Serotypes and Shiga toxin genotypes among Escherichia coli isolated from animals and food in Argentina and Brazil

Veterinary Microbiology 92 (2003) 335±349

Serotypes and Shiga toxin genotypes among Escherichia coli isolated from animals and food in Argentina and Brazil Beatriz E.C. Gutha, Isabel Chinenb, Elizabeth Miliwebskyb, Aloysio M.F. Cerqueiraa,c, GermaÂn Chillemib, JoaÄo R.C. Andraded, Ariela Baschkierb, Marta Rivasb,* a

Departamento de Microbiologia, Imunologia e Parasitologia, Universidade Federal de SaÄo Paulo, SaÄo Paulo, Brazil b Servicio Fisiopatogenia, Instituto Nacional de Enfermedades Infecciosas, ANLIS ``Dr. Carlos G. MalbraÂn'', Av. VeÂlez Sars®eld 563, 1281 Buenos Aires, Argentina c Departamento de Microbiologia e Parasitologia, Universidade Federal Fluminense, Niteroi, Brazil d Disciplina de Microbiologia e Imunologia, Universidade do Estado do Rio de Janeiro, Rio de Janeiro, Brazil Received 11 June 2002; received in revised form 22 October 2002; accepted 1 November 2002

Abstract Shiga toxin (Stx)-producing Escherichia coli (STEC) strains isolated from animals and food in Argentina …n ˆ 44† and Brazil …n ˆ 20† were examined and compared in regard to their phenotypic and genotypic characteristics to evaluate their pathogenic potential. The clonal relatedness of STEC O157 isolates …n ˆ 22† was established by phage typing (PT) and pulsed-®eld gel electrophoresis (PFGE). All O157 strains studied carried eae and enterohemorrhagic E. coli (EHEC)-hly sequences. In Argentina, these strains occurred both in cattle and meat, and 50% of them carried stx2/stx2vh-a genes, whereas in Brazil the O157 strains were isolated from animals, and most harbored the stx2vh-a sequence. At least 13 different O:H serotypes were identi®ed among the nonO157 strains studied, with serotype O113:H21 being found in both countries. All but one non-O157 strains did not carry eae gene, but EHEC-hlyA gene was found in 85.7% of them, and the stx2 genotype was also more prevalent in Argentina than in Brazil …P < 0:01†, where stx1 alone or in association was most common (68.8%). One STEC strain isolated from a calf in Brazil harbored the new variant referred to as stx2-NV206. PFGE analysis showed that STEC O157 strains were grouped in four clusters. One Brazilian strain was considered possibly related (80%) to Argentinean strains of cluster I. Differences in the pathogenic potential, especially in regard to

* Corresponding author. Tel.: ‡54-11-4303-1801; fax: ‡54-11-4303-1801. E-mail address: [email protected] (M. Rivas).

0378-1135/02/$ ± see front matter # 2002 Elsevier Science B.V. All rights reserved. doi:10.1016/S0378-1135(02)00420-0

336

B.E.C. Guth et al. / Veterinary Microbiology 92 (2003) 335±349

serotypes and stx genotypes, were observed among the STEC strains recovered from animals and food in both countries. # 2002 Elsevier Science B.V. All rights reserved. Keywords: STEC; Virulence; Animals; Food; Argentina; Brazil

1. Introduction Shiga toxin (Stx)-producing Escherichia coli (STEC) can cause a broad spectrum of human diseases, including diarrhea, hemorrhagic colitis (HC), and the hemolytic uremic syndrome (HUS) worldwide (Grif®n and Tauxe, 1991; Paton and Paton, 1998). STEC strains causing HC and HUS have been collectively referred to as enterohemorrhagic E. coli (EHEC) (Levine, 1987; WHO, 1997). Although the main virulence property of STEC is the production of one or more types of Stx (Stx1, Stx2, or Stx2 variants), the presence of additional virulence factors have been described in close association with human disease, and are often used to distinguish putative EHEC (Nataro and Kaper, 1998). A well-studied potential STEC virulence marker is the eae gene, which encodes an outer membrane protein (intimin) required for intimate attachment and effacement of microvilli of colonic mucosa (AE lesion) (Kaper et al., 1998). Another marker is the enterohemolysin, or enterohemorrhagic E. coli hemolysin (EHEC-Hly), a member of the repeat in toxin (RTX) family of pore-forming cytolisins (Schmidt et al., 1995). Although most EHEC carry this gene, the role of enterohemolysin in the pathogenesis of HC or HUS is uncertain (Nataro and Kaper, 1998). Escherichia coli O157:H7 is by far the most prevalent serotype associated with sporadic cases and large outbreaks of HC and HUS in many countries. However, there are more than 100 other serotypes that share a similar pathogenic potential for humans, especially O26:H11, O111:non-motile (NM), O113:H21, O128:NM and O145:NM (WHO, 1998). The natural hosts of STEC are wildlife and domestic animals, but cattle appear to be the main reservoir (Beutin et al., 1993; Bonardi et al., 1999; Parma et al., 2000; Pradel et al., 2000; Cobbold and Desmarchelier, 2001), and cattle-derived foods have been the cause of many outbreaks (CDC, 1993, 1995; Paton and Paton, 1998; Ammon et al., 1999; Grif®n et al., 2002). Most human outbreaks and sporadic cases of HC and HUS have been reported from industrialized nations of the northern hemisphere, but their occurrence in a number of countries in the southern hemisphere, including Argentina, Australia, Chile, and South Africa, has also been described (Nataro and Kaper, 1998). In Argentina and Brazil, the occurrence and frequency of human STEC infections is very different, despite their geographical proximity. In Argentina, HUS is endemic (Rivas et al., 1998), with approximately 300 new cases reported annually by Nephrology Hospital Units. In 2001, the estimated annual incidence rate for HUS was 10.4 per 100,000 in children under 5 years of age. STEC O157:H7 is the most prevalent serotype associated with HUS (Miliwebsky et al., 1999), but several other serotypes (O26:H11; O91:NM, O103:NM, O111:NM, O113:H21, O121:H19 and O145:NM) have also been isolated from human patients (Rivas, unpublished data).

