Vaccine 33 (2015) 100107

Contents lists available at ScienceDirect

Vaccine
journal homepage: www.elsevier.com/locate/vaccine

Development of a Salmonella cross-protective vaccine for food animal

production systems
Douglas M. Heithoff a , John K. House b , Peter C. Thomson b , Michael J. Mahan a,
a
b

Department of Molecular, Cellular and Developmental Biology, University of California, Santa Barbara, CA 93106, USA
University of Sydney, Faculty of Veterinary Science, Camden, NSW 2570, Australia

a r t i c l e

i n f o

Article history:
Received 4 September 2014
Received in revised form 7 November 2014
Accepted 8 November 2014
Available online 20 November 2014
Keywords:
Cross-protective Salmonella vaccine
DNA adenine methylation
Dam
Salmonellosis

a b s t r a c t
Intensive livestock production is associated with increased Salmonella exposure, transmission, animal
disease, and contamination of food and water supplies. Modied live Salmonella enterica vaccines that
lack a functional DNA adenine methylase (Dam) confer cross-protection to a diversity of salmonellae in
experimental models of murine, avian, ovine, and bovine models of salmonellosis. However, the commercial success of any vaccine is dependent upon the therapeutic index, the ratio of safety/efcacy. Herein,
secondary virulence-attenuating mutations targeted to genes involved in intracellular and/or systemic
survival were introduced into Salmonella dam vaccines to screen for vaccine candidates that were safe in
the animal and the environment, while maintaining the capacity to confer cross-protective immunity to
pathogenic salmonellae serotypes. Salmonella dam mgtC, dam sifA, and dam spvB vaccine strains exhibited
signicantly improved vaccine safety as evidenced by the failure to give rise to virulent revertants during
the infective process, contrary to the parental Salmonella dam vaccine. Further, these vaccines exhibited a low grade persistence in host tissues that was associated with reduced vaccine shedding, reduced
environmental persistence, and induction of cross-protective immunity to pathogenic serotypes derived
from infected livestock. These data indicate that Salmonella dam double mutant vaccines are suitable for
commercial applications against salmonellosis in livestock production systems. Reducing pre-harvest
salmonellae load through vaccination will promote the health and productivity of livestock and reduce
contamination of livestock-derived food products, while enhancing overall food safety.
2014 Elsevier Ltd. All rights reserved.

1. Introduction
Salmonellosis is a common cause of morbidity and mortality
in intensive livestock production systems. Many Salmonella infections are subclinical with disease more commonly observed in
young, debilitated and parturient animals which are most susceptible to infection [1]. Disease outbreaks compromise animal welfare,
promote antimicrobial use and subsequently lead to selection for
antimicrobial resistance in zoonotic pathogens, compromising productivity and at times resulting in substantial mortality (2060%
of the susceptible population [2]). There are over 2500 Salmonella
serotypes, all of which are potentially pathogenic, however, 80%
of clinical disease in a given region is frequently attributed to less
than 2% of serotypes with those of serogroups B, C, D and E most frequently incriminated [3]. Sources of infection for livestock include
feed, water, wildlife, insects, people and contaminated equipment
[4]. Infected animals amplify environmental contamination and

Corresponding author. Tel.: +61 018058937160.

E-mail address: mahan@lifesci.ucsb.edu (M.J. Mahan).
http://dx.doi.org/10.1016/j.vaccine.2014.11.012
0264-410X/ 2014 Elsevier Ltd. All rights reserved.

represent a source of infection for other stock. Efuent from intensive animal production systems is often utilized as a fertilizer to
grow crops, and Salmonella-contaminated efuent provides a reservoir of infection and a vehicle for contamination of waterways,
fruits and vegetables. Thus, Salmonella contamination of animal
and plant food products by a variety of mechanisms provides an
ongoing source of infection for human populations.
Salmonella control efforts continue to be problematic for the following reasons: (1) most livestock infections are subclinical [2,5];
(2) disease outbreaks are sporadic and frequently caused by specic serotypes although many serotypes are endemic to livestock
production systems [57]; (3) environmental persistence provides
an ongoing reservoir for livestock infection [811]; (4) the recent
emergence of strain variants that are more virulent and can kill
vaccinated animals [12]; (5) some strains derived from human
salmonellosis patients are distinct from those of animal origin [13];
(6) management and environmental events can increase pathogen
exposure and/or compromise host immunity [1,2,8,9,14,15].
Vaccination represents a sustainable approach for promoting animal health, welfare, and food safety through reducing
pathogen exposure at the outset of the food production chain [16].

D.M. Heithoff et al. / Vaccine 33 (2015) 100107

However, the immunity conferred by conventional vaccines is

restricted to a narrow range of closely-related strains, and on-farm
control requires the development of vaccines that elicit protection against many pathogenic serotypes [16]. Recent advancements
have resulted in the development of modied live Salmonella
cross-protective vaccines, many of which contain mutations in
global regulatory networks that favor antigen production, and
that are also suitable for the expression of heterologous antigens
[1721]. The molecular basis of cross-protective vaccine efcacy
may be attributed to several parameters including: the expression
of multiple antigens shared among pathogenic serotypes; diminished vaccine-induced immunosuppression; targeted removal of
immunodominant antigens to expose cross-protective epitopes;
type III secretion of recombinant antigens; and/or delayed vaccine attenuation for enhanced stimulation of immune responses
(reviewed in [16,18,22,23]). Modied live attenuated S. enterica
serovar Typhimurium that harbor loss of function mutations in the
gene encoding the DNA adenine methylase (dam) are capable of
eliciting protection against a diversity of salmonellae and are welltolerated when applied as modied live vaccines in mice [17,24],
poultry [25,26], sheep [27] and calves [2830]. Induction of immunity is rapid and the vaccine can be administered via drinking water
for low-cost and low-stress immunization of livestock populations
[27,31]. The commercial success of any vaccine is dependent on the
therapeutic index, the ratio of safety/efcacy, and safety is of particular concern for modied live vaccines that have the potential
to revert to heightened virulence. Herein, we evaluated whether
the introduction of attenuating mutations into a Salmonella dam
mutant vaccine strain signicantly improves vaccine safety without
compromising efcacy in immunized mice. The vaccine candidates
that offer the best safety and protection characteristics are the
most suitable for a commercially viable product against livestock
salmonellosis.

