Content uploaded by Kim Stanford
Author content
All content in this area was uploaded by Kim Stanford on Jul 22, 2014
Content may be subject to copyright.
Variable Efficacy of a Vaccine and Direct-Fed Microbial
for Controlling Escherichia coli O157:H7
in Feces and on Hides of Feedlot Cattle
Kim Stanford,
1
Sherry Hannon,
2
Calvin W. Booker,
2
and G. Kee Jim
2
Abstract
To evaluate the efficacy of a type-III secreted proteins vaccine and a Lactobacillus-acidophilus–based direct-
fed microbial (DFM) for controlling Escherichia coli O157:H7, cattle (n=864) were allocated to the following
groups: DFM, finishing diets containing 10
9
colony-forming units (CFU)/animal/day L. acidophilus and Pro-
pionibacterium freudenreichii; VAC, finishing diets and 2 mL intramuscular injection of vaccine at allocation
and 28 days later; or CON, finishing diets only. Cattle within replicates were stratified by initial levels of E. coli
O157:H7 and randomized to experimental groups, with 30 pens allocated on June 15, 2011 (AS1), 18 pens
allocated on June 28, 2011 (AS2), and 18 cattle per pen. Rectal fecal samples and perineal swabs were collected
at 28-day intervals until shipment to slaughter (103–145 days on trial). Numbers of cattle with enumerable
E. coli O157:H7 ( ‡1.6 CFU/g feces) were reduced in AS1 and AS2 by VAC ( p=0.008), although interventions
had no impact on numbers of E. coli O157:H7 shed. For AS1, VAC reduced prevalence of E. coli O157:H7 in
feces ( p=0.03) and perineal swabs ( p=0.04) in the feeding period but not at shipment to slaughter. For AS2,
prevalence of E. coli O157:H7 was not reduced in either feces or perineal swabs by VAC at any time. For AS1,
DFM reduced prevalence of E. coli O157:H7 in perineal swabs ( p=0.01) during the feeding period. For AS2,
DFM increased E. coli O157:H7 detection in feces ( p=0.03) and perineal swabs ( p=0.01) at shipment to
slaughter. Seventy-five percent of AS1 E. coli O157:H7 isolates had only stx1, while 87% of AS2 isolates had
stx1 and stx2 genes. Of the two interventions, VAC shows the most potential for pre-harvest control of E. coli
O157:H7, but due to variable efficacy of both DFM and VAC, additional product development is necessary to
ensure more consistent pre-harvest control of E. coli O157:H7.
Introduction
Cattle are known E. coli O157:H7 reservoirs, and on-
farm methods investigated have included Lactobacillus
acidophilus–based direct-fed microbials (DFM) (Brashears
et al., 2003; Younts-Dahl et al., 2005; Peterson et al., 2007a;
Cull et al., 2012) and vaccines against type III secreted
proteins (Potter et al., 2004; Peterson et al., 2007b; Moxley
et al., 2009) or siderophore receptor and porin proteins
(Thomson et al., 2009; Fox et al., 2009; Cull et al., 2012).
Ideal on-farm methods would control this zoonotic pathogen
in feces to reduce environmental contamination (Berger
et al., 2010), on hides at slaughter to potentially reduce
contamination of beef products (Elder et al., 2000), and have
relatively long durations of control (feedlot entry to slaugh-
ter). Vaccination for on-farm control of E. coli O157:H7 has
recently been reviewed (Snedeker et al., 2012; Varela et al.,
2012; Matthews et al., 2013), but studies comparing vaccines
and DFM are rare. Accordingly, the objective of this study
was to compare the efficacy of a DFM to two doses of a type
III secreted-proteins vaccine for controlling E. coli O157:H7
in western-Canadian feedlot cattle over a complete feeding
period.
Materials and Methods
Facilities and feed
The project was conducted in a small pen research facility
( June–November 2011), using cattle housed in open air, dirt-
floor pens with central feed alleys and porosity fencing. Feed
and water were available ad libitum. Diets were standardized
within replicate and included approximately 92% barley
grain, 5% forage, 2% supplement, and 1% corn dried distiller
grains (cDDG) on a dry-matter basis to meet beef cattle
1
Alberta Agriculture and Rural Development, Agriculture Centre, Lethbridge, Alberta, Canada.
2
Feedlot Health Management Services Ltd., Okotoks, Alberta, Canada.
FOODBORNE PATHOGENS AND DISEASE
Volume 11, Number 5, 2014
ªMary Ann Liebert, Inc.
DOI: 10.1089/fpd.2013.1693
379
nutrient requirements (NRC, 2000). Diets were blended in
truck-mounted mixer boxes equipped with electronic load
cells, and contained monensin (Elanco Animal Health,
Guelph, ON), tylosin (Elanco Animal Health), and melen-
gestrol acetate (Pfizer Animal Health, heifer diets only).
The DFM was shipped on ice and frozen (-20C) until use.
The DFM was measured daily to target 1 g/animal/d and
added to diets using cDDG carrier. Pens within a replicate
were fed equal amounts of cDDG, and VAC and CON pens
were fed prior to DFM pens to avoid cross-contamination.
Residual DFM diet in feed trucks was removed by feeding
nonstudy cattle.
