Prevalence and characteristics of Shiga toxin-producing Escherichia coli in finishing pigs: Implications on public health

Accepted Manuscript Prevalence and characteristics of Shiga toxin-producing Escherichia coli in finishing pigs: Implications on public health

Wonhee Cha, Pina M. Fratamico, Leah E. Ruth, Andrew S. Bowman, Jacqueline M. Nolting, Shannon D. Manning, Julie A. Funk PII: DOI: Reference:

S0168-1605(17)30447-6 doi:10.1016/j.ijfoodmicro.2017.10.017 FOOD 7706

To appear in:

International Journal of Food Microbiology

Received date: Revised date: Accepted date:

10 April 2017 11 October 2017 12 October 2017

Please cite this article as: Wonhee Cha, Pina M. Fratamico, Leah E. Ruth, Andrew S. Bowman, Jacqueline M. Nolting, Shannon D. Manning, Julie A. Funk , Prevalence and characteristics of Shiga toxin-producing Escherichia coli in finishing pigs: Implications on public health. The address for the corresponding author was captured as affiliation for all authors. Please check if appropriate. Food(2017), doi:10.1016/j.ijfoodmicro.2017.10.017

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Prevalence and characteristics of Shiga toxin-producing Escherichia coli in finishing pigs: implications on public health

T

Wonhee Chaa, Pina M. Fratamicob, Leah E. Ruthb, Andrew S. Bowmanc, Jacqueline M. Noltingc

CR

Department of Microbiology and Molecular Genetics, Michigan State University, East Lansing,

US

a

IP

Shannon D. Manninga, Julie A. Funkd*

MI 48824, USA

Eastern Regional Research Center, Agricultural Research Service, U.S. Department of

Department of Veterinary Preventive Medicine, The Ohio State University, Columbus, OH

ED

c

M

Agriculture, Wyndmoor, PA 19038, USA

43210, USA

Department of Large Animal Clinical Sciences, Michigan State University, East Lansing, MI

PT

d

AN

b

AC

CE

48824, USA

* Corresponding author: Julie A. Funk, [email protected] 1

ACCEPTED MANUSCRIPT Abstract Shiga toxin-producing Escherichia coli (STEC) are important food-borne pathogens, which can cause serious illnesses, including hemorrhagic colitis and hemolytic uremic syndrome. To study the epidemiology of STEC in finishing pigs and examine the potential risks they pose for human

IP

T

STEC infections, we conducted a longitudinal cohort study in three finishing sites. Six cohorts of pigs (2 cohorts/site, 20 pigs/cohort) were randomly selected, and fecal samples (n=898) were

CR

collected every two weeks through their finishing period. Eighty-two pigs (68.3%) shed STEC at

US

least once, and the proportion of STEC-positive pigs varied across sites (50~97.5%) and cohorts (15~100%). Clinically important serotypes, O157:H7 (stx2c, eae) and O26:H11 (stx1a, eae), were

AN

recovered from two pigs at sites C and A, respectively. The most common serotype isolated was

M

O59:H21 (stx2e), which was particularly prevalent in site B as it was recovered from all STEC positive pigs (n=39). Each cohort showed different patterns of STEC shedding, which were

ED

associated with the prevalent serotype. The median shedding duration of STEC in pigs was 28

PT

days, consistent with our prior study. However, among pigs shedding O59:H21 at least once, pigs in cohort B2 had a significantly longer shedding duration of 42 days (P < 0.05) compared to

CE

other cohorts. Stx2e was the most commonly observed stx variant in finishing pigs (93.9%), in

AC

accordance with the previous studies. Stx2e has been reported to be significantly associated with edema disease in pigs, however, the pathogenicity in humans warrants further investigations. Nonetheless, our findings affirm that pigs are an important reservoir for human STEC infections, and that the circulating serotypes in a cohort and site management factors may significantly affect the prevalence of STEC. Molecular characterization of STEC isolates and epidemiological studies to identify risk factors for shedding in pigs are strongly warranted to further address the significance to public health and to develop mitigation strategies.

2

ACCEPTED MANUSCRIPT Keywords: Shiga toxin-producing Escherichia coli, finishing pigs, cohort, prevalence, shedding duration.

T

1. Introduction

IP

Shiga toxin-producing Escherichia coli (STEC) are a leading cause of food-borne infections

CR

worldwide (Majowicz et al., 2014). Infection commonly causes symptoms of gastroenteritis; however, it can also lead to serious conditions such as hemorrhagic colitis (HC), hemolytic

US

uremic syndrome (HUS) (Karmali et al., 1985; Nataro and Kaper, 1998), and neurological

AN

involvement (Trachtman et al., 2012). Overall, it was estimated that STEC cause more than 265,000 illnesses every year in the U.S., resulting in more than 3,600 hospitalizations and 30

M

deaths (Scallan et al., 2011).

ED

The STEC bacterium is characterized by the ability to produce a cytotoxin, known as Shiga toxin

PT

(Stx), which is encoded by stx genes carried on bacteriophages (O’Brien et al., 1984). Stx is divided into two major types, Stx1 and Stx2, and can be further classified into three subtypes for

CE

Stx1 (Stx1a, Stx1c, Stx1d) and seven subtypes for Stx2 (Stx2a, Stx2b, Stx2c, Stx2d, Stx2e,

AC

Stx2f, and Stx2g) based on the encoding genes (Scheutz et al., 2012). The most common subtypes associated with the development of HC and HUS are Stx2a, Stx2c, and Stx2d (Friedrich et al., 2002; Melton-Celsa, 2014). Intimin, an outer membrane protein encoded by eae, is another important virulence factor that is essential for forming the characteristic attaching and effacing lesion of enteropathogenic E. coli (EPEC) and enterohemorrhagic E. coli (EHEC) (Kaper et al., 2004).

3

ACCEPTED MANUSCRIPT Among more than 300 known serotypes, STEC O157:H7 has been most commonly associated with outbreaks and more severe STEC infections (Gyles, 2007); however strains representing other serotypes (non-O157 STEC) are increasing in frequency (Gould et al., 2013) and have been linked to both HUS and HC, as well as outbreaks (Bettelheim, 2007; Kuehne et al., 2016; Tseng

T

et al., 2016). In the U.S., the annual burden of non-O157 STEC was estimated to be twice that of

IP

O157 STEC (Scallan et al., 2011), with strains of serogroups O26, O45, O103, O111, O121, and

CR

O145 predominating among clinical cases (CDC, 2016). Nonetheless, less is known about the epidemiology of non-O157 STEC, particularly with regard to the source of infections and

US

transmission routes.

