- Top of page
- Materials and methods
The prevalence of Shiga-toxin-producing Escherichia coli (STEC) in healthy dairy ruminants was investigated between 1996 and 1998 by a multiplex polymerase chain reaction (mPCR) technique. A total of 13 552 E. coli colonies from 726 cows, 28 sheep and 93 goats out of 112 randomly selected dairy farms in Hessia, Germany were analysed. STEC strains were recovered from 131 (18·0%) cows, nine (32·1%) sheep and 70 (75·3%) goats. Further characterization of the STEC isolates showed that 89 (0·66% of the investigated colonies) of animal field strains carried stx1 gene, 64 (0·47%) stx2 gene and 57 (0·42%) stx1 and stx2 gene. Sixty (93·8%) out of 64 stx2 field strains were harboured by cows. In contrast, 74 (83·1%) out of 89 stx1 dairy animal field strains were from ovine or caprine origin. Only 17 (8·1%) stx-positive isolates (13 from cattle, three from sheep and only one from goat) were positive for eaeA gene. Eight (9·0%) of the stx1, five (7·8%) of the stx2 and four (7·0%) of the stx1/stx2 gene-positive field strains carried the eaeA gene. The prevalence of EHEC-haemolysin (EHEC-hlyA) gene sequence was 88·8% (79 isolates) of the stx1 and 68·8% (44 isolates) of the stx2 isolates. Out of 57 stx1- and stx2-positive field-strains, 34 (59·6%) carried the EHEC-hlyA gene. E. coli O serovars O:157 and O:111 were not found. Only one isolate was positive with O26 antiserum.
- Top of page
- Materials and methods
Shiga toxin-producing strains of Escherichia coli (STEC) are now recognized as an important human pathogen of public health concern.
Whereas STEC isolates belong to many different serotypes, E. coli O157:H7 and an occasional non-motile variant O157:H– are the most common serotypes associated with human illness. Isolates of this pathogen are a major cause of haemorrhagic colitis and mild diarrhoeal illness and are the major aetiological agent of haemolytic uraemic syndrome (HUS). HUS is characterized by a prodrome of gastroenteritis, frequently including bloody diarrhoea, followed by acute haemolytic anaemia, thrombocytopenia and renal failure. Infections are usually linked to the consumption of STEC-contaminated improperly cooked beef, faeces-contaminated vegetables, apple cider, water and direct transmission of STEC from animals to man ( Griffin and Tauxe 1991). Humans also became infected by the consumption of improperly heated or unpasteurized milk ( Bockemühl et al. 1990 ).
During a milk-borne outbreak of gastroenteritis and HUS caused by E. coli O157:H7 in Ontario, Canada, the organism was recovered from faeces of patients and young animals of dairy herds associated with the outbreak ( Borczyk et al. 1987 ). In Germany O157 STEC were also identified as important pathogens causing HC and HUS ( Bockemühl et al. 1998 ).
Furthermore, even though non-O157 STEC serovars have been associated with HUS, most of the STEC have properties in common, which are known or supposed to be related to their virulence. So STEC strains seem to be pathogenic for humans only if they possess accessory virulence factors. Production of a special type of haemolysin, named EHEC haemolysin, possession of a large 60 MDa plasmid and induction of attaching and effacing lesions in host intestinal epithelial cells were closely related to human pathogenic STEC.
While several STEC serotypes can cause disease in pigs and calves, most are harboured by asymptomatic, healthy animals. Certain types of STEC found in animals, which are designated enterohaemorrhagic E. coli (EHEC) were shown to behave as pathogens in humans ( Levine 1987).
The objective of the present study was to investigate the prevalence of STEC in faeces of healthy dairy ruminants, such as cows, sheep and goats in randomly selected stocks producing raw milk or raw milk products. To determine the isolates' potential as human pathogens, they were characterized for the presence for different virulence genes (stx1, stx2, eaeA and EHEC-hlyA).
Materials and methods
- Top of page
- Materials and methods
A single faecal swab was obtained from 726 randomly selected lactating cows in 103 dairy farms in Hessia, Germany, between April 1996 and August 1998. Additionally, faecal swabs from 28 sheep and 93 goats, also randomly selected in nine dairy farms were investigated.
Swabs were placed in Stuart transport medium (Merck) and taken to the laboratory for immediate processing. At first samples were plated on Gassner agar (Merck) and from each plate incubated for 18–24 h at 37 °C, 16 lactose-positive E. coli-like colonies were chosen and examined for stx genes. Species identification of isolates was performed by standard bacteriological methods.
