Microbiol. Immunol., 51(8), 733–740, 2007
Evaluation of Repetitive Extragenic
Palindromic-PCR for Discrimination of Fecal
Escherichia coli from Humans, and Different
Domestic- and Wild-Animals
Bidyut R. Mohapatra*, 1, Klaas Broersma2, Rick Nordin1, and Asit Mazumder1
1
Water and Watershed Research Program, Department of Biology, University of Victoria, 3800 Finnerty Road, Victoria,
British Columbia V8P5C2, Canada, and 2Kamloops Range Research Station, Agriculture and Agri-Food Canada, 3015 Ord
Road, Kamloops, British Columbia V2B8A9, Canada
Received January 26, 2007; in revised form, May 14, 2007. Accepted May 24, 2007
Abstract: The objective of this study was to investigate the potential of repetitive extragenic palindromic
anchored polymerase chain reaction (rep-PCR) in differentiating fecal Escherichia coli isolates of human,
domestic- and wild-animal origin that might be used as a molecular tool to identify the possible source(s) of
fecal pollution of source water. A total of 625 fecal E. coli isolates of human, 3 domestic- (cow, dog and
horse) and 7 wild-animal (black bear, coyote, elk, marmot, mule deer, raccoon and wolf) species were characterized by rep-PCR DNA fingerprinting technique coupled with BOX A1R primer and discriminant
analysis. Discriminant analysis of rep-PCR DNA fingerprints of fecal E. coli isolates from 11 host sources
revealed an average rate of correct classification of 79.89%, and 84.6%, 83.8%, 83.3%, 82.5%, 81.6%,
80.8%, 79.8%, 79.3%, 77.4%, 73.2% and 63.6% of elk, human, marmot, mule deer, cow, coyote, raccoon,
horse, dog, wolf and black bear fecal E. coli isolates were assigned to the correct host source. These results
suggest that rep-PCR DNA fingerprinting procedures can be used as a source tracking tool for detection of
human- as well as animal-derived fecal contamination of water.
Key words: Escherichia coli, Fecal pollution, Molecular typing, Rep-PCR DNA fingerprinting
The quality of surface waters used for drinking,
recreation and irrigation has been deteriorating due to
increases in fecal pollution. Direct discharge of domestic waste, leaching from poorly maintained septic tanks,
sewer line leakage, inefficient sewage treatment, agricultural and urban runoff, and domestic- and wild-animal defecation are suspected as the major sources of
fecal contamination. Generally, human fecal input into
surface waters has been considered as a greater health
risk because of the presence of an array of human-specific enteric pathogens (Salmonella enterica serovar
Typhi, Shigella spp., hepatitis A virus and Norwalkgroup viruses). However, the emergence of zoonotic
enteric pathogens (Cryptosporidium, Giardia,
Escherichia coli O157:H7, different serotypes of Salmonella and others) in surface waters indicate the
potential public health risk also associated with animals-derived contamination, and better management
practices should be developed to reduce the environmental loading of pathogens associated with waterborne disease transmission (11, 20, 21).
Escherichia coli, a thermotolerant coliform universally present in the stools of endothermic animals has
been used as the regulatory indicator of fecal pollution
(3). However, the presence and/or increase in density of
E. coli provide no information about the fecal pollution
source. In an effort to adequately protect surface water
quality from enteric pathogens, programs are underway
to develop bacterial source tracking (BST) methods for
E. coli, which will allow pinpoint identification of
source(s) of fecal contamination.
Abbreviations: ARCC, average rate of correct classification;
BST, bacterial source tracking; DFAJ, discriminant function
analysis with Jackknife algorithm; RCC, rate of correct classification; Rep-PCR, repetitive extragenic palindromic anchored
polymerase chain reaction; UPGMA, unweighted pair group
method with arithmetic averages.
