JOURNAL OF CLINICAL MICROBIOLOGY, Nov. 2008, p. 3569–3575 0095-1137/08/$08.00⫹0 doi:10.1128/JCM.01095-08 Copyright © 2008, American Society for Microbiology. All Rights Reserved.
Vol. 46, No. 11
Phenotypic and Genotypic Characterization of Verotoxin-Producing Escherichia coli O103:H2 Isolates from Cattle and Humans䌤 Musafiri Karama,1 Roger P. Johnson,2 Robert Holtslander,2 and Carlton L. Gyles1* Department of Pathobiology, Ontario Veterinary College, University of Guelph, Guelph, Ontario N1G 2W1,1 and Laboratory for Foodborne Zoonoses, Public Health Agency of Canada, Guelph, Ontario N1G 3W4,2 Canada Received 9 June 2008/Returned for modification 2 August 2008/Accepted 21 August 2008
Characterization of important non-O157 verotoxin-producing Escherichia coli (VTEC) has lagged considerably behind that of O157:H7 strains. This study characterized 91 VTEC O103:H2 strains from bovine and human sources and of North American and European origins by virulence or putative virulence genes, pulsed-field gel electrophoresis (PFGE) patterns, plasmid profiles, antimicrobial resistance, and colicin production. All strains were positive for vt1 and eae-; 97% were positive for ehxA; and all were negative for hlyA. Two strains carried vt2. There were 66 PFGE patterns grouped in six clusters, and there were 25 different plasmid profiles. Plasmid-encoded katP and etp genes were significantly more frequent in European than in North American human strains. The distribution of selected phenotypes was as follows: enterohemorrhagic E. coli (EHEC) hemolysin, 95%; colicin production, 38%; antimicrobial resistance, 58%. All the strains were negative for the alpha-hemolytic phenotype. In conclusion, the VTEC O103:H2 strains were diverse, as shown by PFGE, plasmid profiles, virulence markers, and antimicrobial resistance patterns, and all strains showed an EHEC hemolytic phenotype instead of the alpha-hemolytic phenotype that has been shown previously. Verotoxin-producing Escherichia coli (VTEC) strains, also termed Shiga toxin-producing E. coli strains, are important food-borne pathogens that have been associated with outbreaks of diarrhea and hemorrhagic colitis, which may lead to life-threatening complications such as the hemolytic-uremic syndrome (HUS) in humans (27). More than 500 VTEC serotypes of animal and human origins have been isolated worldwide (www.microbionet.com.au/VTECtable.htm; www.lugo .usc.es/ecoli). These serotypes differ in their frequency of association with human disease and the severity of disease, suggesting differences in virulence characteristics (20). Most of the human illnesses reported have been attributed to VTEC O157:H7 (19). The large body of research on the epidemiology and virulence of VTEC has been concentrated on VTEC O157:H7, and only a few studies have examined non-O157 VTEC collections. However, the importance of non-O157 VTEC in human disease has been increasingly recognized (19), and data from surveillance studies across the United States show that nonO157 VTEC strains account in some instances for 50% or more of the VTEC strains detected in fecal samples from human patients (11, 16, 23). According to Enter-net, an international surveillance network that tracks human cases of enteric infections in 35 countries, there was a 60.5% increase in the frequency of non-O157 VTEC infections between 2000 and 2005, compared to a 13% increase in O157 VTEC disease in humans during the same period (1). A non-O157 VTEC serotype that has recently received particular attention and is considered an emerging food-borne
pathogen in continental Europe and North America is VTEC O103:H2 (3, 15, 35, 44). The first mention of VTEC O103 came in 1988, when members of this serogroup were associated with diarrhea in calves (7) and humans (32). In 1993, Mariani-Kurkdjian et al. (24) isolated vt1-positive E. coli O103:H2 strains from fecal samples of children suffering from HUS in France. In the following years, VTEC O103:H2 strains were isolated from healthy and diseased humans, cattle, sheep, goats, and various foods worldwide (2, 8, 13, 14, 18, 41, 43). VTEC O103:H2 strains appear to differ from other VTEC strains with regard to some of their phenotypic and genotypic characteristics (35, 41). For example, Schmidt et al. (41) observed that VTEC O103:H2 presented an alpha-hemolytic phenotype instead of the enterohemorrhagic E. coli (EHEC) hemolytic phenotype commonly expressed by VTEC strains. However, this conclusion was based on a small collection of VTEC O103:H2 isolates. Other studies (3, 35) observed that VTEC O103:H2 strains lacked the plasmid marker espP, and most of the strains lacked the plasmid-encoded katP gene. However, most of the observations concerning VTEC O103:H2 in the studies mentioned above have been based on strain collections from humans, and mainly from Germany (3, 35, 41). The objective of this study was to characterize a collection of 91 VTEC O103:H2 strains isolated from cattle and humans in North America and Europe from 1991 to 2002 in order to gain insight into the relationships among these strains of different sources and geographic origins. We used molecular subtyping methods, including pulsed-field gel electrophoresis (PFGE) and plasmid profiling, and searched for recognized chromosomal virulence genes and plasmid-encoded virulence markers, the presence of hemolysins, the production of colicins, and antimicrobial resistance in order to investigate relationships among the strains, with emphasis on human versus bovine
* Corresponding author. Mailing address: Department of Pathobiology, Ontario Veterinary College, University of Guelph, Guelph, Ontario, N1G 2W1, Canada. Phone: (519) 824-4120, ext. 54657. Fax: (519) 824-5930. E-mail:
[email protected]. 䌤 Published ahead of print on 3 September 2008. 3569
3570
KARAMA ET AL.
