Four-Year Experience with Simultaneous Culture and Shiga Toxin Testing for Detection of Shiga Toxin-Producing Escherichia coli in Stool Samples

We report our experience with universal Shiga toxin-producing Escherichia coli (STEC) screening using culture and Shiga toxin antigen testing over 4 years. Twelve cases were detected—8 detected by both culture and Shiga toxin immunochromatographic assay (IA), 3 by culture, and 1 by IA only. The addition of Shiga toxin testing is of questionable benefit over culture alone for detection of STEC in areas of low prevalence.     

Evaluation of Enzyme Immunoassays and Real-Time PCR for Detecting Shiga Toxin-Producing Escherichia coli in Southern Alberta,Canada

Two immunoassays (Shiga Toxin Chek and Shiga Toxin Quik Chek) and real-time PCR were used to detect Shiga toxin-producing Escherichia coli. For enriched culture, the sensitivity and specificity of the three methods ranged from 80.0% to 98.2% and 98.0% to 100.0%, respectively. STEC isolates were identified in 2.6% of the 784 samples.     

Rapid and Sensitive Detection of Shiga Toxin-Producing Escherichia coli from Nonenriched Stool Specimens by Real-Time PCR in Comparison to Enzyme Immunoassay and Culture

Shiga toxin (Stx)-producing Escherichia coli (STEC) bacteria are a frequent cause of food-borne gastroenteritis, hemorrhagic colitis, and hemolytic uremic syndrome. Because antimicrobial agents are generally contraindicated in patients infected with STEC, a sensitive and specific diagnostic test with rapid turnaround is essential. Current culture methods may fail to detect non-O157 STEC. We evaluated a Stx gene real-time PCR assay using hybridization probes and the LightCycler instrument with 204 prospectively collected stool specimens, which were also tested for Stx by enzyme immunoassay (EIA) (ProSpecT STEC; Remel, Lenexa, KS) and by culturing on chromogenic agar (Chromagar O157; BD BBL, Sparks, MD). In addition, 85 archived stool specimens previously tested for Stx (by EIA) and/or E. coli O157:H7 (by culture) were tested by PCR. Sample preparation for PCR included mixing the stool in sterile water and extraction of nucleic acid using the MagNA Pure LC instrument (Roche Diagnostics). The PCR assay had 100% sensitivity and specificity compared to EIA and culture for specimens collected prospectively (4 of 204 specimens were positive) and compared to culture and/or EIA for archival specimens (42 of 85 specimens were positive). Both the EIA and PCR produced positive results from a specimen containing an O103 serotype STEC in the prospective specimens, and the PCR test detected three positive specimens that contained nonviable STEC in the archived specimens. The PCR assay demonstrated 100% sensitivity and specificity compared to EIA and/or culture and more rapid turnaround than either EIA or culture.Shiga toxin (Stx)-producing Escherichia coli (STEC) is a frequent cause of food-borne outbreaks of diarrhea (15). Disease caused by STEC is characterized by abdominal pain and bloody diarrhea, and 5 to 15% of those individuals infected with serotype O157:H7 develop hemolytic uremic syndrome (HUS), a potentially life-threatening condition consisting of hemolytic anemia, thrombocytopenia, and kidney failure caused primarily by Stx (8). STEC may carry genes for one or both types of Stx, Stx1 and Stx2 (17).Although STEC strains are a diverse group of pathogens, up to the present, the most common serotype in the United States has been O157:H7. A common association is that of E. coli O157:H7 contaminating ground beef (3, 7), but recent large outbreaks have involved a variety of other foods, including leafy greens (6, 29). The diversity of potentially contaminated food means that patients may acquire STEC infection from many foodstuffs, far beyond the stereotypical risk of undercooked ground beef. The common denominator of tainted food products seems to be direct or indirect contamination from bovine feces. To best detect infected patients and potential outbreaks, clinical laboratories must have tools to quickly and accurately detect STEC in stool specimens. Culture on sorbitol MacConkey agar is an inexpensive, effective, and widely used method based on lack of sorbitol fermentation by E. coli O157:H7. Several drawbacks limit the utility of culture, including slow turnaround, false-negative results in antibiotic-treated patients, and false-negative results due to emerging serotypes of non-O157 STEC that ferment sorbitol (1, 14, 16, 29). Alternatively, a method that is increasingly utilized is detection of Stx antigen from stool, either directly or after broth enrichment. Our experience concurs with enzyme immunoassay (EIA) product insert data that optimal sensitivity and specificity are achieved only when a broth enrichment step is employed; this results in slow turnaround.Here, we describe a real-time PCR assay that can detect STEC using nucleic acid extracts of stool specimens. We evaluated the performance of this assay using both archived stool specimens and prospectively collected specimens and compared the results to those of culture and Stx antigen detection.(This study was presented in part at the 48th Annual Interscience Conference on Antimicrobial Agents and Chemotherapy, Washington, DC, 25 to 28 October 2008.)     

