Semi-automated fluorogenic PCR assays (TaqMan) forrapid detection of Escherichia coli O157:H7 and other Shiga toxigenic E. coli

Molecular and Cellular Probes (1999) 13, 291–302 Article No. mcpr.1999.0251, available online at http://www.idealibrary.com on

Semi-automated fluorogenic PCR assays (TaqMan) for rapid detection of Escherichia coli O157:H7 and other Shiga toxigenic E. coli V. K. Sharma,∗ E. A. Dean-Nystrom and T. A. Casey Enteric Diseases and Food Safety Research Unit, National Animal Disease Center, USDA, Agricultural Research Service, Ames, Iowa, 50010 USA (Received 13 March 1999, Accepted 13 May 1999) Semi-automated detection of Enterohaemorrhagic Escherichia coli (EHEC) O157:H7 and nonO157:H7 Shiga toxin-producing E. coli (STEC) was achieved using fluorogenic polymerase chain reaction (PCR). These PCR assays were designed to amplify 80, 120 and 150 bp regions of virulence genes stx1, stx2 and eaeA, respectively, using specific primers. The fluorogenic probes were used for specific detection of amplified products of the stx1 and stx2 genes of STEC, and the eaeA gene of EHEC O157:H7. For multiplex PCR assay, the three sets of primers and fluorogenic probes were included in one reaction to simultaneously amplify and detect any of the three targeted virulence genes. In non-multiplex PCR assay, each of the three virulence genes was amplified and detected in independent reactions. The specificity of these assays was evaluated using suspensions of STEC and other bacterial species lacking stx1, stx2 and eaeA. The multiplex assay detected all STEC harbouring any combination of three virulence genes. Three non-multiplex PCR reactions identified types of Shiga toxin genes carried by a STEC and identified STEC as either EHEC O157:H7 or nonO157:H7 STEC. Sensitivity limits of these assays in beef and faeces inoculated with EHEC O157: H7 were 5.8 to 580 cfu and 1.2 to 1200 cfu, respectively. These assays can be completed within 8–10 h when performed simultaneously or within 13 h if the multiplex assay is used as an initial screen for detecting STEC and the non-multiplex assay is used for subsequent detection of stx1 and stx2 of STEC and eaeA of EHEC O157:H7.

KEYWORDS: STEC, EHEC, haemorrhagic colitis, haemolytic uremic syndrome, intimin, food safety.

INTRODUCTION Enterohaemorrhagic Escherichia coli (EHEC) serotypes, a subset of Shiga toxin-producing E. coli (STEC), are predominantly associated with haemorrhagic colitis, haemolytic-uremic syndrome and thrombotic thrombocytopenic purpura in humans.1,2 Although several EHEC serotypes have been associated with human infections, EHEC O157:H7 is implicated in

the vast majority of outbreaks and sporadic cases of bloody diarrhoea.3–6 The pathogenicity of EHEC O157: H7 is associated with a number of virulence factors, including Shiga toxins 1 and 2 (encoded by genes stx1 and stx2),7–9 and intimin (encoded by the gene eaeA).10,11 Shiga toxins are believed to play a major role in the pathogenesis of haemorrhagic colitis and HUS through cytopathic effect on vascular endothelial cells of the kidneys, intestines, central nervous system and

Disclaimer: Mention of trade names or commercial products in this article is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the US Department of Agriculture. ∗ Author to whom all correspondence should be addressed at: USDA, ARS, National Animal Disease Center, PO Box 70, Ames, IA 50010, USA. Tel: +1515 663 7406/7279; Fax: +1515 663 7458; E-mail: [email protected]

0890–8508/99/040291+12 $30.00/0

 1999 Academic Press

292

V. K. Sharma et al.

other organs.3,12 Intimin facilitates the adherence to intestinal villi by the process of attachment and effacement.13 Enterohaemorrhagic E. coli strains which produce one or more Shiga toxins and intimin are also considered more virulent for calves than strains that produce Shiga toxins only.14,15 Cattle are a major reservoir for EHEC O157:H7 and many other non-O157:H7 STEC.16 Most disease outbreaks have involved foods of bovine origin, such as beef and raw milk, that become contaminated with bovine faeces at slaughter houses or dairy farms. Methods for automated and high-throughput screening of foods and faeces of bovine origin to detect the presence of EHEC O157:H7 and STEC are needed for improving food safety assurance and studying the ecology and epidemiology of these pathogenic bacteria. Several methods have been developed for the detection of STEC and EHEC O157:H7. Detection of STEC can be accomplished either by testing broth culture supernatants of suspected foods or faeces in Vero cell cytotoxicity assays7,17 or by enzyme-linked immunosorbent assays using anti-Shiga toxin antibodies.18 However, immunological methods are unable to detect all Shiga toxins with equal efficiency and some Shiga toxins may not be detected at all because of the existence of antigenic variants of these toxins.3,19,20 Moreover, methods that rely on Shiga toxin detection alone cannot differentiate more virulent STEC such as EHEC O157:H7 from other less virulent STEC. Special biochemical media21,22 and diagnostic kits containing latex reagents directed against O157 and H7 antigens23 have been developed for selective isolation and specific detection, respectively, of EHEC O157:H7. These methods, however, require 4–5 days to complete and are difficult to automate. Genotypic detection of STEC and EHEC may be accomplished by DNA colony blot hybridization to identify genes encoding Shiga toxins and intimin.24 The use of radioactive isotopes and the time required make this method unsuitable for many diagnostic laboratories. Polymerase chain reaction has become a useful diagnostic tool because it is quick, specific, sensitive and relatively inexpensive. Several PCR assays have been developed for the detection of genes stx1 and stx2.25–27 These assays, however, do not distinguish EHEC from other STEC. The detection of EHEC by PCR has been achieved by amplification of the sequences located in the 5′ two-thirds of the gene eaeA.28 This region of eaeA gene is highly conserved among EHEC serotypes. Polymerase chain reaction assays that allow differentiation of EHEC O157:H7 from other EHEC use primers that have homology in the 3′ one-third of the eaeA gene.28,29 The 3′ end of the eaeA gene is less conserved among EHEC serotypes. Although the use of these primers in PCR facilitated the detection of all