With respect to the stx-genotype, only one isolate per animal was selected. In animals harbouring colonies with different stx-genotypes one of each was selected.
stx gene detection and investigation of virulence factors was performed with a multiplex polymerase chain reaction (mPCR) method. Briefly, strains were cultivated overnight on sheep blood agar (Merck) at 37 °C. A loop of colony growth was suspended in 100 µl of H2O, boiled for 10 min and centrifuged briefly in a microcentrifuge. Two microlitres of crude cell lysat provided sufficient target DNA for PCR amplification (Perkin Elmer). Each PCR was carried out in 20-µl volumes containing 2 µl 10 × PCR buffer and 0·1 µl (10 mmol l−1) of each desoxynucleoside triphosphate (dNTP), 0·2 µl (KS7-KS8/GK5-GK6), and 0·8 µl (SK1-SK2/Ehly1-Ehly2) primer and 0·32 µl (2 IU) of Taq polymerase (Sigma).
After initial denaturation for 5 min at 94 °C 30 cycles of 30 s at 94 °C, 90 s at 57 °C and 90 s at 72 °C followed. The reaction was finished by incubation for 7 min at 72 °C.
The oligonucleotide primers used for amplification of stx1, stx2, eaeA and EHEC-haemolysin (EHEC-hlyA) gene (MWG Biotech, Ebersberg, Germany) are summarized in Table 1.
Table 1. Primers used for detection of different virulence factors in ruminant STEC by multiplex polymerase chain reaction (mPCR)
|Primer||Virulence factor||PCR-product [length (bp)]||Reference|
|KS7–KS8||Shiga toxin 1 (stx1) ||282||Beutin et al. (1995) |
|GK5–GK6||Shiga toxin 2 (stx2) ||270||Beutin et al. (1995) |
|SK1–SK2||E. coli attaching and effacing gene (eaeA) ||863||Schmidt et al. (1994a) |
|HlyA1–hlyA4 ||Enterohaemolysin plasmid (EHEC hlyA) ||1551||Schmidt et al. (1994b) |
PCR products were loaded into single wells of 2% agarose gel (Roth, Karlsruhe, Germany). After subsequent electrophoresis at 100 V, the gel was stained with ethidium bromide and u.v. analysed (Bio-Rad, Munich, Germany).
stx gene-positive colonies were subcultured on enterohaemolysin agar (Oxoid, Wesel, Germany) for detection of enterohaemolysin phenotype.
Additionally, selected STEC field strains were tested with rabbit antiserum against E. coli serovars O:157, O:111 and O:26 (Sifin, Berlin, Germany) by slide agglutination technique.
- Top of page
- Materials and methods
A total of 13 552 E. coli colonies from 726 healthy cows, 28 healthy sheep and 93 healthy goats from 112 different dairy farms were analysed by mPCR.
Table 2 shows the number and frequencies of stx gene-positive isolates. STEC strains were recovered from 131 (18·0%) cows, nine (32·1%) sheep and 70 (75·3%) goats.
Table 2. Prevalence of STEC in dairy cattle, sheep and goats
| ||Investigated||STEC positive|
| ||No. of farms||No. of animals||No. of farms||No. of animals|
|Cattle *||103||726||51 (49·5%)||131 (18·0%)|
|Sheep †||3||28||3 (100%)||9 (32·1%)|
|Goat †||6||93||6 (100%)||70 (75·3%)|
|Σ||112||847||60 (53·6%)||210 (24·8%)|
The stx genotypes, eaeA gene and EHEC-haemolysin (EHEC hlyA) occurrence is summarized in Table 3.
Table 3. Shiga toxin genotypes, eaeA gene and EHEC-hlyA gene of STEC (n = 210)
| ||No. of strains||Virulence factor (gene)|
|stx1 ||stx2 ||eaeA||EHEC-hlyA|
|Cattle (n = 131) ||4||X|| || || |
|4||X|| || ||X|
|39|| ||X|| ||X|
|Sheep (n = 9) ||1||X|
|5||X|| || ||X|
|Goat (n = 70) ||67||X|| || ||X|
Fifteen (11·5%) of the bovine, six (66·7%) of the ovine and 68 (97·1%) of the caprine STEC strains were positive only for stx1. stx2 gene alone was detected in 60 (45·8%) bovine, three (33·3%) ovine and only one (1·4%) caprine faecal isolate. stx1 and stx2 gene in combination was present in 56 (42·7%) bovine, no ovine and only one (1·4%) caprine field strain.
Only 17 (8·1%) of the stx-positive isolates from healthy dairy animals (13 from cattle, three from sheep and only one from goats) were positive for the eaeA gene sequence when tested by mPCR using the universal primer pair SK1/SK2.
