Helicobacter pylori and progressive gastric pathology that predisposes to gastric cancer.

Evidence is presented suggesting that infection by Helicobacter pylori triggers and continuously contributes to the pathophysiology of progressive gastric changes that can ultimately lead to gastric cancer. In Peru, especially in population groups of low socioeconomic status, infection by H. pylori begins earlier in life and is more prevalent and persistent than in developed countries. The infection produces a destructive lesion of the mucinous surface epithelium which probably enables other aggressive luminal factors to cause further mucosal damage. As a consequence, active chronic gastritis appears. The gastritis is of the superficial type at the beginning but may progressively change to atrophic. Chronic atrophic gastritis is found more frequently and at a younger age in dyspeptic patients with low socioeconomic status--that is, in patients having higher prevalence of persistent infection by H. pylori since earlier in life. When chronic atrophic gastritis becomes severe and extensive, hypochlorhydria ensues. Hypochlorhydria favors the appearance of bacterial overgrowth, nitrites, and N-nitroso compounds in the gastric lumen. N-nitroso compounds, because of their mutagenic-carcinogenic properties, probably induce gastric premalignant lesions like intestinal metaplasia and dysplasia of the gastric mucosa. Oral bismuth therapy apparently reverses H. pylori-associated gastric dysplasia. It is proposed that future programs designed for the control of gastric cancer would be incomplete if they do not include further evaluation of the many effects of infection by H. pylori on the gastric mucosa and of cost-effective methods to eradicate the infection.

In vitro and in vivo effects of three bismuth compounds on fermentation by colonic bacteria.

Three bismuth compounds (tripotassium dicitrate bismuthate, bismuth subsalicylate, and bismuth subnitrate) were tested in vitro and in vivo for their effect on fermentation by colonic bacteria. The studies in vitro were done with use of a technique designed to determine the effect of each one of the bismuth compounds on the fermentation of several stool samples that had been mixed with lactose as additional fermentable substrate (fermentation of lactose-enriched stools, FLES). The three bismuth compounds reduced FLES significantly in 47 (81%) of 58 of the stool samples used to test their effect. Bismuth subsalicylate, which reduced FLES in 10 of 10 stool samples, showed the greatest reduction (mean reduction, 74%; P less than .0001). The in vivo studies, done in six flatulent patients, showed significant reduction (P less than .01) of colonic fermentation of ingested raffinose by oral bismuth subnitrate given for 8 days.

Decreased gastric acid secretion and bacterial colonization of the stomach in severely malnourished Bangladeshi children.

To assess the effect of malnutrition on gastric acidity and gastric bacterial colonization, we studied 35 severely malnourished Bangladeshi children before (0 wk) and after (3 wk) they received nutritional rehabilitation for 3 wk. These results were compared with those obtained from a similarly examined group of 20 better-nourished Bangladeshi children. Gastric acid output, both basal and after betazole stimulation, was significantly lower in the malnourished group at 0 wk compared with the better-nourished children (p less than 0.01): basal 0.22 vs. 0.52 mEq HCl/h and stimulated 0.90 vs. 2.5 mEq HCl/h. Both the concentration of acid and the rate at which gastric juice was secreted were decreased in the malnourished group but serum gastrin levels were not significantly different. After 3 wk, the malnourished children had improved from 61% (+/- 9.0%; SD) to 81% (+/- 8.1%) of expected weight-for-height and were not significantly different than the better-nourished group (86% +/- 11%). Nevertheless, gastric acid concentration remained depressed in the 3-wk group, although the rate of gastric juice secretion equaled levels observed in the better-nourished group. None of the better-nourished children had detectable gram-negative bacterial colonization of their gastric juice. In contrast, 26 of 32 (81%) malnourished children at 0 wk were colonized--even after betazole stimulation, 11 of 33 (33%) gastric juice samples yielded viable organisms--suggesting that the decrease in gastric acid output greatly reduced the gastric acid barrier. Interestingly, only 9 of 20 (45%) better-nourished children had gastric juice with basal pH values below 4.0, suggesting that the gastric acid barrier may be an intermittent defense factor in Bangladeshi children.

