veterinary microbiology

ELSEVIER Veterinary Microbiology 59 (1998) 183-192

Distribution and characterization of faecal necrotoxigenic Escherichia cdi CNFl ’ and

CNF2+ isolated from healthy cows and calves

Miguel Blanco, Jestis E. Blanco, Azucena Mora, Jorge Blanc0 *

Reference Laboratory of E. coli, Department of Microbiology and Parasitology. School of Veterinary Science,

Uniuersil?: of Santiago de Compostela. 27002 Lugo. Spain

Received 15 November 1996; accepted 16 September 1997

Abstract

Faecal swabs obtained from a random sample of 268 cows and 90 calves on 19 Lugo (northwestern Spain) farms were examined for necrotoxigenic Escherichia coli (NTEC) producing the cytotoxic necrotizing factors type 1 (CNFl) and type 2 (CNF2). We found NTEC CNFI + and CNF2+ on 11% and 95% of the farms, respectively. NTEC producing CNF2 were significantly more frequently isolated from calves (58%) than from cows (17%) (P < 0.001). The proportion of animals colonized with CNF2+ strains on each farm ranged from 0% to 60%. NTEC strains producing CNF2 isolated from healthy cattle belonged to 27 0 serogroups; however, 64% were of one of 12 serogroups (02, 08, 08-075, 014, 015, 055, 086, 088, 0115, 0121, 0147, and 0168). Furthermore, the serogroups determined in CNF2+ strains isolated from cows (02, 08, and 014) were different from those found in NTEC producing CNF2 isolated from calves (08-075, 015, 055, 086, 088, 0115 and 0147). 0 1998 Elsevier Science B.V.

Keywords: Escherichia coli; Cytotoxic necrotizing factors: Toxins; Cattle

1. Introduction

Necrotoxigenic Escherichia coli (NTEC) strains are able to elaborate two types of cytotoxic necrotizing factors (CNFl and CNF2) that induce cell multinucleation in Vero and HeLa tissue cultures, cause necrosis in rabbit skin and are lethal to mice (De Rycke et al., 199Oa). Both necrotic toxins are cell-associated products easy to detect when

* Corresponding author. Tel.: + 34-82-252231; fax: + 34-82-252195; e-mail: jba@lugo.usc.es

0378-l 135/98/$19.00 0 1998 Elsevier Science B.V. All rights reserved.

PII SO378-1 135(97)00191-O

184 M. Blanco et al. / Veteritmy Microbiology 59 (199S) 183-192

bacteria are sonicated or when grown in the presence of mitomycin C (Blanc0 et al., 1990b). CNFl and CNF2 are heat-labile monomeric proteins of 115 and 110 KDa, respectively, that are immunologically related (Caprioli et al., 1983; De Rycke et al., 1989; Oswald and De Rycke, 1990). Whereas CNFl is encoded by chromosomic genes (Falbo et al., 1993), CNF2 is encoded by transmissible plasmids (Oswald and De Rycke, 1990). The genes encoding CNFl and CNF2 have been recently cloned and sequenced, and showed a very similar (85, 7% of homology) nucleotide sequence (Falbo et al., 1993; Oswald et al., 1994): but CNFl and CNF2 sequences can be distinguished using specific primers by the polymerase chain reaction (PCR) (Blanc0 et al., 1996b). NTEC strains producing CNFl and CNF2 belong to different serogroups and serotypes (Blanc0 et al., 1992a, 1994a,b).