EHEC O157:H7 strains, these primers also detected enteropathogenic E. coli (EPEC) strains of serotype O55:H7 and an EHEC strain of serotype O145:NM. Most of these methods are based on the use of 18h cultures and lengthy DNA isolation protocols, and require laborious, less sensitive and specific gel-based analysis of PCR amplified products. New and improved PCR methods incorporate fluorogenic probes for automated and specific detection of amplified products. One such system is the TaqMan PCR detection system developed for detection of Salmonella30 spp. and Listeria monocytogenes.31 In TaqMan detection assays a fluorogenic oligonucleotide probe, conjugated to a fluorescent reporter dye at the 5′ end and a fluorescent quencher dye at the 3′ end, binds to its complement in the region of the target gene selected for amplification by flanking primers. During the amplification, the 5′ to 3′ nuclease activity of Taq DNA polymerase displaces the probe from its complementary sequence and cleaves off the reporter dye. The fluorescence emission intensity of the free reporter dye increases because it is no longer quenched by the proximal quencher dye. The emission intensity is measured by a fluorometer and the emission data is analysed by the detection software to provide a ‘+’ or ‘−’ response, indicating amplification or no amplification of target genes. The objectives of the present study were to develop semi-automated TaqMan-based PCR that could be used in a multiplex format for specific detection of all STEC (except STEC that carry the gene stx2e and cause edema disease in swine32) and in a non-multiplex format for deducing the Shiga toxin profile of these strains and for distinguishing EHEC O157:H7 from nonO157:H7 STEC. We described the effects of different culture conditions, the duration of growth before extracting DNA and the presence of non-target bacterial flora on the sensitivity of fluorogenic PCR assay in detecting EHEC O157:H7 in foods and faeces of bovine origin. We optimized these assays to obtain reproducible and efficient amplification and detection of targeted genes using crude DNA preparations thereby eliminating the need to use high-purity DNA that could take additional time to prepare. Our entire assay can be completed in 8–13 h.

MATERIALS AND METHODS Bacterial strains, culture media and growth conditions The bacterial strains used in this study, relevant characteristics and sources of these strains are listed in Table 2. Strains were propagated and maintained on

Fluorogenic PCR assays for STEC and EHEC

Trypticase soy agar (TSA) plates. Liquid cultures were obtained by growing bacteria in Trypticase soy broth (TSB) for 18–20 h at 37°C with continuous agitation (200 rpm). Trypticase soy agar and MacConkey agar were used to enumerate bacteria. Trypticase soy agar, TSB and MacConkey agar were purchased from BBL (Becton Dickson Microbiology Systems, Cockeysville, MD, USA). Novobiocin (Sigma Chemical Co., St Louis, MO, USA) was used at 20 lg ml−1 in modified TSB (mTSB).33

293

Isolation of DNA Pure cultures of EHEC O157:H7 strain 2409, grown for 18 h at 37°C, were diluted 10-fold and 0·2 ml of each dilution was centrifuged to pellet bacteria. The pellets were suspended in 0·2 ml of lysis buffer (10 m Tris HCl, 1 m Na2 EDTA, pH 8·0 and 0·5% Triton X-100).34 Bacteria were recovered from 0·2 ml supernatants of spiked beef and faecal cultures that were grown as described above. The pellets were suspended in 0·2 ml of InstaMatrix (Bio-Rad Laboratories, Richmond, CA, USA). The bacterial suspensions were heated at 100°C for 10 min and then centrifuged at 12,000×g for 5 min. The supernatant containing DNA was stored at −70°C.

Culture of spiked beef and faecal samples Ground beef (80% lean) was purchased on three different occasions from a retail grocery store. The faecal samples were obtained from cattle that were housed at the National Animal Disease Center, Ames, IA, USA. One gram of each sample type was added to 9 ml of TSB, samples were vortexed for 30 s, and then 10-fold dilutions of these samples were plated on TSA and MacConkey plates and incubated for 18 h at 37°C to determine total aerobic and coliform bacteria. The remainder of the beef- or faeces-broth mixture was incubated for 18 h at 37°C with shaking. These samples were centrifuged for 2 min at 1000×g to remove large particles. Culture supernatant (0·2 ml) was added to 0·8 ml of TSB and centrifuged at 12,000×g for 3 min. The DNA was isolated (see below) from pelleted bacteria and it was used as a template in the multiplex PCR assay containing primers to amplify 80, 120, 150 bp fragments of stx1, stx2 and eaeA, respectively. The amplification products were analysed by electrophoresis through a 4% agarose gel (FMC BioProducts, Rockland, ME, USA) followed by ethidium bromide staining. The molecular size of visible bands was estimated from a DNA-sizing ladder (50 bp ladder, Gibco-BRL). The beef and faecal samples (1 g portions) determined negative for the predicted size fragments of stx1, stx2 and eaeA genes by the multiplex PCR assay were spiked with 0·1 ml of 10-fold dilutions of EHEC O157: H7 strain 2409 followed by the addition of 8·9 ml of TSB or mTSB. Some beef and faecal specimens were also spiked with the non-target E. coli strain 63 to obtain different ratios of target (EHEC O157:H7 strain 2409) to non-target strains. These spiked samples were incubated for 0, 2, 4, 6 or 18 h at 37°C. Aliquots (0·2 ml) from these cultures were diluted 1:5 in TSB, bacteria were recovered by centrifugation and processed for DNA isolation. All centrifugation steps were performed at 4°C.

Synthetic oligonucleotide primers and fluorogenic reporter probes The published nucleotide sequences of stx1,35 stx236 and eaeA37 were imported into the Primer Express Software (Perkin-Elmer, Foster City, CA, USA) to design primer pairs to amplify 80 and 120 bp regions of the stx1 and stx2 genes, respectively, of STEC and 150 bp fragment of the eaeA gene. This software was also used in designing fluorogenic reporter probes to specifically detect amplified sequences of stx1 and stx2 of STEC and eaeA of EHEC O157:H7. The eaeA probe was designed to distinguish EHEC O157:H7 from other EHEC serotypes and from bacterial species, such as Citrobacter freundii and Hafnia alvei, that may harbour an eaeA-like gene.28,29 However, this probe was not capable of distinguishing the eaeA gene of EHEC O157:H7 from the eaeA gene of EPEC O55:H7 and other related EPEC serotypes because eaeA genes of these serotypes share extensive nucleotide sequence homology with the eaeA gene of O157:H7.38 The nucleotide sequence and location of these primers and probes in the genes of their origin is shown in Table 1. Fluorogenic probes were generated by conjugating a reporter dye (FAM or 6carboxy-fluorescein) at the 5′ end and a quencher dye (TAMRA or 6-carboxytetramethyl-rhodamine) at the 3′ end (Integrated DNA Technologies, Coralville, IA, USA).

