Accepted Manuscript Campylobacter jejuni capsular genotypes are related to Guillain-Barré syndrome Astrid P. Heikema, Zhahirul Islam, Deborah Horst-Kreft, Ruth Huizinga, Bart C. Jacobs, Jaap A. Wagenaar, Frédéric Poly, Patricia Guerry, Alex van Belkum, Craig T. Parker, Hubert P. Endtz PII:
S1198-743X(15)00563-7
DOI:
10.1016/j.cmi.2015.05.031
Reference:
CMI 290
To appear in:
Clinical Microbiology and Infection
Received Date: 5 November 2014 Revised Date:
5 February 2015
Accepted Date: 22 May 2015
Please cite this article as: Heikema AP, Islam Z, Horst-Kreft D, Huizinga R, Jacobs BC, Wagenaar JA, Poly F, Guerry P, van Belkum A, Parker CT, Endtz HP, Campylobacter jejuni capsular genotypes are related to Guillain-Barré syndrome, Clinical Microbiology and Infection (2015), doi: 10.1016/ j.cmi.2015.05.031. 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 1
Campylobacter jejuni capsular genotypes are related to Guillain-Barré syndrome
2 Astrid P. Heikema1*‡, Zhahirul Islam2‡, Deborah Horst-Kreft1, Ruth Huizinga3, Bart C.
4
Jacobs3,4, Jaap A. Wagenaar5,6, Frédéric Poly7, Patricia Guerry7, Alex van Belkum8, Craig T.
5
Parker9 and Hubert P. Endtz1, 2, 10
6
1
7
Medical Centre Rotterdam, The Netherlands; 2International Centre for Diarrhoeal Disease
8
Research Bangladesh, Dhaka, Bangladesh; 3Department of Immunology, Erasmus MC,
9
University Medical Centre Rotterdam, The Netherlands; Department of Neurology, Erasmus
RI PT
3
SC
Department of Medical Microbiology and Infectious Diseases, Erasmus MC, University
MC, University Medical Centre Rotterdam, The Netherlands; 5Department of Infectious
11
Diseases and Immunology, Faculty of Veterinary Medicine, Utrecht University, The
12
Netherlands; 6 Central Veterinary Institute of Wageningen UR, Lelystad, The Netherlands;
13
7
14
Microbiology, La Balme Les Grottes, France; 9Produce Safety and Microbiology Research
15
Unit, Agricultural Research Service, United States Department of Agriculture, Albany
16
California, USA; 10Fondation Mérieux, Lyon, France.
17
‡ These authors contributed equally to this work
M AN U
10
EP
TE D
Naval Medical Research Center, Silver Spring, Maryland, USA; 8bioMérieux, R&D
18
Running title: LOS classes and capsular types are correlated in GBS
AC C
19 20 21
Keywords: Campylobacter jejuni, capsular genotype, lipooligosaccharide, Guillain-Barré
22
syndrome, MLST
23 24 25
*
Corresponding author:
1
ACCEPTED MANUSCRIPT Astrid P. Heikema
2
3015 CE Rotterdam, The Netherlands
3
Erasmus MC, University Medical Centre Rotterdam
4
Department of Medical Microbiology and Infectious Diseases
5
‘s Gravendijkwal 230
6
3015 CE, Rotterdam, the Netherlands
7
Phone: +31-10-7032176
8
Fax: +31-10-4703875
9
Email: [email protected]
SC
AC C
EP
TE D
M AN U
10
RI PT
1
2
ACCEPTED MANUSCRIPT Abstract
2
In about one in thousand cases, a Campylobacter jejuni infection results in the severe
3
polyneuropathy Guillain-Barré syndrome (GBS). It is established that sialylated
4
lipooligosaccharides (LOS) of C. jejuni are a crucial virulence factor in GBS development.
5
Frequent detection of C. jejuni with sialylated LOS in stools derived from patient with
6
uncomplicated enteritis implies that additional bacterial factors should be involved. To assess
7
whether the polysaccharide capsule is a marker for GBS, the capsular genotype of two
8
geographical distinct GBS-associated C. jejuni strains collections, and an uncomplicated
9
enteritis control collection was determined. Capsule genotyping of C. jejuni strains from the
SC
RI PT
1
Netherlands revealed that three capsule types, HS1/44c, HS2 and HS4c, were dominant in
11
GBS-associated strains and capsular types HS1/44c and HS4c were significantly associated
12
with GBS (P = 0.05 and P = 0.01 respectively) when compared to uncomplicated enteritis. In
13
a GBS-associated strain collection from Bangladesh, capsular types HS23/36c, HS19, and
14
HS41 were most prevalent and the capsular types HS19 and HS41 associated with GBS (P =
15
0.008 and P = 0.02 respectively). Next, specific combinations of the LOS class and capsular
16
genotypes were identified that were related to the occurrence of GBS. Multilocus sequence
17
tying revealed restricted genetic diversity for strains populations with the capsular types HS2,
18
HS19 and HS41. We conclude that capsular types HS1/44c, HS2, HS4c, HS19, HS23/36c and
19
HS41 are markers for GBS. Besides a crucial role for sialylated LOS of C. jejuni in GBS
20
pathogenesis, the identified capsules may contribute to GBS susceptibility.