The role of CNFl and CNF2 in the pathogenesis of E. cob infections is still uncertain, but the necrotic and lethal properties in vivo (De Rycke et al., 1990a) and the ability to alter the cytoskeleton (Oswald et al., 1994) of cultured cells in vitro suggest that CNFl and CNF2 may be true virulence factors. Furthermore, experimental infec- tions in neonatal calves and pigs with CNFl- or CNF2-producing strains indicate that CNFl and CNF2 contribute to the pathogenesis of colibacillosis (De Rycke and Plassiart, 1990b; Wray et al., 1993). NTEC strains producing CNFI were associated with human extraintestinal infections (urinary tract infections and sepsis) (Caprioli et al., 1989; Blanc0 et al., 1990a, 1992b, 1996a) and with intestinal or extraintestinal infections of cats and dogs (Pohl et al., 1993a; Pohl et al., 1993b; Prada et al., 1991), whereas CNF2-positive strains were isolated from calves with diarrhoea or septicaemia (De Rycke et al., 1987; Blanc0 et al., 1988; Oswald et al., 1991; Pohl et al., 1993a). However, we have recently found that NTEC strains can be isolated from faeces of a representative percentage of healthy children (CNFl+ strains), cats (CNFl+ strains), dogs (CNF 1 + strains) and calves (CNF2+ strains); consequently, the association of NTEC strains with human and animal diarrhoea must be carefully investigated (Blanc0 et al., 1993a,b; Blanc0 et al.. unpublished results).

The epidemiology of the bovine NTEC remains unclear. Thus, to date, the prevalence of NTEC infection in adult cattle was only investigated in one study (Bums et al., 1996) and all surveys have expressed the prevalence of infection as the proportion of infected animals in the population. In this study the prevalence of NTEC infection in healthy cattle in the province of Lugo (Galicia, northwestern Spain) was investigated at the herd and individual animal levels. Furthermore, the 0 serogroups of NTEC strains isolated from calves were compared with those found among the strains from cows.

2. Materials and methods

2.1. Specimen collection and E. coli strains

All 19 farms in the province of Lugo (Galicia, northwestern Spain) were sampled once between March and October of 1993. Only healthy animals, both beef and dairy cattle, were examined. A single faecal swab was obtained from each animal. All calves aged 4 months or less and a representative percentage (> 50%) of the cows being on

M. Blanc0 et al. / Veterinary Microbiology 59 (1998) 183-192 185

each farm were sampled. Faecal swabs were placed in transport medium and taken to the laboratory for immediate processing. They were plated on lactose-MacConkey agar and on cefixime tellurite sorbitol MacConkey (CT-SMAC) medium, and from each sample 10 E. co/i colonies (lactose positive, lactose negative or sorbitol negative) were chosen and examined for toxin production in Vero and HeLa cells. Identification of E. coli was performed by standard biochemical tests, including hydrogen sulphide, citrate, urease and indole. All NTEC strains were confirmed as E. coli using the API-20E system (bioMCrieux, France). In animals in which all original isolates were identical with respect to the toxic genotype and 0 serogroup, only one colony was selected as strain. When one cow or calf yielded colonies with different toxic genotypes or 0 serogroups, one of each was selected as strain. Reference E. coli strains used as positive and negative controls were: FVL4 (075:Kl:H7, CNFl+), CCB21 (054:K?:H21, CNF2+) and K12-185 (nontoxigenic) (Mouriho et al., 1996). Strains were stored at room temperature in nutrient broth with 0.75% of agar.

2.2. Production and detection ef CNFI and CNF2 in Vero and HeL.a cells

For production of toxins, bacteria were grown for 20 h in tryptone soya broth with mitomicin C at 37°C (shaken at 200 rpm) and then centrifuged (6000 X g) for 30 min at 4°C. The Vero and HeLa cell culture assays were performed using nearly confluent cell monolayers grown in plates with 24 wells. At the time of assay, the growth medium (RPM1 with polymixin sulphate) was changed (0.5 ml per well) and 75 ~1 undiluted culture supematant added. Cells were incubated at 37°C in a 5% CO, atmosphere and the morphological changes in cells observed after 24 h and 48 h of incubation using a phase contrast inverted microscope (Blanc0 et al., 1990b).