B.E.C. Guth et al. / Veterinary Microbiology 92 (2003) 335±349

337

In Brazil, by contrast, human STEC infections in general have been restricted to sporadic cases of non-bloody diarrhea, particularly in younger children, and although O157:H7 strains were recently identi®ed (Irino et al., 2002), the serotypes mostly associated with these infections have been O26:H11, O111:NM and O111:H8 (Giraldi et al., 1990; Irino et al., 2000; Guth et al., 2002a). Moreover, a HUS case associated with a STEC O26:H11 infection was recently identi®ed (Guth et al., 2002b). Despite this, apparently low occurrence of associated human infections, a high prevalence of STEC strains in animals and food has been observed in this country (Cerqueira et al., 1997; Cerqueira et al., 1999). The main aim of this study was to examine and to compare the phenotypic and genotypic characteristics of Argentinean and Brazilian STEC strains isolated from animals and food in order to evaluate their pathogenic potential. The clonal relatedness of STEC O157 isolates recovered in both countries was established by phage typing and pulsed-®eld gel electrophoresis (PFGE). 2. Materials and methods 2.1. Bacterial strains Sixty-four Argentinean and Brazilian STEC strains isolated from animals …n ˆ 29† and food …n ˆ 35† were studied. The serotype, origin and time of isolation of the strains are shown in Table 1. The 44 Argentinean strains were designated ARG and the 20 Brazilian strains were named BRA. Ten ARG strains were submitted to the National Reference Laboratory between 1996 and 2000 by different food and veterinary testing laboratories for full characterization, and four BRA strains were isolated from diarrheic calves in SaÄo Paulo State (kindly supplied by Dr. T. Yano, UNICAMP, SaÄo Paulo, Brazil). Twenty-one strains belonged to serotype O157:H7, one was O157:NM, and 42 were non-O157. The presence of stx1 and stx2 genes was identi®ed by a multiplex PCR using the primers described by Pollard et al. (1990) in Argentina, and by Paton et al. (1993) in Brazil. The reference E. coli strains EDL933 (stx1, stx2), E40705 (stx1), E30138 (stx2) and DH5a were used as positive and negative controls. 2.2. Phenotypic characterization of isolates Fermentation of sorbitol within 24 h was tested in Andrade peptone water containing 1.0% D-sorbitol (Farmer and Kelly, 1991). Production of b-D-glucuronidase was assessed by hydrolysis of p-nitrophenyl b-D-glucuronide using Colibritania discs (Laboratorios Britania, Buenos Aires, Argentina) (Edberg and Trepeta, 1983). Serotyping of non-O157 STEC strains was performed by tube agglutination (Ewing, 1986) using antisera O1±O172 and H1±H51, kindly supplied by the Centers for Diseases Control and Prevention (CDC), Atlanta, GA, USA. For detection of EHEC-Hly, bacteria were grown on tryptose blood agar base (Difco Laboratories, Detroit, MI, USA) supplemented with 10 mM CaCl2 and 5% de®brinated sheep blood washed three times in phosphate-buffered saline, pH 7.2. After 24 h of incubation, the EHEC-Hly

338

B.E.C. Guth et al. / Veterinary Microbiology 92 (2003) 335±349

Table 1 Origin of the collection of Shiga toxin-producing Escherichia coli strains studied Country

Serotype

Argentina

Non-O157

Brazil

Number of strains

Origin

Time of isolation

Reference

8 7

Healthy steers Healthy calves

1997/1998 1999/2000

Gioffre et al., 2002 Meichtri et al., 2000

O157:H7

2

1998/1999

Otero et al., 2001

Non-O157

9 1

Healthy and diarrheic Veal Holstein calves Frozen hamburgers Soft cheese samples

1998/2000

Gomez et al., 2002

O157:H7 Non-O157 O157:H7

6 1 10

Ground beef samples

2000

Chinen et al., 2001

Animal and meat samples

1996/2000

This study

O157:H7

3

1996/1997

Cerqueira et al., 1999

Non-O157 O157:NM Non-O157 Non-O157

4 1 3 9

Healthy calves, heifers and cows Diarrheic calves

Unknown

This study

Ground beef and hamburger samples

1995/1996

Cerqueira et al., 1997

phenotype was established (Beutin et al., 1989). Production of Stx was determined by a cytotoxicity assay on Vero cells with bacterial supernatant ®ltrates and periplasmic cell extracts obtained from Penassay broth (Antibiotic Medium 3; Difco) cultures (Karmali et al., 1985), and con®rmed by neutralization assays with Stx1- and Stx2-speci®c monoclonal antibodies (MAb 13C4 and BC5BB12, respectively), kindly provided by Dr. N.A. Strockbine, CDC. The antimicrobial susceptibility to ampicillin, ce®xime, cefotaxime, cefuroxime, cephalothin, colistin, chloramphenicol, gentamicin, nalidixic acid, nor¯oxacin, streptomycin, tetracycline and trimethoprim-sulfamethoxazole was performed according to NCCLS standards and methods (NCCLS, 2000). For phage typing, a culture from the O157 strains grown in aerated phage broth (Difco) for 1.5 h at 37 8C, was ¯ooded onto phage agar plates (Difco) and 16 phages (kindly provided by R. Ahmed of the Laboratory Centre for Disease Control (LCDC), Winnipeg, Man., Canada) were applied. Lytic reactions were observed after overnight incubation at 37 8C (Ahmed et al., 1987; Khakhria et al., 1990). The results were interpreted according to a phage type list of 72 con®rmed and 13 provisional phage types. A react but does not conform (RDNC) result was con®rmed by testing ®ve additional colonies of the same strain. 2.3. Genotypic characterization of isolates The presence of stx1 and stx2 was con®rmed by colony blot hybridization assays using speci®c UTP-digoxigenin-labelled gene probes under stringent washing conditions