Cattle, interventions
Cattle were from two feedlots; with most allocation set 1
(AS1) cattle from Feedlot A and most allocation set 2
(AS2) from Feedlot B. Using new gloves for each animal,
feces (3–5 g) were rectally collected at feedlots A and B for
detection and enumeration of E. coli O157:H7. Cattle were
blocked by weight to facilitate marketing and stratified by
E. coli O157:H7 level (Negative, no E. coli O157:H7 de-
tected; Lower, insufficient for enumeration; Higher, ‡1.6
log colony-forming units (CFU)/g feces) and randomized to
equate levels across pens within replicates (Table 1). Cattle
were then transported 110 km to the research feedlot and
penned as per previous randomization. Each pen contained
at least 1 animal from each feedlot, with an average of 70%
of AS1 cattle from Feedlot A and 80% of AS2 cattle from
Feedlot B.
Study experimental groups were as follows: DFM, fin-
ishing diets containing BovamineCulture Complex
(Nutrition Physiology Company, LLC) with 10
9
CFU
L. acidophilus and Propionibacterium freudenreichii fed/
animal/d; VAC, standard finishing diets and 2 mL intramus-
cular injection of Econiche(Bioniche Life Sciences Inc.) at
allocation and 28 days later; or CON, standard finishing diets
only. The vaccine is licensed in Canada for 3 doses (28 days
apart), but the 2-dose regimen was approved by the manu-
facturer. Cattle were allocated on June 15 (AS1) and June 28
(AS2) 2011, respectively. Cattle (n=864) were housed by
allocation set and experimental group in 48 pens with 18
cattle/pen and 1 pen from each experimental group consti-
tuting a replicate. Pens of AS1 cattle (n=30) were steers
(initial weight 416.0 –27.2 kg), while pens of AS2 cattle
(n=18) were heifers (initial weight 402.1 –28.3 kg). Ex-
perimental groups were in contiguous pens in separate feedlot
alleys to avoid cross-contamination between interventions.
Cattle were observed daily for animal health, managed using
standard feedlot procedures, and followed for 103–145 days
until shipment to slaughter.
Sample collection
At 28-day intervals, rectal fecal samples were collected
from all cattle and perineal swabs were obtained from 4
random cattle/pen. All cattle were swabbed the day prior to
slaughter (last handling event). To collect perineal swabs,
a sterile SpongeSicle
(Med-Ox Diagnostics Inc., Ottawa,
ON) scrubbed an area 100 cm
2
below the anus and was
transported in 45 mL modified E. coli broth (mEC). Samples
were shipped chilled for processing within 24 h and stored at
5C until completion of analyses.
E. coli O157:H7 enumeration, detection
Each fecal sample was mixed to promote uniformity, and
two 1-g subsamples were enriched in 9 mL mEC (6 h at 37C).
Perineal swabs were incubated in mEC transport media (18 h at
37C). Enriched samples were subjected to immunomagnetic
separation (IMS) using anti-E. coli O157 Dynabeads
(In-
vitrogen, Carlsbad, CA). A 50-lL bead–bacteria mixture was
plated on sorbitol MacConkey agar with 2.5mg/L potassium
tellurite and 0.05mg/L cefixime (CT-SMAC) and incubated
(18–24 h at 37C). Three sorbitol-negative colonies/plate were
evaluated using E. coli O157 latex tests (Oxoid, Nepean, On-
tario, Canada), with positive colonies frozen in glycerol.
For enumeration, 1:10 dilutions from 1-g fecal subsamples
positive by IMS were prepared in mEC with 20 mg/L novo-
biocin (EMD, Gibbstown, NJ), and 100-lL duplicates were
plated on CT-SMAC. Plates contained 30–300 sorbitol-
negative colonies, with 5 colonies latex tested for O157
Table 1. Balancing Escherichia coli O157:H7
Within Pens of 18 Cattle Across Experimental
Groups at Allocation
Allocation
set E. coli O157 Control
a
DFM
a
Vaccine
a
Total
number
pens
110
6
-10
7
CFU 1 1 1 3
LS 3 3 3
Negative 14 14 14
110
5
-10
6
CFU 1 1 1 3
LS 3 3 3
Negative 14 14 14
110
4
-10
5
CFU 1 1 1 3
LS 3 3 3
Negative 14 14 14
110
3
-10
4
CFU 1 1 1 3
LS 3 3 3
Negative 14 14 14
110
2
-10
3
CFU 1 1 1 3
LS 3 3 3
Negative 14 14 14
1<10
2
CFU 1 1 1 6
LS 3 3 3
Negative 14 14 14
1LS7773
Negative 11 11 11
1LS6663
Negative 12 12 12
1LS5553
Negative 13 13 13
210
5
-10
6
CFU 1 1 1 3
Negative 17 17 17
210
4
-10
5
CFU 1 1 1 3
Negative 17 17 17
2LS3339
Negative 15 15 15
2LS2223
Negative 16 16 16
a
Number of animals per experimental group.
DFM, direct-fed antimicrobials; CFU, colony-forming units; LS,
low-shedder detected by immunomagnetic separation.
380 STANFORD ET AL.
antigen. Number of colonies/plate was adjusted as described
by Stephens et al. (2009). From O157 colonies frozen in
glycerol, polymerase chain reaction (PCR) confirmed E. coli
O157:H7 (Gannon et al., 1997; Paton and Paton, 1998).
Colonies were considered E. coli O157:H7 positive with
eaeA, fliC, and at least one Shiga-toxin gene.