PCR amplification A PCR master mixture suitable for use in both multiplex and non-multiplex PCR assays was developed by testing different concentrations of MgCl2, different primer and probe ratios and different amounts of DNA polymerase. Polymerase chain reactions were

V. K. Sharma et al.

294

Table 1. Nucleotide sequence of primers and fluorogenic probes Sequence (5′ → 3′)

Primer or probea

VS1 (Forward) VS2 (Reverse) VS4 (Forward) VS5 (Reverse) VS8 (Forward) VS9 (Reverse) VS3 (Probe) VS6 (Probe) VS10 (Probe)

CAT TTT GGG TGT GGC CGT TGT CTA AAC

AGT GCC CAG TGC GGA TTT GGC TCA GCC

GGA GAA TTA CGT TTA GGC AAG GGC GAT

ACC AAC TTT ATT GAC ACT AGC GCG ACC

TCA GTA TGC AAC TTC ATT GAT TTT ATT

CG ACG CAG T AAG CTT CA TGT GGA GAA CCC GGC TA TGC CC GTT ACG GTT TG TGA CCA TCT TCG ACT TAT ACC GCG ACG

Gene detected

Locationb within the gene

GenBankc accession number

stx1 stx1 stx2 stx2 eaeA eaeA stx1 stx2 eaeA

830–854 917–886 442–462 562–542 1896–1916 2047–2028 856–881 481–507 1925–1954

M16625 M16625 X07865 X07865 X60439 X60439 M16625 X07865 X60439

a

Probes were conjugated at 5′ and 3′ ends with fluorescent dye FAM and TAMRA, respectively. The positions of the oligonucleotides are listed relative to the initiation codon (+1 adenine) of the respective gene. c The nucleotide sequences submitted with these accession numbers were used in designing primers and probes for targeted gene amplification and detection. b

performed by adding 5 ll of DNA to 45 ll of PCR master mixture. The master mixture for multiplex PCR assay contained 10 m Tris-HCl (pH 8·3), 50 m KCl, 10 m Na2 EDTA, 5 m MgCl2, 0·2 m deoxynucleoside triphosphates, 100 n of stx1, 150 n of stx2 and 200 n of eaeA primer pairs, 65 n of stx1and stx2-specific probes, 45 n of O157:H7-specific eaeA probe, 60 n of reference dye ROX (carboxyX-rhodamine) and 2·5 units of AmpliTaq Gold DNA polymerase (Perkin-Elmer, Foster City, CA, USA). The pre-PCR fluorescence intensity of this master mixture was in the range of 4000 to 4500 fluorescent units. The non-multiplex PCR assay was performed in three independent PCR reactions using the same master mix as was used in the multiplex assay except that each non-multiplex reaction contained only one primer pair and the corresponding probe. The prePCR fluorescence intensity of each non-multiplex assay was in the range of 2000 to 2500 units. The PCR reactions were dispensed into a 96-well MicroAmp plate and amplified using a PTC-200 thermal cycler (MJ Research Inc., Watertown, MA, USA). Samples were heated to 95°C for 10 min to denature the DNA and activate the AmpliTaq DNA polymerase. This was followed by 40 cycles of denaturation at 94°C for 20 s, annealing at 60°C for 30 s and polymerization at 72°C for 30 s. The final extension was carried out at 72°C for 5 min followed by cooling of samples to 4°C.

Fluorogenic detection of amplified fragments and data interpretation The PCR-amplified products were detected by performing fluorescence readings on LS-7200 Sequence Detection System (Perkin-Elmer, Applied Biosystems).

The fluorescence of each sample was measured prior to (pre-PCR read) and after PCR cycling (post-PCR read). A series of six no-template controls were included in each reaction plate to establish the baseline emission intensity of the quenched reporter dye. This baseline intensity measured at 600nm is subtracted from pre- and post-PCR read for each sample to yield the normalized pre- and post-PCR fluorescence intensity. The differential of normalized post- and prePCR values represented the net change in fluorescence (DRn). A threshold DRn was established based on the standard deviation of fluorescence intensity of six no-template controls at [99·5% confidence interval. Samples with DRn higher than the threshold were assigned a ‘+’ score indicating that targeted gene(s) were amplified. Samples with DRn equal to or less than the threshold were assigned a ‘−’ score indicating no detectable amplification of targeted gene(s) occurred in these samples.

DNA sequence analysis The nucleotide sequences of 80, 120 and 150 bp fragments of stx1, stx2 and eaeA, respectively, were determined from the same E. coli O157:H7 (strain 2409) that was used in spiking beef and faecal samples. The nucleotide sequence of these amplicons was determined by Sanger dideoxynucleotide chain termination method using a Dye Terminator Thermosequenase Kit (Amersham Pharmacia Biotech, Inc., Piscataway, NJ, USA). The sequenced samples were analysed using ABI 377 DNA sequencer (PerkinElmer, Applied Biosystems).

Fluorogenic PCR assays for STEC and EHEC 1

2

3

4

5

M

200 bp eae

150 bp

295

the eaeA-specific primer set amplified a 150-bp fragment in non-multiplex PCR assays. All three amplicons were easily detected even when three primer sets were combined in the multiplex assay. The amplified products were obtained only from EHEC O157: H7 strain 2409 and not from E. coli strain 63. The nucleotide sequence analysis revealed that each amplified fragment matched the predicted sequence amplified by each flanking primer pair (data not shown).

stx2 100 bp stx1

Specificity of fluorogenic multiplex PCR assay for STEC 50 bp

Fig. 1. Specificity of stx1-, stx2- and eaeA-specific primers to amplify 80, 120 and 150 bp fragments, respectively. The DNA from Escherichia coli strain 63 that lacks stx1, stx2 and eaeA genes and an EHEC O157: H7 strain 2409 that has stx1, stx2 and eaeA genes were used as a template in multiplex and non-multiplex polymerase chain reaction (PCR) assays and amplified products were analysed on 4% agarose gels. Lanes: M, molecular size markers (50 bp ladder; Gibco-BRL); 1, DNA from E. coli strain 63 amplified with all three primer pairs in the multiplex PCR assay; 2, DNA from EHEC O157:H7 strain 2409 amplified with stx1-specific primer pair; 3, DNA from strain 2409 amplified with stx2-specific primer pair; 4, DNA from strain 2409 amplified with eaeA-specific primer pair; 5, DNA from 2409 amplified with three primer pairs in multiplex PCR assay. The relative position in the gel of three predicted size PCR products is indicated by arrowheads on the left side of the gel. Arrowheads on the right side of the gel mark the position of 50, 100, 150 and 200 bp bands of the molecular size markers.

