Download Comparison of different PCR tests for detecting Shiga toxin

Journal of Microbiological Methods 55 (2003) 383 – 392
www.elsevier.com/locate/jmicmeth
Comparison of different PCR tests for detecting Shiga
toxin-producing Escherichia coli O157 and development of
an ELISA-PCR assay for specific identification of the bacteria
Patrick Fach *, Sylvie Perelle, Joël Grout, Francßoise Dilasser
Agence Francßaise de Sécurité Sanitaire des Aliments (AFSSA), Laboratoire d’Etudes et de Recherches sur l’Hygiène et la Qualité des Aliments,
Unité: Atelier de Biotechnologie, 1-5 rue de Belfort, 94700 Maisons-Alfort, France
Received 3 January 2003; received in revised form 29 April 2003; accepted 28 May 2003
Abstract
In an attempt to develop a standard for ELISA – PCR detection of Shiga toxin producing Escherichia coli (STEC) O157, six
published PCR tests were tested in a comparative study on a panel of 277 bacterial strains isolated from foods, animals and
humans. These tests were based on the detection of the genes rfbE [J. Clin. Microbiol. 36 (1998) 1801] and rfbB [Appl. Environ.
Microbiol. 65 (1999) 2954], the 3Vend of the eae gene [Epidemiol. Infect. 112 (1994) 449], the region immediately flanking the 5V
end of the eae gene [Int. J. Food. Microbiol. 32 (1996) 103], the flicH7 gene [J. Clin. Microbiol. 35 (1997) 656], or a part of the
recently described 2634-bp Small Inserted Locus (SILO157 locus) of STEC O157 [J. Appl. Microbiol. 93 (2002) 250]. Unlike the
other PCR assays, those amplifying the rfb sequences were unable to distinguish toxigenic from nontoxigenic O157. These
assays were relatively specific to STEC O157, giving essentially a cross reaction with clonally related E. coli O55 and to a lesser
extent with E. coli O145, O125, O126. They also detected the Shiga toxin (stx)-negative derivatives of STEC O157. Based on
these results, an ELISA-PCR assay consisting of the solution hybridization of amplicons with two probes that ensured the
specificity of the amplification was developed. The ELISA-PCR assay, which used an internal control (IC) of inhibition, was able
to detect 1 to 10 copies of STEC O157 in the PCR tube. Adaptation of PCR into ELISA-PCR assay format facilitates specific and
sensitive detection of PCR amplification products and constitutes a method of choice for screening STEC O157.
D 2003 Elsevier B.V. All rights reserved.
Keywords: Enterohemorrhagic Escherichia coli; O157; ELISA-PCR
1. Introduction
Since its recognition in 1982, enterohemorrhagic
Escherichia coli O157 (EHEC O157) has been recognized as an important human pathogen predominantly
* Corresponding author. Tel.: +33-1-43-76-30-99; fax: +33-143-76-26-30.
E-mail address: [email protected] (P. Fach).
associated with hemorrhagic colitis and the more
severe complications of hemolytic uremic syndrome.
It causes high morbidity and mortality especially
among the young and elderly. In the last few years,
EHEC O157 has been responsible for a number of
well-publicized outbreaks in many parts of the world,
and various methods have been developed to detect the
bacteria. Several selective differentiating media, immunological assays, and DNA probes have been used.
0167-7012/03/$ - see front matter D 2003 Elsevier B.V. All rights reserved.
doi:10.1016/S0167-7012(03)00172-6
384
P. Fach et al. / Journal of Microbiological Methods 55 (2003) 383–392
The cultural methods have exploited the biochemical
characteristics of the O157:H7 strains, which do not
produce h-glucuronidase and usually do not ferment
D-sorbitol rapidly (Thompson et al., 1990; Nataro and
Kaper, 1998). The agar medium most commonly used
to isolate E. coli O157:H7 is sorbitol-MacConkey
(SMAC) agar containing cefixime and tellurite. It
supports stx-producing E. coli O157:H7 (STEC
O157) and inhibits the growth of most of the other E.
coli strains. The disadvantage of these culture methods
is that they do not differentiate between toxigenic and
stx-negative derivatives of STEC O157, found in
patients, animals or subcultures. In addition, they fail
to detect some atypical strains of STEC O157 that may
occur (Gunzer et al., 1992). Most of the immunological
assays available for identification of the bacteria detect
either O and H antigens or Shiga-like toxin. However,
cross-reactivity with organisms other than E. coli O157
has been reported in immunoassays. Anti-O157 sera
may cross-react with Escherichia hermanii, Salmonella O group N, Hafnia alvei, Yersinia enterocolitica or
to a lesser extent with Citrobacter freundii (Bettelheim
et al., 1993; Borczyk et al., 1987; Chart et al., 1992;
Nataro and Kaper, 1998; Shimada et al., 1992) and the
suspected colonies have to be confirmed as E. coli by
biochemical identification. To achieve accurate and
rapid epidemiological investigation of outbreaks, it is
important to be able to fully distinguish EHEC O157
from other bacteria and from bacteria of serogroup
O157 that are not pathogenic. Nucleic acid amplification technologies offer the potential for improved detection of EHEC O157 in the environment
providing greater sensitivity and dramatically speeding
up detection to improve the management of outbreaks
through more-rapid confirmation of the vehicle of
infection.
