Original Article
FOODBORNE PATHOGENS AND DISEASE Volume 0, Number 0, 2015 ª Mary Ann Liebert, Inc. DOI: 10.1089/fpd.2014.1860
Multilocus Sequence Typing and Virulence Gene Profiles Associated with Escherichia coli from Human and Animal Sources Amee R. Manges,1 Jose´e Harel,2 Luke Masson,3 Thaddeus J. Edens,4 Andrea Portt,5 Richard J. Reid-Smith,6,7 George G. Zhanel,8 Andrew M. Kropinski,6,9 and Patrick Boerlin 9
Abstract
We investigated whether specific sequence types, and their shared virulence gene profiles, may be associated with both human and food animal reservoirs. A total of 600 Escherichia coli isolates were assembled from human (n = 265) and food-animal (n = 335) sources from overlapping geographic areas and time periods (2005– 2010) in Canada. The entire collection was subjected to multilocus sequence typing and a subset of 286 E. coli isolates was subjected to an E. coli–specific virulence gene microarray. The most common sequence type (ST) was E. coli ST10, which was present in all human and food-animal sources, followed by ST69, ST73, ST95, ST117, and ST131. A core group of virulence genes was associated with all 10 common STs including artJ, ycfZ, csgA, csgE, fimA, fimH, gad, hlyE, ibeB, mviM, mviN, and ompA. STs 73, 92, and 95 exhibited the largest number of virulence genes, and all were exclusively identified from human infections. ST117 was found in both human and food-animal sources and shared virulence genes common in extraintestinal pathogenic E. coli lineages. Select groups of E. coli may be found in both human and food-animal reservoirs.
recovered from human urinary tract infection (UTI) cases, isolates recovered from the stool of a convenience sample of healthy individuals, retail meat (beef, pork, and chicken), and directly from food animals at slaughter (pig, beef cattle, and chicken). E. coli isolates were from overlapping geographic areas in Canada and time periods. The entire collection was subjected to multilocus sequence typing (MLST) and a subset was subjected to an E. coli–specific virulence gene microarray (Hamelin et al., 2007; Jakobsen et al., 2011) to investigate whether specific sequence types (STs), and their shared virulence gene profiles, may be associated with human or food-animal reservoirs or shared between both.
Introduction
E
xtraintestinal pathogenic Escherichia coli (ExPEC) are the major cause of extraintestinal infections such as urinary tract and bloodstream infections in humans. These infections are extremely common in the community and in health care institutions and are frequently multidrug resistant (Foxman, 2002; Russo and Johnson, 2003). There is emerging evidence that some E. coli that cause extraintestinal infections have a food-animal reservoir (Racicot Bergeron et al., 2012; Vincent et al., 2010). If infections caused by ExPEC are attributable to the introduction of new E. coli via contaminated food product(s), the relevance to public health, food animal production, and food safety would be significant. With the aim of comparing E. coli from human extraintestinal infections and from potential animal sources, a representative and systematically collected sample of E. coli isolates was assembled. This collection included E. coli
Materials and Methods Bacterial strains and DNA extraction
A total of 600 E. coli isolates were systematically assembled from human and food-animal sources (Table 1). E. coli
1
School of Population and Public Health, University of British Columbia, Vancouver, Canada. De´partement de Pathologie et Microbiologie, Universite´ de Montre´al, Faculte´ de Me´decine Ve´te´rinaire, Montre´al, Canada. 3 National Research Council of Canada, Montre´al, Canada. 4 Devil’s Staircase Consulting, Vancouver, Canada. 5 McGill University Health Centre Research Institute, Montre´al, Canada. 6 Laboratory for Foodborne Zoonoses, Public Health Agency of Canada, Guelph, Canada. 7 Department of Population Medicine, University of Guelph, Ontario Veterinary College, Guelph, Canada. 8 Department of Medical Microbiology, University of Manitoba, Winnipeg, Canada. 9 Department of Pathobiology, University of Guelph, Guelph, Canada. 2
1
2
MANGES ET AL.
Table 1. Assembly of Escherichia coli Collection from Human and Food Animal Sources Source PHAC abattoir (cecal sampling) Bovine
Total Microarray sample N subset N
Geographic location
56
26
Across Canada
Chicken
56
25
Across Canada
Porcine
56
26
Across Canada
56
0
Chicken
54
0
Pork
57
0
75
70
167
116
Quebeca
23
23
Quebec
600
286
PHAC CIPARS (retail meat sampling) Beef
Human clinical Hospital infections (UTI, bloodstream, kidney infections) Community-acquired UTIs Healthy stool Total
Years
Sampling
2005–2007 Random sample from CIPARS collection 2005–2007 Random sample from CIPARS collection 2005–2007 Random sample from CIPARS collection
Ontario & Quebec 2005–2007 Random sample from CIPARS collection Ontario & Quebec 2005–2007 Random sample from CIPARS collection Ontario & Quebec 2005–2007 Random sample from CIPARS collection Across Canada
2007–2008 Systematic sampling of E. coli from CANWARD hospital infections program 2005–2007 Consecutive sampling of all unique UTI episodes from two community clinics 2009 Convenience sample from healthy subjects
a Twelve isolates from California, United States (1999–2001) were included in this group, as they are representatives of important extraintestinal pathogenic Escherichia coli lineages. CANWARD, Canadian Hospital Ward Antibiotic Resistance Surveillance study; CIPARS, Canadian Integrated Program for Antimicrobial Resistance Surveillance; PHAC, Public Health Agency of Canada; UTI, urinary tract infection.
causing community-acquired UTIs were collected from two community health clinics from all consecutive UTI cases over a defined 2-year period (2005–2007) in Montre´al, Que´bec (Manges et al., 2008); these E. coli represent isolates from community-acquired infections. Isolates collected as part of the Canadian Hospital Ward Antibiotic Resistance Surveillance (CANWARD) study (2007–2008) represent E. coli from hospital-associated extraintestinal infections across Canada (Karlowsky et al., 2013). Commensal E. coli cultured from stool samples from a convenience sample of 23 healthy university students in Montre´al, Que´bec were included (2009–2010), as representatives of normal commensal E. coli. The Public Health Agency of Canada, Canadian Integrated Program for Antimicrobial Resistance Surveillance (CIPARS) provided 168 E. coli isolates from food animals (cecal) at slaughter and 167 isolates from retail meat (Government of Canada, 2009) (Table 1). Abattoirs in Canada, especially for beef cattle and pigs, are centralized; isolates were selected on the basis of the annual slaughter volume rather than on geographic location, and therefore were sampled across Canada. Cecal E. coli isolates from chicken were primarily recovered from abattoirs in Que´bec and Ontario, with a few additional isolates coming from abattoirs across Canada. Randomly sampled E. coli from the CIPARS retail meat isolate archive (2005–2007) were selected, with equal representation from chicken, pork, and beef meats, from Que´bec and Ontario. This sampling strategy was used to maximize the number of E. coli coming from the same area as the human source E. coli isolates. From within this 600 iso-
late collection, a subsample of 286 E. coli isolates was systematically selected for evaluation by E. coli microarray. These 286 isolates were selected to represent a range of diverse STs. Isolates were selected at random when multiple isolates of the same ST were available from within one source. The sample was assembled to include one third of isolates from food-animal sources, one third from hospitalized infections, and one third from community-acquired infections (although this group is slightly over-represented). We also included all intestinal E. coli isolates as we had relatively few isolates from healthy subjects. From the foodanimal sources, we elected to include the cecal isolates only in this subsample, as the retail meat isolates could be associated with both animal and human-source E. coli (via slaughter and meat processing and handling). MLST
MLST testing was carried out using the http:// mlst.warwick.ac.uk/mlst scheme. Sequencing of all loci was carried out at the McGill University and Ge´nome Que´bec Innovation Centre. Assignment of ST was made by algorithm on the MLST website. E. coli–specific virulence gene microarray hybridization
E. coli DNA was extracted by bacterial cell lysis, labeled, and hybridized on an E. coli–specific microarray containing 315 virulence genes and gene variants as described previously (Bruant et al., 2006; Jakobsen et al., 2011). Following
HUMAN AND FOOD ANIMAL E. COLI