101

2.3. Statistical analysis

Continuous repeated measures data were analyzed using residual (or restricted) maximum likelihood (REML) analysis (GenStat,
15th Edition, VSN International, UK). A single variate, repeated
measures model was used to analyze CFU on a log base 10 scale. The
xed effects of the model were the factors time, treatment and their
interaction, and repeated observations on the same mouse were
allowed for in the residual term specication. The Wald chi-square
test was used to determine signicant main effects and/or signicant interactions between factors. Any non-signicant terms were
dropped from the model and analysis repeated. Following analysis,
data are presented as predicted model-based means, i.e. predicted
means are those obtained from the tted model rather than the
raw sample means. P-values less than 0.05 were considered to be
statistically signicant. The number of CFU present in tissues at
necropsy was analyzed using analysis of variance (ANOVA, GenStat, 15th Edition, VSN International, UK). Differences between the
individual means calculated using REML and ANOVA were determined by calculating an approximate least signicant difference
(LSD). A difference of means that exceeded the calculated LSD was
considered signicant.
Binomial data (shedding [yes/no] and outcome [live/dead])
were analyzed using a logistic regression model (GenStat, 15th
Edition, VSN International, UK). Vaccine status was specied as a
predictor in the model. Overall signicance was assessed using the
Wald statistic (P < 0.05). Signicance of xed effects (vaccine) was
assessed according to the t parameter estimates relative to the reference group. P-values less than 0.05 were considered statistically
signicant.

2.4. Ethics statement

2. Materials and methods
2.1. Bacterial strains and growth conditions
Salmonella animal isolates were derived from different outbreaks, individual cases, or surveillance submissions to diagnostic
laboratories [13]. Virulent Salmonella Typhimurium UK-1 was used
in all studies for comparison [32]. Unless otherwise specied, bacteria were derived from stationary phase cultures aerated at 37 C
containing Luria-Bertani (LB) medium [33]. Antibiotics were used
at the following concentrations: kanamycin (Kn), 50 g/ml; ampicillin (Ap), 50 g/ml.

2.2. Construction of S. Typhimurium dam vaccine candidates

comprising an additional attenuating mutation
S. Typhimurium UK-1 dam was constructed by introducing
an in-frame 300 bp deletion of dened dam sequence, termed
dam232 [34], using standard genetic protocols [35]. Secondary
virulence-attenuating deletion mutations were introduced into
the parental S. Typhimurium UK-1 dam232 strain utilizing
ampicillin-resistant suicide vector pCVD442 as described [35],
resulting in the construction of in-frame deletions of dened coding sequence in the following targeted genes: aroA (MT3138;
1056 bp); htrA (MT3142; 1341 bp); mgtC (MT3146; 606 bp);
sifA (MT3150; 807 bp); spiC (MT3154; 306 bp); spvB (MT3158;
1563 bp); and ssaV (MT3162; 1959 bp). The resultant genetic constructs were conrmed by PCR using primers that ank the deleted
sequences.

All animal experimentation was conducted following the

National Institutes of Health guidelines for housing and care of
laboratory animals and performed in accordance with Institutional
regulations after pertinent review and approval by the Institutional
Animal Care and Use Committee at the University of California,
Santa Barbara.

3. Results
3.1. Evaluation of Salmonella dam double mutant vaccine
candidates for colonization and persistence in mucosal and
systemic tissues
Colonization/persistence of the Salmonella dam double mutant
vaccine candidates (dam aroA, dam htrA, dam mgtC, dam sifA,
dam spiC, dam spvB, dam ssaV) in BALB/c mice was assessed in
mucosal (Peyers patches; mesenteric lymph nodes) and systemic
tissues (liver and spleen) at 2 and 4 weeks post oral immunization (Fig. 1a and b). The Salmonella dam double mutant candidates
were classied into two groups: Class (I) those that showed
similar colonization/persistence relative to that of the parental
Salmonella dam vaccine strain (dam mgtC, dam sifA, dam spvB);
and Class (II) those that exhibited reduced colonization/persistence
relative to that exhibited by the parental dam vaccine (dam
aroA, dam htrA, dam spiC, dam ssaV). These data indicate that
Class I vaccines sustained a low-grade persistence in host tissues that may be necessary for elicitation of robust protection,
whereas Class II vaccines showed rapid clearance in vaccinated
animals.

102

D.M. Heithoff et al. / Vaccine 33 (2015) 100107

1.00E+06
106

2 weeks post oral immunization

A
100
90
80

% Surviving Mice

Average CFU/g

1.00E+05
105

PP
MLN
S
L

1.00E+04
104

1.00E+03
103

1.00E+02
102

70
60
50

**

40

***

30

***

20

***

10

Average CFU/g

1.00E+05
105

4 weeks post oral immunization

1.00E+04
104

B
100

1.00E+03
103

90

1.00E+02
102

Fig. 1. Evaluation of Salmonella dam double mutant vaccine candidates for colonization and persistence in mucosal and systemic tissues. BALB/c mice (n = 5 per time
point) were orally immunized with S. Typhimurium UK-1 dam232 double mutant
vaccine candidates (dam aroA [MT3138], dam htrA [MT3142], dam mgtC [MT3146],
dam sifA [MT3150], dam spiC [MT3154], dam spvB [MT3158], dam ssaV [MT3162]) or
the parental UK-1 dam232 vaccine strain (MT3134) (109 CFU). At 2 weeks (A) and
4 weeks (B) post oral immunization, bacteria recovered from Peyers patches (PP),
mesenteric lymph nodes (MLN), liver (L) and spleen (S) were assessed for colony
forming units (CFU) on LB medium. Limits of detection: PP, MLN, spleen < 100 CFU;
liver < 50 CFU.