strains and on North American versus European human strains. MATERIALS AND METHODS VTEC O103:H2 strains and culture conditions. Ninety-one VTEC O103:H2 strains from frozen (⫺70C) culture collections maintained by the University of Guelph and the Laboratory for Food-Borne Zoonoses in Guelph, Ontario, Canada, were characterized. The strains originated from three countries in Europe (Germany, Switzerland, and Belgium) and from North America (the United States and Canada). Forty-six strains were of human origin, and 45 originated from cattle. All cattle strains were from Canada; the 46 human strains consisted of 23 strains isolated in North America and 23 that originated in continental Europe. Cattle strains were recovered from individual healthy animals in geographically separate herds during routine prevalence studies, while human strains were from patients presenting with clinical disease in independent outbreaks. All strains were isolated between 1991 and 2002. Some clinical information was available for 5 German strains and 12 Belgian strains. Frozen cultures were resuscitated by overnight growth at 37C on LuriaBertani (LB) agar. Fresh cultures derived from the frozen stock were used each time for each of the procedures carried out in this study. PFGE. To determine the clonal relationships among the VTEC O103:H2 strains, DNA was extracted, digested with XbaI, and subjected to PFGE according to the CDC/PulseNet protocol (http://www.cdc.gov/pulsenet/protocols.htm). Salmonella enterica serotype Braenderup (strain H9812; American Type Culture Collection catalog no. BAA-664) DNA digested with XbaI was used as the DNA marker. To assess reproducibility, PFGE was carried out at least twice. The XbaI PFGE patterns were analyzed for similarity, and a dendrogram was generated by Bionumerics software, version 3.5 (Applied Maths, Sint Martens-Latem, Belgium) with Dice similarity indices (complete linkage; optimization, 1.5%; position tolerance, 1.5%) and the unweighted-pair group method with arithmetic means. Plasmid profiles. Plasmid DNA was prepared using the Qiagen (Mississauga, Ontario, Canada) plasmid minikit, separated on a 0.7% agarose gel, stained with ethidium bromide, and photographed under UV light. Two ladders (2 to 165 kb) were mixed and used as molecular mass standards: 1.5 l of the Supercoiled DNA ladder (Invitrogen Life Technologies, Burlington, Ontario, Canada) was added to 10 l Bac-Tracker Supercoiled DNA (Epicentre, Madison, WI). VTEC O157:H7 strain EDL 933 was used as a plasmid extraction control. Plasmid gel images were entered into Bionumerics software, version 3.5 (Applied Maths, Saint Martens-Latem, Belgium) for size estimation. Southern hybridization. To identify the pO157 equivalent in VTEC O103:H2 strains, plasmid DNAs from 11 strains representing a range of plasmid profiles that included large and/or medium-sized plasmids were transferred to a nylon membrane and probed for the ehxA gene, which is a marker for plasmid pO157 (39). These strains comprised three European human strains (HE4, HE19, and HE90), four North American bovine strains (BNA26, BNA30, BNA31, and BNA44), and four North American human strains (HNA43, HNA67, HNA78, and HNA84). The ehxA probe was generated by PCR by amplifying the ehxA gene of VTEC O157:H7 strain EDL 933 with primers hlyA1 and hlyA4 (39). The PCR product was extracted from the gel, purified (with a QIAquick gel extraction kit; Qiagen, Mississauga, Ontario, Canada), and labeled with digoxigenin (PCR DIG probe synthesis kit; Roche Applied Science, Laval, Quebec, Canada). Hybridization was done overnight at 42°C in DIG Easy Hyb hybridizing buffer (Roche Diagnostics GmbH, Mannheim, Germany). Washing was carried out under high-stringency conditions. The nylon membrane was submitted to two 15-min washes at 68°C in each of three solutions: 2⫻ SSC (1⫻ SSC is 0.15 M NaCl plus 0.015 M sodium citrate), 0.5⫻ SCC, and 0.1⫻ SCC. The DIG detection starter kit II (for chemiluminescent detection with CSPD) was used according to the manufacturer’s instructions (Roche Diagnostics GmbH, Mannheim, Germany) to detect the plasmid DNA bands with the ehxA gene. PCR for virulence genes and markers. DNA was extracted from all strains by the boiling method, and PCR cycling parameters for previously described primers were used to amplify genes that encode virulence factors and markers. A multiplex PCR was used for the detection of vt1, vt2, eae, and ehxA (34). Amplification reactions for the other genes—eae-ε (31), hlyA (4), katP (10), espP (9), and etp (38)—were conducted individually. The VTEC O157:H7 strain EDL 933 (E. coli O157:H7) was used as a positive control for all the genes except hlyA and eae-ε. Porcine O149:H10:F4 enterotoxigenic E. coli (ETEC) strain JG280 (Gyles Laboratory, University of Guelph, Guelph, Ontario, Canada) and E. coli O103:H2 strain UTI (43) were used as positive controls for hlyA and eae-ε, respectively. PCR was carried out in an Eppendorf Mastercycler (Eppendorf AG, Brinkman Instruments, Westbury, NY) by using 25-l reaction mixtures
