J. Med. Microbiol. - Vol. 48 (1999), 907-916
0 1999 The Pathological Society of Great Britain and Ireland
ISSN 0022-26 15
BACTE R I AL PATHOG EN I CITY
Intestinal colonisation of gnotobiotic pigs by
Salmonella organisms: interaction between
isogenic and unrelated strains
M.A. LOVELL and P.A. B A R R O W
Division of Environmental Microbiology, Institute for Animal Health, Compton Laboratory, Newbury, Berkshire
RG20 7NN
The effect of intestinal colonisation by a Salmonella strain on the establishment in the
gut of an isogenic mutant administered orally 24 h after the first strain was studied in
gnotobiotic pigs. Irrespective of the clinical outcome of the infection, the extensive
colonisation of one Salmonella strain prevented a similar degree of colonisation by an
otherwise isogenic antibiotic resistant strain; in some cases the second strain was hardly
detectable. The poor colonisation of the challenge Salmonella strains was generally
reflected in very low counts of organisms in the tissues. Colonisation by a strain of
Escherichia coli reduced the rate of establishment of an isogenic E. coli, strain but did
not prevent colonisation by an S. Typhimurium strain. S. Typhimurium with mutations
in the tsr (serine chemotaxis receptor protein) or oxrA (transcriptional regulator of
anaerobic metabolism) genes did not inhibit colonisation. Mutations in cya (adenylate
cyclase), tar and trg (chemotaxis receptor proteins for aspartate and ribose respectively)
genes were less inhibitory, while motB (non-motile) and cheR (impaired motility)
mutants were fully inhibitory.
Introduction
Although poultry are currently regarded as the main
source of salmonella infections for man in the developed world, pigs have also traditionally been regarded
as a major reservoir of zoonotic infection [ 1,2].
As with poultry, salmonellosis in pigs is of both animal
and public health significance. Most disease in pigs
caused by Salmonella occurs in the post-weaning
period although an uncommon severe systemic infection in neonatal pigs has been described [3]. The
relatively high frequency of infection but low frequency of disease in sucking pigs is attributed to the
protective effects of maternal antibody. In North
America and in many developing countries, Salmonella
serotype Choleraesuis is the most frequently isolated
serotype [4] and is responsible for a septicaemia which
is probably similar to typhoid in man [3, 51. However,
S. Typhimurium is probably responsible for most
zoonotic infections from pigs. Other serotypes such
as S. Derby, S. Bredeney, S. Brandenburg and S.
Received 18 June 1998; revised version received 18 Dec.
1998; accepted 8 Feb. 1999.
Corresponding author: Dr P. A. Barrow (e-mail: PauLBarrow
@BBSRC .AC .UK).
Bovismorbificans, which produce largely asymptomatic
infections in swine [3], have also been implicated in
human disease.
Because of the complex epidemiology of salmonella
infections, control must, of necessity, be multi-factorial
[4, 6-81. With increasing problems associated with
extensive antibiotic use and abuse, and the high cost of
upgrading housing to reduce environmental problems,
it is likely that vaccination with live, attenuated
vaccines will be increasingly used both in poultry
[9, 101 and in pigs [ 11, 121. In chickens, live vaccines
induce immunity [13-151 and, if given orally to newlyhatched chickens, confer, within a few hours of
administration, a high level of resistance to intestinal
colonisation by subsequently inoculated field strains
[16, 171. This exclusion caused by the extensive
colonisation of the alimentary tract by the salmonella
vaccine occurs after oral infection of young animals
[ 17-20]. This resistance to colonisation is neither the
result of a rapid host immune response, nor the result
of bacteriophage or bacteriocin activity [16]. It is
thought to be similar to the genus-specific down
regulation of bacterial growth that occurs in early
stationary phase nutrient broth cultures of Salmonella
and other enteric bacteria [21]. Therefore, live vaccines
administered orally to newly hatched chickens could
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 12:36:26
908
M. A. LOVELL AND P. A. BARROW
confer resistance to infection within a few hours by this
mechanism and would also continue to colonise
sufficiently to stimulate immunity.
Because of the inherent value of such an approach to
the use of live vaccines against salmonellosis in poultry
and the desirability of obtaining early protection in
pigs, and even more so in calves, experiments were
carried out to determine whether similar specific
interactions occurred between SaZmoneZZa organisms
in the intestines of young animals reared normally on a
whole milk diet. Such a study could also contribute to
understanding of the mechanisms of intestinal colonisation of young animals by Salmonella organisms by
characterising the interaction between related and
unrelated bacteria when present in the gut at high
density. Gnotobiotic animals were used in these initial
experiments to avoid any variability in the extent of
colonisation or in other factors that might occur
because of different levels of acquired maternal
antibody. Pigs were used rather than calves for logistic
reasons. Because of the high cost of these experiments,
relatively small group sizes were used in some cases;
however, the results of quantitative bacteriological
analysis were remarkably consistent between individual
animals.