AC C
EP
TE D
M AN U
10
3
ACCEPTED MANUSCRIPT Introduction
2
Campylobacter jejuni is one of the most frequent causes of bacterial gastroenteritis in humans
3
worldwide [1, 2]. In general, C. jejuni diarrhea is self-limiting though a wide spectrum of
4
serious complications may occur, including bacteremia, reactive arthritis and peripheral
5
neuropathies e.g. Guillain-Barré syndrome (GBS) and Miller Fisher syndrome (MFS) [3].
6
GBS is a post-infectious, immune-mediated disease of the peripheral nerves, which is
7
characterized by a rapidly progressive weakness of the limb muscles [4]. About one-quarter of
8
patients develop additional respiratory failure and autonomic dysfunction and 2-10% of the
9
patients die [5, 6]. MFS is regarded a variant of GBS in which the weakness is restricted to
SC
RI PT
1
the oculomotor muscles [7].
11
Various studies have investigated which risk factors of C. jejuni predispose to the
12
development of GBS [8, 9]. Currently, sialylation of lipooligosaccharides (LOS) of C. jejuni
13
is the best characterized virulence factor associated with GBS [10, 11]. The structural
14
similarity (molecular mimicry) between sialylated C. jejuni LOS and ganglioside structures
15
present on human peripheral nerves drives a cross-reactive immune response resulting in
16
immune-mediated nerve damage [10]. In C. jejuni, genes involved in LOS sialylation are
17
located in the LOS biosynthesis locus. Variability in gene content in C. jejuni LOS loci has
18
led to the assignment of several LOS locus classes (A through S) [12]. Genes involved in the
19
synthesis of sialylated LOS and ganglioside mimics are present in the LOS classes A, B, C, M
20
and R [12]. Of these, C. jejuni strains with LOS classes A, B and C are frequently isolated
21
from stools of patients with GBS and the LOS classes A and B are associated with GBS and
22
MFS respectively [11].
23
The presence of sialic acid in the LOS, however, does not appear to be the only bacterial
24
factor involved in the pathogenesis of GBS. In patients with uncomplicated enteritis, the
25
prevalence of sialylated C. jejuni strains has been reported to be up to 63%[11]. We
AC C
EP
TE D
M AN U
10
4
ACCEPTED MANUSCRIPT hypothesize that the expression of additional C. jejuni virulence factors in combination with
2
the presence of sialylated LOS contributes to the induction of GBS. Factors that increase the
3
virulence of C. jejuni include the bipolar flagella, surface and secreted proteins involved in
4
the invasion of the intestinal epithelium [13], and cytolethal distending toxin (CDT) [14].
5
Genes that encode these virulence factors, however, are generally conserved and present in
6
strains derived from both clinical and environmental sources [15, 16].
7
The polysaccharide capsule is a virulence factor that is associated with cell invasion [17],
8
complement resistance [18] and immune modulation [19]. The capsule biosynthesis locus is a
9
highly variable region in the genome of C. jejuni. Diversity in gene content and gene function,
SC
RI PT
1
combined with the presence of phase variable genes in the capsule locus allows the
11
production of a broad repertoire of capsular structures [20]. This feature is the basis of the
12
classical Penner serotyping scheme which divides C. jejuni into 47 serotypes [21]. Previous
13
studies based on Penner serotyping showed that the capsular serotypes HS:19, HS:23/36c and
14
HS:41 are associated with GBS in Japan, Bangladesh and South Africa, respectively [22-24].
15
In contrast, no particular association between the capsule serotype and neuro-pathogenicity
16
was observed for GBS-associated strains from Europe or North America [8, 25]. This
17
discrepancy may be explained by the observation that GBS-associated C. jejuni strains in
18
Japan, Bangladesh and South Africa had a genetically related background whereas the strains
19
in Europe and the USA were more heterogeneous.
20
Recently, capsular genotyping was introduced as a fast, readily available and reliable method
21
to assess the capsular type in C. jejuni [26]. Compared to capsule based serotyping, capsular
22
genotyping is not sensitive to variations in capsule gene expression or influenced by genes or
23
gene products outside the capsule locus. With the availability of two large and geographically
24
distinct GBS-associated strain collections, we determined whether variation in C. jejuni
25
capsule genotype is involved in the risk to develop GBS. The capsule genotype defined by
AC C
EP
TE D
M AN U
10
5
ACCEPTED MANUSCRIPT multiplex PCR was determined in collections of C. jejuni strains isolated from Dutch
2
GBS/MFS patients, Dutch enteritis control patients and Bangladeshi GBS/MFS patients. The
3
capsular type was related to the LOS class in order to determine whether particular LOS
4
class/capsular type combinations were more prevalent amongst GBS/MFS-associated strains
5
compared to enteritis-associates strains. In addition, multilocus sequence typing (MLST) was
6
performed to assess the genetic relation between strains.