*Address correspondence to Dr. Bidyut Ranjan Mohapatra,
Water and Watershed Research Program, Department of Biology,
University of Victoria, Petch Building 116, 3800 Finnerty Road,
Victoria, British Columbia V8P5C2, Canada. Fax: 1–250–
721–7120. E-mail: [email protected]
733
734
B.R. MOHAPATRA ET AL
Presently, DNA fingerprinting methods, such as ribotyping, pulsed-field gel electrophoresis, amplified fragment length polymorphism, PCR analysis of the 16S23S rRNA intergenic spacer region, and repetitive
extragenic palindromic anchored PCR (rep-PCR) of E.
coli isolates are used in the BST studies for differentiation and identification of host sources (reviewed in 9,
13, 16). These fingerprinting methods rely on a library
which consists of a collection of DNA fingerprints of E.
coli from different fecal sources. The aim of these
methods is to compare the DNA fingerprints of environmental E. coli isolates to the library, which would
indicate if the fecal pollution in the environment
derived from a particular host group represented in the
library.
Rep-PCR DNA fingerprinting method targets the
interspersed conserved repetitive DNA sequences, present in various locations within the E. coli genome (18).
Three families of repetitive sequences have been identified: the repetitive extragenic palindromic (REP)
sequences, the enterobacterial repetitive intergenic consensus (ERIC) sequences, and the BOX sequences (19).
These repetitive units are considered to be highly conserved because rep sites are essential protein-DNA
interaction sites and/or these sequences may propagate
themselves as selfish DNA by gene conversion (18).
Amplification of the distinct genomic regions located
between these repetitive sequences is useful in the generation of a unique strain pattern (12).
Rep-PCR has been applied in genotyping of agricultural, environmental, industrial and medicinal bacteria
(19). It has also been proposed as a promising analytical tool for identification of fecal E. coli sources in surface waters (1, 4, 15, 17). Most of the rep-PCR BST
studies used E. coli isolates from relatively few animal
hosts. Rep-PCR, like other genotypic BST methods,
requires a known-animal source (host origin) DNA fingerprint library. It is essential to characterize the fecal
E. coli from various animal hosts for unbiased source
identification.
In this paper, we have assessed the potential of repPCR with BOX A1R primer to discriminate fecal E.
coli isolates obtained from humans, and different
domestic- and wild-animal sources. Fecal E. coli isolates of human (Homo sapiens), 3 domestic animals:
cow (Bos taurus), dog (Canis familiaris) and horse
(Equus caballus), and 7 wild animals: black bear
(Ursus americanus), coyote (Canis latrans), elk
(Cervus elaphus), marmot (Marmota caligata), mule
deer (Odocoileus hemionus), raccoon (Procyon lotor),
and wolf (Canis lupus) were used for testing the robustness of rep-PCR analysis. The E. coli rep-PCR DNA
fingerprint library developed here will be useful to track
the sources of fecal contamination in four watersheds,
Sooke Lake and Leech River (city: Victoria), Lakelse
Lake (town: Smithers), and Duteau Creek (city: Vernon), located in British Columbia, Canada. The surface
waters of these watersheds are used as the source of
drinking water and crop irrigation.
Materials and Methods
E. coli isolation and identification. The fresh fecal
Fig. 1. Map of British Columbia, Canada showing the fecal samples collection sites.
Table 1. Fecal E. coli isolates used in this study
Source
Horse
Cow
Dog
Mule deer
Elk
Wolf
Coyote
Black bear
Marmot
Raccoon
Human
Total
a)
Number of fecal samples
20
31
26
30
34
34
25
20
28
22
49
319
Number of isolates
41 (38)a) (32)b)
63 (57) (51)
52 (48) (44)
60 (56) (49)
68 (66) (60)
69 (66) (59)
51 (48) (41)
41 (38) (32)
56 (54) (50)
45 (42) (36)
112 (112) (98)
658 (625) (552)
Values in parenthesis indicate the number of isolates generated distinct rep-PCR DNA fingerprints.
b)
Values in boldface indicate the number of isolates used for
rep-PCR DNA fingerprinting statistical analysis (after exclusion
of clonal isolates).