J. CLIN. MICROBIOL. consisting of 2.5 l of DNA, 2.5 l of 10⫻ PCR buffer, 1.5 or 2 mM MgCl2, 200 M each deoxynucleoside triphosphate, 2 U of Taq DNA polymerase, and water. Detection of EHEC hemolysin and alpha-hemolysin. All VTEC O103:H2 strains were tested for expression of an EHEC hemolytic or alpha-hemolytic phenotype on blood agar containing unwashed sheep erythrocytes and on blood agar containing washed sheep erythrocytes (5). An overnight blood agar culture of VTEC O103:H2 was touched with an inoculating loop and used to make a 4-cm line on blood agar and on washed blood agar (5). The plates were incubated at 37°C and monitored for presence of a hemolytic zone after 3 and 16 h. A large zone of clear hemolysis on both types of agar after 3 h was interpreted as the alpha-hemolytic phenotype, while a narrow zone of incomplete hemolysis on washed blood agar after 16 h but not after 3 h and a lack of hemolysis on unwashed blood agar was interpreted as the EHEC hemolytic phenotype (5). E. coli O157:H7 strain EDL 933 and porcine ETEC strain JG280 were used as positive controls for the EHEC hemolysin and alpha-hemolysin, respectively. E. coli C600 was used as a negative control. Assay for colicin production. Production of colicin by VTEC O103 strains was detected as described by Pugsley and Oudega (36). Briefly, heavy inocula of fresh colonies of VTEC O103:H2 were transferred with a sterile toothpick onto approximately 1-cm-wide circles on a freshly made lawn of E. coli C600 indicator bacteria on a Trypticase soy agar plate. After overnight incubation at 37°C, the plates were examined for clear zones surrounding the areas of VTEC growth. Strains showing clear zones caused by an absence of growth of the indicator strain were scored as colicin positive. The EHEC O157:H7 strain EDL 933 was used as a positive control, and E. coli strain DH5␣ was used as a negative control. Antimicrobial susceptibility test. VTEC O103:H2 strains were tested to determine their patterns of resistance to 12 antimicrobials by the disk diffusion method as described by the Clinical and Laboratory Standards Institute (12). The panel of antimicrobials consisted of amikacin (30 g), amoxicillin-clavulanic acid (20 and 10 g, respectively), ampicillin (10 g), ceftiofur (30 g), cephalothin (30 g), chloramphenicol (30 g), enrofloxacin (5 g), gentamicin (10 g), spectinomycin (100 g), sulfisoxazole (250 g), tetracycline (30 g), and trimethoprimsulfamethoxazole (1.25 and 23.75 g, respectively). Antimicrobial disks (BBL Sensi Disk) were obtained from Becton Dickinson & Company, Sparks, MD. E. coli ATCC 25922 was used as the control strain. Isolates were classified as susceptible, resistant, or intermediate to each antimicrobial agent, and the intermediate readings were assigned to the resistant category. Statistical analysis. Fisher’s exact test was used to determine if there were statistically significant differences between the proportions of genes and the phenotypes detected in groups of the VTEC O103:H2 strains. Software produced by William Sears, Department of Population Medicine, Ontario Veterinary College, Guelph, Ontario, Canada, was used to perform the statistical calculations. P values of ⬍0.05 were considered statistically significant. Comparisons were made between North American cattle and human VTEC O103:H2 strains and between North American and European human VTEC O103 strains.
RESULTS PFGE profiles. All strains were typeable by PFGE (Fig. 1), and the PFGE patterns obtained in replicate tests were always the same. The 91 strains were classified into 66 subtypes, which showed a 64.4% Dice similarity index and were organized into six dissimilar clusters, designated A, B, C, D, E, and F (Fig. 1). Among the 66 subtypes, 49 were single-strain subtypes, 14 contained two strains each, 1 subtype contained four strains, and 2 subtypes contained five strains each. Cluster A contained only human VTEC O103:H2 strains from Europe. Clusters B, C, D, and E contained combinations of bovine and human strains isolated from both continents, and cluster F consisted of North American strains isolated from both cattle and humans. Three strains constituted single-strain subtypes that were not part of these six major clusters. Of particular interest were five strains from different sources and geographical origins in cluster B (BNA49, BNA53, HE82, HE87, and HE90) that shared an identical PFGE profile. Although these five strains had a common PFGE profile, only two (HE82 and HE90), which belonged to the same country, had identical plasmid profiles, while the other three differed by plasmid profiling.
VOL. 46, 2008
CHARACTERIZATION OF CATTLE AND HUMAN VTEC O103:H2
FIG. 1. Dendrogram showing relationships among VTEC O103:H2 strains based on PFGE patterns.
3571
3572
KARAMA ET AL.
J. CLIN. MICROBIOL. TABLE 1. Distribution of virulence-associated chromosomal and plasmid genes in 93 VTEC O103:H2 isolates from North America and Europe % (no.) of strains positive for the gene
FIG. 2. Southern blot of VTEC O103:H2 plasmids hybridized with a probe for the EHEC hemolysin gene (ehxA), located on the large virulence plasmid of E. coli O157:H7 (pO157). M, plasmid DNA markers. Sizes in kilobases are shown to the left. (A) Agarose gel of the plasmid DNA. (B) Results of the Southern hybridization. Negative results in lanes 3 and 8 were given by strains that lacked the ehxA gene by PCR.