Clinical Evaluation of a Real-Time PCR Assay for Identification of Salmonella,Shigella, Campylobacter (Campylobacter jejuni and C. coli), and Shiga Toxin-Producing Escherichia coli Isolates in Stool Specimens

Enteric illness affects millions of individuals annually in the United States and results in >50,000 hospitalizations. The rapid and accurate identification of bacterial pathogens associated with gastroenteritis can aid acute patient management decisions, including the use of antibiotic therapy and infection control. This study compared the ProGastro SSCS multiplex real-time PCR assay (Gen-Probe Prodesse, San Diego, CA) to culture for the identification of Campylobacter spp. (Campylobacter jejuni and Campylobacter coli), Salmonella spp., and Shigella spp. and to broth enrichment followed by an FDA-cleared enzyme immunoassay (EIA) for the identification of Shiga toxin-producing Escherichia coli (STEC) isolates in stool specimens. Stool samples submitted in preservatives for routine culture and EIA were prospectively enrolled and tested at four clinical centers. Discrepancies between the ProGastro SSCS assay and culture or EIA were resolved using bidirectional sequencing. The overall prevalence of the pathogens as detected by culture was 5.6% (1.8% Campylobacter, 1.8% Salmonella, 1.3% Shigella, and 0.8% STEC). When results based on the ProGastro SSCS assay and bidirectional sequencing were applied, the overall prevalence increased to 8.3% (2.3% Campylobacter, 2.6% Salmonella, 1.8% Shigella, and 1.6% STEC). Following resolution of the discrepant results, the sensitivity of the ProGastro SSCS assay was 100% for all pathogens, and the specificities ranged from 99.4% to 100%. The sensitivity of culture compared to sequence-confirmed ProGastro SSCS results ranged from 52.9% to 76.9%, with the specificities ranging from 99.9% to 100%. Overall, these results suggest that the ProGastro SSCS assay is highly sensitive and specific in a clinical setting.     

Comparison of a Commercial Real-Time PCR Assay for tcdB Detection to a Cell Culture Cytotoxicity Assay and Toxigenic Culture for Direct Detection of Toxin-Producing Clostridium difficile in Clinical Samples