hybridization, arrays were scanned with a ScanArray LITE (Perkin Elmer, Foster City, CA), and acquisition and quantification of background subtracted fluorescent spot intensities were performed using ScanArray Express software, version 2.1 (Perkin-Elmer). The median value of each set of duplicate spotted oligonucleotides was then compared to the median value of the negative control spots present on the array. Oligonucleotides with a signal-to-noise fluorescence ratio of 3.0 (based on a comparison of the negative control spots) were considered positive. Any gene that was present in fewer than two isolates was removed from further analyses. Statistical analyses
We present proportions of MLST sequence types in Table 2 arranged by source. The proportions of each virulence gene in human versus animal isolates are presented in Appendix Table A1. Odds ratios (and 95% CI) are presented indicating which virulence genes are over- or under-represented. Additional multivariable analyses were not performed due to limited sample sizes. We performed a nonmetric multidimensional scaling analysis to examine how similar the E. coli are by source (Fig. 1a) and by source for those E. coli representing the most common STs. All analyses were conducted in R (version 2.15.1) using the isoMDS command. Ninety-five percent confidence ellipses were estimated using the ellipse package in R. Results
A total of 600 E. coli isolates from food animals, collected from cecal swabs and from retail meat, representing beef/beef cattle, pork/pigs and chicken, and from human infections and healthy human stool, were assembled (Table 1). A detailed breakdown of MLST results for all E. coli is presented in Table 2. The most common ST was E. coli ST10, which was present in all human and food-animal sources. The next most common STs included ST69, ST73, ST95, ST117, and ST131. The common STs—73, 95, and 92—were exclusively associated with human source E. coli; ST69 and ST131 were largely associated with humans, though a few isolates were identified in chicken (ST131) or pork meat (ST69). STs 10, 117, 101, 38, 394, and 746 were identified in humans, retail meat, and food animals, while STs 641, 1490, 278, 401, and 48 were only identified in retail meat or food-animal sources. Over one third of E. coli isolates from each source contained a variety of unknown or nonannotated STs. A subsample of 286 E. coli representing diverse STs was subjected to an E. coli–specific virulence gene microarray. The distribution of STs among these 286 isolates from food animals and humans characterized by microarray (Table 1) was similar to the full sample of 600 isolates (data not shown). Specific sets of virulence genes were exclusively present or over-represented in animal-source isolates (exclusively in animal isolates: eae, escJ, escN, espA, espG, f17d-A, ler, nleA, cif, nleA(EHEC), and nleB(O103) and over-represented in animal isolates: lpfA, yjaA, astA, gafD, csgA, and flmA). These gene profiles suggest the presence of potential diarrheagenic E. coli pathotypes. The enteropathogenic E. coli–related genes present in some of these isolates have been reported previously (Racicot Bergeron et al., 2012). Specific sets of virulence genes were exclusively present or over-represented in human source isolates (ex-
3
clusively in human isolates: senB, sat, cnf, sfaD, z4184, papA(10), papGIII, sfaHII, wzxO6, kfiB, papA(48), focG, and focA and over-represented in human isolates: tspE, vat, neuA, neuC, irp(2), iha, b1432, usp, irp(1), fyuA, gimB (orf1), ibeA, kpsM-II, papC, malX, chuA, sitD, and many others). The complete summary of the distribution of virulence gene distribution by isolate source is provided in Appendix Table A1. Nonmetric multidimensional scaling (NMDS) analysis, using Jaccard distances, of virulence genes by source and ST was investigated. Figure 1a shows all isolates with complete virulence gene data and colored by source; 95% confidence ellipses are included. The fit-based R2 was 0.97. A large portion of the human isolates are distinct from the animalsource isolates. However, there is a region where the E. coli from all four sources overlap. Figure 1b shows the NMDS results for source and the most common STs in our data (fitbased R2 was 0.98). These results show that ST117 E. coli isolates from a chicken and porcine source cluster with a ST117 human isolate. Human source ST131 and ST69 and a few ST95 isolates, based on virulence gene profiles, also appear within the confidence ellipse defined by the chickensource isolates. It is also important to note that virtually all of the ST73 isolates cluster together and lie primarily in the human-isolate-only space. A core group of genes was associated with all 10 common STs including artJ, ycfZ, csgA, csgE, fimA, fimH, gad, hlyE, ibeB, mviM, mviN, and ompA. Unsurprisingly, the STs comprising primarily human infection–related E. coli had a higher number of virulence genes (range 36–50). Notably, STs 73, 92, and 95 exhibited the largest number of virulence genes, and all were exclusively identified from human infections. ST117 shared the virulence genes fliC, chuA, fepC, iss, irp, fyuA, sit, ompT, iucD, iut, usp, deoK, and malX, which are common to ExPEC lineages ( Johnson and Russo, 2005). However, the kpsMII gene was absent. STs 38, 10, and 641 exhibited fewer classic ExPEC virulence genes ( Johnson and Russo, 2005) and all were identified in both human and animal samples. E. coli ST131 is currently the most important ST among human ExPEC (Denisuik et al., 2013; Nicolas-Chanoine et al., 2014). In this study, ST131 was identified in nine community-acquired UTI samples, four hospital-acquired infections, and from two retail chicken meat samples. Of the 9 ST131 isolates examined by microarray, all exhibited the 10 core genes mentioned above and the majority also exhibited fliC, chuA, fepC, iss, irp, fyuA, kpsMII, sitA, sitD, ompT, and ccdB. Discussion
There is increasing concern that E. coli causing extraintestinal infections in humans may be transmitted via the foodborne route. E. coli isolates for this study were systematically selected from human infection (community- and hospital-associated infections) and food-animal sources from a specific time period and geographic areas in Canada. ST10, ST69, ST73, ST95, ST117, and ST131 were the most common E. coli STs in this study. Representative information on the distribution of these common STs by human and animal source is difficult to find. To get an idea of the range of sources for these STs, we queried the MLST database (http://
4
167
77
16
75
3 18 12 16 2 9 1 2 2 0 3 2 0 2 0 0 1 0 0 2 0 1 1 0 2 0 0 0 1 1 0 0 0 1 8
UTI
1 4 9 5 0 4 1 2 4 0 2 1 0 1 0 0 2 0 0 1 1 1 0 1 1 1 0 2 0 0 1 1 1 0 12
Hospital infections
UTI, urinary tract infection.
Total
10 69 73 95 117 131 101 38 92 641 493 394 1490 58 278 401 843 48 216 565 602 648 746 789 964 59 409 681 14 62 88 372 404 405 Other singletons Unknown STs
MLST ST
Human count
23
8
1 1 2 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 2 0 0 0 0 2 0 0 2 0 0 0 0 0 0 0 3
Stool
101
5 23 23 21 2 13 2 4 7 0 5 3 0 3 0 0 4 0 2 3 1 2 1 3 3 1 2 2 1 1 1 1 1 1 23
N
38.1
11.1 95.8 100.0 100.0 9.5 86.7 18.2 50.0 100.0 0.0 83.3 60.0 0.0 75.0 0.0 0.0 100.0 0.0 66.7 100.0 33.3 66.7 33.3 100.0 100.0 50.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 27.1
%
Total % human
56
31
5 0 0 0 0 0 1 0 0 1 0 1 0 1 3 2 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 10
Beef
54
12
8 0 0 0 7 2 2 1 0 2 0 0 1 0 0 0 0 2 1 0 2 1 1 0 0 0 0 0 0 0 0 0 0 0 12
Chicken
57
32
5 1 0 0 1 0 3 0 0 0 1 0 1 0 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 11
Pork
Retail meat count
75
18 1 0 0 8 2 6 1 0 3 1 1 2 1 3 4 0 3 1 0 2 1 1 0 0 0 0 0 0 0 0 0 0 0 33
N
28.3
40.0 4.2 0.0 0.0 38.1 13.3 54.5 12.5 0.0 42.9 16.7 20.0 40.0 25.0 75.0 100.0 0.0 100.0 33.3 0.0 66.7 33.3 33.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 38.8
%
Total % retail meat
56
34
2 0 0 0 0 0 1 3 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 15
Bovine
56
28
13 0 0 0 9 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 5
Chicken
56
27
7 0 0 0 2 0 2 0 0 4 0 1 3 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 9
Porcine
Abattoir (cecal) count
89
22 0 0 0 11 0 3 3 0 4 0 1 3 0 1 0 0 0 0 0 0 0 1 0 0 1 0 0 0 0 0 0 0 0 29
N
33.6
48.9 0.0 0.0 0.0 52.4 0.0 27.3 37.5 0.0 57.1 0.0 20.0 60.0 0.0 25.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 33.3 0.0 0.0 50.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 34.1
%
Total % abattoir
Table 2. Multilocus Sequence Typing (MLST) Sequence Type (ST) Classification for Human and Food Animal Source Escherichia coli
%
600
265 44.2
45 7.5 24 4.0 23 3.8 21 3.5 21 3.5 15 2.5 11 1.8 8 1.3 7 1.2 7 1.2 6 1.0 5 0.8 5 0.8 4 0.7 4 0.7 4 0.7 4 0.7 3 0.5 3 0.5 3 0.5 3 0.5 3 0.5 3 0.5 3 0.5 3 0.5 2 0.3 2 0.3 2 0.3 1 0.2 1 0.2 1 0.2 1 0.2 1 0.2 1 0.2 85 14.2
N
Total