Pulsed-field gel electrophoresis (PFGE)
A random subset of isolates confirmed E. coli O157:H7
(n=42) balanced by allocation set; treatment and Shiga-toxin
profile were subtyped by PFGE using XbaI restriction and
standard 1-d protocol (Ribot et al., 2006). One E. coli
O157:H7 isolate/positive sample was PFGE typed using a
CHEF DR II electrophoresis unit (Bio-Rad Laboratories,
Mississauga, ON). Banding patterns were viewed with UV
illumination and photographed using Speedlight Platinum
Gel Documentation System (Bio-Rad).
Statistical analyses
Analyses used SAS 9.2 (SAS Institute, Cary, NC) and were
considered significant at p<0.05 with pen the experimental
unit. Repeated-measure analyses for E. coli O157:H7 enu-
meration used mixed-models including experimental group,
allocation set (AS), and days-on-trial as fixed effects, with
three-way interaction of experimental group, AS and repli-
cate as a random effect and days-on-trial as repeated variable.
Prevalence of E. coli O157:H7 was analyzed using logistic
methodology within GLIMMIX and as significant two- and
three-way interactions with AS were present, analyses were
performed by AS, with experimental group, sample type, and
days-on-trial as fixed effects, and interaction of experimental
group and replicate as random effects. Model-adjusted means
(back-transformed to original scale) were reported, with ef-
ficacy estimated using the formula:
Efficacy ¼[1 (prevalence for intervention=
prevalence for control)] ·100
The PFGE patterns were classed as unique or grouped into
restriction endonuclease digestion pattern clusters (REPC;
‡90% similarity) using Dice similarity coefficients, un-
weighted-pair group methods arithmetic average algorithm,
1% position tolerance, and 0.5% optimization (BioNumerics
6.6, Applied Maths BVBA, Sin-Martens-Latem, Belgium).
Within- and between-group PFGE profile similarities were
tested using Dimensioning Techniques. For these analyses,
band-matched binary character profiles were created after
clustering using Dice coefficients. Bootstrap analysis (n=
1000) then assessed the similarity of PFGE profiles.
Results and Discussion
Study timing and location
Based on previous Alberta studies (Stanford et al., 2005a;
Stephens et al., 2009), peak E. coli O157:H7 prevalence
usually occurs during August and September. Consequently,
initial screening began in May, with cattle allocation by June.
FIG. 1. Fecal samples and perineal swabs positive for Escherichia coli O157:H7 and fecal samples with sufficient E. coli
O157:H7 for enumeration (higher-shedder, ‡1.6 log colony-forming units/g feces) from steers (n=540) allocated on June
15, 2011 (AS1) and heifers (n=324), allocated on June 28, 2011 (AS2).
z
Day of sampling, pre-treat, prior to initiation of interventions, no perineal swabs collected; day 103 (AS1 only slaughter)
n=216; day 118 (AS2 only slaughter) n =151; day 131 slaughter n =158 (AS2), n =264 (AS1); day 145 (AS1 only
slaughter) n =53.
A VACCINE AND DFM FOR E. COLI O157:H7 381
At allocation, E. coli O157 shedding was balanced across
replicates and experimental groups (Table 1), with an average
of 10% fecal positives in AS1 and AS2 (Fig. 1). Monitoring
occurred during E. coli O157:H7 season, with cattle
slaughtered September through early November. Although
prevalence was highest in July ( p<0.001; Table 2), E. coli
O157:H7 was sufficiently prevalent for intervention com-
parison throughout the feeding cycle (Fig. 1).
Cattle were housed in small rather than in commercial pens
(200–400 animals/pen). Animal behavior and disease trans-
mission are known to differ in large- and small-pen com-
parisons ( Jim et al., 1993; Perrett et al., 2008); however,
identifying sufficient E. coli O157:H7-positive cattle is a
challenge in large-pen studies. The small-pen design used
ensured adequate and initially balanced prevalence of E. coli
O157:H7 for experimental group comparisons and prevented
cross-contamination of interventions. The DFM dosage was
chosen to control E. coli O157:H7 as per past studies
(Younts-Dahl et al., 2004; Peterson et al., 2007a; Vascon-
celos et al., 2008) and has not previously improved feedlot
performance (Elam et al., 2003; Peterson et al., 2007a;
Vasconcelos et al., 2008).
Previous work evaluating pre-harvest interventions against
E. coli O157:H7 ensured sufficient prevalence by inoculating
cattle with E. coli O157:H7 (Allen et al., 2011) or using
naturally colonized cattle sampled only during peak season
(Cull et al., 2012). In the present study, a single fecal sample
was used to balance E. coli O157:H7 levels across experi-
mental groups and ensure sufficient prevalence as each pen
had at least one animal shedding E. coli O157:H7 prior
to initiation of interventions. Use of perineal swabs also en-
hanced detection as E. coli O157:H7 was 1.9 times more
likely ( p<0.001) to be detected from perineal swabs than
from feces (Table 2), even though swabs were collected from
4 of 18 cattle during the feeding period and feces were col-
lected from all cattle per pen.
E. coli O157:H7 enumeration
‘‘Super-shedders’’ (10
4
or greater CFU E. coli O157:H7/g
feces; Chase-Topping et al., 2008) have increased contami-
nation of hides of pen mates (Stephens et al., 2009) and beef
carcasses ( Jacob et al., 2010). In contrast to super-shedders,
most cattle positive for E. coli O157:H7 shed at relatively low
levels ( <100 CFU/g of feces) (Ferens and Hovde, 2011). As
less than 10% of cattle may be super-shedders (Chase-
Topping et al., 2008), cattle in this study were divided
between ‘‘lower-shedders’’ (detectable by IMS) and ‘‘higher-
shedders’’ (sufficient E. coli O157:H7 to enumerate; ‡1.6
log CFU/g feces).