AN

Swine have been reported to shed STEC at a similar rate as cattle in various geographic locations

M

(Borie et al., 1997; Johnsen et al., 2001; Nakazawa and Akiba, 1999). In the U.S., a national cross-sectional survey conducted in 2000 reported a 28.5% isolation rate (196/687) of STEC in

ED

pigs, all of which were classified as non-O157 serogroups (Fratamico et al., 2004). Further

PT

molecular characterization revealed a wide diversity of STEC serogroups and virulence gene profiles in the clinically healthy pigs, warranting the need to examine if pigs may contribute to

CE

the burden of human illness associated with STEC. Furthermore, pork products have been linked

AC

to outbreaks associated with O157:H7 (Cheng et al., 2015; Trotz-Williams et al., 2012; Williams et al., 2000) as well as non-O157 STEC (Paton et al., 1996). A recent outbreak in Canada, which ranked as the second largest food-borne E. coli O157:H7 outbreak in the country, was traced to a pork product (Cheng et al., 2015), supporting the likelihood that pigs were the source of STEC contamination. With an overall aim to examine the role of pigs in human STEC infections, the main objective of the current study was to examine the prevalence and transmission dynamics of STEC in healthy 4

ACCEPTED MANUSCRIPT pigs over time. To confirm the previous findings of a high prevalence of STEC in swine (Tseng et al., 2015) and to expand our knowledge on the serotypes and virulence profiles (stx1/2/eae) of STEC strains circulating in healthy pigs, we conducted a longitudinal cohort study. We hypothesized that there was a high prevalence of STEC, particularly of non-O157 STEC in

T

finishing pigs, which included O-serogroups that have been associated with human illness. We

US

CR

attributable to variation in the prevalence of different serotypes.

IP

also hypothesized that the level and duration of STEC shedding will vary between cohorts,

AN

2. Methods and Materials 2.1. Study design

M

A longitudinal cohort study was conducted at three different wean-to-finish sites (A, B, and C)

ED

within one pig production company in Ohio, U.S. This production company was selected to

PT

represent typical pig production farms in the U.S.: multi-site production system with an all-in, all-out pig flow. Three sites were selected based on their willingness to participate in the study.

CE

For each of the three sites, two cohorts were selected based on the temporal aspects of when pigs

AC

were placed. Cohorts A1, B1, and C1 were sampled from June 2014 to October 2014, while cohorts A2, B2, and C2 were sampled from December 2014 to April 2015. Cohorts A1 and A2 were from site A, while cohorts B1 and B2, and cohorts C1 and C2 were from sites B and C, respectively. Each cohort of pigs was selected from total barn inventories of about 2,370 pigs, which were placed in the sites at approximately 3 weeks of age until they were shipped for harvest at approximately 24 weeks.

5

ACCEPTED MANUSCRIPT 2.2. Sample collection For each cohort, 20 pigs were randomly selected and individually identified using ear tags when pigs were aged 10 weeks. A sample size of 20 pigs for each cohort was calculated based on a previous study, which found the STEC incidence to be 65.3% (Tseng et al., 2015). In a

IP

T

population of 2,370 animals, 18 pigs allow an 80% confidence to estimate an incidence of 60% 15%, thus we conservatively rounded up the number to 20. A total of 120 pigs (20 pigs x 6

CR

cohorts) were followed through their finishing period, from 10 weeks of age until marketing,

US

with a sampling interval of 14 days. Individual fecal samples were collected directly from the rectum, and placed into sterile VWR® microbiology/urinalysis specimen containers (VWR

AN

International, USA). The specimen containers were placed into a cooler with ice, until they were

M

shipped overnight on ice to the USDA Eastern Regional Research Center in Wyndmoor, PA. All

ED

fecal samples were stored at 4°C until they were processed.

PT

2.3. Detection, isolation, and characterization of STEC from swine fecal samples

CE

All fecal samples were processed within 6 h of receipt. Samples were initially enriched using a

AC

modification of the protocol described by Grant et al. (Grant et al., 2009). Five grams of each fecal sample were placed in an individual Stomacher bag (Seward Ltd, West Sussex, UK) containing 95 ml of tryptic soy broth (TSB) pH 3, and then subjected to pummeling for 30 sec in a Stomacher apparatus (Seward Laboratory Systems, Bohemia, NY). After incubating for 10-15 min at room temperature, 100 ml of TYTP (TSB + 12g/L yeast extract, 12.5g/L Trizma Base, and 1g/L sodium pyruvate, with a final pH of 8.7) were added, and then samples were incubated without rotation for 15 h at 41°C.

6

ACCEPTED MANUSCRIPT DNA extraction of the enrichment was performed using the PrepSEQ Rapid Spin Sample Preparation Kit (Life Technologies Corporation, Carlsbad, CA) following manufacturer’s instructions. A multiplex PCR assay on the extracted DNA was performed using primers and probes targeting the eae gene and all stx1 and stx2 variants except stx2f (Wasilenko et al., 2012)

T

with TaqMan® Environmental Master Mix 2.0 (Life Technologies Corporation). The cycling

IP

conditions were 50°C for 120 sec and 95°C for 10 min, followed by 40 cycles of 95°C for 15 sec

CR

and 60°C for 60 sec. Samples giving a cycle threshold less than 35 were considered positive.

US

Enriched samples positive for stx were plated onto both CHROMagar STEC (CHROMagar) (DRG International, Inc., Mountainside, NJ) and modified Rainbow Agar O157 (mRBA) plates.

AN

mRBA was prepared by adding potassium tellurite (0.15mg/L), novobiocin (5mg/L), and

M

cefixime trihydrate (0.05mg/L) to Rainbow Agar O157 (Biolog, Hayward, CA). Single colonies with different morphologies and color on each plate were picked and subjected to the same real-

ED

time PCR protocol for stx1/2 and eae as described. The stx subtyping was conducted using

PT

primers and PCR conditions as described previously (Scheutz et al., 2012). A subset of confirmed STEC isolates were sent to the E. coli Reference Center at the Pennsylvania State

CE

University (University Park, PA) for O and H serotype characterization. In brief, O serotyping

AC

was performed using antisera generated against O1-O187 with the exceptions of O31, O47, O72, O94, O122, following the procedure as previously described (Orskov et al., 1977): H typing was performed by PCR-RFLP of fliC gene responsible for flagella (Machado et al., 2000).

7

ACCEPTED MANUSCRIPT 2.4. Statistical analysis If at least one STEC isolate was recovered from a given pig sample, then that pig was considered positive for STEC. Clinically important isolates were defined as those belonging to O157 or six major non-O157 serogroups, O26, O45, O103, O111, O121, and O145. In addition, based on

IP

T

previous literature on pathogenicity, isolates harboring stx2a, stx2c, stx2d, were also classified as clinically important. All data were recorded and managed in Microsoft Office Excel 2013, in

CR

which descriptive statistics, i.e. proportion of STEC positive pigs by pig age, cohort, and barn

US

were conducted. Further statistical analyses including survival analysis were conducted using SAS 9.3 (SAS Institute, Cary, NC, USA). Differences in the frequencies of STEC positive

AN

animals across cohort and barn were examined by χ2 and Fisher’s exact tests; P < 0.05 was

M

considered significant.