Bacterial strains
The bacterial strains used in this study are listed in
Table 1 [13, 16, 18,22-301. Mutants of Thax-1 and
F98 were produced with mutations in the oxrA or cya
and cheR and motB genes. These mutations were
transduced into the required strain by bacteriophage
P22 [13] by standard procedures. The cya phenotype
was checked for very poor growth on Luria-Bertani
(LB) agar. The oxrA phenotype was checked for its
Table 1. Bacterial strains
Phenotype/
Genotype
S. Typhimurium F98
S. Typhimurium Thax- 1
Wild-type
Non-inducer of fluid
secretion
S.
S.
5'.
S.
motB::TnlO
cheR::TnlO
oxrA ::TnlO
cya::Tnl 0
Typhimurium
Typhimurium
Typhimurium
Typhimurium
BJ32
BJ14
TN2336
P3395
S. Typhimurium F98 motB
S. Typhimurium F98 cheR
S. Typhimurium Thax-1 oxrA
S. Typhimurium Thax Acya
S.
S.
S.
S.
Typhimurium
Typhimurium
Typhimurium
Typhimurium
ST1
ST330
ST328
ST832
In all cases, mutants of the above strains, resistant to
nalidixic acid (Nal') or spectinomycin (Spc') were
used. These were selected by standard methods [22].
Neither of these mutations affected the in-vivo colonisation or colonisation inhibition in young chickens
[16, 171. In all cases, Spc' was used for challenge
strains.
Unless stated otherwise, all strains were cultured for
24 h in 10 ml LB broth (Difco) and generally reached
counts of (3-5) X lo9 cfu/ml. This was also the case
for the cya mutant.
Materials and methods
Strain
inability to grow anaerobically with glycerol as sole
carbon source and with fumarate as electron acceptor
and by the absence of stationary phase growth
inhibition in vitro using a standard method [21].
Previous work (Lovell, Rychlik and Barrow, unpublished observations) had found that mutations in oxrA
abolished the growth inhibition phenotype. Mutations
in cya were unstable through loss of TnlO and,
therefore, a deletion mutant was produced by positive
selection for tetracycline sensitivity [3 11 selecting for
the same small colony phenotype. This mutation also
abolishes stationary phase growth inhibition (Lovell,
Rychlik and Barrow, unpublished observations). However, the TcS cya phenotype also proved to be unstable,
although relatively much less so, but presumably by the
occurrence of secondary, suppressor mutations. The
motB transcipient was non-motile and the cheR mutant
showed impaired motility through soft agar [32].
motB::TnlO
cheR::TnlO
oxrA ::TnI 0
A cya(Tc')
Ref.
16,18,22
23,24
25,26
25,26
27
28
This
This
This
This
paper
paper
paper
paper
Wild-type
ST1 tsr
STl tar
STl trg
29
29
29
29
Escherichia coli P4
Wild-type commensal
30
Bacteriophage P22
HT int
13
Experimental animals
Gnotobiotic Large-White pigs were obtained by hysterotomy and reared in metal mesh-floored cages in
positive-pressure isolators [33, 341. They were reared at
a temperature of 25"-28"C and were given increasing
amounts of equal volumes of sterilised condensed milk
and water mixed with a mineral supplement. They were
checked for freedom from bacterial contamination by
culture of a rectal swab at 4 days of age aerobically
and anaerobically on blood agar and were then used at
5-7 days.
Bacterial cultures in metal screw-capped bottles were
taken in and out of the isolators via a port sterilised
with a peracetic acid-HzO2 mixture for 30 min.
Infection and quantitative bacteriological analysis
Pigs were infected orally with 1-ml volumes of
bacterial culture by introducing a syringe into the back
of the mouth and allowing the animal to swallow the
inoculum. Where the effect of colonisation of one
strain on the establishment of another was to be
studied, the first strain was inoculated at a dose of
lo6 cfu and the second (challenge) strain was inoculated 1 day later at a dose of lo3 cfu. In some cases,
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 12:36:26
SALMONELLA COLONISATION IN PIGS
bacterial strains were also inoculated singly in which
case the dose was lo3 c h .
Faeces samples were collected daily from the anus of
each animal. In a few cases, milk was sampled from
the feed trough either during or immediately after
feeding. Clinical condition and temperature were
monitored closely. Animals were killed for postmortem analysis when they had reached specific
humane end-points (moderate dehydration, unwilling
to take further food over a 24-h period and disinclination to stand).
Post-mortem analysis was done within 15 min of death.
After cleaning the skin thoroughly with alcohol, the
abdomen was opened and tissue samples were taken in
the order cardiac blood, kidney, spleen and liver. The
small intestine was tied off with string in several places
to reduce movement of contents and samples were
removed in the order duodenum, jejunum, ileal lymph
node, ileum, caecum, colon and stomach. In most
experiments, contents only were sampled. However, in
one experiment intestinal mucosal samples were taken.
In this case a section of the intestine was removed and,
after the removal of the contents, the mucosa was
washed gently in running water to remove adherent
contents, dried gently and samples collected by
scraping with a scalpel blade [ 181.
All samples were diluted in phosphate buffered saline
(PBS) and homogenised by vigorous mixing or with
Griffith's tubes. The number of viable bacteria in the
samples was estimated by plating serial decimal
dilutions on LB agar containing spectinomycin
50 pg/ml or sodium nalidixate 20 pg/ml.
