ORIGINAL RESEARCH

published: 12 March 2015

doi: 10.3389/fcimb.2015.00023

A systematic review and

meta-analysis of the epidemiology of
pathogenic Escherichia coli of calves
and the role of calves as reservoirs
for human pathogenic E. coli
Rafał Kolenda 1 , Michał Burdukiewicz 2 and Peter Schierack 1*
1
Faculty of Natural Sciences, Brandenburg University of Technology Cottbus–Senftenberg, Senftenberg, Germany,
2
Department of Genomics, Faculty of Biotechnology, University of Wrocław, Wrocław, Poland

Escherichia coli bacteria are the most common causes of diarrhea and septicemia in
Edited by:
Nora Lía Padola, calves. Moreover, calves form a major reservoir for transmission of pathogenic E. coli
Universidad Nacional del Centro de la to humans. Systematic reviews and meta-analyses of publications on E. coli as calf
Provincia de Buenos Aires, Argentina
pathogens and the role of calves as reservoir have not been done so far. We reviewed
Reviewed by:
studies between 1951 and 2013 reporting the presence of virulence associated factors
Michael L. Vasil,
University of Colorado School of (VAFs) in calf E. coli and extracted the following information: year(s) and country of
Medicine, USA sampling, animal number, health status, isolate number, VAF prevalence, serotypes,
Analía Inés Etcheverría,
Universidad Nacional del Centro de la diagnostic methods, and biological assays. The prevalence of VAFs or E. coli pathotypes
Provincia de Buenos Aires, Argentina was compared between healthy and diarrheic animals and was analyzed for time
Teresa Estrada-Garcia,
courses. Together, 106 papers with 25,982 E. coli isolates from 27 countries tested for
Centro de Investigación y de Estudios
Avanzados del Instituto Politécnico VAFs were included. F5, F17, and F41 fimbriae and heat-stable enterotoxin (ST) – VAFs
Nacional, Mexico of enterotoxigenic E. coli (ETEC) were significantly associated with calf diarrhea. On the
*Correspondence: contrary, ETEC VAF F4 fimbriae and heat-labile enterotoxin as well as enteropathogenic
Peter Schierack,
Faculty of Natural Sciences, (EPEC), Shiga toxin-producing (STEC), and enterohemorrhagic E. coli (EHEC) were not
Brandenburg University of Technology associated with diarrhea. The prevalence increased overtime for ST-positive isolates, but
Cottbus–Senftenberg, Großenhainer
decreased for F5- and STEC-positive isolates. Our study provides useful information
Str. 57, D-01968 Senftenberg,
Germany about the history of scientific investigations performed in this domain so far, and helps
[email protected] to define etiological agents of calf disease, and to evaluate calves as reservoir hosts for
human pathogenic E. coli.
Received: 08 October 2014
Accepted: 23 February 2015 Keywords: calves, diarrhea, ETEC, EPEC, STEC, EHEC, systematic review, meta-analysis
Published: 12 March 2015

Citation:
Kolenda R, Burdukiewicz M and Introduction
Schierack P (2015) A systematic
review and meta-analysis of the Escherichia coli is one of the most common causes of diarrhea and septicemia in calves, affecting
epidemiology of pathogenic
dairy, and beef production. Testing for E. coli is part of the routine epidemiological examination
Escherichia coli of calves and the role
of calves as reservoirs for human
for diagnosis of calf diseases, alongside testing for rota- and coronavirus, Cryptosporidium spp., and
pathogenic E. coli. Salmonella spp. Nevertheless, diarrhea in calves remains a major cause of annual financial losses for
Front. Cell. Infect. Microbiol. 5:23. farmers. Moreover, as a very important reservoir of human pathogenic E. coli, calves can transmit
doi: 10.3389/fcimb.2015.00023 this pathogen to humans (Cobbold et al., 2007).

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org 1 March 2015 | Volume 5 | Article 23
Kolenda et al. Pathogenic Escherichia coli in calves