Plasmid profiles. The 91 VTEC O103:H2 strains had 25 different plasmid profiles. Twelve of the profiles consisted exclusively of human strains isolated on both continents, six were specific for cattle strains, and seven consisted of both human and cattle strains from both continents. The 12 plasmid profiles that comprised exclusively human strains included 4 and 8 profiles containing European and North American human isolates, respectively. Three classes of plasmids were identified according to their approximate sizes: large (60 kb, 75 kb, 95 kb, or 120 kb), medium (55 kb or 40 kb), or small (⬍40 kb). Thirty-two (35.1%) strains harbored only a single large plasmid, while the remaining 59 (64.8%) carried multiple plasmids. A 75-kb plasmid alone or in combination with other plasmids was present in 80 (87.9%) strains. A 120-kb plasmid was carried by 23 (25.2%) strains, while a 95-kb plasmid alone or in combination with other plasmids was present in 34 (37.3%) strains. One or more small plasmids were present in 40 strains (43.9%). Three strains that were negative for ehxA also lacked a large plasmid. Southern hybridization with an ehxA probe revealed that the pO157 equivalent was either a 75-kb or a 95-kb plasmid (Fig. 2). Distribution of chromosomal virulence genes and plasmidencoded virulence markers. The distributions of vt1, vt2, eae, ehxA, hlyA, katP, espP, and etp among the 91 VTEC O103:H2 strains and the three categories of strains are presented in Table 1. All strains were vt1 and eae positive, and two strains carried vt2 as well as vt1. The intimin type carried by all 91 strains was eae-ε. The ehxA gene was present in 88 (96.7%) strains. All the strains were hlyA negative. The distribution of other plasmid-encoded genes was as follows: 29 (31.8%) strains had katP, 14 (15.3%) had espP, and 71 (78.2%) had etp. The frequency of the katP gene was significantly higher (P ⫽ 0.007) in North American human strains (30.4%) than in North American cattle strains (8.8%) but significantly lower (P ⫽ 0.007) than in European human strains (78.2%). The proportion of strains with etp was higher for North American cattle strains than for North American human strains (P ⫽ 0.001), and this gene was more prevalent in European than in North American human strains (P ⫽ 0.005). The frequencies of espP were not significantly different (P ⬎ 0.05) in strains of different geographic and/or host origins. Two main plasmid marker combinations were observed: (i) ehxA, katP, and etp
Gene
vt1 vt2 eae eae-ε ehxA hlyA katPa espP etpa
All strains (n ⫽ 91)
North American cattle strains (n ⫽ 45)
North American human strains (n ⫽ 23)
European human strains (n ⫽ 23)
100 2.1 (2) 100 100 96.7 (88) 0 31.8 (29) 15.3 (14) 78 (71)
100 0 100 100 100 0 8.8 (4) 13.3 (6) 86.6 (39)
100 0 100 100 86.9 (20) 0 30.4 (7) 26.08 (6) 47.8 (11)
100 8.6 (2) 100 100 (23) 100 0 78.2 (18) 8.6 (2) 91.3 (21)
a Statistically significant difference between North American cattle and human strains.
and (ii) ehxA and etp. These combinations were present in 20 (21.9%) and 48 (52.7%) isolates, respectively. Fifteen (71.4%) of 21 strains with the combination of ehxA, katP, and etp were European human strains, while 34 (70.8%) of the strains possessing the combination of ehxA and etp were North American cattle strains. No strains harbored both espP and etp together. Hemolysin and colicin phenotypes. Eighty-six strains (94.5%) showed the EHEC hemolysin phenotype, and none showed the alpha-hemolytic phenotype (Table 2). Colicin production was detected in 35 (38.4%) VTEC O103:H2 strains, of which 30 carried at least one small plasmid. All colicinogenic isolates produced small clear zones that extended 2 to 3 mm beyond the edge of VTEC growth. The colicinogenic phenotype was more prevalent in North American cattle strains (P ⫽ 0.01) than in North American human strains. There was no significant difference between the proportions of colicinogenic human strains from North America and Europe. Antimicrobial resistance. Resistance to at least one of the antimicrobials tested was detected in 53 (58.2%) of the VTEC O103:H2 strains (Table 2), with no significant differences between the frequencies in North American cattle strains (57.7%), North American human strains (47.8%), and European human strains (69.5%). The highest frequencies of resistance were observed for sulfisoxazole (34%) and tetracycline
TABLE 2. Distribution of hemolytic, colicinogenic, and antimicrobial resistance phenotypes among O103:H2 VTEC isolates from North America and Europe % (no.) of strains positive for the phenotype Phenotype
All strains (n ⫽ 91)
North American cattle strains (n ⫽ 45)
North American human strains (n ⫽ 23)
European human strains (n ⫽ 23)
EHEC hemolysin Alpha-hemolysin Colicina production AMRb
94.5 (86) 0 38.4 (35) 58.2 (53)
100 0 55.5 (25) 57.7 (26)
82.6 (19) 0 30.4 (7) 47.8 (11)
95.6 (22) 0 13 (3) 69.5 (16)
a Statistically significant difference between North American cattle and human strains. b AMR, resistance to at least one antimicrobial.