A number of PCR methods have been reported for
the detection of EHEC O157. Usually, the assays
involve multiplex PCR targeting different genes that
are not specific to EHEC O157 alone (Cebula et al.,
1995; Deng and Fratamico, 1996; Fratamico et al.,
1995; Gannon et al., 1997; Nagano et al., 1998; Paton
and Paton, 1998). These tests are suited to the identification of isolated strains of bacteria and have been
shown to be specific and sensitive for the examination
of complex matrices such as foods and stools (Sharma
and Dean-Nystrom, 2003). However, in a mixture of
bacteria, the genetic profile obtained by multiplex PCR
may arise from more than one strain, hindering interpretation. Ambiguity arises when both stx and rfbO157
genes are detected in a sample containing different
strains of E. coli. In this situation, it is not possible to
ensure that the signal obtained in multiplex PCR is
displayed by STEC O157. The mixture may contain
non-Shiga toxin-producing E. coli O157 (EC O157)
together with STEC from another serogroup. The
additive effect of different strains in multiplex PCR
described by Deng and Fratamico (1996) underlines
the need to find a single specific DNA sequence to
identify EHEC O157. Moreover, one of the major
drawbacks of the published PCR-based approaches
for detection of STEC O157 is lack of internal amplification control (IAC) which is required in order to
monitor false negatives that may be caused by PCR
inhibitors.
Among the more recent PCR techniques used for
detecting STEC O157, real-time PCR using either
SYBR Green I or TaqMan assay is a powerful technique. Real-time PCR techniques provide results available immediately after completion of the amplification
reaction, with no need for any further processing of the
samples and without opening the test tubes, greatly
reducing the risk of carry-over contamination. Another
advantage of real-time PCR methods is that they do not
need ethidium bromide, which is subject to strict and
constraining regulations in many countries, is increasingly restricted in the food industry and less and less
common in routine laboratory use.
In the present study, we evaluated on the same large
collection of bacteria six different nonmultiplexed
PCR systems used for detecting EHEC O157. Based
on these results, an ELISA-PCR assay including a
solution hybridization step was developed for confirmation of PCR amplification products. Furthermore,
an internal control (IC) of inhibition was designed in
order to increase the reliability of the diagnostic
ELISA-PCR assay. Comparing our data with those of
the literature, the ELISA-PCR test yielded results
comparable to those obtained in many other published
E. coli O157 PCR detection systems. There is no
evidence for significant differences in terms of sensitivity and efficiency compared with the real-time
PCR TaqMan assays described in the literature for
E. coli O157 (Oberst et al., 1998; Sharma, 2002).
The threshold sensitivity of both ELISA-PCR and
real-time PCR systems is approximately 102 CFU
P. Fach et al. / Journal of Microbiological Methods 55 (2003) 383–392
ml 1, which corresponds to less than 10 cells in the
PCR tube.
2. Materials and methods
2.1. Bacterial strains and culture conditions
A total of 277 strains were used for the specificity
study. They comprised 67 strains of Shiga toxinproducing E. coli O157 (STEC O157), 93 non-Shiga
toxin-producing E. coli O157 (EC O157), 91 STEC of
other serogroups: O3 (2), O4 (1), O5 (7), O6 (9), O8
(1), O15 (1), O22 (2), O26 (8), O45 (1), O53 (2), O55
(1), O76 (3), O77 (3), O79 (1), O86 (1), O88 (1), O91
(6), O103 (7), O110 (2), O111 (11), O113 (3), O117
(2), O118 (2), O125 (1), O126 (1), O128 (1), O136
(2), O138 (1), O139 (2), O141 (1), O145 (4), O147
(1), 12 non-stx E. coli: O26 (2), O55 (3), O103 (1),
O111 (2), O125 (1), O126 (1), O127 (1), O128 (1);
and 18 strains of other bacterial species. The strains
were obtained from several laboratories (see Acknowledgments) or isolated in our laboratory (Fach et al.,
2001). All the strains were grown in tryptone soy broth
(TSB) and incubated at 37 jC overnight. Enumeration
of STEC O157 strains used to determine the threshold
sensitivity of the ELISA-PCR was performed by
double plating as previously described (Fach et al.,
2001).
2.2. DNA extraction
One milliliter of pure culture of bacteria was placed
in an Eppendorf tube, centrifuged at 12 000 g for 2
min, and the supernatant decanted. The pellet was
washed in 1 ml PBS, pH 7.5 and mixed with 200 Al of
InstaGenek Matrix (Bio-Rad, Marnes-la-Coquette,
France). Bacterial DNA was released by heating the
sample for 15 min at 56 jC and 8 min at 100 jC.
After vortexing and centrifugation at 14 000 g for 2
min, 10 Al of the supernatant was then used in the
PCR reactions.
2.3. Characterization of isolated bacteria
Bacterial strains were identified by the API-20 test
(bioMérieux, Marcy-l’Etoile, France). Overnight cultures of E. coli were transferred to SMAC plates
385
containing sorbitol MacConkey agar (Difco) and incubated for 18 to 24 h at 37 jC to identify sorbitolfermenting activity of clones. The O157 phenotype of
strains was examined by the E. coli O157 Latex Test
from Oxoid, according to the manufacturer’s instructions (Oxoid, Dardilly, France). Other serogroups were
characterized at the Danish Veterinary Laboratory
(Copenhagen, Denmark) by O-typing, using rabbit
E. coli antisera. The Shiga toxin type was identified
by the Vero cell assay (VCA) as described below, and
by specific PCR of the stx1 and stx2 genes as described by Pollard et al. (1990a,b). The presence of
additional virulence factors responsible for attaching/
effacing lesions (eaeA gene), enterohemolysin (hlyA
gene) and catalase-peroxidase (katP gene) production
was determined by previously described PCR (Brunder et al., 1996; Sandhu et al., 1996; Schmidt et al.,
1995).