Since the end of the nineteenth century, the role of E. coli as of the intestinal microbiota of most mammals and birds. Only
a pathogen of calves has been of interest to the scientific com- a small fraction of the E. coli population belongs to pathovars
munity worldwide. Reports concerning calf colibacteriosis are and pathotyping is based on the occurrence of VAF and viru-
present in the works of Jensen (1903), Titze and Weichel (1908), lence mechanisms (intestinal pathogenic E. coli) (Croxen et al.,
Christiansen (1917); Carpenter and Woods (1924), and Smith 2013) or by occurrence of E. coli in organs and tissues which
and Orcutt (1925). The first scientific studies focused on calf diar- are sterile in healthy hosts (extraintestinal pathogenic E. coli, e.g.,
rhea of very young calves (up to the first 4 days), which is called uropathogenic E. coli). There are six major diarrheagenic E. coli
calf scours (in German: Kälberruhr, in French: diarhée des veaux). pathotypes (Croxen et al., 2013). Enterotoxigenic E. coli (ETEC)
The general outcome of these studies was that “Bacillus coli” is is the confirmed main causative agent of neonatal calf diar-
a main etiological factor of calf diarrhea. They divided the dis- rhea. Enteropathogenic E. coli (EPEC), Shiga toxin-producing
ease into enteritis with and without septicemia. Additionally, a E. coli (STEC including enterohemorrhagic E. coli/EHEC) are
dose-dependent protective role was described for the colostrum. also often isolated from diarrheic and healthy calves, but their
Colostrum-deprived calves or calves with delayed colostrum role in calf disease remains controversial. However, EPEC, STEC,
ingestion died after 1–3 days and large amounts of E. coli were and EHEC are important human pathogens for which cattle con-
isolated from their small and large intestines and other inter- stitute a major reservoir. Enteroaggregative E. coli (EAEC), dif-
nal organs. Calves fed with an insufficient protective dose of fusely adherent E. coli (DAEC), and enteroinvasive E. coli (EIEC;
colostrum developed various forms of bacterial infections, such including Shigella) are less known in cattle.
as arthritis and omphalitis, alone or together with diarrhea. On ETEC is characterized by the presence of specific adhesins
the other hand, calves receiving a sufficient dose of colostrum and toxins. Adhesins involved in farm animal infection are F4,
had the highest odds of staying healthy. The main characteris- F5, F6, F17, and F41 fimbriae, all encoded by fimbrial operons
tics of E. coli isolated from diarrheic calves were: higher counts (Nagy and Fekete, 1999, 2005). Relevant toxins are divided into
of bacteria from different parts of the intestine, and isolation of two groups: heat-stable enterotoxin I and II (STI and STII) and
bacteria from internal organs or pathologically changed parts of heat-labile enterotoxin I and II (LTI and LTII) (Fleckenstein et al.,
the body (e.g., joints) as compared to healthy controls. These 2010).
changes were proposed to stem from toxins released by bacteria EPEC as well as enterohemorrhagic E. coli (EHEC) form
that eventually altered the intestinal physiology. These assump- attaching and effacing (A/E) lesions (Lai et al., 2013). This mul-
tions were confirmed partially and extended in the later studies tistage process begins with initial attachment to the epithelial
of Smith (1962). However, he disputed the role of E. coli as a cell with surface associated filaments (EspA filaments) and bun-
causative agent of the disease in colostrum-fed calves, because dle forming pili (Bfp). In a next step these bacteria use a type
of the failures encountered in establishing experimental infection three secretion system to deliver several effector proteins includ-
and development of the disease in them. His later work, in which ing translocated intimin receptor (TIR) into the host cell followed
he developed a test for toxin production with calf ligated intesti- by intimate attachment, mediated via intimin (eaeA)-TIR bind-
nal loops and oral inoculation of calves, helped to elucidate this ing. Actin rearrangements then result in the formation of pedestal
issue (Smith and Halls, 1967). Using this model, Smith was able structures of A/E lesions (Clarke et al., 2003).
to induce fluid accumulation in intestinal loops and diarrhea in Shiga toxin-producing E. coli were first described as a patho-
6–20 h old infected calves. type 37 years ago (Konowalchuk et al., 1977). These bacterial
The development of new laboratory tools for the analysis strains are considered to be STEC if they are producing at least
of pathogenic E. coli allowed for the verification of hypotheses one of the Shiga toxins Stx1 or Stx2. The role of these bacte-
raised by pioneers in this field. The main goal was to find a pat- ria as calf pathogens has not been conclusively elucidated. There
tern for grouping isolates and to use this knowledge in disease have been cases of fatal STEC infections in cattle, despite their
diagnosis and prevention. Many approaches grouped bacteria lack of expression of globotriaosylceramide (Gb3), a receptor
by their reactivity with sera raised against standard strains. One for Stx cellular internalization, on their vascular endothelium
of the first attempts to O-serogroup isolates from calf diarrhea (Pruimboom-Brees et al., 2000). Recent studies show that recep-
was done by Bokhari and Ørskov (1952). Later, the antigen K99 tor binding is different between Stx1 and Stx2 and might involve
as virulence-associated factor (VAF) associated with the patho- more than one glycan (Gallegos et al., 2012). STEC that are able
genesis of neonatal calf diarrhea was defined by Orskov et al. to induce A/E lesions are grouped as enterohemorrhagic E. coli
(1975) and Guinée et al. (1976). Also, biochemical features of (EHEC). Cattles are frequent shedders of EHEC with approxi-
isolates were investigated extensively, but without any remark- mately 75% of human disease outbreaks linked to bovine derived
able success to specifically identify pathogenic E. coli. The emer- products or cattle (Nguyen and Sperandio, 2012).
gence of DNA- and protein-based assays for the detection of VAF EAEC was initially recognized by a specific adhesion pattern
opened the door for much faster and more accurate evaluations on HEp-2 cells forming cobble-stone-like aggregates (Weintraub,
of E. coli isolates (Grunstein and Hogness, 1975; Orskov et al., 2007). EAEC adherence is mediated by aggregative adherence
1977; Moseley et al., 1980; Burnette, 1981; Espy et al., 2006). fimbriae (AAF) coded on a virulence pAA plasmid. In addi-
The progress of science had a great impact on analytical typing tion, EAEC form mucoid biofilms and secret cytotoxins (e.g.,
methods for E. coli. It is now well accepted, that due to their high plasmid-encoded toxin Pet) that are toxic to epithelial cells. The
genotypic and phenotypic variation, E. coli can be subgrouped enterotoxins “enteroaggregative E. coli ST (EAST1)” and “Shigella
into many pathotypes. Commensalic E. coli is a common part enterotoxin 1 (ShET1)” were associated with EAEC, but they can

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org 2 March 2015 | Volume 5 | Article 23
Kolenda et al. Pathogenic Escherichia coli in calves

be found also in other pathotypes (Croxen and Finlay, 2010; Ruan Materials and Methods
et al., 2012).
DAEC are defined by their diffuse adherence (DA) pattern Literature Search and Eligibility Criteria
on HEp-2 or HeLa cells (Servin, 2005). DAEC express fimbriae The PubMed database was searched for studies published from
like Dr and F1845 and/or afimbrial adhesins (Afa) which are 1st of January 1951 to 31st of December 2013 with the following
responsible for adhesion to the epithelium and which are con- phrases: “E. coli calves,” “E. coli calves virulence genes,” “E. coli
sidered as the main VAFs for this pathotype. Adhesion results in calf,” and “E. coli calf virulence genes.” Manual revision was con-
an effacement of microvilli and a disruption of enzymes involved ducted on all displayed publications and first selections were
in intestinal secretion, which contribute to diarrhea. based on information in the titles and/or abstracts. Selected pub-
NTEC produce several powerful toxins. Originally, NTEC lications had to be available for downloading and had to contain
were mainly characterized by the secretion of cytotoxic necro- extractable data in English about the presence of VAF in E. coli
tizing factors (CNF 1, 2, 3) (Orden et al., 2007). Additionally isolated from calves. Selection of studies and extraction of data
they can produce toxins like hemolysin and cytolethal distending was done independently by the authors RK and MB and then
toxin (cdt). NTEC were associated with intestinal and extrain- compared. All discrepancies were reviewed by the third author
testinal infections in both humans and animals (Kaper et al., PS. The study selection workflow is presented in Figure 1.
2004).
Four types of hemolysin have been identified in E. coli (Lorenz Data Extraction
et al., 2013). Alpha-hemolysin (hlyA) is produced by many strains Data was extracted by RK and MB. The compiled information
associated with urinary tract infections. Phage-carried enterohe- contains: year(s) and country of sampling, number of animals
molysin (ehxA) is frequently associated with the STEC pathotype. sampled, presence of diarrhea, number of isolates analyzed, num-
The role of the bacteriophage carried enterohemolysin (e-hlyA) ber of isolates positive for VAFs, serotypes, diagnostic meth-
is so far not known. Presence of the silent hemolysin (sheA) ods, and biological assays. Isolates were divided into four groups
has been confirmed in most of the E. coli pathotypes. Alpha- depending on the animals’ health status: diarrheic, healthy, mixed
hemolytic E. coli were frequently isolated from healthy as well as (studies did not distinguish between diarrheic and healthy),
diseased animals and humans and their specific physiological or unknown. Collected raw data was stored in a relational database
pathophysiological role remains unclear (Schierack et al., 2006, created specifically for this study.
2011).
Common diagnostic methods for pathotyping E. coli rely on Data Preprocessing
the detection of genes/gene products and/or proteins which focus Due to the non-homogeneous nomenclature the naming con-
on toxins and adhesion factors. Biological assays are occasion- vention of VAFs was unified. For example, isolates originally
ally added to investigate and test for virulence mechanisms or annotated as eae, eaeA, and/or intimin were re-coded as eaeA-
to quantify biological activities of VAFs. One popular method positive. Another reason for unification was the diversity in ana-
for the prediction of pathogenicity is typing of somatic (O) and lyzed genes. In order to improve data processing we aggregated
flagellar (H) antigens. Actually, there are 181 O and 53 H anti- subtypes of genes as one gene (e.g., all Stx2e and Stx2 are grouped
gens (Lacher et al., 2014). Several O groups are more prevalent in as Stx2). These two aforementioned modifications are shown in
single pathotypes (Ewers et al., 2014). Other tools for the charac- Supplementary Table 1. Whenever isolates were collected over
terization of isolates are various genotyping methods e.g., pulsed more than 1 year, a consensus year was the average between the
field gel electrophoresis (PFGE), random amplified polymorphic year of beginning and end of the gathering of isolates.
DNA (RAPD), multilocus sequence typing (MLST), multiple loci
VNTR analysis (MLVA), and the Clermont method (Clermont Missing Data Imputation
et al., 2000; Foley et al., 2004). Forty-one (39%) publications did not contain precise informa-
Over the last decades numerous studies were published about tion about the time frame in which isolates were collected. The
E. coli in calves describing single events in circumscribed areas probable year of collection for such a study was approximated
with controversial results. However, systematic reviews and even by the linear model y = a ∗ x fitted to data from all publica-
meta-analyses of previous studies are still rare endeavors, reflect- tions with clearly indicated collection periods. The isolation year
ing a trend in the field of veterinary medicine. Such meta-analyses was treated as the response variable and the publication year was
are strong tools to clarify the role of bacteria in disease of a spe- used as the explanatory variable. This approximation resulted in
cific host group, the role of a host species as reservoir for other the factor of 0.9977: isolation_year = publication_year ∗ 0.9977
species and epidemiological trends which can help in the evalua- (p-value: <2.2e-16, adjusted R2 = 1).
tion of disease defense strategies (e.g., vaccination), optimization
of existing diagnostic protocols and indication of new research Pathotyping
topics. Our meta-analysis reviews all published and available data We analyzed data as follows: (1) Data for all VAFs were analyzed
between 1951 and 2013 concerning the presence of VAFs in E. coli regarding the prevalence of single VAFs with no mentioned affili-
isolated from diarrheic and healthy calves and discusses the role ation to VAF patterns (e.g., 54% of all isolates were eaeA-positive,
of E. coli pathotypes in disease, the role of calves as pathogen 37% of all isolates were Stx-positive, but with no information
reservoir and the epidemiological trends in this field over the past about the prevalence of EHEC = Stx + eaeA). (2) For EPEC,
decades. STEC, and EHEC we analyzed and presented the prevalence of

