic and local antitoxin responses and was not obviously explained by either. These studies reveal methods for induction of a mucosal antitoxin response, but leave in question the mechanism of prolonged protection induced by s.c.-oral immunization of dogs.

The immunogenicity and safety of procholeragenoid, a minimally toxic, heat-induced aggregate of cholera toxin (CT), were studied in enterically immunized rats and dogs. Although 99% less toxic than CT, procholeragenoid was only slightly less efficient in causing jejunal anti-CT responses in rats; in contrast, choleragenoid, the nontoxic B subunit pentamer of CT, was much less effective. The immunogenicity of procholeragenoid was due almost entirely to its large-molecular-weight components (MW = 10(6) to 10(7)) and was markedly reduced by preincubation with GM1 ganglioside or treatment with Formalin to eliminate residual toxicity. These findings suggest that molecular aggregation, binding to GM1 receptors on cell membranes, and stimulation of cellular adenylate cyclase each contributed to the effectiveness of procholeragenoid as a mucosal immunogen. In dogs, oral immunization with five 500-micrograms doses of procholeragenoid evoked vigorous anti-CT responses in jejunal mucosa without causing significant diarrhea. When subsequently challenged with virulent Vibrio cholerae, immunized dogs showed 83% protection against the development of severe or lethal diarrhea compared with non-immunized controls. These results confirm a protective role for mucosal antitoxin in experimental cholera and show that procholeragenoid is both safe and effective as an oral immunogen. Procholeragenoid, combined with other antigens of V. cholerae, may constitute a simple, safe, and effective oral vaccine for cholera.

The immunogenicity and safety of purified cholera toxin (CT), its B subunit, and a crude culture filtrate of toxigenic Vibrio cholerae (CrT) were compared in dogs immunized orally and challenged with virulent V. cholerae. CT and CrT caused marked protection in two- or three-dose regimens. Protection due to CT occurred only with doses that caused transient, sometimes severe, diarrhea in most dogs; this protection was proportional to the peak antitoxin response in jejunal mucosa and lasted at least 15 weeks. In contrast, minimum protective doses of CrT contained much less cholera toxin, caused very mild diarrhea in only 21% of the dogs, and evoked protection that was greater than predicted from the modest jejunal antitoxin response. B subunit caused smaller jejunal antitoxin responses than did similar doses of CT and was poorly protective, the 50% protective dose being >40-fold greater than that of CT. Two observations indicated that protection due to CrT involved synergy between antibacterial and antitoxic immune responses. First, the 50% protective dose of CrT was 24-fold and >36-fold smaller than the 50% protective doses of its CT and non-CT antigenic components, respectively, when tested separately. Second, protection was greater in CrT-immunized dogs than in CT-immunized dogs for a given mucosal antitoxin response. Low doses of CrT evoked serotype-specific protection, indicating that the serotype-specific O somatic antigen contributed significatly to antibacterial protection. These results suggest that a simple, effective, nonliving oral vaccine for cholera based on combined antibacterial and antitoxic immunity can probably be achieved. However, further studies are needed to determine how a protective antitoxic response can be evoked without causing diarrhea during immunization.

The ability of tetracycline and clioquinol to prevent intestinal colonization of Vibrio cholerae and Escherichia coli was tested in a rabbit model. In the model 10(10) bacteria are given via oro-gastric tube following intravenous cimetidine and oral sodium bicarbonate and prior to intraperitoneal tincture of opium. Eighteen hours after challenge the rabbits are sacrificed, and the numbers of the challenge strain remaining in the jejunum and ileum are determined. Tetracycline interrupted the intestinal colonization of V. cholerae and E. coli. Clioquinol however, had minimal effect on the colonization process. Our studies demonstrate the efficacy of prophylactic tetracycline but do not support the use of clioquinol to prevent intestinal infection due to these organisms. This rabbit model may also be useful to study the efficacy of other antibiotics against these bacterial infections.