2.4. Vero cell assay
Detection of Shiga toxin-producing bacteria was
performed using a Vero cell assay (VCA) technique, as
described previously in the literature (Konowalchuk et
al., 1977; Strockbine et al., 1986). Briefly, 1 ml of the
enrichment culture was centrifuged and the pellet was
treated with 2000 U ml 1 of polymyxin B (Sigma,
France) at 37 jC for 45 min. After centrifugation and
filtration of the supernatant through a 0.22-Am membrane filter, the cytotoxicity was evaluated on Vero
cells. Cell viability was observed for 4 days. When
cytotoxicity was found, the Shiga toxin production
was confirmed by a seroneutralization test using the
reference monoclonal antibodies 13C4 (ATCC No.
CRL 1794) (Strockbine et al., 1985) and 11F11 (ATCC
No. CRL 1908) (Perera et al., 1988).
2.5. PCR assays for detecting Shiga toxin-producing
E. coli O157
The six PCR assays amplifying the genes rfbE
(Desmarchelier et al., 1998) and rfbB (Maurer et al.,
1999), the 3Vend of the eae gene (Louie et al., 1994),
the region immediately flanking the 5Vend of the eae
gene (Meng et al., 1996), the flicH7 gene (Gannon et
al., 1997), or a part of the recently described SILO157
locus of STEC O157 (Perelle et al., 2002) were
performed in the GeneAmp PCR System 9700 (Ap-
386
P. Fach et al. / Journal of Microbiological Methods 55 (2003) 383–392
plied Biosystems, Courtaboeuf, France) in a final
volume of 50 Al containing 200 AM of each dNTP,
600 nM of each primer, 5 Al of the GeneAmpR 10 PCR buffer (Applied Biosystems), 2.5 U of AmpliTaq
Goldk (Applied Biosystems) and 10 Al of the
extracted DNA. The PCR cycling conditions used with
the six sets of primers (Table 1) were as follows: an
initial denaturation at 94 jC for 10 min followed by 35
cycles, each consisting of 30 s at 94 jC, 20 s at the
annealing temperature (see Table 1), and 30 s at 72 jC.
After a final elongation step at 72 jC for 10 min, the
PCR products were electrophoresed on 1 – 2% agarose
gels containing ethidium bromide in 1 TBE buffer.
SF6 primer binding regions was cloned in the
pMOSBlue plasmid and formed the synthetic IC.
Simultaneous amplification of these two different
DNA fragments flanked by the same primer sites
results in a competitive PCR, depending on the molar
ratio of those DNA fragments. The number of IC
copies in each PCR tube was limited to 10 copies to
ensure competition advantages for STEC O157. IC
was developed to monitor potential PCR inhibitors and
ensure successful amplification. Owing to the competition of the two PCR reactions, if the SILO157 locus
was amplified but not the IC, then it was assumed that
STEC O157 DNA was present in a proportionally
greater amount. If neither the IC nor the SILO157 locus
was amplified, then it was assumed that inhibition of
the PCR had occurred and the test for that sample was
not valid.
After amplification, the PCR products were detected
in a sandwich hybridization assay as previously described (Fach et al., 2001). This step was performed in
parallel on two microtiter plates; one for the specific
detection of the SILO157 locus and one for the IC
detection. The set of internal capture and detection
probes for detection of the SILO157 locus was 388S
oligonucleotide 5V-end labeled with biotin and 388R
oligonucleotide 3V-end labeled with digoxigenin (Table 1). The set of internal capture and detection probes
for detection of IC was CatCap oligonucleotide 5Vend-labeled with biotin and CatRev oligonucleotide
2.6. ELISA-PCR assay to detect the SILO157 locus
Amplification reactions with the RJD3 and SF6
primers were performed as described above, using 10
copies of a synthetic IC in each reaction. IC was
synthesized as previously described (Fach et al.,
2001), in a one-step PCR reaction with primers
bearing 5Voverhanging ends identical to the published
primers RJD3 and SF6 used in STEC O157 diagnostic
reaction, and 3’ ends complementary to a DNA
sequence of the Cm gene from Tn9. This feature
allowed RJD3 and SF6 primers to co-amplify the
SILO157 locus from STEC O157 and the recombinant
Cm gene sequence in the same reaction. The recombinant Cm gene sequence flanked by the RJD3 and
Table 1
Sequence of primers and probes
Primers and
probes
Sequence (5V– 3V)
Annealing
temperature
Amplicon
size
Target gene
Reference
O157 A-F
O157 A-R
O157 P-F8
O157 P-R8
P1EH
P2EH
FLIC H7-F
FLIC H7-R
SZ I
SZ II
RJD3
SF6
388S
388R
CatCap
CatRev
AAGATTGCGCTGAAGCCTTTG
CATTGGCATCGTGTGGACAG
CGTGATGATGTTGAGTTG
AGATTGGTTGGCATTACTG
AAGCGACTGAGGTCACT
ACGCTGCTCACTAGATGT
GCGCTGTCGAGTTCTATCGAGC
CAACGGTGACTTTATCGCCATTCC
CCATAATCATTTTATTTAGAGGGA
GAGAAATAAATTATATTAATAGATCGGA
TTAAAACCGGTGACGTGATGATGGTG
CGCAGAAATACCGGCTTTAAGTACC
AGCGGGAGCGGGAACCTTCAACGGTGAATCT
GTGTCTGTCGGAACATCATTTAATGCCGGAA
TCGCAAGATGTGGCGTGTTACGGTGAAAAC
CAAGGCGACAAGGTGCTGATGCCGCTGGCG
66 jC
497 bp
55 jC
420 bp
55 jC
476 bp
65 jC
625 bp
60 jC
632 bp
71 jC
125 bp
–
–
–
–
–
–
–
–
rfbEO157
rfbEO157
rfbBO157
rfbBO157
eaeAO157
eaeAO157
fliCH7
fliCH7
OrfuLEE O157
OrfuLEE O157
SILO157
SILO157
SILO157
SILO157
Cm
Cm
Desmarchelier et al., 1998
Desmarchelier et al., 1998
Maurer et al., 1999
Maurer et al., 1999
Louie et al., 1994
Louie et al., 1994
Gannon et al., 1997
Gannon et al., 1997
Meng et al., 1996
Meng et al., 1996
Perelle et al., 2002
Perelle et al., 2002
This study
This study
Fach et al., 2001
Fach et al., 2001
P. Fach et al. / Journal of Microbiological Methods 55 (2003) 383–392
3V-end-labeled with digoxigenin (Fach et al., 2001)
(Table 1). Thus using two different sets of internal
capture and detection probes in two distinct microtiter
plates, it was possible to differentiate amplicons derived from the SILO157 locus from those derived from
the IC.