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org 3 March 2015 | Volume 5 | Article 23
Kolenda et al. Pathogenic Escherichia coli in calves

FIGURE 1 | Flow diagram for the selection of manuscripts.

single VAFs and affiliation of all isolates to possible VAF pat- status. Correlations between the prevalence of a VAF and the year
terns (e.g., one isolate was eaeA-positive = EPEC, one isolate was of isolation of E. coli was determined with the Pearson correla-
Stx-positive = STEC, one isolate was eaeA- and Stx-positive = tion coefficient on normally distributed data. The false discovery
EHEC). This was possible due to a large number of relevant rate of all statistical tests was controlled by applying Benjamini
publications. and Hochberg correction to all obtained p-values (Benjamini and
Hochberg, 1995). Confidence intervals of the prevalence for each
Data Processing and Statistical Analysis VAF were estimated using the Wilson score interval with con-
The prevalence of VAFs or E. coli pathotypes was compared tinuity correction. This method was chosen because of its good
between healthy and diarrheic animals using the chi-square test coverage properties (Brown et al., 2001).
of independence with Yates’s continuity correction. The same This study was conducted following the guidelines for report-
statistical test was used to determine relationships between the ing meta-analysis of observational studies in epidemiology
prevalence of ETEC, EPEC, STEC, EHEC, and the animal health (MOOSE) (Stroup et al., 2000) and preferred reporting items

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org 4 March 2015 | Volume 5 | Article 23
Kolenda et al. Pathogenic Escherichia coli in calves

for systematic reviews and meta-analyses (PRISMA) statement publications, respectively. F4 was not found in analyzed samples
(Moher et al., 2009) with PRISMA 2009 Checklist (Supplemen- (810 isolates tested). F6 fimbriae were detected in 12.3% (82 of
tary File 1). 666) of diarrheic isolates, but no data about prevalence in healthy
All calculations were performed using the R software animals was available.
(R Development Core Team, 2013). All figures were made using The prevalence of ST-positive isolates in diarrheic and healthy
the ggplot2 package (Wickham, 2010) implemented in the R calves increased over time (p < 0.05). F5 prevalence decreased
programming environment. over time in diseased and healthy animals (p < 0.0005), while
prevalence did not change for the other VAFs within this time
Results (Figure 2; Supplementary Figure 3).

Our systematic review and meta-analysis includes data from 106 Enteropathogenic Escherichia coli
studies from 2552 citations identified in the Pubmed database The presence of EPEC VAFs (one eaeA gene + absence of Shiga
(Supplementary File 2). These studies represent data about bac- toxins or Shiga toxin genes) was investigated in 12,246 isolates
teria isolated in 27 countries (Supplementary Figure 1) and pub- from 18 countries in 37 publications.
lished in 37 journals (Supplementary Figure 2). The results of EPECs were found more often in healthy than in diarrheic ani-
our search are summarized in Supplementary Table 2. In all, we mals with 7.5% of isolates from diarrheic animals (709 from 9448
gathered information from 25,982 isolates tested for VAFs. isolates) and 14.6% of isolates from healthy animals (383 from
2629 isolates) being EPEC (Table 3).
Methods for Characterization of Isolates
The methods used for VAF detection in E. coli are summarized in Shiga Toxin-Producing Escherichia coli
Supplementary Table 3. PCR was used most often for VAF detec- The presence of STEC VAFs was investigated in 8053 isolates
tion. More than a half of papers contained information about from 19 countries in 61 publications (Table 3). The average
serotypes. Clonal relationship assays and antimicrobial suscep- prevalence of STECs isolated from healthy animals was 19.4%
tibility tests were analyzed in nearly one-fifth of all publications and from diseased animals 18.2%. Only 51 publications includ-
under review. ing 5676 isolates determined co-occurrence of toxins: 344 isolates
were Stx1-positive (6.1%), 405 isolates were Stx2+ (7.1%), and in
VAF: Overall Prevalence and Association with 259 isolates both Shiga toxin genes were present (4.6%). STEC
Health Status was not associated with diarrhea. STEC prevalence decreased
The database covers information about 61 VAFs (genes, operons, over time (p-value < 0.001, Figure 3). Interestingly, STEC preva-
and/or expressed antigens). Each publication presented informa- lence is remarkably smaller in contemporary than in older studies
tion about 1–27 VAFs (median = 3.5 VAFs per publication). Only (Figure 4).
the most prevalent and pathotype-associated VAFs were chosen
for further analysis and discussion. However, all other VAFs have Enterohemorrhagic Escherichia coli
also been listed in Table 1 and Supplementary Table 4. In addi- The presence of EHEC VAFs (one eaeA gene + Shiga toxins
tion, the relation between particular VAFs and an animal’s health or Shiga toxin genes) was investigated in 13,826 isolates from
status is shown in Table 1. In summary 16 VAFs were associated 18 countries in 50 publications. The prevalence of EHEC inde-
with animals’ health status with p-value lower than 0.05 (lowest p pendent of health status was 8.7% (1197 isolates were EHEC) with
in these group 10−52 ). Identification of five VAFs was associated 6% in diarrheic and 10.7% in healthy calves (p < 10−15 , Table 3).
with healthy calves, while 11 of them with diseased. Forty four papers including 12,378 isolates differentiated between
Stx1 and Stx2: 659 isolates were Stx1 + eaeA-positive, 180 iso-
Enterotoxigenic Escherichia coli lates were Stx2 + eaeA-positive and 118 isolates were Stx1 +
The presence of ETEC VAFs was investigated in 9597 isolates Stx2 + eaeA-positive. EHEC prevalence did not decreased over
from 27 countries in 51 publications. ST and LT were present in time (Figures 3, 4).
4.3% (254/5849 isolates) and 0.3% (7/2672) isolates, respectively
(Table 2). Other VAFs
ST was found in 7% of isolates from diseased and in 0.3% from Afa and F1845 are typical for DAEC. The presence of DAEC
healthy animals. STI (4.9%) was isolated more often than STII VAFs was investigated in 1568 isolates from 10 countries in nine
(3.2%). LT was almost absent in isolates from both diarrheic and publications. Afa was found significantly more often in isolates
healthy calves. EAST was found with similar prevalence in iso- from diarrheic (30.6% of isolates) as compared to isolates from
lates from healthy (32.1%) and diarrheic calves (30.4%, Table 1). healthy animals (1.3%; p < 10−6 ). F1845 was investigated only in
The prevalence of various ETEC adhesins is shown in Table 2. two publications and three from 136 isolates (2.2%) were pos-
F5, F17, and F41 fimbriae were the most frequently studied. They itive (Table 1). Aggregative adhesion fimbria (aaf) expression
were isolated from 9.5, 30.4, and 11.1% animals, respectively. All in EAEC was not found in any of 147 isolates tested in three
of them were found more often in diarrheic, than in healthy ani- publications from three countries.
mals (Table 1). F5, F17, and F41 fimbriae were isolated 5.7, 1.2, The presence of the cytotoxic necrotizing factor (CNF) was
and 23.7 times more frequently from diarrheic calves than from investigated in 2125 isolates from nine countries in 14 publica-
healthy ones. F4 and F6 were tested for only in four and two tions. CNF was reported 1.8 times more often in isolates from