Rapid detection of toxin-producing strains of Clostridium difficile is essential for optimal management of patients with C. difficile infection. The BD GeneOhm (San Diego, CA) Cdiff assay, a real-time PCR assay that amplifies tcdB, was compared to a cell culture neutralization assay (Wampole C. difficile Toxin B [TOX-B] test; TechLab, Blacksburg, VA) and to toxigenic culture. Using liquid (n = 273) and soft (n = 131) stool specimens from 377 symptomatic patients, all testing was performed on the same day by independent laboratory staff according to the manufacturers' protocols. Toxigenic bacterial culture was performed as follows. A 0.5-ml aliquot of stool was heated to 80°C for 10 min, followed by inoculation onto modified cycloserine cefoxitin fructose agar with and without horse blood (Remel, Lenexa, KS) and into prereduced chopped-meat broth. Of the 404 stool specimens tested, 340 were negative and 40 were positive (10.0% prevalence) both by PCR for tcdB and by cytotoxin production. The overall agreement between the BD GeneOhm Cdiff assay and the TOX-B test was 94.8% (380/401). When the TOX-B test was used as the reference method, the initial sensitivity, specificity, and positive and negative predictive values of the BD GeneOhm Cdiff assay were 90.9% (40/44), 95.2% (340/357), 70.2% (40/57), and 98.8% (340/344), respectively. When toxigenic culture was used as the “gold standard,” the sensitivity, specificity, and positive and negative predictive values of the BD GeneOhm Cdiff assay were 83.6%, 98.2%, 89.5%, and 97.1%, respectively, and those of the TOX-B test were 67.2%, 99.1%, 93.2%, and 94.4%, respectively. PCRs for three samples were inhibited upon initial testing; one sample was resolved upon retesting. One sample produced nonspecific cytotoxin results. The BD GeneOhm Cdiff assay performed well compared to a standard cell culture neutralization assay and to toxigenic culture for the detection of toxigenic C. difficile directly from fecal specimens.     

Multiplex PCR for Enterotoxigenic, Attaching and Effacing, and Shiga Toxin-Producing Escherichia coli Strains from Calves

A multiplex PCR was developed to identify enterotoxigenic, attaching and effacing, and Shiga toxin-producing Escherichia coli strains by amplifying genes encoding K99 and F41 fimbriae, heat-stable enterotoxin a, intimin, and Shiga toxins 1 and 2. This multiplex PCR was specific and sensitive. It will be useful for identification of E. coli strains which cause diarrhea in calves.     

Pathogenesis and Diagnosis of Shiga Toxin-Producing Escherichia coli Infections

Since their initial recognition 20 years ago, Shiga toxin-producing Escherichia coli (STEC) strains have emerged as an important cause of serious human gastrointestinal disease, which may result in life-threatening complications such as hemolytic-uremic syndrome. Food-borne outbreaks of STEC disease appear to be increasing and, when mass-produced and mass-distributed foods are concerned, can involve large numbers of people. Development of therapeutic and preventative strategies to combat STEC disease requires a thorough understanding of the mechanisms by which STEC organisms colonize the human intestinal tract and cause local and systemic pathological changes. While our knowledge remains incomplete, recent studies have improved our understanding of these processes, particularly the complex interaction between Shiga toxins and host cells, which is central to the pathogenesis of STEC disease. In addition, several putative accessory virulence factors have been identified and partly characterized. The capacity to limit the scale and severity of STEC disease is also dependent upon rapid and sensitive diagnostic procedures for analysis of human samples and suspect vehicles. The increased application of advanced molecular technologies in clinical laboratories has significantly improved our capacity to diagnose STEC infection early in the course of disease and to detect low levels of environmental contamination. This, in turn, has created a potential window of opportunity for future therapeutic intervention.     

Comparison of a Commercial Reversed Passive Latex Agglutination Assay to an Enzyme Immunoassay for the Detection of Shiga Toxin-Producing Escherichia coli

A multicenter study was performed to compare the performance of a prototypic reversed passive latex agglutination assay (VTEC Screen Seiken; Denka-Seiken, Japan) with the Premier EHEC Enzyme Immunoassay (Meridian Diagnostics, USA) for the detection of Shiga toxin in 554 diarrheal stool samples. Standard culture on sorbitol MacConkey agar and the use of latex agglutination reagents were included to identify the Escherichia coli O157, O26 and O111 serotypes. There was 99% agreement between the VTEC screen and enzyme immunoassay (kappa=0.823). Seventeen samples were positive for toxin by one or both assays. One toxin-positive sample using the enzyme immunoassay and four positive samples using the VTEC Screen could not be confirmed. Serotypes identified included: O157:H7 (n=8), O26 (n=2), O111 (n=1) and O45:H2 (n=1). The VTEC screen is easy to perform and comparable to the Meridian EHEC test for detection of Shiga toxin in clinical samples.     

