necrotic enteritis-producing strains of clostridium perfringens displace non-necrotic enteritis...
TRANSCRIPT
www.elsevier.com/locate/vetmic
Veterinary Microbiology 126 (2008) 377–382
Short communication
Necrotic enteritis-producing strains of Clostridium perfringens
displace non-necrotic enteritis strains from the gut of chicks
Angelique J. Barbara, Hien T. Trinh, Robert D. Glock, J. Glenn Songer *
Department of Veterinary Science and Microbiology, The University of Arizona, Tucson, AZ 85721, USA
Received 17 May 2007; received in revised form 17 July 2007; accepted 20 July 2007
Abstract
We inoculated broiler chicks with mixtures of Clostridium perfringens strains to investigate the single strain dominance
observed in natural cases of necrotic enteritis (NE) [Nauerby, B., Pedersen, K., Madsen, M., 2003. Analysis by pulsed-field gel
electrophoresis of the genetic diversity among Clostridium perfringens isolates from chickens. Vet. Microbiol. 94, 257–266].
Pre-inoculation bacteriologic culture of chick intestines yielded up to six pulsed-field gel electrophoresis (PFGE) types of C.
perfringens. Birds developed typical NE lesions in response to administration (2� per day for 4 days) of a combined inoculum
comprising one NE strain (JGS4143, PFGE pattern 8) and four non-NE strains (from piglet necrotizing enteritis, chicken normal
flora, human gas gangrene, and bovine neonatal enteritis). After inoculation commenced, only the NE strain was recovered
through the first post-inoculation day, in spite of intense efforts to recover pre-challenge flora strains and the other challenge
strains. Thereafter, pre-inoculation and previously undetected PFGE types were found, and JGS4143 became undetectable.
Birds inoculated simultaneously with five NE strains (from disease in chickens or turkeys, and including JGS4143) also
developed lesions, but again only JGS4143 was recovered through the 1st day post-challenge. At that time, birds began to be
repopulated with pre-challenge PFGE types. Two NE strains (JGS4143 and JGS4064) produced bacteriocins, which inhibited
each other and normal flora strains (n = 17), while normal flora strains inhibited neither NE strains nor each other. Thus, it
appears that naturally occurring dominance of the gut by NE strains can be reproduced experimentally. Bacteriocins directed
against normal flora could possibly provide the necessary advantage, although inhibition of one NE strain by another suggests
that other factors may be partially or completely responsible for the dominance.
# 2007 Elsevier B.V. All rights reserved.
Keywords: Poultry necrotic enteritis; Clostridium perfringens; Bacteriocin
* Corresponding author. Tel.: +1 520 621 2962;
fax: +1 520 621 6366.
E-mail address: [email protected] (J. Glenn Songer).
0378-1135/$ – see front matter # 2007 Elsevier B.V. All rights reserved
doi:10.1016/j.vetmic.2007.07.019
1. Introduction
Poultry necrotic enteritis (NE) is commonly caused
by Clostridium perfringens type A (Parish, 1961;
Elwinger et al., 1992; Songer, 1996; Williams, 2005)
and occurs worldwide (Songer, 1996; Williams et al.,
2003). C. perfringens type A is a member of the
.
A.J. Barbara et al. / Veterinary Microbiology 126 (2008) 377–382378
intestinal flora of birds, but normal flora strains do not
produce lesions following experimental inoculation.
Strains from cases of NE are virulent, producing
disease of greater-or-lesser severity in experimentally
inoculated chicks (unpublished data).
C. perfringens is genetically diverse (Myers et al.,
2006), but isolates from birds in an NE outbreak are
usually genetically identical at the discriminatory
level of pulsed-field gel electrophoresis (PFGE). Up to
five PFGE types were isolated from normal birds in
Danish broiler flocks (Nauerby et al., 2003), but
examination of isolates from NE-affected chickens
revealed a single PFGE type in each flock. PFGE types
detected during NE outbreaks were not detected
before or after the outbreak, suggesting that, under
appropriate (but unknown) conditions, NE strains have
a competitive advantage over normal flora.
Elucidation of mechanisms by which NE strains
displace non-NE strains from the chicken intestine
may, in time, reveal information about pathogenesis
and provide targets for management of NE. We
present here the results of experimental coinfection
studies, the findings of which parallel those in natural
outbreaks and demonstrate that NE strains displace
non-NE strains from the gut of affected chicks.