EHEC hlyA gene was found in six of the stx1- and 41 of the stx2-positive bovine isolates. The EHEC hlyA gene was detected in 34 bovine field strains with stx1 and stx2 gene sequences. In sheep all of one ST-positive field strain (stx1) showed the EHEC-haemolysin gene. With the exception of only one stx2 strain and one stx1/stx2 strain also in caprine STEC strains EHEC hlyA gene could be detected.
Figure 1 shows typical amplification fragments of stx1, stx2, eaeA gene and EHEC hlyA gene in mPCR.
Figure 1. Agarose gel electrophoresis patterns showing typical amplification fragments in multiplex PCR for stx1, stx2, eaeA and EHEC hlyA gene. M, marker; 1, Aqua destillata; 3, E. coli field strain (stx1, eaeA, EHEC-hlyA); 4, E. coli negative control K12; 5, reference strain E. coli (EDL 933; stx1, stx2, eaeA, EHEC-hlyA); 6, E. coli field strain (stx1, stx2, EHEC-hlyA)
Download figure to PowerPoint
One bovine E. coli-field strain (stx1+, eaeA+, EHEC-hlyA+) was positive in slide agglutination test with O26 antiserum.
None of the investigated E. coli isolates showed positive agglutination with O157 or O111 antiserum.
All EHEC hlyA gene-positive STEC-strains showed a small and turbid zone of haemolysis after 24 h incubation on enterohaemolysin agar (enterohaemolysin phenotype).
- Top of page
- Materials and methods
The aim of this study was to investigate the prevalence of STEC in healthy dairy cattle, sheep and goat. The results of this study indicate that STEC occurrence is widespread among healthy dairy animals in the investigated area. We found STEC strains in 53·6% of the dairy farms and they were present in 24·8% of the examined animals.
Similar results are reported from a study of healthy cattle in Wisconsin and Washington ( Wells et al. 1991 ). Here, STEC was found in 80% of the investigated farms. The prevalence of STEC infections in cows was estimated at 8·4% for adult cows and 19·5% for heifers and calves. In a study of randomly selected cattle in Ontario, Canada, STEC strains were detected in 10·5% of beef cattle, 19·5% of dairy cows and 3·5% of calves ( Wilson et al. 1992 ). In another study of healthy cattle in Germany Montenegro et al. (1990) recovered STEC strains from 17% of dairy cows and 9% of bulls at slaughter. In Spain faecal swabs obtained from a random sample of 268 cows and 90 calves in 19 dairy farms were examined for STEC by Blanco et al. (1997) . They found STEC in 95% of the observed farms. The prevalence of STEC infection in asymptomatic cows was estimated to be 35%. The proportion of animals infected on each farm ranged from 0 to 100%.
The prevalence of bovine Shiga toxin-producing E. coli in Argentina was reported by Sanz et al. (1998) . Twenty-three percent of calves with diarrhoea, 29% of healthy calves, 44% of dairy cows at the slaughterhouse and 22% of grazing cattle carried STEC strains.
Miyao et al. (1998) examined 2152 faecal E. coli isolates from 387 cattle in Japan. The detection rate of STEC was 24·3%.
Only a few data are available concerning STEC prevalence in small ruminants. In Germany Beutin et al. (1993) isolated STEC from sheep (66·6%) and goats (56·1%). Randall et al. (1997) used a monoclonal sandwich ELISA for the detection of Shiga toxigenic E. coli in animal faeces. The STEC prevalence was 63% in sheep and 45% in goats.
Serovar O157 is most commonly found among STEC strains that cause HC and HUS in humans ( Barett et al. 1992 ; Sharp et al. 1994 ). Furthermore, non-O157 STEC serotypes such as O26:H4, O103:H2 and O111:H– have been associated with human diseases ( Beutin et al. 1998 ). In this study only one bovine strain belonged to serovar O:26. None of the isolated STEC strains in this study belonged to serovars O157 or O111. Our results agree with previous reports ( Wells et al. 1991 ; Wilson et al. 1992 ; Beutin et al. 1995 ; Blanco et al. 1997 ) which suggested that STEC O157 are uncommon in asymptomatic animals. Previous studies indicate that STEC serovars O26:H11, O111:H– and O157:H7/H– are sporadically isolated from the faeces of cattle and appear more frequently in ruminants with diarrhoea than in healthy asymptomatic animals. The methods used for isolation of E. coli O157 greatly influence the isolation rates. More sensitive detection methods, such as immunomagnetic separation, have shown that this serovar can be detected at moderate frequency from healthy cattle ( Chapman et al. 1994 ). Using a sensitive ELISA method, Zhao et al. (1995) detected STEC O157:H7 in 22% of investigated control herds, from which E. coli O157:H7 had not been isolated previously. In a Dutch study, using immunomagnetic separation technique, STEC O157 occurred in seven out of 10 dairy farms ( Heuvelink et al. 1998 ).