3.2. Efcacy evaluation of Salmonella dam double mutant

vaccine candidates
The Salmonella dam double mutant vaccine candidates (Class
I and Class II) were examined to discern whether a low-grade
persistence is necessary to confer protective immune responses
similar to that elicited by the parental Salmonella dam vaccine.
BALB/c mice were orally immunized with each of the seven vaccine
candidates, and orally challenged with a 200-fold LD50 infection dose of the virulent parental strain, S. Typhimurium UK-1.
Vaccination with all seven candidates provided signicant protection against homologous challenge (200 LD50 s) in comparison
to non-vaccinated mice (P < 0.001). Similar levels of protection
were afforded by the Class I vaccines incorporating the secondary
deletions (mgtC, sifA, and spvB) as compared to the parent dam
vaccine (Fig. 2a). Conversely, a signicant reduction in the efcacy of the Class II vaccines was observed following introduction
of the secondary deletions (aroA, htrA, spiC, and ssaV) (**P < 0.01;
***P < 0.001).
Next, the Class I vaccine candidates were evaluated for the
capacity to elicit heterologous protection. BALB/c mice were orally
immunized and orally challenged with a 100-fold LD50 infection dose of livestock industry-relevant pathogenic Salmonella
strains derived from sheep (Salmonella Bovismorbicans 174, S.
Typhimurium 131) and cattle (Salmonella Dublin 8895), comprising

% Surviving Mice

80

***

S. Typhimurium UK-1
S. Bovismorbificans 225
S. Dublin 8895
S. Typhimurium 131

***

***

70
60
50
40

30
20
10
0

No
No vaccine
vaccine

dam mgtC

dam sifA

dam spvB

Vaccine Strain
Fig. 2. Efcacy evaluation of Salmonella dam double mutant vaccine candidates.
BALB/c mice (1625 per cohort) were orally immunized with S. Typhimurium UK1 dam232 double mutant vaccine candidates (dam aroA [MT3138], dam htrA
[MT3142], dam mgtC [MT3146], dam sifA [MT3150], dam spiC [MT3154], dam spvB
[MT3158], dam ssaV [MT3162]) or the parental UK-1 dam232 vaccine strain
(MT3134) (109 CFU). Eleven weeks post-immunization, vaccinated mice were challenged with an oral dose of (A), 200 LD50 of homologous strain, wild-type S.
Typhimurium UK-1 (3761) or (B), 100 LD50 of heterologous Salmonella serotypes
of clinical relevance to the livestock industry (S. Dublin 8895 [cattle], S. Bovismorbicans 225 [sheep], S. Typhimurium 131 [sheep]). The oral and i.p. LD50 for
S. Typhimurium UK-1 is 105 and <10 organisms, respectively [12]. Non-vaccinated
control mice all died by day 21 post-infection. Differences in the proportion of mice
surviving virulent challenge were analyzed using logistic regression (GenStat 15th
Edition [53]). Homologous protection: Vaccination with all candidates provided signicant protection against homologous challenge in comparison to non-vaccinated
mice (P < 0.001). A signicant reduction in the efcacy of the dam vaccine was
observed following introduction of the secondary deletions (aroA, htrA, spiC, and
ssaV) (**P < 0.01; ***P < 0.001). Heterologous protection: Vaccination with each of the
dam vaccines incorporating the secondary deletions (mgtC, sifA, and spvB) provided
signicant protection against heterologous challenge strains assessed relative to
that observed in non-vaccinated mice (***P < 0.001).

serogroups C2C3, B, and D, respectively. All three Class I vaccine

candidates conferred robust cross-protection to the three heterologous virulent strains tested relative to non-vaccinated mice (Fig. 2b;
***P < 0.001), similar to the levels of cross-protection exhibited previously against these three heterologous challenge strains in mice
vaccinated with a S. Typhimurium 14028 dam232 vaccine strain
[17]. Taken together, these data indicate that low-grade vaccine
persistence is congruent with high levels of cross-protection.

D.M. Heithoff et al. / Vaccine 33 (2015) 100107

10

10

10

10

Day 5 Liver

CFU per gram

CFU per gram

Day 5 Spleen

103

10

10

10

10

dam

dam mgtC dam sifA

dam spvB

dam mgtC dam sifA dam spvB

dam

Fig. 3. Vaccine safety evaluation (reversion to 2-AP resistance) among Salmonella dam double mutant vaccine candidates. BALB/c mice (5 per cohort) were intraperitoneally
infected with 105 CFU of S. Typhimurium UK-1 dam232 double mutant vaccine candidates (dam mgtC [MT3146], dam sifA [MT3150], dam spvB [MT3158]) or the parental UK-1
dam232 vaccine strain (MT3134). The number of 2-AP sensitive (open boxes) or 2-AP resistant (closed boxes) Salmonella organisms in the spleen or liver was enumerated
at day 5 post-infection. The symbols below the zero CFU value represent the number of mice in which the bacterial load in spleen and liver was below the limit of detection
(<25 CFU). Statistical signicance for Salmonella dam double mutant vaccine candidate persistence (2-APs ) and reversion to heightened virulence (2-APr ) in comparison to
the parental Salmonella dam232 vaccine strain was determined using analysis of variance (*P < 0.05).

3.3. Vaccine safety evaluation via assessment of reversion to 2-AP

resistance
Salmonella dam mutant vaccines have the capacity to undergo
reversion to a more virulent state via acquisition of a mutation(s)
in methyl-directed mismatch repair genes [36]. Such reversion can
be evaluated using the purine analog 2-aminopurine (2-AP), which
is toxic to bacteria lacking Dam function [37]. BALB/c mice were
i.p. infected with Salmonella dam double mutant vaccine candidates
(dam mgtC, dam sifA, dam spvB) or the parental S. Typhimurium UK-1
dam232 strain (105 CFU). Five days following infection, bacteria recovered from the spleen and liver were assessed for 2-APs

