Search

A gene probe derived from the colonization factor antigen I (CFA/I) operon cross-hybridized at very low stringency to plasmid DNA from coli surface antigen 17 (CS17)-producing enterotoxigenic Escherichia coli (ETEC) and from the ETEC strain F595C, which was negative for previously described CFAs, CSs, and putative colonization factors (PCFs). A 16-kDa protein was identified in sodium dodecyl sulfate-polyacrylamide gel electrophoresis of heat extracts prepared after growth of strain F595C at 37 degrees C on CFA agar containing bile salts. Transmission electron microscopy revealed bile salt- and temperature-dependent expression of fimbriae with a diameter of 7 nm. After transformation with a recombinant plasmid harboring the cfaR gene, which encodes a positive regulator of several CFAs, PCFs, and CSs, the 16-kDa protein was hyperexpressed. Polyclonal antibodies raised against this protein bound to the fimbriae and inhibited the adhesion of F595C bacteria to tissue-cultured Caco-2 cells. Nucleotide sequence determination of the gene encoding the 16-kDa fimbrial subunit revealed a high degree of amino acid sequence homology to the CFA/I, CS1, CS2, CS4, CS14, and CS17 polypeptides. The term CS19 is proposed for the new fimbria.

Colonization factor antigens (CFAs) mediate attachment of enterotoxigenic Escherichia coli (ETEC) to the intestinal mucosa and induce protective immunity against ETEC diarrhea. ETEC strains (n = 111) isolated from North Indian children from 1985 to 1989 were examined for CFAs and putative colonization factors (PCFs). CFA/IV was the most common factor (26%), followed by coli surface antigen 17 (CS17) (19%), CFA/I (14%), PCFO166 (7%), and CFA/II (5%), while 24% of the isolates were negative for CFAs and PCFs. Among the strains producing heat-stable and heat-labile toxin (ST+LT+ strains), the STaI gene was strongly associated with the absence of known CFAs and PCFs, making the STaI+LT+ isolates an interesting target for the identification of previously undescribed factors. Repetitive sequence--based polymerase chain reaction revealed that the CS17+ strains, although clonally related, represented endemically circulating strains with a diversity greater than that of the CFA/I+ strains, which showed a substantial clonal clustering.

An enterotoxigenic Escherichia coli (ETEC) strain producing a previously undescribed putative colonization factor was isolated from a child with diarrhea in India. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of bacterial heat extracts revealed a polypeptide band of 20.8 kDa when the bacteria were grown at 37 degrees C which was absent after growth at 22 degrees C. A specific rabbit antiserum raised against the purified 20.8-kDa protein bound specifically to the fimbriae, as shown by immunoelectron microscopy, and inhibited bacterial adhesion to tissue-cultured Caco-2 cells. Transformation with a recombinant plasmid harboring the cfaD gene, which encodes a positive regulator for several ETEC fimbriae, induced hyperexpression of the 20.8-kDa fimbrial subunit and a substantial increase in the proportion of bacterial cells that were fimbriated. The N-terminal amino acid sequence of the polypeptide showed 65 and 60% identity to the PCFO20 and 987P fimbriae of human and porcine ETEC, respectively. We propose the term CS20 for this new putative colonization factor of human ETEC.

Enteroaggregative Escherichia coli (EAggEC) have been implicated as diarrheal pathogens in several settings. Some EAggEC produce a distinct heat-stable enterotoxin named EAST1. The distribution and prevalence of the EAST1 gene in selected groups of bacterial enteropathogens were determined by colony hybridization. One hundred percent of 75 O157:H7 enterohemorrhagic E. coli (EHEC), 41% of 227 EAggEC, 41% of 149 enterotoxigenic E. coli, 22% of 65 enteropathogenic E. coli (EPEC), and 38% of 47 E. coli stool isolates from asymptomatic children hybridized with an EAST1 DNA probe. None of 55 enteroinvasive E. coli, 12 Yersinia enterocolitica, or 20 Vibrio cholerae non-O1 strains were EAST1 probe-positive. Concordance between EAST1 genotype and enterotoxicity was shown in examined strains of EAggEC, EHEC, and EPEC. The gene encoding EAST1 is more broadly distributed among diarrheogenic E. coli than previously known and may represent an additional determinant in the pathogenesis of E. coli diarrhea.