2.7. Determination of the ELISA-PCR assay cut-off
The RJD3-SF6 amplified fragment from the SILO157 locus was cloned into pMOSBlue vector using
the pMOSBlue blunt ended cloning kit (Amersham
Biosciences, Saclay, France). The resulting recombinant plasmid DNA was purified using QIAGENR
plasmid kit (QIAGEN, Courtaboeuf, France) according to the manufacturer’s instructions. Its concentration was determined by fluorimetry and the copy
number of the plasmid was calculated according to
its size. The cloned sequence was used as positive
control and to determine the detection limit (cut-off) of
the ELISA-PCR. The threshold selection criterion
determining positive versus negative samples was
established by testing triplicates of serial 10-fold
dilutions (106 to 1 copies ml 1) of the plasmid in
ELISA-PCR. The mean values were calculated for
each dilution and for the negative controls, and standard deviations were also determined.
3. Results and discussion
3.1. Comparison of PCR assays for detecting Shiga
toxin-producing E. coli O157
Several simplex (nonmultiplex) PCR assays were
compared for detecting STEC O157. Specificity of the
PCR assays was evaluated with 277 bacterial strains
isolated from foods, animals or humans and six different simplex PCR assays amplifying the genes rfbE and
rfbB, the 3Vend of the eae gene, the region immediately
flanking the 5Vend of the eae gene, the flicH7 gene, or a
part of the SILO157 locus of STEC O157. Data showed
that the PCR assays amplifying the rfb sequences
indifferently detected all the 67 strains of toxigenic
E. coli O157 and the 93 strains of nontoxigenic E. coli
O157 that were isolated essentially from animals and
foods. These nontoxigenic E. coli O157 frequently
encountered in certain samples, such as bovine sam-
387
ples, were usually sorbitol-positive, had an H-type
other than H7, and were negative for the eaeA, hlyA,
katP EHEC virulence genes. Such strains, which have
to be unequivocally distinguished from toxigenic E.
coli O157, could be differentiated using the four other
PCR systems tested. The stx-negative derivatives of
STEC O157 (eight strains) that were sorbitol-negative
and positive for the eaeA, hlyA, katP EHEC virulence
genes could not be differentiated from toxigenic E. coli
O157 and tested positive in the six PCR assays. These
PCR systems, when evaluated on 34 serogroups of E.
coli (including STEC and non-STEC), showed a high
specificity for E. coli O157, giving cross reactions with
only a few E. coli strains of STEC or EPEC belonging
to other serogroups. Thus PCR tests amplifying the 3V
end of the eae gene and the region immediately
flanking the 5Vend of the eae gene detected strains of
E. coli O55 and O145. The EPEC strains O125:H6 and
O126:H6 also tested positive by amplification of the
region immediately flanking the 5Vend of the eae gene.
PCR assay targeting the SILO157 locus was found to be
more specific, detecting toxigenic O157 and only
clonally related E. coli O55. Amplification of the
flicH7 gene of STEC O157 allowed the detection of
toxigenic E. coli O157 of H7 type and strains of STEC
or EPEC belonging to the O55:H7 serogroup while the
nontoxigenic E. coli O157 of H types other than H7
such as HND, H19 or H45 were not detected. Thus the
simplex PCR assays for detecting toxigenic E. coli
O157 were relatively specific to STEC O157, essentially giving cross reactions with O55:H7 and to a
lesser extent with O125:H6, O126:H6 and O145
(Tables 2 and 3). They also detected the stx-negative
derivatives of STEC O157, occurring in patients,
animals or subcultures (Table 2). However, they did
not detect the 14 strains of other bacteria such as C.
freundii, Enterobacter sakasaki, H. alvei and Salmonella enterica group O301 which cross react with the E.
coli O157 Latex Test (Table 4).