Rep-PCR FINGERPRINTING OF FECAL E. COLI
samples of 3 domestic- and 7 wild-animals were collected from the surrounding areas of Sooke Lake (latitude 48º33', longitude 123º41'), Leech River (latitude
48º31', longitude 123º45'), Lakelse Lake (latitude 54º21',
longitude 128º35') and Duteau Creek (latitude 50º15',
longitude 119º15'), located in British Columbia (BC),
Canada (Fig. 1). The fecal samples were collected by
sterile container and transported on ice to the Water and
Watershed Research Laboratory, University of Victoria,
Victoria, BC. The human fecal samples were obtained
from healthy volunteers. All the animal and human
fecal samples were collected during June 2004 and
December 2005. The number of fecal samples collected
and number of E. coli isolated per host source is mentioned in Table 1. Fecal samples were processed within
24 to 48 hr of collection. In general, 1 g of fecal sample
was homogenized with 100 ml of saline (0.85% NaCl).
The samples were further diluted depending on the
source. The homogenate was filtered through a 47 mm
white gridded 0.45 µm membrane filter paper (Millipore, Inc., U.S.A.) with gentle vacuum. After filtration
the filter paper was placed on a Petri dish containing mColiBlue24TM medium (Millipore, Inc.) and incubated
at 37 C for 18 to 24 hr. E. coli forms blue color
colonies in m-ColiBlue24TM medium. Between two to
three E. coli colonies per fecal sample were isolated
and purified by restreaking thrice in MacConkey agar.
The identities of Escherichia coli isolates were reconfirmed by biochemical tests (3) and also by a PCRbased assay targeting the gene encoding the universal
stress protein (uspA) (2). The pure cultures of E. coli
were resuspended in 40% glycerol solution and stored at
80 C until further use.
Rep-PCR DNA fingerprinting of fecal E. coli. The
fecal E. coli isolates were cultured in tryptic soy broth at
37 C and harvested at late exponential growth phase
(OD600 nm1). The genomic DNA of the bacteria was
extracted by using the InstaGeneTM DNA extraction kit
(Bio-Rad Laboratories, Canada) as per the instructions
of manufacturer. Briefly, 1 ml bacterial cell culture was
centrifuged (10,000g, 10 min and 4 C) and the resulting cell pellet was washed with sterile saline (0.85%
NaCl). Two hundred microliters of InstaGeneTM matrix
(premixed on a magnetic stirrer) was added to the cell
pellet. The suspension was vortexed (10 sec) and incubated at 56 C for 30 min. Subsequently, the mixture
was vortexed for 10 sec and placed in a heating block at
100 C for 8 min. Afterwards the suspension was centrifuged at 12,000g for 10 min, and the supernatant
was used as the template DNA for rep-PCR fingerprinting. The rep-PCR assay mixture was prepared as
described by Rademaker and de Bruijn (12) with 50 ng
template DNA and 2 µM BOX A1R primer (5'-CTACG-
735
GCAAGGCGACGCTGACG-3'). Amplification was
performed in a PxE0.2 thermal cycler (Thermo Electron
Corporation, U.S.A.) with an initial denaturation at 95 C
for 2 min; 30 cycles consisting of 94 C for 3 sec, 92 C
for 30 sec, 50 C for 1 min, and 65 C for 8 min; followed by a single step extension at 65 C for 8 min. A
negative (0.22 µm filtered autoclaved distilled water)
and a positive (genomic DNA of E. coli K-12 strain)
control was included in the PCR experiments. The
PCR products were separated by horizontal gel electrophoresis on a 1.5% agarose gel. Gels were stained in
0.5 µg/ml ethidium bromide solutions for 30 min, and
the gel images were taken in an EpiChemi3 darkroom
bioimaging system (UVP, Inc., U.S.A.). The rep-PCR
fingerprints of the isolates were normalized with a 1-kb
DNA ladder, and the banding patterns were captured by
a LabWorksTM image acquisition and analysis software
(UVP, Inc.) accompanying the EpiChemi3 darkroom
bioimaging system. Only rep-PCR DNA fragments
from 0.3-kb to 6-kb were used for the statistical analysis
because the bands were distinct in this range.