2.3. Detection of CNFI and CNF2 genes by PCR

All NTEC strains detected on Vero and HeLa cells were tested for CNF-genes by PCR. DNA to be amplified was released from whole cells by boiling. Bacteria were harvested from CFA agar, suspended in 200 ~1 of sterile water, incubated at 100°C for 10 min, and centrifuged. The supematant was used in the PCR reaction as described below. Base sequences, locations and predicted sizes of amplified products for the specific oligonucleotide primers used in this study are shown in Table 1. Oligonu- cleotide primers were synthesized by using a Gene Assembler Special (Pharmacia, LKB

Table 1

Primers used in PCR to amplify specific fragments from genes for CNFl and CNF2

Primer” Oligonucleotide sequence (5’ -3’) Location within geneh Size of amplified product (bp)

CNFl-A GAACTTATTAAGGATAGT 1585-1602 CNF 1 -B CATTATITATAACGCTG 2112-2128 543 CNF2-A AATCTAATTAAAGAGAAC 841-858 CNF?-B CATGCTTTGTATATCTA 1368-1384 543

“Reference: (Blanc0 et al., 1996b). bin nucleotides.

186 M. Blanco et al./ Veterinary Microbiology 59 (1998) 183-192

Biotechnology) according to the protocol provided by the manufacturer. Amplification of bacterial DNA was performed with 50 ~1 volumes containing 10 ~1 of the prepared sample supernatant; the oligonucleotide primers (450 ng for CNFl primers and 900 ng for CNF2 primers); 0.2 mM (each) dATP, dGTP, dCTP, and dTTP; 10 mM Tris-HCI (pH 8.8); I, 5 mM MgCl,; 50 mM KC]; and I U of DynaZyme DNA polymerase (Finnzymes OY, Finland). The reaction mixtures were overlaid with an equal volume of mineral oil. The PCR was performed with a thermal cycler (model Gene ATAQ Controller; Pharmacia, LKB Biotechnology) at 94°C for 2 min for 1 cycle followed by 30 cycles of 94°C for 1 min, 50°C for 1 min, and 72°C for 1 min. The amplified product was visualized by standard submarine gel electrophoresis using 10 ~1 of the final reaction mixture on a 2% agarose (agarose MP, Boehringer-Mannheim, Germany) gels in TBE buffer (89 mM Tris, 89 mM boric acid, 2.5 mM EDTA). The samples were electrophoresed for 40 min at 140 v. Amplified DNA fragments of specific sizes were located by UV fluorescence after staining with ethidium bromide (0.5 pg/ml). Molecu- lar size markers (HaeIII digest of 4 X 174 DNA) were included in each gel (Blanc0 et al., 1996b).

2.4. Serotyping

The serotyping of NTEC strains was carried out by the method described by GuinCe et al. (1972) and modified by us (Blanc0 et al., 1992a), employing all available 0 (01-017 1) antisera. All antisera were obtained and absorbed with the corresponding cross-reacting antigens to remove the nonspecific agglutinins.

2.5, Statistical methods

Results were compared by the x2 test with Yates’ correction for continuity.

3. Results

3.1. Prevalence of NTEC in healthy cattle

A total of 3093 E. coli colonies from 358 healthy animals from 19 farms were investigated for production of CNFI and CNF2 necrotic toxins in Vero and HeLa cells. NTEC producing CNF2 were recovered from 45 (17%) of 268 cows and from 52 (58%) of 90 calves (P < 0.001) (Table 2) from 18 (95%) of 19 farms examined (Table 3). The estimated proportion of animals infected with CNF2+ strains on each farm ranged from

Table 2

Prevalence of NTEC strains in healthy cattle

Age group No. of animals No. of animals with NTEC strains

Total CNFl+ CNF? +

COWS 268 47 (18%) 2 (1%) 45 (17%)

Calves 90 53 (59%) I (1%) 52 (58%)

Table 3

M. Blanco et al./Veterinary Microbiology 59 (19981 183-192 187

Prevalence of CNF2+ strains in the 19 farms investigated

Farm No. of CNF2+ /No. of animals Serogroups of CNF2+ strains

cows 1 2/16 (12%)

Calves

7/12 (58%)

2 o/7 (0%) 4/5 (80%)