Colera toxin was evaluated as an oral immunogen against experimental canine cholera. Dogs were immunized orally with 100-microgram doses of purified cholera toxin or comparable doses of crude toxin. Both doses caused moderate diarrhea in most nonimmune dogs. Repeated oral doses (12 doses in 54 days) gave marked protection against the diarrheal effect of oral toxin, provoked a vigorous antitoxic response in jejunal mucosa, and gave nearly complete protection against subsequent oral challenge with living virulent Vibrio cholerae. Protection appeared to be due largely to the antitoxic response in intestinal mucosa. The effectiveness of cholera toxin as an oral vaccine contrasts with the previously described ineffectiveness of toxoid given orally. This study provides an example of mucosal immunity due to a nonreplicating vaccine given orally and suggests that cholera toxin may be useful as a component of an oral vaccine for cholera.

The role of cholera toxin (CT) in mucosal colonization by Vibrio cholerae O1 was studied in rabbits by using toxinogenic V. cholerae and nontoxinogenic (A-B+ or A-B-) recombinant mutants derived from them. After oral inoculation, toxinogenic strains colonized intestinal mucosa significantly more efficiently than did either A-B- or A-B+ mutants; average colonization was increased 1.5- to 30-fold with toxinogenic strains, depending on the inoculum used and the portion of intestine studied. Additionally, colonization by an A-B- mutant was increased to the levels of its toxinogenic parent by coadministration of CT with the inoculum. We conclude that CT contributes significantly to mucosal colonization by V. cholerae and that this effect is not due to an interaction of the CT B subunit with its mucosal receptor. The possibility that this effect contributes to the in vivo selection of hypertoxinogenic variants of V. cholerae is considered.

Viable Vibrio cholerae O1 were inoculated into the intestinal lumen of nonimmune rabbits. The vibrios were phagocytosed by M cells over Peyer's patch lymphoid follicles, carried in vesicles through the epithelium, and discharged among underlying lymphocytes and macrophages. Autoradiography of V. cholerae labeled with [2-3H]adenine confirmed transport. Indigenous bacteria with and without capsules were also taken up from control loops and carried through M cells into Peyer's patches. V. cholerae killed by acidification, formalin, heat, or UV irradiation were not taken up, a result that may have relevance for development of oral vaccines. Ruthenium red stain revealed gaps in the layer of mucus over M cells, glycocalyx bridging the space between vibrios and M cell microvilli, and knobby projections over membranes of M cell microvilli; these projections were not found over absorptive cells. M cells thus convey viable enteric microbes, including V. cholerae that are not otherwise invasive, into intestinal lymphoid tissue, where mucosal immune responses are initiated. Uptake and transport by M cells may also assist certain pathogenic bacteria in traversing the mucosal barrier and initiating systemic infection.

Adult rabbits were inoculated orally (or duodenally) with virulent Vibrio cholerae O1. Jejunal colonization occurred only when hypoperistalsis was induced at the time of inoculation by tincture of opium given intraperitoneally (or by temporary ileal obstruction). For oral inoculation, prior neutralization of gastric acid was also required. Inoculation with 10(9) V. cholerae caused jejunal colonization for 1 to 2 days and ileal colonization for 5 to 6 days. The extent of small bowel colonization 18 h after oral inoculation was related to inoculum size but also reflected limited multiplication of small inoculum sizes and net death, clearance of large inoculum sizes, or both. Serious diarrhea occurred only in rabbits fed large inoculum sizes, i.e., 10(10) V. cholerae, and then rarely. Rabbits colonized once with 10(10) V. cholerae became highly resistant to recolonization with either the same or opposite serotype. After 18 weeks, these rabbits were still partially protected, whereas twice-colonized rabbits were highly protected. Protection against recolonization appeared to be due, at least partly, to interference with the adherence of V. cholerae to the bowel mucosa, thus allowing rapid removal of V. cholerae when peristalsis resumed. Prior colonization also protected against cholera-like diarrhea in rabbits challenged by the removable intestinal tie-adult rabbit diarrhea technique, the 50% effective dose for severe or lethal diarrhea being increased more than 100-fold, and probably more than 10,000-fold, for challenge with either the homologous or heterologous serotype of V. cholerae. The described rabbit model appears well suited for the study of immunity evoked by enteric colonization with V. cholerae O1.