B.E.C. Guth et al. / Veterinary Microbiology 92 (2003) 335±349

339

(Thomas et al., 1991). The eae gene was determined by colony blot hybridization, using as positive and negative control the E. coli strains 2348/69 and HS, respectively. The EHEChlyA gene was identi®ed by PCR using the primers hlyA1 and hlyA4 as described by Schmidt et al. (1995), using as positive and negative control the E. coli strains E32511 and HS, respectively. For differentiation of the Stx2 variants, the genotyping method of Tyler et al. (1991) was performed. Brie¯y, after a PCR using Stx2 and Stx2 variant-speci®c oligonucleotide primers, 10 ml of the ampli®cation fragments were digested with HaeIII, RsaI, and NciI. The restriction fragment length polymorphism (RFLP) obtained by conventional gel electrophoresis using 2.5% agarose gels, was analyzed. The PCR-RFLP method of Tyler et al. (1991) was extended to include the primers and restriction enzymes described by PieÂrard et al. (1998). The reference E. coli strains 92-3580 O157:H7 (stx2vha) and 93-016 O113:H21 (stx2d) were kindly provided by Dr. D. Woodward, LCDC, Winnipeg, Man., Canada, and E. coli strain EH250 ONT:H12 (stx2vh-d) by Dr. D. PieÂrard, Department of Microbiology, Academisch Ziekenhuis Vrije Universiteit Brussel, Brussels, Belgium. 2.4. DNA sequencing and sequence analysis DNA sequencing was performed on both strands by the method of Sanger et al. (1977), using the BigDye terminator methodology (Applied Biosystem/Perkin-Elmer, Foster City, CA, USA) with the VT2c±VT2d primer pair (Tyler et al., 1991). The sequences were analyzed in an ABI Prism 377 DNA sequencer (Applied Biosystem/Perkin-Elmer). Nucleotide sequence analysis was performed with BioEdit sequence alignments editor software (Hall, 1999). Homology searches were performed using the BLAST GenBank database. 2.5. Clonal relatedness of STEC O157 strains PFGE was performed using the 24 h CDC protocol (1998), with minor modi®cations. Digestion was carried out with 25U of XbaI (Promega Corporation, Madison, WI, USA) at 37 8C for 18 h. The standard strain used was E. coli O157:H7 G5244 and was provided by CDC. DNA fragments were resolved in 1% agarose gel in 0.5 Tris borate EDTA electrophoresis buffer at 14 8C, in a contour clamped homogeneous electric ®eld (CHEF) DR-III electrophoresis chamber (Bio-Rad Laboratories, Hercules, CA, USA). The run time was 23 h, with a constant voltage of 200 V, using a linear pulse ramp of 2.2±54.2 s. The relatedness among the patterns was estimated by the proportions of shared bands, after applying the DICE coef®cient. Data recording and calculation were performed using the RAPDistance program (Armstrong et al., 1994). The resulting matrix of pairwise distance was used to generate a phenogram based on the UPGMA method, included in MEGA software (Kumar et al., 1993). 2.6. Statistical analysis The data were analyzed by the chi-square test with Yates' correction for continuity. P < 0:01 was considered to be statistically signi®cant.

340

B.E.C. Guth et al. / Veterinary Microbiology 92 (2003) 335±349

3. Results 3.1. Phenotypic characteristics of STEC strains Independently of the country of origin, the STEC O157 strains were D-sorbitol and b-Dglucuronidase negative, whereas all the 42 non-O157 strains fermented D-sorbitol, and 39 (92.8%) were also b-D-glucuronidase positive. Thirteen different O:H serotypes were found among the non-O157 strains isolated in both countries, comprising 12 O serogroups and 17 H serogroups (Table 2). Seven strains were O non-typeable, and ®ve were O rough; two strains were H non-typeable, and one was non-motile. In Argentina, serotype O8:H19 was frequently identi®ed among animal and food samples (12/26, 46.1%), but it was not found in Brazil, and the only common serotype identi®ed in both countries was O113:H21. In Brazil, serotypes O82:H8 and O113:H21 were identi®ed in calves and ground beef. EHEC-Hly production was observed in all O157 strains and in 36 of the 42 (85.7%) nonO157 strains (Table 2). In general, the STEC strains studied were susceptible to the antimicrobial agents assayed, except for two Argentinean strains, one of serotype O157:H7 isolated from a beef sample and one O8:H19 from a steer, that were resistant to ampicillin and to ampicillin/cephalotin, respectively. The phage types most frequently identi®ed among the STEC O157 strains were PT4 and PT87 (four strains each), followed by PT14 (three strains), PT26 (one strain), and PT49 (one strain). Nine strains (41%) did not conform to any published phage type, and were designated RDNC. The Brazilian strains did not belong to any recognized PT, except for one strain (BRA YB20) identi®ed as PT14 (Table 3). 3.2. Genotypic characteristics of STEC strains Different stx types occurred among the non-O157 STEC strains from Argentina and Brazil. The stx2 genotype was more prevalent in Argentina than in Brazil …P < 0:01†, representing 76.9% of the isolates, including three strains of the stx2vh-b genotype. The remaining non-O157 STEC strains carried the stx1 (four strains), and stx1/stx2 (two strains) genes. In Brazil, stx1 alone or in association with stx2 was most common (68.8%), while only ®ve STEC strains harbored the stx2 gene. The stx2 variant identi®ed was stx2vh-b, alone or in combination with stx1 or stx2 (three strains). One strain isolated from a calf in Brazil and serotyped as O4:H4 did not ®t the patterns of the strains with Stx2vh-a, Stx2vh-b and Stx2d subtypes (Fig. 1A and B). The restriction pattern of this isolate using HaeIII, RsaI and NciI endonucleases was the same as those described by Bertin et al. (2001), and was referred to as stx2-NV206. The nucleotide sequencing of the fragments obtained after ampli®cation using the VT2c±VT2d primer pair con®rmed it had 100% the homology with the sequence of the stx2-NV206 variant (GenBank accession number AF329817). In regard to the 22 STEC O157 strains, the PCR-RFLP genotyping method showed that nine (40.9%) strains harbored stx2/stx2vh-a genes, ®ve carried stx2vh-a, and in the other ®ve strains stx1/stx2vh-a sequences were identi®ed. In the three remaining strains, no stx2

