HAL Id: hal-00532525
https://hal.archives-ouvertes.fr/hal-00532525
Submitted on 4 Nov 2010
HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.
Genomic, phenotypic changes of Campylobacter jejuni strains after passage of the chicken gut
I. Hänel, E. Borrmann, J. Müller, W. Müller, B. Pauly, E.M. Liebler-Tenorio, F. Schulze
To cite this version:
I. Hänel, E. Borrmann, J. Müller, W. Müller, B. Pauly, et al.. Genomic, phenotypic changes of Campylobacter jejuni strains after passage of the chicken gut. Veterinary Microbiology, Elsevier, 2009, 136 (1-2), pp.121. �10.1016/j.vetmic.2008.10.018�. �hal-00532525�
Accepted Manuscript
Title: Genomic, phenotypic changes of Campylobacter jejuni strains after passage of the chicken gut
Authors: I. H¨anel, E. Borrmann, J. M¨uller, W. M¨uller, B.
Pauly, E.M. Liebler-Tenorio, F. Schulze
PII: S0378-1135(08)00494-X
DOI: doi:10.1016/j.vetmic.2008.10.018
Reference: VETMIC 4247
To appear in: VETMIC Received date: 27-6-2008 Revised date: 16-10-2008 Accepted date: 21-10-2008
Please cite this article as: H¨anel, I., Borrmann, E., M¨uller, J., M¨uller, W., Pauly, B., Liebler-Tenorio, E.M., Schulze, F., Genomic, phenotypic changes of Campylobacter jejuni strains after passage of the chicken gut, Veterinary Microbiology (2008), doi:10.1016/j.vetmic.2008.10.018
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Accepted Manuscript
Genomic and phenotypic changes of Campylobacter jejunistrains after passage of the
1
chicken gut
2
3
I. Hänel1*, E. Borrmann2, J. Müller1, W. Müller1, B. Pauly1, E. M. Liebler-Tenorio2, F. Schulze1
4
5
Friedrich-Loeffler-Institute (1Institute of Bacterial Infections and Zoonoses, Jena, 2Institute of
6
Molecular Pathogenesis, Jena,) Naumburger Str. 96a, 07743 Jena Germany
7
*Corresponding author. Tel.: +49-3641 804-375; fax: +49-3641 804-228.
8
E-mail address: ingrid.haenel@fli.bund.de(I. Hänel)
9
10 11
Abstract
12
13
The ability to colonize the chicken gut was determined for 17 Campylobacter jejunistrains of
14
human and bovine origin. The level of colonization varied according to the strain used for
15
experimental infection. Two Campylobacterisolates from patients suffering from
16
gastroenteritis were found in the group of non-colonizing strains, suggesting that other
17
reservoirs as poultry are also important sources of human Campylobacterinfections. Bovine
18
Campylobacter isolates can also effective colonize the chicken intestine and may be a
19
source for poultry infection. The invasion ability of the strains as determined in the cell culture
20
model using Caco-2 cells correlates with their colonization capacity in the chicken gut. The
21
genomic and phenotypic stability of the selected strains were evaluated by analysis of their
22
pulsed-field gel electrophoresis (PFGE) patterns, flaA-typing and in vitrodetermination of
23
motility, adhesion and invasion abilities after colonizing chickens for up to 21 days. Changes
24
were identified in flaA-types of six isolates and three isolates from chicken showed different
25
patterns by PFGE using SmaI or KpnI as restriction enzymes. One isolate showed
26
phenotypic differences after in vivo passage which were seen in enhancement of adherence
27
* Manuscript
Accepted Manuscript
to eukaryotic cells, decrease of motility and changes in morphology. These phenotypic
28
changes were not associated with the observed genomic instabilities.
29 30
Keywords: Campylobacter jejuni, Invasion, Chicken colonization, Typing
31
32
1. Introduction
33
34
The food-borne pathogen Campylobacter (C.) jejuniand, to a smaller extent, the closely
35
related C. colihave become recognized as probably the most common cause of sporadic
36
bacterial gastroenteritis. Most infections by Campylobacterhave a zoonotic background
37
associated with consumption of contaminated animal products or animal contact. Poultry
38
products are now widely accepted as a common source of human infection with
39
Campylobacter.C. jejunistrains differ markedly in their ability to colonize the chicken gut
40
ranging from strains with immediately sustained colonization to completely non-colonizing
41
strains (Ringoir and Korolik 2003, Hänel et al., 2004). The differences in colonization types
42
may be due to genetic differences, or differences in gene expression of colonization/invasion
43
related genes. C. jejuniexhibits several potential virulence properties of which the best
44
characterized include flagellum-mediated motility, adhesion and invasion capability (Young et
45
al., 2007, Wassenaar and Blaser, 1999). In the chick colonization model could be shown that
46
virulence-associated genes such as ciaB, cadF (Ziprin et al., 2001), flhA (Carrillo et al.,
47
2004), docA, docB, docC (Hendrixson and DiRita, 2004), and the two-component systems
48
racRS (Bras et al., 1999) and dccRS (MacKichan et al., 2004) are involved in colonization of
49
the chick gastrointestinal tract by C. jejuni.
50
Both phenotypic and genomic changes of C. jejunistrains after experimental infection of
51
chickens may occur. An increase of colonization potential for C. jejunistrains after passage
52
in vivothrough a chicken was reported (Cawthraw et al., 1996, Ringoir and Korolik, 2003).
53
Several reports have described genomic changes of C. jejuniafter passage through chick
54
intestine because of genomic rearrangement, insertions, deletions or point mutations using
55
Accepted Manuscript
pulse-field gel electrophoresis (PFGE) (Hänninen et al. 1999, Wassenaar et al., 1998, de
56
Boer et al., 2002). However, there is a lack of information if the genetic instabilities correlates
57
with virulence related phenotypic properties such as adhesion or invasion.