The number of higher-shedder cattle was reduced 71.4%,
by VAC ( p=0.008; Fig. 2), which likely reduced the load of
E. coli O157:H7 in VAC pens, although VAC impact on
shedding duration was not assessed. While overall incidence
of higher-shedders was reduced, VAC did not prevent cattle
from becoming higher-shedders; VAC cattle shed up to 10
7
CFU E. coli O157:H7/g feces 1 month after booster and be-
fore final sampling (data not shown). Similar reductions in
higher-shedder cattle have been reported for siderophore
receptor-based vaccines (Fox et al., 2009; Cull et al., 2012).
For cattle shedding E. coli O157:H7, average number of
E. coli O157:H7 in feces did not differ by intervention, al-
though AS2 feces were approximately 1 log higher ( p=0.004)
than AS1 (Fig. 3). Previous response to VAC for numbers of
E. coli O157:H7 shed has varied. Potter et al. (2004) reported
significant reductions in cattle inoculated with the organism,
while Moxley et al. (2009) suggested that naturally colonized
cattle shed relatively similar numbers regardless of VAC. For
DFM, previous responses have also been mixed, with fecal
E. coli O157:H7 numbers reduced at dosages of 10
9
CFU/d
(Stephens et al., 2007a), while a companion study found no
impact on numbers in cattle feces with DFM doses from 10
7
to
10
9
CFU/d (Stephens et al., 2007b).
E. coli O157:H7 detection
E. coli O157:H7 prevalence in samples and intervention
responses differed ( p=0.02) by AS (Figs. 4 and 5). For AS2,
neither intervention reduced E. coli O157:H7 prevalence.
Compared to CON, AS2 VAC and DFM cattle had increased
prevalence of E. coli O157:H7 in perineal swabs within the
feeding period and at exit sampling ( p<0.05). It is unlikely
that interventions promoted E. coli O157:H7 shedding
in AS2 cattle and would instead reflect high variability of
Table 2. Probability of Detecting Escherichia coli O157:H7 from Allocation Set 1
a
and Allocation Set 2
b
Cattle as Influenced by Sample Type (Fecal Grab
or Perineal Swab)and Days on Trial During the Feeding Period
Allocation set 1 Allocation set 2
Variable Odds ratio 95% CI Significance Odds ratio 95% CI Significance
Sample p<0.001 p<0.001
Perineal swab 1.9 1.5–2.5 2.7 1.8–4.0
Fecal grab 1.0 Referent 1.0 Referent
Days on trial
c
p<0.001 p<0.001
28 12.5 8.3–16.7 12.5 7.1–20.0
56 5.0 3.4–7.1 2.1 1.2–3.7
84 1.4 1.0–2.2 1.5 0.8–2.9
112 1.0 Referent 1.0 Referent
a
Allocation set 1, steers (n=540) allocated on June 15, 2011 into 30 pens.
b
Allocation set 2, heifers (n=324) allocated June 28, 2011 into 18 pens.
c
Includes both perineal swab and fecal grab positives.
CI, confidence interval.
382 STANFORD ET AL.
shedding among individual AS2 cattle due to negligible
impact of interventions. In contrast, VAC reduced E. coli
O157:H7 prevalence in feces ( p=0.03) and in perineal swabs
(p=0.04) in AS1 cattle during the feeding period, although
neither reduction was significant at exit sampling. Efficacy of
VAC in AS1 during the feeding period for reducing E. coli
O157:H7 prevalence was 60% for feces and 56% for perineal
swabs (Fig. 4). For DFM, E. coli O157:H7 prevalence was
reduced ( p=0.01) in AS1 perineal swabs during the feeding
period, although not at exit sampling (Fig. 5). However, fecal
E. coli O157:H7 prevalence was never reduced in AS1
DFM-fed cattle. Having >10% of cattle positive for E. coli
O157:H7 prior to intervention initiation (Fig. 1) likely neg-
atively influenced DFM efficacy as proposed DFM mecha-
nisms may be most effective prior to gastrointestinal tract
colonization with E. coli O157:H7 (Callaway et al., 2004).
Although AS1 and AS2 were steers and heifers, respec-
tively, gender would not likely impact intervention efficacy.
A 2-week sampling schedule difference may have contrib-
uted to differing efficacies, but if minor changes in sample
collection timing were responsible for these differences,
DFM and VAC would be of limited utility for controlling
E. coli O157:H7 in commercial livestock. Possibly, differ-
ences in E. coli O157:H7 populations for AS1 and AS2
influenced intervention impacts. From PCR analyses, toxin
gene frequencies differed between AS ( p=0.004; data not
shown), with 87% of isolates from AS2 having stx1 and stx2,
while 75% of isolates from AS1 cattle had only stx1. Al-
though stx2 has suppressed development of cellular immune
responses to E. coli O157:H7 in cattle (Hoffman et al., 2006),
evaluation of a subset of AS2 data excluding isolates with
stx2 did not improve efficacy (data not shown). Matthews
0
10
20
30
40
50
60
Experimental group
a
b
a
CON DFM VAC
Number of samples
FIG. 2. Incidence of fecal samples with
sufficient Escherichia coli O157:H7 for
enumeration ( ‡1.6 log colony-forming units
(CFU)/g) from day 28 until shipment to
slaughter for cattle
z
fed the direct-fed mi-
crobial Bovamine
(Nutrition Physiology
Company, LLC) at a dosage of 10
9
CFU/
animal/d Lactobacillus acidophilus and
Propionibacterium freudenreichii starting
on day 0 (DFM), given a 2-mL intramus-
cular injection of Econiche
(Bioniche Life
Sciences Inc., Belleville, Ontario, Canada)
on day 0, and a booster on day 28 (VAC), or
given standard feedlot diets without DFM
and not vaccinated (CON).