B.E.C. Guth et al. / Veterinary Microbiology 92 (2003) 335±349

341

Table 2 Characteristics of non-O157 Shiga toxin-producing Escherichia coli strains isolated from cattle and food in Argentina and Brazil Country

Origin

Argentina

Cattle

Brazil

Number of strains

Stx genotype

Number of EHEC-hlyA positive strainsa

Serotype (number of strains)b,c,d

4

1

1

10

2

9

O8:NT, O26:H11, O116:NM, NT:H16 O2:H25, O8:H19 (7), O11:H14, NT:H14 NT:H21

1

2vh-b

1

Ground beef and hamburgers

6

2

6

2 2

1/2 2vh-b

2 1

O8:H16, O8:H19 (3), O39:H49, NT:NT O8:H19, O60:H19 O113:H21, NT:H7

Soft cheese

1

2

1

O8:H19

Cattle

2 1 2 1 1

1 2 1/2 2vh-b 2-NV206

1 1 2 1 1

NT:H11, R:H18 O113:H21 O82:H8, NT:H28 R:H19 O4:H4

Ground beef and hamburgers

3 1 3

1 2 1/2

3 1 3

1 1

1/2vh-b 2/2vh-b

1 1

R:H7, R:H14, R:H47 O113:H21 O46:H38, O65:H48, O116:H21 R:H42 O82:H8

a

This result presented 100% of correlation with the enterohemolytic phenotype, except for the strain O4:H4. NT, non-typeable; NM, non-motile, R, rough. None of the STEC strains carried eae genes, except for one O26:H11 strain. c Underlined serotypes correspond to strains with a positive EHEC-hlyA result. d Serotypes found in humans are in bold. b

variants were found (Table 3). The stx results were in agreement with the data obtained with the Vero cells cytotoxicity assays. Hybridization results showed that the eae sequence was present in all O157 strains, and in one O26:H11 strain isolated from a steer in Argentina. The EHEC-hlyA sequence was identi®ed in all the STEC strains studied that presented the enterohemolytic phenotype, except for the strain YB33, isolated in Brazil, that harbored the EHEC-hlyA gene but did not express the phenotype (Tables 2 and 3). 3.3. Clonal relatedness of STEC O157 strains Twenty-two distinct genomic patterns were obtained when the STEC O157 strains were processed by XbaI-PFGE. These patterns showed 13±20 discernible fragments ranging approximately from 48 to 590 kb in molecular weight (Fig. 2). The patterns with less than

342

B.E.C. Guth et al. / Veterinary Microbiology 92 (2003) 335±349

Table 3 Characteristics of Shiga toxin-producing Escherichia coli O157 strains isolated from animals and food in Argentina and Brazil Strain numbera

Origin

Sampling year

Stx genotypeb

Phage type

PFGE Pattern number

c

ARG 215 ARG 159 ARG 71e ARG 168 BRA YB20f ARG 214

Ground beef Ground beef Beef Ground beef Diarrheic calf Ground beef

2000 2000 1996 2000 Unknown 2000

1/2vh-a 2vh-a 2vh-a 1/2vh-a 1/2vh-a 1

ARG 325 ARG 329 ARG 238

Beef Beef Lamb meat

1999 1999 1999

2/2vh-a 2/2vh-a 2/2vh-a

4 4 26

BRA B1/1 BRA B18/1

Healthy calf Healthy calf

1997 1997

2vh-a 2vh-a

c

ARG ARG ARG ARG

Beef Beef Diarrheic calf Ground beef

1996 2000 1998 2000

2/2vh-a 2/2vh-a 2/2vh-a 1/2vh-a

Porcine meat Sheep Healthy calf Beef Beef Healthy calf Ground beef

1996 1998 1999 1997 1999 1997 2000

1/2 2/2vh-a 2vh-a 2/2vh-a 2/2vh-a 2 1/2vh-a

109 163 145 250

ARG 94 ARG 33 ARG 438 ARG 148 ARG 237 BRA GC 148 ARG 213

Cluster numberd

1 2 3 4 5 6

I

7 8 9

II

10 11

III

87 87 87

12 13 14 15

IV

c

16 17 18 19 20 21 22

UR UR UR UR UR UR UR

4 14 c

14 c

c

c

4 49 87 14 c c

a

ARG, strains isolated in Argentina; BRA, strains isolated in Brazil. All the strains carried eae and EHEC-hlyA sequences. c Reacted with phage set but did not correspond to any recognized phage type. d UR, unrelated. e Ampicillin-resistant strain. f Diarrheic E. coli O157:NM. b

six band differences were grouped in clusters when they were considered possibly related, according to criteria of Tenover et al. (1995) (Fig. 3). Fifteen patterns were grouped in four clusters (Table 3). Four of the six strains included in cluster I were isolated from ground beef and beef samples obtained in Argentina during 2000. The two remaining strains of this cluster, ARG 71 and BRA YB20 isolated from a beef sample and a diarrheic calf in Argentina and Brazil, respectively, belonged to PT14 and were 84% related. The major genetic relatedness was observed between the strains ARG 215 (pattern #1) and ARG 159 (pattern #2) isolated from ground beef samples. They shared 97% of the XbaI-PFGE restriction fragments, with only one band difference. Strains ARG 215 (pattern #1), ARG 168 (pattern #4) and ARG 214 (pattern #6) presented the same non-recognized lytic reaction pattern.

B.E.C. Guth et al. / Veterinary Microbiology 92 (2003) 335±349

343

Fig. 1. (A) RFLP analysis of Stx2 genotypes by digestion of the 285-bp amplified product, after PCR using c and d primers from Tyler et al. (1991), with HaeIII (a), RsaI (b) and NciI (c). Lanes 1, 2, 3: E. coli BRA YB33 O4:H4 (stx2-NV206); lanes 4, 5, 6: E. coli EDL933 O157:H7 (stx2); lanes 8, 9, 10: E. coli 92-3580 O157:H7 (stx2vh-a); lanes 11, 12, 13: E. coli 93-016 O113:H21 (stx2vh-b); lane 7: 100-bp DNA ladder. (B) RFLP analysis of Stx2 genotypes by digestion of the 348-bp amplified product, after PCR using e and f primers from PieÂrard et al. (1998), with HaeIII (a) and PvuII (b). Lanes 1, 2: E. coli BRA YB33 O4:H4 (stx2-NV206); 3, 4 E. coli EH250 ONT:H12 (stx2d); 5, 100-bp DNA ladder.