2. Materials and methods
2.1. Bacteria and cultivation
Strains of C. perfringens used in these studies
(Table 1) were cultivated routinely on brain heart
infusion (BHI) agar (Difco, Detroit, MI) with 5%
citrated bovine blood. Plates were incubated at 37 8C
Table 1
Clostridium perfringens strains used in these studies
Strain number Study Source
JGS1235 2 Chicken
JGS1473 1 Chicken
JGS1521 2 Chicken
JGS1882 1 Porcine N
JGS1936 1 Bovine n
JGS4064 2 Chicken
JGS4104 2 Turkey N
JGS4143 1/2 Chicken
JGS4151 1 Strain 13
None N/A Normal fl
in an atmosphere of 5% H2:5% CO2:90% N2. Storage
of strains was in 50% glycerol at �80 8C.
2.2. Experimental reproduction of NE
Day-old, female Jumbo Cornish � Rock broiler
chicks (n � 25 per group; McMurray Hatchery,
Webster City, IA), housed in one-third of 1.5 m
circular brooders, were fed a commercial chick starter
on days 1–7 and a 28% protein wheat-based ration
mixed 1:1 with fishmeal thereafter. Water was
available ad libitum. Bedding (fine wood shavings)
was replaced at 2-week intervals. Environmental
temperature was maintained by allowing birds to self-
regulate their proximity to infrared heat lamps. On day
14, food was withheld for 20 h, followed by challenge.
Challenge inoculum was prepared by serial passage
through increasing volumes of cooked meat medium
(CMM; Difco) and fluid thioglycollate medium (FTG;
Difco), to a final FTG culture volume of 1 L.
Individual challenge inocula were administered alone
or combined to provide equal numbers of colony-
forming units of each challenge strain. In either case,
inocula were mixed 1.25:1 with high protein feed and
offered to chicks ad libitum, twice daily for 4 days.
Based upon loss of an estimated 25% of feed:inoculum
mixture due to scratching behavior and the routine
consumption of all remaining mixture, we estimate
that each chick consumed �3 � 109 colony-forming
units at each of the eight feedings.
2.3. Individual chick inoculation studies
In Study 1, the principal group was inoculated with
one NE strain (JGS4143) and four non-NE strains
Genotype
NE, hepatitis A, cpb2 pos
normal flora A, cpb2 pos
NE A, cpb2 pos
E case A, cpb2 pos
eonatal enteritis case A, cpb2 neg
NE A, cpb2 neg
E A, cpb2 pos
NE A, cpb2 pos
(human gas gangrene) A, cpb2 pos
ora from chicks in these studies A, cpb2 neg
A.J. Barbara et al. / Veterinary Microbiology 126 (2008) 377–382 379
(JGS1882, JGS1473, JGS4151, and JGS1936). In
Study 2, principal group chicks were challenged with
five NE strains (JGS4143, JGS1235, JGS1521,
JGS4104, and JGS4064). Relative virulence of NE
strains (based upon percent of birds developing lesions
after inoculation) was 4143 > 4104 > 1521 >4064 > 1235. In both studies, positive control chicks
were challenged with JGS4143 and negative controls
were offered feed mixed with uninoculated FTG.
2.4. Bacteriologic and pathologic examination of
inoculated chicks
Two to four chicks were euthanized by CO2
asphyxiation, at intervals beginning 5 days before
challenge and continuing through 5 days post-
challenge. Intestines were removed and opened
aseptically, and jejunum and ileum were examined
for gross lesions. Segments (�10 mm) were fixed in
10% phosphate-buffered formalin and paraffin-
embedded. Five micrometre sections were stained
with haemotoxylin and eosin and examined for
microscopic lesions. To maximize the rigor of
attempts to recover PFGE types representative of
the population at the time of necropsy, the entire
mucosal layer was removed from a 10 cm segment of
jejunum and ileum by scraping with sterile micro-
scope slides. This material was examined by
bacteriologic culture on SPS agar (Merck, Rahway,
NJ), with results expressed as 1+ (growth in the initial
streak only) to 4+ (growth in the initial area of the
streak, plus the three parts of the streak for isolation).
All isolated colonies were collected for examination
by PFGE and for genotyping by a multiplex PCR
method (Meer and Songer, 1997).