(persistence) and reversion to the 2-APr phenotype (Fig. 3). All

three vaccine candidates showed signicantly reduced colonization/persistence (2-APs ) and reduced reversion to 2-AP resistance
in the spleen and liver relative to that of the parental Salmonella
dam vaccine (*P < 0.05).
Next, the 2-APr derivatives were evaluated for enhanced virulence in nave mice via oral and i.p. lethal-dose 50 (LD50 ) virulence
assays. 2-APr isolates derived from either the Salmonella dam double mutant vaccine candidates (11 of 11), or the parental Salmonella
dam vaccine (5 of 5), were avirulent by the oral route of administration (Table 1). In contrast, all (11/11) 2-APr isolates derived from
Salmonella dam double mutant vaccine candidates were highly

Table 1
In vivo-selected 2-APr derivatives of dam double mutant Salmonella are avirulent via the intraperitoneal or oral routes of infection.
Challenge strain

Relevant genotype

S. Typhimurium UK-1

Wild type

MT3134
MT3243
MT3244
MT3245
MT3246
MT3247

dam232
dam232 2-APr
dam232 2-APr
dam232 2-APr
dam232 2-APr
dam232 2-APr

MT3146
MT3248
MT3249
MT3250

IP virulence (LD50 )a
<10

Oral virulence (LD50 )a

105

>104
103
<102
<102
<102
<102

1010
109 1010
109
109
109
1010

dam232 mgtC
dam232 mgtC 2-APr isolate #1
dam232 mgtC 2-APr isolate #2
dam232 mgtC 2-APr isolate #3

>104
103
104
104

1010
1010
1010
109 1010

MT3150
MT3251
MT3252
MT3253
MT3254
MT3255

dam232 sifA
dam232 sifA 2-APr
dam232 sifA 2-APr
dam232 sifA 2-APr
dam232 sifA 2-APr
dam232 sifA 2-APr

>104
104
104
104
104
104

1010
1010
109 1010
1010
109 1010
1010

MT3158
MT3256
MT3257
MT3258

dam232 spvB
dam232 spvB 2-APr isolate #1
dam232 spvB 2-APr isolate #2
dam232 spvB 2-APr isolate #3

>104
104
104
104

1010
1010
109 1010
109 1010

isolate #1
isolate #2
isolate #3
isolate #4
isolate #5

isolate #1
isolate #2
isolate #3
isolate #4
isolate #5

a
Independently isolated, in vivo selected, 2-aminopurine resistant (2-APr ) derivatives of Salmonella dam mutant vaccines strains were isolated from the spleens of infected
mice, and evaluated for oral and intraperitoneal (IP) virulence in nave mice [36]. The LD50 assay for each of these strains was compared to that of the wild type (UK-1). The
IP LD50 was determined by infecting ve mice per challenge dose; the peroral LD50 via gastrointubation was determined by infecting ten mice per challenge dose. The oral
and i.p. LD50 s for wild-type UK-1 are 105 and <10 CFU, respectively [12]. Surviving mice were scored >2 weeks post-infection.

D.M. Heithoff et al. / Vaccine 33 (2015) 100107

3.4. Vaccine and challenge strain shedding evaluation of

Salmonella dam double mutant vaccine strains
BALB/c mice were orally immunized with the Salmonella dam
double mutant vaccine candidates or the parental Salmonella dam
vaccine (109 CFU). Vaccine strain shedding: Fecal shedding of mice
immunized with Salmonella dam double deletion vaccine candidates vs. the parental Salmonella dam232 vaccine strain was
analyzed. All Salmonella dam double deletion vaccine candidates
exhibited signicantly reduced vaccine strain fecal shedding compared to that of the parental Salmonella dam vaccine, as both vaccine
and time following vaccination were signicant (Fig. 4a; *P < 0.05).
Challenge strain shedding: Fecal shedding of immunized mice challenged with virulent wild-type strain UK-1 was analyzed. Pairwise
comparisons revealed signicant differences between Salmonella
dam mgtC, dam sifA and dam spvB cohorts vs. the Salmonella
dam parent cohort at different times following virulent challenge (Fig. 4b). Both vaccine and time following vaccination were
signicant (P < 0.05) and there was a trend for a signicant interaction between time and vaccine (P = 0.075). These data indicate
that immunization with Salmonella dam double mutant vaccines
resulted in less vaccine and challenge strain shedding relative to
that of the dam parent vaccine.

CFU/g of feces

attenuated via i.p. infection, whereas four of ve isolates derived

from the parental dam vaccine were associated with reversion
to a more virulent state as demonstrated previously [36]. These
data indicate that Salmonella dam mgtC, dam sifA, and dam spvB
vaccine strains exhibited signicantly improved vaccine safety as
evidenced by the failure to give rise to virulent revertants during the
infective process, contrary to the parental Salmonella dam vaccine.

240

dam
dam mgtC
dam sifA
dam spvB

180

*
*

120
*
*

*
*

60
2

11

14

21

Day post-vaccination

B 60000

CFU/g of feces

104

11

14

21

6000

600

60
2

Day post-challenge

3.5. Environmental persistence evaluation (deionized water and

sheep feces) of Salmonella dam double mutant vaccine strains
Salmonella dam double mutant vaccine candidates were evaluated for environmental persistence in deionized water and in sheep
feces. Deionized water: Deionized water was inoculated with Knr
derivatives of Salmonella dam double mutant vaccine candidates,
parental dam UK-1 vaccine, or the wild-type parent strain, UK-1
(104 CFU/ml). The number of CFU/ml present in water over time
showed a signicant interaction between vaccine group and time
(P < 0.001), and pairwise comparisons revealed signicant differences between groups at different times (Fig. 5a). Importantly, the
low-level vaccine persistence in water may be compatible with
trough water vaccine administration as has been shown with a
Salmonella dam vaccine strain [27,31]. Sheep feces: Twenty per cent
dry matter sheep feces was inoculated with Knr derivatives of
Salmonella dam double mutant vaccine candidates, parental dam
UK-1 vaccine, or the wild-type parent strain, UK-1 (104 CFU/ml).
The number of CFU/ml present in feces over time showed a signicant interaction between vaccine group and time (P < 0.001),
and pairwise comparisons revealed signicant differences between
groups at different times (Fig. 5b). All vaccine strains exhibited less
viability than the wild-type parent UK-1 for all time points except
for day 1 and 6. These data indicate that Salmonella vaccine candidates show reduced environmental persistence in both deionized
water and sheep feces in comparison to that of the wild type UK-1
strain.
4. Discussion
Despite good husbandry practices, salmonellosis continues to
be a signicant problem in intensive production systems that favor
fecal-oral transmission. Disease is principally caused by increased