Fig. 2. PFGE analysis of XbaI digested genomic DNA. Lanes 1, 13, 25: standard strain E. coli G5244; lanes 2± 7: cluster I strains; lanes 8±10: cluster II strains; lanes 11, 12: cluster III strains; lanes 14±17: cluster IV strains; lanes 18±24: unrelated strains.

344

B.E.C. Guth et al. / Veterinary Microbiology 92 (2003) 335±349

Fig. 3. XbaI-PFGE phenogram based on DICE coefficient of similarity for STEC O157 strains isolated from animals and food samples, in Argentina and Brazil.

The strains included in cluster II were isolated in Argentina in 1999, and all carried stx2/ stx2vh-a sequences. Two strains, ARG 325 (pattern #7) and ARG 329 (pattern #8), recovered from beef samples, were of PT4 and shared 82% similarity. Strain ARG 238 (pattern #9), recovered from lamb meat, was of PT26 and showed 94% identity with strain ARG 329. The strains BRA B1/1 (pattern #10) and BRA B18/1 (pattern #11), isolated from healthy calves in Brazil in 1997, were grouped in cluster III. They harbored stx2vh-a genes and showed 87% similarity by XbaI-PFGE. Neither belonged to any recognized PT to date, but presented the same lytic reaction pattern. All the strains in cluster IV were isolated in Argentina. Strain ARG 145 (pattern #14) isolated from a diarrheic calf in 1998, was characterized as stx2/stx2vh-a, and belonged to PT87 with two other strains (patterns #12 and 13) isolated from beef samples, having at least 90% similarity. The calf strain was 94% genetically similar to the strain ARG 250 (pattern #15), stx1/stx2vh-a, showing a non-recognized PT. Seven strains (patterns #16±22) were located very distant in the phenogram and showed to be genetically unrelated. 4. Discussion In the present study, STEC strains recovered from animals and food in Argentina and Brazil presented several phenotypic and genotypic similarities, however important differences especially in regard to serotypes and stx genotypes could be observed.

B.E.C. Guth et al. / Veterinary Microbiology 92 (2003) 335±349

345

Serotype O157:H7 was more frequently detected among animal and food strains isolated in Argentina (40.9%), whereas in Brazil, the few strains identi®ed so far were from the animal reservoir. All the O157 strains, independently of the origin and country, were eae and EHEC-hlyA positive. Interestingly, in Argentina, 50% of these strains carried the stx2/ stx2vh-a genes, as do most of the STEC O157:H7 strains isolated from HUS cases in this country (Chinen et al., 2000). On the other hand, in Brazil, this stx genotype was not observed, and except for one O157 strain that was stx2, the others harbored stx2vh-a alone or associated with stx1. Nishikawa et al. (2000) recently reported that STEC strains carrying only stx2vh-a were likely less virulent and caused bloody diarrhea less frequently than strains which harbored stx2 or both stx2/stx2vh-a genes. Moreover, Boerlin et al. (1999) have modeled the association between the virulence factors and the presence of human disease using a logistic regression model. The only signi®cant interaction at the 5% level was between eae and stx2, associated with isolates from serotypes found in severe human disease. Several different O:H serotypes were found among the non-O157 strains isolated in both countries, but the only common serotype identi®ed was O113:H21. Although some of the serotypes presently identi®ed have never, or only sporadically, been found in humans in Argentina and Brazil, others like O26:H11, O82:H8, O113:H21, NT:H21 and NT:H28 have been found in association with human disease in our countries as well as in other regions. To our knowledge, serotype O11:H14 corresponding to one strain recovered from a steer, harboring stx2 gene, has not previously been reported as STEC (Bettelheim, 2002; http://www.microbionet.com.au/frames/feature/vtec/brief06.html). The stx2 genotype was also more prevalent among the non-O157 strains isolated in Argentina than in Brazil, where stx1 alone or in association with stx2 was most common. In addition, except for one O26:H11 strain isolated from a healthy steer, all the non-O157 strains did not carry eae, but EHEC-hlyA gene was found in 85.7% of them. Moreover, subtype stx2d was not observed among isolates of both countries. This result was in accordance with those reported in Australia, where stx2d subtypes were very rarely found among STEC isolates from bovine sources (Ramachandran et al., 2001). Phage typing and PFGE appear to be the most useful typing techniques for E. coli O157:H7 strains. XbaI is the most discriminatory endonuclease used in PFGE for this purpose. Despite its lower discriminatory power, phage typing is helpful in interpreting PFGE data, and it is also useful as a simple and rapid screening method (Barret et al., 1994). In this study, a marked correlation between the results obtained with phage typing and XbaI-PFGE was observed. PT4 and PT87 were the predominant PTs observed among the strains studied, and in Argentina, these PTs are also prevalent among the STEC O157 strains of human origin (Rivas, unpublished data). However, the PT distribution found in the present study was different to that described in several other countries (Saari et al., 2001). PFGE analysis grouped the STEC O157 strains in four clusters. Interestingly, one Brazilian strain (YB20) was considered possibly related (80%) to the Argentinean strains included in cluster I. Moreover, the PFGE-patterns corresponding to the cluster IV presented a high clonal relatedness, with the most prevalent patterns being associated with strains isolated from human cases in different geographical areas of Argentina (Chinen et al., 2000).