ED

Survival analysis using the Kaplan-Meier method was performed to estimate the duration of STEC shedding in positive pigs, which was defined as the time interval (days) between the first

PT

and last sampling date of STEC isolation from a pig. Fourteen days were added to the observed

CE

duration, accounting for the sampling interval. Three different datasets were prepared for the analysis. The first dataset included pigs shedding only one serotype with the following inclusion

AC

criteria; (1) survival until marketing, (2) no more than one period from which a sample was not collected, (3) no more than two negative cultures between two positive culture results. The second data set included only pigs shedding O59:H21 with the same criteria, however, including pigs that shed more than one serotype. A third dataset was made for all pigs shedding clinically important isolates with the same criteria as the second dataset. Right-censoring was applied to the pigs that were marketed before three consecutive STEC negative results. The homogeneity of survivor curves in each dataset was examined by the Log-rank test and Wilcoxon test; P < 0.05 8

ACCEPTED MANUSCRIPT was considered significant. Specifically, the difference by cohort, barn, and serotype was examined.

T

3. Results

IP

3.1. Sample collection

CR

Out of 120 enrolled pigs, one died after the first visit, and 34 pigs were sent for marketing before

US

24 weeks of age. Complete sample sets, i.e. eight farm visits, were obtained for 67 (55.8%) pigs, with of those 11 pigs also sampled on a 9th visit. Two pigs were sampled five times, while eight

AN

pigs were sampled six times. Besides early marketing and death, inability to locate individual pigs at the time of sampling precluded complete sample sets. By site, 62.5% of pigs (25/40) from

M

site B, 57.5% (23/40) from site C, and 47.5% (19/40) from site A completed eight farm visits. In

PT

ED

total, 898 samples were collected for culture (Table 1).

CE

3.2. Prevalence and distribution of STEC in pigs over the finishing period Eighty-two pigs (68.3%) among a total of 120 pigs shed STEC at least once during their

AC

finishing period. There was a significant difference in the proportion of STEC positive pigs by site, ranging from 50% (site A) to 97.5% (site B; P < 0.0001). The proportion of STEC in pigs from sites A and B remained similar between different cohorts, while site C had significant changes across cohorts, from 100% to 15% (Figure 1). Specifically, 13 pigs in cohort A1, seven pigs in cohort A2, all 20 pigs in cohort B1, 19 pigs in cohort B2, 20 pigs in cohort C1, and three pigs in cohort C2 were positive for STEC (Figure 1).

9

ACCEPTED MANUSCRIPT When the proportion of STEC positive pigs were examined by age over the finishing period, peaks at different ages were observed for each cohort (Figure 2A). While the proportion of positive pigs in cohorts A1 and B1 peaked at the end of the finishing period (24 weeks and 22 weeks, respectively), cohorts A2, B2, and C2 all showed the highest positive rate at 12 weeks of

T

age. Cohort C1 had the highest proportion at 18 weeks of age, while cohort B2 had at least one

IP

pig positive for STEC at all ages. Among STEC positive pigs, the median age at the initial time

CR

of STEC isolation was 16 weeks, ranging from 10 weeks to 24 weeks. The frequency distribution of age at which STEC was first detected in individual pigs within each cohort was similar as the

AN

US

proportion of STEC positive pigs for each cohort (Figure 2B).

M

3.3. Detection, isolation and characterization of the STEC isolates

ED

Among all 898 samples, 397 (44.2%) tested positive for at least one stx gene following enrichment. The positive samples were subcultured to CHROMagar and mRBA plates to isolate

PT

individual stx-positive colonies, which resulted in 352 isolates with at least one stx gene detected

CE

by real-time PCR. One hundred and seventy-six isolates were recovered from CHROMagar and 176 from mRBA plates. For further characterization, at least one isolate was selected per pig for

AC

each visit, in addition to isolates that showed different morphology on plates. For instance, if a sample resulted in STEC-positive colonies on both plates, isolates from both plates were selected for serotyping, including any additional colonies showing different colors/characteristics on either plate. A total of 302 isolates, representing 176 samples, were selected from the 352 colonies identified on the two media types, and sent for O:H characterization, from which 15 different O:H

10

ACCEPTED MANUSCRIPT serotypes were identified. Eleven different O-serogroups, including one non-typeable (O-neg), and 9 H-types were identified (Table 2). O59:H21 was the most common serotype as it was recovered in 58 of the 82 pigs (70.7%), followed by O-neg:H30 (n=13; 15.9%) and O174:H2 (n=9; 11.0%). Significant associations were observed between cohort and serotypes.

T

Specifically, O-neg:H30 was more likely to be found in cohort A1 (Fisher’s P < 0.0001), while

IP

O86:H32 predominated in cohort A2 (Fisher’s P < 0.0001), and all of the O174:H2 isolates were

CR

exclusive to cohort C1 (Fisher’s P < 0.0001). The most prevalent serotype, O59:H21 was observed in only cohorts B1, B2, and C1, thereby contributing to a significantly higher recovery

US

rate for site B (OR: 20.15, 95% CI: 8.83, 45.96). Serotypes O157:H7 and O26:H11, which are

AN

clinically important in humans, were recovered from two pigs at sites C and A, respectively.

M

Shedding of two and three different serotypes were observed in 23 pigs and three pigs, respectively, at the same or different time points. Most of the isolates carried stx2e only

ED

(288/302), while other stx variant genes, stx1a, stx2a, stx2c and stx2d were also observed in 8, 2, 4,

PT

and 2 isolates, respectively. As shown in Table 2, there was a strong association between a serotype and the virulence profile (P < 0.001). Particularly, all 224 isolates of O59:H21 harbored

CE

only stx2e, while the O26:H11 (n=2) and O157:H7 (n=4) isolates carried stx1a and stx2c,

AC

respectively. Furthermore, these were the only two serotypes that harbored eae along with the stx genes and could be classified as EHEC. Similarly, all of the isolates belonging to serogroup O163 carried stx1a (Fisher’s P < 0.0001). Interestingly, a significant difference in the number of various serotypes recovered was observed between the two plating agars. Five serotypes were recovered from CHROMagar plates, while mRBA recovered all of the serotypes identified in this study, except for O26:H11 (Table 2). Significant associations were found between mRBA and recovery of specific serotypes including 11

ACCEPTED MANUSCRIPT O86:H32 (Fisher’s P < 0.01) and O174:H2 (Fisher’s P < 0.0001). On the contrary, all isolates of O26:H11 (n=2) were recovered from only CHROMagar. Four serotypes, O59:H21, O157:H7, O100:H30, and O-neg:H30 were able to be cultured from both plates, although the most common serotype identified in this study, O59:H21, was more likely to be recovered from CHROMagar

CR

IP

T

(OR: 3.78; 95% CI: 2.12, 6.76).