3.2. ELISA-PCR test with cloned sequence of the
SILO157 locus and pure culture of bacteria
The threshold sensitivity of the ELISA-PCR assay
was determined by testing 10 Al of 10-fold serial
dilutions of the cloned sequence of the SILO157 (data
not shown). DNA plasmid concentrations of 106 and
105 copies ml 1 (104 and 103 copies in the PCR tube)
388
P. Fach et al. / Journal of Microbiological Methods 55 (2003) 383–392
Table 2
Examination of E. coli O157 isolates by PCR
Bacteria
Number
tested
Shiga toxin-producing E. coli O157 (STEC
O157:H7, NSF, Stx1 positive
O157:H7, NSF, Stx2 positive
O157:H7, NSF, Stx1 and Stx2 positive
O157:H-, NSF, Stx2 positive
O157:H-, SF, Stx2 positive
O157:H-, NSF, Stx1 and Stx2 positive
O157:Hnd, NSF, Stx2 positive
O157)
3a
16a
28a
4a
1a
2a
13a
Non-Shiga toxin-producing E. coli O157 (EC O157)
O157:Hnd, SF, Stx negative
76b
O157:Hnd, NSF, Stx negative
4b
O157:H7, NSF, Stx negative
4a
O157:Hnd, NSF, Stx negative
4a
O157:H19, SF, Stx negative
2b
O157:H45, SF, Stx negative
1c
O157:Hnd, SF, Stx negative
2b,d
PCR assays
rfbEO157
rfbBO157
eaeAO157
fliCH7
orfULEE
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
O157
SILO157
NSF: non-sorbitol fermenting, SF: sorbitol fermenting, ND: not determined.
a
Positive for the EHEC virulence eaeA, hlyA, katP genes.
b
Negative for the EHEC virulence eaeA, hlyA, katP genes.
c
Positive for the EHEC virulence eaeA gene.
d
Positive for the ETEC virulence ST.
gave an absorbance value of 4.0, and DNA plasmid
concentrations ranging from 104 to 103 copies ml 1
(102 – 10 copies in the PCR tube) gave absorbance
values between 3.620 and 4.0. Concentrations of 102
copies ml 1 (approximately one copy in the PCR
tube) usually yielded absorbance values lower than
0.03, except for one sample which was 0.140. For
concentrations under 10 copies ml 1 (less than one
copy in the PCR tube) absorbance values were lower
than 0.03. Mean values and standard deviation of PCR
Table 3
Examination of non-O157 E. coli isolates by PCR
Bacteria
Number
tested
PCR assays
rfbEO157
rfbBO157
E. coli giving a negative reaction with the O157 RPLA test
O55:H7, SF, Stx2 positive
1a,b
O55:H7, SF, Stx negative
3a,b
O125:H6, SF, Stx negative
1a,b
O126:H6, NSF, Stx negative
1a,b
O145:H-, SF, Stx1 positive
1c
O145:H-, SF, Stx2 positive
1d
O145:H-, SF, Stx negative
1c
O145:Hnd, SF, Stx2 positive
1d
Other serogroups of E. coli
93
(see Material and methods)
NSF: non-sorbitol fermenting, SF: sorbitol fermenting, ND: not determined.
a
Negative for the EHEC virulence hlyA, katP genes.
b
Positive for the EHEC virulence eaeA gene.
c
Positive for the EHEC virulence eaeA, hlyA, katP genes.
d
Positive for the EHEC virulence eaeA, hlyA genes.
eaeAO157
fliCH7
orfULEE
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
O157
SILO157
+
+
P. Fach et al. / Journal of Microbiological Methods 55 (2003) 383–392
389
Table 4
Examination of other bacterial species isolates by PCR
Bacteria
Number
tested
PCR assays
rfbEO157
rfbBO157
eaeAO157
fliCH7
orfULEE
O157
SILO157
Bacteria other than E. coli giving a positive reaction with the O157 RPLA test
Citrobacter freundii, NSF, Stx negative
1
Enterobacter sakasaki, SF, Stx negative
1
Hafnia alvei, NSF, Stx negative
4
Salmonella enterica group O301 (group N)
S. aqua, SF, Stx negative
1
S. aschersleben, SF, Stx negative
1
S. bodjonegoro, SF, Stx negative
1
S. landau, SF, Stx negative
1
S. morehead, SF, Stx negative
1
S. ramatgan, SF, Stx negative
1
S. soerenga, SF, Stx negative
1
S. urbana, SF, Stx negative
1
Other Stx negative species of bacteria
Pseudomonas aeruginosa, Pseudomonas fluorescens
4
Table 5
ELISA-PCR with pure culture of STEC O157
STEC O157 strains
VCA
Numeration (CFU ml
Theoretical
O157:H7
(ATCC 43890)
Stx1
O157:H(E-32511)
Stx2
O157:H7
(ATCC 43895)
Stx1, Stx2
0
1
10
102
103
104
105
106
0
1
10
102
103
104
105
106
0
1
10
102
103
104
105
106
1
)
Estimated
by plating
Estimated number
of copies in the
PCR tube
0
2.4
24
2.4 102
2.4 103
2.4 104
2.4 105
2.4 106
0
3.1
31
3.1 102
3.1 103
3.1 104
3.1 105
3.1 106
0
3
30
3 102
3 103
3 104
3 105
3 106
0
0.12
1.2
12
1.2 102
1.2 103
1.2 104
1.2 105
0
0.15
1.55
15.5
1.55 102
1.55 103
1.55 104
1.55 105
0
0.15
1.5
15
1.5 102
1.5 103
1.5 104
1.5 105
ELISA-PCR
SILO157
IC
Test 1
Test 2
Test 1
Test 2
0.001
3.565
4.000
4.000
4.000
4.000
4.000
4.000
0.004
0.001
4.000
4.000
4.000
4.000
4.000
4.000
0.011
0.005
3.677
4.000
4.000
4.000
4.000
4.000
0.005
4.000
4.000
4.000
4.000
4.000
4.000
4.000
0.008
4.000
4.000
4.000
4.000
4.000
4.000
4.000
0.004
0.003
3.737
4.000
4.000
4.000
4.000
4.000
3.699
3.694
3.428
3.391
1.217
0.585
0.029
0.028
4.000
3.827
3.491
3.078
1.457
0.468
0.049
0.033
3.786
3.778
3.628
3.434
2.143
0.343
0.078
0.028
3.794
3.697
3.589
3.501
2.506
0.497
0.029
0.025
4.000
3.630
3.607
3.582
3.349
2.399
0.074
0.012
4.000
4.000
3.796
3.516
3.065
0.107
0.058
0.025
390
P. Fach et al. / Journal of Microbiological Methods 55 (2003) 383–392
negative controls (sterile water) were also measured,
and the cut-off of the test was established on the basis
of mean values plus three standard deviations. Accordingly, absorbance values greater than or equal to
0.100 were considered positive, and values lower than
0.100 negative.