346

B.E.C. Guth et al. / Veterinary Microbiology 92 (2003) 335±349

Most of the STEC strains presently studied were isolated from dairy and beef cattle, and food during different surveys conducted in Argentina and Brazil (Cerqueira et al., 1997; Cerqueira et al., 1999; Meichtri et al., 2000; Chinen et al., 2001; Otero et al., 2001; Gioffre et al., 2002; Gomez et al., 2002), and thus should be representative of the distribution and type of isolates present in both countries. Differences in the pathogenic potential, especially in regard to serotypes and stx genotypes were presently observed. However, there are no studies comparing animal husbandry, feeding regimens, source and type of livestock, and slaughter practices in Argentina and Brazil that could possibly explain those differences and their impact on the incidence of STEC-related diseases. Acknowledgements The technical assistance of Ana Garbini and Natalia Martinez is acknowledged. This study was partially supported by grants from Consejo Nacional de Investigaciones Cientõ®cas y TecnoloÂgicas (CONICET) (PIP No. 0020/98) and FundacioÂn Alberto J. Roemmers, Argentina, and by grants from Financiadora de Estudos e Projetos/MinisteÂrio da CieÃncia e Tecnologia/Programa de Apoio a NuÂcleos de ExceleÃncia (FINEP/MCT/ PRONEX), Brazil. References Ahmed, R., Bopp, C., Borczyk, A., Kasatiya, S., 1987. Phage-typing scheme for Escherichia coli serotype O157:H7. J. Infect. Dis. 155, 806±809. Ammon, A., Peterson, L.R., Karch, H., 1999. A large outbreak of haemolytic uremic syndrome caused by an unusual sorbitol-fermenting strain Escherichia coli O157:H-. J. Infect. Dis. 179, 1274±1277. Armstrong, J., Gibbs, A., Peakall, R., Weiler, G., 1994. RAPDistance Program, Version 1.04 for the Analysis of Patterns of RAPD Fragments. Australian National University, Canberra, Australia. Barret, T.J., Lior, H., Green, J.H., Khakhria, R., Wells, J.G., Bell, B.P., Greene, K.D., Lewis, J., Griffin, P.M., 1994. Laboratory investigation of a multistate food-borne outbreak of Escherichia coli O157:H7 by using pulsed-field gel electrophoresis and phage typing. J. Clin. Microbiol. 32, 3013±3017. Bertin, Y., Boukhors, K., Pradel, N., Livrelli, V., Martin, C., 2001. Stx2 subtyping of Shiga toxin-producing Escherichia coli isolated from cattle in France: detection of a new Stx2 subtype and correlation with additional virulence factors. J. Clin. Microbiol. 39, 3060±3065. Beutin, L., Montenegro, M.A., érskov, I., érskov, F., Prada, J., Zimmermann, S., Stephan, R., 1989. Close association of Verotoxin (Shiga-like toxin) production with enterohemolysin production in strains of Escherichia coli. J. Clin. Microbiol. 27, 2559±2564. Beutin, L., Geier, D., Steinruck, H., Zimmermann, S., Scheutz, F., 1993. Prevalence and some properties of verotoxin (Shiga-like toxin)-producing Escherichia coli in seven different species of healthy domestic animals. J. Clin. Microbiol. 31, 2483±2488. Boerlin, P., McEwen, S.A., Boerlin-Petzold, F., Wilson, J.B., Johnson, R.P., Gyles, C.L., 1999. Associations between virulence factors of Shiga toxin-producing Escherichia coli and disease in humans. J. Clin. Microbiol. 37, 497±503. Bonardi, S., Maggi, E., Bottarelli, A., Pacciarini, M.L., Ansuini, A., Vellini, G., Morabito, S., Caprioli, A., 1999. Isolation of verocytotoxin-producing Escherichia coli O157:H7 from cattle at slaughter in Italy. Vet. Microbiol. 67, 203±211. Centers for Disease Control and Prevention, 1993. Update: multistate outbreak of Escherichia coli O157:H7 infections from hamburgers, Western United States, 1992±1993. Morb. Mortal. Wkly. Rep. 42, 258±263.