a,b
Experimental Groups with different su-
perscripts differ, P=0.008.
z
No effect of allocation set (P>0.05) on
incidence of enumerable samples ( ‡1.6 log
CFU E. coli O157:H7/g feces).
0
1
2
3
4
5
6
a a
b
a
a,b
a,b
E. col i O157:H7 log CFU/ g feces
Experimental group by allocation set
CON-1 CON-2DFM-1 DFM-2VAC-1 VAC-2
z
FIG. 3. Average number of Escherichia coli O157:H7 per gram of feces by experimental group
y
and allocation set (1 or
2)
z
for complete study period.
a,b
Columns with different superscripts differ (P=0.03); overall effect of allocation set (P=0.004).
z
Allocation set 1 (steers, n =540, allocated June 15, 2011); Allocation set 2 (heifers, n =324, allocated June 28, 2011).
y
Experimental groups: CON, control; DFM, fed the direct-fed microbial Bovamine
(Nutrition Physiology Company, LLC)
at a dosage of 10
9
CFU/animal/day Lactobacillus acidophilus and Propionibacterium freudenreichii starting on day 0;
VAC, given a 2 mL intramuscular injection of Econiche
(Bioniche Life Sciences, Belleville ON) on day 0 and a booster on
day 28.
A VACCINE AND DFM FOR E. COLI O157:H7 383
FIG. 4. Overall impact of two doses of Econiche vaccine
(Bioniche Life Sciences Inc., Belleville, Ontario, Canada) on day
0 and a booster on day 28 on detection of Escherichia coli O157:H7 in feces and hide swabs from allocation set 1 (AS1; steers
allocated June 15, 2011) and allocation set 2 (AS2; heifers allocated June 28, 2011) during the feeding period (cumulative
incidence from day 56 to day 112) until shipment to slaughter (single-day incidence occurring day 103 to day 145).
z
Efficacy of VAC.
y
NS, not significant (P>0.05).
x
Negative, VAC prevalence of E. coli O157:H7 >Control (P<0.05).
FIG. 5. Overall impact of the direct-fed microbial (DFM) Bovamine
(Nutrition Physiology Company, LLC) fed to
allocation set 1 (AS1; steers allocated June 15, 2011) and allocation set 2 (AS2; heifers allocated June 28, 2011) at a dosage
of 10
9
colony-forming units (CFU) Lactobacillus acidophilus and Propionibacterium freudenreichii per animal per day
during the feeding period from day 0 until shipment to slaughter on detection of Escherichia coli O157:H7 in feces and hide
swabs during the feeding period (cumulative incidence from day 28 to day 112) until shipment to slaughter (single-day
incidence occurring day 103 to day 145).
z
NS, not significant (P>0.05).
y
Efficacy of DFM.
x
Negative, DFM prevalence of E. coli O157:H7 >Control (P<0.05).
384
et al. (2013) proposed that cattle shedding E. coli O157:H7
with stx2 were more likely to be super-shedders and that
presence of stx2 was a primary risk factor for human disease.
Consequently, reduced efficacy of VAC in AS2 as compared
to AS1 cattle may limit impacts of VAC on human health.
Additional study is required regarding estimated impacts of
interventions on human health and the role of stx2 on inter-
vention efficacy.
Variable efficacy of VAC and DFM in different populations
of cattle would be consistent with other studies. A review of
vaccines to reduce ruminant fecal shedding E. coli O157:H7
(Snedeker et al., 2012) concluded that care should be taken
interpreting findings due to variability among studies. When
pre-harvest and at-harvest outcomes were considered for either
type-III secreted proteins or siderophore receptor and porin
protein vaccines, meta-analysis also showed heterogeneous
results (Varela et al., 2012), although these authors found ho-
mogeneity in pre-harvest outcomes in two-dose studies of
VAC. As E. coli O157:H7 is a commensal organism not
causing disease in cattle, variable immune responses by ani-
mals may contribute to heterogeneous vaccine efficacy (Sne-
deker et al., 2012). In the present study, VAC response varied
by AS for pre-harvest measures and did not impact at-harvest
measures of E. coli O157:H7 contamination.
A large-scale feedlot evaluation of DFM efficacy (Cull
et al., 2012) also reported DFM ineffective for reducing fecal-
shedding of E. coli O157:H7, although these authors did not
measure hide contamination and used 10
6
CFU L. acid-
ophilus/head/d. Feeding the 10
9
CFU dosage, DFM previ-
ously reduced E. coli O157:H7 detection in feces and hides
(Brashears et al., 2003) or in feces but not hides (Younts-Dahl
et al., 2005; Stephens et al., 2007b). In this study, it is in-
triguing that both DFM and VAC interventions were most
effective in AS1 and did not control E. coli O157:H7 prev-
alence in AS2. More universally effective pre-harvest inter-
ventions are required, and the vaccine manufacturer in this
study has recently announced initiation of research for a
second-generation E. coli O157 vaccine (Anonymous, 2013).