Statistical analysis of rep-PCR DNA fingerprints.
The rep-PCR DNA fingerprints were subjected to hierarchical cluster analysis (SPSS, ver. 12.0 for Windows).
A dendrogram was constructed by using the Jaccard
similarity coefficient and UPGMA (unweighted pair
group method with arithmetic averages) algorithm provided in the SPSS software (ver. 12.0 for Windows).
Discriminant function analysis (prior probabilities,
equal, covariance matrix, pooled and leave-one-out
classification) with Jackknife algorithm was also performed with the cluster analysis results to find out the
number and percentage of isolates from each known
source that were classified in each source category
(SPSS, ver. 12.0 for Windows). The percentage of isolates from a given source that were placed in the correct
source category is termed as rate of correct classification
(RCC). The average rate of correct classification
(ARCC) is defined as the percent of the isolates correctly classified in all host categories. In the cluster analysis
and discriminant function analysis, the clonal E. coli
isolates (isolates having identical fingerprints) obtained
from the same host source were not included because
the previous BST study suggested that the presence of
identical fingerprints hinder the fidelity of the results of
cluster analysis and discriminant function analysis (8).
Results and Discussion
A total of 658 fecal E. coli isolates of horse, cow,
dog, mule deer, elk, wolf, coyote, black bear, marmot,
raccoon and human were subjected to rep-PCR DNA
fingerprinting with BOX A1R primer (Table 1). Thirty-
736
B.R. MOHAPATRA ET AL
three (5%) isolates did not generate distinct banding
patterns and therefore were not included in the statistical
analysis. Six hundred and twenty-five E. coli isolates
produced high quality rep-PCR DNA fingerprints and
composed of 10 to 28 amplimers ranging in sizes from
0.2-kb to 6-kb. A typical banding pattern of selected E.
Fig. 2. Rep-PCR fingerprint patterns of fecal E. coli isolates of
humans, and different domestic- and wild-animals. Lanes: 1, 1kb DNA ladder (Sigma); 2, horse isolate; 3, cow isolate; 4, dog
isolate; 5, mule deer isolate; 6, elk isolate; 7, wolf isolate; 8,
coyote isolate; 9, black bear isolate; 10, marmot isolate; 11, raccoon isolate; 12, human isolate; 13, E. coli K-12 strain.
coli strains from different host sources is shown in Fig.
2. The reproducibility of the rep-PCR banding patterns
was observed by conducting experiments with randomly
selected 3 E. coli isolates per host source. Although the
rep-PCR patterns of quadruplicate of an E. coli isolate
were not 100% identical, they were very tightly clustered among themselves with a similarity value of
greater than 92% (Fig. 3).
The cluster analysis was conducted with the rep-PCR
DNA fingerprinting patterns of 552 fecal E. coli isolates
after removal of the clonal isolates (Table 1). The dendrogram, constructed with Jaccard similarity coefficients and UPGMA algorithm, produced two clusters (I
and II) separating human-group isolates from animalgroup isolates (cow, dog, horse, black bear, coyote, elk,
marmot, mule deer, raccoon and wolf). Cluster I contained the majority of human-group isolates (81%),
while cluster II contained mainly animal-group isolates
(79%). The isolates of various hosts were not separated
by these two clusters. Therefore, a discriminant function analysis with Jackknife algorithm (DFAJ) was performed on the cluster analysis results to evaluate how
accurately the rep-PCR DNA fingerprints were able to
predict a host source. Both cluster analysis and DFAJ
are suitable statistical methods for genotyping of bacteria. Each method can be tested alone, but in combination usually provide results with additional confidence
(4, 5). The dendrogram formed by cluster analysis
shows the level of separation of source and the relatedness of the isolates tested. DFAJ can easily determine
the percentage of the isolates that has been correctly
identified to its source and the ARCC. The DFAJ of
552 E. coli isolates obtained from 11 host sources
yielded an ARCC of 79.89% (Table 2). The probability
Fig. 3. Rep-PCR fingerprint patterns of quadruplicate of a fecal E. coli isolate. Lanes: 1, 1-kb DNA ladder (Sigma); 2a,
b, c, d, horse isolate; 3a, b, c, d, cow isolate; 4a, b, c, d, dog isolate; 5a, b, c, d, mule deer isolate; 6a, b, c, d, elk isolate;
7a, b, c, d, wolf isolate; 8a, b, c, d, coyote isolate; 9a, b, c, d, black bear isolate; 10a, b, c, d, marmot isolate; 11a, b, c,
d, raccoon isolate; 12a, b, c, d, human isolate.