In-vitro inhibition assays
The ability of early stationary phase cultures of
different bacterial strains made in milk or LB broth
to inhibit the multiplication of other inoculated
organisms was assayed by standard procedures
[16,21]. In brief, 10-ml volumes of LB broth or milk,
identical to that given to the pigs, were inoculated with
lo6 cfu of the bacterial strain to be tested. This culture
was incubated with shaking at 37°C for 24 h. A second
strain, inoculated at lo3 cfu and with a distinguishing
antibiotic resistance marker was inoculated into this
culture and the mixture was re-incubated. At selected
times thereafter the numbers of both organisms were
counted on LB agar containing appropriate antibiotics.
Results
Experiments with wild-type Salmonella strains
and E. coli
Pigs inoculated orally with either S. Typhimurium F98
Nal' followed by F98 Spc' (animals 1-4) or F98 Spc'
alone (animals 5-9) became ill within 24 h of
909
infection. The faecal bacterial counts of the inoculated
strains were high at both times of sampling (Fig. la, b).
The counts of the challenge strain (F98 Spc') in the
pigs already infected with F98 Nal' were very low for
the one sampling time before the pigs had to be killed
because of the severity of the illness. The disease took
the form of a diarrhoea and vomiting accompanied in
the later stages by inappetence and lethargy.
When these animals were killed there was a fibrinous
layer over the spleen and some oedema in the colonic
submucosa and between the coils of the colon. Pigs 59, inoculated with only lo3 cfh of F98 Spc', had counts
in excess of lo5 cfu/g of liver and spleen and evidence
of bacteraemia with Salmonella present in the blood
and kidneys (Table 2). High counts of this strain were
present throughout the gut. In pigs 1-4, inoculated with
F98 Nal' followed by F98 Spc', similarly high counts of
F98 Nal' were found in the tissues and gut samples.
However, very few organisms of F98 Spcr were isolated
from the tissues of these pigs and the counts in the gut
were much lower than those of the first strain.
Oral inoculation with E. coli P4 resulted in faecal
counts of this strain in excess of lo1' cfu/g and
occasionally 10" cfu/g whether this was the Nal'
mutant (Fig. lc, d) or the Spc' mutant (Fig. le). These
high counts had little effect on the colonisation of F98
Spc' given 1 day later (Fig. lc). These pigs were
healthy at the time they were killed, although the viable
numbers of the S. Typhimurium strain in the intestine
were high. Few E. coli organisms were found in the
non-intestinal tissues with small numbers of F98 Spc'
only from the liver and spleen (Table 2, animals 10 and
11). Very high counts of E. coli were present
throughout the gut. The counts of F98 Spc' were as
high in the terminal intestine as in the pigs given this
strain only. They were somewhat lower in the stomach,
duodenum and jejunum.
In contrast, pre-colonisation with E. coli P4 retarded
the establishment of P4 Spc' (Fig. Id) compared with
the rate of colonisation of P4 Spc' in pigs given this
strain alone (Fig. le). However, numbers of P4 Spc' in
the faeces of the two pigs of between lo4 and
lo7 cfu/g were reached (Fig. Id). The numbers of the
challenge strain stopped increasing and stabilised 3
days after challenge. When these animals were killed
the numbers of the first strain in the gut were much
higher than those of the challenge strain. Small
numbers of the first strain were isolated from the liver
and spleen. The second strain was not isolated from the
tissues although the cut-off in sensitivity was lo2 cfu/g
(Table 2, animals 12 to 15).
Experiments with S. Typhimurium strains Thax-1
and STl and mutants of these strains
Because strain F98 was highly virulent in this model,
strains with a reduced capacity to produce gastro-
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 12:36:26
910
M. A. LOVELL AND P. A. BARROW
12r a
10 8-
642012r
0
4
2
6
Time (days)
10'
8,
6'
4,
20,
Time (days)
Fig. 1. Inhibition of intestinal colonisation by isogenic and non-isogenic strains of Salmonella or E. coli in gnotobiotic
pigs assessed by counting numbers of viable bacteria in faeces (a) S. Typhimurium F98 Nal' (o), S. Typhimurium F98
Spc' (0);mean and SEM from four pigs. (b) S. Typhimurium F98 Spc' (0); mean and SEM from five pigs. (c) E. coli
P4 Nal' (o), S. Typhimurium F98 Spc' (0);individual values from two pigs. (a) E. coli P4 Nal' (o), E. coli P4 Spc'
(0);
individual values from two pigs. ( e ) E. coli Spc' (0).V indicates day on which pigs were killed because of
disease; dotted line indicates limit of sensitivity of bacterial counting method (loglo 2/g of gut contents).
enteritis infections in monkeys [24] were used to study
the effects of different mutations on the interaction
between bacteria in the gut. All pigs infected with
these strains remained completely healthy. When given
alone at a dose of lo3 cfb, the faecal counts of Spc'
mutant of Thax-1 were very high throughout the
duration of the experiment, whereas in the presence
of the Nal' mutant colonisation by the Spc' mutant was
inhibited (Fig. 2a, b). The Spc' mutant of STl was not
tested on its own because of problems of virulence, but
had to be used because the mutations of interest in it
were not linked to an antibiotic resistance marker and
could not be transduced to an avirulent strain. In this
case, the pigs were infected 24 h before with E. coli P4
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 12:36:26
SALMONELLA COLONISATION IN PIGS
m
-3
-0
9
2
%
a
.3
m
c
a
5
2
.-8
a
-0
9
2
%
.-a
91 1
whose establishment did not prevent intestinal colonisation by S. Typhimurium ST1, but which reduced its
virulence for the host.