Enteroaggregative Escherichia coli (EAggEC) is a distinct category of diarrheal pathogen implicated as the cause of persistent diarrhea. The pathogen exhibits a characteristic "stacked-brick" pattern of aggregation when incubated with HEp-2 cells. The outer membrane protein (OMP) profile of a prototype EAggEC strain (F03) reflected the presence of one major 30-kDa protein. The OMP is expressed in the presence of the 60-MDa plasmid that the strain harbors. Antibodies were raised against the OMP by injecting the protein into a rabbit. The manifestation of an adherence phenotype on HEp-2 cells was observed for F03 and other strains that express OMP in the presence and absence of anti-OMP serum. Clumps of bacteria forming an aggregative pattern were observed in the HEp-2 cell assay in the absence of OMP antibodies, whereas a few bacteria attached to the cells in the presence of OMP antibodies. Mannose-resistant hemagglutination of human erythrocytes observed in the presence of EAggEC strains was inhibited in the presence of anti-OMP serum. Sequence analysis of a peptide generated by protease digestion of OMP exhibited 90% homology to a peptide of flagellin protein encoded by the hag gene of Serratia marcescens. Immunolabeling of the outer membrane by colloidal gold confirmed the protein to be an OMP. Our results suggest that the OMP of EAggEC have common antigenic properties. Antibodies raised against the protein can prevent adherence in vitro and could potentially interrupt the natural disease.

Enteroaggregative Escherichia coli (EAggEC) is a recently discovered diarrhoeal pathogen implicated as a cause of persistent diarrhoea in children. EAggEC strains exhibit a characteristic pattern of adherence when incubated with HEp-2 cells. Because of the difficulty in identifying this group of bacteria, the epidemiological significance of this pathogen as a diarrhoeal agent has not been fully realised. A gene probe was developed from the 60-MDa plasmid associated with EAggEC strains that encodes the genes for adherence and fimbriae. The sensitivity of the gene probe was 93% and the specificity 98% for detecting EAggEC isolates and is potentially useful for diagnostic and epidemiological studies.

A polynucleotide probe comprising the gene encoding a major structural subunit protein of coli surface antigen 6 (CS6) of enterotoxigenic Escherichia coli (ETEC) was developed. Eighty-nine ETEC isolates were examined in parallel with the probe in a colony hybridization assay and in a recently developed polyclonal-antibody-based inhibition enzyme-linked immunosorbent assay (ELISA). The two assays showed a high level of concordance in the detection of CS6-positive ETEC (kappa = 0.84, P < 0.00001). Thus, 36 of the 89 ETEC isolates were identified as CS6-positive by both assays. Six strains that were negative for other colonization factor antigens were positive with the CS6 probe but negative in the ELISA, suggesting lack of surface CS6 expression in these strains. One strain was probe negative but positive in the ELISA, while the remaining 46 strains were negative in both assays. The phenotypic and genotypic assays will prove useful in vaccine-oriented studies of ETEC disease.

Enteropathogenic Escherichia coli (EPEC) strains designated on the basis of their serotypes are epidemiologically associated with diarrhea. They adhere to the intestinal mucosa, producing the characteristic attaching and effacing (AE) lesion in an in vitro organ culture system. EPEC manifest localized adherence (LA) in the HEp-2 cell assay, and this is commonly used for clinical diagnosis. Recently, the fluorescence actin staining (FAS) test was proposed for the identification of E. coli causing the AE lesion. We therefore compared the FAS test with the HEp-2 cell assay and the EPEC adherence factor (EAF) probe assay for the detection of EPEC strains. Among 240 stool samples from children with diarrhea examined, 176 yielded E. coli and 14 of these strains showed the LA pattern in the HEp-2 cell assay; 11 of these were positive by both the EAF and the FAS tests. By using the HEp-2 cell assay as the "gold standard," the FAS test gave a sensitivity of 78.5% and a specificity of 100%. The three localized adherent FAS-negative strains tested subsequently were positive by the enteroaggregative E. coli DNA Probe and failed to produce the AE lesions characteristic of EPEC. When these strains were not considered, the sensitivity of the FAS test for detecting isolates that manifest LA was 100%. Against the EAF probe, the sensitivity and specificity of the FAS test were 91.6 and 100%, respectively. The FAS test avoids infrequent but nevertheless important phenotypic misclassifications in the HEp-2 cell assay, and it may therefore serve as a confirmatory test for EPEC.