FIG. 1. Visualization of nonmetric MDS analysis using virulence gene data to cluster Escherichia coli isolates by source (human, chicken, pork/porcine, or beef/bovine). (a) Groupings of E. coli as defined by virulence gene profiles for all four sources. (b) Groupings of E. coli as defined by virulence gene profiles for source and the most common multilocus sequence typing sequence types; 95% confidence ellipses are presented. This analysis was done using Jaccard distances in R (the isoMDS command). 5
6
mlst.warwick.ac.uk/mlst/dbs/Ecoli, accessed October 16, 2014). This database is dependent on investigator-initiated data deposition and therefore the proportions of each ST cannot be interpreted as the true distribution of STs in human and animal sources. Isolates from this study are not yet included in the MLST database. In this study, ST10 E. coli was found in human and all food-animal sources (Table 2). Similarly, we found 218 ST10 E. coli entries in the MLST database; 142 (65%) were recovered from humans and 27 (12%), 8 (4%), and 9 (4%) were recovered from beef/bovine, chicken, or pork/porcine sources, respectively. Other E. coli members of ST10 were isolated from companion or wild animals. In this study, ST69 was found primarily in humans, aside from a lone chicken isolate. In the MLST database, we found 92 ST69 E. coli entries, of which 84% were human, 7% were from poultry, and 1% were from beef cattle. We found 33 ST73 E. coli MLST database entries, of which 52% were associated with human sources and 39% were identified from companion animals; this mirrors our results, which identified only human ST73 isolates (although we did not sample companion animals). All ST95 E. coli were identified from humans in our study; however, we found 175 ST95 E. coli MLST database entries, of which 63% were human associated, 32% were poultry associated, and none were associated with bovine or porcine samples. We found 39 ST117 E. coli MLST database entries, of which 13% were from human sources and 41% were from poultry sources; none were associated with bovine or porcine samples. In our study, ST117 was likewise associated with human and poultry sources. Finally, we found 118 ST131 E. coli entries in the MLST database, of which 78% were recovered from human sources and 4% were recovered from poultry sources; there were no reports of ST131 from bovine sources and only 1 isolate reported from a pig. Our ST131 isolates were also primarily recovered from human and poultry sources. The shared pattern of ST distribution by source suggests that with additional sampling we would be able to capture the underlying distribution of specific human ExPEC lineages in food-animal and other sources. The distribution of virulence genes by source suggests that there are few genes confined exclusively to food-animal isolates; moreover, the virulence genes found exclusively in foodanimal isolates occur in a relatively small proportion of all food-animal isolates (range 2.6–7.8%). Similarly, the genes confined to human source isolates are also relatively infrequent in all human source isolates (with some exceptions such as senB, sat, cnf, and sfaD, range 1–34%). These results suggest that although distinct and host-adapted or host-specific E. coli populations exist among E. coli causing human infections and those inhabiting the food-animal gut, there are many genes common in isolates from both sources ( Johnson et al., 2003; Jakobsen et al., 2010, 2011). Moreover, the classic ExPEC genes, which were more frequently found among the human infection isolates, were also found in some animal isolates, though the proportion varied (e.g., iss found in 55% of animal isolates and ibe found in 1.3% of animal isolates) (Appendix Table A1). Nonmetric MDS analysis based on virulence genes also shows that specific and select groups of E. coli in food-animal reservoirs may serve as a source for human ExPEC infections. Conclusions
MLST results suggest that specific MLST lineages are strongly associated with human sources only (ST73 and
MANGES ET AL.
ST95), while some STs (e.g., ST117 and chicken, ST278 and beef cattle/beef, and ST1490 and pigs/pork) may be associated with specific food-animal species. A smaller number of STs were found in all reservoirs (ST10 and ST117) (Manges and Johnson, 2013). ST117 also shared multiple virulence genes common to ExPEC lineages, which has been observed previously (Vincent et al., 2010; Racicot Bergeron et al., 2012; Manges and Johnson, 2013). MLST and virulence gene profiling demonstrate that certain groups of E. coli may be strongly associated with human infections, whereas other E. coli are more strongly associated with food-animal reservoirs. A smaller, select group of E. coli lineages are found in both human and food-animal sources, indicating possible transmission from food-animal reservoirs to humans. Acknowledgments
We wish to thank members of the surveillance team of the Canadian Integrated Program for Antimicrobial Resistance Surveillance (Brent Avery) and the CANWARD study team. Funding for this study was contributed by the Health Canada Genomic Research and Development Initiative to P.B. and A.M.K., Public Health Agency of Canada and the Canadian Institutes of Health Research (CIHR), Institute of Infection and Immunity (MOP-114879) to A.R.M. Disclosure Statement
No competing financial interests exist. References
Bruant G, Maynard C, Bekal S, Gaucher I, Masson L, Brousseau R, et al. Development and validation of an oligonucleotide microarray for detection of multiple virulence and antimicrobial resistance genes in Escherichia coli. Appl Environ Microbiol 2006;72:3780–3784. Denisuik AJ, Lagace´-Wiens PR, Pitout JD, Mulvey MR, Simner PJ, Tailor F, et al. Molecular epidemiology of extendedspectrum b-lactamase-, AmpC b-lactamase- and carbapenemaseproducing Escherichia coli and Klebsiella pneumoniae isolated from Canadian hospitals over a 5 year period: CANWARD 2007-11. J Antimicrob Chemother 2013;68:i57–i65. Foxman B. Epidemiology of urinary tract infections: incidence, morbidity, and economic costs. Am J Med 2002;113(Suppl 1A):5S–13S. Government of Canada. Canadian Integrated Program for Antimicrobial Resistance Surveillance (CIPARS), 2008. Guelph, Ontario: Public Health Agency of Canada, 2009. Hamelin K, Bruant G, El-Shaarawi A, Hill S, Edge TA, Fairbrother J, et al. Occurrence of virulence and antimicrobial resistance genes in Escherichia coli isolates from different aquatic ecosystems within the St. Clair River and Detroit River areas. Appl Environ Microbiol 2007;73:477–484. Jakobsen L, Garneau P, Kurbasic A, Bruant G, Stegger M, Harel J, et al. Microarray-based detection of extended virulence and antimicrobial resistance gene profiles in phylogroup B2 Escherichia coli of human, meat and animal origin. J Med Microbiol 2011;60:1502–1511. Jakobsen L, Spangholm DJ, Pedersen K, Jensen LB, Emborg HD, Agersø Y, et al. Broiler chickens, broiler chicken meat, pigs and pork as sources of ExPEC related virulence genes and resistance in Escherichia coli isolates from communitydwelling humans and UTI patients. Int J Food Microbiol 2010;142:264–272.
HUMAN AND FOOD ANIMAL E. COLI
7
Johnson JR, Kuskowski MA, Owens K, Gajewski A, Winokur PL. Phylogenetic origin and virulence genotype in relation to resistance to fluoroquinolones and/or extended-spectrum cephalosporins and cephamycins among Escherichia coli isolates from animals and humans. J Infect Dis 2003;188:759–768. Johnson JR, Russo TA. Molecular epidemiology of extraintestinal pathogenic (uropathogenic) Escherichia coli. Int J Med Microbiol 2005;295:383–404. Karlowsky JA, Lagace-Wiens PRS, Simner PJ, DeCorby MR, Adam HJ, Walkty A, et al. Antimicrobial resistance in urinary tract pathogens in Canada from 2007 to 2009: CANWARD Surveillance Study. Antimicrob Agents Chemother 2013;55:3169–3175. Manges AR, Johnson JR. Food-borne origins of Escherichia coli causing extraintestinal infections. Clin Infect Dis 2013; 55:712–719. Manges AR, Tabor H, Tellis P, Vincent C, Tellier PP. Endemic and epidemic lineages of Escherichia coli that cause urinary tract infections. Emerg Infect Dis 2008;14:1575–1583. Nicolas-Chanoine MH, Bertrand X, Madec JY. Escherichia coli ST131, an intriguing clonal group. Clin Microbiol Rev 2014; 27:543–574.
Racicot Bergeron C, Prussing C, Boerlin P, Daignault D, Dutil L, Reid-Smith RJ, et al. Chicken as reservoir for human extraintestinal pathogenic Escherichia coli, Canada. Emerg Infect Dis 2012;18:415–421. Russo TA, Johnson JR. Medical and economic impact of extraintestinal infections due to Escherichia coli: Focus on an increasingly important endemic problem. Microbes Infect 2003;5:449–456. Vincent C, Boerlin P, Daignault D, Dozois CM, Dutil L, Galanakis C, et al. Food reservoir for Escherichia coli causing urinary tract infections. Emerg Infect Dis 2010; 16:88–95.