VOL. 46, 2008
CHARACTERIZATION OF CATTLE AND HUMAN VTEC O103:H2
(32.9%). Resistance was also recorded for ampicillin (21.5%), cephalothin (17.2%), trimethoprim-sulfamethoxazole (7.6%), and chloramphenicol (4.3%). Among the 91 VTEC O103:H2 strains, 14 (15.3%) strains were resistant to one antimicrobial, 29 (31.8%) were resistant to two antimicrobials, 3 (3.2%) were resistant to three antimicrobials, 1 (1%) was resistant to four antimicrobials, 4 (4.3%) were resistant to five antimicrobials, and 2 (3.2%) were resistant to six antimicrobials. Combined resistance to sulfisoxazole and tetracycline (48.8%) was observed in most North American cattle strains, while European and North American human strains were mostly resistant to cephalothin and/or ampicillin. DISCUSSION Although VTEC O103:H2 strains were consistent in their possession of the chromosomal virulence genes vt1 and eae and of the plasmid-encoded ehxA gene, they were very diverse, as indicated by their PFGE patterns, plasmid profiles, and plasmid-encoded genes. PFGE was highly discriminatory: it identified 66 different profiles among the 91 VTEC O103:H2 strains. The high diversity among the PFGE profiles was also evident by the low Dice similarity index of 64.4%. Higher diversity has been reported for VTEC O103 than for other VTEC serogroups, including VTEC O26 and O157 (35). The present study showed higher diversity among VTEC O103:H2 strains than the study by Prager et al. (35), in which only 19 PFGE subtypes were observed among 145 VTEC O103:H2/H⫺ strains isolated in Germany during the period 1997 to 2000. This is not surprising, given the differences in origins and time frames for the isolation of strains in the two collections. Beutin et al. (3) also demonstrated high diversity among 54 O103 strains, consisting of 44 VTEC O103 and 10 enteropathogenic E. coli O103 strains, that were examined by PFGE and multilocus sequence typing. However, a previous report by Schmidt et al. (41) recorded 80% similarity among randomly amplified polymorphic DNA patterns of a small collection of epidemiologically unrelated VTEC O103:H2 strains from Germany. High heterogeneity in PFGE patterns has been also observed in other VTEC serotypes and serogroups, including O26:H11, O157:H7/H⫺, O111, and O145 (6, 17, 22, 26, 42). VTEC O157:H7, the best-characterized VTEC serotype, has more than 25,000 PFGE profiles in the PulseNet database (17). Among the six dissimilar PFGE clusters identified in the present study, clusters A and F were the only ones in which strains were related by source or geographic origin. All the strains in cluster A originated from humans in Europe and might be strains that are widespread in the three European countries of origin. The North American cattle and human strains in cluster F were from the same Canadian province and may have shared a common ancestor in the recent past or, alternatively, may have had a common cattle reservoir linked by geographic proximity. The high number of single-strain PFGE profiles and the scarcity of shared PFGE subtypes among strains from the same species and geographic origins, as well as the results of other studies (3, 35), support the existence of independent populations of VTEC O103:H2 strains circulating among humans and/or cattle on both continents. A number of studies have shown that the high heterogeneity among VTEC O157 and non-O157 serotypes is determined by inser-
3573
tions and deletions of diverse prophage and prophage-related elements, which recombine with resident lysogenic phages or incoming phages and are scattered across the chromosome (28, 29). Concerning the five strains from different sources and geographical origins in cluster B that shared an identical PFGE profile, we believe that three of these strains are not clonal, as evidenced by their differing plasmid profiles. However, PFGE with a second enzyme may be a better way to ascertain conclusively the difference between these strains. As for the common PFGE and plasmid patterns exhibited by strains HE82 and HE90, which were isolated in the same country, we believe that these patterns may be representative of widespread or common PFGE and plasmid patterns in the country of origin of the two strains. Plasmid profiles and analysis also revealed extensive diversity among VTEC O103:H2 strains. There was a correlation between plasmid markers of VTEC O103:H2 strains according to their human or animal source and their geographic origin, although the strains in the collection showed high diversity in the number and size of plasmids. In contrast to VTEC O157: H7, in which pO157 is a 90-kb plasmid possessing the full complement of plasmid-encoded genes ehxA, katP, espP, and etp, the pO157 equivalent in VTEC O103:H2 strains was a smaller, 75-kb plasmid that instead encoded a combination of either ehxA, katP, and etp or ehxA and etp in most strains. Plasmids with this size and gene organization have been observed by Beutin et al. (3) and Prager et al. (35) in German VTEC O103:H2 strains. According to Prager et al. (35), in addition to the pO157-equivalent plasmid, some of the plasmids of VTEC O103:H2 with sizes ranging from 60 to 105 kb have been associated with multidrug resistance, while small plasmids have been associated with colicins or antibiotic resistance in O157:H7 VTEC strains (30, 33, 35, 37). A high proportion of VTEC O103:H2 strains that carried small plasmids were colicinogenic (30/40) and antibiotic resistant (28/40). Considerable variability in plasmid number and size and in gene composition and arrangement has also been observed in other VTEC serotypes, including O26 and O145 (42, 46), and has been attributed to plasmid loss or exchange among bacteria. The EHEC hemolytic phenotype was always associated with the presence of ehxA, except for two strains that were positive for ehxA but did not produce the EHEC hemolytic phenotype. Our results differ from those reported by Schmidt et al. (41), who observed that VTEC O103:H2 strains manifest an alphahemolysin phenotype rather than an EHEC hemolysin phenotype. Two other studies (3, 35) that characterized VTEC O103 collections did not report the presence or absence of the alphahemolytic phenotype in these strains. Both the plasmid-encoded EHEC hemolysin and the chromosomally encoded alpha-hemolysin are powerful RTX (repeats-in-toxin) toxins with similar pore-forming capacities (40). The role played by hemolysins in the pathogenesis of disease caused by VTEC has never been established, but most patients who develop HUS after a VTEC O157 infection produce serum antibodies to EHEC hemolysin, suggesting that this hemolysin is expressed during VTEC disease (39). Most studies of VTEC have found a high frequency of resistance to sulfonamides and tetracycline among various sero-
3574
KARAMA ET AL.