Kudva et al. (1997) isolated and characterized E. coli O157:H7 and non-O157 STEC strains from sheep. Variation in the occurrence of E. coli O157 was observed in animals being culture-positive only during the summer months. The predominant non-O157 STEC serotype found was O91:NM.
STEC strains seem to be pathogenic for humans only if they possess accessory virulence factors. Another objective of this study was to characterize STEC from healthy dairy ruminants for their virulence markers and thereby for possible relationships with the known human pathogenic types of E. coli. The eaeA gene is responsible for attachment and effacement lesions similar to those in enteropathogenic E. coli ( Gannon et al. 1993 ). In most human STEC strains belonging to enterohaemorrhagic serotypes eaeA genes are present. In this study only 17 (8·1%) of 210 STEC strains isolated from healthy dairy ruminants were positive for the eaeA gene sequence.
The low prevalence of eaeA-positive STEC is also in agreement with previous studies ( Beutin et al. 1993 ; Blanco et al. 1997 ). eaeA gene-positive STEC were found more frequently in cattle with diarrhoea ( Blanco et al. 1997 ). However, recent results indicate that many bovine STEC strains isolated from healthy cattle would not be pathogenic for humans. If the eaeA gene is an essential virulence factor in the pathogenicity of human STEC strains, it is probable that many bovine STEC strains do not influence human health. However, recent results indicate that the eaeA gene is essential for some STEC strains (O26, O103, O111, O157). In contrast O91 and O104 STEC strains were eaeA gene-negative and able to cause HUS or HC. Therefore the pathogenetic role of eaeA gene is still uncertain and the pathogenicity of the eaeA gene-negative STEC from healthy dairy ruminants for humans could not be excluded.
Subtyping of the stx genes of bovine strains isolated in the present study by using stx1- and stx2-specific primers ( Schmitt et al. 1994 ) showed that only 15 strains (11·5%) possessed the stx1 gene, 60 (45·8%) the stx2 gene and 56 (42·7%) both stx1 and stx2. Similar results were obtained by Beutin et al. (1993) and Blanco et al. (1977) . Caprioli et al. (1995) reported that most patients developing HUS were infected with strains harbouring the stx2 gene.
In contrast, Wieler et al. (1992) found a significantly higher percentage of stx1-producing E. coli in diarrhoeic calves, suggesting a pathogenic role in neonatal calve diarrhoea.
stx2 genes were rarely found in our field strains from small ruminants, which is in agreement with results of Beutin et al. (1995) . In contrast, Kudva et al. (1997) detected stx1 and stx2 gene carrying O157:H7 and non-O157:H7 isolates.
Different types of haemolysins have been described for E. coli ( Beutin 1991). Besides the well-characterized α-haemolysin, another type of haemolysin named EHEC haemolysin was shown to be encoded on a large virulence plasmid of EHEC O157 ( Schmidt et al. 1994 ). Its possible clinical significance is pointed out by specific immune response of HUS patients to the EHEC haemolysin.
Using the mPCR technique the EHEC haemolysin was found in 157 STEC isolates. Eighty-one STEC strains of cattle, eight from sheep and 68 from goat showed positive reactions. In this work we could demonstrate that almost all STEC field strains from dairy ruminants showing the EHEC haemolysin sequence were also positive for enterohaemolysin phenotype.
Beutin et al. (1995) found enterohaemolysin phenotype only in 128 out of 208 STEC, including field strains from cattle, sheep and goats.
We found a high prevalence of STEC field strains in the gut of healthy dairy ruminants, pointing to their importance as a reservoir of these toxigenic E. coli. The high incidence of STEC strains found in ruminants poses the question of whether humans are at a risk of acquiring these infections. Faecal contamination of raw milk during milking cannot be totally excluded. In the present study the virulence factor analysis of a large number of STEC field strains reveals a very heterogeneous group of E. coli with different combinations of virulence markers ( Table 3). It was previously reported that the eaeA gene may be required for the expression of full virulence of STEC for humans ( Barett et al. 1992 ). Considering this, only the eaeA-positive strains would present a health hazard for humans. The absence of eaeA gene sequence in most dairy ruminant STEC strains could indicate that these strains are less virulent for humans. However, some strains lacking the eaeA gene sequence such as O113:H21 and O104:H2 seem to be clearly associated with human disease ( Strockbine et al. 1997 ).
To prevent faecal contamination of raw milk for human consumption accurate milking hygiene is absolutely essential.