RESULTS Primer specificity The DNA from EHEC O157:H7 strain 2409 containing stx1, stx2 and eaeA genes or E. coli strain 63 lacking these three genes was used as a template in multiplex and non-multiplex PCR assays to determine the ability of primers, listed in Table 1, to amplify 80, 120 and 150 bp conserved regions of stx1, stx2 and eaeA, respectively. As shown in Fig. 1, the stx1-specific primer set amplified an 80-bp fragment, the stx2specific primer set amplified a 120-bp fragment, and

The specificity of the fluorogenic multiplex PCR assay was evaluated by using 26 STEC and 21 non-STEC strains. When DNA from STEC strains was subjected to PCR amplification and detection in fluorogenic multiplex PCR assay, DRn greater than the threshold was generated for all STEC strains. These samples were scored positive for amplification by the fluorogenic detection system (Table 2). On the other hand, all non-STEC isolates, except EPEC O55:NM and O55:H6, resulted in DRn values less than or equal to the threshold values and these samples were scored negative for amplification by the detection system.

Specificity of fluorogenic non-multiplex PCR assay for stx1 and stx2 of STEC and eaeA of EHEC O157:H7 We evaluated the ability of the stx1-, stx2- and EHEC O157:H7-specific eaeA non-multiplex fluorogenic PCR assays to identify the virulence gene profile of 27 STEC strains. As shown in Table 3, the stx1-specific PCR detected only those STEC possessing stx1, the stx2-specific PCR detected STEC that carried the gene stx2, and the EHEC O157:H7-specific eaeA PCR detected EHEC O157:H7, EHEC O157:NM, and three of the four EPEC isolates belonging to the serogroup O55. The cross-reactivity of EHEC O157:H7-specific eaeA probe with the eaeA gene of EHEC O157:NM and EPEC O55:H7 has been observed in other PCR assays.39,40 However, the EHEC O157:H7-specific probe did not cross react with the eaeA gene of several other serotypes of EHEC or Hafnia alvei.

Sensitivity limits of fluorogenic PCR assays The detection limits of fluorogenic multiplex PCR performed with DNA from 18 h broth cultures of EHEC O157:H7 strain 2409 were 2×104 cfu ml−1 (Fig. 2). The detection limits of three non-multiplex

V. K. Sharma et al.

296

Table 2. Specificity of fluorogenic multiplex polymerase chain reaction assay for detecting STEC Bacterial strains tested

Non-toxigenic Escherichia coli Enterotoxigenic Escherichia coli (ETEC) Enteropathogenic Escherichia coli (EPEC)

Shiga toxigenic Escherichia coli (STEC)

Strain number 63 912 431 1477 1861 1987 1988 1989 5722 5702 3883 5709 3128 3244 5705 5701 1524 2409 2725 2871 3081 4700 4718 5354 5570 2799 2873 3048 5264 5981 1043 5710 5708 5699 3861

Serotype

078:H16 OX3:H11 O101 O149:H10 O119 O111:H2 O26:NM O55:NM O55:H6 O103:H2 O111:NM O111:NM O113 O119:H16 O121:H19 O126:H27 O132 O157:H7 O157:H7 O157:H7 O157:H7 O157:H7 O157:H7 O157:H7 O157:H7 O157:NM O157:NM O157:NM O157:NM O157:NM O22 O26:H11 O45:H2 O91:H21 OX3:H2

Hafnia alvei c Citrobacter freundii Enterobacter cloacae Klebsiella pneumoniae Proteus vulgaris Pseudomonas aeruginosa S. cholerae-suis S. dublin S. enteritidis Salmonella typhimurium Staphylococcus aureus Yersinia enterocolitica

Genotypea

staP elt estB staP eaeA eaeA eaeA eaeA eaeA stx1 stx1, eaeA stx1, stx2, eaeA stx1, stx2 stx2 stx1, stx2 stx1 eaeA stx1, eaeA stx1, stx2, eaeA stx2, eaeA stx1, stx2, eaeA stx1, stx2, eaeA stx1, stx2, eaeA stx1, eaeA stx2, eaeA stx2, eaeA stx1, stx2, eaeA stx1, stx2, eaeA stx1, stx2, eaeA eaeA stx1, stx2 stx1, stx2, eaeA stx2 stx2 stx1, stx2 eaeA

Origin/ Disease Dog Pig/normal Pig/diarrhoea Pig/diarrhoea Human/diarrhoea Human Human Human Human Human/HUS Calf/normal Human/HUS Calf/normal Calf/normal Human/HUS Human/HUS Rabbit/diarrhoea Food HUS Calf/normal Calf/normal Calf Human/HUS Calf/normal Human/diarrhoea Lab Calf/normal Calf/normal Calf/normal Calf Calf/diarrhoea Human/HUS Human/HUS Human/HUS Calf/normal Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown

Amplificationb − − − − − − − + + + + + + + + + − + + + + + + + + + + + + + + + + + + − − − − − − − − − − − −

a

The presence of these genes was determined by colony blot hybridization using DNA probes specific for these genes (unpubl. data). +or −indicates presence or absence, respectively, of amplified products that resulted in DRn greater than threshold in DRn. c Harley Moon, Veterinary Medical Research Institute, Iowa State University, Ames, IA, USA. b

PCR assays were also 104 cfu ml−1 (data not shown). Since DNA equivalent to 1/400 of the total sample was used in each PCR reaction, the actual number of cfu required to produce a positive signal were 50 for the multiplex PCR and 17–170 for the nonmultiplex PCR assays.

Sensitivity limits of fluorogenic PCR assays to detect EHEC O157:H7 in beef The infective dose of EHEC O157:H7 and probably of other EHEC serotypes is believed to be in the range of 10 to 100 cfu.41 However, as shown in Fig. 2, a

Fluorogenic PCR assays for STEC and EHEC

297

Table 3. Specificity of fluorogenic non-multiplex polymerase chain reaction assays to detect stx1 and stx2 of shiga toxin-producing Escherichia coli (STEC) and eaeA of EHEC O157:H7 STEC serotype tested

O22 O26:H11 O26:NM O45:H2 O55:H30 O55 O55:NM O55:H6 O91:H21 O103:H2 O111:H2 O111:NM O111:NM O113 O119:H16 O121:H19 O126:H27 O157:H7 O157:H7 O157:H7 O157:H7 O157:H7 O157:H7 O157:H7 O157:H7 O157:NM O157:NM O157:NM O157:NM O157:NM OX3:H2