Oral immunization of rabbits with enterotoxigenic Escherichia coli protects against intraintestinal challenge.

The development of a successful oral vaccine against enterotoxigenic Escherichia coli depends upon the identification of appropriate protective antigens which can be delivered effectively to intestinal mucosa. We have determined in a modified RITARD model the relative protection against intraintestinal challenge afforded by oral immunization with live enterotoxigenic E. coli carrying different candidate antigens. Studies were done with both wild-type strains and genetically manipulated strains of enterotoxigenic E. coli (parent strain E1392/75 2A) which carried plasmids containing intact heat-labile toxin (LT) gene sequences or various mutations of the LT genes. Immunizations were done by orogastric tube inoculation on days 0, 7, and 14; challenges were done on day 33. Protection against diarrhea with a homologous challenge was found to be 84 to 100% (P less than 0.01). Protection against diarrhea with challenges in which specific antigens could be tested included the following: (i) O and H antigens (O6:H16), 87 to 100% protection with different E. coli strains with identical O and H antigens (P less than 0.01) but no protection against a heterologous challenge; (ii) LT or the B subunit of LT only, approximately 50% protection (P less than 0.02). These findings suggest that O antigens are highly protective in this model but afford only serotype-specific protection and that the B subunit (with or without the A subunit) affords less protection but confers cross-protection against heterologous strains producing LT. This model should be useful in further defining appropriate protective antigens for candidate enterotoxigenic E. coli vaccine strains.

Epidemiology of rotavirus diarrhea in a prospectively monitored American Indian population.

Rotaviral diarrhea is endemic in most areas of the world, yet community-wide epidemics have not been reported in prospectively monitored populations. This prospective study of the etiology of diarrhea included biweekly visits to the homes of 10% of the population of the White Mountain Apache Indians and began in April 1981. During a three-week period beginning 21 October, 1981, 342 new cases of diarrhea were identified on different parts of the reservation. Rotaviral antigen, detected by an enzyme-linked immunosorbent assay, was identified in 169 (73%) of the 233 stool samples that were tested. Rotavirus was not detected in any of the stool samples taken six months before or after the epidemic. During the epidemic, respiratory symptoms were present in 44 (33%) of 135 rotavirus-positive patients compared with 17 (17%) of 98 rotavirus-negative patients (P less than .05). This rapidly spreading epidemic involving all areas of the reservation, in the absence of a common source of exposure of ill persons, suggests the possibility of respiratory transmission.

Outpatient use of oral rehydration solutions in Apache population: effect of instructions on preparation and contamination.

Oral Rehydration Solutions (ORS) containing 90 and 50 mmol/L sodium have recently been recommended for use in ambulatory children in the U.S. These solutions are now marketed in powder form by some commercial companies. However, few data are available in the U.S. on the accuracy with which the solutions are mixed at home or on the bacterial contamination that may occur during mixing. We evaluated the effect of various forms of instructions on the occurrence of bacterial contamination and accuracy of mixing ORS at home by mothers of patients who were dispensed the dry ingredients of an ORS containing 90 mmol/L sodium at the U.S. Public Health Service Hospital, Whiteriver, Arizona. Patients were randomized to one of the four following groups: group I (23 patients) was given written instructions for mixing the solution along with a pre-marked container; group II (22 patients) was given written instructions only; group III (22 patients) was given a premarked container only; and group IV (19 patients) was given neither. All patients were given oral instructions in the preparation of ORS and were asked to refrigerate the reconstituted ORS. We collected samples of ORS at the patient's home 1 day after the clinic visit, to measure their electrolyte content and to identify any bacterial contamination. Mean Na+ concentrations were significantly lower in the ORS prepared by mothers/guardians in groups that were not given a premarked container [82 +/- 13 (II) and 79 +/- 21 (IV) mmol/L vs. 88 +/- 13 (I) and 92 +/- 14 (III) mmol/L; p less than 0.01].(ABSTRACT TRUNCATED AT 250 WORDS)

Purification of enterotoxins from Vibrio mimicus that appear to be identical to cholera toxin.