3 4/12 (33%) 4/6 (67%)

4 l/13 (8%) O/l (0%)

5 o/15 (0%) 9/l 1(82%)

6 3/9 (33%) 7 8/21 (38%)

8 9/31 (29%) 9 l/5 (20%)

10 2/16 (12%) 11 5/24 (21%) 12 o/9 (0%) 13 2/13 (15%) 14 o/3 (0%)

15 l/7 (14%) 16 o/31 (0%)

17 o/7 (0%) 18 o/9 (0%) 19 7/20 (35%)

2/4 (50%)

3/5 (60%)

2/7 (29%)

5/5 (100%)

O/l (0%)

3/5 (60%)

l/2 (50%)

2/3 (67%)

o/2 (0%)

l/3 (33%)

3/6 (50%)

2/4 (50%)

2/4 (50%)

2/4 (50%)

02 (3)a. 08 (1). 08-075 (2), 015 (2) 039 (1).

049 (2). 086 (1). 088 (1) 014 (I), 015 (l), 055 (1). NTb (3) 05 (1). 08 (4), 015 (1). 086 (1). 0120 (1). 0153 (2X NT (4)

NT(l) 02 (2), 08 (1). 08-075 (l), 055 (21,064 (1).

088 (2). 0128 (11, NT (3)

02 (3). 08 (1). OS-075 (11, 015 (I), 086 (2)

02 (2), 014 (8),0123 (I), NT (3) 02 (5), 012 (2), 025 (1). 0121 (3). 0141 (I), NT (3)

02 (2). 08 ( l), 08-075 (6), 0 115 (4). 0147 (3)

02 (2) 01 (l), 02 (3). 08-075 (1). 0123 (l), 0147 (3)

NT (1) 08 (2). 088 (2)

OZ(l),NT(l)

098 (1). 0168 (2). NT (1)

092 (1). 0121 (2). NT (3)

088 (1). NT(l)

0162 (2). 0168 (11, NT(6)

aNumber of animals yielding CNF2+ strains belonging to indicated serogroups.

bNot typeable strains.

0% (in farm no. 14) to 60% (in farm no. 9) (Table 3). NTEC strains producing CNFl were isolated from only three (0.8%) of 358 sampled animals (Table 2). The toxin genotypes determined by PCR corresponded to the phenotypic results of cell assays. A total correlation between phenotypic and genotypic toxin detection was observed.

3.2. Serogroups of NTEC

A total of 139 CNF2+ strains were identified in this study, 51 from cows and 88 from calves. NTEC strains producing CNF2 belonged to 27 different 0 serogroups; however, 64% (89 of 139) were among 12 serogroups (02, 08, 08-075, 014, 01.5, 055, 086, 088, 0115, 0121, 0147 and 0168). In 27 (28%) of 97 animals with NTEC strains producing CNF2, two (16 animals), three (nine animals) or four (two animals) serogroups were identified. Furthermore, in many farms a variety of CNF2+ strains with different 0 serogroups were detected (Table 3). The serogroups determined in CNF2+ strains isolated from cows were different from those found in NTEC producing CNF2 isolated from calves (Table 4). Thus, the three serogroups (02, 08 and 014) most frequently detected in CNF2+ strains isolated from cows (59%; 30 of 51) were found in only 12 (14%) of 88 CNF2+ strains from calves (P < 0.001). In contrast, the serogroups 08-075, 015, 055, 086, 088, 0115 and 0147 were more commonly detected among

188 M. Bhco et al./ Veterinap Microbiology 59 (1998) 183-192

Table 4

Comparison between the 0 serogroups of CNF2 + strains isolated from healthy cows and calves

Serogroup No. of CNF2+ strains from

cows Calves

01 02

05

08

08-075

012

014

015

025

039

049

055

064

086

088

092

098

0115

013-O

0121

0113

0128

0141

0147

0153

0162

0168

NT”

0

0

0

0

0

0

0

0

0

0

3 0

0

0

0

2

8

2 2

‘Not typeable strains.