Fig. 4. Vaccine and challenge strain shedding evaluation of Salmonella dam double mutant vaccine candidates. Kanamycin-resistant derivatives of S. Typhimurium
UK-1 dam232 double mutant vaccine candidates (dam mgtC [MT3183], dam sifA
[MT3184], dam spvB [MT3186]) and the parental UK-1 dam strain (MT3180) were
used to vaccinate BALB/c mice (1115 per cohort) by the oral route (109 CFU). (A)
Vaccine fecal shedding: Feces was collected from individual mice and plated for CFU/g
on kanamycin 50 g/ml LB plates on days 2, 4, 7, 11, 14, and 21 post-immunization.
Fecal shedding of the Salmonella dam double deletion vaccine candidates vs. the
parental Salmonella dam232 vaccine strain was analyzed using REML repeated
measures analysis. Both vaccine and time following vaccination were signicant
(*P < 0.05). The interaction between vaccine and time did not indicate that the
signicant difference in shedding between the dam double deletion vaccines and
the dam Salmonella parent was consistent over time. No signicant differences
in fecal shedding were observed between the different double deletion dam vaccines. Values given are the model predicted mean number of CFU/g in feces of mice
following vaccination. Limit of detection is 60 CFU. (B) Challenge strain shedding:
Eleven weeks post-immunization, vaccinated mice were challenged with a dose
of 100 LD50 of a kanamycin-resistant derivative of wild-type S. Typhimurium UK-1
(MT2315; 107 CFU). Feces was collected from individual mice and plated for CFU/g
on kanamycin 50 g/ml LB plates on days 2, 4, 7, 11, 14, and 21 post-immunization.
Fecal shedding of the wild-type challenge strain following challenge of vaccinated
mice was analyzed using REML repeated measures analysis. Both vaccine and time
following vaccination were signicant (P < 0.05) and there was a trend for a signicant interaction between time and vaccine (P = 0.075). Pairwise comparisons
revealed signicant differences between groups at different times following virulent challenge (P < 0.05): a = shedding of double deletion vaccines signicantly less
than dam; b = shedding of dam sifA and dam spvB less than dam and dam sifA shedding is signicantly less than dam mgtC and dam spvB; c = shedding of dam sifA and
dam spvB signicantly less than dam, and dam sifA shedding is signicantly less than
dam mgtC and dam spvB, and dam spvB shedding is signicantly less than dam mgtC;
d = shedding of dam sifA, dam spvB and dam mgtC is signicantly less than dam; dam
sifA shedding is signicantly less than dam spvB; and dam mgtC shedding is signicantly less than dam spvB. Values given are model-based predicted mean CFU/g of
wild type challenge strain in feces following challenge. The limit of detection for the
culture technique is 60 CFU.

D.M. Heithoff et al. / Vaccine 33 (2015) 100107

60000
UK 1
dam
dam mgtC
dam sifA
dam spvB

CFU/ml

6000
a
b

a
b

a
c

a
b

a
c

600

60
0

10

12

14

Days in deionized water

CFU/ml

a
e

b
f

c
d

a
d
e

a
e
f

a
f

Days in 20% dry matter sheep feces

Fig. 5. Environmental vaccine persistence evaluation (deionized water and sheep
feces) of Salmonella dam double mutant vaccine candidates. Kanamycin-resistant
derivatives of S. Typhimurium UK-1 dam232 double mutant vaccine candidates
(dam mgtC [MT3183], dam sifA [MT3184], dam spvB [MT3186]), parental dam UK-1
vaccine (MT3180), and wild-type parent strain, UK-1 (MT2315) were used to inoculate (A), 20 ml of deionized water or (B), 20% fecal dry matter (104 CFU/ml). Triplicate
assays were performed in 50 ml conical tubes with loose caps at room temperature. Samples were vortexed and plated for CFU/ml for a 2-week period at the
time points indicated. Values given are the average CFU/ml with error bars indicating standard error of the mean (SEM). Deionized water: The number of CFU/ml
present in water over time was analyzed using REML repeated measures analysis.
A signicant interaction between vaccine group and time was observed (P < 0.001).
Pairwise comparisons revealed signicant differences between groups at different
times (P < 0.05). a = All vaccines have lower CFU/ml than the parent UK-1 wild type;
b = dam is signicantly less than dam mgtC, dam sifA, and dam spvB; c = dam and dam
spvB are signicantly less than dam sifA and dam mgtC. 20% fecal dry matter: Twenty
per cent fecal dry matter was generated by adding 20 ml of deionized water to 5 g of
dried sheep feces (gift from Barbara Byrne, University of California, Davis; [54,55]).
The number of CFU/ml present in feces over time was analyzed using REML repeated
measures analysis. A signicant interaction between vaccine group and time was
observed (P < 0.001). Pairwise comparisons revealed signicant differences between
groups at different times (P < 0.05). All vaccine strains were less than the parent UK1 wild type for all time points except for day 1 and 6. a = dam sifA signicantly less
than dam, dam mgtC, and dam spvB; b = dam sifA signicantly less than dam and dam
mgtC; c = dam sifA signicantly less than mgtC; d = dam spvB signicantly less than
dam mgtC; e = dam spvB signicantly less than dam; f = dam mgtC signicantly less
than dam.

pathogen exposure and disease susceptibility. Fluctuations in environmental conditions cause shifts in the environmental pathogen
load and subsequently host challenge. Physiological changes associated with pregnancy and parturition increase susceptibility to
disease as does the nave immune status of neonates. Management practices may also negatively impact host immunity with
the cumulative stressors experienced by stock on farm (mustering, yarding, food and water deprivation prior to transport), during