Address correspondence to: Amee R. Manges, MPH, PhD School of Population and Public Health University of British Columbia 137-2206 East Mall Vancouver, BC V6T 1Z3, Canada E-mail: [email protected]
Appendix Table A1. Complete List of Virulence Genes by Isolate Source Distribution in source Genes over-represented among human isolates
Gene
% Animal isolates
% Human isolates
% Total
Odds ratio
tspE vat neuA neuC irp(2) iha b1432 usp irp(1) fyuA gimB (orf1) ibeA kpsM-II papC malX chuA sitD sitA fepC papGII pic hlyA ompT agn iss mcbA iut(A2) deoK iron(2) iut(UPEC) iron iucD traT ompT(2) iutA iss(3)
0.039 0.026 0.013 0.013 0.182 0.013 0.013 0.052 0.182 0.184 0.013 0.013 0.130 0.052 0.130 0.247 0.286 0.286 0.260 0.039 0.026 0.091 0.403 0.247 0.338 0.013 0.169 0.221 0.169 0.182 0.169 0.169 0.169 0.403 0.156 0.545
0.627 0.507 0.239 0.239 0.833 0.225 0.220 0.517 0.809 0.799 0.163 0.163 0.684 0.397 0.641 0.789 0.804 0.780 0.756 0.244 0.172 0.373 0.794 0.579 0.679 0.048 0.421 0.483 0.397 0.416 0.383 0.378 0.349 0.603 0.282 0.670
0.469 0.378 0.178 0.178 0.657 0.168 0.164 0.392 0.640 0.635 0.122 0.122 0.535 0.304 0.503 0.643 0.664 0.647 0.622 0.189 0.133 0.297 0.689 0.490 0.587 0.038 0.353 0.413 0.336 0.353 0.325 0.322 0.301 0.549 0.248 0.636
41.43 38.59 23.90 23.90 22.37 21.76 21.45 19.51 19.01 17.61 14.77 14.77 14.52 12.02 11.97 11.45 10.24 8.86 8.83 7.96 7.80 5.95 5.73 4.20 4.16 3.82 3.58 3.30 3.24 3.21 3.05 2.99 2.64 2.25 2.13 1.69
95% CI 12.63 9.23 3.24 3.24 11.29 2.95 2.90 6.88 9.69 9.00 1.98 1.98 7.03 4.23 5.81 6.19 5.62 4.90 4.85 2.41 1.83 2.61 3.25 2.34 2.39 0.48 1.86 1.81 1.68 1.69 1.58 1.55 1.36 1.32 1.07 0.99
135.88 161.30 176.28 176.28 44.31 160.71 158.44 55.34 37.30 34.46 109.84 109.84 29.99 34.15 24.64 21.19 18.68 16.03 16.08 26.35 33.25 13.60 10.08 7.54 7.24 30.34 6.90 6.03 6.26 6.09 5.90 5.78 5.12 3.84 4.23 2.88
(continued)
8
MANGES ET AL.
Appendix Table A1. (Continued) Gene
% Animal isolates
% Human isolates
% Total
Odds ratio
Genes present at similar levels in animal and human isolates
rfc ce1A shf pilL f165(1)A papA(11) hraI fliC(H7) capU fliC ccdB cia bmaE cdtB(3) f17c-A b1121 eivG cvaC rtx fimA fimH cba virK tia cma cib eibA wzyO86 paa(ETEC) csgE cnf(2) eae(beta) paa
0.013 0.052 0.026 0.039 0.052 0.052 0.234 0.078 0.026 0.900 0.429 0.130 0.013 0.013 0.013 0.909 0.182 0.169 0.117 0.935 0.961 0.091 0.052 0.053 0.091 0.104 0.013 0.026 0.039 0.987 0.026 0.039 0.039
0.033 0.120 0.057 0.072 0.091 0.091 0.330 0.115 0.038 0.900 0.416 0.100 0.010 0.010 0.010 0.876 0.120 0.105 0.062 0.876 0.919 0.043 0.024 0.024 0.038 0.043 0.005 0.010 0.010 0.933 0.005 0.005 0.005
0.028 0.101 0.049 0.063 0.080 0.080 0.304 0.105 0.035 0.900 0.420 0.108 0.010 0.010 0.010 0.885 0.136 0.122 0.077 0.892 0.930 0.056 0.031 0.032 0.052 0.059 0.007 0.014 0.017 0.948 0.010 0.014 0.014
2.63 2.48 2.28 1.91 1.83 1.83 1.62 1.54 1.49 1.03 0.95 0.75 0.73 0.73 0.73 0.70 0.61 0.58 0.50 0.49 0.46 0.45 0.45 0.44 0.40 0.39 0.37 0.36 0.24 0.18 0.18 0.12 0.12
0.32 0.83 0.50 0.54 0.60 0.60 0.89 0.60 0.31 0.42 0.56 0.34 0.07 0.07 0.07 0.29 0.30 0.28 0.21 0.18 0.13 0.16 0.12 0.12 0.14 0.14 0.02 0.05 0.04 0.02 0.02 0.01 0.01
21.76 7.37 10.45 6.78 5.55 5.55 2.95 3.91 7.19 2.56 1.61 1.67 8.22 8.22 8.22 1.69 1.25 1.22 1.22 1.32 1.61 1.25 1.71 1.69 1.14 1.05 5.91 2.62 1.45 1.42 2.02 1.16 1.16
Genes overrepresented among animal isolates
lpfA yjaA astA(2) astA(1) lpfA(O113) gafD csgA flmA wzyO98 leoA lpfA(O157) tsh eprJ spaS tlrA toxB
0.169 0.779 0.234 0.230 0.558 0.078 0.987 0.221 0.273 0.052 0.052 0.506 0.792 0.753 0.187 0.753
0.081 0.526 0.077 0.060 0.187 0.014 0.909 0.033 0.043 0.005 0.005 0.072 0.211 0.096 0.005 0.057
0.105 0.594 0.119 0.100 0.287 0.031 0.930 0.084 0.105 0.017 0.017 0.189 0.367 0.273 0.053 0.245
0.44 0.31 0.27 0.20 0.18 0.17 0.13 0.12 0.12 0.09 0.09 0.08 0.07 0.03 0.02 0.02
0.20 0.17 0.13 0.09 0.10 0.04 0.02 0.05 0.05 0.01 0.01 0.04 0.04 0.02 0.00 0.01
0.95 0.58 0.57 0.44 0.32 0.71 1.00 0.31 0.28 0.80 0.80 0.15 0.13 0.07 0.16 0.04
Genes present in animal isolates only
eae escJ escN espA espG f17d-A ler nleA cif nleA(EHEC) nleB(O103)
0.078 0.078 0.078 0.078 0.078 0.078 0.078 0.078 0.052 0.052 0.052
0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
0.021 0.021 0.021 0.021 0.021 0.021 0.021 0.021 0.014 0.014 0.014
NE NE NE NE NE NE NE NE NE NE NE
Distribution in source
95% CI
(continued)
HUMAN AND FOOD ANIMAL E. COLI
9
Appendix Table A1. (Continued) Gene
% Animal isolates
% Human isolates
% Total
Odds ratio
espB-2 espB-3 etpD map(1) map(2) nleB(O157) nleE set tir-1 eae(gamma) nleA(EPEC) nleF tccP wzy(O146) wzy(O66)
0.039 0.039 0.039 0.039 0.039 0.039 0.039 0.039 0.039 0.026 0.026 0.026 0.026 0.026 0.026
0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
0.010 0.010 0.010 0.010 0.010 0.010 0.010 0.010 0.010 0.007 0.007 0.007 0.007 0.007 0.007
NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE
Genes present in human isolates only
senB sat cnf(1) sfaD z4184 papA(10) papGIII sfaHII wzxO6 kfiB papA(48) focG focA cdtB(1) papA(13) papA(16) papGIV kpsM-III sfaA papA(12) wb(O8) mcjA nfaE pixA afaE cdtB(4) facA papA(40) aap cka mccB papA(15) papA(8) papA(9) wzyO22 aatA wzyO15
0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
0.335 0.292 0.282 0.273 0.196 0.191 0.177 0.158 0.134 0.100 0.096 0.081 0.067 0.048 0.048 0.048 0.043 0.038 0.038 0.033 0.033 0.024 0.024 0.024 0.019 0.019 0.019 0.019 0.014 0.014 0.014 0.014 0.014 0.014 0.014 0.010 0.010
0.245 0.213 0.206 0.199 0.143 0.140 0.129 0.115 0.098 0.073 0.070 0.059 0.049 0.035 0.035 0.035 0.031 0.028 0.028 0.024 0.024 0.017 0.017 0.017 0.014 0.014 0.014 0.014 0.010 0.010 0.010 0.010 0.010 0.010 0.010 0.007 0.007
NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE
Common in both human and animal isolates
ompA hlyE mviM ibeB artJ gad mviN
1.000 1.000 1.000 1.000 1.000 1.000 1.000
0.866 0.895 0.904 0.957 0.971 0.981 0.986
0.902 0.923 0.930 0.969 0.979 0.986 0.990
NE NE NE NE NE NE NE
Distribution in source
The genes included in this table were present in at least two isolates within the entire dataset. NE, not estimatable.