58
In the present study we investigated the colonization ability of 17 C. jejuniisolates of human
59
and bovine origin using the chicken gut model. Further the genomic and phenotypic stability
60
of the C. jejunistrains were investigated after passage of the chicken gut. For this purpose
61
strains used for inoculation were compared with the corresponding isolates after three weeks
62
lasting colonization in the chicken gut with regard to adhesion and invasion in the cell culture
63
model using Caco-2 cells and the detection of 23 putative virulence genes by PCR. The
64
genomic stability of the strains was studied after passage through chicken intestine by PFGE
65
and flagellin gene typing.
66 67
2. Material and methods
68
2.1. Bacterial isolates
69
17 C. jejuniisolates of human and bovine origin (Table 1) were investigated. The strains
70
were characterized by different abilities to adhere to and to invade into Caco-2-cells. All
71
strains were cultivated on Müller-Hinton (MH) agar plates supplemented with 10 % bovine
72
citrated blood under microaerobic (85 % N2, 10 % CO2, 5 % O2) conditions at 37 oC. In
73
investigations of invasion on epithelial cells, the non-invasive E. coliK12 strain was included
74
as a negative control. This strain shows a percentage entry in the range of 0.001 % - 0.01 %
75
although 21 % of the inoculum was found to adhere to Caco-2 cells.
76
2.2. Experiments on colonization of C. jejuniin the chicken gut
77
White Leghorn chicks used in this study were hatched from specific pathogen-free eggs
78
obtained from the companies Lohmann and Charles River. The groups of animals were
79
housed in different, but adjacent rooms under comparable conditions. In each experiment on
80
day of hatch chicks were randomly separated into groups of 20 birds each. Chicks were
81
reared in cages and fed ad libitum with a commercial starter ration devoid of antibiotic or
82
coccidiostat. They were provided with chlorinated drinking water. Cloacal swab samples
83
Accepted Manuscript
were taken before inoculation to confirm the Campylobacternegative status of the animals at
84
the start of the experiment. At 9 days of age, the chickens of a group received approximately
85
1 x 108colony forming units (cfu) of one C. jejunistrain via oesophageal gavage. Bacterial
86
inoculations were prepared from frozen stock cultures of C. jejuniwhich were thawed,
87
inoculated on MH blood agar plates and incubated microaerobically for 48 h at 37 °C.
88
Bacteria were harvested and resuspended in sterile phosphate-buffered saline (PBS), pH
89
7.4. The suspension was adjusted using spectrophotometry and the mean numbers of cfu
90
per ml were determined by performing standard plate counts. 1, 2, 6, 14 and 21 days after
91
inoculation, 4 chicks of each group were sacrificed by cervical dislocation and liver and
92
faeces were collected for processing. Faecal samples were obtained from intact ceca,
93
weighed, homogenized with sterile buffer (stomacher), diluted and inoculated onto CCDA
94
medium and modified Skirrow medium supplemented with 30 µg/ml Cefoperazone, 10 µg/ml
95
Amphotericin B and 10 % calf blood. The liver samples were processed in the same manner.
96
The plates were incubated microaerobically at 37 °C for 48 h, after which time the
97
Campylobactercolonies were enumerated. A liquid enrichment based on CCDA medium
98
without agar was also used. At the end of each experiment, the colonizing strains were
99
reisolated from the ceca of 4 chickens. A plate sweep was recovered from each bird for
100
comparison with the corresponding C. jejunistrain used for inoculation.
101
The chicken colonization experiments were performed according to the law and approved by
102
the ethical committee of Thuringia (04-01/03).
103
2.3. Adherence and invasion assay
104
Caco-2 cells (ACC 169, German Collection of Microorganisms and Cell Cultures
105
Braunschweig) were grown in Dulbecco's Modified Eagle's Medium supplemented with 10 %
106
fetal bovine serum (FBS) and 1 % non-essential amino acids without the use of antibiotics at
107
37 °C in a 5 % CO2, humidified atmosphere. For the experimental assays, Caco-2 cells were
108
grown in 24-well plastic plates. The cells were seeded at 4.5 x 104cells per well and
109
incubated for 7 days at 37 °C in a 5 % CO2, humidified atmosphere. Prior the assay, the cell
110
Accepted Manuscript
monolayers were once washed with PBS, pH 7.2. Adhesion and invasion assays were
111
performed as described by Hänel et al. (2004), except that the bacteria were grown 24 h
112
before they were harvested from plates. Briefly, for determination of adherent bacteria cells
113
infected with approximately 1 x 108cfu/ml were incubated for 3 h at 37 °C in a 5 % CO2,
114
humidified atmosphere. In order to quantify bacterial invasion the cells were treated for
115
another 2 h with 150 µg/ml gentamicin to kill extracellular bacteria. Adherent or intracellular
116
bacteria were evaluated after cell lysis with 1 % Triton X-100 by plating serial dilutions of the
117
cell lysates on MH agar plates and by counting the resulting colonies. Significance between
118
samples was determined with t test and a Pvalue of < 0.05 was considered significant.
119
2.4. Detection of virulence-associated genes by PCR
120
DNA preparation from a 24 h culture on MH agar and PCR detection of 18 virulence
121
associated genes (flaA, flaB, flhA, flhB, flgB, flgE2, fliM, fliY, ciaB, iamA, virB11, cadF, docA-
122
C, cdtA-C) were carried out according to Müller et al. 2006. Additionally, virulence genes
123
secD, secF, Cj1470c, Cj1471c, and Cj1474cwere detected using primers and annealing
124
temperatures listed in Table 2. For the generation of PCR primers, the Primer Express
125
software (Applied Biosystems) was used.