3.4. Shedding duration of STEC in pigs

US

Among 82 pigs with at least one positive STEC isolate, 56 pigs shed only one serotype during

AN

their finishing period. Two pigs had more than one missed sampling while two other pigs had more than two consecutive negative results between positive STEC cultures. Thus, 52 pigs were

M

included in the first data set, among which 15 pigs (28.8%) were right-censored. The median

ED

shedding duration was 28 days with the probability of shedding for 21 days at 71.2% and maximum of 56 days at 12.9% (Figure 3). No significant difference was observed in survival

PT

curves when stratified by site (P = 0.76) or cohort (P = 0.16). Eight serotypes remained in this

CE

data set, and although there was no significant difference in the median duration by serotype (P = 0.19), isolates belonging to O59:H21 had a longer shedding duration of up to 56 days compared

AC

to other serotypes at 28 days (Supplementary Figure 1). O59:H21 was the most prevalent serotype found in this study in concordance with our previous study (Tseng et al., 2015), thus we analyzed pigs that shed O59:H21 for at least at one time point (n=58) separately. Two pigs that had more than one missed sampling and three pigs with more than two negative cultures in between were excluded from the analysis. The median duration of O59:H21 shedding was 28 days, and there was no significant difference between pigs shedding

12

ACCEPTED MANUSCRIPT only O59:H21 (n=34; 4 cases censored) and O59:H21 plus another serotype, e.g. O59:H21 and O174:H2 (n=19; 0 cases censored). However, it is noteworthy that shedding duration varied among the cohorts; the median shedding duration for cohort B2, for instance, was significantly longer at 42 days compared to 28 days for cohorts B1 and C1 (P < 0.05) (Figure 4). The

T

shedding duration of clinically important serotypes, O157:H7 (n=2), O26:H11 (n=2), and O8:H7

US

CR

serotypes, but showed a narrower range from 21 days to 28 days.

IP

(n=1) based on the virulence profile (stx2a, stx2d), was not significantly different from other

AN

3.5. Seasonality of STEC shedding in pigs

By season, 165, 225, 224, and 284 samples were collected in spring, summer, fall, and winter,

M

respectively. A significantly higher proportion of STEC positive samples were recovered in fall

ED

(36.16%) compared to other seasons (P < 0.0001) (Figure 5). Among cohorts that were sampled from late June to late October (cohorts A1, B1, C1), cohorts A1 and B1 showed significantly

PT

higher recovery rate of STEC isolates in the fall compared to summer (Fisher’s P < 0.05). The

CE

association was even stronger in cohort A1, in which all of the recovered isolates (n=33) were detected in fall. Isolates belonging to serotype O-neg:H30 were significantly more likely to be

AC

recovered in the fall months (Fisher’s P < 0.0001). Meanwhile, clinically important serotypes, O26:H11 and O157:H7, were observed in the spring and winter months, respectively.

4. Discussion This study confirms a high prevalence of STEC in commercial pigs in the U.S. Indeed, 68.3% of pigs (82/120) were shedding STEC at least once during their finishing period, consistent with our 13

ACCEPTED MANUSCRIPT previous finding of 65.3% STEC prevalence in a different Midwestern state (Tseng et al., 2015). Also, similar to our prior study, O59:H21was the most prevalent serotype observed among finishing pigs. However, clinically important serotypes, O157:H7 and O26:H11, were also recovered in this study. In the U.S., previous national surveys reported a lower STEC prevalence

T

of 28.5% and no recovery of O157:H7 serotype in clinically healthy pigs (Bush, 1997; Fratamico

IP

et al., 2004). As these surveys were cross-sectional studies we cannot make a direct comparison,

CR

however, the varying prevalence of STEC and specific serotypes observed over time in both our studies (Tseng et al., 2015) emphasizes the importance of longitudinal sampling for determining

US

STEC shedding status in pigs.

AN

Use of different isolation protocols can affect the ability to recover different STEC serotypes. In

M

the present study, we used both mRBA and CHROMagar plates to recover STEC from enriched samples. mRBA has a reduced level of tellurite (0.15mg/L) compared to CHROMagar, and it

ED

has been shown to have an increased ability to detect STEC, particularly of O26, O45, O103,

PT

O111, O121, and O145 in a study of raw ground beef (Tillman et al., 2012). With pig fecal samples, mRBA showed a higher capacity to recover various serotypes, as 14 out of 15 serotypes

CE

were detected on mRBA while only five serotypes were recovered from CHROMagar. In

AC

contrast to the study results with ground beef, O26:H11 was not recovered on mRBA, and O59:H21, the most prevalent serotype was more likely to be isolated from CHROMagar (P < 0.05). Correspondingly, our previous study, which found O59:H21 as the most prevalent serotype used only CHROMagar for STEC isolation (Tseng et al., 2015). However, there were 10 additional isolates of O59:H21 recovered in this study that were not cultured on CHROMagar but only on mRBA. These data suggest that there is variation in tellurite tolerance, e.g., presence of terB - tellurite resistance gene, even within a serotype (Verhaegen et al., 2015) that may

14

ACCEPTED MANUSCRIPT impact detection. Based on the results in this study and prior studies reporting differences in the ability to detect non-O157 STEC on various agars (Tillman et al., 2012; Verhaegen et al., 2015), the paired use of agars like mRBA and CHROMagar, for instance, is strongly recommended when the sampling goal is to maximize recovery of non-O157 STEC.