The sensitivity of the ELISA-PCR was then evaluated with pure cultures of STEC O157 10-fold
diluted from approximately 106 to 1 CFU ml 1 in
buffered peptone water. The results indicated absorbance values greater than the cut-off for dilutions of
approximately 24– 30 CFU ml 1 and above while
under these dilutions, i.e. for samples theoretically
containing less than one copy in the PCR tube,
alternatively negative and positive results were observed (Table 5). As expected for high dilution levels,
the copy number of target per PCR tube may vary
from 0 to 10. In this situation, results of the assay are
randomly positive and negative. STEC O157 and IC
were co-amplified with one common set of primers, in
the same conditions, and in the same PCR tube so that
simultaneous amplification sometimes resulted in the
inhibition of the IC, depending on the molar ratio of
STEC O157 and IC. The number of IC copies in each
PCR tube was limited to 10 to ensure competition
advantages for STEC O157. IC was never detected in
samples containing proportionally greater amounts of
STEC O157. For samples containing more than
1.2 104 STEC O157 in the PCR tube, IC was unable
to compete for primers and the IC signal was negative.
As expected, IC was detected in all STEC O157
negative samples.
4. Conclusion
Using E. coli O157 immunoassays, a few bacterial
cells can be detected following enrichment in selective
medium and subsequent isolation by immunomagnetic
separation. However, these tests are prone to crossreactivity of O antisera with organisms other than E.
coli O157 (Bettelheim et al., 1993; Borczyk et al.,
1987; Chart et al., 1992; Nataro and Kaper, 1998;
Shimada et al., 1992). Also, O157 has been reported to
cross-react with 12 O-antigens of other E. coli Oserogroups (Aleksic et al., 1992). Thus reliable detection of STEC O157 needs confirmation of the suspected colonies by biochemical identification and Stx
demonstration using either Vero cell assay, enzymelinked immunosorbent assay (ELISA) kits or PCR
based on the stx genes. Multiplex PCR assays have
been developed for identification of suspected clones
isolated from water, food, or fecal samples (Cebula et
al., 1995; Deng and Fratamico, 1996; Fratamico et al.,
1995; Gannon et al., 1997; Nagano et al., 1998; Paton
and Paton, 1998). Usually, they are based on detection
of both stx and rfb or flicH7 genes and offer a rapid and
reliable way to test isolated clones. However, in a
mixture of bacteria they are sometimes difficult to
manage because the genetic profile obtained by multiplex PCR may arise from more than one strain,
hindering interpretation. When both stx and rfbO157
genes are detected by multiplex PCR in a mixture of
bacteria, it is not possible to ensure that the signal is
displayed by STEC O157. The mixture may contain
non-Shiga toxin-producing E. coli O157 (EC O157)
together with STEC from another serogroup. In that
case, only isolation of E. coli clones and individual
testing of the isolates by PCR can demonstrate the
presence or absence of STEC O157 in the sample.
Thus for rapid screening of these samples there was a
need to select one simplex (nonmultiplex) PCR assay
based on a single maximally specific DNA sequence
identifying STEC O157. In this study, we compared,
on the same large collection of bacteria, different nonmultiplexed PCR systems used for detecting EHEC
O157 and chose the most appropriate system to adapt
an ELISA-PCR assay. An IC was also included in this
test to monitor false negative samples. The ELISAPCR assay consists of the solution hybridization of
amplicons with two probes that ensure the specificity
of the amplification. The first one is used as a capture
probe and the second as a detection probe. The capture
probe is labeled with biotin and is bound to streptavidin-coated microtiter plates, while the detection probe
is labeled with digoxigenin. After alkaline denaturing
of PCR products, hybridization occurs in parallel on
two microtiter plates: one for the specific detection of
STEC O157 and one for the internal control (IC)
detection. Thus by using two different sets of internal
capture and detection probes in two separate microtiter
plates, it is possible to differentiate amplicons derived
from STEC O157 from those derived from the IC.
Similarly to the ELISA technology, hybrids are
detected using a peroxidase anti-digoxigenin conjugate. The final enzymatic reaction of the test gives a
P. Fach et al. / Journal of Microbiological Methods 55 (2003) 383–392
colorimetric signal measured by spectrophotometry.