The simplicity of enterotoxigenic Escherichia coli (ETEC) stool blot hybridization, where the total bacterial growth of a fecal inoculum is examined directly for the presence of enterotoxin genes, has been marred by reports of unsatisfactory sensitivity and/or specificity. To assess the accuracy of stool blot hybridization and to study the effect of varying proportions of ETEC among fecal E. coli (ETEC/E. coli) on test performance, a detailed 'blind' study of 166 stool specimens from children with diarrhea was performed. Oligonucleotide probes were found to be superior to polynucleotide probes, having a sensitivity of 80%, a specificity of 99%, a positive predictive value of 89% and a negative predictive value of 97%. The sensitivity was found to be 100% when ETEC/E. coli > 2/12, as compared with 20% when ETEC/E. coli < or = 2/12 (p = 0.001), showing that the proportion of ETEC among fecal E. coli is of paramount importance for test sensitivity.

The role of Hep-2 cell adherent Escherichia coli (EAEC) of localized (EAEC-L), diffuse (EAEC-D) and aggregative (EAggEC) phenotype, was investigated in 254 children aged < or = 48 months seeking treatment for non-bloody diarrhea at an outpatient clinic, and in 107 age-matched controls. The stool excretion rates of single isolates from patients/controls for the different phenotypes of Hep-2 cell adherent E. coli were: EAEC-L, 5.9/2.8%, p = 0.33; EAEC-D, 7.5/12.1%, p = 0.22; and EAggEC, 11.8/3.7%, p = 0.03. When patients were categorized by pre-admission diarrheal duration > or = 14/< 14 days, the excretion rates were EAEC-L, 0/7.1%; EAEC-D, 9.5/7.1%, and EAggEC, 21.4/9.9%, the difference approaching significance only for EAggEC (p = 0.06). These findings suggest that EAggEC may be an important cause of diarrhea in children, with a predilection to cause prolonged disease.

The ability of HEp-2 cell adherent Esch. coli of aggregative phenotype (EA-Agg EC) to cause diarrhoea and to colonize the bowel of rabbits was studied. Thirty six rabbits were challenged with one of three EA-Agg EC strains (F23A; H766C and F17A-15, 3 and 3 rabbits respectively) or a control strain (K12-15 rabbits) in reversible ileal-tie in adult rabbit diarrhoea (RITARD) model. The animals were sacrificed 72 h post challenge. Severe diarrhoea occurred in greater number of F23A challenged rabbits than the controls (P < 0.05). Mucosal cultures from proximal and distal small intestine and colon yielded about 1000 times more Esch. coli in the test than control rabbits (P < 0.001 in each case). EA-Agg EC were consistently grown from mucosa in the test rabbits who commonly showed mild to moderate villous stunting and grade + to nuclear fragmentation (karyorrhexis) in the small and large bowel epithelium. The control animals had either normal villi or very mild villous stunting. Results comparable to F23A were obtained with the other two EA-Agg EC strains tested in a smaller number of animals.

Plasmid DNA from two strains of enterotoxigenic Escherichia coli harboring genes encoding coli surface antigen 4 (CS4) and from seven Indian enterotoxigenic E. coli isolates cross-hybridized at low stringency but not at high stringency with two polynucleotide probes derived from the colonization factor antigen I (CFA/I) operon. Low-stringency Southern blot hybridization of PstI-digested plasmid DNA from the seven Indian isolates yielded characteristic restriction fragment patterns, distinct from those of CS4- and CFA/I-associated plasmid DNA. Two of the Indian strains were transformed with a recombinant plasmid harboring the cfaD gene, which encodes a positive regulator of CFA/I and CS4 genes. The cfaD transformants produced large amounts of putative colonization factor O166 (PCFO166) irrespective of whether the nutrient agar contained bile salts, a growth factor otherwise required for adequate PCFO166 expression. A considerable interstrain variation in the level of PCFO166 production could be explained by differences in the proportion of bacteria that were fimbriated, as visualized by electron microscopy. The N-terminal amino acid sequence of PCFO166 fimbrial protein showed a high degree of homology with the corresponding sequences of CFA/I and CS4.