IP

T

In addition to O157, isolates belonging to serogroups of O26, O45, O91, O103, O111, O113, O121, O128, and O145 as well as O104 are of significant public health concern (Bielaszewska et

CR

al., 2011). In this study, we identified two pigs (1.7%) with O157:H7 and O26:H11, which is

US

similar to prior U.S. studies reporting a STEC prevalence of 1.9% and 1.2% O157:H7 STEC in swine colon samples (n=305) and agricultural fair pigs (n=1102), respectively (Feder et al., 2003;

AN

Keen et al., 2006). This is the first study, however, to report the natural shedding of O157:H7 in

M

commercial pigs in the U.S. In addition, eae-negative O174:H2, the third most common serotype found in this study (n=9 pigs), was reported to have been associated with HUS (Blanco et al.,

ED

2004). Tarr et al. (Tarr et al., 2008) reported the isolation of eae-negative stx2-producing

PT

O174:H2 strains from human clinical samples, and two stx1 + stx2-producing strains were associated with HC. However, the O174:H2 strains in the current study carried stx2e, and it is

CE

not known if O174 strains carrying this stx subtype can cause serious illness. A stx1/stx2-

AC

positive and eae-negative OX3:H2 strain (prior designation of O174 was OX3), previously isolated from a patient with a urinary tract infection, also caused HUS in the same patient (Scheutz et al., 2000). Moreover, O174:H2 has been isolated from cattle and sheep (Blanco et al., 2004), suggesting that this serotype can colonize different animal reservoirs representing multiple potential sources for zoonotic transmission to humans. Including O174:H2 isolates, the majority of isolates (288/302) recovered from most pigs (77/82) harbored stx2e only, in accordance with our previous study (Tseng et al., 2015). By serotype, all 15

ACCEPTED MANUSCRIPT the isolates belonging to serotypes O8:H28, O59:H21, O71:H21, O86:H19, O86:H32, O174:H2, O184:H48, O100:H30, O-neg:H7, O-neg:H30 carried stx2e. Stx2e is the most frequently reported Shiga toxin subtype from pigs (Fratamico et al., 2004; Meng et al., 2014), and certain STEC serogroups occasionally cause edema disease in weaned pigs. Only a few human STEC

T

infections have been associated with strains harboring stx2e, however these included cases with

IP

severe symptoms, as well as HUS cases (Thomas et al., 1994). Thomas et al. (1994), for

CR

example, reported that a patient with HUS was infected with two strains carrying stx2e and belonging to serotypes O9ab:H- and O101:H-. A strain carrying stx2e and eae of serotype

US

O51:H49 caused HUS in a 65-year-old immunocompromised male (Fasel et al., 2014); the

AN

O51:H59 strain was isolated from the blood and stool, and the patient died. Indeed, it has been suggested that infections with stx2e-producing E. coli strains cause illness via different

M

pathogenic mechanisms as they often lack virulence factors like eae and ehxA (enterohemolysin)

ED

(Franke et al., 1995). Further molecular characterization studies are therefore warranted to

potential risk to humans.

PT

identify virulence factors, including ehxA, that are unique to these strains and to elucidate the

CE

Unlike cattle, pigs can present with disease caused by STEC, which results in edema disease in

AC

post-weaning and young finishing pigs. Certain serogroups like O8, O138, O139, O141, and O147, have been documented in these cases (Friedrich et al., 2002; Kaper et al., 2004; MeltonCelsa, 2014). In this study, two pigs were found to shed serotypes O8:H7, and O8:H28. However, due to the lack of health information on the pigs and the small sample size, we cannot examine associations between these serotypes and the possible clinical impact. The O8:H7 isolate had a virulence profile of stx2a and stx2d, which makes the serotype a potential human pathogen, as well. In fact, strains expressing stx2a, stx2c, and stx2d have been frequently associated with HC and 16

ACCEPTED MANUSCRIPT HUS, and in vitro experiments with the purified toxin has shown that stx2a and stx2d are 25 times more potent than stx2b and stx2c for inhibiting protein synthesis and metabolic activity (Fuller et al., 2011). Including O8, several serogroups (e.g., O26, O59, O100, O163) and H-types (e.g., H19, H30)

IP

T

were previously reported from pigs in the U.S. (Fratamico et al., 2004; Tseng et al., 2015; Wells et al., 2013), among which O59:H21 and O100:H30 were observed in highest frequency. In this

CR

study, four pigs in cohort 1 shed O100:H30, and all isolates belonging to the serotype (n=7) had

US

the same virulence gene profile of stx2e. On the contrary, a previous study that examined more diverse STEC isolates from different geographical locations, showed that most of the O100:H30

AN

isolates harbored stx2e, but they also reported one isolate carrying an stx2 subtype that was not

M

stx2e (Fratamico et al., 2004). It is therefore possible that some serotypes, including O100:H30 and O59:H21, express certain virulence characteristics that are more prevalent and more readily

ED

transmissible in commercial pigs. However, since stx genes are mobile and carried on

PT

bacteriophages, it is possible that the STEC virulence profiles can change within an animal over time through a coinfection with other STEC strains or exposure to stx-encoding bacteriophages.

CE

In fact, we found a total of 14 pigs to be shedding two different serotypes, e.g. O59:H21 (stx2e)

AC

and O163:H32 (stx1a), at the same time. This finding highlights the importance of a comprehensive detection method for STEC and additional molecular characterization studies to fully understand the transmission dynamics of STEC. In addition to transmission dynamics of mobile virulence factors within an animal, transmission of STEC isolates between animals determines the population transmission dynamics. Similar to our previous study (Tseng et al., 2015), a variation of STEC prevalence rates was observed in different cohorts. Particularly, the proportion of positive pigs in cohorts A1 and B1 peaked at the 17

ACCEPTED MANUSCRIPT end of the finishing period (24 weeks and 22 weeks, respectively), adding more concern to the STEC contamination in pork products. In fact, a recent study by Colello et al. (Colello et al., 2016) reported an increase of stx detection rate through pork production chain, from 2.86% at farms to 4.08% at slaughters, and 6% at boning rooms, implicating high cross-contamination rate

T

during the process. In the present study, we also observed that the distribution of serotypes was

IP

significantly associated with cohort, suggesting the presence of different sources and associated

CR

risk factors for STEC infection in each cohort. STEC prevalence over time in each cohort resembled the pattern of a point-source outbreak curve, similar to what we observed in our

US

previous study (Tseng et al., 2015) (Figure 2A). Notably, there were cohorts showing two peaks,

AN

which may be explained by the emergence and dynamics of different serotypes. For instance, the prevalence curve for cohort C1, which showed a first peak at 14 weeks of age, and then a higher

M

peak at 18-20 weeks, was a combination of two different point-source outbreak curves for

ED

O174:H2 and O59:H21, which contributed significantly to each peak, respectively. On the other hand, cohort B1 showed three different peaks, with O59:H21 predominating the cohort (isolate

PT

n=33/35), implying that there were multiple introductions of a same strain or that different

CE

strains of the same serotype, O59:H21, were introduced to the cohort at different time points.