The ELISA-PCR test proved specific and highly
sensitive; a positive signal was detected in samples
containing less than 10 genomes of STEC O157 in the
PCR tubes. Comparing our data with those of the
literature, the ELISA-PCR test gave results comparable to those obtained in many other published E. coli
O157 PCR detection systems. There is no evidence for
significant differences in terms of sensitivity and
efficiency compared with the real-time PCR TaqMan
assays described in the literature for E. coli O157
(Oberst et al., 1998; Sharma, 2002). This method
would constitute a valuable PCR screening method
to detect STEC O157 in complex matrices like foods
or stools. Experiments are in progress to evaluate it on
naturally contaminated samples. However, as the number of STEC O157 positive samples is very low, such
contamination was undetected in samples investigated
at the moment in our laboratory.
Acknowledgements
We thank E. Oswald from INRA-ENVT, Toulouse,
France, who provided most of the reference strains of
STEC and the hybridoma 13C4 and 11F11, C. Lapeyre
who produced and purified monoclonal antibodies,
and M. Maire and C. Crucière from AFSSA for help in
the Vero Cell Assay. We are grateful to K. Frydendahl
from Danish Veterinary Laboratory, Copenhagen,
Denmark, for help in O-typing of strains, and to the
following for providing strains of bacteria: P. Pohl,
NIDO-INRV, Brussels, Belgium; D. Piérard, AZVUB, Brussels, Belgium; B. China and J. Mainil,
University of Liège, Belgium; G. Duffy, The National
Food Centre, Dublin, Ireland; M. Doyle, University of
Georgia, USA; D. Woodward and R. Caldeira, LCDC,
Ottawa, Ontario, Canada; P. Fratamico, USDA,
Philadelphia, PA, USA; A. Caprioli, Istituto Superiore
di Sanità, Rome, Italy; J. Blanco, LREC, Universidad
de Santiago de Compostela, Lugo, Spain; L. Beutin,
RKI, Berlin, Germany; H. Schmidt, University of
Würzburg, Germany; Y. Wasteson, The Norwegian
School of Veterinary Science, Oslo, Norway; C.
Vernozy-Rozand, ENVL, Marcy-l’Etoile, France; C.
Doit, Hôpital Robert-Debré, Paris, France; B. Andral,
AFSSA-Lyon, France and M. Bohnert, AFSSAPloufragan, France.
391
References
Aleksic, S., Karch, H., Bockemuhl, J., 1992. A biotyping scheme
for Shiga-like (Vero) toxin-producing Escherichia coli O157
and a list of serological cross-reactions between O157 and other
gram-negative bacteria. Zbl. Bakt. 276, 221 – 230.
Bettelheim, K.A., Evangelidis, H., Pearce, J.L., Sowers, E., Strockbine, N.A., 1993. Isolation of a Citrobacter freundii strain which
carries the Escherichia coli O157 antigen. J. Clin. Microbiol.
31, 760 – 761.
Borczyk, A.A., Lior, H., Ciebin, B., 1987. False positive identifications of Escherichia coli O157 in foods. Int. J. Food. Microbiol. 4, 347 – 349.
Brunder, W., Schmidt, H., Karch, H., 1996. KatP, a novel catalaseperoxidase encoded by the large plasmid of enterohaemorrhagic
Escherichia coli O157:H7. Microbiology 142, 3305 – 3315.
Cebula, T.A., Payne, W.L., Feng, P., 1995. Simultaneous identification of strains of Escherichia coli serotype O157:H7 and their
Shiga-like toxin type by mismatch amplification mutation assaymultiplex PCR. J. Clin. Microbiol. 33, 248 – 250 (published
erratum appears in J. Clin. Microbiol. 33, 1048).
Chart, H., Okubadejo, O.A., Rowe, B., 1992. The serological relationship between Escherichia coli O157 and Yersinia enterocolitica O9 using sera from patients with brucellosis. Epidemiol.
Infect. 108, 77 – 85.
Deng, M.Y., Fratamico, P.M., 1996. A multiplex PCR for rapid
identification of Shiga-like toxin-producing Escherichia coli
O157:H7 isolated from foods. J. Food. Prot. 59, 570 – 576.
Desmarchelier, P.M., Bilge, S.S., Fegan, N., Mills, L., Vary Jr., J.C.,
Tarr, P.I., 1998. A PCR specific for Escherichia coli O157 based
on the rfb locus encoding O157 lipopolysaccharide. J. Clin.
Microbiol. 36, 1801 – 1804.
Fach, P., Perelle, S., Dilasser, F., Grout, J., 2001. Comparison between a PCR-ELISA test and the Vero cell assay for detecting
Shiga toxin-producing Escherichia coli in dairy products and
characterization of virulence traits of the isolated strains.
J. Appl. Microbiol. 90, 809 – 818.
Fratamico, P.M., Sackitey, S.K., Wiedmann, M., Deng, M.Y., 1995.
Detection of Escherichia coli O157:H7 by multiplex PCR.
J. Clin. Microbiol. 33, 2188 – 2191.
Gannon, V.P.J., D’Souza, S., Graham, T., King, R.K., Rhan, K.,
Read, S., 1997. Use of the flagellar H7 gene as a target in
multiplex PCR assays and improved specificity in identification
of enterohemorrhagic Escherichia coli strains. J. Clin. Microbiol. 35, 656 – 662.
Gunzer, F., Bohm, H., Russmann, H., Bitzan, M., Aleksic, S.,
Karch, H., 1992. Molecular detection of sorbitol-fermenting Escherichia coli O157 in patients with hemolytic-uremic syndrome. J. Clin. Microbiol. 30, 1807 – 1810.