Genotype

stx1, stx2 stx1, stx2, eaeA eaeA stx2 eaeA eaeA eaeA stx2 stx1 eaeA stx1, eaeA stx1, stx2, eaeA stx1, stx2 staP, stx2 stx1, stx2 stx1 stx1, eaeA stx1, stx2, eaeA stx2, eaeA stx1, stx2, eaeA stx1, stx2, eaeA stx1, stx2, eaeA stx1, eaeA stx2, eaeA stx2, eaeA stx1, stx2, eaeA stx1, stx2, eaeA eaeA stx1, stx2, eaeA stx1, stx2

minimum of 104 cfu ml−1 of EHEC O157:H7 strain 2409 culture were required to generate a positive amplification signal in our fluorogenic PCR assays. In order to improve the sensitivity of the fluorogenic PCR assay to detect very low numbers of EHEC O157: H7 (less than 100 cfu) in a food sample, 1-g portions of ground beef (containing 103 cfu of endogenous bacteria per gram of beef; these bacterial counts were determined as described in Materials and Methods) were spiked with 3×10−3 to 104 cfu of EHEC O157: H7 strain 2409 and was cultured for 0, 2, 4, 6 or 18 h before isolation of DNA. The isolated DNA was tested in fluorogenic PCR assays to detect EHEC O157:H7. Table 4 shows the type of amplification signal generated for each sample when tested in the multiplex fluorogenic PCR assay designed for detection of any STEC or in the non-multiplex fluorogenic PCR assay designed for detection of EHEC O157:H7-specific eaeA gene. The sensitivity limits of

Detection of

O157:H7-specific eaeA

stx1

stx2

− − − − − + + + − − − − − − − − − + + + + + + + + + + + + + −

+ + − − − − − − − + − + + + − + + + + − + + + + − − + + − + +

+ + − + − − − − + − − − + + + + − − + + + + + − + + + + − + +

these assays were 3 cfu g−1 of beef that was cultured in TSB or mTSB for 6–18 h.

Effect of non-target bacteria on the detection of EHEC O157:H7 in beef The fluorogenic PCR assays were able to detect 3 cfu of EHEC O157:H7 in ground beef, containing very low numbers of non-target bacteria, that was cultured in broth for 6 h (Table 4). Next, we determined the sensitivity of the fluorogenic PCR assays to detect very low numbers of EHEC O157:H7 in food samples containing very high numbers of non-target bacteria. Bacterial DNA was isolated from ground beef, spiked with different ratios of target (O157:H7) to non-target (E. coli strain 63) strains, after 6 h of growth in TSB and was tested in the non-multiplex fluorogenic PCR

V. K. Sharma et al.

298 80 70 60

∆Rn

50 40 30 20 10 0 0

0

1

2

3

4

5

6

7

2 × 10 2 × 10 2 × 10 2 × 10 2 × 10 2 × 10 2 × 10 2 × 10 –1

cfu ml

Fig. 2. Sensitivity limits of the multiplex fluorogenic polymerase chain reaction (PCR) assay. The DNA from 10-fold dilutions of an overnight culture of EHEC O157:H7 strain 2409 was used in the multiplex fluorogenic PCR assay. The fluorescence measurement and data analysis were performed as described under Materials and Methods. The fluorescent values (DRn) were plotted against cfu per ml. The threshold DRn is indicated by the dashed line. The dilutions with DRn values above the threshold value were scored positive indicating the amplification of targeted genes. Errors bar indicate the standard deviation of the mean (n=3).

assay designed for detection of EHEC O157:H7-specific eaeA gene. The sensitivity limits of this assay were 5·8, 58 and 580 cfu of EHEC O157:H7 in beef that contained 106, 107 and 108 cfu of the non-target strain, respectively.

of faecal coliforms and other bacteria. We evaluated the ability of fluorogenic PCR assays to detect very low numbers of EHEC O157:H7 in bovine faeces containing different ratios of target to non-target bacteria. Faeces, spiked with different ratios of target (EHEC O157:H7 strain 2409) to non-target (E. coli strain 63) strains, were cultured for 6 h and DNA from these samples was tested in the non-multiplex fluorogenic PCR assay designed for detection of EHEC O157:H7-specific eaeA gene. The sensitivity limits of this assay were 1·2, 12 and 1200 cfu of EHEC O157: H7 per gram of faeces that contained 0, 108 and

Sensitivity limits of fluorogenic PCR assay to detect EHEC O157:H7 in bovine faeces Detection of low numbers of pathogens in faeces is a challenge because of the presence of large number

Table 4. Sensitivity limits of fluorogenic polymerase chain reaction assays to detect EHEC O157:H7 in beefa Detection of O157:H7 in beef-broth mixtures that were cultured for O157:H7 cfu ml−1 in beef-broth mixture

0 3×10−3 3×10−2 3×10−1 3×100 3×101 3×102 3×103 3×104 a

0h

2h

4h

6h

18 h

TSB

m-TSB

TSB

m-TSB

TSB

m-TSB

TSB

m-TSB

TSB

m-TSB

− − − − − − − − −

− − − − − − − − +

− − − − − − − + +

− − − − − − − + +

− − − − − − + + +

− − − − − − + + +

− − − − + + + + +

− − − − + + + + +

− − − − + + + + +

− − − − + + + + +

Endogenous bacterial count in beef before inoculation with O157:H7 was 103 cfu ml−1.

Fluorogenic PCR assays for STEC and EHEC 80 70

299

6

1 × 10 E. coli strain 63 1 × 107 E.coli strain 63 1 × 108 E.coli strain 63

60

∆Rn

50 40 30 20 10 0 0

–1

5.8 × 10

0

5.8 × 10

1

5.8 × 10

2

5.8 × 10

3

5.8 × 10

5.8 × 10

4

5

5.8 × 10

EHEC O157:H7 cfu g–1 beef

Fig. 3. Sensitivity limits of the fluorogenic polymerase chain reaction (PCR) assay for detecting EHEC O157:H7 in beef containing different levels of non-target bacteria. One-gram portions of beef were spiked with 106, 107, or 108 cfu of Escherichia coli strain 63 (non-target strain) and that was followed by the addition of 5·8×10−1 to 5·8×105 cfu of EHEC O157:H7 strain 2409 (target strain). The DNA isolated from these samples was used in the non-multiplex fluorogenic PCR assay designed for detecting enterohaemorrhagic E. coli (EHEC) O157:H7-specific eaeA. The fluorescence measurement and data analysis were performed as described under Materials and Methods. The threshold DRn is indicated by the dashed line. The dilutions with DRn values above the threshold value were scored positive indicating the amplification of targeted genes. Errors bar indicate the standard deviation of the mean (n=3).