Characterization of Urinary Tract Infection-Associated Shiga Toxin-Producing Escherichia coli

Enterohemorrhagic Escherichia coli (EHEC), a subgroup of Shiga toxin (Stx)-producing E. coli (STEC), is a leading cause of diarrhea and hemolytic-uremic syndrome (HUS) in humans. However, urinary tract infections (UTIs) caused by this microorganism but not associated with diarrhea have occasionally been reported. We geno- and phenotypically characterized three EHEC isolates obtained from the urine of hospitalized patients suffering from UTIs. These isolates carried typical EHEC virulence markers and belonged to HUS-associated E. coli (HUSEC) clones, but they lacked virulence markers typical of uropathogenic E. coli. One isolate exhibited a localized adherence (LA)-like pattern on T24 urinary bladder epithelial cells. Since the glycosphingolipids (GSLs) globotriaosylceramide (Gb3Cer) and globotetraosylceramide (Gb4Cer) are well-known receptors for Stx but also for P fimbriae, a major virulence factor of extraintestinal pathogenic E. coli (ExPEC), the expression of Gb3Cer and Gb4Cer by T24 cells and in murine urinary bladder tissue was examined by thin-layer chromatography and mass spectrometry. We provide data indicating that Stxs released by the EHEC isolates bind to Gb3Cer and Gb4Cer isolated from T24 cells, which were susceptible to Stx. All three EHEC isolates expressed stx genes upon growth in urine. Two strains were able to cause UTI in a murine infection model and could not be outcompeted in urine in vitro by typical uropathogenic E. coli isolates. Our results indicate that despite the lack of ExPEC virulence markers, EHEC variants may exhibit in certain suitable hosts, e.g., in hospital patients, a uropathogenic potential. The contribution of EHEC virulence factors to uropathogenesis remains to be further investigated.     

Usability and Performance of CHROMagar STEC Medium in Detection of Shiga Toxin-Producing Escherichia coli Strains

The performance and usability of CHROMagar STEC medium (CHROMagar Microbiology, Paris, France) for routine detection of Shiga toxin-producing Escherichia coli (STEC) strains were examined. The ability of the medium to selectively propagate STEC strains differing by their serotypes and virulence genes was studied with a collection of diarrheagenic E. coli isolates (n = 365) consisting of 49 different serotypes and with non-STEC and other bacterial isolates (n = 264). A total of 272 diarrheagenic E. coli (75.0%) isolates covering 24 different serotypes grew on CHROMagar STEC. The highest detection sensitivities were observed within the STEC serogroups O26 (90.0%), O111 (100.0%), O121 (100.0%), O145 (100.0%), and O157 (84.9%), and growth on CHROMagar STEC was highly associated with the presence of the tellurite resistance gene (terD). The specificity of the medium was 98.9%. In addition, CHROMagar STEC was used in parallel with a Shiga toxin-detecting immunoassay (Ridaquick Verotoxin/O157 Combi; R-biopharm, Darmstadt, Germany) to screen fecal specimens (n = 47) collected from patients suffering from hemorrhagic diarrhea. Positive growth on CHROMagar STEC was confirmed by the Premier EHEC enzyme immunoassay (Meridian Bioscience, Inc., Cincinnati, OH), and discrepant results between the two screening methods were confirmed by stx gene-detecting PCR. All 16 of the 47 stool samples that showed positive growth on CHROMagar STEC were also positive in the confirmatory tests. CHROMagar STEC proved to be an interesting option for STEC screening, allowing good detection sensitivity and specificity and permitting strain isolation for further outbreak investigations when required.     