types or serogroups (21, 25, 45). In agreement with the studies mentioned above, VTEC O103:H2 strains had high frequencies of resistance to sulfisoxazole (34%) and tetracycline (32.9%). Typically, resistance to sulfisoxazole and tetracycline occurred in the same strains and was mainly observed in North American cattle strains, consistent with recent Canadian reports that resistance to sulfonamides and tetracyline is present at high frequencies in generic E. coli (www.phac-aspc.gc.ca /cipars-picra/2006_pr_e.html) and VTEC O157:H7 (45) isolates from slaughter cattle. In conclusion, this is the first time a large collection of VTEC O103:H2 strains from cattle and humans and of North American and European origins has been analyzed. The strains were homogeneous for the presence of stx1 and eae, and all except three strains carried ehxA. All hemolytic strains showed an EHEC hemolysin phenotype rather than the alpha-hemolysin phenotype, which has been observed previously for VTEC O103:H2 (41). Considerable genomic diversity was demonstrated by PFGE, plasmid profiles, virulence-related genes, and antimicrobial resistance. This diversity probably reflects both the adaptation of VTEC O103:H2 strains to various niches and their responses to different selective pressures. Although plasmid profiling was less discriminatory than PFGE, it proved to be an adequate method for differentiating VTEC O103:H2 strains by source and geographic origin. An implication of the findings is that PFGE can be expected to be effective in tracking outbreak EHEC strains of this serotype. PFGE could be combined with another rapid and simple technique such as plasmid profile determination to enhance the discrimination of strains.
J. CLIN. MICROBIOL.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
ACKNOWLEDGMENTS Funding for this research was provided by the Natural Sciences and Engineering Research Council of Canada (NSERC) and the Ontario Ministry of Agriculture, Food and Rural Affairs (OMAFRA). We thank D. Pierard, L. Beutin, S. Alesic, A. Burnens, W. Johnson, and P. Tarr, who kindly donated VTEC O103:H2 strains.
21.
22.
REFERENCES 1. Anonymous. 2005. European Commission Annual Report. Surveillance of enteric pathogens in Europe and beyond. 1786/2002/EC. International surveillance network for the enteric infections—Salmonella, VTEC O157 and Campylobacter. European Commission, Brussels, Belgium. 2. Bettelheim, K. A., M. A. Hornitzky, and S. P. Djordjevic. 2001. First bovine and ovine isolations of Shiga toxin-producing Escherichia coli O103:H2 in Australia. Aust. Vet. J. 79:289–290. 3. Beutin, L., S. Kaulfuss, S. Herold, E. Oswald, and H. Schmidt. 2005. Genetic analysis of enteropathogenic and enterohemorrhagic Escherichia coli serogroup O103 strains by molecular typing of virulence and housekeeping genes and pulsed-field gel electrophoresis. J. Clin. Microbiol. 43:1552–1563. 4. Beutin, L., G. Krause, S. Zimmermann, S. Kaulfuss, and K. Gleier. 2004. Characterization of Shiga toxin-producing Escherichia coli strains isolated from human patients in Germany over a 3-year period. J. Clin. Microbiol. 42:1099–1108. 5. Beutin, L., M. A. Montenegro, I. Orskov, F. Orskov, J. Prada, S. Zimmermann, and R. Stephan. 1989. Close association of verotoxin (Shiga-like toxin) production with enterohemolysin production in strains of Escherichia coli. J. Clin. Microbiol. 27:2559–2564. 6. Bielaszewska, M., H. Schmidt, A. Liesegang, R. Prager, W. Rabsch, H. Tschape, A. Cizek, J. Janda, K. Blahova, and H. Karch. 2000. Cattle can be a reservoir of sorbitol-fermenting shiga toxin-producing Escherichia coli O157:H⫺ strains and a source of human diseases. J. Clin. Microbiol. 38: 3470–3473. 7. Blanco, J., E. A. Gonzalez, S. Garcia, M. Blanco, B. Regueiro, and I. Bernardez. 1988. Production of toxins by Escherichia coli strains isolated from calves with diarrhoea in Galicia (north-western Spain). Vet. Microbiol. 18: 297–311. 8. Brooks, J. T., E. G. Sowers, J. G. Wells, K. D. Greene, P. M. Griffin, R. M.
23. 24.
25.
26.
27. 28.
29.
30.