126
2.5. Flagellin gene typing and RFLP analysis
127
PCR-RFLP profiles of the flaA gene were produced as previously described (Hänel et al.,
128
2004) using restriction endonucleases DdeI or AluI (New England Biolabs). Typing and RFLP
129
analysis of flaB gene were additionally performed using following primers (Smith et al.,
130
1999): FlaB3 (5'-ATA AAC ACC AAC ATC GGT GCA-3') and FlaB4 (5'-GTT ACG TTG ACT
131
CAT AGC ATA-3') with the same PCR and restriction conditions such as used forflaA typing.
132
2.6. Pulsed-field gel electrophoresis (PFGE)
133
Preparation of unsheared DNA for PFGE, digestion using the restriction endonucleases
134
SmaI and KpnI (Amersham Biosciences Europe GmbH) and electrophoresis were performed
135
as described by CAMPYNET [http://campynet.vetinst.dk/PFGE.html] and Hänel et al. (2007).
136
The DNA fragments produced were visualized by ethidium bromide staining.
137
Accepted Manuscript
A lambda ladder (size range from 48.5 to 727.5 kb, New England Biolabs) used to determine
138
fragment size was run on three lanes (both edges and middle) of each gel.
139 140
2.7. Motility assay
141
Motility phenotypes of strains were tested in Brucella media containing 0.5 % agar. Bacterial
142
cells were harvested from a 24 h culture on MH agar plates incubated at 37 oC under
143
microaerophilic conditions into PBS to obtain an optical density at 600 nm of 0.45
144
(approximately 1 x 109cfu/ml). 1 µl of a bacterial suspension of approximately 1 x 108cfu/ml
145
was stabbed into motility agar. Plates were incubated at 37 oC or 42 oC under microaerophilic
146
conditions and the diameter of the resulting swarms was measured after 24 h. Isolates from
147
chickens were scored against the motility phenotype of the strain used for inoculation of the
148
chickens. The results were the mean of at least six separate determinations of three
149
experiments. Significance of differences between samples was determined with t test and a
150
Pvalue of < 0.05 was considered significant.
151 152
2.8. Electron microscopy
153
For negative staining, 24 h broth-grown cultures of Campylobacterstrains were used. The
154
broth was gently swirled, 50 µl of the suspension collected and placed as drop on a glass
155
slide. 400-mesh copper grids that had been filmed with formvar, coated with carbon and
156
hydrophilisated by glow discharge were floated for 10 min on the suspension. Then grids
157
were removed and excess liquid drained with filter paper. Finally grids were contrasted on a
158
drop of 2% phosphotungstic acid for 10 minutes. After drying, grids were examined in a
159
transmission electron microscope (Tecnai 12, FEI, Netherlands).
160 161
3. Results
162
3.1. Colonization of C. jejuniin the chicken gut
163
The colonization potential in the chicken gut was assessed for 17 C. jejunistrains of human
164
or bovine origin with different abilities to adhere to and to invade into Caco-2 cells. None of
165
Accepted Manuscript
the control chickens was colonized with Campylobacterat the end of experiment. The
166
colonization ability of the strains in 9-day-old chickens differed in a wide range (Table 1). The
167
colonization potential was assessed by means of the following criteria: first detection after
168
inoculation, number of C. jejuniper gram of sample, duration of excretion (up to the end of
169
experiment at three weeks after inoculation), number of positive animals out of the group at
170
different investigation times, tendency to invasion of the liver. According to these criteria the
171
isolates were divided into the following three groups: (1) No colonization - the isolates could
172
not be reisolated. According to this criterion three strains (128/94, 129/94, 972 dJ) were
173
arranged in this group. (2) Weak or delayed colonization - the strains only achieved low
174
counts per gram caecal content and were intermittently excreted (not all animals positive at
175
different investigations). Two strains (292/94, 315/94) corresponded to these criteria. They
176
showed a delayed colonization and strain 292/94 could not be reisolated at the end of
177
experiment. (3) Strong colonization - these strains caused a strong colonization up to the end
178
of experiment at three weeks after inoculation. According to these characteristic features, 12
179
strains (167/96, 44/96, 170/96, 288/94, 320/94, 305/94, 158/96, 277/94, 164/96, 165/96,
180
163/96, 216/04) were assigned to this group. Already two days after inoculation, the first six
181
strains achieved more than 108cfu per gram of ceacal content. Six days after inoculation, the
182
other strains were detectable in high concentrations in this organ and all strains remained
183
detectable in these high concentrations up to the end of experiment. Although some C. jejuni
184
could be reisolated from the liver, the tendency to invade the liver showed no clear
185
correlation to the concentration of the pathogen in caeca.
186 187
3.2. Adhesion and invasion abilities of C. jejunistrains
188
Adherence and invasion studies were carried out with monolayers of Caco-2 cells. All C.