The mutants of Thax-1 and STl tested showed varying
degrees of colonisation-inhibition for the parental
strains. In these experiments only two pigs were used
to test each combination of strains, but the pattern of
excretion of the strains was consistent in both animals
used. In the pigs colonised by either the oxrA::TnlO or
cya::TnlO mutant of Thax-1 Nal' the numbers of the
Spc' mutant of the parent challenge strain in the faeces
increased during the course of the experiment (Fig. 2c,
d). In the pigs colonised by ST1 tsr Nal', the Spc'
mutant of the parent challenge strain increased rapidly
and eventually equalled the faecal counts of the mutant
(Fig. 3b). The tar mutant in both pigs and the trg in
one animal were less inhibitory than the Nal' mutant of
the parent (Fig. 3c, d).
A further experiment was performed in which two
groups of two pigs each were infected with E. coli P4
followed 24 h later (as above) by mutants of S.
Typhimurium F98 Nal' which were cheR::TnlO or
motB::TnlO and were then challenged 24 h later with
F98 Spc'. The cheR and motB mutants were fully
inhibitory for the parental F98 strain (results not
shown).
Analysis of contents and mucosal counts of pigs
infected with inhibitory and non-inhibitory strains
m
m
I
8
.e
a
-
d
v)
.-aM
A comparison of the bacterial counts in the contents
and mucosa of different parts of the gut in pigs infected
with non-inhibitory mutants (oxrA and cya) of the
Thax-1 strain showed that mucosal counts were always
lower than the contents counts (Table 3). Neither the
contents nor the mucosal counts of the less-inhibitory
mutants were lower than those of the inhibitory parent
strain.
The viable counts of the Thax-1 Nal' and Thax-1 oxrA
Nal' mutant in milk samples taken from the troughs
were similar to those of the stomach samples from
these pigs. The counts of the Spc' challenge strains
were low in these samples.
In-vitro stationary phase growth inhibition assays
in milk
The ability of Spc' mutants of S. Typhimurium strains
F98, Thax-1, ST1 and E. coli P4 to multiply in 24-h
cultures of Nal' mutants grown in milk or broth was
tested. The Nal' mutants were either of the parental
phenotype or were cheR, motB, trg, tar, tsr, oxrA or
cya. The full results are not shown but representative
results for interactions between derivatives of Thax-1 in
milk are shown in Fig. 4. Multiplication was monitored
for three days after inoculation of the Spc' challenge
strain. The viable counts of the first strain in the assay
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 12:36:26
M. A. LOVELL AND P. A. BARROW
912
b
a
5
d
10
6
4
2
0
0
2
4
Time (days)
6
a
Time (days)
Fig. 2. Inhibition of intestinal colonisation between isogenic mutants of S. Typhimurium strain Thax- 1 using faecal
bacterial counts. (a) S. Typhimurium Thax-1 Nal' (o), S. Typhimurium Thax-1 Spc' (0);mean and SEM from four
pigs. (b) S. Typhimurium Thax-1 Spc' (0);mean and SEM from four pigs. (c) S. Typhimurium Thax-1 oxrA::TnlO Nal'
(o), S. Typhimurium Thax-1 Spc' (0);individual values from two pigs. (d) S. Typhimurium Thax-1 A cya Nal' (o), S.
Typhimurium Thax-1 Spc' (0);values for two pigs. For other details see legend to Fig. 1.
(the Nal' culture into which the challenge Spc' strain
was inoculated) remained high during the course of the
experiment in both milk (between 8.0 and 8.3 cfu/ml
for Thax-1 and between 8.8 and 9.5 cfu/ml for STl,
F98 and E. coli P4) and broth (between 8.3 and
9.7 c h / m l for all strains) and the rate of increase in
the viable numbers of these single strains was high
(Fig. 4). In milk and broth, cultures of the parent
strains prevented multiplication of Spc' mutants of the
same strain inoculated into them (Fig. 4). Several of
the mutants, including the cheR and motB mutants of
F98 and the trg and tar mutants of ST1 behaved very
much like the parent strains. In contrast, the oxrA and
cya mutants of Thax-1 and the tsr mutant of ST1
allowed some growth of the Spc' mutant of the parental
strain (the results of incubating Thax-1 Spc' in a milk
culture of the oxrA derivative of the Nal' mutant are
shown in Fig. 4). In broth, E. coli P4 did not prevent
multiplication of S. Typhimurium Thax-1 Spc' and the
rate of increase in the numbers of the challenge strain
was similar to the growth of Thax-1 on its own (Fig.