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org 5 March 2015 | Volume 5 | Article 23
Kolenda et al. Pathogenic Escherichia coli in calves

TABLE 1 | Comparison of study groups (healthy and diarrheic).

Group Diarrheic Healthy P-value

VAF name Number of Isolates Isolates Proportion of Number of Isolates Isolates Proportion of
publications positive tested positives publications positive tested positives

aaf 3 0 155 0.0000 1 0 56 0.0000 NA

Aerobactin 5 114 174 0.6552 1 49 56 0.8750 0.0085
Afa 7 423 1384 0.3056 2 1 79 0.0127 2.356e-07
bfp 6 3 353 0.0085 3 0 117 0.0000 0.9756
cdt 6 78 1271 0.0614 2 7 77 0.0909 0.6468
cif 1 15 255 0.0588 1 0 21 0.0000 0.7666
CNF 10 135 1880 0.0718 6 55 1356 0.0406 9.089e-04
CS31 7 183 646 0.2833 2 1 145 0.0069 1.404e-11
eaeA 33 1203 9779 0.1230 29 856 4295 0.1993 7.766e-31
EAF 3 3 160 0.0187 2 0 61 0.0000 0.934
EAST 5 45 148 0.3041 1 18 56 0.3214 1
EFA1 1 15 16 0.9375 1 23 23 1.0000 0.9756
eibG 1 0 16 0.0000 1 0 23 0.0000 NA
ent 1 0 61 0.0000 1 0 56 0.0000 NA
escV 1 14 61 0.2295 1 16 56 0.2857 0.9005
F17 15 735 2328 0.3157 4 231 854 0.2705 0.0437
F1845 1 0 14 0.0000 1 3 122 0.0246 1
F41 15 293 2030 0.1443 6 6 991 0.0061 2.521e-31
F5 36 719 5566 0.1292 12 42 1855 0.0226 1.580e-37
fim 1 60 61 0.9836 1 56 56 1.0000 1
fyuA 2 36 76 0.4737 1 29 56 0.5179 0.9756
hcp 1 3 16 0.1875 1 0 23 0.0000 0.2138
hly 25 455 1661 0.2739 15 283 1368 0.2069 9.274e-05
ibe 1 45 61 0.7377 1 39 56 0.6964 0.9756
iha 2 21 58 0.3621 1 23 23 1.0000 3.281e-06
inv 2 0 75 0.0000 1 0 56 0.0000 NA
ipa 3 0 107 0.0000 1 0 56 0.0000 NA
kpsMII 2 4 103 0.0388 1 1 56 0.0179 0.9756
ldaE 1 7 16 0.4375 1 0 23 0.0000 0.0069
LT 5 3 1141 0.0026 1 0 56 0.0000 1
LTI 7 0 409 0.0000 3 0 806 0.0000 NA
LTII 3 4 146 0.0274 1 0 20 0.0000 1
malX 2 3 103 0.0291 1 5 56 0.0893 0.3334
paa 1 15 16 0.9375 1 23 23 1.0000 0.9756
Pap 7 299 1248 0.2396 1 42 112 0.3750 0.0071
pic 2 0 77 0.0000 1 0 56 0.0000 NA
saa 4 119 248 0.4798 3 60 863 0.0695 9.715e-52
sfa 4 37 998 0.0371 1 3 56 0.0536 0.9756
STI 17 172 2178 0.0790 8 7 1261 0.0056 1.531e-19
STII 7 50 1043 0.0479 3 0 806 0.0000 3.924e-09
stx 4 27 94 0.2872 2 20 130 0.1538 0.0584
stx1 34 745 3562 0.2092 31 677 3577 0.1893 0.0877
stx2 34 402 3562 0.1129 31 654 3577 0.1828 7.275e-16
toxB 1 11 16 0.6875 1 22 23 0.9565 0.1343
traT 1 58 61 0.9508 1 54 56 0.9643 1
uidA 1 59 61 0.9672 1 51 56 0.9107 0.5763
etpD NA NA NA NA 1 2 13 0.1538 NA
aap 1 0 16 0.0000 NA NA NA NA NA

(Continued)

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org 6 March 2015 | Volume 5 | Article 23
Kolenda et al. Pathogenic Escherichia coli in calves

TABLE 1 | Continued

Group Diarrheic Healthy P-value

VAF name Number of Isolates Isolates Proportion of Number of Isolates Isolates Proportion of
publications positive tested positives publications positive tested positives

esp 3 32 212 0.1509 NA NA NA NA NA

F18 1 0 14 0.0000 NA NA NA NA NA
F4 3 0 716 0.0000 NA NA NA NA NA
F6 1 82 666 0.1231 NA NA NA NA NA
HPI 1 27 42 0.6429 NA NA NA NA NA
iroN 1 4 42 0.0952 NA NA NA NA NA
modD 1 4 42 0.0952 NA NA NA NA NA
shf 1 3 16 0.1875 NA NA NA NA NA

NA, not applicable; P-value, value for difference between healthy and diarrheic.