189
jejunistrains examined in the chicken colonization model adhered to Caco-2-cells at a range
190
of 0.15 - 2.03 % of the starting viable inoculum (Table 1). The invasiveness of
191
Campylobacterstrains into Caco-2 cells was found to be in a wide range of 0.0013 % - 0.75
192
% of the starting bacterial inoculum. E. colistrain K12 was used as non-invasive control
193
Accepted Manuscript
strain in the invasion assay showing 0.001-0.01 % entry into Caco-2 cells. In comparison
194
with the non-invasive control-strain we found three of the tested strains (128/94, 129/94 and
195
972dJ) which were not invasive for Caco-2 cells. Two strains (292/94 and 315/94) showed a
196
minor enhancement of entry into the cells compared to the non-invasive control strain. All
197
other Campylobacterstrains included in the study were considered to be invasive, having at
198
least 0.1 % intracellular bacteria of the inoculum which survived gentamicin treatment (Table
199
1). There was no correlation between adhesion to Caco-2 cells and the colonization groups
200
in the chicken model. In contrast a correlation was observed between invasion and
201
colonization as all good colonizers invaded Caco-2 cells whereas weak and no colonizers did
202
not invade into the cells. A measure of invasiveness is the invasion index, which is the
203
number of invaded bacteria as a percentage of the number of adhered bacteria. Hence, the
204
invasion index reflects the organism’s ability to adhere to a cell and how likely it is that that
205
interaction will lead to internalization. It is a useful parameter comparing invasiveness
206
between similar pathogens (Elsinghorst, 1994, Hänel et al., 2004). The invasion index
207
formed the basis for the differentiation of the strains into three groups. Strains which were
208
characterized by an invasion index below 1 % (128/94, 129/94 and 972dJ) could not be
209
reisolated from the chicks. Two strains (292/94 and 315/94) causing a weak or delayed
210
colonization were characterized by an invasion index about ten (9.7 % and 10.6 %
211
respectively). For most of the strains we determined an invasion index over 20 %. These
212
strains showed a strong, 21 days lasting colonization in chickens.
213
Adhesion and invasion differences between strains used for inoculation of chickens and the
214
isolates from the chickens were not seen except for strain C. jejuni305/94. The results for
215
this strain, shown in Fig. 1a, indicated that the isolates from chickens adhered to Caco-2
216
cells at significantly (P< 0.05) higher levels compared to the strain used for inoculation. The
217
effect on invasion of Caco-2 cells was more modest and not significant. The invasion index
218
(Fig.1b), was decreased (below 10 %) in comparison to the inoculation strain (31.5 %).
219 220
3.3. Genotyping
221
Accepted Manuscript
Described virulence-associated genes (flaA, flaB, flhA, flhB, flgB, flgE2, fliM, fliY, ciaB, iamA,
222
virB11, cadF, docA-C, cdtA-C, secD, secF, Cj1470c, Cj1471c, and Cj1474c) were
223
investigated by PCR in C. jejunistrains used for inoculation and in the reisolated strains of 4
224
chickens (chicken 17 – chicken 20) 21 days after inoculation. Except of differences in flaA-
225
types we could not identify any differences of the virulence-associated genes between the
226
strain used for inoculation of chickens and reisolated strains of each isolate. Additionally,
227
there were no differences in the detection of the selected genes among the three
228
colonization and invasion groups of isolates.
229
Plate sweeps taken from ceacal isolates recovered at 21 days post inoculation showed
230
different flaA-types for six of 17 tested isolates (158/96, 163/96,164/96, 165/96, 216/04 and
231
305/94) of at least for one chicken in comparison to the inoculation strain (Fig. 2). In the
232
cases of strains 163/96 and 165/96 the isolates of all four chickens had a different flaA-type
233
compared to the inoculum strain but the same type among themselves. Plate sweeps of one
234
chicken for strains 158/96 and 305/94 and plate sweeps of two chickens for strains 164/96
235
and 216/04 showed different flaA-types in comparison to the inoculation strain (Fig. 2). In the
236
case of 216/04 the two different reisolated strains (216/04-chicken 17 and 216/04-chicken
237
20) had not the same flaA-type. Furthermore, we were not able to detect a flaA PCR product
238
in strains 158/96 and 165/96 after reisolation from two chickens.
239
The genetic variability of Campylobacterstrains throughout a passage through chicken
240
intestine was further investigated by means of PFGE. The genotypes of 14 strains remained
241
unchanged after passage through chicken intestines and showed identical patterns by PFGE
242
using SmaI and KpnI as restriction enzymes. For three strains (315/94, 288/94, 305/94)
243
changed PFGE types were detected in individual chickens indicating that DNA
244
rearrangement had occurred in these strains (Fig. 3). The inoculum of strain 315/94 and
245
three isolates from chickens showed identical patterns for SmaI, while for the fourth isolate
246
(chicken 18) one band at 425.0 kb was missing. For this isolate with novel SmaI PFGE type
247
also a distinctKpnI PFGE genotype was obtained. Using the restriction enzyme SmaI, two
248
isolates from strain 288/94 showed the same patterns as the inoculum, while the isolates of
249
Accepted Manuscript
chicken 18 and 20 differed in their SmaI profiles. However, all isolates showed four bands in
250
the range between 97.0 and 194.0 kb. UsingKpnI as restriction enzyme, all isolates
251
exhibited identical patterns. On the other hand, strain 305/94 showed identical PFGE pattern
252
for inoculum and the corresponding chicken isolates usingSmaI, but differences in the
253
fingerprints for KpnI for one chicken (305/94-chicken 20).FlaA-typing of these strains with
254
changed PFGE types in individual chickens showed no differences in comparison to the
255
strains inoculated, demonstrating that observed PFGE changes did not involve this known
256
variable region. However, strains that varied in flaA-types had constant PFGE patterns.
257 258
3.4. Motility assays
259
The motility phenotypes ofCampylobacterstrains before and after passage through the
260
chicken were determined for strains showing alterations in the PFGE pattern or flaA-type.
261
Despite absence of genomic changes for strains 305/94-chicken 17 and 19, all isolates from
262
chickens infected withCampylobacterstrain 305/94 displayed significantly (P < 0.05)
263
reduced motility in Brucella motility agar comparing to the strain used for inoculation (Fig. 4).