4), whereas inhibition occurred in milk. The pH values
of the E. coli cultures at the time of challenge were 7.0
(broth) and 5.2 (milk) although there was no indication
of gross clotting in the milk.
Discussion
In these experiments oral inoculation of gnotobiotic
pigs with Salmonella or E. coli strains resulted in
extensive colonisation with bacterial counts in the
faeces exceeding 10" cfu/g and in the case of the E.
coli strain, 10' c h / g . Colonisation by wild-type
Salmonella strains inhibited establishment by subsequently inoculated isogenic strains over the relatively
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 12:36:26
SALMONELLA COLONISATION IN PIGS
9 13
r b
12r a
r d
-1
0
2
4
Time (days)
6
a
-1
0
2
4
6
8
Time (days)
Fig. 3. Inhibition of intestinal colonisation between isogenic mutants of S. Typhimurium ST 1 using faecal bacterial
counts in pigs pre-colonised with non-inhibitory E. coli P4 (A).(a) S. Typhimurium STl Nal' (o), S. Typhimurium STl
Spc' (0).(b) S. Typhimurium ST330 (ST1 tsr) Nal' (o), S. Typhimurium ST1 Spc' (0).(c) S. Typhimurium ST328
(ST1 tar) Nal' (o), S. Typhimurium STl Spc' (0).(d) S. Typhimurium ST832 (ST1 trg) Nal' (o), S. Typhimurium STl
individual values from two pigs. For other details see legend to Fig. 1.
Spc' (0);
short observation period. This is very similar to the
colonisation-inhibition that occurs between related
members of the Enterobacteriaceae in the alimentary
tract of newly hatched chickens [16] where, in the
absence of an adult gut flora, very high bacterial counts
of these organisms are also attained [18]. Despite the
relatively small number of animals tested in the present
study, the variation in counts was generally small.
Colonisation of gnotobiotic pigs by S. Typhimurium
strains resulted in very high bacterial numbers in the
intestine which inhibited subsequent establishment by
isogenic mutants inoculated 24 h later. A less pronounced inhibition also occurred between isogenic
antibiotic resistant E. coli strains whereas E. coli had
no effect on the establishment of S. Typhimurium F98.
This degree of specificity was also found to occur with
similar organisms in conventional chicks where a
relatively poor level of colonisation-inhibition was also
found between E. coli strains [ 161. In that case, it was
attributed to the difficulty in establishing an E. coli
strain in birds which, in all probability, were already
completely or partially colonised naturally with E. coli.
Similar incomplete inhibition between heterologous E.
coli strains has also been observed in gnotobiotic mice
and pigs [35-371 although, in this case, it was unclear
how far the absence of complete inhibition was the
result of using non-isogenic strains. Much greater
inhibition was observed when isogenic mutants of a
K12 strain were used in gnotobiotic mice [38].
Therefore, it seems that the degree of inhibition of
E. coli may be strain-dependent as can occur with
Salmonella strains [39].
Because of the short interval between oral infection
with the first and second strains the inhibition effect is
likely to be purely microbiological, rather than an
immunological phenomenon. Because the mutants used
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 12:36:26
914
M. A. LOVELL AND P. A. BARROW
Table 3. The effect of intestinal colonisation of gnotobiotic pigs with S. Typhimurium strains Thax-1 Nal' or mutants
of this strain on the colonisation and mucosal association by the spectinomycin-resistant mutant of the parent given
orally 24 h later
Loglo cfu/g* of tissues or contents of pre-colonising and challenge strains in
~~~
~
Pigs 18 and 19
Pigs 16 and 17
Site sampled
Pigs 20 and 21
Pre-colonising
strain
Thax- 1
Nal'
Challenge
strain
Thax- 1
Spc'
Pre-colonising
strain
Thax- 1oxrA
Nal'
Challenge
strain
Thax-1
Spc'
Pre-colonising
strain
Thax- 1cya
Nal'
Challenge
strain
Thax- 1
Spc'
5.82
5.73
4.64
5.76
3.74
6.95
4.61
9.15
6.97
9.43
7.72
5.66
5.5 1
<2
<2
<2
<2
<2
<2
<2
2.30
5.16
4.88
4.44
5.72
4.13
6.37
4.01
9.44
6.08
8.49
6.14
5.66
5.51
5.19
4.57
3.55
5.18
3.13
6.44
3.49
8.83
5.62
8.62
5.71
1.69
2.00
6.0
5.56
3.91
7.28
4.61
7.69
4.97
9.1 1
6.97
8.81
7.25
2.78
2.48
<2
3.59
<2
3.90
<2
7.53
5.1 1
6.75
5.61
Stomach
contents
Duodenum contents
mucosa
Jejunum
contents
mucosa
Ileum
contents
mucosa
Colon
contents
mucosa
Caecum
contents
mucosa
Milk in trough Day 6
Day 7
<2
<2
<2
1.69
2.00
...
...
...
...
For doses see footnote to Table 2.
*Geometric mean of values from two pigs.
assessing the reduction in inhibition caused by mutations in genes affecting metabolism in the environment
of the gut has begun. Such genes might include those
affecting the switching between aerobic and anaerobic
metabolism, motility and chemotaxis. The latter could
be essential in the maintenance of high numbers of
bacteria in mucus lining the epithelium.