Cholera toxin-like (CT-like) enterotoxins produced by two strains of Vibrio mimicus, 61892 and 63616, isolated from diarrhea patients in Bangladesh, were purified, and their physicochemical, biological, and immunological properties were compared with those of CT produced by classical Vibrio cholerae 569B. The CT-like toxins were produced by lincomycin-resistant mutants grown in the presence of lincomycin at 200 micrograms/ml for strain 63616 and 250 micrograms/ml for strain 61892 and were purified by coprecipitation with hexametaphosphate followed by chromatography on phosphocellulose. The pure CT-like toxins were indistinguishable from 569B CT in sodium dodecyl sulfate-polyacrylamide gel electrophoresis, rabbit intestinal loop and Y-1 adrenal cell assays, antiserum neutralization and binding inhibition assays, and Ouchterlony immunodiffusion, except that the CT-like toxins appeared to consist almost entirely of A subunit which was proteolytically unnicked. Trypsin nicking, however, resulted in fragments that appeared to be identical to those of 569B CT. These results indicate that at least one species of Vibrio other than V. cholerae can produce enterotoxins which are virtually identical to CT.

Production of cholera toxin-like toxin by Vibrio mimicus and non-O1 Vibrio cholerae: batch culture conditions for optimum yields and isolation of hypertoxigenic lincomycin-resistant mutants.

Vibrio mimicus 61892, isolated in 1977 from a case of watery diarrhea in Bangladesh, produces an enterotoxin which possesses activity in Y-1 mouse adrenal cells and in rabbit ileal loops which is identical to the prototype cholera toxin (CT) produced by Vibrio cholerae 569B. The neutralization of the adrenal cell activity of 61892 toxin and 569B CT by homologous and heterologous antisera generates parallel titration curves which show complete neutralization in all cases. Paired titrations in the ganglioside GM1 enzyme-linked immunosorbent assay (using either CT or Escherichia coli heat-labile toxin antitoxin) of both toxins indicates that 61892 toxin is antigenically indistinguishable from 569B CT. The specific activity of the two toxins in the rabbit ileal loop is virtually identical. Batch culture production of CT-like toxin and CT by isolates of V. mimicus and different biotypes of V. cholerae was found to be highest in shake flask cultures of Casamino Acids-yeast extract broth grown at 27 degrees C with vigorous aeration. Incorporation of lincomycin into the growth medium at a concentration of 50 micrograms/ml increased yields from wild-type strains. Dramatically higher yields were obtained when a spontaneous resistance mutant of strain 61892 was grown in the presence of 200 to 300 micrograms of lincomycin per ml. Under these conditions, yields of CT-like toxin were increased by 300- to 500-fold, and the highest yields reached more than 100 micrograms/ml after 44 h of culture. This is substantially higher than that reported in the literature for CT production by any strain of V. cholerae, including hypertoxigenic strain 569B.

Colonization of the rabbit small intestine by clinical and environmental isolates of non-O1 Vibrio cholerae and Vibrio mimicus.

We examined the capability of 12 isolates of non-cholera toxin-producing O1 and non-O1 Vibrio cholerae to colonize the small intestine of adult rabbits and cause diarrhea. Using the removable intestinal tie-adult rabbit diarrhea model, we found that eight environmental isolates that showed no or marginal biological activity in other diarrhea models (rabbit ileal loop, infant rabbit, and suckling mouse) appeared to be incapable of attaching to and colonizing, even transiently, the small intestinal mucosa of animals with normal clearance mechanisms. In contrast, three clinical isolates attached, proliferated rapidly, and colonized mucosal surfaces of the entire small intestine within 8 h of challenge. This led to diarrhea with strikingly high rates of mortality compared with that of rabbits given similar challenges doses with strains of O1 V. cholerae that produce cholera toxin and Vibrio mimicus, which produces a toxin similar to cholera toxin. We have further demonstrated that multiple exposures to enteric infection by these strains elicited local and serum antibodies that reacted strongly with cell surface antigens of the homologous strain and showed a high degree of cross-reactivity against the cell surface antigens of the two heterologous strains. The enteric infections appeared to engender protection against subsequent infection as well, as evidenced by reduced incidence of diarrhea and duration of fecal shedding of the challenge organism upon subsequent challenges.