737
Rep-PCR FINGERPRINTING OF FECAL E. COLI
that an isolate fell into 1 of 11 host sources by chance
alone was 9.1%. The ARCC obtained here is within
the range of values reported by Dombek et al. (4) with
154 E. coli isolates from 7 host sources (human, duck,
geese, chicken, pig, sheep and cow), Carson et al. (1)
with 482 E. coli isolates from 8 host sources (human,
cattle, pig, horse, dog, chicken, turkey and goose),
Seurinck et al. (15) by using 267 E. coli isolates from 5
host sources (raw sewage, dog, cow, horse and gull) and
Somarelli et al. (17) with 123 E. coli isolates from 4
host sources (cow, goose, human and deer).
Elk, human, marmot, mule deer and cow isolates
were highly classified with a RCC of 84.6%, 83.8%,
83.3%, 82.5% and 81.6%, respectively, while 63.6%,
73.2%, 77.4%, 79.3%, 79.8% and 80.8% of black bear,
wolf, dog, horse, raccoon and coyote isolates, respectively, were classified correctly into their host source
(Table 2). RCC value of dog was lower than that stated
earlier by Carson et al. (1) and Seurinck et al. (15).
However, the RCC values of cow, horse and deer were
higher than those documented by others (1, 8, 17).
These differences in RCC values might be attributed to
the geographical variations of host animals, genetic
diversity among the E. coli isolates, and/or the feeding
habitat of the host animals (6, 8). A comparison of our
results for the isolates of black bear, coyote, marmot,
raccoon and wolf with literature was not possible
because the rep-PCR genotypic characterization of E.
coli isolates of these wild animals has not been studied
before.
When mule deer, elk, wolf, coyote, black bear, marmot and raccoon isolates were pooled together as the
wildlife-group, horse, cow and dog isolates were pooled
together as the domestic-group, and all the human isolates were combined as the human-group, the ARCC
was improved to 85.14%. The RCC was 87.7% for
wildlife-group, 85.4% for human-group and 80.3% for
domestic-group (Table 3). Several of the previous studies by using rep-PCR (1, 4, 15), amplified fragment
length polymorphism (5) and ribotyping (1) indicated
that pooling the data into a smaller group substantially
increases the ARCC. Although there is not an established standard of accuracy that has been defined for
any bacterial source tracking method, any method with a
RCC of 60% to 70% has been considered useful for the
water quality regulatory agencies for development of
Table 2. Percentage of fecal E. coli isolates assigned to the correct host source by using discriminant analysis (Jackknife algorithm) of
rep-PCR DNA fingerprints
% of fecal E. coli classified asa):
Host source
Horse
Cow
Dog
Mule deer
Elk
Wolf
Coyote
Black bear
Marmot
Raccoon
Human
a)
Horse
Cow
Dog
79.3
3.8
2.2
0
0
0
0
0
1.5
6.8
3.2
4.6
81.6
2.9
2.6
0
3.1
0
4.6
0
0
4.8
3.8
1.7
77.4
0
0
3.7
4.8
0
2.6
5.2
0
Mule
deer
0
4.7
0
82.5
7.8
3.4
2.1
6.8
4.6
0
2.8
Elk
Wolf
Coyote
0
0
0
3.1
84.6
3
0
6.2
0
0
0
0
0
4.8
0
3.9
73.2
6.1
0
0
3.6
0
0
0
4.9
4.1
3.7
7.9
80.8
9.6
6.8
0
0
Black
bear
0
5.6
3.8
0
0
2.7
0
63.6
0
4.6
5.4
Marmot Raccoon
5.4
0
0
4.1
0
0
0
0
83.3
0
0
5.1
0
2.4
2.5
0
3
6.2
5.6
0
79.8
0
Values in boldface indicate the rate of correct classification (RCC). The ARCC was 79.89%.