109 cfu, respectively, of the non-target strain plus 105 cfu of faecal bacteria (Fig. 4).

DISCUSSION The fluorogenic PCR assays reported in this study were optimized to provide excellent specificity for semi-automated detection of EHEC O157:H7 and other STEC and for determining virulence gene profiles of these strains. Although the specificity of these assays is comparable to numerous other gel-based PCR assays facilitating the detection of STEC25–27,42 and EHEC O157:H7,28,29,34,43,44 the assays described in this report offer two unique advantages as compared to other PCR-based assays. First, the fluorogenic multiplex PCR assay can be used as a screen to test large number of samples as it allows rapid detection of STEC carrying any individual or any combination of stx1, stx2 and eaeA genes. Although this method does not identify the individual gene or gene combination harboured by the detected STEC (unless samples are analysed on an agarose gel), this method provides immediate warning as to the number of samples that may contain STEC. Second, the three individual non-multiplex PCR reactions can be used when identity of Shiga toxin genes harboured by a STEC needs to be determined, and when it is essential

to characterize STEC as EHEC O157:H7 and nonO157:H7 STEC. Several parameters were optimized to enable the fluorogenic PCR assays to detect EHEC O157:H7 and other STEC with specificity approaching 100%. First, we designed novel sets of primers that generated short PCR products (less than 200 bp as compared to primer pairs reported and evaluated previously for PCR amplification of stx1, stx2 and eaeA to detect STEC25,26,42 and EHEC O157:H728,29,34,43,44 in gel-based PCR assays) so that PCR cycling could be completed in a short time, enzyme and other reaction components would not become limiting factors and poor template integrity in crude DNA preparations would not compromise the yield of amplified products. The ability of these primers to amplify predicted size fragments of targeted genes was demonstrated both in multiplex and in non-multiplex PCR assays. Second, we designed fluorogenic probes that allowed specific detection of EHEC O157:H7 and other STEC. The fluorogenic probes specific for stx1 and stx2 detected all STEC (except STEC that harbour stx2e32 were not detected by the stx2 probe; data not shown) harbouring these genes and the fluorogenic probe specific for EHEC O157:H7 eaeA gene detected EHEC O157:H7 and EHEC O157:NM but not other STEC. The cross-reactivity of the EHEC O157:H7-specific eaeA probe with the eaeA gene of EHEC O157:NM

V. K. Sharma et al.

300 80

1 × 105 faecal bacteria (fb) 8 fb + 1 × 10 E.coli strain 63 9 fb + 1 × 10 E.coli strain 63

70 60

∆Rn

50 40 30 20 10 0 0

0

1

2

3

4

5

6

7

1.2 × 10 1.2 × 10 1.2 × 10 1.2 × 10 1.2 × 10 1.2 × 10 1.2 × 10 1.2 × 10 –1

EHEC O157:H7 cfu g

faeces

Fig. 4. Sensitivity limits of the fluorogenic polymerase chain reaction (PCR) assay for detecting EHEC O157:H7 in faeces containing different levels of non-target bacteria. One-gram portions of faeces were first spiked with 0, 108 and 109 cfu of Escherichia coli strain 63 (non-target strain) followed by the addition of 1·2×100 to 1·2×107 cfu of EHEC O157:H7 strain 2409 (target strain). The DNA isolated from these samples was used in the non-multiplex fluorogenic PCR assay designed to detect enterohaemorrhagic E. coli (EHEC) O157:H7-specific eaeA. The threshold DRn is indicated by the dashed line. The dilutions with DRn values above the threshold value were scored positive indicating the amplification of targeted genes. Errors bar indicate the standard deviation of the mean (n=3).

has been observed in other eaeA-based PCR assays39,40 and it is considered advantageous because EHEC O157:NM has been increasingly isolated from patients with HUS.41 The identification of O157:NM by O157:H7-specific eaeA probe is plausible in light of the recent studies showing the presence of the gene fliC encoding the H7 flagellar antigen in O157: NM isolates, and suggesting that many O157:NM isolates are the non-motile variants of O157:H7.45 The EHEC O157:H7-specific eaeA probe also crossreacted with EPEC O55, O55:H6 and O55:NM. Similar cross-reactivity of the EHEC O157:H7-specific eaeA probe with the eaeA gene of EPEC O55:H7 has been observed in other eaeA-based PCR assays.34,39 This cross-reactivity is due to the presence of virtually identical eaeA genes in these strains,38 and that precludes the design of primers and probes to successfully differentiate O157:H7 from EPEC, especially O55:H7 and other motile and non-motile members of the serogroup O55. However, EPEC are not detected by stx1- and stx2-based PCR assays because these strains lack genes encoding Shiga toxins.46 Moreover, O55:H7 lacks O157:H7-specific rfbE, responsible for O157-specific O-antigen biosynthesis.47 The gene rfbE has been used as an additional target gene in PCR assays to distinguish O157:H7 from O55:H7.45 Thus, new primers and fluorogenic probes can easily be incorporated into our fluorogenic PCR assays to distinguish Shiga toxin-negative EPEC belonging to the serogroup O55 from EHEC O157:H7.

The fluorogenic PCR assays described in this study were highly sensitive in detecting EHEC O157:H7 in beef and faecal cultures containing very high numbers of non-target bacteria. The detection limit of less than 10 cfu for EHEC O157:H7 in beef cultures is equivalent to the detection limits of a few recently reported fluorogenic PCR assays for detecting STEC in beef.26,48 Unlike our assays, previously described assays were unable to distinguish EHEC O157:H7 from other STEC and did not provide specific identification of Shiga toxin genes. A 5′ nuclease assay employing a fluorogenic EHEC O157:H7-specific eaeA probe for the detection of EHEC O157:H7 and EHEC O157:NM was able to detect 100 cfu after secondary culture of target cells that were captured using EHEC O157:H7-specific immunomagnetic beads from primary beef cultures.39 Another advantage of the fluorogenic PCR assays described in this study is that they provide automated and specific detection of amplified products thereby eliminating the need for agarose gel analysis of amplified samples. Coupling of this automated PCR amplification detection system with automated procedures for DNA extraction and liquid transfer will make these methods suitable for complete automation and high-throughput screening of food and faecal samples for rapid detection of EHEC O157:H7 and other STEC. These methods will be useful in helping the meat industry satisfy federal regulations, such as Hazard Analysis and Critical Control Points

Fluorogenic PCR assays for STEC and EHEC

(HACCP),49 concerning the presence of harmful bacteria in foods. These methods will help veterinary and clinical microbiologists to detect EHEC O157: H7 and STEC in samples of interest and to monitor the shedding of these pathogenic strains in cattle.