AC

Further molecular characterization of these isolates is warranted to clarify the genetic relatedness. There was also dominance of O59:H21 at site B, with the cohorts associated with this site (B1 and B2) having 100% STEC prevalence (except for one pig that died after the first visit). Interestingly, the shedding duration of O59:H21 in cohort B2 (42 days) was significantly longer than that of cohorts B1 or C1, which was the same as the median shedding duration (28 days) of all of the serotypes. However, the peak of cohort B2 was observed at a notably earlier time point, compared to those of cohorts B1 and C1, resulting in a lower number of pigs to be censored for

18

ACCEPTED MANUSCRIPT the survival analysis. In fact, all these sites were wean-to-finish operations, implying earlier exposure to STEC before the study period began. Yet, the significant difference observed in prevalence and shedding duration of STEC over cohort and site warrants further analytical studies to examine the associated factors in cohort and site level. In cattle, for instance, the

T

prevalence of STEC varies across different geographic regions, however, a few consistent factors

IP

associated with the prevalence, i.e. season, herd management practices, and diet, have been

CR

identified (Cernicchiaro et al., 2009; Cho et al., 2013; Dunn et al., 2004).

US

In the U.S., two prominent peaks in July and October have been observed for human STEC infections according to the active surveillance data by the CDC (CDC, 2016). Also, a different

AN

seasonality of human STEC infections between serogroups, O157 and O26, has been reported

M

(Garvey et al., 2016). The authors hypothesized that the finding might be due to different primary animal reservoirs and different survival characteristics for each serogroup, and also

ED

specific human behavior, e.g. seasonal food, resulting in more frequent exposure to sources of

PT

each serogroup at different times of the year. The primary STEC reservoir, cattle, shows the highest prevalence in summer months (Hancock et al., 1994; Stanford et al., 2005), however, the

CE

seasonality of STEC in other animal reservoirs is largely unknown. In the course of following

AC

the cohorts, no samples were collected in May and November in the present study; however, a significantly higher rate of STEC recovery was observed in colder months, showing a sharp increase between August and September (Supplementary Figure 2). Additionally, cohort A1, in which STEC O-neg:H30 (15/18) was the predominant serotype, had a significantly higher frequency of STEC in the fall compared to summer. We therefore hypothesize that the change to a low thermal environment in the fall and winter may contribute to increased stress in pigs leading to lowered immunity and increased susceptibility to new STEC infections. A similar

19

ACCEPTED MANUSCRIPT relationship was previously suggested in pigs and Salmonella shedding (Pires et al., 2013), though further studies with a higher number of cohorts in different geographic locations are needed to confirm this seasonality and investigate the association between serotypes and season. In summary, a high prevalence of STEC was observed in finishing pigs, most of which were

IP

T

non-O157 STEC and showed a virulence profile of stx2e only. However, five pigs (4.2%) were found to shed clinically important serotypes, including O157:H7, highlighting the possibility of

CR

commercial pigs being a source of human STEC infections. Also, significant differences were

US

observed in prevalence and serotypes by cohort and site, which indicates a high attributable risk of environmental factors. Further monitoring and epidemiological studies related to pigs

AN

including molecular characterization and multi-level risk factor analysis are strongly needed to

ED

M

assess the importance of pigs as an animal reservoir for human STEC infections.

PT

Acknowledgements

We thank the producers who participated in this research, and Michele Zentkovich, Nola Bliss,

CE

Teresa Mills, Jessica Higgins, and Dr. Jeff Workman who assisted with sample collection. We

AC

thank Dr. Chitrita DebRoy and the E. coli Reference Center at the Pennsylvania State University for assistance with isolate characterization. This research was supported by the United States Department of Agriculture, National Institute of Food and Agriculture (Grant number 2013-67005-21189).

20

ACCEPTED MANUSCRIPT References Bettelheim, K.A., 2007. The Non-O157 Shiga-Toxigenic (Verocytotoxigenic) Escherichia coli ; Under-Rated Pathogens. Crit. Rev. Microbiol. 33, 67–87. doi:10.1080/10408410601172172

T

Bielaszewska, M. et al., 2011. Characterisation of the Escherichia coli strain associated with an

IP

outbreak of haemolytic uraemic syndrome in Germany, 2011: a microbiological study.

CR

Lancet Infect. Dis. 11, 671–676. doi:10.1016/S1473-3099(11)70165-7 Blanco, M. et al., 2004. Serotypes, virulence genes, and intimin types of Shiga toxin (verotoxin)-

US

producing Escherichia coli isolates from cattle in Spain and identification of a new intimin

AN

variant gene (eae-xi). J. Clin. Microbiol. 42, 645–51.

Borie, C. et al., 1997. Prevalence and characterization of enterohaemorrhagic Escherichia coli

M

isolated from healthy cattle and pigs slaughtered in Santiago, Chile. Arch. Med. Vet. 29,

ED

205–212.

PT

Bush, E., 1997. U.S. swine herd appears free of Escherichia coli O157:H7. Food Saf. Dig.

CE

January-Fe.

Center for Disease Control and Prevention (CDC). Foodborne Disease Active Surveillance

AC

Network (FoodNet), 2016. FoodNet Surveillance Report 2014 (Final report). Atlanta, GA. Cernicchiaro, N. et al., 2009. Risk factors associated with Escherichia coli O157:H7 in Ontario beef cow-calf operations. Prev. Vet. Med. 92, 106–15. doi:10.1016/j.prevetmed.2009.07.004 Cheng, J. et al., 2015. When Pigs Fly: A Large Outbreak of E.coli O157:H7 Infections Associated with Exposure to Contaminated Pork Products in Alberta, Canada, in: Council 21

ACCEPTED MANUSCRIPT of State and Territorial Epidemiologists. Council of State and Territorial Epidemiologists, Boston. Cho, S. et al., 2013. Herd-level risk factors associated with fecal shedding of Shiga toxinencoding bacteria on dairy farms in Minnesota, USA. The Canadian Veterinary Journal. 54,

IP

T

693–7.

CR

Colello, R. et al., 2016. From Farm to Table: Follow-Up of Shiga Toxin-Producing Escherichia coli Throughout the Pork Production Chain in Argentina. Front. Microbiol. 7, 93.

US

doi:10.3389/fmicb.2016.00093

AN

Dunn, J.R. et al, 2004. Prevalence of Shiga-toxigenic Escherichia coli O157:H7 in adult dairy cattle. J. Am. Vet. Med. Assoc. 224, 1151–8.

M

Fasel, D. et al., 2014. Hemolytic uremic syndrome in a 65-Year-old male linked to a very

ED

unusual type of stx2e- and eae-harboring O51:H49 shiga toxin-producing Escherichia coli.