Enterotoxin production by Vibrio cholerae and Vibrio mimicus grown in continuous culture with microbial cell recycle.

We have examined the effect of complete cell recycle on the production of cholera toxin (CT) by Vibrio cholerae and CT-like toxin by Vibrio mimicus in continuous culture fermentations. Complete cell recycle was obtained by filtering culture fluids through Amicon hollow fibers with an exclusion limit of 100,000 daltons (H1P100-20) and returning the concentrated cell slurry to the fermentor. A single 1-liter laboratory fermentor system modified with this recycle loop was capable of producing over 20 liters of cell-free culture filtrate per day. Toxin production in this system was compared with yields obtained in traditional continuous cultures and in shake flask cultures. Yields of CT from V. cholerae 569B in the recycle fermentor were highest at the highest dilution rate employed (1.0 vol/vol per h). The use of complete cell recycle dramatically increased yields over those obtained in continuous culture and equaled those obtained in shake flasks. The concentration of CT in the filtrate was slightly less than half of that measured in culture fluids sampled at the same time. Similarly, V. mimicus 61892 grown in the presence of 50 micrograms of lincomycin per ml produced 280 ng of CT per ml in the recycle fermentor, compared with 210 ng/ml in shake flasks under optimal conditions. The sterile filtrate from this fermentation contained 110 ng/ml.

Chronic giardiasis: studies on drug sensitivity, toxin production, and host immune response.

UNLABELLED: We investigated a patient (W.B.) with chronic symptomatic Giardiasis despite seven separate courses of either quinacrine or metronidazole who was cured by combined quinacrine and metronidazole. After isolating Giardia lamblia from W.B., we cultured the trophozoites to make the following observations. In vitro drug testing showed that (a) W.B.'s organisms were not more drug resistant than three other isolates and that (b) W.B.'s organisms were more sensitive to combined quinacrine and metronidazole than to either drug alone. Isolates of Giardia lamblia from W.B. and 3 other patients did not produce detectable enterotoxin in four different assays. W.B. had normal levels of circulating immunoglobulins, detectable intestinal immunoglobulin A, circulating immunoglobulin G anti-Giardia lamblia antibodies, and lymphocyte responsiveness to solubilized giardia lamblia. However, monocytes-macrophages from W.B. exhibited reduced killing for Giardia lamblia compared with normal subjects. CONCLUSIONS: (a) The chronicity of our patient's infection was not due to the organism having unique properties of drug resistance. (b) Combined quinacrine and metronidazole, which cured our patient's chronic giardiasis, should be tried in patients in whom infection persists after single drug therapy. (c) The diarrhea in our patient was probably caused by a mechanism other than Giardia lamblia-induced secretion by currently recognized enterotoxins. (d) Reduced cellular cytotoxicity for Giardia lamblia may have contributed to the persistence of our patient's infection and should be suspected in other patients with chronic giardiasis.

Kinetics of early cholera infection in the removable intestinal tie-adult rabbit diarrhea model.

The colonization of the small intestine of adult rabbits challenged with 5 X 10(7) cells of Vibrio cholerae strain Ogawa 395 has been examined in the removable intestinal tie-adult rabbit diarrhea (RITARD) model. During the first 6 h of infection, numbers of both free and adherent vibrios increased at a rate representing a generation time of about 71 min. Detectable fluid output in response to infection began at about 4 to 5 h postchallenge, and overt diarrhea was observed as early as 11 h. By 8 h after challenge, adherent V. cholerae reached a saturation concentration on the intestinal epithelium of approximately 5 X 10(8) cells per g of intestine, whereas numbers of free cells continued to increase at an exponential rate for at least 12 to 14 h. The concentration of adherent cells remained relatively constant at the saturation level during this period. This saturation level was similar in all parts of the small intestine. The concentration of adherent organisms increased significantly in moribund animals, suggesting that factors responsible for the earlier saturation equilibrium began changing as animals neared death.

