Search

Escherichia coli from 138 fecal samples from aboriginal children, in whom no other enteric pathogen was isolated (including enterovirulent E. coli), were examined for HEp-2 cell adhesion. Twenty-five (36.8%) of 68 children with diarrhea and 32 (45.7%) of 70 without diarrhea had diffusely adherent isolates, which were thus not associated with diarrhea (P > .25). However, after age stratification, children > or = 18 months showed a significant association of diffusely adherent E. coli with diarrhea (P < or = .05). Enteroaggregative E. coli were isolated from 12 children with diarrhea (17.6%) and 15 without diarrhea (21.4%); thus, there was no association with diarrhea (P > or = .5). Sixteen children with diarrhea (23.5%) and 6 without diarrhea (8.6%) carried isolates that caused detachment of the HEp-2 cell monolayer from the glass coverslip when examined in the adhesion assay and were significantly associated with diarrhea (P < or = .05). These isolates, termed cell-detaching E. coli, were different from all recognized classes of enterovirulent E. coli.

Infectious diarrhoea is common in young Australian Aborigines and is one of the main causes for their unsatisfactory health standards with consequent widespread failure to thrive and undernutrition. Most published reports relate to patients in hospital or to hospital admission statistics and give little indication of the extent or severity of diarrhoeal disease in children in Aboriginal communities. The present investigation involved more than 100 Aboriginal children up to 5 years of age living in remote communities in the tropical north of Western Australia who were studied prospectively over a 12-month period.

Cytotoxicity of faecal extracts was demonstrated in 47 of 88 children (54%) referred for microbiological investigation of stools. Cytotoxic Clostridium difficile and vertotoxigenic Escherichia coli (VTEC) were the pathogens identified most commonly but cytotoxicity was also found in association with Campylobacter jejuni, Salmonella spp, Shigella sonnei, Giardia lamblia, rotavirus, adenovirus and poliovirus type 1 which had been acquired by oral immunization. In two patients, one of whom had cystic fibrosis, cytotoxicity of faecal extracts was associated with isolation of Pseudomonas aeruginosa. In five of 13 patients with diarrhoea and cytotoxic C. difficile, other pathogens were also present, in agreement with the view that C. difficile are more readily recovered when the intestinal flora have been altered by colonization with other micro-organisms. There was no correlation between previous treatment with antibiotics and isolation of C. difficile. Cytotoxicity neutralized by antitoxin, usually to C. sordellii, is used to detect cytotoxic C. difficile. We suggest that cytotoxicity not neutralized in this way should be an indication for further investigation of stools for the presence of other pathogens such as VTEC or viruses.

A total of 112 EPEC strains isolated from children with diarrhoea in New Zealand were examined for mannose-resistant HEp-2 cell adherence and production of exotoxins. Enterotoxin production was not detected in any of the strains examined. Verotoxin production was detected in 13 (11.6%) strains and of these 4 were also found to adhere to HEp-2 cells. HEp-2 cell adherence was displayed by a total of 29 (25.8%) strains of which 22 were diffusely adherent. Only 3 (2.7%) strains were shown to belong to the new virulence phenotype, entero-aggregative adherence, when examined in the adherence assay. We identified one strain with the novel characteristic of causing detachment of HEp-2 cells from glass coverslips and are further investigating this possible virulence mechanism. These results suggest that if EPEC strains are to be considered as a cause of diarrhoea, the search for new virulence factors must be extended.

Direct plating of faecal specimens on blood agar was compared with the use of enrichment culture for isolation of Aeromonas spp. from faeces during a large epidemiological study. Of enterotoxigenic strains isolated by direct plating, 89% were associated with acute diarrhoea and 7% with an episode of diarrhoea during the month before collection, but 79% of enterotoxigenic strains isolated only after enrichment were not associated with acute diarrhoea. With Aeromonas spp., as with intestinal pathogens, it appears that enrichment allows isolation of the bacteria when in low faecal concentrations likely to be found in convalescent patients, carriers and those with subclinical infection. The routine use of enrichment for isolation of faecal aeromonads, by detecting Aeromonas spp. in low numbers in patients without diarrhoea, is likely to confuse interpretation of epidemiological studies seeking to clarify the relationship between Aeromonas spp. and acute diarrhoea.

Aeromonas hydrophila was isolated from the faeces of 32 patients during a nine month period in three hospitals in Western Australia. All 32 isolates produced enterotoxin which was detected by the suckling-mouse test, and all patients except one had diarrhoea. Treatment should be considered in patients with chronic diarrhoea and in those with malignant disease or with hepatobiliary disease who are at risk of developing aeromonas septicaemia.