105

transport (food and water deprivation, environmental stress), and

in sale yards (co-mingling, pathogen exposure). The development
of an efcacious Salmonella vaccine for livestock is needed to
promote animal health and welfare and human food safety. The
commercial viability of vaccination strategies in livestock production systems is largely inuenced by the cost of disease as well
as vaccine safety, cost and efcacy. If an affordable and effective
Salmonella vaccine is made available to the commercial sector, the
vaccine could be applied broadly across animal production industries as vaccination is simple, understood by producers, and likely
to be adopted. As a potential means to address this issue, modied
live Salmonella dam vaccines have been shown to be effective and
well-tolerated in several species of stock. However, the principal
concerns of live vaccines are safety, shedding, and environmental persistence. Herein, secondary virulence-attenuating mutations
were introduced into a Salmonella dam vaccine strain to improve
vaccine safety. A subset of Salmonella dam vaccines containing an
additional attenuating mutation (dam mgtC, dam sifA, and dam
spvB) exhibited signicantly improved vaccine safety, attenuated
replication in the environment, and the capacity to elicit crossprotective immunity against pathogenic salmonellae serotypes
derived from infected livestock. These data indicate that Salmonella
dam double mutant vaccines exhibit a favorable therapeutic index
(safety/efcacy) for commercial applications in intensive livestock
production systems.
There are a number of promising features of modied live
Salmonella dam vaccines that suggest they are likely to provide an
effective means of salmonellae prophylaxis in livestock industries.
It has been established that Salmonella dam vaccines: (1) confer
robust cross-protection against pathogenic serotypes that are commonly associated with disease in livestock and humans [13,17];
(2) are well-tolerated by a diversity of species including mice
[17,24], poultry [25,26], sheep [27] and calves [2830]; (3) confer
rapid immunity, a requisite for neonates and the high-risk period
following feedlot entry [26,28]; (4) can be administered in drinking water, providing a low-cost, low-stress protocol for livestock
immunization [27,31]. Herein, the safety of Salmonella dam vaccine strains containing additional attenuating mutations targeted
to genes involved in intracellular and/or systemic survival were
evaluated in vaccinated animals and in conditions mimicking the
environment. Salmonella dam mgtC, dam sifA, and dam spvB vaccine
strains exhibited improved vaccine safety (vaccine shedding; challenge strain shedding; persistence in systemic tissues; persistence
in the environment), while maintaining robust efcacy against virulent challenge with homologous and heterologous pathogenic
serotypes. These data are consistent with the hypothesis that introduction of certain virulence-attenuating mutations result in an
optimal balance of vaccine safety combined with the maintenance
of a low-grade persistence of the vaccine strain, which may provide
a stable source of antigens over the time needed to transition to the
development of strong adaptive immune responses [17,24,34].
The mechanistic basis of Salmonella dam vaccine crossprotection is likely associated with the delivery of an expanded
repertoire of antigens over the time needed to transition to the
development of cross-protective immune responses [17,24,34].
This is accomplished by a multi-factorial process including an
increased production/release of a number of antigens [24,38]; a
reduction in the ability of the bacteria to invade non-phagocytic
cell types [39,40]; minimal cytotoxicity to M cells of the Peyers
patch [39]; and increased stimulation of B and T cell responses
congruent with a lack of vaccine-mediated immune suppression
[17,24]. This is in contrast to other Salmonella live vaccines (e.g.
aroA [4144]) that confer a transient state of immunosuppression attributed to nitric oxide (NO) production by myeloid derived
suppressor cells (MDSCs) that may restrict protective immunity
only to closely-related strains [4547]. Thus, pathways involved in

106

D.M. Heithoff et al. / Vaccine 33 (2015) 100107

antigen processing and presentation by antigen presenting cells

may be uniquely stimulated in animals infected with Salmonella
dam vaccines [17,48,49].
From a farm-management perspective, it is sensible to dene
the management and environmental events that lead to increased
pathogen transmission in the context of the food production
chain so appropriate risk-management strategies can be implemented to prevent disease. It has been established in livestock that
host susceptibility and shedding are inuenced by management
and environmental events (herd size, adverse weather conditions,
equipment failure, labor issues, surface water management) that
contribute to compromised host immunity and increased pathogen
exposure [1,2,8,9,14,15]. Additionally, changes in environmental
conditions inherent to livestock production facilities (temperature and moisture content of the fecal pack), dietary management
(macro- and micro-nutrient composition), and/or exposure to
antimicrobials may also favor pathogen proliferation, pathogen
virulence, and/or host susceptibility [12]. These management and
environmental events in combination with Salmonellas high tolerance to environmental stress, adaptability, multi-drug resistance,
and widespread distribution and diversity in animal and environmental reservoirs present an ongoing and formidable challenge to
animal health and food safety.
Although this information has led to a more complete understanding of the risk factors in the food production chain, effective
and reliable strategies for salmonellae control have not been established as evidenced by the continuing high pathogen prevalence
in livestock populations [50] and recent abundance of Salmonella
outbreaks in the U.S. and worldwide [7]. A thorough review of food
safety practices from farm to fork and has led to the following recommendations [16,51,52]. (1) Improved surveillance/traceability
for source product identication; (2) improved monitoring by the
FDA and USDA; (3) irradiation of post-harvest foods; (4) suspension
of growth-promoting antibiotics in animal feeds; (5) implementation of more effective livestock immunization programs. Together,
these practices formulate a comprehensive food safety plan that
will likely limit the size, frequency, and severity of foodborne
outbreaks, and reduce the risk of false implication of other food
products that can devastate a given industry.
Acknowledgement
This work was supported by Meat and Livestock Australia
(W.LIV.0168) (to M.J.M.).
References
[1] House JK, Ontiveros MM, Blackmer NM, Dueger EL, Fitchhorn JB, McArthur
GR, et al. Evaluation of an autogenous Salmonella bacterin and a modied live
Salmonella serotype Choleraesuis vaccine on a commercial dairy farm. Am J Vet
Res 2001;62:8971902.
[2] Anderson R, House J, Smith B, Kinde H, Walker R, Steeg B, et al. Epidemiologic
and biological characteristics of salmonellosis in three dairy herds. J Am Vet
Med Assoc 2001;219:31022.
[3] National Veterinary Services Laboratories. Salmonella and Arizona serotypes
from animals and related sources for the scal year 19941995.
[4] Mohler V, Izzo M, House J. Salmonella in calves. Vet Clin North Am Food Anim
Pract 2009;25:3754.
[5] Centers for Disease Control and Prevention. National Salmonella Surveillance
Annual Summary, 2008. Atlanta, Georgia: US Department of Health and Human
Services; 2011. p. 15.
[6] Centers for Disease Control and Prevention. National Salmonella Surveillance
Overview. Atlanta, Georgia: US Department of Health and Human Services;
2011. p. 112.
[7] Centers for Disease Control and Prevention. Reports of selected Salmonella
outbreak investigations. Atlanta, Georgia: CDC; 2014. Available at:
http://www.cdc.gov/salmonella/outbreaks.html
[8] Fossler C, Wells S, Kaneene J, Ruegg P, Warnick L, Bender J, et al. Herd-level
factors associated with isolation of Salmonella in a multi-state study of conventional and organic farms I. Salmonella shedding in cows. Prevent Vet Med
2005;70:25777.