95% CI
FOODBORNE PATHOGENS AND DISEASE Volume 0, Number 0, 2015 ª Mary Ann Liebert, Inc. DOI: 10.1089/fpd.2014.1860
Multilocus Sequence Typing and Virulence Gene Profiles Associated with Escherichia coli from Human and Animal Sources Amee R. Manges,1 Jose´e Harel,2 Luke Masson,3 Thaddeus J. Edens,4 Andrea Portt,5 Richard J. Reid-Smith,6,7 George G. Zhanel,8 Andrew M. Kropinski,6,9 and Patrick Boerlin 9
Abstract
We investigated whether specific sequence types, and their shared virulence gene profiles, may be associated with both human and food animal reservoirs. A total of 600 Escherichia coli isolates were assembled from human (n = 265) and food-animal (n = 335) sources from overlapping geographic areas and time periods (2005– 2010) in Canada. The entire collection was subjected to multilocus sequence typing and a subset of 286 E. coli isolates was subjected to an E. coli–specific virulence gene microarray. The most common sequence type (ST) was E. coli ST10, which was present in all human and food-animal sources, followed by ST69, ST73, ST95, ST117, and ST131. A core group of virulence genes was associated with all 10 common STs including artJ, ycfZ, csgA, csgE, fimA, fimH, gad, hlyE, ibeB, mviM, mviN, and ompA. STs 73, 92, and 95 exhibited the largest number of virulence genes, and all were exclusively identified from human infections. ST117 was found in both human and food-animal sources and shared virulence genes common in extraintestinal pathogenic E. coli lineages. Select groups of E. coli may be found in both human and food-animal reservoirs.
recovered from human urinary tract infection (UTI) cases, isolates recovered from the stool of a convenience sample of healthy individuals, retail meat (beef, pork, and chicken), and directly from food animals at slaughter (pig, beef cattle, and chicken). E. coli isolates were from overlapping geographic areas in Canada and time periods. The entire collection was subjected to multilocus sequence typing (MLST) and a subset was subjected to an E. coli–specific virulence gene microarray (Hamelin et al., 2007; Jakobsen et al., 2011) to investigate whether specific sequence types (STs), and their shared virulence gene profiles, may be associated with human or food-animal reservoirs or shared between both.
Introduction
E
xtraintestinal pathogenic Escherichia coli (ExPEC) are the major cause of extraintestinal infections such as urinary tract and bloodstream infections in humans. These infections are extremely common in the community and in health care institutions and are frequently multidrug resistant (Foxman, 2002; Russo and Johnson, 2003). There is emerging evidence that some E. coli that cause extraintestinal infections have a food-animal reservoir (Racicot Bergeron et al., 2012; Vincent et al., 2010). If infections caused by ExPEC are attributable to the introduction of new E. coli via contaminated food product(s), the relevance to public health, food animal production, and food safety would be significant. With the aim of comparing E. coli from human extraintestinal infections and from potential animal sources, a representative and systematically collected sample of E. coli isolates was assembled. This collection included E. coli
Materials and Methods Bacterial strains and DNA extraction
A total of 600 E. coli isolates were systematically assembled from human and food-animal sources (Table 1). E. coli
1
School of Population and Public Health, University of British Columbia, Vancouver, Canada. De´partement de Pathologie et Microbiologie, Universite´ de Montre´al, Faculte´ de Me´decine Ve´te´rinaire, Montre´al, Canada. 3 National Research Council of Canada, Montre´al, Canada. 4 Devil’s Staircase Consulting, Vancouver, Canada. 5 McGill University Health Centre Research Institute, Montre´al, Canada. 6 Laboratory for Foodborne Zoonoses, Public Health Agency of Canada, Guelph, Canada. 7 Department of Population Medicine, University of Guelph, Ontario Veterinary College, Guelph, Canada. 8 Department of Medical Microbiology, University of Manitoba, Winnipeg, Canada. 9 Department of Pathobiology, University of Guelph, Guelph, Canada. 2
1
2
MANGES ET AL.
Table 1. Assembly of Escherichia coli Collection from Human and Food Animal Sources Source PHAC abattoir (cecal sampling) Bovine
Total Microarray sample N subset N
Geographic location
56
26
Across Canada
Chicken
56
25
Across Canada
Porcine
56
26
Across Canada
56
0
Chicken
54
0
Pork
57
0
75
70
167
116
Quebeca
23
23
Quebec
600
286
PHAC CIPARS (retail meat sampling) Beef
Human clinical Hospital infections (UTI, bloodstream, kidney infections) Community-acquired UTIs Healthy stool Total
Years
Sampling
2005–2007 Random sample from CIPARS collection 2005–2007 Random sample from CIPARS collection 2005–2007 Random sample from CIPARS collection
Ontario & Quebec 2005–2007 Random sample from CIPARS collection Ontario & Quebec 2005–2007 Random sample from CIPARS collection Ontario & Quebec 2005–2007 Random sample from CIPARS collection Across Canada
2007–2008 Systematic sampling of E. coli from CANWARD hospital infections program 2005–2007 Consecutive sampling of all unique UTI episodes from two community clinics 2009 Convenience sample from healthy subjects
a Twelve isolates from California, United States (1999–2001) were included in this group, as they are representatives of important extraintestinal pathogenic Escherichia coli lineages. CANWARD, Canadian Hospital Ward Antibiotic Resistance Surveillance study; CIPARS, Canadian Integrated Program for Antimicrobial Resistance Surveillance; PHAC, Public Health Agency of Canada; UTI, urinary tract infection.
causing community-acquired UTIs were collected from two community health clinics from all consecutive UTI cases over a defined 2-year period (2005–2007) in Montre´al, Que´bec (Manges et al., 2008); these E. coli represent isolates from community-acquired infections. Isolates collected as part of the Canadian Hospital Ward Antibiotic Resistance Surveillance (CANWARD) study (2007–2008) represent E. coli from hospital-associated extraintestinal infections across Canada (Karlowsky et al., 2013). Commensal E. coli cultured from stool samples from a convenience sample of 23 healthy university students in Montre´al, Que´bec were included (2009–2010), as representatives of normal commensal E. coli. The Public Health Agency of Canada, Canadian Integrated Program for Antimicrobial Resistance Surveillance (CIPARS) provided 168 E. coli isolates from food animals (cecal) at slaughter and 167 isolates from retail meat (Government of Canada, 2009) (Table 1). Abattoirs in Canada, especially for beef cattle and pigs, are centralized; isolates were selected on the basis of the annual slaughter volume rather than on geographic location, and therefore were sampled across Canada. Cecal E. coli isolates from chicken were primarily recovered from abattoirs in Que´bec and Ontario, with a few additional isolates coming from abattoirs across Canada. Randomly sampled E. coli from the CIPARS retail meat isolate archive (2005–2007) were selected, with equal representation from chicken, pork, and beef meats, from Que´bec and Ontario. This sampling strategy was used to maximize the number of E. coli coming from the same area as the human source E. coli isolates. From within this 600 iso-
late collection, a subsample of 286 E. coli isolates was systematically selected for evaluation by E. coli microarray. These 286 isolates were selected to represent a range of diverse STs. Isolates were selected at random when multiple isolates of the same ST were available from within one source. The sample was assembled to include one third of isolates from food-animal sources, one third from hospitalized infections, and one third from community-acquired infections (although this group is slightly over-represented). We also included all intestinal E. coli isolates as we had relatively few isolates from healthy subjects. From the foodanimal sources, we elected to include the cecal isolates only in this subsample, as the retail meat isolates could be associated with both animal and human-source E. coli (via slaughter and meat processing and handling). MLST
MLST testing was carried out using the http:// mlst.warwick.ac.uk/mlst scheme. Sequencing of all loci was carried out at the McGill University and Ge´nome Que´bec Innovation Centre. Assignment of ST was made by algorithm on the MLST website. E. coli–specific virulence gene microarray hybridization
E. coli DNA was extracted by bacterial cell lysis, labeled, and hybridized on an E. coli–specific microarray containing 315 virulence genes and gene variants as described previously (Bruant et al., 2006; Jakobsen et al., 2011). Following
HUMAN AND FOOD ANIMAL E. COLI