Hoekstra, and N. A. Strockbine. 2005. Non-O157 Shiga toxin-producing Escherichia coli infections in the United States, 1983–2002. J. Infect. Dis. 192:1422–1429. Brunder, W., H. Schmidt, and H. Karch. 1997. EspP, a novel extracellular serine protease of enterohaemorrhagic Escherichia coli O157:H7 cleaves human coagulation factor V. Mol. Microbiol. 24:767–778. Brunder, W., H. Schmidt, and H. Karch. 1996. KatP, a novel catalaseperoxidase encoded by the large plasmid of enterohaemorrhagic Escherichia coli O157:H7. Microbiology 142:3305–3315. Centers for Disease Control and Prevention. 2007. Laboratory-confirmed non-O157 Shiga toxin-producing Escherichia coli—Connecticut, 2000–2005. MMWR Morb. Mortal. Wkly. Rep. 56:29–31. Clinical and Laboratory Standards Institute. 2005. Performance standards for antimicrobial disk and dilution susceptibility tests for bacteria isolated from animals; approved standards, 2nd ed. M31–A2. Clinical and Laboratory Standards Institute, Wayne, PA. DesRosiers, A., J. M. Fairbrother, R. P. Johnson, C. Desautels, A. Letellier, and S. Quessy. 2001. Phenotypic and genotypic characterization of Escherichia coli verotoxin-producing isolates from humans and pigs. J. Food Prot. 64:1904–1911. Duhamel, G. E., R. A. Moxley, C. W. Maddox, and E. D. Erickson. 1992. Enteric infection of a goat with enterohemorrhagic Escherichia coli (O103: H2). J. Vet. Diagn. Investig. 4:197–200. Eklund, M., F. Scheutz, and A. Siitonen. 2001. Clinical isolates of non-O157 Shiga toxin-producing Escherichia coli: serotypes, virulence characteristics, and molecular profiles of strains of the same serotype. J. Clin. Microbiol. 39:2829–2834. Fey, P. D., R. S. Wickert, M. E. Rupp, T. J. Safranek, and S. H. Hinrichs. 2000. Prevalence of non-O157:H7 Shiga toxin-producing Escherichia coli in diarrheal stool samples from Nebraska. Emerg. Infect. Dis. 6:530–533. Gerner-Smidt, P., K. Hise, J. Kincaid, S. Hunter, S. Rolando, E. HyytiaTrees, E. M. Ribot, B. Swaminathan, and the PulseNet Taskforce. 2006. PulseNet USA: a five-year update. Foodborne Pathog. Dis. 3:9–19. Guth, B. E., T. M. Vaz, T. A. Gomes, S. H. Chinarelli, M. M. Rocha, A. F. de Castro, and K. Irino. 2005. Re-emergence of O103:H2 Shiga toxin-producing Escherichia coli infections in Sao Paulo, Brazil. J. Med. Microbiol. 54: 805–806. Johnson, K. E., C. M. Thorpe, and C. L. Sears. 2006. The emerging clinical importance of non-O157 Shiga toxin-producing Escherichia coli. Clin. Infect. Dis. 43:1587–1595. Karmali, M. A., M. Mascarenhas, S. Shen, K. Ziebell, S. Johnson, R. ReidSmith, J. Isaac-Renton, C. Clark, K. Rahn, and J. B. Kaper. 2003. Association of genomic O island 122 of Escherichia coli EDL 933 with verocytotoxin-producing Escherichia coli seropathotypes that are linked to epidemic and/or serious disease. J. Clin. Microbiol. 41:4930–4940. Kijima-Tanaka, M., K. Ishihara, A. Morioka, A. Kojima, T. Ohzono, K. Ogikubo, T. Takahashi, and Y. Tamura. 2003. A national surveillance of antimicrobial resistance in Escherichia coli isolated from food-producing animals in Japan. J. Antimicrob. Chemother. 51:447–451. Liesegang, A., U. Sachse, R. Prager, H. Claus, H. Steinruck, S. Aleksic, W. Rabsch, W. Voigt, A. Fruth, H. Karch, J. Bockemuhl, and H. Tschape. 2000. Clonal diversity of Shiga toxin-producing Escherichia coli O157:H7/H⫺ in Germany—a ten-year study. Int. J. Med. Microbiol. 290:269–278. Lockary, V. M., R. F. Hudson, and C. L. Ball. 2007. Shiga toxin-producing Escherichia coli, Idaho. Emerg. Infect. Dis. 13:1262–1264. Mariani-Kurkdjian, P., E. Denamur, A. Milon, B. Picard, H. Cave, N. Lambert-Zechovsky, C. Loirat, P. Goullet, P. J. Sansonetti, and J. Elion. 1993. Identification of a clone of Escherichia coli O103:H2 as a potential agent of hemolytic-uremic syndrome in France. J. Clin. Microbiol. 31:296–301. Mora, A., J. E. Blanco, M. Blanco, M. P. Alonso, G. Dhabi, A. Echeita, E. A. Gonzalez, M. I. Bernardez, and J. Blanco. 2005. Antimicrobial resistance of Shiga toxin (verotoxin)-producing Escherichia coli O157:H7 and non-O157 strains isolated from humans, cattle, sheep and food in Spain. Res. Microbiol. 156:793–806. Morabito, S., H. Karch, H. Schmidt, F. Minelli, P. Mariani-Kurkdjian, F. Allerberger, K. A. Bettelheim, and A. Caprioli. 1999. Molecular characterisation of verocytotoxin-producing Escherichia coli of serogroup O111 from different countries. J. Med. Microbiol. 48:891–896. Nataro, J. P., and J. B. Kaper. 1998. Diarrheagenic Escherichia coli. Clin. Microbiol. Rev. 11:142–201. Ogura, Y., K. Kurokawa, T. Ooka, K. Tashiro, T. Tobe, M. Ohnishi, K. Nakayama, T. Morimoto, J. Terajima, H. Watanabe, S. Kuhara, and T. Hayashi. 2006. Complexity of the genomic diversity in enterohemorrhagic Escherichia coli O157 revealed by the combinational use of the O157 Sakai OligoDNA microarray and the Whole Genome PCR scanning. DNA Res. 13:3–14. Ohnishi, M., K. Kurokawa, and T. Hayashi. 2001. Diversification of Escherichia coli genomes: are bacteriophages the major contributors? Trends Microbiol. 9:481–485. Ostroff, S. M., P. I. Tarr, M. A. Neill, J. H. Lewis, N. Hargrett-Bean, and J. M. Kobayashi. 1989. Toxin genotypes and plasmid profiles as determi-
VOL. 46, 2008
31.