PFGE genetic variability
From PFGE analyses, seven REPC and nine unique iso-
lates were detected, with most isolates from AS1 and AS2 in
separate REPC (data not shown). In REPC with co-mingling
of AS1 and AS2, only atypical AS2 isolates lacking stx2 were
present. Generally, isolates in REPC were from a single AS
and had similar stx genes; the only exception was REPC 6,
where AS1 cattle with stx1 or both stx1 and stx2 were present.
As AS1 and AS2 cohorts were mostly from different feedlots
and would have been penned separately at those feedlots as
well as in the present study, PFGE subtype bifurcation by AS
is not surprising. E. coli O157:H7 subtypes can vary by farm
(Rice et al., 1999; Stanford et al., 2005b) and by pen within
commercial feedlots (Sargeant et al., 2006; Stanford et al.,
2012), possibly related to the farm of origin of the cattle.
From bootstrapping analyses, AS1 isolates had more
similar PFGE profiles compared to all isolates ( p<0.001),
while AS2 isolates showed less uniformity ( p=0.08; Table
3). The PFGE profiling revealed that within-group similarity
of isolates with stx1 alone was higher than when compared to
all isolates ( p<0.001), while within-group similarity of
isolates with stx1 and stx2 genes was lower when compared
to all isolates ( p=0.11). Subdividing isolates with stx1 and
stx2 genes by AS, both sets of PFGE profiles were diverse.
Overall, PFGE analyses demonstrated that AS1 E. coli
O157:H7 subtypes were relatively homogeneous compared
to increased diversity of AS2 isolates. This may at least
partially explain the limited efficacy of interventions for AS2.
Subtypes of E. coli O157:H7 from AS1 and AS2 may dif-
ferently express genes in the locus of enterocyte effacement
(LEE) encoding the type III proteins secretion system (Roe
et al., 2004). Alternatively, proteins other than those encoded
by LEE also contribute to E. coli O157:H7 adherence in the
bovine gastrointestinal tract (Kudva et al., 2012) and may be
more common in AS2. Studies to further characterize E. coli
O157:H7 isolates from AS1 and AS2 are in progress.
Conclusions
Although VAC reduced the numbers of cattle shedding
enumerable E. coli O157:H7, it did not prevent individuals
shedding up to 10
7
CFU/g feces. Neither VAC nor DFM
consistently reduced prevalence of E. coli O157:H7 in feces
or perineal swabs from either AS of cattle, and neither in-
tervention lowered numbers of E. coli O157:H7 shed in feces.
Intervention efficacy may have been negatively influenced by
an initial shedding rate ( >10% of cattle) and presence of
different E. coli O157:H7 subtypes. As VAC and DFM had
variable efficacy, additional development is required prior to
recommending either pre-harvest E. coli O157:H7 interven-
tion to the commercial beef industry.
Acknowledgments
This work was funded by the Alberta Livestock and Meat
Agency.
Disclosure Statement
Bioniche Life Sciences Inc. and Nutrition Physiology
Company LLC. contributed product and funding to Feedlot
Health Management Services Ltd. to collectively offset 9%
of project costs.
Table 3. Average Group Similarity of Pulsed-Field
Gel Electrophoresis (PFGE) Profiles by Source
of Escherichia coli O157:H7 Isolates Compared
to the Similarity of PFGE Profiles for All Isolates
Grouping n
Average
similarity
within
group (%)
Average
similarity
between
groups (%)
Significance
of group
By allocation set
AS1 20 70.72 66.02 p<0.001
AS2 22 67.95 66.02 p=0.08
By Stx genes
Stx1 only 20 71.30 66.05 p<0.001
Stx2 only 1 NA NA NA
Stx1 and Stx2 21 59.09 57.39 p=0.11
Having both Stx genes by allocation set
AS1 6 69.11 68.75 p=0.49
AS2 15 67.04 66.04 p=0.28
NA, not applicable as minimum of two needed for comparisons.
A VACCINE AND DFM FOR E. COLI O157:H7 385
References
Allen KJ, Rogan D, Finlay BB, Potter AA, Asper DJ. Vacci-
nation with type III secreted proteins leads to decreased
shedding in calves after experimental infection with Escher-
ichia coli O157. Can J Vet Res 2011;75:98–105.
Anonymous. Bioniche Life Sciences Inc. receives government
of Canada support for second generation E. coli O157 vac-
cine. Bioniche News 2013. Available at: www.bioniche@
bioniche.com, accessed March 27, 2013.
Berger CN, Sodha SV, Shaw RK, Griffin PM, Pink D, Hand P,
Frankel G. Fresh fruit and vegetables as vehicles for the
transmission of human pathogens. Environ Microbiol 2010;
12:2385–2397.
Brashears MM, Galyean ML, Loneragan GH, Mann JE,
Killinger-Mann K. Prevalence of Escherichia coli O157:H7
and performance by beef feedlot cattle given Lactobacillus
direct-fed microbials. J Food Protect 2003;66:748–754.
Callaway TR, Anderson RC, Edrington TS, Genovese KJ,
Bischoff KM, Poole TL, Jung YS, Harvey RB, Nisbet DJ.