264
Interestingly, the increase of incubation temperature up to 42 oC resulted in an increase of
265
motility only of the inoculum strain. The analysis of growth zones for strains 158/96, 163/96,
266
164/96, 165/96, 216/04, 288/94 and 315/94 as well as the corresponding chicken isolates
267
showed no differences between the strain used for inoculation and the corresponding
268
isolates at 37oC and an enhancement of motility at 42 oC for all strains.
269 270
3.5. Morphology
271
Due to the decrease of motility of the chicken isolates from strain 305/94 we assumed a
272
possible alteration of the flagella. Strain C. jejuni305/94 used for inoculation into chickens
273
and the isolates had one long flagella at each end (Fig. 5) and the isolates were comparable
274
in diameter and length. Otherwise the original strain C. jejuni305/94 had the expected spiral
275
shape (Fig. 5a, b), but surprisingly all isolates from the chickens presented as straight rods
276
(Fig. 5 c, d).
277
Accepted Manuscript
278
4. Discussion
279
Results from the present study demonstrated that colonization of nine days old chickens with
280
C. jejunivaries according to the Campylobacterstrain used. Furthermore, we confirmed
281
results of previous investigations (Hänel et al., 2004) suggesting a relationship between
282
invasion of C. jejuniin Caco-2 cells and their colonization ability in the chicken gut. Three
283
colonization phenotypes were identified ranging from non-colonizing strains to strains
284
producing a strong and sustained colonization. Only Campylobacterstrains with strong ability
285
to colonize the gut of chickens were found to be invasive in the Caco-2 cell culture model.
286
Such differences in colonization of the chick gut may be related to differences in the source
287
of isolation (Ringoir and Korolik, 2003). The strains in our study were selected from a
288
collection of isolates of human and bovine origin. Campylobacterstrains previously isolated
289
from faeces of humans suffering from gastroenteritis were found in the group of non-
290
colonizers as well as in the group of isolates successful colonizing chickens (Table 1). Others
291
(Korolik et al., 1998, Ringoir and Korolik, 2003) showed that strains isolated from patients
292
were less successful in colonizing chickens in comparison with C. jejunistrains isolated from
293
chicken faeces. Our results support the opinion that other reservoirs as poultry could be
294
important sources of human Campylobacterinfections. Up to now there are only few studies
295
of the possible importance of non-poultry sources of human C. jejuniinfections (Siemer,
296
2004, Wilson et al. 2008). As shown, bovine Campylobacterisolates can also effective
297
colonize the chicken intestine. The importance of cattle as reservoir and potential source of
298
Campylobacterinfections in poultry production has not been investigated in detail. Our
299
results suggest that bovine Campylobacterstrains may be a source of poultry colonization.
300
Campylobacterstrains are considered to have a high frequency of intragenomic
301
recombination and other events that can change the pheno- and genotypes (Hänninen et al.,
302
1999, Wassenaar et al., 1998, Ridley et al., 2008). Genomic instability has been
303
demonstrated to affect both flaAtyping (Harrington et al., 1997) and PFGE (Wassenaar et
304
al., 1998, Hänninen et al., 1999). The occurrence of genomic instabilities suggests that there
305
Accepted Manuscript
are, as yet unknown, selective benefits to such novel subpopulatins (Ridley et al., 2008). Up
306
to now, phenotypic changes followed by genomic alterations were not investigated. In the
307
present study the genomic and phenotypic stability of C. jejunistrains colonizing the chicken
308
gut were studied. We were able to show that novel PFGE genotypes and flaA-types can be
309
formed and phenotypic changes occur during a three weeks lasting passage through the
310
chicken gut. We also determined a constant bacterial genetic fingerprint by PFGE for 14 and
311
by flaA-typing for 11 of 17 Campylobacterstrains at the beginning and the end of
312
colonization in chickens. The changes in PFGE profiles seem to be caused by various
313
mechanisms, as we determined enzyme specific changes and changed PFGE patterns for
314
both enzymes tested. Only strain 315/94 showing a delayed colonization was found to
315
change the PFGE type for two restriction enzymes in one chicken (C-18) indicating a large
316
genome rearrangement. These genomic changes were not linked with the determined flaA-
317
type or changes in adherence, invasion or motility. Any other observed changes in the PFGE
318
patterns for the various chicken isolates have never correlated with alterations in phenotypic
319
properties like motility, adherence or invasion and did not affect the flaA-gene as indicated by
320
determination of a constantflaA-type of inoculation strains and chicken isolates throughout.
321
FlaA-type changes detected in one or two of four identical infected chickens with four strains
322
(158/96, 164/96, 216/04, 305/94), suggesting randomly distributed events of nucleotide
323
polymorphism, could possibly induced by the chicken gut environment. The identical change
324
of the flaA-type in isolates from all four chickens (163/96) or from three chickens (165/96)
325
compared to the corresponding inoculation strains could indicate a specific influence of the
326
chicken passage. The detected variations in flaA-types did not cause changes in adherence,
327
invasion or motility of the reisolated strains.
328
Phenotypic changes of the strains in comparison to the respective chicken isolates could be
329
observed only with the invasive strain C. jejuni305/94 where all isolates from the chickens
330
exhibit dramatic differences in morphology, motility and their adhesion ability. The isolates
331
from all four chickens adhered to Caco-2 cells at significantly higher numbers than the
332
inoculation strain. We determined an invasion index below 10 %, showing that the increase
333
Accepted Manuscript
of binding to the epithelial cells was not associated with a subsequent increase of
334
internalization. These results indicated that internalization of adherent bacteria is impaired.