PT

J. Clin. Microbiol. 52, 1301–3. doi:10.1128/JCM.03459-13 Feder, I. et al., 2003. Isolation of Escherichia coli O157:H7 from intact colon fecal samples of

CE

swine. Emerg. Infect. Dis. 9, 380–3. doi:10.3201/eid0903.020350

AC

Franke, S. et al., 1995. Clonal relatedness of Shiga-like toxin-producing Escherichia coli O101 strains of human and porcine origin. J. Clin. Microbiol. 33, 3174–8. Fratamico, P.M. et al., 2004. Prevalence and characterization of shiga toxin-producing Escherichia coli in swine feces recovered in the National Animal Health Monitoring System’s Swine 2000 study. Appl. Environ. Microbiol. 70, 7173–8. doi:10.1128/AEM.70.12.7173-7178.2004

22

ACCEPTED MANUSCRIPT Friedrich, A.W. et al., 2002. Escherichia coli harboring Shiga toxin 2 gene variants: frequency and association with clinical symptoms. J. Infect. Dis. 185, 74–84. doi:10.1086/338115 Fuller, C.A. et al, 2011. Shiga Toxin Subtypes Display Dramatic Differences in Potency. Infect.

T

Immun. 79, 1329–1337. doi:10.1128/IAI.01182-10

IP

Garvey, P. et al., 2016. Serogroup-specific Seasonality of Verotoxigenic Escherichia coli ,

CR

Ireland. Emerg. Infect. Dis. 22, 742–744. doi:10.3201/eid2204.151160 Gould, L.H. et al., Emerging Infections Program Foodnet Working Group, 2013. Increased

US

recognition of non-O157 Shiga toxin-producing Escherichia coli infections in the United

AN

States during 2000-2010: epidemiologic features and comparison with E. coli O157 infections. Foodborne Pathog. Dis. 10, 453–60. doi:10.1089/fpd.2012.1401

M

Grant, M.A. et al., 2009. Comparison of enrichment procedures for shiga toxin-producing

ED

Escherichia coli in wastes from commercial swine farms. J. Food Prot. 72, 1982–6.

PT

Gyles, C.L., 2007. Shiga toxin-producing Escherichia coli: an overview. J. Anim. Sci. 85, E45-

CE

62. doi:10.2527/jas.2006-508

Hancock, D.D. et al., 1994. The prevalence of Escherichia coli O157.H7 in dairy and beef cattle

AC

in Washington State. Epidemiol. Infect. 113, 199–207. doi:10.1017/S0950268800051633 Johnsen, G. et al., 2001. Escherichia coli O157:H7 in faeces from cattle, sheep and pigs in the southwest part of Norway during 1998 and 1999. Int. J. Food Microbiol. 65, 193–200. Kaper, J.B. et al., 2004. Pathogenic Escherichia coli. Nat. Rev. Microbiol. 2, 123–40. doi:10.1038/nrmicro818 Karmali, M.A. et al., 1985. The association between idiopathic hemolytic uremic syndrome and 23

ACCEPTED MANUSCRIPT infection by verotoxin-producing Escherichia coli. J. Infect. Dis. 151, 775–82. Keen, J.E. et al., 2006. Shiga-toxigenic Escherichia coli O157 in agricultural fair livestock, United States. Emerg. Infect. Dis. 12, 780–6. doi:10.3201/eid1205.050984

T

Kuehne, A. et al., 2016. Estimating true incidence of O157 and non-O157 Shiga toxin-producing

IP

Escherichia coli illness in Germany based on notification data of haemolytic uraemic

CR

syndrome. Epidemiol. Infect. 1–11. doi:10.1017/S0950268816001436

Machado, J. et al., 2000. Identification of Escherichia coli flagellar types by restriction of the

US

amplified fliC gene. Res. Microbiol. 151, 535–46.

AN

Majowicz, S.E. et al., 2014. Global incidence of human Shiga toxin-producing Escherichia coli infections and deaths: a systematic review and knowledge synthesis. Foodborne Pathog.

M

Dis. 11, 447–55. doi:10.1089/fpd.2013.1704

ED

Melton-Celsa, A.R., 2014. Shiga Toxin (Stx) Classification, Structure, and Function. Microbiol

PT

Spectr 2, EHEC-0024-2013. doi:10.1128/microbiolspec.EHEC-0024-2013

CE

Meng, Q. et al., 2014. Characterization of Shiga toxin-producing Escherichia coli isolated from healthy pigs in China. BMC Microbiol. 14, 5. doi:10.1186/1471-2180-14-5

AC

Nakazawa, M., Akiba, M., 1999. Swine as a potential reservoir of shiga toxin-producing Escherichia coli O157:H7 in Japan. Emerg. Infect. Dis. 5, 833–4. doi:10.3201/eid0506.990618 Nataro, J.P., Kaper, J.B., 1998. Diarrheagenic Escherichia coli. Clin. Microbiol. Rev. 11, 142– 201. O’Brien, A.D. et al., 1984. Shiga-like toxin-converting phages from Escherichia coli strains that 24

ACCEPTED MANUSCRIPT cause hemorrhagic colitis or infantile diarrhea. Science 226, 694–6. Orskov, I. et al., 1977. Serology, chemistry, and genetics of O and K antigens of Escherichia coli. Bacteriol. Rev. 41, 667–710.

T

Paton, A.W. et al., 1996. Molecular microbiological investigation of an outbreak of hemolytic-

CR

producing Escherichia coli. J. Clin. Microbiol. 34, 1622–7.

IP

uremic syndrome caused by dry fermented sausage contaminated with Shiga-like toxin-

Pires, A.F.A. et al., 2013. Longitudinal study to evaluate the association between thermal

US

environment and Salmonella shedding in a midwestern US swine farm. Prev. Vet. Med.

AN

112, 128–37. doi:10.1016/j.prevetmed.2013.07.001

Scallan, E. et al., 2011. Foodborne illness acquired in the United States--major pathogens.

M

Emerg. Infect. Dis. 17, 7–15. doi:10.3201/eid1701.091101p1

ED

Scheutz, F. et al., 2000. Two Cases of Human Urinary Tract Infection Complicated by

PT

Hemolytic Uremic Syndrome Caused by Verotoxin-Producing Escherichia coli. Clin.

CE

Infect. Dis. 31, 815–816. doi:10.1086/314019 Scheutz, F. et al., 2012. Multicenter evaluation of a sequence-based protocol for subtyping Shiga

AC

toxins and standardizing Stx nomenclature. J. Clin. Microbiol. 50, 2951–63. doi:10.1128/JCM.00860-12 Stanford, K. et al., 2005. Ecology of Escherichia coli O157:H7 in Commercial Dairies in Southern Alberta. J. Dairy Sci. 88, 4441–4451. doi:10.3168/jds.S0022-0302(05)73131-3 Tarr, C.L. et al., 2008. Molecular characterization reveals similar virulence gene content in unrelated clonal groups of Escherichia coli of serogroup O174 (OX3). J. Bacteriol. 190, 25

ACCEPTED MANUSCRIPT 1344–9. doi:10.1128/JB.01317-07 Thomas, A. et al., 1994. Isolation of Vero cytotoxin-producing Escherichia coli serotypes O9ab:H- and O101:H-carrying VT2 variant gene sequences from a patient with haemolytic

T

uraemic syndrome. Eur. J. Clin. Microbiol. Infect. Dis. 13, 1074–6.