What are we doing about Escherichia coli O157:H7 in cattle?
J Anim Sci 2004;82(Suppl):E93–E99.
Chase-Topping M, Gall D, Low C, Matthews L, Woolhouse M.
Super-shedding and the link between human infection and
livestock carriage of Escherichia coli O157. Nature Rev
Microbiol 2008;6:904–912.
Cull CA, Paddock ZD, Nagaraja TG, Bello NM, Babcock AH,
Renter DG. Efficacy of a vaccine and a direct-fed microbial
against fecal shedding of Escherichia coli O157:H7 in a
randomized pen-level field trial of commercial feedlot cattle.
Vaccine 2012;30:6210–6215.
Elam NA, Cleghorn JF, Rivera JD, Galyean ML, Defoor PJ,
Brashears MM, Younts-Dahl SM. Effects of live cultures
of Lactobacillus acidophilus (strains NP45 and NP51 and
Propionibacterium freudenreichii) on performance, carcass
and intestinal characteristics and Escherichia coli O157
shedding of finishing feedlot steers. J Anim Sci 2003;81:
2686–2698.
Elder RO, Keen JE, Siragusa GR, Barkocy-Gallagher GA,
Koohmaraie M, Laegreid WW. Correlation of entero-
hermorrhagic Escherichia coli prevalence in feces, hides, and
carcasses of beef cattle during processing. Proc Natl Acad Sci
U S A 2000;97:2999–3003.
Ferens WA, Hovde CJ. Escherichia coli O157:H7: Animal
reservoir and sources of human infection. Foodborne Pathog
Dis 2011;8:465–487.
Fox JT, Thomson DU, Drouillard JS, Thornton AB, Burkard
DT, Emery DA, Nagaraja TG. Efficacy of Escherichia coli
O157:H7 siderophore receptor/porin proteins-based vaccine
in feedlot cattle naturally shedding E. coli O157. Foodborne
Pathog Dis 2009;6:893–899.
Gannon VPJ, D’Souza S, Graham T, King RK, Rahn K, Read S.
Use of the flagellar H7 gene as a target in multiplex PCR
assays and improved specificity in identification of en-
terohemorrhagic Escherichia coli strains. J Clin Microbiol
1997;35:656–662.
Hoffman MA, Menge C, Casey TA, Laegreid W, Bosworth
BT, Dean-Nystrom EA. Bovine immune response to Shiga-
toxigenic Escherichia coli O157:H7. Clin Vac Immunol
2006;13:1322–1327.
Jacob ME, Renter DG, Nagaraja TG. Animal and truckload-
level associations between Escherichia coli O157:H7 in feces
and on hides at harvest and contamination of pre-evisceration
beef carcasses. J Food Protect 2010;73:1030–1037.
Jim GK, Booker CW, Ribble CS, Guichon PT, Thorlakson BE.
A field investigation of the economic impact of respiratory
disease in feedlot calves. Can Vet J 1993;34:668–673.
Kudva IT, Griffin RW, Krastins B, Sarracino DA, Calderwood
SB, John M. Proteins other than the locus of enterocyte ef-
facement-encoded proteins contribute to Escherichia coli
O157:H7 adherence to bovine rectoanal junction stratified
squamous epithelium. BMC Microbiol 2012;12:103.
Matthews L, Reeve R, Gally DL, Low JC, Woolhouse MEJ,
McAteer SP, Locking ME, Chase-Topping ME, Haydon DT,
Allison LJ, Hanson MF, Gunn GJ, Reid SWJ. Predicting the
public health benefit of vaccinating cattle against Escherichia
coli O157. Proc Natl Acad Sci U S A 2013;110:16265–
16270.
Moxley RA, Smith DR, Luebbe M, Erickson GE, Klopfenstein
TJ, Rogan D. Escherichia coli O157:H7 vaccine dose—
Effect in feedlot cattle. Foodborne Pathog Dis 2009;6:879–
884.
[NRC] National Research Council. Nutrient Requirements of
Beef Cattle, 7th revised ed. Washington, DC: National Re-
search Council, National Academy Press, 2000.
Paton AW, Paton JC. Detection and characterization of Shiga
toxigenic Escherichia coli by using multiplex PCR assays for
Stx1,Stx2,eaeA, enterohemorrhagic E. coli hylA, rfb
0111
and
rfb
0157
. J Clin Microbiol 1998;36:598–602.
Perrett T, Wildman BK, Jim GK, Vogstad AR, Fenton RK,
Hannon SJ, Schunicht OC, Abutarbush SM, Booker CW.
Evaluation of the efficacy and cost-effectiveness of me-
lengestrol acetate in feedlot heifer calves in western Canada.
Vet Ther 2008;9:223–240.
Peterson RE, Klopfenstein TJ, Erickson GE, Folmer J, Hinkley
S, Moxley RA, Smith DR. Effect of Lactobacillus acid-
ophilus strain NP51 on Escherichia coli O157:H7 fecal
shedding and finishing performance in beef feedlot cattle.
J Food Protect 2007a;70:287–291.
Peterson RE, Klopfenstein TJ, Moxley RA, Erickson GE,
Hinkley S, Bretschneider G, Berbervov EM, Rogan D, Smith
DR. Effect of a vaccine product containing type III secreted
proteins on the probability of Escherichia coli O157:H7 fecal
shedding and mucosal colonization in feedlot cattle. J Food
Protect 2007b;70:2568–2577.