Gauze filtration and enrichment procedures for recovery of Vibrio cholerae from contaminated waters.

Gauze filtration followed by 18-h enrichment in alkaline bile-peptone water is a simple, inexpensive, and efficient method for isolation Vibrio cholerae biotype eltor from contaminated surface waters.

Uptake of Vibrio cholerae biotype eltor from contaminated water by water hyacinth (eichornia crassipes).

Vibrio cholerae biotype eltor appears to concentrate on the surface of the water hyacinth (Eichornia crassipes), thereby enhancing its survival and its potential for transmission through waterways of cholera-endemic regions such as Bangladesh.

Simple adult rabbit model for Vibrio cholerae and enterotoxigenic Escherichia coli diarrhea.

We developed an adult rabbit model for enteric infection by Vibrio cholerae and enterotoxigenic Escherichia coli. The cecum of each animal was first ligated to prevent it from retaining fluid secreted by the small intestine. A temporary reversible obstruction (a slip knot tie) of the small bowel was introduced at the time of challenge and removed 2 h later. With this modification, we were able to elicit a massive and usually fatal cholera-like diarrhea in adult (3.5- to 6-lb [1.6- to 2.7-kb]) animals challenged with V. cholerae. Animals challenged with enterotoxigenic E. coli also developed diarrhea which was severe and watery but less explosive and less rapidly fatal than that produced by V. cholerae. The susceptibility of animals in this model to infection by V. cholerae was similar to the susceptibility of infant rabbits challenged intraintestinally. The death rate was almost 25% when 10(3) Vibrio cells were given and 90% or more when the dose was greater than or equal to 10(6) cells per animal. We designated this procedure the RITARD (for removable intestinal tie-adult rabbit diarrhea) model.

Isolation of drug-resistant Aeromonas hydrophila from aquatic environments.

Antibiotic-resistant strains of Aeromonas hydrophila have been isolated from the natural environment in the Chesapeake Bay and areas surrounding Dacca and the Matlab region of Bangladesh. The Bangladesh strains carried resistance to chloramphenicol, streptomycin, and tetracycline, and 57% of them had a multiple streptomycin-tetracycline resistance phenotype correlated with the presence of a large plasmid. The Chesapeake Bay strains were resistant to polymyxin B ane tetracycline, but showed neither multiple resistance nor R-factor carriage. Twenty-five percent of the environmental strains were toxigenic in a Y-1 adrenal cell assay. Toxigenicity showed no positive correlation with drug resistance or with plasmid carriage. Environmental areas of heavy human impact appear to be associated with a higher incidence of antibiotic-resistant strains of aeromonads.

R-factor carriage in a group F vibrio isolated from Bangladesh.

Two group F vibrio organisms have been identified among a collection of vibrio strains isolated from the aquatic environment in Bangladesh. Neither group F strain produced a cholera-like enterotoxin. One of the isolates, BV12, contained an R plasmid conferring resistance to streptomycin and chloramphenicol.

Microtiter ganglioside enzyme-linked immunosorbent assay for vibrio and Escherichia coli heat-labile enterotoxins and antitoxin.

We have developed a microtiter enzyme-linked immunosorbent assay method for detecting the heat-labile enterotoxins of Vibrio cholerae and Escherichia coli using GM1 ganglioside as the base coat. This method compares favorably with a similar assay using anticholera toxin as the base coat, and with the Y1 adrenal cell assay. The assay should be useful in detecting enterotoxin production in E. coli and vibrios (including non-agglutinating Vibrio), in quantitating the toxin, and in determining binding properties of enterotoxins to ganglioside. The assay can also be used to quantitate antibodies which block the attachment of the toxin to the ganglioside.