[9] Fossler C, Wells S, Kaneene J, Ruegg P, Warnick L, Bender J, et al. Herd-level

factors associated with isolation of Salmonella in a multi-state study of conventional and organic farms II. Salmonella shedding in calves. Prevent Vet Med
2005;70:27991.
[10] Ruzante J, Lombard J, Wagner B, Fossler C, Karns J, Van Kessel J, et al.
Factors associated with Salmonella presence in environmental samples
and bulk tank milk from US dairies. Zoonoses and Public Health; 2010,
http://dx.doi.org/10.1111/j.1863-2378.2010.01333.x.
[11] Smith B, Da Roden L, Thurmond M, Dilling G, Konrad H, Pelton J, et al. Prevalence
of salmonellae in cattle and in the environment on California dairies. J Am Vet
Med Assoc 1994;205:46771.
[12] Heithoff DM, Shimp WR, House JK, Xie Y, Weimer BC, Sinsheimer RL, et al.
Intraspecies variation in the emergence of hyperinfectious bacterial strains in
nature. PLoS Pathogens 2012;8:e1002647.
[13] Heithoff DM, Shimp WR, Lau PW, Badie G, Enioutina EY, Daynes RA, et al. Human
Salmonella clinical isolates distinct from those of animal origin. Appl Environ
Microbiol 2008;74:175766.
[14] Fossler CP, Wells SJ, Kaneene JB, Ruegg PL, Warnick LD, Eberly LE, et al. Cattle and
environmental sample-level factors associated with the presence of Salmonella
in a multi-state study of conventional and organic dairy farms. Prev Vet Med
2005;67:3953.
[15] Huston CL, Wittum TE, Love BC, Keen JE. Prevalence of fecal shedding of
Salmonella spp in dairy herds. J Am Vet Med Assoc 2002;220:6459.
[16] Mahan MJ, Heithoff DM, House JK. Salmonella cross-protective vaccines: fastforward to the next generation of food safety. Future Microbiol 2012;7:
8058.
[17] Heithoff DM, Enioutina EY, Bareyan D, Daynes RA, Mahan MJ. Conditions that
diminish myeloid-derived suppressor cell activities stimulate cross-protective
immunity. Infect Immun 2008;76:51919.
[18] Hegazy WAH, Hensel M. Salmonella enterica as a vaccine carrier. Future Microbiol 2012;7:11127.
[19] Kong Q, Yang J, Liu Q, Alamuri P, Roland KL, Curtiss III R. Effect of deletion of
genes involved in lipopolysaccharide core and O-antigen synthesis on virulence and immunogenicity of Salmonella enterica serovar Typhimurium. Infect
Immun 2011;79:422739.
[20] Li Y, Wang S, Scarpellini G, Gunn B, Xin W, Wanda SY, et al. Evaluation of new
generation Salmonella enterica serovar Typhimurium vaccines with regulated
delayed attenuation to induce immune responses against PspA. Proc Natl Acad
Sci U S A 2009;106:5938.
[21] Nagy G, Palkovics T, Otto A, Kusch H, Kocsis B, Dobrindt U, et al. Gently rough:
the vaccine potential of a Salmonella enterica regulatory lipopolysaccharide
mutant. J Infect Dis 2008;198:1699706.
[22] Curtiss R, Xin W, Yuhua L, Kong W, Wanda SY, Gunn B, et al. New technologies
in using recombinant attenuated Salmonella vaccine vectors. Crit Rev Immunol
2010;30:25570.
[23] Singh B. Salmonella vaccines for animals and birds and their future perspective.
Open Vaccine J 2009;2:10012.
[24] Heithoff DM, Enioutina EY, Daynes RA, Sinsheimer RL, Low DA, Mahan MJ.
Salmonella DNA adenine methylase mutants confer cross-protective immunity. Infect Immun 2001;69:672530.
[25] Dueger EL, House JK, Heithoff DM, Mahan MJ. Salmonella DNA adenine methylase mutants elicit protective immune responses to homologous and heterologous serovars in chickens. Infect Immun 2001;69:
79504.
[26] Dueger EL, House JK, Heithoff DM, Mahan MJ. Salmonella DNA adenine methylase mutants prevent colonization of newly hatched chickens by homologous
and heterologous serovars. Int J Food Microbiol 2003;80:1539.
[27] Mohler VL, Heithoff DM, Mahan MJ, Walker KH, Hornitzky MA, Gabor L,
et al. Protective immunity conferred by a DNA adenine methylase decient
Salmonella enterica serovar Typhimurium vaccine when delivered in-water
to sheep challenged with Salmonella enterica serovar Typhimurium. Vaccine
2011;29:357182.
[28] Dueger EL, House JK, Heithoff DM, Mahan MJ. Salmonella DNA adenine methylase mutants elicit early and late onset protective immune responses in calves.
Vaccine 2003;21:324958.
[29] Mohler V, Heithoff D, Mahan M, Walker K, Hornitzky M, McConnell C, et al.
Cross-protective immunity in calves conferred by a DNA adenine methylase decient Salmonella enterica serovar Typhimurium vaccine. Vaccine
2006;24:1339.
[30] Mohler V, Heithoff D, Mahan M, Walker K, Hornitzky M, Shum L, et al.
Cross-protective immunity conferred by a DNA adenine methylase decient
Salmonella enterica serovar Typhimurium vaccine in calves challenged with
Salmonella serovar Newport. Vaccine 2008;26:17518.
[31] Mohler V, Heithoff D, Mahan M, Hornitzky M, Thomson P, House J. Development
of a novel in-water vaccination protocol for DNA adenine methylase decient Salmonella enterica serovar Typhimurium vaccine in adult sheep. Vaccine
2012;30:148191.
[32] Hassan JO, Curtiss 3rd R. Development and evaluation of an experimental vaccination program using a live avirulent Salmonella typhimurium strain to protect
immunized chickens against challenge with homologous and heterologous
Salmonella serotypes. Infect Immun 1994;62:551927.
[33] Davis RW, Botstein D, Roth JR. Advanced bacterial genetics. Plainview, N.Y.:
Cold Spring Harbor Laboratory Press; 1980.
[34] Heithoff DM, Sinsheimer RL, Low DA, Mahan MJ. An essential role for DNA
adenine methylation in bacterial virulence. Science 1999;284:96770 [see
comments].