Recent Progress of Shiga Toxin Neutralizer for Treatment of Infections by Shiga Toxin-Producing Escherichia coli

Infection with Shiga toxin (Stx)-producing Escherichia coli (STEC), including O157:H7, causes bloody diarrhea and hemorrhagic colitis in humans, occasionally resulting in fatal systemic complications, such as neurological damage and hemolytic-uremic syndrome. Because Stx is a major virulence factor of the infectious disease, a series of Shiga toxin neutralizers with various structural characteristics has been developed as promising therapeutic agents. Most of these agents function to bind to the toxin directly and inhibit the binding to its receptor present on the target cells. Other neutralizers do not inhibit receptor binding but induce aberrant intracellular transport of the toxin, resulting in effective detoxification. Such a novel type of Stx neutralizer provides a new therapeutic strategy against STEC infections. Here, recent progress of the development of Stx neutralizers is reviewed.     

Identification and Characterization of a Newly Isolated Shiga Toxin 2-Converting Phage from Shiga Toxin-Producing Escherichia coli

Shiga toxins 1 (Stx1) and 2 (Stx2) are encoded by toxin-converting bacteriophages of Stx-producing Escherichia coli (STEC), and so far two Stx1- and one Stx2-converting phages have been isolated from two STEC strains (A. D. O’Brien, J. W. Newlands, S. F. Miller, R. K. Holmes, H. W. Smith, and S. B. Formal, Science 226:694–696, 1984). In this study, we isolated two Stx2-converting phages, designated Stx2Φ-I and Stx2Φ-II, from two clinical strains of STEC associated with the outbreaks in Japan in 1996 and found that Stx2Φ-I resembled 933W, the previously reported Stx2-converting phage, in its infective properties for E. coli K-12 strain C600 while Stx2Φ-II was distinct from them. The sizes of the plaques of Stx2Φ-I and Stx2Φ-II in C600 were different; the former was larger than the latter. The restriction maps of Stx2Φ-I and Stx2Φ-II were not identical; rather, Stx2Φ-II DNA was approximately 3 kb larger than Stx2Φ-I DNA. Furthermore, Stx2Φ-I and Stx2Φ-II showed different phage immunity, with Stx2Φ-I and 933W belonging to the same group. Infection of C600 by Stx2Φ-I or 933W was affected by environmental osmolarity differently from that by Stx2Φ-II. When C600 was grown under conditions of high osmolarity, the infectivity of Stx2Φ-I and 933W was greatly decreased compared with that of Stx2Φ-II. Examination of the plating efficiency of the three phages for the defined mutations in C600 revealed that the efficiency of Stx2Φ-I and 933W for the fadL mutant decreased to less than 10⁻⁷ compared with that for C600 whereas the efficiency of Stx2Φ-II decreased to 0.1% of that for C600. In contrast, while the plating efficiency of Stx2Φ-II for the lamB mutant decreased to a low level (0.05% of that for C600), the efficiencies of Stx2Φ-I and 933W were not changed. This was confirmed by the phage neutralization experiments with isolated outer membrane fractions from C600, fadL mutant, or lamB mutant or the purified His₆-tagged FadL and LamB proteins. Based on the data, we concluded that FadL acts as the receptor for Stx2Φ-I and Stx2Φ-II whereas LamB acts as the receptor only for Stx2Φ-II.     

Real-Time Multiplex PCR Assay and Melting Curve Analysis for Identifying Diarrheagenic Escherichia coli

A real-time multiplex PCR assay was designed to amplify the virulence genes eae, pEAF, aatA, daaC, elt, est, ipaH, stx₁, and stx₂ for the detection of all diarrheagenic Escherichia coli pathotypes. This assay proved to be more sensitive and rapid than a conventional multiplex PCR for diarrheagenic E. coli isolates from children with diarrhea.     