335
While C. jejuni305/94 showed the expected spiral shape with two flagella, one at each end,
336
the isolates from the chickens lost this shape and appeared as a straight rod with two polar
337
flagella. The reisolated strains were significantly impaired in their motility. The results of
338
genotyping analysis of strain 305/94 indicated that no causal relationship exists between the
339
determined genomic changes for several chicken isolates and the uniform changes of
340
phenotypic properties for all isolates after chicken passage. It is possible that the phenotypic
341
differences caused by other more subtle genetic changes detectable only by whole-genome
342
approaches. Gaynor et al. (2004) reported that the genome-sequenced variant of C. jejuni
343
NCTC 11168 and the original clinical isolate from which it was derived exhibit dramatic
344
differences in numerous virulence-associated phenotypes, including colonization, invasion,
345
translocation, and motility. Despite these differences, these clonally derived strains appear
346
indistinguishable by molecular genotyping techniques like AFLP and PFGE. Using
347
sequencing they were able to show that genome-sequenced variant of C. jejuniNCTC 11168
348
contains very subtle genetic changes, which clearly engender marked differences in
349
transcription and phenotype.
350
The flagella of Campylobacterare essential for motility and virulence. Recombinations,
351
deletions and duplications between flagellin gene sequences of Campylobacterhave been
352
demonstrated in chickens (Nuijten et al., 2000). They found that during colonization of the
353
chicken intestine a nonmotile flaAmutant of C. jejuniunderwent rearrangements within its
354
flagellin locus, thereby regaining its motility and colonization capacity. In contrast to these
355
findings we identified loss of motility of the strain 305/94 during colonization of the chicken
356
gut. The most striking finding in this connection was the change in morphology of this strain
357
from a spiral shape to a straight rod without loss of the flagella. Because in addition to the
358
flagella the spiral morphology of Campylobacteris one important factor in the organism’s
359
ability to move (Shigematsu et al., 1998), the morphological change to a straight rod is likely
360
the reason for the observed reduction of motility. Furthermore, some reports describe
361
Accepted Manuscript
flagella as adhesins while others describe flagella as factor for Campylobacterpenetration of
362
host epithelial cells (Grant et al., 1993, Guerry, 2007). These functions of flagella may be
363
linked with the determined changes in in vitroadhesion and invasion properties of the
364
chicken isolates of strain 305/94. Further investigations are required to elucidate how and
365
why the observed phenomenon occurs.
366
Genetic changes, resulting in small differences in restriction patterns were demonstrated
367
during the course of colonization of chickens (Hänninen et al., 1999, Wassenaar et al.,
368
1998), of mice (Mixter et al., 2003) and also during the course of a human infection
369
(Steinbrückner et al., 2001). Otherwise the genetic stability of four C. jejunistrains after
370
gastrointestinal passage through experimentally colonized chickens after seven days was
371
shown using LPS genes (LG) genotyping (Knudson et al., 2005). Nielsen et al. (2001) could
372
show that many Campylobacterstrains are indeed genetically stable during in vitroand mice
373
passage according to a spectrum of typing methods. Our study confirms these results
374
showing that during intestinal colonization in the chicken gut behind apparently genomic
375
stable C. jejunistrains genomic rearrangement may occur. These genomic instabilities were
376
dependent on the individual bird, suggesting the presence of host-specific factors. However,
377
the observed genetic changes were not followed by examined phenotypic changes as in vitro
378
adhesion or invasion. Our data suggest that although there is increasing evidence that
379
genetic recombination can occur in Campylobacterduringin vivopassage, the subtype
380
pattern variations, leading to phenotypic changes, were not common events and occur only
381
occasionally.
382 383 384
Acknowledgements
385
386
We thank Peggy Methner, Waltraud Wilhelm, Susann Bahrmann, Byrgit Hofmann, Christine
387
Muselmann, Sabine Lied and Wolfram Maginot for excellent technical assistance.
388
389
Accepted Manuscript
References
390
Bras, A.M., Chatterjee, S., Wren, B.W., Newell, D.G., Ketley, J.M.,1999. A novel
391
Campylobacter jejunitwo-component regulatory system important for temperature-
392
dependent growth and colonization. J. Bacteriol. 181, 3298-3302.
393
Carrillo, C.D., Taboada, E., Nash, J.H.E., Lanthier, P., Kelly, J., Lau, P.C., Verhulp, R.,
394
Mykytczuk,.O., Sy, J., Findlay, W.A., Amoako, K., Gomis, S., Willson, P., Austin, J.W., Potter,
395
A., Babiuk, L., Allan, B., Szymanski, C.M., 2004. Genome-wide expression analyses of
396
Campylobacter jejuniNCTC 11168 reveals coordinate regulation of motility and virulence by
397
flhA. J. Biol. Chem. 279, 20327-20338.
398
Cawthraw, S.A., Wassenaar, T.M., Ayling, R., Newell, D.G., 1996. Increased colonisation
399
potential of Campylobacter jejunistrain 81116 after passage through chickens and its
400
implication on the rate of transmission within flocks. Epidemiol. Infect. 117, 213-215.
401
de Boer, P., Wagenaar, J.A., Achterberg, R.P., van Putten, J.P.M., Schouls, L.M., Duim, B.,
402
2002. Generation of Campylobacter jejunigenetic diversity in vivo. Mol. Microbiol. 44, 351-
403
404
359.Elsinghorst, E.A., 1994. Measurement of invasion by gentamicin resistance. Meth. in
405
Enzymology 236, Academic Press. Inc., pp. 405-420.