RI PT
1
7 Materials and methods
9
Strain collection, culture conditions and DNA isolation
SC
8
In this study, a Dutch- and a Bangladeshi GBS-associated C. jejuni strain collections and a
11
Dutch enteritis-associated control strain collections were used. The Dutch GBS collection
12
comprises 32 GBS- and 5 MFS-associated strains which were isolated from stools of
13
primarily Dutch patients. Among these are three GBS-associated strains from patients from
14
the Netherlands Antilles and one from Belgium. The GBS-associated strain collection from
15
Bangladesh consists of 30 strains.The Dutch enteritis-associated strains were isolated from
16
patients with diarrhoea in The Netherlands and were from two different provenances. The
17
strains in the first enteritis collection (n = 21) were isolated between 1990 and 1999 and have
18
been previously described [11]. A second enteritis collection originates from a Dutch study
19
performed by the Rijksinstituut voor Volksgezondheid en Milieu in 2002-2003, on behalf of a
20
national Campylobacter and Salmonella surveillance study [27, 28]. From this large
21
collection, we selected 122 enteritis strains that matched to our GBS/MFS-associated strains
22
based on age (decade) and sex. Thus, we selected 3-4 age and sex matched enteritis control
23
strains per GBS/MFS case. There were no significant differences in the distribution of the
24
various capsule types between the two enteritis collections. Therefore, the two enteritis strain
25
collections were combined in this study and referred to as the uncomplicated enteritis-
AC C
EP
TE D
M AN U
10
6
ACCEPTED MANUSCRIPT associated strain collection (n = 143). For the Dutch GBS-, MFS- and enteritis-associated
2
strains, LOS class typing (A, B, C, D and E) was performed previously [11] (Heikema et al.,
3
manuscript in preparation). The LOS class of 10 of the GBS-associated C. jejuni strains
4
derived from Bangladesh was determined previously [24]. For the remaining 20 strains, the
5
LOS class was newly assessed. C. jejuni strains were cultured from -80°C stocks and
6
maintained on Colombia blood agar (BA) plates (Becton Dickinson BV, Alphen aan den Rijn,
7
The Netherlands) supplemented with 10 µg/ml vancomycin in a micro-aerobic atmosphere at
8
37°C. DNA was isolated from overnight grown C. jejuni, using a DNeasy blood & tissue kit
9
(Qiagen, Venlo, The Netherlands) as described by the manufacturer’s protocol.
SC
RI PT
1
M AN U
10 LOS locus class PCR
12
Genotyping was performed for LOS classes A, B, C, D and E using standard PCR. Primers
13
used are based on the DNA sequence of unique genes or regions of the LOS locus classes
14
involved and were previously described[11].
15
TE D
11
Multiplex PCR-based capsule analysis
17
Capsule genotyping was performed as described [26]. In short, based on the sequence of 17
18
CPS loci, unique regions were identified and specific primers pairs that recognize each
19
serotype were designed. The primer pairs were divided in two mixes in such a way that the
20
expected amplicons differed at least 20 bp from one another. Standard PCR was performed
21
and the PCR products were separated on 2% agarose gels using a 50-bp DNA ladder
22
(Fermentas, Landsmeer, The Netherland) to assess the size of each amplicon.
AC C
EP
16
23 24
Multi Locus Sequence Typing (MLST)
7
ACCEPTED MANUSCRIPT 1
MLST, which relies on the partial sequencing of seven housekeeping genes, was performed as
2
described [8]. A dendrogram was constructed using BioNumerics software. Pairwise
3
alignment of the sequences of the separate MLST alleles was performed. The dendrogram
4
was generated using UPGMA clustering.
RI PT
5 Statistical analysis
7
Fisher exact and Chi square tests were performed using Prism software (GraphPad, La Jolla,
8
CA, USA). Two-sided P ≤ 0.05 was considered to be statistically significant.
SC
6
AC C
EP
TE D
M AN U
9
8
ACCEPTED MANUSCRIPT Results
2
Validation of the capsular genotyping method
3
The capsular genotype determined by PCR was compared to the serotype for our C. jejuni
4
strains with known Penner serotype (n = 40). The serotype did not match the genotype in six
5
cases and in one case the serotype was not included in the multiplex capsule PCR (data not
6
shown). Full genome sequence data were available for 4/6 strains with a mismatch between
7
capsular serotype and genotype. Blast analysis of the capsule locus region revealed homology
8
to strains with serotypes that matched the identified genotype and not the serotype in all four
9
cases (data not shown). Overall, we found a specificity of 95% and a sensitivity of 100% for
SC
RI PT
1
strains (n = 39) with serotypes covered in the genotyping method. Therefore, we concluded
11
that the genotyping method is a reliable method to determine the capsular type in C. jejuni.
M AN U
10
12
Six capsular genotypes are dominant in GBS/MFS-associated C. jejuni strains
14
In the Dutch GBS/MFS-associated strains, a genocapsular type could be assigned for all
15
strains whereas in the Dutch enteritis-associated strains a capsular genotype could be assigned
16
for 120/143 (83%) of the strains.. Twenty-two strains in the enteritis group failed to produce
17
amplicons and in one case a mix of two capsular types (HS1/HS41) was found. In the Dutch
18
GBS/MFS collection, nine different capsular types were identified of which three capsular
19
types; HS4c, HS2 and HS1/44c were dominant (Fig. 1). In the Dutch enteritis strain
20
collection, a broader panel of 15 different capsular types was identified of which capsules
21
HS4c, HS8/17 and HS2 were most prevalent (Fig. 1). When GBS/MFS-associated C. jejuni
22
strains and enteritis-associated strains were compared, capsular type HS4c (P = 0.01; Chi
23
square test) and HS1/44c (P = 0.05; Chi square test) significantly associated with GBS/MFS
24
(Fig. 1). No association with a specific capsular type was observed for the MFS strains. In the
25
five MFS associated strains, four different capsular types were detected; HS2, HS4c (in two
AC C
EP
TE D
13
9
ACCEPTED MANUSCRIPT cases), HS10 and HS23/36c. When combined, 28/37 (76%) of the GBS/MFS-associated
2
strains from The Netherlands either had capsular type HS1/44c, HS2 or HS4c, compared to
3
50/143 (35%) in the enteritis control strains (P < 0.0001; Chi square test).