Table 3. Percentage (number) of fecal E. coli isolates assigned to the correct host-group (human vs.
domestic vs. wildlife) by using discriminant analysis (Jackknife algorithm) of rep-PCR DNA fingerprints
Host-group (number of isolates)
Domestic (127)
Human (98)
Wildlife (327)
a)
% (number) of fecal E. coli classified asa):
Domestic
Human
Wildlife
80.3 (102)
5.9 (7)
13.8 (18)
8.8 (9)
85.4 (82)
5.8 (7)
9.5 (32)
2.8 (9)
87.7 (286)
Values in boldface indicate the rate of correct classification (RCC). The ARCC was 85.14%.
Human
1.8
2.6
1.6
1.1
0
0
0
3.6
1.2
0
83.8
738
B.R. MOHAPATRA ET AL
better management practices for control and/or mitigation of fecal pathogens (7). As the ultimate goal of this
study is to identify the non-point source (human vs.
domestic vs. wildlife) of fecal contamination into the
surface water, the results obtained here may be promising for predicting the source(s) of human- and animalderived fecal pollution in the Sooke Lake, Leech River,
Lakelse Lake and Duteau Creek, located in British
Columbia.
A canonical discriminant function plot displayed that
127 E. coli strains isolated from domestic-group clearly
clustered into their specific host source (Fig. 4). The
two-function model showed significant group difference
(χ2622.36, P0.005) with functions 1 and 2 having
canonical correlation values of 0.959 and 0.904, respectively. The DFAJ among the E. coli isolates of domes-
tic-group resulted an ARCC of 84.25%, and 85.3%,
84.6% and 83.1% of cow, horse and dog isolates,
respectively were identified correctly (Table 4).
Rep-PCR fingerprinting methods could also distinguish the fecal E. coli isolates of wildlife-group. A
canonical discriminant function plot of first two functions is shown in Fig. 5. The χ2 was found to be
1,429.63 and the first two discriminant functions
accounted for 81.3%. The canonical correlation coefficients of the first two functions were 0.927 and 0.902,
respectively. The DFAJ of rep-PCR DNA fingerprints
of 327 fecal E. coli of wildlife-group indicated an
ARCC of 81.65%, and the RCC was 85.2%, 84.6%,
84.3%, 82.6% and 81.5% for elk, marmot, mule deer,
coyote and raccoon isolates, respectively (Table 5). E.
coli isolates of black bear and wolf were more often
Fig. 4. Canonical discriminant function plot of rep-PCR DNA fingerprints of fecal E. coli isolates of domestic-group.
Table 4. Percentage (number) of fecal E. coli isolates assigned to the correct host source in
domestic-group by using discriminant analysis (Jackknife algorithm) of rep-PCR DNA fingerprints
Host source (number of isolates)
Horse (32)
Cow (51)
Dog (44)
a)
% (number) of fecal E. coli classified asa):
Horse
Cow
Dog
84.6 (27)
13.2 (4)
2.2 (1)
8.2 (4)
85.3 (44)
6.5 (3)
10.8 (5)
6.1 (3)
83.1 (36)
Values in boldface indicate the rate of correct classification (RCC). The ARCC was 84.25%.
739
Rep-PCR FINGERPRINTING OF FECAL E. COLI
misidentified. The RCC of black bear isolates into
black bear-group was 74.8%, whereas their misclassification rate was 6.3% as a member of mule deer, coyote,
marmot and raccoon. Wolf isolates were correctly classified at a rate of 78.4% with misclassification as a
member of the black bear, raccoon and mule deer at a
rate of 7.5%, 7.2% and 6.9%, respectively.