1
2
3
4
5
Time (days)
Fig. 4. Multiplication of S. Typhimurium Thax-1 Spc' in
24-h milk cultures of mutants of Thax-1 Nal'. 0, growth
curve of Thax-1 Spc' on its own; multiplication of Thax1 Spc' (m) in culture of Thax-1 Nal'; multiplication of
Thax-1 Spc' (a) in culture of oxrA mutants of Thax-1
Nal'.
were isogenic, the inhibition was not the result of
lysogenic bacteriophage or bacteriocin activity, which
has also been discounted as an explanation for the
colonisation-inhibition between Salmonella strains observed in chickens [ 161.
An investigation into the basis of the phenomenon by
The fact that the Thax-1 strain, which is non-motile,
and the motB and cheR mutants of strain F98 were
fully inhibitory suggests that motility per se is not
directly involved, although the inhibitory ability of
additional non-motile TnphoA mutants of another
strain of S. Typhimurium were found to be significantly
reduced [40]. Therefore, the reduced inhibition of
colonisation shown by the tsr mutant of strain STl was
of interest. This gene is responsible for the chemotactic
response to serine, alanine and glycine. However, the
viable counts of this strain in the faeces were 10-fold
lower than those of the parent and other strains. It is
possible that this may have been a factor in the absence
of inhibition produced. The fact that bacterial numbers
are not the only factor is indicated by the absence of
inhibition produced by the oxrA and cya mutants of
Thax-1, which colonised almost as well as the parent
strain. Mutations in oxrA [41] affect regulation of
carbon metabolism under different redox conditions
and are thus likely to affect a number of in-vivo
characteristics. Mutations in cya are pleiotropic [42461 and affect utilisation of carbon sources other than
glucose but also induce avirulence in mice [28].
The involvement of cya and oxrA suggests that these
particular mutants might have been less efficient at
using carbon sources, other than glucose, in the gut.
Even in the presence of these mutants in the gut, such
compounds would have been available for utilisation by
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 12:36:26
SALMONELLA COLONISATION IN PIGS
the parent strain allowing it to multiply. This is thought
to contribute to the inability of stationary phase broth
cultures of such mutants to inhibit multiplication of the
parent strain [46]. This also suggests that the organisms
in the large intestine, at least in this model, are present
in stationary phase for much of the time.
Analysis of bacterial counts in gut contents and those
associated with the mucosa indicated that the numbers
of bacteria in the latter niche generally reflect their
numbers in the lumen in the same way that occurs in
chicken gut [18]. In the chicken, and in the present
study, the number of bacteria of non-invasive strains
associated with the mucosa was always considerably
less than that in the lumen, supporting the contention
that the lumen is the main site of primary intestinal
colonisation (establishment) and that, unlike enterotoxigenic E. coli [47], extensive preferential colonisation of the mucosa does not occur. There seems little
ecological advantage in this occurring either in the
caecum, where semi-batch incubation conditions occur,
or in the colon, where the flow rate of the chyme is
relatively slow [20]. However, in both organs, as in the
chicken caeca, micro-organisms embedded in the
mucus must act as inoculum for fresh contents entering
these organs.
Given the high faeces counts, the bacterial counts in
the feeding troughs were sufficiently low to rule out the
possibility that inhibition may have been occurring in
vitro in the small amount of contaminated milk left
between feeds, as the pigs generally drank their milk
soon after being fed. The fact that the results obtained
in wholly liquid phase media in vitro were very similar
to those obtained in vivo suggested that the main site of
interaction between the strains is in the lumen of the
gut rather than at the level of the mucosa.
The practical issues that come out of these studies are
potentially very exciting. The results suggest that in
young animals fed milk diets the oral administration of
live attenuated vaccines, which colonise sufficiently
well, could induce a rapid inhibition of colonisation of
pathogenic organisms by a microbiological process
while remaining in the gut to stimulate the immune
system in the usual way. This sort of process would
depend on the chosen vaccine strain being able to
inhibit a sufficiently wide variety of pathogenic field
strains. Work with chickens suggests that at the
moment such strains have to be selected on an
empirical basis [17,39]. This aspect of vaccine use
for Salmonella in chickens is now being increasingly
utilised in Europe [48], although the importance of
selecting inhibitory strains from which to develop
vaccines is indicated by the poor inhibitory nature of
the live vaccines currently available [49].
We acknowledge the assistance of Miss K. Page and Mr A. Gray and
the financial support of the Ministry of Agriculture, Fisheries and
Food. Miss J. Howard typed the manuscript.
915
References
1. Buxton A. Salmonellosis in animals: a review. Farnham Royal,
Commonwealth Agricultural Bureaux. 1957.
2. Smith HW. Salmonella food poisoning in human beings. The
part played by domestic animals. R Soc Health J 1969; 89:
271-275.
3. Lava1 A, Morvan H, Desprez G, Corbion B. Salmonellosis in
swine. In: Proceedings of the International Symposium on
Salmonella and Salmonellosis, Ploufragan/St. Brieuc, France
1992: 164-175.