4
In order to determine whether the dominant capsular types identified in the GBS/MFS-
5
associated strain collection from the Netherlands were also prominent in GBS-associated
6
strains from another geographical area, capsule genotyping was performed on 30 GBS-
7
associated strains isolated from patients from Bangladesh. In these strains, a capsular type
8
could be assigned for 28/30 (93%) of the strains and capsular types HS23/36c, HS19 and
9
HS41 were dominant (Fig. 1). In a randomly selected uncomplicated enteritis control group
SC
RI PT
1
from Bangladesh, 0/39 (0%) of the C. jejuni strains had the capsular type HS19, 11/39 (28%)
11
had HS23/36 and in 1/39 (3%) capsular type HS41 was identified (Z. Islam, personal
12
communications). The capsular types HS19 and HS41 significantly associated with GBS in
13
Bangladesh when compared to enteritis (P = 0.008 and P = 0.02 respectively; Chi square test).
14
Capsular type HS41 was only found in strains from Bangladesh whereas capsular types
15
HS23/36c and HS19 were also observed in the Dutch strain collections. When combined,
16
56/67 (84%) of the GBS/MFS-associated C. jejuni strains either had capsular type HS1/44c,
17
HS2, HS4c, HS19, HS23/36c or HS41 (from here on referred to as GBS/MFS-related capsular
18
types).
TE D
EP
AC C
19
M AN U
10
20
Sialylated LOS classes and specific capsular genotypes are correlated
21
Upon correlation of LOS class and capsular genotype in the GBS/MFS strain collection from
22
The Netherlands, we observed that all but one strain with the LOS class A, B or C had the
23
GBS/MFS-related capsular types (Fig. 2A). A particular distribution pattern was observed
24
when LOS class and capsular type were combined. In the GBS/MFS collection, strains with
25
LOS class A were typed HS1/44c, HS2, HS4c or HS19, in strains with LOS class B, capsular
10
ACCEPTED MANUSCRIPT types HS2 and HS4c were the most prominent and LOS class C strains, mostly had capsular
2
type HS1/41 or HS2 (Fig. 2A). In the enteritis control collection, approximately the same
3
capsule distribution was observed for LOS class A strains. In contrast, for the class B and C
4
strains, similar capsular types were only found for a subset of the enteritis strains when
5
compared to the GBS/MFS strains. A distinct capsular type was found in 9/28 (32%) of the
6
LOS class B strains and 18/34; 53% of the class C strains had a capsular type that did not
7
belong to the GBS/MFS-related capsule group (P = 0.01; GBS/MFS vs enteritis; Chi square).
8
In enteritis strains with LOS class B, there was diversity in the remaining capsular types but in
9
the LOS class C strains, capsular type HS8/17 was detected in 9/18 (50%) of the strains which
SC
RI PT
1
did not belong to the GBS/MFS-related capsule group.
11
In enteritis strains with LOS class D, capsular type HS53 was dominant (9/31; 29%; HS53 is
12
not present in class D strains of the GBS/MFS collection) whereas capsular types within class
13
E strains were diverse (data not shown). It should be noted that LOS class D and E loci do not
14
possess the sialylation genes necessary to produce ganglioside mimics.
15
LOS genotyping of 20 newly isolated GBS-associated C. jejuni strains from Bangladeshi
16
patients resulted in the identification of 10 strains with LOS class A and 10 strains with LOS
17
class B. When LOS class and capsular type of the GBS-associated strains derived from
18
Bangladeshi patients (n = 30) were correlated, all strains with capsular type HS41 or HS19
19
correlated with LOS class A, whereas all the strains with the HS23/36c capsular type
20
correlated with LOS class B (Fig. 2B).
TE D
EP
AC C
21
M AN U
10
22
A genetic linkage between LOS/capsule combinations
23
MLST was performed on all (n = 67) GBS- and MFS-associated strains and on 84/143 (59%)
24
of the enteritis-associated strains. The 84 enteritis control strains on which MLST was
25
performed comprise 60% sialylated strains and 40% non-sialylated strains. These percentages
11
ACCEPTED MANUSCRIPT are representative for the whole enteritis control collection of 143 strains. A total of 76
2
different sequence types were identified which were classified into 21 clonal complexes. C.
3
jejuni clonal complex ST-21 was most prominent (29/151; 19%), followed by ST-206
4
(16/151; 11%) and ST-22 (15/151; 10%). In 13 cases, the MLST sequence type represented a
5
singleton. In the GBS collection, the clonal complexes ST-21 and ST-22 were the most
6
common (both 10/67; 15%) followed by ST-403 (9/67; 13%). A minimum spanning tree
7
(MST), in which our MLST data were combined with MLST data of C. jejuni isolates from
8
sporadic cases of human enteritis present in the Oxford C. jejuni MLST database (n = 4066),
9
showed that our GBS/MFS- associated and enteritis-associated strains were widely spread
10
within the total C. jejuni MLST population (Supplemental figure 1). GBS/MFS- as well as
11
enteritis-associated control strains were present within the same clonal complexes.