A soluble haemagglutinin has been identified in cell-free culture supernates of human diarrhoeal isolates of Aeromonas sobria, A. hydrophila and A. caviae. It was oligomeric; a major peak of haemagglutinating activity had an apparent mol. wt of 780,000 but there was haemagglutinating activity throughout the mol. wt range less than 40,000- greater than 10(6). Human group O, A and B, horse, rabbit, chicken and rat erythrocytes, but not those of sheep and cow, were agglutinated by the soluble haemagglutinin, in contrast to the cell-bound agglutinin. Agglutination was inhibited by fetuin, a complex glycoprotein, but not by simple sugars. The haemagglutinating activity was not affected by 0.5 M NaCl, dithiothreitol or the presence or absence of Ca++. It was unrelated to the haemolytic, enterotoxigenic and proteolytic activities present in cell-free extracts of A. sobria. All A. sobria, 73% of A. hydrophila and 68% of A. caviae strains tested produced this soluble haemagglutinin. A. caviae does not appear to be an enteric pathogen, therefore this soluble haemagglutinin alone is unlikely to be a virulence factor in Aeromonas spp.

Chlorpromazine or aspirin alone, when given to rats parenterally, reduced intestinal fluid secretion induced by cell-free preparations of enterotoxigenic organisms including Escherichia coli, Aeromonas hydrophila, Staphylococcus pyogenes, and Salmonella typhimurium. A combination of chlorpromazine-aspirin given parenterally caused much more marked reduction of fluid secretion. Indomethacin also had significant antisecretory effects against a range of bacterial enterotoxins, while loperamide was effective against heat-labile toxin (LT)-positive E. coli and A. hydrophila. Nicotinamide increased net fluid absorption in the presence of E. coli LT, A. hydrophila, and S. typhimurium. Of the adsorbents tested, aluminum hydroxide showed a positive effect only with E. coli LT and A. hydrophila, while cholestyramine affected net fluid flux only with E. coli ST (heat-stable toxin). Charcoal was effective against all microorganisms tested but only when premixed with the perfusate before the experiments. Aspirin and chlorpromazine probably act at multiple sites to decrease intestinal secretion, and the combination of low doses of these drugs with possibly different sites of action may have advantages over a single agent used in high dosage.

A year-long, prospective study of Aboriginal children with diarrhoea showed a high rate of isolation of bacterial pathogens, viruses and parasites in faecal specimens. Seventeen per cent of patients had enterotoxigenic bacteria in their stools; none of the 132 Aboriginal children without diarrhoea studied at the same time had enterotoxigenic bacteria isolated. Salmonellae and shigellae were also much more common in the diarrhoea group. Bacterial pathogens including enterotoxigenic E. coli and Aeromonas, salmonella, shigella, and campylobacter occurred in 39% of the diarrhoea group but were found in only 4.5% of the controls. Rotaviruses were detected in 12% of children with diarrhoea and in 8% of controls. Intestinal parasites, particularly Giardia lamblia, were present in over 25% of children in each group.

Biochemical characteristics and virulence factors were compared in 147 Aeromonas spp. isolated from patients with diarrhea and in 94 strains isolated from metropolitan water supplies in the same area during the same period. Fermentation of arabinose occurred with 58.5% of the environmental strains and 15% of the clinical isolates; 39.4% of the strains from water and 6.8% of the fecal isolates fermented salicin. The frequency of esculin hydrolysis was the same in both groups. Ninety-one percent of clinical isolates and 70.2% of environmental strains were enterotoxigenic and, except for four clinical isolates, all of these strains also produced hemolysins. Hemagglutination that was inhibited by fucose and mannose but not by galactose was found in 67% of the water isolates and 10.2% of the clinical strains. Although the distribution of several characteristics differs in clinical and environmental strains, many of the strains found in water have properties identical with those of the clinical isolates. We suggest that such strains may be potential enteric pathogens.

Aeromonas spp. show patterns of hemagglutination with human group O cells in the presence of fucose, galactose, and mannose. These patterns are related to biotype as well as to the source of isolates. There was good correlation between hemagglutination pattern and the presence of diarrhea among strains isolated in Western Australia, which was the only source with adequate data for classification of children with an without diarrhea. Most of the environmental and other nonfecal isolates produced patterns different from those in strains associated with diarrhea. These results suggest that hemagglutinins should be considered with enterotoxins as virulence factors in Aeromonas spp.

The study included 138 A. sobria and 182 A. hydrophila isolated in Perth from samples of diarrhoeal or non-diarrhoeal faeces or from domestic water. Strains were grouped in relation to agglutination of human, horse, rat and guinea pig erythrocytes and the effect of sugars on haemagglutination. Agglutination of red cells of all four species (primary group 1) was most commonly associated with A. sobria, particularly those strains isolated from faeces of patients with diarrhoea. Most A. hydrophila associated with diarrhoea also belonged to group 1 but A. hydrophila from non-diarrhoeal stools or from water most commonly agglutinated human and guinea pig cells but not horse erythrocytes (primary groups 2 and 3). Fucose-resistant haemagglutination (FRHA) of strains in primary group 1 occurred with about 68% of the strains of A. sobria associated with diarrhoea. Mannose-resistant haemagglutination (MRHA) was limited to 29% of strains of A. sobria associated with diarrhoea. The predominance of primary group 1 among strains of Aeromonas spp. associated with diarrhoea and the proportion of these strains showing FRHA suggest that haemagglutination of cells from human, horse, rat and guinea pig, particularly if fucose-resistant, should be considered in a search for characteristics, which possibly contribute to virulence of Aeromonas spp.