ACKNOWLEDGEMENTS The authors thank Drs Steve Carlson and Jay Ellingson for critical review of this manuscript, Robert Morgan for his technical assistance and Sandy Johnson for her help in preparation of this manuscript.

13.

14.

15.

1999 US Government 16.

REFERENCES 1. Karmali, M. A., Petric, M., Lim, C., Fleming, P. C., Arbus, G. S. & Lior, H. (1985). The association between idiopathic hemolytic uremic syndrome and infection by verotoxin-producing Escherichia coli. Journal of Infectious Diseases 151, 775–81. 2. Riley, L. W., Remis, R. S., Helgerson, S. D. et al. (1983). Hemorrhagic colitis associated with a rare Escherichia coli serotype. New England Journal of Medicine 308, 681–5. 3. Karmali, M. A. (1989). Infection by verocytotoxinproducing Escherichia coli. Clinical Microbiology Reviews 2, 15–38. 4. Centers for Disease Control (1993). Update: multistate outbreak of Escherichia coli O157:H7 infections from hamburgers in western United States, 1992–1993. Morbidity and Mortality Weekly Report 42, 258–63. 5. Swinbanks, D. (1996). Outbreak of EHEC infection in Japan renews concerns. Nature 382, 290. 6. Bradbury, J. (1997). Report on Scottish EHEC outbreak released. Lancet 349, 1073. 7. Konowalchuk, J., Spiers, J. L. & Starvic, S. (1977). Vero response to a cytotoxin of Escherichia coli. Infection and Immunity 18, 775–9. 8. Strockbine, N. A., Marques, L. R. M., Newland, J. W., Smith, H. W., Holmes, R. K. & O’Brien, A. D. (1986). Two toxin-converting phages from Escherichia coli O157:H7 strain 933 encode antigenically distinct toxins with similar biologic activities. Infection and Immunity 53, 135–40. 9. O’Brien, A. D., Tesh, V. L., Donohue-Rolfe, A. et al. (1992). Shiga toxin: biochemistry, genetics, mode of action, and role in pathogenesis. Current Topics in Microbiology and Immunology 180, 65–94. 10. Donnenberg, M. S., Tzipori, S., McKee, M. L., O’Brien, A. D., Alroy, J. & Kaper, J. B. (1993). The role of eae gene of enterohemorrhagic Escherichia coli in intimate attachment in vitro and in a porcine model. Journal of Clinical Investigation 92, 1418–24. 11. Tzipori, S., Gunzer, F. & Donnenberg, M. S. et al. (1995). The role of eaeA gene in diarrhea and neurological complications in a gnotobiotic piglet model of enterohemorrhagic Escherichia coli infection. Infection and Immunity 63, 3621–7. 12. Obrig, T. G. (1992). Pathogenesis of Shiga toxin (Verotoxin)-induced endothelial cell injury. In Hemolytic

17.

18.

19.

20.

21.

22.

23.

24.

25.

301

Uremic Syndrome and Thrmobocytpenic Purpura (Kaplan, B. S., Trompeter, R. S. & Moake, J. L., eds) pp. 405–19. New York: Marcel Dekker, Inc. Moon, H. W., Whipp, S. C., Argenzio, R. A., Levine, M. M. & Giannella, R. A. (1983). Attaching and effacing activities of rabbit and human enteropathogenic Escherichia coli in pig and rabbit intestines. Infection and Immunity 41, 1340–51. Wieler, L. H., Schwanitz, A., Vieler, E. et al. (1998). Virulence properties of Shiga toxin-producing Escherichia coli (STEC) strains of serogroup O118, a major group of STEC pathogens in calves. Journal of Clinical Microbiology 36, 1604–7. Dean-Nystrom, E. A., Bosworth, B. T., Moon, H. W. & O’Brien, A. D. (1998). Escherichia coli O157:H7 requires intimin for enteropathogenicity in calves. Infection and Immunity 66, 4560–3. Hancock, D. D., Besser, T. E., Kinsel, M. K., Tarr, P. I., Rice, D. H. & Paros, M. G. (1994). The prevalence of Escherichia coli O157:H7 in dairy and beef cattle in Washington State. Epidemiology and Infection 113, 199–207. Smith, H. R. & Scotland, S. M. (1993). Isolation and identification methods for Escherichia coli O157:H7 and other Vero cytotoxin producing strains. Journal of Clinical Pathology 46, 10–17. Acheson, D. W. K., DeBreucker, S., Donohue-Rolfe, A., Kozak, K., Yi, A. & Keusch, G. T. (1994). Development of a clinically useful diagnostic enzyme immunoassays for enterohemorrhagic Escherichia coli infection. In Recent Advances in Verocytotoxin Producing Escherichia coli Infections (Karmali, M. A. & Goglio, A. G., eds) pp. 109–12 Amsterdam, The Netherlands: Elsevier Science B. V. Gannon, V. P. J., Teerling, C., Masri, S. & Gyles, C. L. (1990). Molecular cloning and nucleotide sequence of another variant of the Escherichia coli Shiga-like toxin family. Journal of General Microbiology 136, 1125–35. Ito, H., Nishibuchi, M. & Takeda, Y. (1991). Analysis of the antigenic differences between Verotoxin 2 (VT2) and VT2 (VT2vh) of Verotoxin-producing Escherichia coli by site directed mutagenesis. FEMS Microbiology Letters 79, 27–30. March, S. B. & Ratnam, S. (1986). Sorbitol-MacConkey medium for detection of Escherichia coli O157:H7 associated with hemorrhagic colitis. Journal of Clinical Microbiology 23, 869–72. Ojeda, A., Prado, V. & Martinez, J. et al. (1995). Sorbitol-negative phenotype among enterohemorrhagic Escherichia coli strains of different serotypes and from different sources. Journal of Clinical Microbiology 33, 2199–201. Sowers, E. G., Wells, J. G. & Strockbine, N. A. (1996). Evaluation of commercial latex reagents for identification of O157 and H7 antigens of Escherichia coli. Journal of Clinical Microbiology 34, 1286–9. Samadpour, M., Liston, J., Ongerth, J. E. & Tarr, P. I. (1990). Evaluation of DNA probes for detection of Shiga-like toxin producing Escherichia coli in foods and calf fecal samples. Applied and Environmental Microbiology 56, 1212–15. Karch, H. & Meyer, T. (1989). Single primer pair for amplifying segments of distinct Shiga-like toxin genes by polymerase chain reaction. Journal of Clinical Microbiology 27, 2751–7.