Therefore, the present study was undertaken to summarize data

TABLE 2 | Data about ETEC VAFs.
from over 60 years of research in this field. We included all eligi-
Group Number of Isolates Isolates Percent ble studies from the Pubmed database. Data were extracted and
name publications positive tested positive stored in a relational database created specifically for this research
and analyzed with R software (R Development Core Team, 2013).
F4 4 0 810 0.0 ETECs are regarded as major agents in the etiology of calf
F5 44 795 8346 9.5 diarrhea. One of the first steps of ETEC virulence mechanisms
F6 2 82 760 10.8 is adhesion to the intestinal epithelium, which is mediated by
F17 17 966 3182 30.4 fimbrial adhesins, mainly the F4, F5, F6, F17, and F41 fimbriae.
F41 19 391 3515 11.1 According to our analysis the presence of F4 is not associated
LT 6 3 1291 0.2 with diarrhea. F4 fimbriae-positive strains were first reported in
LTI 9 0 1215 0.0 pigs and are the most common VAF in porcine ETEC. There-
LTII 3 4 166 2.4 fore, absence of F4 in calf isolates is not surprising. The role
LT overall 15 7 2672 0.3 of F6 fimbriae in the etiology of calf diarrhea remains unclear,
STI 22 192 3906 4.9 owing to the lack of information about healthy animals. F17 have
STII 10 62 1943 3.2 been implicated in the pathogenesis of calf diarrhea and we show
ST overall 22 254 5849 4.3 that F17 is found more often in diarrheic than in healthy calves
(p < 0.05); yet, there is high prevalence of F17 positive isolates
in healthy animals (27%). We propose two explanations: Firstly,
the F17 fimbrium requires the presence of other VAFs to partici-
diarrheic (7.2%) than in isolates from healthy animals (4.1%; pate in the etiology of diarrhea. Secondly, F17 fimbriae genes are
p < 0.001). E. coli hemolysins (hly, including all four types) were present in E. coli, as detected by the most common method PCR,
more often reported in isolates from diarrheic animals (27.4%) but are not functionally expressed. Unfortunately, resolving this
than in isolates from healthy animals (20.7%; p < 10−4 ). issue requires more in-depth research.
F5 and F41 fimbriae were highly associated with the presence
Biological Assays of diarrhea (p < 10−36 and p < 10−30 ). F41 prevalence has the
Phenotypic characterization can improve pathotyping of E. coli. highest diarrheic to healthy ratio, which justifies a major role of
In vitro cell-based or ex vivo assays are the primary choice for this this pilus in ETEC pathogenesis. F5 fimbriae were the first VAF
purpose. In Supplementary Table 5 we summarize such biologi- identified in ETEC from diarrheic isolates (Guinée et al., 1976).
cal assays. The most often used was the Vero cytotoxicity assay, They are also the most often investigated VAF in our ETEC meta-
which is found in 27 manuscripts. Cells or cell lines were rarely analysis. This gave us the opportunity to show changes in preva-
used to investigate adhesion/adhesion pattern of E. coli. Entero- lence of F5 over time. A significant drop in F5 prevalence over the
toxin production was done in 10 publications for which ligated past years was observed. This outcome can be explained by the
intestinal loops or mouse infant assays were used. commonness of vaccination against this ETEC antigen (Moon
and Bunn, 1993; Crouch et al., 2001), which could have led to neg-
Discussion ative selection against F5 fimbriae-positive bacteria coincident
with success in disease control.
Systematic epidemiological reviews and meta-analyses on publi- After attachment to the intestinal mucosa, ETEC initiate their
cations of E. coli as calf pathogen and the role of calves as human pathogenic actions by secretion of toxin(s). It has been proposed,
pathogenic E. coli reservoir have not been carried out so far. that ST plays a major role in the pathogenesis of calf diarrhea

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org 7 March 2015 | Volume 5 | Article 23
Kolenda et al. Pathogenic Escherichia coli in calves

(Nagy and Fekete, 1999). Indeed, ST toxin was detected fre-

quently only in diarrheic isolates, while LT was reported only in
seven of 2672 isolates. The role of EAST toxin in the pathogen-
esis of calf diarrhea seems to be doubtful. Although the overall
prevalence of EAST was high, it was represented in diarrheic and
healthy samples in similar frequencies. Recently, published data
by Ruan et al. (2012) indicated that an EAST1-positive isolate
alone is not able to cause diarrhea in a 5-day old gnotobiotic pig
model, and suggests that EAST1 is not likely a VAF in ETEC-
associated diarrhea. Additionally, EAST1 was detected with high
prevalence in clinically healthy individuals of other domestic and
wild animal species (Römer et al., 2012; Frömmel et al., 2013)
which also puts to doubt whether EAST1 has a prominent role
in diarrhea.
The readers of this manuscript should raise an important
question at this moment. If the highest prevalence of VAF from
E. coli isolated from diarrhea is between 5 and 15%, what is
happening with the rest of isolates? Are those isolates commen-
salic E. coli or are there still many unidentified adhesins and
toxins which may contribute to pathogenesis? E. coli can be
isolated from each diarrheic sample but might not be the ade-
quate causative agent of disease. A favorable differential diag-
nosis of infectious agents of calf diarrhea consists of testing
for rota- and coronavirus, Salmonella spp., and Cryptosporid-
ium spp.. Unfortunately, information about such investigations
was sparse in the context of our meta-analysis. Only 22 publica-
tions mentioned additional testing for at least one of the above
mentioned pathogens. Other important causes of diarrhea are
non-infectious, and related to housing, management, and feed-
ing routines (Ortiz-Pelaez et al., 2008). Isolation of pathogenic
E. coli from most of these cases should be rare and are not associ-
ated with disease. The inclusion of a questionnaire into relevant
studies could help in obtaining valuable information on how the
environment influences animal health, but such additional data
were mentioned only in 15 publications.
The next interesting topic raised in our work is the prevalence
of STEC, EPEC, and EHEC in calves. The high prevalence
of STEC, EPEC, and EHEC in healthy and diarrheic animals
shows, that these animals can be considered as reservoirs for
human pathogenic E. coli. The role of STEC, EPEC, and EHEC
as causative agents of diarrhea in calves have been widely
investigated (Dean-Nystrom et al., 1998; Menge et al., 1999;
Pruimboom-Brees et al., 2000; Hoey et al., 2002). Lack of Stx
receptor in vascular endothelium makes cattle a perfect candi-
date as a reservoir of STEC/EHEC as it is not possible for them
to develop systemic disease. Other actions of STEC/EHEC like
immunomodulation (Menge et al., 1999) and intestinal coloniza-
tion (Etcheverría and Padola, 2013) help them to survive and
propagate in the host intestine. Major questions remain to be
answered as to whether these processes contribute to pathogen-
esis of diarrhea. Our comparison of the prevalence of STEC,
FIGURE 2 | ETEC VAF prevalence in between 1962 and 2013. Each EPEC, and EHEC in diarrheic and healthy calves showed that
subpanel corresponds to one VAF. The color of dots describes health status of none of these pathotypes is associated with disease, and these
animals: red, diarrheic animals; blue, healthy animals. The lines show general pathotypes were more often detected in healthy animals. But
tendencies by approximating the relationship between the year of isolation and
can cattle have any profit from the presence of these bacteria?
the prevalence of a VAF using a linear model y = ax + b. The shaded areas
around lines represent confidence intervals.
An interesting hypothesis was raised in the works of Ferens and
Hovde (2000), Basu et al. (2003) and Ferens et al. (2004, 2006),

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org 8 March 2015 | Volume 5 | Article 23
Kolenda et al. Pathogenic Escherichia coli in calves

TABLE 3 | Data about EPEC, STEC, EHEC.