D.M. Heithoff et al. / Vaccine 33 (2015) 100107

[35] Donnenberg MS, Kaper JB. Construction of an eae deletion mutant of
enteropathogenic Escherichia coli by using a positive-selection suicide vector.
Infect Immun 1991;59:43107.
[36] Heithoff D, Badie G, Julio S, Enioutina E, Daynes R, Sinsheimer R, et al. In vivoselected mutations in methyl-directed mismatch repair suppress the virulence
attenuation of Salmonella dam mutant strains following intraperitoneal, but
not oral, infection of nave mice. J Bacteriol 2007;189:470817.
[37] Glickman B, van den Elsen P, Radman M. Induced mutagenesis in dam- mutants
of Escherichia coli: a role for 6-methyladenine residues in mutation avoidance.
Mol Gen Genet 1978;163:30712.
[38] Pucciarelli M, Prieto A, Casadess J, Garca-del Portillo F. Envelope instability in DNA adenine methylase mutants of Salmonella enterica. Microbiology
2002;148:117182.
[39] Garca-Del Portillo F, Pucciarelli MG, Casadess J. DNA adenine methylase
mutants of Salmonella typhimurium show defects in protein secretion, cell invasion, and M cell cytotoxicity. Proc Natl Acad Sci U S A 1999;96:1157883.
[40] Lpez-Garrido J, Casadess J. Regulation of Salmonella enterica pathogenicity
island 1 by DNA adenine methylation. Genetics 2010;184:63749.
[41] al-Ramadi B, Brodkin M, Mosser D, Eisenstein T. Immunosuppression induced
by attenuated Salmonella. Evidence for mediation by macrophage precursors.
J Immunol 1991;146:273746.
[42] Al-Ramadi B, Chen Y-W, Meissler J, Eisenstein T. Immunosuppression induced
by attenuated Salmonella. Reversal by IL-4. J Immunol 1991;147:195461.
[43] Al-Ramadi B, Meissler J, Huang D, Eisenstein T. Immunosuppression
induced by nitric oxide and its inhibition by interleukin-4. Eur J Immunol
1992;22:224954.
[44] Eisenstein T, Huang D, Meissler Jr J, Al-Ramadi B. Macrophage nitric oxide
mediates immunosuppression in infectious inammation. Immunobiology
1994;191:493502.
[45] Harrison JA, Villarreal-Ramos B, Mastroeni P, Demarco de Hormaeche R,
Hormaeche CE. Correlates of protection induced by live Aro- Salmonella
typhimurium vaccines in the murine typhoid model. Immunology
1997;90:61825.

107

[46] Hormaeche CE, Joysey HS, Desilva L, Izhar M, Stocker BA. Immunity
conferred by Aro- Salmonella live vaccines. Microb Pathog 1991;10:
14958.
[47] Hormaeche CE, Mastroeni P, Harrison JA, Demarco de Hormaeche R, Svenson
S, Stocker BA. Protection against oral challenge three months after i.v. immunization of BALB/c mice with live Aro Salmonella typhimurium and Salmonella
enteritidis vaccines is serotype (species)-dependent and only partially determined by the main LPS O antigen. Vaccine 1996;14:2519.
[48] Shtrichman R, Heithoff D, Mahan M, Samuel C. Tissue selectivity of
interferon-stimulated gene expression in mice infected with Dam+ versus
Dam- Salmonella enterica serovar Typhimurium strains. Infect Immun
2002;70:557988.
[49] Simon R, Heithoff D, Mahan M, Samuel C. Comparison of tissue-selective proinammatory gene induction in mice infected with wild-type, DNA adenine
methylase-decient, and agellin-decient Salmonella enterica. Infect Immun
2007;75:562739.
[50] Haneklaus A, Harris K, Grifn D, Edrington T, Lucia L, Savell J. Salmonella
prevalence in bovine lymph nodes differs among feedyards. J Food Prot
2012;75:11313.
[51] Maki DG. Coming to grips with foodborne infectionpeanut butter, peppers, and nationwide Salmonella outbreaks. New Engl Med 2009;360:
94953.
[52] U.S. Food and Drug Administration FDAs strategy on antimicrobial resistance
questions and answers. Silver Spring, MD: U.S. Department of Health and
Human Services; 2012.
[53] International V. GenStat for Windows. 15th ed. Hemel Hempstead, UK: VSN
International; 2012.
[54] Griggs T. Determining forage dry matter concentration with a microwave oven
AG/Forage & Pasture/2005-01; 2005.
[55] Pitt R, Brusewitz G, Chase L, Collins M. Forage moisture determination. NRAES
(59)(Cooperative Extension); 1993.