Konowalchuk, J., Speirs, J.I., Stavric, S., 1977. Vero response to a
cytotoxin of Escherichia coli. Infect. Immun. 18, 775 – 779.
Louie, M., De Azavedo, J., Clarke, R., Borczyk, A., Lior, H.,
Richter, M., Brunton, J., 1994. Sequence heterogeneity of
the eae gene and detection of verotoxin-producing Escherichia
coli using serotype-specific primers. Epidemiol. Infect. 112,
449 – 461.
Maurer, J.J., Schmidt, D., Petrosko, P., Sanchez, S., Bolton, L., Lee,
392
P. Fach et al. / Journal of Microbiological Methods 55 (2003) 383–392
M.D., 1999. Development of primers to O-antigen biosynthesis
genes for specific detection of Escherichia coli O157 by PCR.
Appl. Environ. Microbiol. 65, 2954 – 2960.
Meng, J., Zhao, S., Doyle, M.P., Mitchell, S.E., Kresovich, S.,
1996. Polymerase chain reaction for detecting Escherichia coli
O157:H7. Int. J. Food. Microbiol. 32, 103 – 113.
Nagano, I., Kunishima, M., Itoh, Y., Wu, Z., Takahashi, Y., 1998.
Detection of verotoxin-producing Escherichia coli O157:H7 by
multiplex polymerase chain reaction. Microbiol. Immunol. 42,
371 – 376.
Nataro, J.P., Kaper, J.B., 1998. Diarrheagenic Escherichia coli.
Clin. Microbiol. Rev. 11, 142 – 201 (published erratum appears
in Clin. Microbiol. Rev. 11, 403).
Oberst, R.D., Hays, M.P., Bohra, L.K., Phebus, R.K., Yamashiro,
C.T., Paszko-Kolva, C., Flood, S.J., Sargeant, J.M., Gillespie,
J.R., 1998. PCR-based DNA amplification and presumptive detection of Escherichia coli O157:H7 with an internal fluorogenic
probe and the 5Vnuclease (TaqMan) assay. Appl. Environ. Microbiol. 64, 3389 – 3396.
Paton, A.W., Paton, J.C., 1998. Detection and characterization of
Shiga toxigenic Escherichia coli by using multiplex PCR assays
for stx1, stx2, eaeA, enterohemorrhagic E. coli hlyA, rfbO111, and
rfbO157. J. Clin. Microbiol. 36, 598 – 602.
Perelle, S., Fach, P., Dilasser, F., Grout, J., 2002. A PCR test for
detecting Escherichia coli O157:H7 based on the identification
of the small inserted locus (SILO157). J. Appl. Microbiol. 93,
250 – 260.
Perera, L.P., Marques, L.R., O’Brien, A.D., 1988. Isolation and
characterization of monoclonal antibodies to Shiga-like toxin
II of enterohemorrhagic Escherichia coli and use of the monoclonal antibodies in a colony enzyme-linked immunosorbent
assay. J. Clin. Microbiol. 26, 2127 – 2131.
Pollard, D.R., Johnson, W.M., Lior, H., Tyler, S.D., Rozee, K.R.,
1990a. Rapid and specific detection of verotoxin genes in Escherichia coli by the polymerase chain reaction. J. Clin. Microbiol. 28, 540 – 545.
Pollard, D.R., Johnson, W.M., Lior, H., Tyler, S.D., Rozee, K.R.,
1990b. Differentiation of Shiga toxin and Vero cytotoxin
type 1 genes by polymerase chain reaction. J. Infect. Dis. 162,
1195 – 1198.
Sandhu, K.S., Clarke, R.C., McFadden, K., Brouwer, A., Louie, M.,
Wilson, J., Lior, H., Gyles, C.L., 1996. Prevalence of the eaeA
gene in verotoxigenic Escherichia coli strains from dairy cattle
in Southwest Ontario. Epidemiol. Infect. 116, 1 – 7.
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.
Sharma, V.K., 2002. Detection and quantitation of enterohemorrhagic Escherichia coli O157, O111, and O26 in beef and bovine
feces by real-time polymerase chain reaction. J. Food Prot. 65,
1371 – 1380.
Sharma, V.K., Dean-Nystrom, E.A., 2003. Detection of enterohemorrhagic Escherichia coli O157:H7 by using a multiplex realtime PCR assay for genes encoding intimin and Shiga toxins.
Vet. Microbiol. 93, 247 – 260.
Shimada, T., Kosako, Y., Isshiki, Y., Hisatsune, K., 1992. Enterohemorrhagic Escherichia coli O157:H7 possesses somatic (O)
antigen identical with that of Salmonella O301. Curr. Microbiol.
25, 215 – 217.
Strockbine, N.A., Marques, L.R., Holmes, R.K., O’Brien, A.D.,
1985. Characterization of monoclonal antibodies against Shigalike toxin from Escherichia coli. Infect. Immun. 50, 695 – 700.
Strockbine, N.A., Marques, L.R.M., Newland, J.W., Smith, H.W.,
Holmes, R.K., O’Brien, A.D., 1986. Two toxin-converting
phages from Escherichia coli O157:H7 strain 933 encode antigenically distinct toxins with similar biologic activities. Infect.
Immun. 53, 135 – 140.
Thompson, J.S., Hodge, D.S., Borczyk, A.A., 1990. Rapid biochemical test to identify verocytotoxin-positive strains of Escherichia
coli serotype O157. J. Clin. Microbiol. 28, 2165 – 2168.