Pathotype EPEC STEC EHEC

Group Isolates Isolates Number of Isolates Isolates Number of Isolates Isolates Number of
tested positive publications tested positive publications tested positive publications

Diarrheic 9448 709 27 3248 592 36 9014 545 31

Healthy 2629 383 19 3147 610 28 4120 442 27
Mixed 0 0 0 130 66 1 130 64 1
Unknown 169 0 3 1528 386 10 562 146 5
Total 12246 1092 37 8053 1654 61 13826 1197 50

FIGURE 3 | STEC, EPEC, EHEC prevalence in between 1986 and 2010. lines show general tendencies by approximating the relationship between the
Each subpanel corresponds to one pathotype. The color of dots describes year of isolation and the prevalence of a pathotype using a linear model y =
health status of animals: red, diarrheic animals; blue, healthy animals. The ax + b. The shaded areas around lines represent confidence intervals.

that Stx production can mitigate bovine leukemia virus (BLV)- DAEC and EAEC are well-known as a diarrhea causing
induced disease in cattle. This hypothesis could explain the strong agents in humans (Servin, 2005; Weintraub, 2007), but there is
species-pathotype association, but to our knowledge no survey scant information about their prevalence in cattle and calves.
study about BLV-STEC/EHEC prevalence correlation has been In the case of DAEC, data from nine studies show that the
conducted so far. prevalence of this pathotype is similar to the prevalence of
Another intriguing observation is the decrease in the preva- ETEC VAFs. Therefore, it seems that cattle might be affected by
lence of STEC over time as shown in Figures 3, 4. We assumed DAEC. Attention should be paid on EAEC and DAEC in next
that prevalence might be influenced by the diagnostic meth- years.
ods used. To verify this, we created additional forest plots with Analysis of the methods used in studies appraised in this
information about methods used in the study (Supplementary work is worth a small separate discussion. Supplementary Fig-
Figure 4). The analysis of this results shows, that our assumption ure 5 shows high impact of DNA-based methods on patho-
was wrong. typing. As these tools are fast and reliable, an additional

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org 9 March 2015 | Volume 5 | Article 23
Kolenda et al. Pathogenic Escherichia coli in calves

FIGURE 4 | Prevalence of STEC, EPEC, EHEC in the manuscripts healthy; D, diarrheic; M, mixed; U, unknown) and the year of isolation (with
under review. The dashed vertical lines represent the mean prevalence of the asterisk if isolation date was extrapolated). Horizontal lines represent the
the given pathotype of all studies. Labels contain information about the first 0.95 confidence intervals. Point size represents number of isolates for a
author of the study, the year of publication, the health status of animals (H, pathotype. Data were sorted according to the year of isolation.

confirmation for pathogenic phenotypes using phenotyping Our study provides useful information about the history of scien-
assays seems to be possible and useful. However, such assays tific investigations performed so far and helps understand their
were rarely applied. Newly developed technologies seem to be influence on forthcoming studies. Our data suggest new trends
an answer for the need of quick, reliable and high-throughput for future work concerning E. coli in calves as an etiological agent
functional analysis of isolated E. coli (Han and Lee, 2006; of disease or in carrier state. We hope that our study will high-
Sorek and Cossart, 2010; Frömmel et al., 2013; Rödiger et al., light the need to develop a unified framework for future research
2013). on pathogenic E. coli of calves.
It is our duty to mention the drawbacks of our work. Due to
the lack of reports about several VAFs (e.g., F6), performance Acknowledgments
of planned analyses was not possible. Unfortunately in many
studies different sets of genes and health status of animals were This work was supported by the Federal Ministry of Edu-
investigated, therefore presented analyses for each VAF are com- cation and Research, Germany [Project InnoProfile-Transfer
posed of sets of various number of isolates. The atypical structure 03IP611X].
of data prohibited fitting with more advanced statistical models
used often in meta-analysis. One of our goals was to evaluate the Supplementary Material
spatial distribution of VAFs and pathotypes, but it was not feasi-
ble due to the geographical sparseness of the data. Still, we think The Supplementary Material for this article can be found
that our systematic review and meta-analysis is a good way to online at: http://www.frontiersin.org/journal/10.3389/fcimb.
summarize epidemiological status of pathogenic E. coli in calves. 2015.00023/abstract

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org 10 March 2015 | Volume 5 | Article 23
Kolenda et al. Pathogenic Escherichia coli in calves