IP

Tillman, G.E. et al., 2012. Isolation of Shiga toxin-producing Escherichia coli serogroups O26,

CR

O45, O103, O111, O121, and O145 from ground beef using modified rainbow agar and post-immunomagnetic separation acid treatment. J. Food Prot. 75, 1548–54.

US

doi:10.4315/0362-028X.JFP-12-110

AN

Trachtman, H. et al., 2012. Renal and neurological involvement in typical Shiga toxin-associated HUS. Nat. Rev. Nephrol. 8, 658–669. doi:10.1038/nrneph.2012.196

M

Trotz-Williams, L.A. et al., 2012. Pork implicated in a Shiga toxin-producing Escherichia coli

ED

O1 7:H7 outbrea in Ontario, Canada. Can. . public Heal.

PT

e322-6.

ev. Can. sant publi ue 103,

Tseng, M. et al., 2015. Shiga toxin-producing E. coli (STEC) in swine: prevalence over the

CE

finishing period and characteristics of the STEC isolates. Epidemiol. Infect. 143, 505–14.

AC

doi:10.1017/S0950268814001095 Tseng, M. et al., 2016. Increasing incidence of non-O157 Shiga toxin-producing Escherichia coli (STEC) in Michigan and association with clinical illness. Epidemiol. Infect. 144, 1394–405. doi:10.1017/S0950268815002836 Verhaegen, B. et al., 2015. Comparison of Six Chromogenic Agar Media for the Isolation of a Broad Variety of Non-O157 Shigatoxin-Producing Escherichia coli (STEC) Serogroups.

26

ACCEPTED MANUSCRIPT Int. J. Environ. Res. Public Health 12, 6965–6978. doi:10.3390/ijerph120606965 Wasilenko, J.L. et al., 2012. Influence of primer sequences and DNA extraction method on detection of non-O157 Shiga toxin-producing Escherichia coli in ground beef by real-time PCR targeting the eae, stx, and serogroup-specific genes. J. Food Prot. 75, 1939–50.

IP

T

doi:10.4315/0362-028X.JFP-12-087

CR

Wells, J.E. et al., 2013. Effects of antimicrobials fed as dietary growth promoters on faecal shedding of Campylobacter, Salmonella and shiga-toxin producing Escherichia coli in

US

swine. J. Appl. Microbiol. 114, 318–28. doi:10.1111/jam.12065

AN

Williams, R.C et al., 2000. Illness outbreak associated with Escherichia coli O157:H7 in Genoa

AC

CE

PT

ED

M

salami. E. coli O157:H7 Working Group. CMAJ 162, 1409–13.

27

ACCEPTED MANUSCRIPT Table 1. Sample collection period and number of pigs and samples used in the study *

One pig died after the first visit.

No. of

T

No. of

IP

2401

20

148

Cohort 2

13 Jan 2015 ~ 21 Apr 2015

2336

20

144

Cohort 3

24 Jun 2014 ~ 14 Oct 2014

2392

20

155

Cohort 4

16 Dec 2014 ~ 7 Apr 2015

2342

20*

148

Cohort 5

8 Jul 2014 ~ 14 Oct 2014

2410

20

146

Cohort 6

6 Jan 2015 ~ 21 April 2015

2343

20

157

Total

14224

120

898

US

AN

M

PT

CE

Site C

22 July 2014 ~ 28 Oct 2014

AC

Site B

Cohort 1

CR

total pigs studied pigs

Site A

samples

ED

Sample collection period

No. of collected

28

ACCEPTED MANUSCRIPT Table 2. Serotypes and virulence profiles of STEC recovered from finishing pigs

Frequency

Virulence

Agar plate

Serotype

O8:H28

stx2e

1

1

O26:H11

stx1a, eae

2

2

X

O59:H21

stx2e

58

224

US

X

O71:H21

stx2e

1

1

X

O86:H19

stx2e

1

1

X

O86:H32

stx2e

7

8

X

O157:H7

stx2c, eae

2

4

O163:H19

stx1a

3

5

X

O163:H32

stx1a

1

1

X

O174:H2

stx2e

9

19

X

O184:H48

stx2e

1

1

X

O100:H30

stx2e

4

7

T

2

IP

1

ED

stx2a, stx2d

AC

O8:H7

AN

CHROM mRBA

M

Isolate

PT

Pig

CE

profile

X

X

X

CR

X

X

X

X

29

ACCEPTED MANUSCRIPT O-neg*:H7

stx2e

1

1

O-neg*:H30

stx2e

13

25

X X

X

X denotes recovery of isolates belonging to the serotype.

AC

CE

PT

ED

M

AN

US

CR

IP

T

O-neg*, O antigen non-typeable.

30

ACCEPTED MANUSCRIPT Figure 1. Proportion (%) of STEC positive pigs for each site and cohort Figure 2. A. Proportion (%) of pigs shedding STEC by pig age over the finishing period by cohort B. Proportion (%) of pigs at the age of first STEC detection by cohort

T

Figure 3. Kaplan-Meier survival curve for STEC shedding in pigs with only one serotype

IP

recovered during the finishing period

CR

Figure 4. Kaplan-Meier survival curves for O59:H21 STEC shedding in finishing pigs by

US

cohort

AC

CE

PT

ED

M

AN

Figure 5. % Frequency of STEC recovery by season

31

AN

US

CR

IP

T

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

M

Fig. 1

32

Fig. 2

AC

CE

PT

ED

M

AN

US

CR

IP

T

ACCEPTED MANUSCRIPT

33

Fig. 3

AC

CE

PT

ED

M

AN

US

CR

IP

T

ACCEPTED MANUSCRIPT

34

Fig. 4

AC

CE

PT

ED

M

AN

US

CR

IP

T

ACCEPTED MANUSCRIPT

35

AC

Fig. 5

CE

PT

ED

M

AN

US

CR

IP

T

ACCEPTED MANUSCRIPT

36

ACCEPTED MANUSCRIPT Highlights



A high prevalence (68.3%) of Shiga toxin-producing Escherichia coli (STEC) was observed

Finishing pigs can naturally shed clinically important serotypes, e.g. O157:H7, O26:H11,

IP



T

in finishing pigs raised for pork production.

Different STEC prevalence and shedding patterns, associated with the circulating

CE

PT

ED

M

AN

US

serotypes, were observed between cohorts.

AC



CR

that have been linked to hemolytic uremic syndrome and hemorrhagic colitis.

37