B.E.C. Guth et al. / Veterinary Microbiology 92 (2003) 335±349

347

Centers for Disease Control and Prevention, 1995. Escherichia coli O157:H7 outbreak linked to commercially distributed dry-cured salami, Washington and California, 1994. Morb. Mortal. Wkly. Rep. 44, 157±160. Centers for Disease Control and Prevention, Foodborne and Diarrheal Diseases Branch, 1998. Standarized Molecular Subtyping of Foodborne Bacterial Pathogens by Pulsed-Field Gel Electrophoresis: A Training Manual. Centers for Disease Control and Prevention, Atlanta, GA, USA. Cerqueira, A.M.F., Tibana, A., Guth, B.E.C., 1997. High occurrence of Shiga-like toxin-producing strains among diarrheagenic Escherichia coli isolated from raw beef products in Rio de Janeiro City. Brazil J. Food Prot. 60, 1±5. Cerqueira, A.M.F., Guth, B.E.C., Joaquim, R.M., Andrade, J.R.C., 1999. High occurrence of Shiga toxinproducing Escherichia coli (STEC) in healthy cattle at Rio de Janeiro State. Brazil Vet. Microbiol. 70, 111± 121. Chinen, I., Miliwebsky, E.S., Chillemi, G.M., Baschkier, A., Rivas, M., 2000. Clonal relatedness of Argentinean Escherichia coli O157:H7 strains by pulsed-field gel electrophoresis. In: Abstracts of the Ninth International Congress on Infectious Diseases. Buenos Aires, Argentina, p. 97. Chinen, I., Tanaro, J.D., Miliwebsky, E.S., Lound, L.H., Chillemi, G.M., Ledri, S., Baschkier, A., Scarpin, M., Manfredi, E., Rivas, M., 2001. Isolation and characterization of Escherichia coli O157:H7 from retail meats in Argentina. J. Food Prot. 64, 1346±1351. Cobbold, R., Desmarchelier, P., 2001. Characterisation and clonal relationships of Shiga-toxigenic Escherichia coli (STEC) isolated from Australian dairy cattle. Vet. Microbiol. 79, 323±335. Edberg, S.C., Trepeta, R.W., 1983. Rapid and economical identification and antimicrobial susceptibility test methodology for urinary tract pathogens. J. Clin. Microbiol. 18, 1287±1291. Ewing, W.H., 1986. Edwards & Ewing's Identification of Enterobacteriaceae, Fourth ed. Elsevier, New York, NY, USA. Farmer III, J.J., Kelly, M., 1991. Enterobacteriaceae, In: Balows, A., Hausler, W.J., Hermann, K.L., Isenberg, H.D., Shadomy, H.J. (Eds.), Manual of Clinical Microbiology, Fifth ed. American Society for Microbiology, Washington, DC, pp. 360±383. GioffreÂ, A., Meichtri, L., Miliwebsky, E., Baschkier, A., Chillemi, G., Romano, M.I., Sosa Estani, S., Cataldi, A., Rodriguez, R., Rivas, M., 2002. Detection of Shiga toxin-producing Escherichia coli by PCR in cattle in Argentina. Evaluation of two procedures. Vet. Microbiol. 87, 301±313. Giraldi, R., Guth, B.E.C., Trabulsi, L.R., 1990. Production of Shiga-like toxin among Escherichia coli strains and other bacteria isolated from diarrhea in SaÄo Paulo. Brazil J. Clin. Microbiol. 28, 1460±1462. Gomez, D., Miliwebsky, E., Fernandez Pascua, C., Baschkier, A., Manfredi, E., Zotta, M., Nario, F., PiquõÂn, A., Sanz, M., EtcheverrõÂa, A., Padola, N., Parma, A., Rivas, M., 2002. Aislamiento y caracterizacioÂn de Escherichia coli productor de toxina Shiga en hamburguesas supercongeladas y quesos de pasta blanda. Rev. Arg. Microbiol. 34, 66±71. Griffin, P.M., Tauxe, R.V., 1991. The epidemiology of infections caused by Escherichia coli O157:H7, other enterohemorrhagic E. coli, and the associated hemolytic uremic syndrome. Epidemiol. Rev. 13, 60±98. Griffin, P.M., Mead, P., Sivapalasingam, S., 2002. Escherichia coli O157:H7 and other enterohemorrhagic Escherichia coli. In: Blaser, M.J., Smith, P.D., Ravdin, J.I., Greenberg, H.B., Guerrant, R.L. (Eds.), Infections of the Gastrointesinal Tract, Second ed. Lippincott Williams and Wilkins, Philadelphia, pp. 627±642. Guth, B.E.C., Ramos, S.R.T.S., Cerqueira, A.M.F., Andrade, J.R.C., Gomes, T.A.T., 2002a. Phenotypic and genotypic characteristics of Shiga toxin-producing Escherichia coli strains isolated from children in SaÄo Paulo, Brazil. Mem. Inst. Oswaldo Cruz. 97, 1085±1089. Guth, B.E.C., Souza, R.L., Vaz, T.M.I., Irino, K., 2002b. First Shiga toxin-producing Escherichia coli isolate from a patient with hemolytic uremic syndrome. Brazil Emerg. Infect. Dis. 8, 535±536. Hall, T.A., 1999. BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucl. Acids Symp. Ser. 41, 95±98. Irino, K., Gomes, T.A.T., Vaz, T.I., Kano, E., Kato, M.A.M.F., Dias, A.M.G., GoncËalves, C.R., Guth, B.E.C., 2000. Prevalence of Shiga toxin and intimin gene sequences among Escherichia coli of serogroups O26, O55, O111, O119 and O157 isolated in SaÄo Paulo, Brazil. In: Abstracts of the Fourth International Symposium and Workshop on Shiga Toxin (Verocytotoxin)-Producing Escherichia coli Infections. Kyoto, Japan, p. 107.