406
Gaynor, E.C., Cawthraw, S., Manning, G., MacKichan, J.K., Falkow, S., and Newell, D.G.,
407
2004. The genome-sequenced variant of Campylobacter jejuniNCTC 11168 and the original
408
clonal clinical isolate differ markedly in colonization, gene expression, and virulence-
409
associated phenotypes. Journal of Bacteriology, 186, 503-517.
410
Grant, C.C.R., Konkel, M.E., Cieplak, JR., W., Tompkins, L.S., 1993. Role of flagella in
411
adherence, internalisation and translocation of Campylobacter jejuniin nonpolarized and
412
polarized epithelial cell cultures. Infect. Immun. 61, 1764-1771.
413
Guerry, P. 2007. Campylobacter flagella: not just for motility. TRENDS in Microbiol. 15, 10,
414
456-461.
415
Accepted Manuscript
Hänel, I., Müller, J., Müller, W., Schulze, F., 2004. Correlation between invasion of Caco-2
416
eucaryotic cells and colonisation ability in the chick gut in Campylobacter jejuni. Vet.
417
Microbiol. 101, 75-82.
418
Hänel, I., Borrmann, E., Müller, J., Alter, T., 2007. Relationships between bacterial genotypes
419
and in vitrovirulence properties of Campylobacter jejuniand Campylobacter coliisolated
420
from turkeys. J. Appl. Microbiol. 102, 433-441.
421
Hänninen, M.-L., Hakkinen, M. and Rautelin, H., 1999. Stability of related human and chicken
422
Campylobacter jejunigenotypes after passage through chick intestine studied by pulsed-field
423
gel electrophoresis. Appl. Environm. Microbiol. 65, 2272-2275.
424
Harrington, C.S., Thomson-Carter, F.M., Carter, P.E., 1997. Evidence for recombination in
425
the flagellin locus of Campylobacter jejuni: Implications for the flagellin gene typing scheme.
426
J. Clin. Microbiol. 35, 2386-2392.
427
Hendrixson, D.R., DiRita, V.J., 2004. Identification of Campylobacter jejunigenes involved in
428
commensal colonization of the chick gastrointestinal tract. Mol. Microbiol. 52, 471-484.
429
Knudsen, K.N., Bang, D.D., Nielsen, E. M., Madsen, M., 2005. Genotyping of Campylobacter
430
jejunistrains from Danish broiler chickens by restriction fragment length polymorphism of the
431
LPS gene cluster. J. Appl. Microbiol. 99, 392-399.
432
Korolik, V., Alderton, M.R., Smith, S.C., Chang, J., Coloe, P.J., 1998. Isolation and molecular
433
analysis of colonising and non-colonising strains of Campylobacter jejuniand Campylobacter
434
colifollowing experimental infection of young chickens. Vet. Microbiol. 60, 239-249.
435
MacKichan, J.K., Gaynor, E.C., Chang, C., Cawthraw, S., Newell, D.G., Miller, J.F., Falkow,
436
S., 2004. The Campylobacter jejuni dccRStwo-component system is required for optimal in
437
vivocolonization but is dispensable for in vitrogrowth. Mol. Microbiol. 54, 1269-1286.
438
Mixter, P.F., Klena, J.D., Flom, G.A., Siegesmund, A.M., Konkel, M.E. 2003. In vivo tracking
439
of Campylobacter jejuniby using a novel recombinant expressing green fluorescent protein.
440
Appl. Environm. Microbiol. 69, 2864-2874.
441
Accepted Manuscript
Müller, J., Schulze, F., Müller, W., Hänel. I. 2006. PCR detection of virulence-associated
442
genes in Campylobacter jejunistrains with differential ability to invade Caco-2 cells and to
443
colonize the chick gut. Vet. Microbiol. 113, 123-129.
444
Nielsen, E. M., Engberg, J., Fussing, V., 2001. Genotypic and serotypic stability of
445
Campylobacter jejunistrains during in vitro and in vivo passage. Int. J. Med. Microbiol. 291,
446
379-385.
447
Nuijten, P.J.M., van den Berg, A.J.G., Formentini, I., van der Zeijst, B.A.M., Jacobs, A.A.C.
448
2000. DNA rearrangements in the flagellin locus of an flaA mutant of Campylobacter jejuni
449
during colonization of chicken ceca. Infect. Immun. 68, 7137-7140.
450
Ridley, A.M., Toszeghy, M.J., Cawthraw, S.A., Wassenaar, T.M., Newell, D.G. 2008. Genetic
451
instability is associated with changes in the colonization potential of Campylobacter jejuniin
452
the avian intestine. J. Appl. Microbiol. 105, 95-104.
453
Ringoir, D.D., Korolik, V., 2003. Colonisation phenotype and colonisation potential
454
differences in Campylobacter jejunistrains in chickens before and after passage in vivo. Vet.
455
Microbiol.92, 225-235.
456
Shigematsu, M., Umeda, A., Fujimoto, S., Amako, K., 1998. Spirochaete-like swimming
457
mode of Campylobacter jejuniin a viscous environment. J. Med. Microbiol. 47, 521-526.
458
Siemer, B.L., Harrington, C.S., Nielsen, E.M., Borck, B., Nielsen, N.L., Engberg, J., On,
459
S.L.W., 2004. Genetic relatedness among Campylobacter jejuniserotyped isolates of diverse
460
origin as determined by numerical analysis of amplified fragment length polymorphism
461
(AFLP) profiles. J. Appl. Microbiol. 96, 795-802.
462
Smith, S.I., Olukoya, D.K., Fox, A.J., Coker, A.O., 1999. Flagellin gene polymorphism
463
analysis of Campylobacter compared with antigen serotyping. Z. Naturforsch. 54c, 946-951.