12
To assess whether the LOS class/capsular type co-occurrence within one strain was due to a
13
closely related genetic background, in particular for the strains that carry genes involved in
14
LOS sialylation, a dendrogram was generated using strains from both, the Netherlands and
15
Bangladesh, which either had LOS loci classes A, B or C (Fig. 3 ). The same genetic
16
background, ST-21, was frequently observed for strains with capsular types HS1/44c and HS2
17
and LOS class C (18/29; 62%). Most HS8/17 strains with LOS class C (6/7; 86%) also
18
belonged to ST-21. The HS4c strains with LOS class A or B were predominantly present in
19
four genetic backgrounds: ST-206, ST-48, ST-61 and ST-508 (30/35; 86%). A strong
20
individual genetic relationship was observed for LOS class A strains with capsular type HS41
21
(n = 6) and HS19 (n = 13). All HS19 strains belonged to ST-22 whereas all HS41 strains
22
belonged to ST-362. The strains with capsular type HS23/36c were more widely spread
23
within the dendrogram. Most (11/14; 78%), however, belonged to two MLST CCs, ST403
24
and ST-42 (Fig 3). One MLST clonal complex, ST-508 contained GBS-associated strains
AC C
EP
TE D
M AN U
SC
RI PT
1
12
ACCEPTED MANUSCRIPT 1
only but overall MLST typing did not discriminate strains carrying genes involved in LOS
2
sialylation when GBS/MFS- and enteritis-associated strains were compared.
3 Discussion
5
We identified a group of six main capsular types HS1/44c, HS2, HS4c, HS19, HS23/36c and
6
HS41 in two large, geographically distinct GBS/MFS-associated C. jejuni strain collections.
7
In a predominantly Dutch collection (n = 37), capsular genotypes HS1/44c, HS2 and HS4c
8
were dominant. Of these, capsular type HS4c and HS1 associated with the neuro-
9
pathogenicity. In a GBS-associated strain collection from Bangladesh (n = 30), three other
10
capsule genotypes, HS19, HS23/36c and HS41 were abundant and capsular type HS19 and
11
HS41 associated with GBS
12
In the GBS/MFS group, a capsular type was assigned to all but two strains whereas twenty-
13
two strains in the enteritis group failed to produce amplicons. The twenty-four strains in total
14
that did not produce amplicons likely have a capsular type distinct from the 17 major capsular
15
types that were tested. Multiple amplification fragments found in one case might be due to a
16
mix of two strains, gene translocations between particular capsular types or multiple capsule
17
operons in one genome.Earlier, we reported that no association was observed between GBS
18
and a particular capsule serotype in a collection of strains from GBS patients in The
19
Netherlands [8]. The discrepancy with the current results may be explained by the fact that
20
serotyping is sensitive to capsule expression whereas genotyping is not. For example, phase
21
variation and intragenic mutations that lead to premature stop codons result in more diversity
22
in capsule serotypes compared to the capsule genotypes [26]. Furthermore, for some
23
serotypes, the LOS plays a role in sero-diversity [29]. Reduction of the number of capsular
24
genotypes compared to serotypes, enhanced sensitivity of the genotyping method compared
25
to serotyping and, in particular, more power due to an expanded GBS/MFS-associated C.
AC C
EP
TE D
M AN U
SC
RI PT
4
13
ACCEPTED MANUSCRIPT jejuni strain collection are factors that explain why an association between capsular types and
2
GBS was found in the current study and not in the study performed by Dingle et al.[8]. The
3
association with GBS and the capsular types HS1/44c and HS4c reported in our study could
4
be biased due to a suboptimal enteritis control group in the case it would underrepresent
5
HS1/44c and HS4c and strains. Geographical differences in capsule type distribution has been
6
reported [30] but with 9% of HS1/44c and 26% of HS4c strains, our Dutch enteritis collection
7
certainly does not underrepresent the capsule proportion in Europe (9.1 % HS1/44c, 17.3%
8
HS4c) and North America (9.3% HS1/44c, 23.4% HS4c) [30].
9
As great diversity between the capsular loci and multiple rearrangements between capsular
10
genes has been reported for the capsular types HS1/44c, HS2, HS4c, HS19, HS23/26c and
11
HS41 [20, 26], it remains unclear whether particular genes or mutations in particular genes
12
within GBS-associated capsular types may promote the development of GBS.
13
Our current data, derived from a three-fold expanded Bangladeshi strain collection, confirm
14
that HS23/36c is a prevalent capsular type amongst GBS-associated strains from Bangladesh
15
[24, 31]. This capsular type, however, was also frequently observed in enteritis-associated
16
strains from Bangladesh and did not associated with GBS when compared to enteritis. When
17
we combined the results of the LOS classification with the capsule genotyping, a striking
18
correlation was observed. Capsular types HS19 and HS41 correlated solely with LOS class A,
19
HS23/36c often with LOS class B, HS1/44c was predominantly found in strains with LOS
20
class A and C and capsular type HS4c mostly associated with LOS class A and B strains. In
21
agreement with our data, strains with the HS19 (ST-22), HS23/36c (ST-403) and HS41 (ST-
22
362) capsular type represent genetically related populations [24, 32, 33]. A correlation
23
between capsular type and LOS class in these strains is therefore to be expected. The frequent
24
sharing of capsular types in GBS-associated strains that differ in MLST, observed in this
25
study, argues more strongly for a role of the capsule in GBS induction. The bi-functionality of
AC C
EP
TE D
M AN U
SC
RI PT
1
14
ACCEPTED MANUSCRIPT some of the enzymes involved in polysaccharide biosynthesis may additionally explain why
2
particular LOS loci and capsules types are linked. In C. jejuni, a phosphatase encoded by the
3
LOS locus gene cj1152 is active in the biosynthesis of both, LOS and capsule [20].