32. 33.
34.
35. 36. 37. 38.
39.
CHARACTERIZATION OF CATTLE AND HUMAN VTEC O103:H2
nants of systemic sequelae in Escherichia coli O157:H7 infections. J. Infect. Dis. 160:994–998. Oswald, E., H. Schmidt, S. Morabito, H. Karch, O. Marches, and A. Caprioli. 2000. Typing of 8472intimin genes in human and animal enterohemorrhagic and enteropathogenic Escherichia coli: characterization of a new intimin variant. Infect. Immun. 68:64–71. Pai, C. H., N. Ahmed, H. Lior, W. M. Johnson, H. V. Sims, and D. E. Woods. 1988. Epidemiology of sporadic diarrhea due to verocytotoxin-producing Escherichia coli: a two-year prospective study. J. Infect. Dis. 157:1054–1057. Paros, M., P. I. Tarr, H. Kim, T. E. Besser, and D. D. Hancock. 1993. A comparison of human and bovine Escherichia coli O157:H7 isolates by toxin genotype, plasmid profile, and bacteriophage lambda-restriction fragment length polymorphism profile. J. Infect. Dis. 168:1300–1303. Paton, A. W., and J. C. Paton. 1998. Detection and characterization of Shiga toxigenic Escherichia coli by using multiplex PCR assays for stx1, stx2, eaeA, enterohemorrhagic E. coli hlyA, rfbO111, and rfbO157. J. Clin. Microbiol. 36:598–602. Prager, R., A. Liesegang, W. Voigt, W. Rabsch, A. Fruth, and H. Tschape. 2002. Clonal diversity of Shiga toxin-producing Escherichia coli O103:H2/H⫺ in Germany. Infect. Genet. Evol. 1:265–275. Pugsley, A. P., and B. Oudega. 1987. Methods of studying colicins and their plasmids, p. 105–161. In K. G. Hardy (ed.), Plasmids: a practical approach. IRL Press, Oxford, England. Ratnam, S., S. B. March, R. Ahmed, G. S. Bezanson, and S. Kasatiya. 1988. Characterization of Escherichia coli serotype O157:H7. J. Clin. Microbiol. 26:2006–2012. Schmidt, H., B. Henkel, and H. Karch. 1997. A gene cluster closely related to type II secretion pathway operons of gram-negative bacteria is located on the large plasmid of enterohemorrhagic Escherichia coli O157 strains. FEMS Microbiol. Lett. 148:265–272. Schmidt, H., L. Beutin, and H. Karch. 1995. Molecular analysis of the
40.
41.
42.
43.
44.
45.
46.
3575
plasmid-encoded hemolysin of Escherichia coli O157:H7 strain EDL 933. Infect. Immun. 63:1055–1061. Schmidt, H., E. Maier, H. Karch, and R. Benz. 1996. Pore-forming properties of the plasmid-encoded hemolysin of enterohemorrhagic Escherichia coli O157:H7. Eur. J. Biochem. 241:594–601. Schmidt, H., C. Geitz, P. I. Tarr, M. Frosch, and H. Karch. 1999. NonO157:H7 pathogenic Shiga toxin-producing Escherichia coli: phenotypic and genetic profiling of virulence traits and evidence for clonality. J. Infect. Dis. 179:115–123. Sonntag, A. K., R. Prager, M. Bielaszewska, W. Zhang, A. Fruth, H. Tschape, and H. Karch. 2004. Phenotypic and genotypic analyses of enterohemorrhagic Escherichia coli O145 strains from patients in Germany. J. Clin. Microbiol. 42:954–962. Tarr, P. I., L. S. Fouser, A. E. Stapleton, R. A. Wilson, H. H. Kim, J. C. Vary, Jr., and C. R. Clausen. 1996. Hemolytic-uremic syndrome in a six-year-old girl after a urinary tract infection with Shiga-toxin-producing Escherichia coli O103:H2. N. Engl. J. Med. 335:635–638. Tozzi, A. E., A. Caprioli, F. Minelli, A. Gianviti, L. De Petris, A. Edefonti, G. Montini, A. Ferretti, T. De Palo, M. Gaido, G. Rizzoni, and the Hemolytic Uremic Syndrome Study Group. 2003. Shiga toxin-producing Escherichia coli infections associated with hemolytic uremic syndrome, Italy, 1988–2000. Emerg. Infect. Dis. 9:106–108. Vidovic, S., and D. R. Korber. 2006. Prevalence of Escherichia coli O157 in Saskatchewan cattle: characterization of isolates by using random amplified polymorphic DNA PCR, antibiotic resistance profiles, and pathogenicity determinants. Appl. Environ. Microbiol. 72:4347–4355. Zhang, W. L., M. Bielaszewska, A. Liesegang, H. Tscha ¨pe, H. Schmidt, M. Bitzan, and H. Karch. 2000. Molecular characteristics and epidemiological significance of Shiga toxin-producing Escherichia coli O26. J. Clin. Microbiol. 38:2134–2140.