348

B.E.C. Guth et al. / Veterinary Microbiology 92 (2003) 335±349

Irino, K., Vaz, T.M.I., Kato, M.A.M.F., Naves, Z.V.F., Lara, R.R., Marco, M.E.C., Rocha, M.M.M., Moreira, T.P., Gomes, T.A.T., Guth, B.E.C., 2002. O157:H7 Shiga toxin-producing Escherichia coli strains associated with sporadic cases of diarrhea in SaÄo Paulo. Brazil Emerg. Infect. Dis. 8, 446±447. Kaper, J.B., Elliot, S., Sperandio, V., Perna, N.T., Mayhew, G.F., Blattner, F.R., 1998. Attaching-and-effacing intestinal histopathology and the locus of enterocyte effacement. In: Kaper, J.B., O'Brien, A.D. (Eds.), Escherichia coli O157:H7 and Other Shiga Toxin-Producing E. coli Strains. American Society Microbiology, Washington, DC, pp. 163±182. Karmali, M.A., Petric, M., Lim, C., Cheung, R., Arbus, G.S., 1985. Sensitive method for detecting low numbers of Verotoxin-producing Escherichia coli in mixed cultures by use of colony sweeps and polymyxin extraction of Verotoxin. J. Clin. Microbiol. 22, 614±619. Khakhria, R., Duck, D., Lior, H., 1990. Extended phage-typing scheme for Escherichia coli O157:H7. Epidemiol. Infect. 105, 511±520. Kumar, S., Tamura, K., Nei, M., 1993. MEGA: Molecular Evolutionary Genetics Analysis, Version 1.02. The Pennsylvania State University, University Park, PA, USA. Levine, M.M., 1987. Escherichia coli that cause diarrhea: enterotoxigenic, enteropathogenic, enteroinvasive, enterohemorrhagic, and enteroadherent. J. Infect. Dis. 155, 377±389. Meichtri, L.H., Miliwebsky, E., Gioffre, A.K., Baschkier, A., Cataldi, A., Romano, M.I., Chillemi, G., Rodriguez, H.R., Rivas, M., 2000. Detection of Shiga toxin-producing Escherichia coli by PCR in beef cattle at slaughtering level in Argentina. In: Abstracts of 46th International Congress of Meat Science and Technology. Buenos Aires, Argentina, pp. 708±709. Miliwebsky, E.S., Balbi, L., GoÂmez, D., Wainsztein, R., Rua, M.C., RoldaÂn, C., Calleti, M.G., Leardini, N.A., Baschkier, A., Chillemi, G.M., Rivas, M., 1999. SõÂndrome ureÂmico hemolõÂtico en ninÄos de Argentina: asociacioÂn con la infeccioÂn por Escherichia coli productor de toxina Shiga. BioquõÂmica y PatologõÂa ClõÂnica (Argentina) 63, 113±121. Nataro, J.P., Kaper, J.B., 1998. Diarrheagenic Escherichia coli. Clin. Microbiol. Rev. 11, 142±201. National Committee for Clinical Laboratory Standards, 2000. Performance Standards for Antimicrobial Disk Susceptibility Tests, Seventh ed. vol. 20 (1). Approved Standard M100-S10. National Committee for Clinical Laboratory Standards, Wayne, PA, USA. Nishikawa, Y., Zhou, Z., Hase, A., Ogasawara, J., Cheasty, T., Haruki, K., 2000. Relationship of genetic type of Shiga toxin to manifestation of bloody diarrhea due to enterohemorrhagic Escherichia coli serogroup O157 isolates in Osaka City, Japan. J. Clin. Microbiol. 38, 2440±2442. Otero, J.L., RoldaÂn, M.L., Chinen, I., Miliwebsky, E., Frizzo, L., NoÂboli, C., Rivas, M., 2001. Escherichia coli O157:H7 aislado de ganado bovino y su relacioÂn clonal con cepas de distintos orõÂgenes. In: Abstracts of Ninth Congreso Argentino de MicrobiologõÂa. Buenos Aires, Argentina, p. 104. Parma, A.E., Sanz, M.E., Blanco, J.E., Blanco, J., Blanco, M., Padola, N.L., Echeverria, A.I., 2000. Virulence genotypes and serotypes of verotoxigenic Escherichia coli isolated from cattle and foods in Argentina. Eur. J. Epidemiol. 16, 757±762. Paton, A.W., Paton, J.C., Goldwater, P.N., Maning, P.A., 1993. Direct detection of Escherichia coli Shiga-like toxin genes in primary fecal cultures by polymerase chain reaction. J. Clin. Microbiol. 31, 3063±3067. Paton, J.C., Paton, A.W., 1998. Pathogenesis and diagnosis of Shiga toxin-producing Escherichia coli infections. Clin. Microbiol. Rev. 11, 450±479. PieÂrard, D., Muyldermans, G., Moriau, L., Stevens, D., Lauwers, S., 1998. Identification of new verocytotoxin type 2 variant B-subunit genes in human and animal Escherichia coli isolates. J. Clin. Microbiol. 36, 3317±3322. Pollard, D.R., Johnson, W.M., Lior, H., Tyler, S.D., Rozee, K.R., 1990. Rapid and specific detection of verotoxin genes in Escherichia coli by the polymerase chain reaction. J. Clin. Microbiol. 28, 540±545. Pradel, N., Livrelli, V., De Champs, C., Palcoux, J.B., Reynaud, A., Scheutz, F., Sirot, J., Joly, B., Forestier, C., 2000. Prevalence and characterization of Shiga toxin-producing Escherichia coli isolated from cattle, food and children during a one-year prospective study in France. J. Clin. Microbiol. 38, 1023±1031. Ramachandran, V., Hornitzky, M.A., Bettelheim, K.A., Walker, M.J., Djordjevic, S.P., 2001. The common ovine Shiga toxin 2-containing Escherichia coli serotypes and human isolates of the same serotypes possess a Stx2d toxin type. J. Clin. Microbiol. 39, 1932±1937. Rivas, M., Balbi, L., Miliwebsky, E.S., Garcia, B., Tous, M., Leardini, N., Prieto, M., Chillemi, G.M., Principi, M., 1998. SõÂndrome UreÂmico HemolõÂtico en ninÄos de Mendoza, Argentina: su asociacioÂn con la infeccioÂn por Escherichia coli productor de toxina Shiga. Medicina (Buenos Aires) 58, 1±7.

B.E.C. Guth et al. / Veterinary Microbiology 92 (2003) 335±349

349

Saari, M., Cheasty, T., Leino, K., Siitonen, A., 2001. Phage types and genotypes of Shiga toxin-producing Escherichia coli O157 in Finland. J. Clin. Microbiol. 39, 1140±1143. Sanger, F., Nicklen, S., Coulson, A.R., 1977. DNA sequencing with chain terminating inhibitors. Proc. Natl. Acad. Sci. 74, 5463±5467. Schmidt, H., Beutin, L., Karch, H., 1995. Molecular analysis of the plasmid-encoded hemolysin of Escherichia coli O157:H7 strain EDL 933. Infect. Immun. 63, 1055±1061. Strockbine, N.E., Sowers, E., Greene, K., Hayes, P., Griffin, P., Wells, J., 1997. Characterization of Shiga toxinproducing non-O157 Escherichia coli from the United States, 1983±1997. In: Abstracts of the Third International Symposium and Workshop on Shiga Toxin (Verocytotoxin)-Producing Escherichia coli Infections. p. 67. Tenover, F.C., Arbeit, R.D., Goering, R.V., Mickelsen, P.A., Murria, B.E., Persing, D.H., Swaminathan, B., 1995. Interpreting chromosomal DNA restriction patterns produced by pulsed-field gel electrophoresis: criteria for bacterial strain typing. J. Clin. Microbol. 33, 2233±2239. Thomas, A., Smith, H.R., Willshaw, G.A., Rowe, B., 1991. Non-radioactively labeled polynucleotide and oligonucleotide DNA probes, for selectively detecting Escherichia coli strains producing Vero cytotoxins VT1, VT2 and VT2 variant. Mol. Cell. Probes 5, 129±135. Tyler, S.D., Johnson, W.M., Lior, H., Wang, G., Rozee, K.R., 1991. Identification of Verotoxin type 2 variant B subunit genes in Escherichia coli by the polymerase chain reaction and restriction fragment length polymorphism analysis. J. Clin. Microbiol. 29, 1339±1343. World Health Organization (WHO), 1997. Prevention and control of enterohaemorrhagic Escherichia coli (EHEC) infections. In: Proceedings of the Report of a WHO Consultation. World Health Organization, Geneva. World Health Organization (WHO), 1998. Zoonotic non-O157 Shiga toxin-producing Escherichia coli (STEC). In: Proceedings of the Report of a WHO Scientific Working Group Meeting. World Health Organization, Geneva.