302

V. K. Sharma et al.

26. Witham, P. K., Yamashiro, C. T., Livak, K. J. & Batt, C. A. (1996). A PCR-based assay for the detection of Escherichia coli Shiga-like toxin genes in ground beef. Applied and Environmental Microbiology 62, 1347– 53. 27. Ramotar, K., Waldhart, B., Church, D., Szumski, R. & Louie, T. J. (1995). Direct detection of verotoxinproducing Escherichia coli in stool samples by PCR. Journal of Clinical Microbiology 33, 519–24. 28. Gannon, V. P. J., Rashed, M., King, R. K. & Thomas, E. J. G. (1993). Detection and characterization of the eae gene of Shiga-like toxin-producing Escherichia coli by using polymerase chain reaction. Journal of Clinical Microbiology 31, 1268–74. 29. Louie, M., De Azavedo, J., Clarke, R. et al. (1994). Sequence heterogeneity of the eae gene and detection of verotoxin-producing Escherichia coli using serotype-specific primers. Epidemiology and Infection 112, 449–61. 30. Matsuura, M., Yamashiro, C., Flood, S. & PaszkoKolva, C. (1997). Detection of Salmonella in food using a fluorogenic 5′ nuclease assay. American Environmental Laboratory March, 24–5. 31. Bassler, H. A., Flood, S. J. A., Livak, K. J., Marmaro, J., Knorr, R. & Batt, C. A. (1995). Use of a fluorogenic probe in a PCR-based assay for the detection of Listeria monocytogenes. Applied and Environmental Microbiology 61, 3724–8. 32. Marques, L. R. M., Cryz, J. S. M. & O’Brien, A. D. (1987). Escherichia coli strains isolated from pigs with edema disease produce a variant of Shiga-like toxin II. FEMS Microbiology Letters 44, 33–8. 33. Doyle, M. P. & Schoeni, J. L. (1987). Isolation of Escherichia coli O157:H7 from retail fresh meats and poultry. Applied and Environmental Microbiology 53, 2394–6. 34. Fratamico, P. M., Sackitey, S. K., Weidmann, M. & Deng, M. Y. (1995). Detection of Escherichia coli O157:H7 by multiplex PCR. Journal of Clinical Microbiology 33, 2188–91. 35. Calderwood, S. B., Auclair, F., Donohue-Rolfe, A., Keusch, G. T. & Mekalanos, J. J. (1987). Nucleotide sequence of the Shiga-like toxin genes of Escherichia coli. Proceedings of National Academy of Sciences USA 84, 4364–8. 36. Jackson, M. P., Neill, R. J., O’Brien, A. D., Holmes, R. K. & Newland, J. W. (1987). Nucleotide sequence analysis and comparison of the structural genes for Shiga-like toxin I and Shiga-like toxin II encoded by bacteriophages from Escherichia coli. FEMS Microbiology Letters 44, 109–14. 37. Beebakhee, G., Louie, M., De Azavedo, J. & Brunton, J. (1992). Cloning and nucleotide sequence of the eae gene homologue from enterohemorrhagic Escherichia coli serotype O157:H7. FEMS Microbiology Letters 91, 63–8.

38. McGraw, E. A., Li, J., Selander, R. K. & Whittam, T. S. (1999). Molecular evolution and mosaic structure of alpha, beta, and gamma intimins of pathogenic Escherichia coli. Molecular Biology and Evolution 16, 12–22. 39. Oberst, R. D., Hays, M. P., Bohra, L. K. et al. (1998). PCR-based amplification and presumptive detection of Escherichia coli O157:H7 with an internal fluorogenic probe and the 5′ nuclease (TaqMan) assay. Applied and Environmental Microbiology 64, 3389–96. 40. Meng, J., Zhao, S., Doyle, M. P., Mitchell, S. E. & Kresovich, S. (1997). A multiplex PCR for identifying Shiga-like toxin-producing Escherichia coli O157:H7. Letters in Applied Microbiology 24, 172–6. 41. Willshaw, G. A., Thirlwell, J., Jones, A. P., Parry, S., Salmon, R. L. & Hickey, M. (1994). Vero cytotoxinproducing Escherichia coli O157:H7 in beef burgers linked to an outbreak of diarrhea, hemorrhagic colitis, and hemolytic uremic syndrome in Britain. Letters in Applied Microbiology 19, 304–7. 42. Read, S. C., Clarke, R. C., Martin, A., De Grandis, S. A., Hii, J., McEwen, S. & Gyles, C. L. (1992). Polymerase chain reaction for detection of verotoxigenic Escherichia coli isolated from animal and food sources. Molecular and Cellular Probes 6, 153–61. 43. Cebula, T. A., Payne, W. L. & Feng, P. (1995). Simultaneous identification of Escherichia coli serotype O157:H7 and their Shiga-like toxin type by mismatch amplification mutation assay-multiplex PCR. Journal of Clinical Microbiology 33, 248–50. 44. Paton, A. W. & Paton, J. C. (1998). Detection and characterization of Shiga toxigenic Escherichia coli by using multiplex PCR assays for stx1, stx2, eaeA, enterohemorrhagic E. coli hlyA, rfb0111, and rfb0157. Journal of Clinical Microbiology 36, 598–602. 45. Gannon, V. P. J., D’Souza, S., Graham, T., King, R. K., Rahn, K. & Read, S. (1997). Use of flagellar H7 gene as a target in multiplex PCR assays and improved specificity in identification of enterohemorrhagic Escherichia coli. Journal of Clinical Microbiology 35, 656–62. 46. Nataro, J. P. & Kaper, J. B. (1998). Diarrheagenic Escherichia coli. Clinical Microbiology Reviews 11, 142–201. 47. Bilge, S. S., Vary, Jr., J. C., Dowell, S. F. & Tarr, P. I. (1996). Role of the Escherichia coli O157:H7 O side chain in adherence and analysis of an rfb locus. Infection and Immunity 64, 4795–801. 48. Chen, S., Xu, R., Yee, A. et al. (1998). An automated fluorogenic PCR method for detection of Shiga toxinproducing Escherichia coli in foods. Applied Environmental Microbiology 64, 4210–16. 49. Department of Agriculture, Food Safety and Inspection Service (1996). Pathogen Reduction: Hazard Analysis and Critical Control Points (HACCP) Systems; Final Rule. Federal Register 61, 38,805–55.