CNF2+ strains isolated from calves (39%: 34 of 88) than from cows (10%; five of 51) (P < 0.001) (Table 4).

The three NTEC strains producing CNFl belonged to serogroups 02 (one strain) and 04 (two strains).

4. Discussion

The results of this study indicate that NTEC colonization caused by CNF2-producing strains is widespread among healthy cattle in the province of Lugo because we found CNF2+ strains on 95% of the farms examined. Furthermore, the presence of CNF2+ strains was significantly more frequent in the faeces from calves (58%) than from cows (17%). Previously, between 1980 and 1983, we examined 78 calves with diarrhoea in the province of La Coruha (northwestern Spain) and we detected CNF2-producing

M. Blanco et al. / Veterinary Microbio1og.v 59 (19981 183-192 189

strains in 13 (17%) animals (Blanco et al., 1988). In that study we did not use a control group of healthy calves, and thus, we did not know whether NTEC were really associated with diarrhoea or formed a part of the normal intestinal flora. Later, in a second study realized in the province of Lugo during 1986 to 1991, we evaluated the prevalence of NTEC strains in ill and healthy calves, and we concluded that CNF2+ strains form part of the normal intestinal flora of cattle for they were more frequently isolated from healthy calves (34%; 38 of 112) than from calves with diarrhoea (20%; 39 of 197) (Blanc0 et al., 1993a). Our observations agree with those made by Bums et al. (1996) in Northern Ireland. The percentage of CNF2-producing strains isolated from normal samples (3 1%) was higher than that obtained from the diarrhoeic samples (19%). Furthermore, there was also a higher percentage of CNF2-producing E. coli strains isolated from healthy animals under one year of age (39%) than from older animals (12%). Therefore, our results and those obtained by Bums et al. (1996) indicate that younger animals are more susceptible to be colonized by CNF2-producing strains.

The 123 bovine CNF2-producing strains identified in the two surveys (Blanc0 et al., 1988, 1993b) previously realized in Galicia belonged to 28 different 0 serogroups (01, 02, 03, 07, 08, 015, 039, 043, 054, 055, 076, 078, 086, 088, 092, 0109, 0113, 0115, 0117,0123,0125, 0126,0128,0132,0139,0147,0153 and 0168); however, only six of them (01, 03, 015, 055, 088 and 0123) accounted for 60% of strains (Blanco et al., 1992a). In the present study, CNF2+ strains demonstrated considerable heterogeneity in their O-groups, because the 139 CNF2-producing strains identified belonged to 27 different serogroups, including 12 serogroups (05, 08-075, 012, 014, 025, 049, 064, 098, 0120, 0121, 0141 and 0162) not previously described among CNF2-producing strains. The most frequent serogroup determined among CNF2’ strains in the two previous studies performed in Galicia was 0123 (26%; 32 of 123). Unexpectedly, the 0123 serogroup was found only in two (1%) of 139 CNF2-producing strains identified in the present study. Furthermore, among the 12 serogroups (02, 08, 08-075, 014, 015, 055, 086, 088, 0115, 0121, 0147 and 0168) most frequently found in this study, only three (015, 055 and 088) of six serogroups most commonly determined in the two previous studies performed in Galicia are included. Bovine CNF2-producing strains isolated by Bums et al. (1996) in Ireland belonged to 19 different 0 serogroups, however, over 50% of these were of one of four serogroups: 03, 015, 026 and 0101. The remaining isolates belonged to groups 02, 06, 08, 011, 020, 045, 064,086, 0115,0137, 0153 and 0157. Therefore, among bovine CNF2+ strains a wide variety of serogroups exists. In fact, in the present study an important variety of CNF2-producing strains with different 0 serogroups were detected in many farms. Interestingly, the serogroups determined in CNF2+ strains isolated from cows (02, 08 and 014) were different from those found in CNF2-producing isolated from calves (08-075, 015, 055, 086, 088, 0115 and 0147). Future studies are necessary to know if the strains belonging to these two groups of strains express different types of adhesins that justify their different age-host specificity.