Assessment of Real-Time PCR Assay for Detection of Rickettsia spp. and Rickettsia rickettsii in Banked Clinical Samples

Two novel real-time PCR assays were developed for the detection of Rickettsia spp. One assay detects all tested Rickettsia spp.; the other is specific for Rickettsia rickettsii. Evaluation using DNA from human blood and tissue samples showed both assays to be more sensitive than nested PCR assays currently in use at the CDC.     

Development of a Multiplex Real-Time PCR for Detection and Differentiation of Malignant Catarrhal Fever Viruses in Clinical Samples

A multiplex real-time PCR was developed using a single pair of primers and fluorescent probes specific for five malignant catarrhal fever viruses and an internal positive control. The assay was able to simultaneously detect and differentiate the viruses in clinical samples with high sensitivity (97.2%) and specificity (100%).Malignant catarrhal fever (MCF), a lymphoproliferative syndrome primarily of ruminants, is caused by gammaherpesviruses included in the MCF virus group (5, 15). MCF viruses (MCFV) exist in nature as unapparent infections in well-adapted hosts but cause an often fatal disease in certain clinically susceptible species (20). Within the MCFV group, six viruses are clearly associated with clinical disease: ovine herpesvirus-2 (OvHV-2) (16, 18, 23, 24, 28), alcelaphine herpesvirus 1 (AlHV-1) (20, 21), caprine herpesvirus-2 (CpHV-2) (3, 8, 14), an MCFV of unknown origin causing disease in white-tailed deer (MCFV-WTD) (9, 11), ibex MCFV (MCFV-ibex) (17), and AlHV-2-like virus (10). MCF is increasingly being recognized as the cause of significant economic losses in several major ruminant species as well as a threat to certain susceptible species held in mixed-species confinement (6, 13, 18).The diagnosis of MCF can still pose a challenge to clinicians and pathologists, even though the classical clinical signs and the histopathology are highly suggestive (18, 22). To confirm a diagnosis, several PCR assays have been used (1, 2, 4, 7, 25, 26); however, none of them is capable of simultaneously differentiating among MCFV, and several reactions have to be performed until the diagnosis can be established, which is time-consuming and expensive.In this work, a multiplex real-time PCR that used one pair of primers in conjunction with fluorescently labeled probes specific for OvHV-2, CpHV-2, MCFV-WTD, MCFV-ibex, and AlHV-1 was optimized and validated for the identification of these pathogenic MCFV in clinical samples using a single reaction. AlHV-2-like virus was not included in the present study due to the unavailability of its sequence information and clinical samples.The finding of a polymorphic region in the viral DNA polymerase gene containing unique sequences for each virus of interest, used as probe targets, flanked by conserved regions was a critical step in the development of the assay (Fig. ​(Fig.1).1). The presence of the flanking conserved sequences allowed a single pair of primers to amplify the expected 80-bp fragment from the DNA polymerase gene from all viruses of interest. This characteristic represented a great advantage to assay optimization because interaction between primers was minimized. A synthetic internal positive control (IPC), consisting of an oligonucleotide of 58 bp containing the primer sequences flanking an irrelevant sequence used for specific probe binding (Fig. ​(Fig.1B1B and Table ​Table1),1), was included in the assay as an indicator of the presence of PCR inhibitory factors in the reaction mixtures. The probes were labeled with fluorescent dyes with different emission spectra to allow simultaneous detection in the multiplex format (Table ​(Table1).1). Due to the limit of five channels in the real-time PCR system used, the probes for AlHV-1 and MCFV-ibex were labeled with the same fluorophore (Cy5); both probes were routinely tested simultaneously, and when a positive result for Cy5 was obtained, the sample was retested using the two probes separately.Open in a separate windowFIG. 1.(A) Alignment of the 80-bp sequences from the DNA polymerase genes of five MCFV known to cause disease in ruminants. Conserved nucleotides among sequences are highlighted, and the primer and probe target sequences are indicated in italic and bold, respectively. GenBank accession numbers are {"type":"entrez-nucleotide","attrs":{"text":"DQ198083","term_id":"78127836","term_text":"DQ198083"}}DQ198083 for OvHV-2, {"type":"entrez-nucleotide","attrs":{"text":"AF283477","term_id":"13447464","term_text":"AF283477"}}AF283477 for CpHV-2, and {"type":"entrez-nucleotide","attrs":{"text":"AF005370","term_id":"2337967","term_text":"AF005370"}}AF005370 for AlHV-1. The sequence from MCFV-WTD was available from our previous studies, and the sequence from MCFV-ibex was obtained in this study. (B) IPC oligonucleotide sequence.