4
Additionally, gmhA in the LOS locus and gmhA2 in the capsule locus have duplicate functions
5
[20]. More in depth studies on gene functionality of LOS and capsule biosynthesis genes are
6
needed to explain why certain LOS classes and capsules are genetically linked together.
7
MLST analysis showed that the Dutch and Bangladeshi HS19 (ST-22) strains and the
8
Bangladeshi HS41 (ST-362) strains originate from a similar genetic lineage as HS19 strains
9
frequently isolated in Japan [34] and HS41 strains that were associated with GBS in South
10
Africa, respectively [35]. Moreover, the Japanese and South African GBS strains were all
11
LOS class A. The preservation of stable lineages in HS19 and HS41 strains is likely do to
12
DNase activity or defects in the DNA transformation mechanism resulting in nonnaturally
13
transformable strains [36]. The described feature of C. jejuni to exchange genes and even
14
entire loci due to horizontal gene transfer [20, 37] explains the observed variations in LOS
15
class and capsular type within identical MLST clonal complexes. Overall, MLST failed to
16
segregate strains associated with GBS from strains associated with gastroenteritis. Does
17
capsule discriminate between neuro-pathogenic and uncomplicated diarrhoeal strains of C.
18
jejuni? In particular for the LOS classes B and C the answer is affirmative. LOS class B and C
19
strains with capsular types other than HS1/44c, HS2, HS4c or HS23/36c were predominantly
20
detected in the enteritis-associated control strain collection. Capsular type HS8/17 and HS31
21
for example were observed in enteritis-associated strains with LOS class A, B or C but not in
22
GBS-associated strains with similar LOS classes. Such strains may therefore be considered to
23
be less potent with regard to GBS induction. Certainly, more comparative capsule studies,
24
including functional studies, are needed to determine if and how a particular capsular type
25
might contribute to GBS development. In conclusion, we demonstrate that the combination of
AC C
EP
TE D
M AN U
SC
RI PT
1
15
ACCEPTED MANUSCRIPT LOS loci that contain genes involved in LOS sialylation with the capsular types HS1/44c,
2
HS2, HS4c, HS19, HS23/36c and HS41 are associated with the development of GBS and
3
MFS and are therefore markers for these neurological diseases. The observation that similar
4
capsular types are prevalent in GBS-associated strains with genetically unrelated lineages
5
strongly argues for a role of the capsule in GBS inductions. A distinct capsular type other than
6
the six capsular types identified in this study may in part explain why not all infections with
7
C. jejuni that carry ganglioside mimics lead to GBS. The geographical differences in C. jejuni
8
capsular type distribution in GBS patients reported here needs to be confirmed in other
9
comparative studies.
AC C
EP
TE D
M AN U
SC
RI PT
1
16
ACCEPTED MANUSCRIPT
6 7 8
9
10
11
12
13
14
15
16
RI PT
4 5
SC
3
M AN U
2
Havelaar AH, Ivarsson S, Lofdahl M, Nauta MJ. Estimating the true incidence of campylobacteriosis and salmonellosis in the European Union, 2009. Epidemiol Infect. 2013; 141: 293-302. Scallan E, Hoekstra RM, Angulo FJ, et al. Foodborne illness acquired in the United States--major pathogens. Emerg Infect Dis. 2011; 17: 7-15. Allos BM. Campylobacter jejuni infections: Update on emerging issues and trends. Clin Infect Dis. 2001; 32: 1201-1206. Hughes RA, Cornblath DR. Guillain-Barré syndrome. Lancet. 2005; 366: 1653-1666. Winer JB, Hughes RAC, Anderson MJ, Jones DM, Kangro H, Watkins RP. A prospective study of acute idiopathic neuropathy. Ii. Antecedent events. J Neurol Neurosurg Psychiatry. 1988; 51: 613-618. Alshekhlee A, Hussain Z, Sultan B, Katirji B. Guillain-Barré syndrome: Incidence and mortality rates in US hospitals. Neurology. 2008; 70: 1608-1613. Fisher M. An unusual variant of acute idiopathic polyneuritis: Syndrome of ophthalmoplegia, ataxia and areflexia. N Engl J Med. 1956; 255: 57-65. Dingle KE, van den Braak N, Colles FM, et al. Sequence typing confirms that Campylobacter jejuni strains associated with Guillain-Barré and Miller Fisher syndromes are of diverse genetic lineage, serotype and flagella type. J Clin Microbiol. 2001; 39: 3346-3349. Louwen R, Horst-Kreft D, de Boer AG, et al. A novel link between Campylobacter jejuni bacteriophage defence, virulence and Guillain-Barre syndromé. Eur J Clin Microbiol Infect Dis. 2012. Yuki N, Susuki K, Koga M, et al. Carbohydrate mimicry between human ganglioside GM1 and Campylobacter jejuni lipooligosaccharide causes Guillain-Barré syndrome. PNAS. 2004; 101: 11404-11409. Godschalk PC, Heikema AP, Gilbert M, et al. The crucial role of Campylobacter jejuni genes in anti-ganglioside antibody induction in Guillain-Barré syndrome. J Clin Invest. 2004; 114: 1659-1665. Parker CT, Gilbert M, Yuki N, Endtz HP, Mandrell RE. Characterization of lipooligosaccharide-biosynthetic loci of Campylobacter jejuni reveals new lipooligosaccharide classes: Evidence of mosaic organizations. J Bacteriol. 2008; 190: 5681-5689. Eucker TP, Konkel ME. The cooperative action of bacterial fibronectin-binding proteins and secreted proteins promote maximal Campylobacter jejuni invasion of host cells by stimulating membrane ruffling. Cellular microbiology. 2012; 14: 226238. Jain D, Prasad KN, Sinha S, Vishwakarma AL. Cell cycle arrest & apoptosis of epithelial cell line by cytolethal distending toxin positive Campylobacter jejuni. The Indian journal of medical research. 2009; 129: 418-423. Talukder KA, Aslam M, Islam Z, et al. Prevalence of virulence genes and cytolethal distending toxin production in Campylobacter jejuni isolates from diarrheal patients in Bangladesh. J Clin Microbiol. 2008; 46: 1485-1488. Datta S, Niwa H, Itoh K. Prevalence of 11 pathogenic genes of Campylobacter jejuni by PCR in strains isolated from humans, poultry meat and broiler and bovine faeces. J Med Microbiol. 2003; 52: 345-348.