References enterotoxigenic Escherichia coli infection. Microbes Infect. 12, 89–98. doi:
10.1016/j.micinf.2009.10.002
Basu, I., Ferens, W. A., Stone, D. M., and Hovde, C. J. (2003). Antiviral activity of Foley, S. L., Simjee, S., Meng, J., White, D. G., McDermott, P. F., and Zhao, S.
shiga toxin requires enzymatic activity and is associated with increased perme- (2004). Evaluation of molecular typing methods for Escherichia coli O157:H7
ability of the target cells. Infect. Immun. 71, 327–334. doi: 10.1128/IAI.71.1.327- isolates from cattle, food, and humans. J. Food Prot. 67, 651–657.
334.2003 Frömmel, U., Böhm, A., Nitschke, J., Weinreich, J., Groß, J., Rödiger, S., et al.
Benjamini, Y., and Hochberg, Y. (1995). Controlling the false discovery rate: a (2013). Adhesion patterns of commensal and pathogenic Escherichia coli from
practical and powerful approach to multiple testing. J. R. Stat. Soc. B (Methodol.) humans and wild animals on human and porcine epithelial cell lines. Gut
57, 289–300. Pathog. 5:31. doi: 10.1186/1757-4749-5-31
Bokhari, S. M. H., and Ørskov, F. (1952). O-grouping of E. coli strains isolated Gallegos, K. M., Conrady, D. G., Karve, S. S., Gunasekera, T. S., Herr, A. B., and
from cases of white scours. Acta Pathol. Microbiol. Scand. 30, 87–89. doi: Weiss, A. A. (2012). Shiga toxin binding to glycolipids and glycans. PLoS ONE
10.1111/j.1699-0463.1952.tb00164.x 7:e30368. doi: 10.1371/journal.pone.0030368
Brown, L. D., Cai, T. T., and DasGupta, A. (2001). Interval estimation for a Grunstein, M., and Hogness, D. S. (1975). Colony hybridization: a method for the
binomial proportion. Stat. Sci. 16, 101–133. doi: 10.1214/ss/1009213286 isolation of cloned DNAs that contain a specific gene. Proc. Natl. Acad. Sci.
Burnette, W. N. (1981). “Western Blotting”: electrophoretic transfer of proteins U.S.A. 72, 3961–3965. doi: 10.1073/pnas.72.10.3961
from sodium dodecyl sulfate-polyacrylamide gels to unmodified nitrocellulose Guinée, P. A., Jansen, W. H., and Agterberg, C. M. (1976). Detection of the K99
and radiographic detection with antibody and radioiodinated protein A. Anal. antigen by means of agglutination and immunoelectrophoresis in Escherichia
Biochem. 112, 195–203. doi: 10.1016/0003-2697(81)90281-5 coli isolates from calves and its correlation with entertoxigenicity. Infect.
Carpenter, C., and Woods, G. (1924). The distribution of the colon-aerogenes Immun. 13, 1369–1377.
group of bacteria in the alimentary tract of calves. Cornell Vet. 14, 218–225. Han, M.-J., and Lee, S. Y. (2006). The Escherichia coli proteome: past,
Christiansen, M. (1917). Bakterien der Typhus-coli-Gruppe im Darm von gesun- present, and future prospects. Microbiol. Mol. Biol. Rev. 70, 362–439. doi:
den Spankkalbern und bei deren Darminfektion. Zbl. Bakt. 79, 196. 10.1128/MMBR.00036-05
Clarke, S. C., Haigh, R. D., Freestone, P. P. E., and Williams, P. H. (2003). Virulence Hoey, D. E. E., Currie, C., Else, R. W., Nutikka, A., Lingwood, C. A., Gally, D. L.,
of enteropathogenic Escherichia coli, a global pathogen. Clin. Microbiol. Rev. 16, et al. (2002). Expression of receptors for verotoxin 1 from Escherichia coli O157
365–378. doi: 10.1128/CMR.16.3.365-378.2003 on bovine intestinal epithelium. J. Med. Microbiol. 51, 143–149.
Clermont, O., Bonacorsi, S., and Bingen, E. (2000). Rapid and simple determina- Jensen, C. O. (1903). Om den infektiese Kalverdiarrhoe og dens Aarsag.
tion of the Escherichia coli phylogenetic group. Appl. Environ. Microbiol. 66, Maanedsskrift Dyrlaeger 4, 140–145.
4555–4558. doi: 10.1128/AEM.66.10.4555-4558.2000 Kaper, J. B., Nataro, J. P., and Mobley, H. L. (2004). Pathogenic Escherichia coli.
Cobbold, R. N., Hancock, D. D., Rice, D. H., Berg, J., Stilborn, R., Hovde, Nat. Rev. Microbiol. 2, 123–140. doi: 10.1038/nrmicro818
C. J., et al. (2007). Rectoanal junction colonization of feedlot cattle by Konowalchuk, J., Speirs, J. I., and Stavric, S. (1977). Vero response to a cytotoxin
Escherichia coli O157:H7 and its association with supershedders and excre- of Escherichia coli. Infect. Immun. 18, 775–779.
tion dynamics. Appl. Environ. Microbiol. 73, 1563–1568. doi: 10.1128/AEM.01 Lacher, D. W., Gangiredla, J., Jackson, S. A., Elkins, C. A., and Feng, P. C. H. (2014).
742-06 A novel microarray design for molecular serotyping of Shiga toxin-producing
Crouch, C. F., Oliver, S., and Francis, M. J. (2001). Serological, colostral and milk Escherichia coli isolated from fresh produce. Appl. Environ. Microbiol. 80,
responses of cows vaccinated with a single dose of a combined vaccine against 4677–4682. doi: 10.1128/AEM.01049-14
rotavirus, coronavirus and Escherichia coli F5 (K99). Vet. Rec. 149, 105–108. Lai, Y., Rosenshine, I., Leong, J. M., and Frankel, G. (2013). Intimate host
doi: 10.1136/vr.149.4.105 attachment: enteropathogenic and enterohaemorrhagic Escherichia coli. Cell.
Croxen, M. A., and Finlay, B. B. (2010). Molecular mechanisms of Escherichia coli Microbiol. 15, 1796–1808. doi: 10.1111/cmi.12179
pathogenicity. Nat. Rev. Microbiol. 8, 26–38. doi: 10.1038/nrmicro2265 Lorenz, S. C., Son, I., Maounounen-Laasri, A., Lin, A., Fischer, M., and Kase,
Croxen, M. A., Law, R. J., Scholz, R., Keeney, K. M., Wlodarska, M., and Finlay, J. A. (2013). Prevalence of hemolysin genes and comparison of ehxA sub-
B. B. (2013). Recent advances in understanding enteric pathogenic Escherichia type patterns in Shiga toxin-producing Escherichia coli (STEC) and non-STEC
coli. Clin. Microbiol. Rev. 26, 822–880. doi: 10.1128/CMR.00022-13 strains from clinical, food, and animal sources. Appl. Environ. Microbiol. 79,
Dean-Nystrom, E. A., Bosworth, B. T., Moon, H. W., and O’Brien, A. D. (1998). 6301–6311. doi: 10.1128/AEM.02200-13
Escherichia coli O157:H7 requires intimin for enteropathogenicity in calves. Menge, C., Wieler, L. H., Schlapp, T., and Baljer, G. (1999). Shiga toxin 1 from
Infect. Immun. 66, 4560–4563. Escherichia coli blocks activation and proliferation of bovine lymphocyte sub-
Espy, M. J., Uhl, J. R., Sloan, L. M., Buckwalter, S. P., Jones, M. F., Vetter, E. A., et al. populations in vitro. Infect. Immun. 67, 2209–2217.
(2006). Real-time PCR in clinical microbiology: applications for routine labo- Moher, D., Liberati, A., Tetzlaff, J., Altman, D. G., and The PRISMA
ratory testing. Clin. Microbiol. Rev. 19, 165–256. doi: 10.1128/CMR.19.1.165- Group. (2009). Preferred reporting items for systematic reviews and meta-
256.2006 analyses: the PRISMA statement. PLoS Med. 6:e1000097. doi: 10.1371/jour-
Etcheverría, A. I., and Padola, N. L. (2013). Shiga toxin-producing Escherichia coli: nal.pmed.1000097
factors involved in virulence and cattle colonization. Virulence 4, 366–372. doi: Moon, H. W., and Bunn, T. O. (1993). Vaccines for preventing enterotoxi-
10.4161/viru.24642 genic Escherichia coli infections in farm animals. Vaccine 11, 213–200. doi:
Ewers, C., Stamm, I., Stolle, I., Guenther, S., Kopp, P. A., Fruth, A., Wieler, L. H., 10.1016/0264-410X(93)90020-X
et al. (2014). Detection of Shiga toxin- and extended-spectrum β-lactamase- Moseley, S. L., Huq, I., Alim, A. R., So, M., Samadpour-Motalebi, M., and Falkow, S.
producing Escherichia coli O145:NM and Ont:NM from calves with diarrhoea. (1980). Detection of enterotoxigenic Escherichia coli by DNA colony hybridiza-
J. Antimicrob. Chemother. 69, 2005–2007. doi: 10.1093/jac/dku042 tion. J. Infect. Dis. 142, 892–898. doi: 10.1093/infdis/142.6.892
Ferens, W. A., Cobbold, R., and Hovde, C. J. (2006). Intestinal Shiga toxin- Nagy, B., and Fekete, P. Z. (1999). Enterotoxigenic Escherichia coli (ETEC) in farm
producing Escherichia coli bacteria mitigate bovine leukemia virus infec- animals. Vet. Res. 30, 259–284.
tion in experimentally infected sheep. Infect. Immun. 74, 2906–2916. doi: Nagy, B., and Fekete, P. Z. (2005). Enterotoxigenic Escherichia coli in veterinary
10.1128/IAI.74.5.2906-2916.2006 medicine. Int. J. Med. Microbiol. 295, 443–454. doi: 10.1016/j.ijmm.2005.07.003
Ferens, W. A., Grauke, L. J., and Hovde, C. J. (2004). Shiga toxin 1 targets Nguyen, Y., and Sperandio, V. (2012). Enterohemorrhagic E. coli (EHEC) patho-
bovine leukemia virus-expressing cells. Infect. Immun. 72, 1837–1840. doi: genesis. Front. Cell. Infect. Microbiol. 2:90. doi: 10.3389/fcimb.2012.00090
10.1128/IAI.72.3.1837-1840.2004 Orden, J. A., Domínguez-Bernal, G., Martínez-Pulgarín, S., Blanco, M., Blanco, J.
Ferens, W. A., and Hovde, C. J. (2000). Antiviral activity of shiga toxin 1: suppres- E., Mora, A., et al. (2007). Necrotoxigenic Escherichia coli from sheep and goats
sion of bovine leukemia virus-related spontaneous lymphocyte proliferation. produce a new type of cytotoxic necrotizing factor (CNF3) associated with the
Infect. Immun. 68, 4462–4469. doi: 10.1128/IAI.68.8.4462-4469.2000 eae and ehxA genes. Int. Microbiol. 10, 47–55. doi: 10.2436/20.1501.01.7
Fleckenstein, J. M., Hardwidge, P. R., Munson, G. P., Rasko, D. A., Orskov, I., Orskov, F., Jann, B., and Jann, K. (1977). Serology, chemistry, and
Sommerfelt, H., and Steinsland, H. (2010). Molecular mechanisms of genetics of O and K antigens of Escherichia coli. Bacteriol. Rev. 41, 667–710.