The comparative numbers of CNFl+ and CNF2+ strains isolated also support other findings which have demonstrated a higher prevalence of CNF2-producing strains than CNFl+ strains in cattle (Blanc0 et al., 1988, 1993a; Bums et al., 1996). CNFl+ strains are rarely identified in cattle: 1% in the three studies of Galicia and 4% in the survey of

190 M. Blanco et al./ Veterinary Microbiology 59 (1998) 183-192

Ireland. The three bovine CNFl+ strains detected in this study belonged to two 0 serogroups (02 and 04) typical of human strains that produce CNFl necrotic factor (Blanc0 et al., 1992a, 1994a).

The isolation of CNF2-producing strains from normal faeces, similar to the findings of verotoxin-producing E. coli (VTEC) in cattle faeces (to be exact, VTEC were recovered from 23 [37%] of 90 calves and from 93 [35%] of 268 cows) (Blanco et al., 1997), suggests either that the endemic colonization of CNF2+ strains among the cattle population results in a high natural immunity to CNF2 necrotic factor in the majority of animals, or, that the production of CNF2 alone may not be sufficient to cause enteric disease (Bums et al., 1996). Although, significant knowledge of the biological and molecular properties of CNFl and CNF2 toxins has accumulated over the years (Oswald et al., 1994; Fiorentini et al., 1995; De Rycke et al., 1996), there is as yet only indirect evidence to substantiate the role of these toxins as true virulence factors. The available evidence is first based on their potent toxic properties, including necrotoxicity in the rabbit skin test and lethality for mice and for young ruminants after parenteral administration (De Rycke et al., 1990a; De Rycke and Plassiart, 1990b). However, a major obstacle to the comprehension of the role of CNFl and CNF2 in the virulence of NTEC is the fact that, in vitro at least, the necrotic factors remains associated with the bacteria and is not detectably released into the culture medium (De Rycke et al., 1990a).

5. Conclusion

The results of this study show that NTEC strains that produce CNF2 are frequently isolated from faeces of healthy cattle, the younger animals being more susceptible to colonization by this type of toxigenic E. coli. A considerable heterogeneity in their 0 serogroups exists even among CNF2-producing strains isolated in the same farm.

Acknowledgements

This study was supported by grants from the Fondo de Investigaci6n Sanitaria (FIS 94/1056), from the Comision Interministerial de Ciencia y Tecnologia (AGF92-0570), and from the Xunta de Galicia (XUGA 26105B97 and 26106B97). J.E.B. and A.M. acknowledge the DGICYT of the Spanish Ministerio de Education y Ciencia, Xunta de Galicia and the Universidad de Santiago de Compostela for research fellowships. We also thank Cristina Prado, Marta Rio and Lucia Femandez for skillful technical assistance.

References

Blanco, .I., G~nzBlez, E.A.. Garcia, S., Blanco, M., Regueiro. B., Berrkdez, I., 1988. Production of toxins by

Escherichia coli strains isolated from calves with diarrhoea in Galicia (Northwestern Spain). Vet.

Microbial. 18, 297-311.

Blanco, J.. Alonso, M.P., GonzBlez. E.A., Blanco, M., Garabal. J.I.. 1990a. Virulence factors of bacteraemic

M. Blanco et al. / Veterinav Microbiology 59 (1998) 183-192 191

Escherichia coli with particular reference to production of cytotoxic necrotising factor (CNF) by

P-fimbriate strains. J. Med. Microbial. 31, 175-183.

Blanco, J., Blanco, M., Gonzalez, E.A., Alonso, M.P., Garabal, J.I., 1990b. Comparative evaluation of three

tests for the detection of Escherichia coli cytotoxic necrotizing factors (CNFl and CNF2) using filtrates of

cultures treated with mitomycin C. FEMS Microbial. Lett. 69, 3 11-3 16.