We developed an accurate nonradioactive colony hybridization assay (NCHA) using a digoxigenin-labeled polynucleotide probe and an antidigoxigenin alkaline phosphatase conjugate for the identification of enterotoxigenic Escherichia coli (ETEC) harboring genes for colonization factor antigen I (CFA/I), coli surface antigen 4 (CS4), or putative colonization factor O166 (PCFO166). In this 2-day assay, visual registration of color intensity could be used to distinguish between CFA/I-positive strains and strains with the genetic potential to express CS4 or PCFO166. A rapid NCHA was developed by which the results could be read visually 7 h and 45 min after inoculation of the bacteria. In the rapid NCHA, densitometry verified the visual discrimination between four groups of E. coli; ETEC with the CFA/I gene, ETEC with the CS4 gene, ETEC with the PCFO166 gene, and E. coli strains that lack such genes. As a confirmatory test, plasmids from ETEC with the CFA/I, CS4, or PCFO166 gene were differentiated by their characteristic restriction fragment patterns in nonradioactive Southern blot hybridization.

The present study describes a nonisotopic DNA-DNA hybridization assay for the detection of enterotoxigenic Escherichia coli (ETEC) in which the probes were labelled with the hapten molecule digoxigenin and after hybridization, DNA hybrids were detected by antidigoxigenin alkaline phosphatase conjugate. A blinded study carried out on a battery of enterotoxigenic and nonenterotoxigenic Esch. coli by dig-probe hybridization assay were compared with the results of a radiolabelled toxin gene probe hybridization assay (performed earlier). The three digoxigenin labelled probes gave a 100 per cent specificity and sensitivities of 95.45, 100 and 100 per cent for LT, STh and STp respectively. These results were comparable to those with the radioactive probes.

A collection of 44 enteroaggregative Escherichia coli (EAggEC) strains isolated from infants with diarrhea in India and the United Kingdom were examined for their ability to adhere in vitro to human intestinal mucosa and by electron microscopy for production of putative adherence factors. None of the strains adhered to human duodenal mucosa, and six strains tested did not adhere to ileal mucosa; all 44 strains, however, adhered to human colonic mucosa in localized aggregates. Electron microscopy of infected colonic mucosa indicated fimbrially mediated adhesion of the EAggEC strains. Four morphologically distinct kinds of fimbriae, including a new morphological type of E. coli fimbriae consisting of bundles of fine filaments, were identified among the EAggEC strains; this new type of fimbria was observed in 43 of the 44 EAggEC strains. Forty-three of the 44 EAggEC strains were positive with a DNA probe developed to identify EAggEC, and most of the strains belonged to serotypes unrelated to the other major classes of diarrheic E. coli. These results suggest that EAggEC may be a large-bowel pathogen and colonize the colon by a fimbrially mediated adhesion mechanism.

We examined the ability of a colonization factor antigen I (CFA/I) polynucleotide probe to identify coli-surface antigen 4 producing (CS4+) strains of enterotoxigenic Escherichia coli (ETEC). At low stringency (LS) the probe hybridized to colony lysates of strains previously shown to produce CS4 or CFA/I fimbriae. Only DNA from CFA/I+ strains maintained a stable probe-target hybrid under high stringency (HS) conditions. On examination of several clones from three previous CS4 producers, identified as positive in LS and negative in HS colony hybridization, spontaneous loss of nucleotide sequences homologous to a gene encoding a positive CFA/I regulator, CfaD, was found to be associated with lacking expression of CS4. Our findings indicate that, on stored or subcultured isolates of ETEC, identification of CS4 strains may benefit from applying gene probe technology.

We compared a new colony hybridization assay with an established enzyme-linked immunosorbent assay for detection of enterotoxigenic Escherichia coli (ETEC) expressing colonization factor antigen I (CFA/I). The tests were applied to 135 human ETEC strains. Of these isolates, 30 had previously been characterized for CFAs. A strain harboring the plasmid vector of the polynucleotide gene probe, nine non-ETEC strains from healthy infants, and eight ETEC strains of animal origin were included for further evaluation of probe specificity. The two assays showed a high level of concordance in the specific detection of ETEC strains expressing CFA/I. A total of 24 strains tested positive in the CFA/I hybridization assay, while 23 of those strains were positive in the CFA/I enzyme-linked immunosorbent assay. The single discrepant result could be explained by the loss of a regulatory gene. The strain harboring the plasmid vector of the probe, the non-ETEC E. coli strains, and the ETEC strains of animal origin were all negative in the CFA/I probe assay.