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org 11 March 2015 | Volume 5 | Article 23
Kolenda et al. Pathogenic Escherichia coli in calves

Orskov, I., Orskov, F., Smith, H. W., and Sojka, W. J. (1975). The establishment Servin, A. L. (2005). Pathogenesis of Afa/Dr diffusely adhering Escherichia coli.
of K99, a thermolabile, transmissible Escherichia coli K antigen, previously Clin. Microbiol. Rev. 18, 264–292. doi: 10.1128/CMR.18.2.264-292.2005
called “Kco,” possessed by calf and lamb enteropathogenic strains. Acta Pathol. Smith, H. W., and Halls, S. (1967). Observations by the ligated intestinal seg-
Microbiol. Scand. B. 83, 31–36. ment and oral inoculation methods on Escherichia coli infections in pigs, calves,
Ortiz-Pelaez, A., Pritchard, D. G., Pfeiffer, D. U., Jones, E., Honeyman, P., and lambs and rabbits. J. Pathol. 93, 499–529. doi: 10.1002/path.1700930211
Mawdsley, J. J. (2008). Calf mortality as a welfare indicator on British cattle Smith, T., and Orcutt, M. L. (1925). The becteriology of the intestinal tract of young
farms. Vet. J. 176, 177–181. doi: 10.1016/j.tvjl.2007.02.006 calves with special referenceto the early diarrhea (“scours”). J. Exp. Med. 41,
Pruimboom-Brees, I. M., Morgan, T. W., Ackermann, M. R., Nystrom, E. D., 89–106. doi: 10.1084/jem.41.1.89
Samuel, J. E., Cornick, N. A., et al. (2000). Cattle lack vascular receptors Smith, W. W. (1962). Observations on the ætiology of neonatal diarrhœa (scours)
for Escherichia coli O157:H7 Shiga toxins. Proc. Natl. Acad. Sci. U.S.A. 97, in calves. J. Pathol. 84, 147–168. doi: 10.1002/path.1700840117
10325–10329. doi: 10.1073/pnas.190329997 Sorek, R., and Cossart, P. (2010). Prokaryotic transcriptomics: a new view on
R Development Core Team (2013). R: A Language and Environment for Statistical regulation, physiology and pathogenicity. Nat. Rev. Genet. 11, 9–16. doi:
Computing. Vienna: R Foundation for Statistical Computing. Available online 10.1038/nrg2695
at: http://www.r-project.org Stroup, D. F., Berlin, J. A., Morton, S. C., Olkin, I., Williamson, G. D., Rennie,
Rödiger, S., Schierack, P., Böhm, A., Nitschke, J., Berger, I., Frömmel, U., et al. D., et al. (2000). Meta-analysis of observational studies in epidemiology: a pro-
(2013). A highly versatile microscope imaging technology platform for the mul- posal for reporting. Meta-analysis of observational studies in Epidemiology
tiplex real-time detection of biomolecules and autoimmune antibodies. Adv. (MOOSE) group. JAMA 283, 2008–2012. doi: 10.1001/jama.283.15.2008
Biochem. Eng. Biotechnol. 133, 35–74. doi: 10.1007/10_2011_132 Titze, C., and Weichel, A. (1908). Die ätiologie der kälberruhr. Berlin. Tierärtzl.
Römer, A., Wieler, L. H., and Schierack, P. (2012). Analyses of intestinal Wochschr. 26, 181–192.
commensal Escherichia coli strains from wild boars suggest adaptation to Weintraub, A. (2007). Enteroaggregative Escherichia coli: epidemiology, virulence
conventional pig production conditions. Vet. Microbiol. 161, 122–129. doi: and detection. J. Med. Microbiol. 56, 4–8. doi: 10.1099/jmm.0.46930-0
10.1016/j.vetmic.2012.07.009 Wickham, H. (2010). A layered grammar of graphics. J. Comput. Graph. Stat. 19,
Ruan, X., Crupper, S. S., Schultz, B. D., Robertson, D. C., and Zhang, W. (2012). 3–28. doi: 10.1198/jcgs.2009.07098
Escherichia coli expressing EAST1 toxin did not cause an increase of cAMP or
cGMP levels in cells, and no diarrhea in 5-day old Gnotobiotic Pigs. PLoS ONE Conflict of Interest Statement: The authors declare that the research was con-
7:e43203. doi: 10.1371/journal.pone.0043203 ducted in the absence of any commercial or financial relationships that could be
Schierack, P., Steinrück, H., Kleta, S., and Vahjen, W. (2006). Virulence construed as a potential conflict of interest.
factor gene profiles of Escherichia coli isolates from clinically healthy
pigs. Appl. Environ. Microbiol. 72, 6680–6686. doi: 10.1128/AEM.02 Copyright © 2015 Kolenda, Burdukiewicz and Schierack. This is an open-access arti-
952-05 cle distributed under the terms of the Creative Commons Attribution License (CC
Schierack, P., Weinreich, J., Ewers, C., Tachu, B., Nicholson, B., and Barth, BY). The use, distribution or reproduction in other forums is permitted, provided the
S. (2011). Hemolytic porcine intestinal Escherichia coli without virulence- original author(s) or licensor are credited and that the original publication in this
associated genes typical of intestinal pathogenic E. coli. Appl. Environ. Micro- journal is cited, in accordance with accepted academic practice. No use, distribution
biol. 77, 8451–8455. doi: 10.1128/AEM.05289-11 or reproduction is permitted which does not comply with these terms.

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org 12 March 2015 | Volume 5 | Article 23