TE D
1
EP
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47
Reference List
AC C
1
17
ACCEPTED MANUSCRIPT
22
23 24
25
26 27
28
29 30 31 32
33
34
RI PT
21
SC
20
M AN U
19
TE D
18
Bacon DJ, Szymanski CM, Burr DH, Silver RP, Alm RA, Guerry P. A phase-variable capsule is involved in virulence of Campylobacter jejuni 81-176. Molecular microbiology. 2001; 40: 769-777. Keo T, Collins J, Kunwar P, Blaser MJ, Iovine NM. Campylobacter capsule and lipooligosaccharide confer resistance to serum and cationic antimicrobials. Virulence. 2011; 2: 30-40. Rose A, Kay E, Wren BW, Dallman MJ. The Campylobacter jejuni NCTC11168 capsule prevents excessive cytokine production by dendritic cells. Medical microbiology and immunology. 2012; 201: 137-144. Karlyshev AV, Champion OL, Churcher C, et al. Analysis of Campylobacter jejuni capsular loci reveals multiple mechanisms for the generation of structural diversity and the ability to form complex heptoses. Mol Microbiol. 2005; 55: 90-103. Penner JL, Hennessy JN, Congi RV. Serotyping of Campylobacter jejuni and Campylobacter coli on the basis of thermostable antigens. Eur J Clin Microbiol. 1983; 2: 378-383. Kuroki S, Saida T, Nukina M, et al. Campylobacter jejuni strains from patients with Guillain-Barré syndrome belong mostly to Penner serogroup 19 and contain beta-nacetylglucosamine residues. Ann Neurol. 1993; 33: 243-247. Lastovica AJ, Goddard EA, Argent AC. Guillain-Barré syndrome in South Africa associated with Campylobacter jejuni O:41 strains. J Infect Dis. 1997; 176: S139-143. Islam Z, van Belkum A, Wagenaar JA, et al. Comparative genotyping of Campylobacter jejuni strains from patients with Guillain-Barré syndrome in Bangladesh. PLoS ONE. 2009; 4: e7257. Allos BM, Lippy FT, Carlsen A, Washburn RG, Blaser MJ. Campylobacter jejuni strains from patients with Guillain-Barré Syndrome. Emerg Infect Dis. 1998; 4: 263268. Poly F, Serichatalergs O, Schulman M, et al. Discrimination of major capsular types of Campylobacter jejuni by multiplex PCR. J Clin Microbiol. 2011; 49: 1750-1757. Doorduyn Y, Van Den Brandhof WE, Van Duynhoven YT, Breukink BJ, Wagenaar JA, Van Pelt W. Risk factors for indigenous Campylobacter jejuni and Campylobacter coli infections in The Netherlands: A case-control study. Epidemiol Infect. 2010; 138: 1391-1404. Doorduyn Y, Van Den Brandhof WE, Van Duynhoven YT, Wannet WJ, Van Pelt W. R Guillain-Barré syndrome (DT104 and non-DT104) infections in The Netherlands: Predominant roles for raw eggs in enteritidis and sandboxes in typhimurium infections. Epidemiol Infect. 2006; 134: 617-626. Preston MA, Penner JL. Characterization of cross-reacting serotypes of Campylobacter jejuni. Can J Microbiol. 1989; 35: 265-273. Pike BL, Guerry P, Poly F. Global distribution of penner serotypes: A systematic review. PLoS ONE. 2013; 8: e67375. Islam Z, van Belkum A, Cody AJ, et al. Campylobacter jejuni HS:23 and GuillainBarré syndrome, bangladesh. Emerg Infect Dis. 2009; 15: 1315-1317. Nachamkin I, Engberg J, Gutacker M, et al. Molecular population genetic analysis of Campylobacter jejuni HS:19 associated with Guillain-Barré syndrome and gastroenteritis. J Infect Dis. 2001; 184: 221-226. Wassenaar TM, Fry BN, Lastovica AJ, Wagenaar JA, Coloe PJ, Duim B. Genetic characterization of Campylobacter jejuni O:41 isolates in relation with Guillain-Barré syndrome. J Clin Microbiol. 2000; 38: 874-876. Yabe S, Higuchi W, Iwao Y, et al. Molecular typing of Campylobacter jejuni and C. Coli from chickens and patients with gastritis or Guillain-Barré syndrome based on
EP
17
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
18
ACCEPTED MANUSCRIPT
36
37
RI PT
35
multilocus sequence types and pulsed-field gel electrophoresis patterns. Microbiology and immunology. 2010; 54: 362-367. Quinones B, Guilhabert MR, Miller WG, Mandrell RE, Lastovica AJ, Parker CT. Comparative genomic analysis of clinical strains of Campylobacter jejuni from South Africa. PLoS ONE. 2008; 3: e2015. Gaasbeek EJ, Wagenaar JA, Guilhabert MR, et al. A DNase encoded by integrated element CJE1 inhibits natural transformation of Campylobacter jejuni. J Bacteriol. 2009; 191: 2296-2306. Gilbert M, Godschalk PCR, Karwaski MF, et al. Evidence for acquisition of the lipooligosaccharide biosynthesis locus in Campylobacter jejuni GB11, a strain isolated from a patient with Guillain-Barré syndrome, by horizontal exchange. Infect Immun. 2004; 72: 1162-1165.