hybridization, arrays were scanned with a ScanArray LITE (Perkin Elmer, Foster City, CA), and acquisition and quantification of background subtracted fluorescent spot intensities were performed using ScanArray Express software, version 2.1 (Perkin-Elmer). The median value of each set of duplicate spotted oligonucleotides was then compared to the median value of the negative control spots present on the array. Oligonucleotides with a signal-to-noise fluorescence ratio of 3.0 (based on a comparison of the negative control spots) were considered positive. Any gene that was present in fewer than two isolates was removed from further analyses. Statistical analyses
We present proportions of MLST sequence types in Table 2 arranged by source. The proportions of each virulence gene in human versus animal isolates are presented in Appendix Table A1. Odds ratios (and 95% CI) are presented indicating which virulence genes are over- or under-represented. Additional multivariable analyses were not performed due to limited sample sizes. We performed a nonmetric multidimensional scaling analysis to examine how similar the E. coli are by source (Fig. 1a) and by source for those E. coli representing the most common STs. All analyses were conducted in R (version 2.15.1) using the isoMDS command. Ninety-five percent confidence ellipses were estimated using the ellipse package in R. Results
A total of 600 E. coli isolates from food animals, collected from cecal swabs and from retail meat, representing beef/beef cattle, pork/pigs and chicken, and from human infections and healthy human stool, were assembled (Table 1). A detailed breakdown of MLST results for all E. coli is presented in Table 2. The most common ST was E. coli ST10, which was present in all human and food-animal sources. The next most common STs included ST69, ST73, ST95, ST117, and ST131. The common STs—73, 95, and 92—were exclusively associated with human source E. coli; ST69 and ST131 were largely associated with humans, though a few isolates were identified in chicken (ST131) or pork meat (ST69). STs 10, 117, 101, 38, 394, and 746 were identified in humans, retail meat, and food animals, while STs 641, 1490, 278, 401, and 48 were only identified in retail meat or food-animal sources. Over one third of E. coli isolates from each source contained a variety of unknown or nonannotated STs. A subsample of 286 E. coli representing diverse STs was subjected to an E. coli–specific virulence gene microarray. The distribution of STs among these 286 isolates from food animals and humans characterized by microarray (Table 1) was similar to the full sample of 600 isolates (data not shown). Specific sets of virulence genes were exclusively present or over-represented in animal-source isolates (exclusively in animal isolates: eae, escJ, escN, espA, espG, f17d-A, ler, nleA, cif, nleA(EHEC), and nleB(O103) and over-represented in animal isolates: lpfA, yjaA, astA, gafD, csgA, and flmA). These gene profiles suggest the presence of potential diarrheagenic E. coli pathotypes. The enteropathogenic E. coli–related genes present in some of these isolates have been reported previously (Racicot Bergeron et al., 2012). Specific sets of virulence genes were exclusively present or over-represented in human source isolates (ex-
3
clusively in human isolates: senB, sat, cnf, sfaD, z4184, papA(10), papGIII, sfaHII, wzxO6, kfiB, papA(48), focG, and focA and over-represented in human isolates: tspE, vat, neuA, neuC, irp(2), iha, b1432, usp, irp(1), fyuA, gimB (orf1), ibeA, kpsM-II, papC, malX, chuA, sitD, and many others). The complete summary of the distribution of virulence gene distribution by isolate source is provided in Appendix Table A1. Nonmetric multidimensional scaling (NMDS) analysis, using Jaccard distances, of virulence genes by source and ST was investigated. Figure 1a shows all isolates with complete virulence gene data and colored by source; 95% confidence ellipses are included. The fit-based R2 was 0.97. A large portion of the human isolates are distinct from the animalsource isolates. However, there is a region where the E. coli from all four sources overlap. Figure 1b shows the NMDS results for source and the most common STs in our data (fitbased R2 was 0.98). These results show that ST117 E. coli isolates from a chicken and porcine source cluster with a ST117 human isolate. Human source ST131 and ST69 and a few ST95 isolates, based on virulence gene profiles, also appear within the confidence ellipse defined by the chickensource isolates. It is also important to note that virtually all of the ST73 isolates cluster together and lie primarily in the human-isolate-only space. A core group of genes was associated with all 10 common STs including artJ, ycfZ, csgA, csgE, fimA, fimH, gad, hlyE, ibeB, mviM, mviN, and ompA. Unsurprisingly, the STs comprising primarily human infection–related E. coli had a higher number of virulence genes (range 36–50). Notably, STs 73, 92, and 95 exhibited the largest number of virulence genes, and all were exclusively identified from human infections. ST117 shared the virulence genes fliC, chuA, fepC, iss, irp, fyuA, sit, ompT, iucD, iut, usp, deoK, and malX, which are common to ExPEC lineages ( Johnson and Russo, 2005). However, the kpsMII gene was absent. STs 38, 10, and 641 exhibited fewer classic ExPEC virulence genes ( Johnson and Russo, 2005) and all were identified in both human and animal samples. E. coli ST131 is currently the most important ST among human ExPEC (Denisuik et al., 2013; Nicolas-Chanoine et al., 2014). In this study, ST131 was identified in nine community-acquired UTI samples, four hospital-acquired infections, and from two retail chicken meat samples. Of the 9 ST131 isolates examined by microarray, all exhibited the 10 core genes mentioned above and the majority also exhibited fliC, chuA, fepC, iss, irp, fyuA, kpsMII, sitA, sitD, ompT, and ccdB. Discussion
There is increasing concern that E. coli causing extraintestinal infections in humans may be transmitted via the foodborne route. E. coli isolates for this study were systematically selected from human infection (community- and hospital-associated infections) and food-animal sources from a specific time period and geographic areas in Canada. ST10, ST69, ST73, ST95, ST117, and ST131 were the most common E. coli STs in this study. Representative information on the distribution of these common STs by human and animal source is difficult to find. To get an idea of the range of sources for these STs, we queried the MLST database (http://
4
167
77
16
75
3 18 12 16 2 9 1 2 2 0 3 2 0 2 0 0 1 0 0 2 0 1 1 0 2 0 0 0 1 1 0 0 0 1 8
UTI
1 4 9 5 0 4 1 2 4 0 2 1 0 1 0 0 2 0 0 1 1 1 0 1 1 1 0 2 0 0 1 1 1 0 12
Hospital infections
UTI, urinary tract infection.
Total
10 69 73 95 117 131 101 38 92 641 493 394 1490 58 278 401 843 48 216 565 602 648 746 789 964 59 409 681 14 62 88 372 404 405 Other singletons Unknown STs
MLST ST
Human count
23
8
1 1 2 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 2 0 0 0 0 2 0 0 2 0 0 0 0 0 0 0 3
Stool
101
5 23 23 21 2 13 2 4 7 0 5 3 0 3 0 0 4 0 2 3 1 2 1 3 3 1 2 2 1 1 1 1 1 1 23
N
38.1
11.1 95.8 100.0 100.0 9.5 86.7 18.2 50.0 100.0 0.0 83.3 60.0 0.0 75.0 0.0 0.0 100.0 0.0 66.7 100.0 33.3 66.7 33.3 100.0 100.0 50.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 27.1
%
Total % human
56
31
5 0 0 0 0 0 1 0 0 1 0 1 0 1 3 2 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 10
Beef
54
12
8 0 0 0 7 2 2 1 0 2 0 0 1 0 0 0 0 2 1 0 2 1 1 0 0 0 0 0 0 0 0 0 0 0 12
Chicken
57
32
5 1 0 0 1 0 3 0 0 0 1 0 1 0 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 11
Pork
Retail meat count
75
18 1 0 0 8 2 6 1 0 3 1 1 2 1 3 4 0 3 1 0 2 1 1 0 0 0 0 0 0 0 0 0 0 0 33
N
28.3
40.0 4.2 0.0 0.0 38.1 13.3 54.5 12.5 0.0 42.9 16.7 20.0 40.0 25.0 75.0 100.0 0.0 100.0 33.3 0.0 66.7 33.3 33.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 38.8
%
Total % retail meat
56
34
2 0 0 0 0 0 1 3 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 15
Bovine
56
28
13 0 0 0 9 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 5
Chicken
56
27
7 0 0 0 2 0 2 0 0 4 0 1 3 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 9
Porcine
Abattoir (cecal) count
89
22 0 0 0 11 0 3 3 0 4 0 1 3 0 1 0 0 0 0 0 0 0 1 0 0 1 0 0 0 0 0 0 0 0 29
N
33.6
48.9 0.0 0.0 0.0 52.4 0.0 27.3 37.5 0.0 57.1 0.0 20.0 60.0 0.0 25.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 33.3 0.0 0.0 50.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 34.1
%
Total % abattoir
Table 2. Multilocus Sequence Typing (MLST) Sequence Type (ST) Classification for Human and Food Animal Source Escherichia coli
%
600
265 44.2
45 7.5 24 4.0 23 3.8 21 3.5 21 3.5 15 2.5 11 1.8 8 1.3 7 1.2 7 1.2 6 1.0 5 0.8 5 0.8 4 0.7 4 0.7 4 0.7 4 0.7 3 0.5 3 0.5 3 0.5 3 0.5 3 0.5 3 0.5 3 0.5 3 0.5 2 0.3 2 0.3 2 0.3 1 0.2 1 0.2 1 0.2 1 0.2 1 0.2 1 0.2 85 14.2
N
Total