2.5. PFGE typing of isolates
Each isolate or strain was cultivated overnight in
10 ml NIH thioglycollate broth with 1% starch and
180 mg chloramphenicol per ml. Cells harvested by
centrifugation (4000 � g, 20 min) were washed once
with SE wash buffer (75 mM NaCl, 25 mM EDTA (pH
8)) and resuspended to an OD600 of 1.3 in cell
suspension buffer (10 mM Tris (pH 7.2), 20 mM
NaCl, 50 mM EDTA). Cells from 1 ml of this
suspension were harvested by centrifugation and
resuspended in 50 ml of the same buffer. The
suspension was equilibrated to 56 8C, combined with
an equal volume of 2% CleanCut agarose (BioRad,
Hercules, CA), mixed gently, and poured into plug
molds. A 1 ml volume of plugs was suspended in 5 ml
of lysozyme buffer (10 mM Tris (pH 7.2), 50 mM
NaCl, 0.2% sodium deoxycholate, 0.5% sodium lauryl
sarcosine, and 1 mg lysozyme per ml) and incubated
for 16 h at 37 8C. Buffer was then removed and plugs
rinsed with proteinase K buffer (100 mM EDTA (pH
8.0), 0.2% sodium deoxycholate, and 1% sodium
lauryl sarcosine). Proteinase K reaction buffer
(100 mM EDTA (pH 8.0), 0.2% sodium deoxycholate,
1% sodium lauryl sarcosine, and proteinase K (1 mg
per ml)) was added (5 ml buffer for each ml of plugs),
followed by overnight incubation at 56 8C without
agitation. Plugs were then washed four times with
50 ml wash buffer (20 mM Tris (pH 8.0), 50 mM
EDTA) and stored at 4 8C.
Plugs suspended in 200 ml of NE buffer 4 (New
England Biolabs, Ipswich, MA) were incubated at
25 8C for 15 min and then 12 U of SmaI (New
England Biolabs) were added, followed by overnight
incubation at 25 8C. Buffer was then removed and
plugs incubated in 1 ml of 0.66� TBE buffer (60 mM
Tris–60 mM boric acid–1.3 mM EDTA (pH 8.3)) for
�30 min with gentle agitation. Plugs were cut to size
with a razor blade and each was placed on a tooth of
the comb. One percent low-melt preparative-grade
agarose (BioRad) was cast around the comb, which
was then removed, leaving the plug in place. PFGE
was performed in a CHEF-DR II apparatus (BioRad)
with pulse times ramped from 0.5 to 40 s over 20 h.
Gels were photographed via UV transillumination and
banding patterns were analyzed visually. PFGE
patterns were given arbitrary numbers.
2.6. Bacteriocin production
Each strain was cultivated to late log phase in BHI,
as above. An aliquot was adjusted to a density
equivalent to McFarland No. 0.5, and 100 ml amounts
were spread on BHI agar plates to prepare lawns. A
colony of each bacteriocin test strain was collected on
the tip of a sterile toothpick and stabbed through the
lawn and partially through the agar beneath. Plates
were incubated for 24 h at 37 8C under anaerobic
conditions and diameters of zones of inhibition were
measured in mm.
A.J. Barbara et al. / Veterinary Microbiology 126 (2008) 377–382380
2.7. Approval of animal use
Study protocols were reviewed and approved a
priori by the University of Arizona Animal Care and
Use Committee.
3. Results
3.1. Genetic characterization of C. perfringens
recovered pre-, during, and post-challenge
Study 1 birds necropsied up to and including the
first inoculation day had no gross or microscopic
lesions. Isolates from these birds comprised six PFGE
types, all distinct from those of the inoculum strains.
Mild hyperemia of jejunal mucosa was observed on
the second inoculation day, without other lesions, but
the striking change was the disappearance of strains
with pre-challenge PFGE types (Table 2). Lesions of
varying severity but typical of NE appeared on day 3
of inoculation and were present through the first day
post-inoculation. In spite of rigorous attempts to
Table 2
Bacteriologic and pathologic examination of Study 1 birds, before, durin
Time Group
Pre-challenge N/A
During challengea Positive control
During challengeb Negative control
During challengea Five strains
Post-challengec Five strains, positive control
a Collected on challenge days 1–3.b Collected on challenge day 3.c Collected on post-challenge days 1–3.
Table 3
Bacteriologic and pathologic examination of Study 2 birds, before, durin
Time Group
Pre-challenge N/A
During challengea Positive control
During challengeb Negative control
During challengea Five strains
Post-challengec Positive control
Post-challengec Five strains
a Collected on challenge days 1–3.b Collected on challenge day 3.c Collected on post-challenge days 1–3.
isolate any strain that was present, only PFGE type 8
(inoculum strain JGS4143) was recovered through the
first post-inoculation day. After the first post-inocula-
tion day, pre-inoculation PFGE types were again
found, accompanied by PFGE type 8 and previously
undetected type 16. Thereafter, PFGE type 8 was not
isolated again, and types 17 and 18 appeared.