4. Newel1 KW, Williams LP. The control of Salmonellae affecting
swine and man. J Amer Vet Med Assoc 1971; 158: 89-98.
5. Wilcock Bl? Salmonellosis. In: Leman AD, Straw B, Glock
RD, Mengeling WL, Penny RHC, Scholl E (eds) Diseases of
swine 6th edn. Ames, Iowa, Iowa State University Press. 1992;
508-520.
6. Wray C. Salmonellosis in cattle and sheep. In: Proceedings of
the International Symposium on Salmonella and Salmonellosis,
Ploufragan/St. Brieuc, France 1992: 176-183.
7. Wray C, Sojka WJ. Review of dairy science: bovine salmonellosis. J Dairy Sci 1977; 44: 383-425.
8. Barrow PA. Salmonella control - past, present and future.
Avian Path01 1993; 22: 651-669.
9. Cooper GL. Salmonellosis-infections in man and the chicken:
pathogenesis and the development of live vaccines - a review.
Vet Bull 1994; 64: 123-143.
10. Barrow PA. Immunity to Salmonella and other bacteria. In:
Davison TF, Morris TR, Payne LN (eds) Poultry immunology.
Poultry Science Symposium Series, 1996; 24: 243-263.
11. Lumsden JS, Wilkie BN, Clarke RC. Resistance to faecal
shedding in pigs and chickens vaccinated with an aromaticdependent mutant of Salmonella typhimurium. Amer J Vet Res
1991; 52: 1784-1787.
12. Roof MB, Doitchinoff DD. Safety, efficacy and duration of
immunity induced in swine by use of an avirulent live
Salmonella choleraesuis-containing vaccine. Amer J Vet Res
1995; 56: 39-44.
13. Barrow PA, Hassan JO, Berchieri A. Reduction in faecal
excretion of Salmonella typhimurium strain F98 in chickens
vaccinated with live and killed S. typhimurium organisms.
Epidemiol Infect 1990; 104: 413-426.
14. Cooper GL, Venables LM, Nicholas RAJ, Cullen GA,
Hormaeche CE. Further studies of the application of live
Salmonella enteritidis aroA vaccines in chickens. Vet Rec 1993;
133: 31-36.
15. Cooper GL, Venables LM, Woodward MJ, Hormaeche CE.
Vaccination of chickens with strain CVL30, a genetically
defined Salmonella enteritidis aroA live oral vaccine candidate.
Infect Immun 1994; 62: 4747-4754.
16. Barrow PA, Tucker JF, Simpson JM. Inhibition of colonization
of the chicken alimentary tract with Salmonella typhimurium
by gram-negative facultatively anaerobic bacteria. Epidemiol
Infect 1987; 98: 3 1 1-322.
17. Berchieri A, Barrow PA. Further studies on the inhibition of
colonization of the chicken alimentary tract with Salmonella
typhimurium by pre-colonization with an avirulent mutant.
Epidemiol Infect 1990; 104: 427-441.
18. Barrow PA, Simpson JM,Lovell MA. Intestinal colonization in
the chicken by food-poisoning Salmonella serotypes; microbial
characteristics associated with faecal excretion. Avian Path01
1988; 17: 571-588.
19. Smith HW, Halls S. The simultaneous oral administration of
Salmonella dublin, S. typhimurium and S. choleraesuis to
calves and other animals. J Med Microbiol 1968; 1: 203-209.
20. Smith HW. Observations on the flora of the alimentary tract of
animals and factors affecting its composition. J Path01
Bacteriol 1965; 89: 95-122.
21. Barrow PA, Lovell MA, Barber LZ. Growth suppression in
early-stationary-phase nutrient broth cultures of Salmonella
typhimurium and Escherichia coli is genus specific and not
regulated by Fo. J Bacteriol 1996; 178: 3072-3076.
22. Smith HW, Tucker JF. The virulence of Salmonella strains for
chickens; their excretion by infected chickens. J Hyg 1980; 84:
479-488.
23. Giannella RA, Formal SB, Dammin GJ, Collins H. Pathogenesis of salmonellosis: studies of fluid secretion, mucosal
invasion, and morphologic reaction in the rabbit ileum. J Clin
Invest 1973; 52: 441-453.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 12:36:26
916
M. A. LOVELL AND P. A. BARROW
24. Wallis TS, Starkey WG, Stephen J, Haddon SJ, Osborne MP,
Candy DCA. Enterotoxin production by Salmonella typhimurium strains of different virulence. J Med Microbiol 1986; 21:
19-23.
25. Aswad D, Koshland DE. Isolation, characterization and
complementation of Salmonella typhimurium chemotaxis mutants. J Mol Biol 1975; 97: 225-235.
26. Stock JB, Surette MG. Chemotaxis. In: Neidhardt FC, Curtiss
R, Ingraham JL et al. (eds) Escherichia coli and Salmonella;
cellular and molecular biology, vol 1, 2nd edn. American
Society of Microbiology, Washington DC, ASM Press. 1996:
1103-1 129.
27. Jamieson DJ, Higgins CF. Anaerobic and leucine-dependent
expression of a peptide transport gene in Salmonella typhimurium. J Bacteriol 1984; 160: 131-136.