FIG. 1. Visualization of nonmetric MDS analysis using virulence gene data to cluster Escherichia coli isolates by source (human, chicken, pork/porcine, or beef/bovine). (a) Groupings of E. coli as defined by virulence gene profiles for all four sources. (b) Groupings of E. coli as defined by virulence gene profiles for source and the most common multilocus sequence typing sequence types; 95% confidence ellipses are presented. This analysis was done using Jaccard distances in R (the isoMDS command). 5
6
mlst.warwick.ac.uk/mlst/dbs/Ecoli, accessed October 16, 2014). This database is dependent on investigator-initiated data deposition and therefore the proportions of each ST cannot be interpreted as the true distribution of STs in human and animal sources. Isolates from this study are not yet included in the MLST database. In this study, ST10 E. coli was found in human and all food-animal sources (Table 2). Similarly, we found 218 ST10 E. coli entries in the MLST database; 142 (65%) were recovered from humans and 27 (12%), 8 (4%), and 9 (4%) were recovered from beef/bovine, chicken, or pork/porcine sources, respectively. Other E. coli members of ST10 were isolated from companion or wild animals. In this study, ST69 was found primarily in humans, aside from a lone chicken isolate. In the MLST database, we found 92 ST69 E. coli entries, of which 84% were human, 7% were from poultry, and 1% were from beef cattle. We found 33 ST73 E. coli MLST database entries, of which 52% were associated with human sources and 39% were identified from companion animals; this mirrors our results, which identified only human ST73 isolates (although we did not sample companion animals). All ST95 E. coli were identified from humans in our study; however, we found 175 ST95 E. coli MLST database entries, of which 63% were human associated, 32% were poultry associated, and none were associated with bovine or porcine samples. We found 39 ST117 E. coli MLST database entries, of which 13% were from human sources and 41% were from poultry sources; none were associated with bovine or porcine samples. In our study, ST117 was likewise associated with human and poultry sources. Finally, we found 118 ST131 E. coli entries in the MLST database, of which 78% were recovered from human sources and 4% were recovered from poultry sources; there were no reports of ST131 from bovine sources and only 1 isolate reported from a pig. Our ST131 isolates were also primarily recovered from human and poultry sources. The shared pattern of ST distribution by source suggests that with additional sampling we would be able to capture the underlying distribution of specific human ExPEC lineages in food-animal and other sources. The distribution of virulence genes by source suggests that there are few genes confined exclusively to food-animal isolates; moreover, the virulence genes found exclusively in foodanimal isolates occur in a relatively small proportion of all food-animal isolates (range 2.6–7.8%). Similarly, the genes confined to human source isolates are also relatively infrequent in all human source isolates (with some exceptions such as senB, sat, cnf, and sfaD, range 1–34%). These results suggest that although distinct and host-adapted or host-specific E. coli populations exist among E. coli causing human infections and those inhabiting the food-animal gut, there are many genes common in isolates from both sources ( Johnson et al., 2003; Jakobsen et al., 2010, 2011). Moreover, the classic ExPEC genes, which were more frequently found among the human infection isolates, were also found in some animal isolates, though the proportion varied (e.g., iss found in 55% of animal isolates and ibe found in 1.3% of animal isolates) (Appendix Table A1). Nonmetric MDS analysis based on virulence genes also shows that specific and select groups of E. coli in food-animal reservoirs may serve as a source for human ExPEC infections. Conclusions
MLST results suggest that specific MLST lineages are strongly associated with human sources only (ST73 and
MANGES ET AL.
ST95), while some STs (e.g., ST117 and chicken, ST278 and beef cattle/beef, and ST1490 and pigs/pork) may be associated with specific food-animal species. A smaller number of STs were found in all reservoirs (ST10 and ST117) (Manges and Johnson, 2013). ST117 also shared multiple virulence genes common to ExPEC lineages, which has been observed previously (Vincent et al., 2010; Racicot Bergeron et al., 2012; Manges and Johnson, 2013). MLST and virulence gene profiling demonstrate that certain groups of E. coli may be strongly associated with human infections, whereas other E. coli are more strongly associated with food-animal reservoirs. A smaller, select group of E. coli lineages are found in both human and food-animal sources, indicating possible transmission from food-animal reservoirs to humans. Acknowledgments
We wish to thank members of the surveillance team of the Canadian Integrated Program for Antimicrobial Resistance Surveillance (Brent Avery) and the CANWARD study team. Funding for this study was contributed by the Health Canada Genomic Research and Development Initiative to P.B. and A.M.K., Public Health Agency of Canada and the Canadian Institutes of Health Research (CIHR), Institute of Infection and Immunity (MOP-114879) to A.R.M. Disclosure Statement
No competing financial interests exist. References
Bruant G, Maynard C, Bekal S, Gaucher I, Masson L, Brousseau R, et al. Development and validation of an oligonucleotide microarray for detection of multiple virulence and antimicrobial resistance genes in Escherichia coli. Appl Environ Microbiol 2006;72:3780–3784. Denisuik AJ, Lagace´-Wiens PR, Pitout JD, Mulvey MR, Simner PJ, Tailor F, et al. Molecular epidemiology of extendedspectrum b-lactamase-, AmpC b-lactamase- and carbapenemaseproducing Escherichia coli and Klebsiella pneumoniae isolated from Canadian hospitals over a 5 year period: CANWARD 2007-11. J Antimicrob Chemother 2013;68:i57–i65. Foxman B. Epidemiology of urinary tract infections: incidence, morbidity, and economic costs. Am J Med 2002;113(Suppl 1A):5S–13S. Government of Canada. Canadian Integrated Program for Antimicrobial Resistance Surveillance (CIPARS), 2008. Guelph, Ontario: Public Health Agency of Canada, 2009. Hamelin K, Bruant G, El-Shaarawi A, Hill S, Edge TA, Fairbrother J, et al. Occurrence of virulence and antimicrobial resistance genes in Escherichia coli isolates from different aquatic ecosystems within the St. Clair River and Detroit River areas. Appl Environ Microbiol 2007;73:477–484. Jakobsen L, Garneau P, Kurbasic A, Bruant G, Stegger M, Harel J, et al. Microarray-based detection of extended virulence and antimicrobial resistance gene profiles in phylogroup B2 Escherichia coli of human, meat and animal origin. J Med Microbiol 2011;60:1502–1511. Jakobsen L, Spangholm DJ, Pedersen K, Jensen LB, Emborg HD, Agersø Y, et al. Broiler chickens, broiler chicken meat, pigs and pork as sources of ExPEC related virulence genes and resistance in Escherichia coli isolates from communitydwelling humans and UTI patients. Int J Food Microbiol 2010;142:264–272.
HUMAN AND FOOD ANIMAL E. COLI
7
Johnson JR, Kuskowski MA, Owens K, Gajewski A, Winokur PL. Phylogenetic origin and virulence genotype in relation to resistance to fluoroquinolones and/or extended-spectrum cephalosporins and cephamycins among Escherichia coli isolates from animals and humans. J Infect Dis 2003;188:759–768. Johnson JR, Russo TA. Molecular epidemiology of extraintestinal pathogenic (uropathogenic) Escherichia coli. Int J Med Microbiol 2005;295:383–404. Karlowsky JA, Lagace-Wiens PRS, Simner PJ, DeCorby MR, Adam HJ, Walkty A, et al. Antimicrobial resistance in urinary tract pathogens in Canada from 2007 to 2009: CANWARD Surveillance Study. Antimicrob Agents Chemother 2013;55:3169–3175. Manges AR, Johnson JR. Food-borne origins of Escherichia coli causing extraintestinal infections. Clin Infect Dis 2013; 55:712–719. Manges AR, Tabor H, Tellis P, Vincent C, Tellier PP. Endemic and epidemic lineages of Escherichia coli that cause urinary tract infections. Emerg Infect Dis 2008;14:1575–1583. Nicolas-Chanoine MH, Bertrand X, Madec JY. Escherichia coli ST131, an intriguing clonal group. Clin Microbiol Rev 2014; 27:543–574.
Racicot Bergeron C, Prussing C, Boerlin P, Daignault D, Dutil L, Reid-Smith RJ, et al. Chicken as reservoir for human extraintestinal pathogenic Escherichia coli, Canada. Emerg Infect Dis 2012;18:415–421. Russo TA, Johnson JR. Medical and economic impact of extraintestinal infections due to Escherichia coli: Focus on an increasingly important endemic problem. Microbes Infect 2003;5:449–456. Vincent C, Boerlin P, Daignault D, Dozois CM, Dutil L, Galanakis C, et al. Food reservoir for Escherichia coli causing urinary tract infections. Emerg Infect Dis 2010; 16:88–95.