Study 2 birds yielded five PFGE types pre-challenge,
none of which matched the patterns of the inoculum
strains (Table 3). The pattern of lesion development and
resolution was the same as in Study 1. During the 4 days
of challenge, PFGE type 8 was isolated from all birds
challenged with five NE strains, and PFGE type 10 (not
an inoculum type) was also found in a few. Strains
JGS1235, JGS1521, JGS4064, and JGS4104 had each
produced signs and lesions of NE in experimentally
inoculated birds (unpublished data). However, they
were not recovered from birds inoculated with these 4 in
the company of JGS4143 (PFGE type 8) in this study.
PFGE type 8 alone was found in positive control birds
during challenge, and PFGE types 8, 10, 12, 15, 16, and
17 were found post-challenge in positive controls and in
birds challenged with the mix of five strains.
g, and after inoculation
Average lesion score PFGE strain number(s)
0 10, 11, 12, 13, 14, 15
2.3 8
0 10, 15
2.4 8
0 8, 12, 13, 14, 15, 16, 17, 18
g, and after inoculation
Average lesion score PFGE strain number(s)
0 10, 11, 12, 13, 14
2.4 8
0 15
2.2 10, 8
0.9 10, 8, 16
1.0 8, 12, 15, 17
A.J. Barbara et al. / Veterinary Microbiology 126 (2008) 377–382 381
Table 4
Bacteriocin production (zone diameters in mm) of NE and non-NE strains
Bacteriocin test strains
4064 4143
Lawns 4064 0 5.6a
4143 5.7 0
Normal flora strains (n = 17) 6.6b (range 3.5–11)c 6.8 (range 4–11)
A zone denoted ‘‘0’’ indicate no inhibition.a Average of three repetitions.b Average across all 17 normal flora strains.c Range of averages (each based upon three repetitions) for normal flora strains.
3.2. Bacteriocin production by NE and non-NE
strains
All NE strains inhibited growth of normal flora, but
normal flora strains did not inhibit any NE strain (i.e.,
there was no zone of inhibition). Representative data
are presented in Table 4. Two NE strains (JGS4143
and JGS4064) inhibited each other and normal flora
strains (n = 17). Normal flora strains did not inhibit
each other (data not shown).
4. Discussion
Replacement of the genetically heterogeneous
(based upon PFGE analysis) enteric population of
C. perfringens, usually by single, chicken-virulent
strains has been noted in natural outbreaks of NE
(Engstrom et al., 2003; Nauerby et al., 2003). We
reasoned that experimental reproduction of naturally
occurring strain dominance would facilitate study of
population dynamics of C. perfringens in infected
chicks and perhaps lead to identification of specific
virulence attributes. Genome hyperplasticity (Myers
et al., 2006) augurs against use of PFGE to draw
definitive conclusions about relatedness among
diverse strains, but disparate PFGE types imply
variety in the population; strains with identical PFGE
types are likely to be identical, especially when
comparing inoculum and output strains from the same
birds: it is rational to assume that appearance of an
inoculum PFGE type in post-inoculation cultures is
the result of inoculation. Thus, PFGE was useful in
these studies for tracking both normal flora and
challenge strains. The disappearance of pre-challenge
PFGE types upon challenge with NE strains and the
disappearance of PFGE type 8 (NE strain JGS4143)
and reappearance of pre-inoculation and new PFGE
types upon recovery has the appearance of the sort of
strain dominance seen in the field (Engstrom et al.,
2003; Nauerby et al., 2003). Apparent inhibition of
other NE strains by JGS4143 suggests that there is no
equality in this group.
It is possible that the pre- and post-inoculation
strains were in the gut throughout the study and simply
went undetected due to small numbers. Our data do
not rule out this possibility, but it seems no more likely
than that birds became re-colonized through exposure
to their previous flora strains in the environment. In
any case, the point is the same: strains which were
readily detectable, or introduced via challenge, were
greatly reduced in numbers or eliminated in the
presence of an NE strain.
These experimental results confirm those of
individual inoculation studies that differences exist
in virulence for chicks among C. perfringens strains. It
seems clear that NE strains have a selective advantage
over non-NE strains and even degrees of selective
advantage over each other. The last finding suggests
that variations in strain virulence may express
themselves not only in the proportion of birds
developing lesions upon challenge, but also in
interstrain competition for a place in the intestine.