Blanco, J., Blanco, M., Alonso, M.P., Blanco, J.E., Garabal, J.I., Gonzalez, E.A., 1992a. Serogroups of

Escherichia coli strains producing cytotoxic necrotizing factors CNFl and CNF2. FEMS Microbial. Lett.

96, 155-160.

Blanco, J., Blanco, M., Alonso, M.P.. Blanco, J.E., Gonzalez. E.A., Garabal, J.I.. 1992b. Characteristics of

haemolytic Escherichia coli with particular reference to production of cytotoxic necrotizing factor type 1

(CNFl). Res. Microbial. 143, 869-878.

Blanco, M., Blanco, J., Blanco, J.E., Ramos, J., 1993a. Enterotoxigenic, verotoxigenic and necrotoxigenic

Escherichia coli isolated from cattle in Spain. Am. J. Vet. Res. 54, 1446-1451.

Blanco, J., Blanco, M., Wong, I.. Blanco, J.E., 1993b. Haemolytic Escherichia coli strains isolated from

stools of healthy cats produce cytotoxic necrotizing factor type 1 (CNFl). Vet. Microbial. 38, 157-165.

Blanco, J.E., Blanco, J., Blanco, M., Alonso, M.P., Jansen, W.H., 1994a. Serotypes of CNFl-producing

Escherichia coli strains that cause extraintestinal infections in humans. Eur. J. Epidemiol. 10, 707-711.

Blanco, M., Blanco, J.E., Blanco, J., Verbruggen, A., Jansen, W.H., 1994b. Serotypes of bovine Escherichia coli producing cytotoxic necrotizing factor type 2 (CNF2). Vet. Microbial. 39, 83-88.

Blanco, M., Blanco, J.E., Alonso, M.P., Blanco, J., 1996a. Virulence factors and 0 groups of Escherichia coli from patients with acute pyelonephritis, cystitis and asymptomatic bacteriuria. Eur. J. Epidemiol. 12,

191-198.

Blanco, M., Blanco, J.E., Blanco, J., Alonso, M.P., Balsalobre, C., Mouriiio, M., Madrid, C., Jutiez, A.,

1996b. Polymerase chain reaction for detection of Escherichia coli strains producing cytotoxic necrotizing

factor type 1 and 2 (CNFl and CNF2). J. Microbial. Meth. 26, 95-101.

Blanco. M., Blanco, J.E., Blanco, J., Mora, A., Prado, C., Alonso, M.P.. Mouriiio, M.. Madrid, C., Balsalobre, C., Juarez, A.. 1997. Distribution and characterization of faecal verotoxin-producing Escherichia coli (VTEC) isolated from healthy cattle. Vet. Microbial. 54, 309-319.

Bums, A.L., Ball. H.J., Finlay, D.A., 1996. CNF producing Escherichia coli isolated from cattle in Northern Ireland. Vet. Microbial. 49. 235-241.

Caprioli. A., Falbo, V., Roda, L.G.. Ruggeri, F.M., Zona. C.. 1983. Partial purification and characterization of an Escherichia coli toxic factor that induces morphological cell alterations. Infect. Immun. 39, 13Ot-1306.

Caprioli, A., Falbo. V.. Ruggeri, F.M., Minelli, F., Brskov, I., Donelli, G., 1989. Relationship between

cytotoxic necrotizing factor production and serotypc in haemolytic Escherichia coli. I. Clin. Microbial. 27, 758-761.

De Rycke, J., Guillot, J.F., Boivin, R., 1987. Cytotoxins in non-enterotoxigenic strains of Escherichiu coli isolated from feces of diarrhoeic calves. Vet. Microbial. 15, 137-150.

De Rycke, J., Phan-Thanh, L., Bernard. S., 1989. Immunochemical identification and biological characteriza- tion of cytotoxic necrotizing factor from Escherichia coli. J. Clin. Microbial. 27, 983-988.

De Rycke, J., Gonzalez, E.A., Blanco. J., Oswald, E., Blanco, M., Boivin, R., 1990a. Evidence for two types of cytotoxic necrotizing factor in human and animal clinical isolates of Escherichia coli. J. Clin. Microbial. 28, 694-699.