Address correspondence to: Amee R. Manges, MPH, PhD School of Population and Public Health University of British Columbia 137-2206 East Mall Vancouver, BC V6T 1Z3, Canada E-mail: [email protected]
Appendix Table A1. Complete List of Virulence Genes by Isolate Source Distribution in source Genes over-represented among human isolates
Gene
% Animal isolates
% Human isolates
% Total
Odds ratio
tspE vat neuA neuC irp(2) iha b1432 usp irp(1) fyuA gimB (orf1) ibeA kpsM-II papC malX chuA sitD sitA fepC papGII pic hlyA ompT agn iss mcbA iut(A2) deoK iron(2) iut(UPEC) iron iucD traT ompT(2) iutA iss(3)
0.039 0.026 0.013 0.013 0.182 0.013 0.013 0.052 0.182 0.184 0.013 0.013 0.130 0.052 0.130 0.247 0.286 0.286 0.260 0.039 0.026 0.091 0.403 0.247 0.338 0.013 0.169 0.221 0.169 0.182 0.169 0.169 0.169 0.403 0.156 0.545
0.627 0.507 0.239 0.239 0.833 0.225 0.220 0.517 0.809 0.799 0.163 0.163 0.684 0.397 0.641 0.789 0.804 0.780 0.756 0.244 0.172 0.373 0.794 0.579 0.679 0.048 0.421 0.483 0.397 0.416 0.383 0.378 0.349 0.603 0.282 0.670
0.469 0.378 0.178 0.178 0.657 0.168 0.164 0.392 0.640 0.635 0.122 0.122 0.535 0.304 0.503 0.643 0.664 0.647 0.622 0.189 0.133 0.297 0.689 0.490 0.587 0.038 0.353 0.413 0.336 0.353 0.325 0.322 0.301 0.549 0.248 0.636
41.43 38.59 23.90 23.90 22.37 21.76 21.45 19.51 19.01 17.61 14.77 14.77 14.52 12.02 11.97 11.45 10.24 8.86 8.83 7.96 7.80 5.95 5.73 4.20 4.16 3.82 3.58 3.30 3.24 3.21 3.05 2.99 2.64 2.25 2.13 1.69
95% CI 12.63 9.23 3.24 3.24 11.29 2.95 2.90 6.88 9.69 9.00 1.98 1.98 7.03 4.23 5.81 6.19 5.62 4.90 4.85 2.41 1.83 2.61 3.25 2.34 2.39 0.48 1.86 1.81 1.68 1.69 1.58 1.55 1.36 1.32 1.07 0.99
135.88 161.30 176.28 176.28 44.31 160.71 158.44 55.34 37.30 34.46 109.84 109.84 29.99 34.15 24.64 21.19 18.68 16.03 16.08 26.35 33.25 13.60 10.08 7.54 7.24 30.34 6.90 6.03 6.26 6.09 5.90 5.78 5.12 3.84 4.23 2.88
(continued)
8
MANGES ET AL.
Appendix Table A1. (Continued) Gene
% Animal isolates
% Human isolates
% Total
Odds ratio
Genes present at similar levels in animal and human isolates
rfc ce1A shf pilL f165(1)A papA(11) hraI fliC(H7) capU fliC ccdB cia bmaE cdtB(3) f17c-A b1121 eivG cvaC rtx fimA fimH cba virK tia cma cib eibA wzyO86 paa(ETEC) csgE cnf(2) eae(beta) paa
0.013 0.052 0.026 0.039 0.052 0.052 0.234 0.078 0.026 0.900 0.429 0.130 0.013 0.013 0.013 0.909 0.182 0.169 0.117 0.935 0.961 0.091 0.052 0.053 0.091 0.104 0.013 0.026 0.039 0.987 0.026 0.039 0.039
0.033 0.120 0.057 0.072 0.091 0.091 0.330 0.115 0.038 0.900 0.416 0.100 0.010 0.010 0.010 0.876 0.120 0.105 0.062 0.876 0.919 0.043 0.024 0.024 0.038 0.043 0.005 0.010 0.010 0.933 0.005 0.005 0.005
0.028 0.101 0.049 0.063 0.080 0.080 0.304 0.105 0.035 0.900 0.420 0.108 0.010 0.010 0.010 0.885 0.136 0.122 0.077 0.892 0.930 0.056 0.031 0.032 0.052 0.059 0.007 0.014 0.017 0.948 0.010 0.014 0.014
2.63 2.48 2.28 1.91 1.83 1.83 1.62 1.54 1.49 1.03 0.95 0.75 0.73 0.73 0.73 0.70 0.61 0.58 0.50 0.49 0.46 0.45 0.45 0.44 0.40 0.39 0.37 0.36 0.24 0.18 0.18 0.12 0.12
0.32 0.83 0.50 0.54 0.60 0.60 0.89 0.60 0.31 0.42 0.56 0.34 0.07 0.07 0.07 0.29 0.30 0.28 0.21 0.18 0.13 0.16 0.12 0.12 0.14 0.14 0.02 0.05 0.04 0.02 0.02 0.01 0.01
21.76 7.37 10.45 6.78 5.55 5.55 2.95 3.91 7.19 2.56 1.61 1.67 8.22 8.22 8.22 1.69 1.25 1.22 1.22 1.32 1.61 1.25 1.71 1.69 1.14 1.05 5.91 2.62 1.45 1.42 2.02 1.16 1.16
Genes overrepresented among animal isolates
lpfA yjaA astA(2) astA(1) lpfA(O113) gafD csgA flmA wzyO98 leoA lpfA(O157) tsh eprJ spaS tlrA toxB
0.169 0.779 0.234 0.230 0.558 0.078 0.987 0.221 0.273 0.052 0.052 0.506 0.792 0.753 0.187 0.753
0.081 0.526 0.077 0.060 0.187 0.014 0.909 0.033 0.043 0.005 0.005 0.072 0.211 0.096 0.005 0.057
0.105 0.594 0.119 0.100 0.287 0.031 0.930 0.084 0.105 0.017 0.017 0.189 0.367 0.273 0.053 0.245
0.44 0.31 0.27 0.20 0.18 0.17 0.13 0.12 0.12 0.09 0.09 0.08 0.07 0.03 0.02 0.02
0.20 0.17 0.13 0.09 0.10 0.04 0.02 0.05 0.05 0.01 0.01 0.04 0.04 0.02 0.00 0.01
0.95 0.58 0.57 0.44 0.32 0.71 1.00 0.31 0.28 0.80 0.80 0.15 0.13 0.07 0.16 0.04
Genes present in animal isolates only
eae escJ escN espA espG f17d-A ler nleA cif nleA(EHEC) nleB(O103)
0.078 0.078 0.078 0.078 0.078 0.078 0.078 0.078 0.052 0.052 0.052
0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
0.021 0.021 0.021 0.021 0.021 0.021 0.021 0.021 0.014 0.014 0.014
NE NE NE NE NE NE NE NE NE NE NE
Distribution in source
95% CI
(continued)
HUMAN AND FOOD ANIMAL E. COLI
9
Appendix Table A1. (Continued) Gene
% Animal isolates
% Human isolates
% Total
Odds ratio
espB-2 espB-3 etpD map(1) map(2) nleB(O157) nleE set tir-1 eae(gamma) nleA(EPEC) nleF tccP wzy(O146) wzy(O66)
0.039 0.039 0.039 0.039 0.039 0.039 0.039 0.039 0.039 0.026 0.026 0.026 0.026 0.026 0.026
0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
0.010 0.010 0.010 0.010 0.010 0.010 0.010 0.010 0.010 0.007 0.007 0.007 0.007 0.007 0.007
NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE
Genes present in human isolates only
senB sat cnf(1) sfaD z4184 papA(10) papGIII sfaHII wzxO6 kfiB papA(48) focG focA cdtB(1) papA(13) papA(16) papGIV kpsM-III sfaA papA(12) wb(O8) mcjA nfaE pixA afaE cdtB(4) facA papA(40) aap cka mccB papA(15) papA(8) papA(9) wzyO22 aatA wzyO15
0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
0.335 0.292 0.282 0.273 0.196 0.191 0.177 0.158 0.134 0.100 0.096 0.081 0.067 0.048 0.048 0.048 0.043 0.038 0.038 0.033 0.033 0.024 0.024 0.024 0.019 0.019 0.019 0.019 0.014 0.014 0.014 0.014 0.014 0.014 0.014 0.010 0.010
0.245 0.213 0.206 0.199 0.143 0.140 0.129 0.115 0.098 0.073 0.070 0.059 0.049 0.035 0.035 0.035 0.031 0.028 0.028 0.024 0.024 0.017 0.017 0.017 0.014 0.014 0.014 0.014 0.010 0.010 0.010 0.010 0.010 0.010 0.010 0.007 0.007
NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE NE
Common in both human and animal isolates
ompA hlyE mviM ibeB artJ gad mviN
1.000 1.000 1.000 1.000 1.000 1.000 1.000
0.866 0.895 0.904 0.957 0.971 0.981 0.986
0.902 0.923 0.930 0.969 0.979 0.986 0.990
NE NE NE NE NE NE NE
Distribution in source
The genes included in this table were present in at least two isolates within the entire dataset. NE, not estimatable.
95% CI