TABLE 1.

Primers and probes used for the multiplex real-time PCR

Pentaplex PCR Assay for Detection of Hemorrhagic Bacteria from Stool Samples

Diarrheal diseases cause illness and death among children younger than 10 years in developing countries. Conventional testing for the detection of hemorrhagic bacteria takes 2 to 5 days to yield complete information on the organism and its antibiotic sensitivity pattern. Hence, in the present study, we developed a molecular-based diagnostic assay that identifies common hemorrhagic bacteria in stool samples. A set of specific primers were designed for the detection of Salmonella spp., Shigella spp., enterohemorrhagic Escherichia coli (EHEC), and Campylobacter spp., suitable for use in a one-tube PCR assay. The assay in the present study simultaneously detected five genes, namely, ompC for the Salmonella genus, virA for the Shigella genus, eaeA for EHEC, 16S rRNA for the Campylobacter genus, and hemA for an internal control. Specific primer pairs were successfully designed and simultaneously amplified the targeted genes. Validation with 20 Gram-negative and 17 Gram-positive strains yielded 100% specificity. The limit of detection of the multiplex PCR assay was 1 × 10³ CFU at the bacterial cell level and 100 pg at the genomic DNA level. Further evaluation of the multiplex PCR with 223 bacterium-spiked stool specimens revealed 100% sensitivity and specificity. We conclude that the developed multiplex PCR assay was rapid, giving results within 4 h, which is essential for the identification of hemorrhagic bacteria, and it might be useful as an additional diagnostic tool whenever time is important in the diagnosis of hemorrhagic bacteria that cause diarrhea. In addition, the presence of an internal control in the multiplex PCR assay is important for excluding false-negative cases.     

Associations between Virulence Factors of Shiga Toxin-Producing Escherichia coli and Disease in Humans

Associations between known or putative virulence factors of Shiga toxin-producing Escherichia coli and disease in humans were investigated. Univariate analysis and multivariate logistic regression analysis of a set of 237 isolates from 118 serotypes showed significant associations between the presence of genes for intimin (eae) and Shiga toxin 2 (stx₂) and isolates from serotypes reported in humans. Similar associations were found with isolates from serotypes reported in hemorrhagic colitis and hemolytic-uremic syndrome. The enterohemorrhagic E. coli (EHEC) hemolysin gene was significantly associated with isolates from serotypes found in severe diseases in univariate analysis but not in multivariate logistic regression models. A strong association between the intimin and EHEC-hemolysin genes may explain the lack of statistical significance of EHEC hemolysin in these multivariate models, but a true lack of biological significance of the hemolysin in humans or in disease cannot be excluded. This result warrants further investigations of this topic. Multivariate analysis revealed an interaction between the eae and stx₂ genes, thus supporting the hypothesis of the synergism between the adhesin intimin and Shiga toxin 2. A strong statistical association was observed between the stx₂ gene and severity of disease for a set of 112 human isolates from eight major serotypes. A comparison of 77 isolates of bovine origin and 91 human isolates belonging to six major serotypes showed significant associations of the genes for Shiga toxin 1 and EspP protease with bovine isolates and an increased adherence on HEp-2 cell cultures for human isolates, particularly from diarrheic patients and healthy persons.