Potter AA, Klashinsky S, Li Y, Frey E, Townsend H, Rogan D,
Erickson G, Hinkley S, Klopfenstein T, Moxley RA, Smith
DR, Finlay BB. Decreased shedding of Escherichia coli
O157:H7 by cattle following vaccination with type III se-
creted proteins. Vaccine 2004;22:362–369.
Ribot EM, Fair MA, Gautom R, Cameron DN, Hunter SB,
Swaminathan B, Barrett TJ. Standardization of pulsed-field
gel electrophoresis protocols for the subtyping of Escherichia
coli O157:H7, Salmonella and Shigella for PulseNet. Food-
borne Pathog Dis 2006;3:59–66.
Rice DH, McMenamin KM, Prichett LC, Hancock DD, Besser
TE. Genetic subtyping of Escherichia coli O157 isolates from
41 Pacific Northwest USA cattle farms. Epidemiol Infect
1999;122:479–484.
Roe AJ, Naylor SW, Spears KJ, Yull HM, Dransfield TA,
Oxford M, McKendrick IJ, Porter M, Woodward MJ, Smith
DE, Gally DL. Co-ordinate single-cell expression of LEE4
and LEE5-encoded proteins of Escherichia coli O157:H7.
Mol Microbiol 2004;54:337–352.
Sargeant JM, Shi X, Sanderson MW, Renter DG, Nagaraja TG.
Pulsed-field gel electrophoresis patterns of Escherichia coli
O157 isolates from Kansas feedlots. Foodborne Pathog Dis
2006;3:251–258.
386 STANFORD ET AL.
Snedeker KG, Campbell M, Sargeant JM. A systematic review
of vaccinations to reduce the shedding of Escherichia coli
O157 in the faeces of domestic ruminants. Zoonoses Publ
Health 2012;59:126–138.
Stanford K, Agopsowicz CA, McAllister TA. Antimicrobial
resistance and genetic diversity among isolates of Escher-
ichia coli O157:H7 from feces and hides of super-shedders
and low-shedding pen-mates. BMC Vet Res 2012;8:178.
Stanford K, Bach SJ, Marx TH, Jones S, Hansen JR, Wallins
GL, Zahiroddini H, McAllister TA. Monitoring Escherichia
coli O157:H7 in inoculated and naturally infected feedlot
cattle and their environment. J Food Protect 2005a;68:26–33.
Stanford K, Croy D, Bach SJ, Wallins GL, Zahiroddini H,
McAllister TA. Ecology of Escherichia coli O157:H7 in
commercial dairies in southern Alberta. J Dairy Sci 2005b;88:
4441–4451.
Stephens TP, McAllister TA, Stanford K. Use of perineum hide
swabs to document the impact of super shedders on the
transmission of Escherichia coli O157 in commercial feed-
lots. J Anim Sci 2009;87:4151–4160.
Stephens TP, Loneragan GH, Karunasena E, Brashears MM.
Reduction of Escherichia coli O157 and Salmonella in feces
and on hides of feedlot cattle using various doses of a direct-
fed microbial. J Food Protect 2007b;70:2386–2391.
Stephens TP, Loneragan GH, Karunasena E, Brashears MM.
Prevalence and enumeration of Escherichia coli O157 in
steers receiving various strains of Lactobacillus-based direct-
fed microbials. J Food Protect 2007a;70:1252–1255.
Thomson DU, Loneragan GH, Thornton AB, Lechtenbergh KF,
Emery DA, Burkdart DT, Nagaraja TG. Use of a siderophore
receptor and porin proteins-based vaccine to control the
burden of Escherichia coli O157:H7 in feedlot cattle. Food-
borne Pathog Dis 2009;6:871–877.
Varela NP, Dick P, Wilson J. Assessing the existing information
on the efficacy of bovine vaccination against Escherichia coli
O157:H7—A systematic review and meta-analysis. Pro-
ceedings of the John Waters Zoonotic Disease Workshop,
October 16, 2012, Banff, Alberta.
Vasconcelos JT, Elam NA, Brashears MM, Galyean ML. Ef-
fects of increasing dose of live cultures of Lactobacillus
acidophilus (Strain NP 51) combined with a single dose of
Propionibacterium freudenreichii (Strain NP 24) on perfor-
mance and carcass characteristics of finishing beef steers.
J Anim Sci 2008;86:756–762.
Younts-Dahl SM, Osborn GD, Galyean ML, Rivera JD, Lone-
ragan GH, Brashears MM. Reduction of Escherichia coli
O157 in finishing beef cattle by various doses of Lactoba-
cillus acidophilus in direct-fed microbials. J Food Protect
2005;68:6–10.
Younts-Dahl SM, Galyean ML, Loneragan GH, Elam NA,
Brashears MM. Dietary supplementation with Lactobacillus-
and Propionibacterium-based direct-fed microbials and
prevalence of Escherichia coli O157 in beef feedlot cattle and
on hides at harvest. J Food Protect 2004;67:889–893.
Address correspondence to:
Kim Stanford, PhD
Alberta Agriculture and Rural Development
Agriculture Centre
5401-1st Avenue South
Lethbridge, Alberta, T1J 4V6
Canada
E-mail: kim.stanford@gov.ab.ca
A VACCINE AND DFM FOR E. COLI O157:H7 387