28. Curtiss R, Kelly SM. Salmonella typhimurium deletion mutants
lacking adenylate cyclase and cyclic AMP receptor protein are
avirulent and immunogenic. Infect Immun 1987; 55: 30353063.
29. Adler J, Hazelbauer GL, Dahl MM. Chemotaxis toward sugars
in Escherichia coli. J Bacteriol 1973; 115: 824-847.
30. Barrow PA, Huggins MB, Love11 MA. Host-specificity of
Salmonella infection in chickens and mice is expressed in vivo
primarily at the level of the reticuloendothelial system. Infect
Immun 1994; 62: 4602-4610.
31. Bochner BR, Huang H-C, Schieven GL, Ames BN. Positive
selection for loss of tetracycline resistance. J Bacteriol 1980;
143: 926-933.
32. Jones BD, Lee CA, Falkow S. Invasion by Salmonella
typhimurium is affected by the direction of flagellar rotation.
Infect Immun 1992; 60: 2475-2480.
33. Tavernor WD, Trexler PC, Vaughan LC, Jones DG. The
production of gnotobiotic piglets and calves by hysterectomy
under general anaesthesia. Vet Rec 1971; 88: 10-14.
34. Trexler PC. Microbiological isolation of large animals. Vet Rec
1971; 88: 15-20.
35. Duval-Iflah Y, Raibaud P, Rousseau M. Antagonisms among
isogenic strains of Escherichia coli in the digestive tracts of
gnotobiotic mice. Infect Immun 1981; 34: 957-969.
36. Duval-Iflah Y, Ovriet M-F, Moreau C, Daniel N, Gabilan JC,
Raibaud I? Implantation precoce d’une souche de Escherichia
coli dans l’intestin de nouveau-nb humains: effet de barriire
vis-a-vis de souches de E. coli antibioresistantes. Ann
Microbiol (Inst Pasteur) 1982; 133A: 393-408.
37. Duval-Iflah Y, Chappuis JP, Ducluzeau R, Raibaud P.
Intraspecific interactions between Escherichia coli strains in
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
human newborns and in gnotobiotic mice and piglets. Prog
Food Nutr Sci 1983; 7: 107-116.
Onderdonk A, Marshall B, Cisneros R, Levy SB. Competition
between congenic Escherichia coli K- 12 strains in vivo. Infect
Immun 1981; 32: 74-79.
Iba AM, Berchieri A, Barrow PA. Interference between
Salmonella serotypes in intestinal colonization of chickens:
correlation with in vitro behaviour. FEMS Microbiol Letts
1995; 131: 153-159.
Methner U, Barrow PA. Zur Bedeutung der Beweglichkeit von
Salmonella Enteritidis und Salmonella Typhimurium als Virulenzfaktor und fiir die Auspragung des Hemmphiinomens in
vitro und in vivo bei SPF- Hiihnerkuken. Berl Munch Tierurztl
Wochenschr 1997; 110: 391-396.
Spiro S, Guest JR. Adaptive responses to oxygen limitation in
Escherichia coli. Trends in Biochem Sci 1991; 16: 310-314.
Yokota R, Gots JS. Requirement of adenosine 3’,5’ -cyclic
phosphate for flagella formation in Escherichia coli and
Salmonella typhimurium. J Bacteriol 1970; 103: 5 13-516.
Komeda Y, Suzuki H, Ishidsu J, Iino T. The role of CAMP in
flagellation of Salmonella typhimurium. Mol Gen Genet 1975;
142: 289-298.
Pastan I, Adhya S . Cyclic adenosine 5’-monophosphate in
Escherichia coli. Bacteriol Rev 1976; 40: 527-55 1.
Saier MH, Schmidt MR, Leibowitz M. Cyclic AMP-dependent
synthesis of fimbriae in Salmonella typhimurium: effect of cya
and pts mutations. J Bacteriol 1978; 134: 356-358.
Mowa RN, Green P, Nakamura K, Inouye M. Interaction of
CAMP receptor protein with the ompA gene, a gene for a
major outer membrane protein of Escherichia coli. FEBS Lett
1981; 128: 186-190.
Smith H-W. The influence of plasmid-determined and other
characteristics of entero-pathogenic Escherichia coli on their
ability to proliferate in the alimentary tracts of piglets, calves
and lambs. J Med Microbiol 1977; 11: 471-492.
Martin G, Barrow PA, Berchieri A, Methner U, Meyer H.
Hemmphanomene zwischen Salmonella-Stbunen-ein neuer
Aspekt der Bekhpfung der Salmonellose beim Geflugel.
[Inhibition phenomena between Salmonella strains - a new
aspect of salmonella infection control in poultry]. DTW
Deutsch Tierarztl Wochenschr 1996; 103: 468-472.
Methner U, Barrow PA, Martin G, Meyer H. Comparative study
of the protective effect against Salmonella colonisation in newly
hatched SPF chickens of using live, attenuated Salmonella
vaccine strains, wild-type Salmonella strains or a competitive
exclusion product. Int J Food Prot 1997; 35: 223-230.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 12:36:26