In conclusion, the results of this study have provided
evidence about the robustness of rep-PCR DNA fingerprinting analysis in differentiating fecal E. coli strains
isolated from humans, and different domesticated and
wild animals. It has been reported that during runoff
events animals contribute significant amount of E. coli
contamination to the multiple fecal input remote watersheds in United States (14, 17) and Canada (10). As
host origin databases for bacterial source tracking
become more comprehensive with regards to number of
E. coli isolates and the animal species, it may be possible to accomplish pinpoint identification of specific animal fecal contributions and thereby prevent further pollution. Rep-PCR DNA fingerprinting is simple and
easy to perform and, thus, may prove to be a cost-effective screening tool for rapid determination of E. coli
isolates identity and tracking the non-point sources of
fecal contamination of surface water.
Fig. 5. Canonical discriminant function plot of rep-PCR fingerprints of fecal E. coli isolates of wildlife-group.
Table 5. Percentage (number) of isolates assigned to the correct host source in wildlife-group by using discriminant analysis (Jackknife
algorithm) of rep-PCR DNA fingerprints
Host source
(number of isolates)
Mule deer (49)
Elk (60)
Wolf (59)
Coyote (41)
Black bear (32)
Marmot (50)
Raccoon (36)
a)
Mule deer
84.3 (41)
8.7 (5)
6.9 (4)
3.7 (2)
6.3 (2)
7.9 (4)
0
Elk
7.8 (4)
85.2 (51)
0
0
0
0
0
% (number) of fecal E. coli classified asa):
Wolf
Coyote
Black bear
4.1 (2)
0
0
4.6 (3)
1.5 (1)
0
78.4 (46)
0
7.5 (5)
7.1 (3)
82.6 (33)
0
0
6.3 (2)
74.8 (24)
0
7.5 (4)
0
0
8.4 (3)
10.1 (4)
Values in boldface indicate the rate of correct classification (RCC). The ARCC was 81.65%.
Marmot
0
0
0
0
6.3 (2)
84.6 (42)
0
Raccoon
3.8 (2)
0
7.2 (4)
6.6 (3)
6.3 (2)
0
81.5 (29)
740
B.R. MOHAPATRA ET AL
We thank Nicole Harris for excellent technical assistance and
Dr. Julian Davies of University of British Columbia, Vancouver
for useful suggestions on genotyping of E. coli. This work was
funded by Canadian Institute of Health Research, and Natural
Sciences and Engineering Research Council of Canada Industrial Research Chair Program grants awarded to AM. We also
acknowledge the support of BC Ministry of Environment, Capital Regional District of Victoria (BC) and Kamloops Range
Research Station, Agriculture and Agri-Food Canada for collection of fecal samples.
10)
11)
12)
References
1) Carson, C.A., Shear, B.L., Ellersieck, M.R., and Schnell,
J.D. 2003. Comparison of ribotyping and repetitive extragenic palindromic-PCR for identification of fecal
Escherichia coli from humans and animals. Appl. Environ.
Microbiol. 69: 1836–1839.
2) Chen, J., and Griffiths, M.W. 1998. PCR differentiation of
Escherichia coli from other Gram-negative bacteria using
primers derived from the nucleotide sequences flanking the
gene encoding the universal stress protein. Lett. Appl.
Microbiol. 27: 369–371.
3) Clesceri, L.S., Greenberg, A.E., and Eaton, A.D. (eds),
1998. Standard methods for the examination of water and
wastewater, 20th edition, American Public Health Association, Washington, D.C.
4) Dombek, P.E., Johnson, L.K., Zimmerley, S.T., and Sadowsky, M.J. 2000. Use of repetitive DNA sequences and the
PCR to differentiate Escherichia coli isolates from human
and animal sources. Appl. Environ. Microbiol. 66:
2572–2577.
5) Guan, S., Xu, R., Chen, S., Odumeru, J., and Gyles, C.
2002. Development of a procedure for discriminating
among Escherichia coli isolates from animal and human
sources. Appl. Environ. Microbiol. 68: 2690–2698.