464
Steinbrueckner, B., Ruberg, F., Kist, M., 2001. Bacterial genetic fingerprint: a reliable factor
465
in the study of the epidemiology of human Campylobacterenteritis? J. Clinical Microbiol. 39,
466
4155-4159.
467
Wassenaar, T.M., Geilhausen, B., Newell, D. G., 1998. Evidence of genomic instability in
468
Campylobacter jejuniisolated from poultry. Appl. Environm. Microbiol. 64, 1816-1821.
469
Accepted Manuscript
Wassenaar, T.M., Blaser, M.J., 1999. Pathophysiology of Campylobacter jejuniinfections of
470
humans. Microb.Infect. 1, 1-11.
471
Wilson, D.J., Gabriel, E., Leatherbarrow, A.,J.,H., Cheesbrough J., Gee, S., Bolton, E., Fox,
472
A., Fearnhead, P., Hart, C.A., Diggle, P.J., 2008. Tracing the source of campylobacteriosis.
473
Plos Genet 4, 9, e1000203.
474
Young, K.T., Davis, L.M., DiRita, V.J., 2007. Campylobacter jejuni: molecular biology and
475
pathogenesis. Microbiol. 5, 665-679.
476
Ziprin, R.L., Young, C.R., Byrd, J.A., Stanker, L.H., Hume, M.E., Gray, S.A., Kim, B.J.,
477
Konkel, M.E., 2001. Role of Campylobacter jejunipotential virulence genes in cecal
478
colonization. Avian Dis. 45, 549-557.
479
Accepted Manuscript
Figure captions
480
481
Fig. 1Adhesion and invasion of C. jejuni305/94 and the corresponding chicken isolates
482
(305/94-17 – 305/94-20), a – adhesion and invasion, b – invasion index
483
484
Fig. 2FlaA-types of C. jejunistrains used for inoculation and isolates after chicken passage;
485
St – 100 bp ladder, WT – inoculation strain, C17-C20 – corresponding chicken isolates, CS –
486
control strain
487
488
Fig. 3PFGE patterns of C. jejunistrains used for inoculation and isolates after chicken
489
passage showing changed genotypes for individual chickens; St - molecular size ladder; WT
490
– inoculation strain, C17-C20 – corresponding chicken isolates
491
492
Fig. 4Motility of strain C. jejuni305/94 used for inoculation and corresponding isolates
493
observed after 21 days of chicken passage (305/94-17 – 305/94-20; black column – motility
494
at 37 oC, grey column – motility at 42 oC
495
496
Fig. 5Morphology of C. jejuni305/94 and representative morphology of corresponding
497
Campylobacter isolates after chicken passage, a – inoculation strain characterized by a
498
spiral shape, b – one long flagella is present at each end of the spiral bacterium, c – all
499
reisolated bacteria appear as straight rods, b – one long flagella is present at each end of the
500
straight bacterium
501
502
503
Accepted Manuscript
Figure 1
Accepted Manuscript
Figure 2
Accepted Manuscript
Figure 3
Accepted Manuscript
Figure 4
Accepted Manuscript
Figure 5
Accepted Manuscript
Table 1
Colonization ability in the chicken gut, adherence to and invasion of Caco-2 cells
Colonization group
Strain Origin of isolate % of inoculum adhered
% of inoculum internalized
Invasion index
Non-colonizing strains
128/94 Human enteritis 0.5 0.0013 0.17
129/94 Human enteritis 0.67 0.004 0.67
972dJ Calf intestine 0.62 0.0025 0.46
Strains with weak or delayed
colonization
292/94 Dairy cow 0.2 0.019 9.7
315/94 Dairy cow 0.15 0.016 10.6
Strains with strong colonization
ability
277/94 Dairy cow 2.03 0.75 37.8
288/94 Dairy cow 0.94 0.2 21.1
320/94 Dairy cow 1.7 0.52 30.3
305/94 Dairy cow 0.87 0.28 31.5
158/96 Human enteritis 1.92 0.52 26.7
167/96 Human enteritis 1.76 0.4 22.8
44/96 Human enteritis 1.26 0.3 23.5
170/96 Human enteritis 1.22 0.32 27.3
164/96 Human enteritis 1.4 0.58 41.8
165/96 Human enteritis 0.56 0.19 32.9
163/96 Human enteritis 0.65 0.13 20.7
216/04 Human enteritis 0.54 0.14 26.3
Accepted Manuscript
Table 2
List of target genes, PCR primers, and annealing temperatures
Target gene/
ORF
Primers Sequence Annealing
tempera- ture (°C)
Ampli- con lenght
(bp)
Reference
secD CjsecD11 CjsecD22
5'-CCC AAA GTC ACA GCA AAA CC-3' 5'-GCA GTC AGA ACG TGG TGC GAA G-3'
55°C 1371 This study
secF CjsecF11 CjsecF22
5'-TCA TAC ATT GCA CGA TTG CG-3' 5'-ATG AGA ATG CGT TTT GCT GC-3'
50°C 920 This study
Cj1470c Cj1470cF1 Cj1470cF2
5'-CCT GAG CTA AGC TCC CAC ATA G-3' 5'-CAG CAC AAG CAA AAG CCT TAT C-3'
50°C 1089 This study
Cj1471c Cj1471cE1 Cj1471cE2
5'-GAT GAC AAC GCT CAC ATC CT-3' 5'-GAC GGC TTT ATG ATT ATT AG-3'
50°C 1129 This study
Cj1474c Cj1474cD1 Cj1474cD2
5'-GGA GCA TTT ACG CTA TCT AC-3' 5'-GCC ATT ATC TTT ATG CTT TA-3'
45°C 1342 This study