These findings are in keeping with those obtained
from studying natural disease (Engstrom et al., 2003;
Nauerby et al., 2003).
There is no evidence that differences in generation
time vary significantly among NE strains or between
NE and non-NE strains (unpublished data). This
should perhaps be explored further. However, bacter-
iocin action is a documented means by which bacteria
compete for resources and C. perfringens bacteriocin
A.J. Barbara et al. / Veterinary Microbiology 126 (2008) 377–382382
production has been reported (Dupuy et al., 2005;
Dupuy and Matamouros, 2006). All NE strains
examined in this study produced substances that
strongly inhibited non-NE strains; the reverse did not
occur, nor did non-NE strains inhibit each other. As far
as we are aware, there is no information on in vivo
bacteriocin production by C. perfringens, but it is
tempting to speculate that elaboration of bacteriocins in
the gut allows displacement of normal flora strains and
anarchic multiplication by NE strains. This is tempered
by the fact that strains JGS4143 and JGS4064 inhibited
each other about equally in vitro and JGS4143 out-
competed JGS4064 in vivo. Thus, the phenomenon may
be due, completely or in part, to factors other than
bacteriocins, such as superior adhesion characteristics,
more rapid multiplication, and production of specific
toxins. Further study is indicated.
Acknowledgements
The authors acknowledge, with gratitude, the
technical assistance of Jeremy Coombs. Supported in
part by a grant from Schering-Plough Animal Health.
References
Dupuy, B., Matamouros, S., 2006. Regulation of toxin and bacter-
iocin synthesis in Clostridium species by a new subgroup of
RNA polymerase sigma-factors. Res. Microbiol. 157, 201–205.
Dupuy, B., Mani, N., Katayama, S., Sonenshein, A.L., 2005. Tran-
scription activation of a UV-inducible Clostridium perfringens
bacteriocin gene by a novel sigma factor. Mol. Microbiol. 55,
1196–1206.
Elwinger, K., Schneitz, C., Berndtson, E., Fossum, O., Teglof, B.,
Engstom, B., 1992. Factors affecting the incidence of necrotic
enteritis, caecal carriage of Clostridium perfringens and bird
performance in broiler chicks. Acta Vet. Scand. 33, 369–
378.
Engstrom, B.E., Fermer, C., Lindberg, A., Saarinen, E., Baverud, V.,
Gunnarsson, A., 2003. Molecular typing of isolates of Clostri-
dium perfringens from healthy and diseased poultry. Vet. Micro-
biol. 94, 225–235.
Meer, R.R., Songer, J.G., 1997. Multiplex polymerase chain reaction
assay for genotyping Clostridium perfringens. Am. J. Vet. Res.
58, 702–705.
Myers, G., Rasko, D., Cheung, J., Ravel, J., Seshadri, R., DeBoy, R.,
Ren, Q., Varga, J., Awad, M., Brinkac, L., Daugherty, S., Haft,
D., Dodson, R., Madupu, R., Nelson, W., Rosovitz, M., Sullivan,
S., Khouri, H., Dimitrov, G., Watkins, K., Mulligan, S., Benton,
J., Radune, D., Fisher, D., Atkins, H., Hiscox, T., Jost, H.,
Billington, S., Songer, J.G., McClane, B., Titball, R., Rood,
J., Melville, S., Paulsen, I., 2006. Skewed genomic variability in
strains of the toxigenic bacterial pathogen, Clostridium perfrin-
gens. Genome Res. 16, 1031–1040.
Nauerby, B., Pedersen, K., Madsen, M., 2003. Analysis by pulsed-
field gel electrophoresis of the genetic diversity among Clos-
tridium perfringens isolates from chickens. Vet. Microbiol. 94,
257–266.
Parish, W.E., 1961. Necrotic enteritis in the fowl (Gallus gallus
domesticus). I. Histopathology of the disease and isolation of a
strain of Clostridium welchii. J. Comp. Pathol. 71, 377–393.
Songer, J.G., 1996. Clostridial enteric diseases of domestic animals.
Clin. Microbiol. Rev. 9, 216–234.
Williams, R.B., 2005. Intercurrent coccidiosis and necrotic enteritis
of chickens: rational, integrated disease management by main-
tenance of gut integrity. Avian Pathol. 34, 159–180.