Esch. coli strains manifesting localised (17), diffuse (8) or aggregative (17) phenotypes of adherence to HEp-2 cells were tested for their ability to adhere to human enterocytes isolated from duodenal biopsies of adult volunteers to obtain further evidence of their enteropathogenecity. Esch. coli strains H10407+; CFAI+ and LT+ STp+ STh+, F 294 B; a localised adherent strain positive with entero-adherent factor probe reported previously to attach to small intestinal enterocytes and F 582 C; LT- STp+ STh+ were the positive controls: H10407P (CFAI- mutant of H10407+) and K12 served as negative control strains. Adherence of variable degree was seen with 35.3 per cent of enteroaggregative Esch. coli (EAggEc) and with 58.8 per cent of enteroadherent Esch. coli localised (EAEC-L); EAEC-diffuse (EAEC-D) did not adhere to the human enterocytes. The possibility that EAgg EC and diffuse phenotypes may adhere better to lower small intestine or the large intestine, needs to be investigated.

The present study describes a colony hybridization setup for identification of enterotoxigenic Escherichia coli obviating the need for advanced equipment and radioactive isotopes. With a modest laboratory arrangement, polynucleotide gene probes were produced in large quantities. The probes were labeled with digoxigenin and, after hybridization, detected with an antidigoxigenin alkaline phosphatase conjugate. With an established isotope-based oligonucleotide hybridization assay as reference, a blinded study on a large battery of enterotoxigenic and nonenterotoxigenic bacteria revealed a satisfactory sensitivity and specificity of the nonradioactive assay.

A hospital-based case-control study including 92 children with diarrhea for longer than 14 days and 92 controls without gastrointestinal symptoms was performed to describe the association between the excretion of enteric pathogens and persistent diarrhea. In patients the most frequently isolated stool pathogens were enteroaggregative Escherichia coli (19.6%), nontyphoidal Salmonella spp. (17.4%), E. coli with diffuse adherence pattern (7.6%), G. lamblia (7.6%) and enterotoxigenic E. coli (5.4%). The excretion rates in patients were significantly greater than in controls only for nontyphoidal Salmonella spp. (P = 0.0006) and enteroaggregative E. coli (P = 0.016).

A cohort of 452 rural children was followed longitudinally for 13 mo to ascertain the role of HEp-2 cell adherent Escherichia coli and other pathogens in causing acute (less than or equal to 14 d) and persistent (greater than 14 d) diarrhea. Aeromonas, Campylobacter jejuni, E. coli manifesting localized adherence to HEp-2 cells and enterotoxigenic E. coli were significantly associated with acute diarrhea. E. coli strains that exhibit aggregative adherence, so-called enteroaggregative E. coli, a newly-described category of diarrheagenic E. coli distinct from enterotoxigenic, enteroinvasive, enterohemorrhagic, and enteropathogenic E. coli, were found significantly more often in patients with persistent diarrhea (29.5%) than with acute diarrhea (12.8%) (P = .0052) or controls (9.9%) (P = .0006).

Restriction endonuclease-generated polynucleotide and synthetically produced oligonucleotide gene probes used in colony hybridization assays proved to be efficient for the detection and differentiation of enterotoxigenic Escherichia coli. To compare their relative efficiencies, these two sets of probes were radiolabeled with 32P and were applied to 74 strains of E. coli with known enterotoxin profiles and to 156 previously unexamined E. coli isolates. The enterotoxigenic bacteria Vibrio cholerae O1, Vibrio cholerae non-O1 (NAG), Yersinia enterocolitica, and E. coli harboring the plasmid vectors of the polynucleotide gene probes were examined for further evaluation of probe specificity. The two classes of probes showed a perfect concordance in their specific detection and differentiation of enterotoxigenic E. coli. In the analysis of six strains, the signal strength on autoradiography after hybridization with oligonucleotides was weaker than that obtained after hybridization with polynucleotide probes. The probes did not hybridize with DNA from V. cholerae O1, V. cholerae non-O1 (NAG), or Y. enterocolitica. The strains of E. coli harboring the plasmid vectors of the polynucleotide gene probes were, likewise, negative in the hybridization assays.