De Rycke, J., Plassiart, G., 1990b. Toxic effects for lambs of cytotoxic necrotising factor from Escherichia coli. Res. Vet. Sci. 49, 349-354.

De Rycke. J., Mazars, P., Nougayrede. J.P., Tasca. C., Boury, M., Herault, F., Valette, A., Oswald, E.. 1996. Mitotic block and delayed lethality in HeLa epithelial cells exposed to Escherichia coli BM2-1 producing

cytotoxic necrotizing factor type 1. Infect. Immun. 64, 1694-1705.

Falbo, V., Pace, T.. Picci, L., Pizzi. E., Caprioli, A., 1993. Isolation and nucleotide sequence of the gene encoding cytotoxic necrotizing factor 1 of Escherichiu coli. Infect. Immun. 61, 4909-4914.

Fiorentini. C.. Donelli, G.. Matarrese. P., Fabbri. A., Paradisi, S., Boquet, P., 1995. Escherichia coli cytotoxic necrotizing factor: 1: Evidence for induction of actin assembly by constitutive activation of the p21 Rho GTPase. Infect. Immun. 63, 3936-3944.

Guinte, P.A.M., Agterberg, C.M., Jansen, W.H., 1972. Escherichia coli 0 antigen typing by means of a mechanized microtechnique. Appl. Microbial. 24, 127- 13 1,

192 M. Blanco et al. / Veterinqv Microbiology 59 (1998) i83- 192

Mourifio, M., Madrid, C., Balsalobre. C., Prenafeta, A., Muhoa, F., Blanco, J., Blanco, M., Blanco. J.E., Ju&reez, A., 1996. The Hha protein as a modulator of the expression of virulence factors in Escherichiu co/i. Infect. Immun. 64. 2881-2884.

Oswald, E., De Rycke, J., 1990. A single protein of 110 KDa is associated with the multinucleating and

necrotizing activity coded by the Vir plasmid of Escherichia co/i. FEMS Microbial. Lett. 68. 279-284.

Oswald, E., De Rycke. J., Lintermans. P.. Van Muylem, K., Mainil, J., Daube, G., Pohl, P., 1991. Virulence

factors associated with cytotoxic necrotizing factor type 2 in bovine diarrhoeic and septicemic strains of Escherichia coli. J. Clin. Microbial. 29, 2522-2527.

Oswald. E.. Sugai, M., Labigne, A., Wu, H.C., Fiorentini. C.. Boquet, P., O’Brien, A.D., 1994. Cytotoxic necrotizing factor type 2 produced by virulent Escherichia coli modifies the small GTP-binding proteins

Rho involved in assembly of actin stress fibers. Proc. Natl. Acad. Sci. U.S.A. 91, 3814-3818.

Pohl, P., Oswald, E., Van Muylem, K., Jacquemin. E., Lintermans, P., Mainil, J., 1993a. Escherichia cdi producing CNFl and CNF2 cytotoxins in animals with different disorders. Vet. Res. 24, 31 l-315.

Pohl, P.. Mainil, J., Devriese, L.A., Haesebrouck, F., Broes, A., Lintermans, P.. Oswald, E., 1993b.

Escherichia coli productrices de la toxine cytotoxique necrosante de type I (CNFI) isoltes a partir de

processus pathologiques chez des chats et des chiens. Ann. Med. Vet. 137, 21-25.

Prada, J., Baljer. G.. De Rycke. J., Steinriick, H., Zimmermannn, S.. Stephan, R., Beutin, L.. 1991.

Characteristics of a-haemolytic strains of Escherichia coli isolated from dogs with gastroenteritis. Vet.

Microbial. 29. 59-73.

Wray, C.. Piercy, D.W.T., Carroll, P.J., Cooley, W.A.. 1993. Experimental infection of neonatal pigs with CNF toxin-producing strains of Escherichia co/i. Res. Vet. Sci. 54, 290-298.