6) Hartel, P.G., Summer, J.D., and Segars, W.I. 2003. Deer
diet affects ribotype diversity of Escherichia coli for bacterial source tracking. Water Res. 37: 3263–3268.
7) Harwood, V.J., Whitlock, J., and Withington, V. 2000. Classification of antibiotic resistance patterns of indicator bacteria by discriminant analysis: use in predicting the source of
fecal contamination in subtropical waters. Appl. Environ.
Microbiol. 66: 3698–3704.
8) Johnson, L.K., Brown, M.B., Carruthers, E.A., Ferguson,
J.A., Dombek, P.E., and Sadowsky, M.J. 2004. Sample size,
library composition, and genotypic diversity among natural
populations of Escherichia coli from different animals
influence accuracy of determining sources of fecal pollution. Appl. Environ. Microbiol. 70: 4478–4485.
9) Meays, C.L., Broersma, K., Nordin, R., and Mazumder, A.
13)
14)
15)
16)
17)
18)
19)
20)
21)
2004. Source tracking fecal bacteria in water: a critical
review of current methods. J. Environ. Manage. 73: 71–79.
Meays, C.L., Broersma, K., Nordin, R., Mazumder, A., and
Samadpour, M. 2006. Spatial and annual variability in concentrations and sources of Escherichia coli in multiple
watersheds. Environ. Sci. Technol. 40: 5289–5296.
Pell, A.N. 1999. Manure and microbes: public and animal
health problem. J. Dairy Sci. 80, 2673–2681.
Rademaker, J.L.W., and de Bruijn, F.J. 1997. Characterization and classification of microbes by rep-PCR genomic
fingerprinting and computer-assisted pattern analysis, p.
151–171. In Caetano-Anollés, G., and Gresshoff, P.M.
(eds), DNA markers: protocols, applications, and
overviews, J. Wiley and Sons, New York.
Scott, T.M., Rose, J.B., Jenkins, T.M., Farrah, S.R., and
Lukasik, J. 2002. Microbial source tracking: current
methodology and future directions. Appl. Environ. Microbiol. 68: 5796–5803.
Scott, T.M., Caren, J., Nelson, G.R., Jenkins, T.M., and
Lukasik, J. 2004. Tracking sources of fecal pollution in a
South Carolina watershed by ribotyping Escherichia coli: a
case study. Environ. Forensics 5: 15–19.
Seurinck, S., Verstraete, W., and Siciliano, S.D. 2003. Use
of 16S-23S rRNA intergenic spacer region PCR and repetitive extragenic palindromic PCR analyses of Escherichia
coli isolates to identify nonpoint fecal sources. Appl. Environ. Microbiol. 69: 4942–4950.
Simpson, J.M., Domingo, J.W.S., and Reasoner, D.J. 2002.
Microbial source tracking: state of the science. Environ.
Sci. Technol. 36: 5279–5288.
Somarelli, J.A., Makarewicz, J.C., Sia, R., and Simon, R.
2007. Wildlife identified as major source of Escherichia
coli in agriculturally dominated watersheds by BOX A1Rderived genetic fingerprints. J. Environ. Manage. 82: 60–65.
Versalovic, J., Koeuth, T., and Lupski, J.R. 1991. Distribution of repetitive DNA-sequences in eubacteria and application to fingerprinting of bacterial genomes. Nucleic Acids
Res. 19: 6823–6831.
Versalovic, J., Schneider, M., de Bruijn, F.J., and Lupski,
J.R. 1994. Genomic fingerprinting of bacteria using repetitive sequence-based polymerase chain reaction. Methods
Mol. Cell. Biol. 5: 25–40.
WHO. 2003. Emerging issues in water and infectious diseases. World Health Organization, Geneva (online),
http://www.who.int/water_sanitation_health/emerging/en/em
erging.pdf, accessed 13 December 2006.
WHO. 2004. Waterborne zoonoses: identification, causes
and control. World Health Organization, Geneva (online),
http://www.who.int/water_sanitation_health/diseases/zoonos
es.pdf, accessed 13 December 2006.