SC
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Acknowledgements
17
This project was supported in part by Navy Work Unit 6000.RAD1.DA3.A0308 and the
18
United States Department of Agriculture, Agricultural Research Service CRIS projects 5325-
19
42000-047.
AC C
EP
TE D
M AN U
16
19
ACCEPTED MANUSCRIPT 1
Figures
P = 0.01
40
30
20
P = 0.05
RI PT
Percentage of total (%)
50
10
0 HS15/31 HS6/7 HS41 HS23/36c HS19
HS4c
HS2 HS1/44c HS31
2
HS5
HS37
HS10 HS8/17 HS53
Capsule genotype
HS3
HS15
HS42
HS22
mix
UT
Figure 1. Capsule genotypes in GBS/MFS- and enteritis associated strains. A multiplex PCR
4
was used to determine the capsular type of 37 GBS/MFS- (black bars) and 143 enteritis-
5
associated (white bars) C. jejuni strains derived from Dutch patients and 30 GBS-associated
6
strains derived from patients with GBS from Bangladesh (grey bars). With this PCR, 17 main
7
capsular types can be distinguished. Chi square and Fisher exact tests were used to compare
8
the frequency of specific capsular types between the Dutch GBS/MFS- patients and the
9
enteritis group. UT, untypable; mix, more than one capsular type detected. P ≤ 0.05 was
TE D
M AN U
SC
3
10
considered as statistically significant.
11
A
GBS/MFS NL (n = 37)
B
Enteritis NL (n = 143)
GBS BD (n = 30)
EP
P = 0.01
HS41
AC C
Capsule type
Others
UT
Capsule type
UT
HS23/36c
HS19
Others HS41 HS23/36c HS19
HS4c
HS4c
HS2
HS2
HS1/44c
HS1/44 A
B
C
D
E
A
B
LOS loci class
C
D
E
A
B
C
D
E
LOS loci class
12
20
ACCEPTED MANUSCRIPT Figure 2. Correlation of LOS class and capsular type. For each strain, the LOS class was
2
correlated to the respective capsular type. The six main capsular types observed in the
3
GBS/MFS-strain collection are depicted. A. C. jejuni strains from The Netherlands. The Chi
4
square tests were used to assess whether there were significant differences in LOS
5
class/capsular type combinations between the GBS/MFS- and enteritis strains. Hereto, the six
6
main capsular types were compered to others and UT. B. C. jejuni strains from BD. Others,
7
capsular types that did not belong to the group of six dominant capsular types observed in the
8
GBS/MFS-associates strain collections but that could be determined with the multiplex
9
capsule PCR; NL, The Netherlands; BD, Bangladesh; UT, untypable; P ≤ 0.05 was
12
AC C
EP
TE D
11
SC
considered as statistically significant.
M AN U
10
RI PT
1
13
Supplemental figure 1. Minimal spanning tree of C. jejuni isolates derived from sporadic
14
cases of human enteritis present in the Oxford MLST data base together with our GBS/MFS-
15
and enteritis-associated strains. Each circle presents one sequence type (ST). Circles of
21
ACCEPTED MANUSCRIPT increased diameter represent higher strain numbers within on ST. Dominant STs are
2
indicated. NL, The Netherlands; BD, Bangladesh.
AC C
EP
TE D
M AN U
SC
RI PT
1
22
SC
Disease- LOS Capsule Country class genotype GBS-BD B HS23/36c GBS-BD B HS23/36c B HS23/36c GBS-BD B HS23/36c GBS-BD B HS23/36c GBS-BD B HS23/36c GBS-BD B HS23/36c GBS-BD B UT GBS-BD B HS4c MFS-NL B HS2 MFS-NL GBS-NL B HS2 B HS23/36c Enteritis A HS19 GBS-NL A HS19 GBS-BD A HS1/44c GBS-NL A GBS-BD HS15/31c A HS19 GBS-NL A HS19 GBS-NL A HS19 Enteritis A HS19 Enteritis A HS19 Enteritis A HS19 Enteritis A HS19 Enteritis A HS19 GBS-BD A HS19 GBS-BD GBS-BD A HS19 GBS-BD A HS19 A HS31 Enteritis GBS-BD B HS23/36c GBS-BD B HS23/36c B MFS-NL HS23/36c GBS-BD B HS23/36c Enteritis C UT C GBS-NL HS2 Enteritis C HS2 Enteritis A HS2 C C HS2 Enteritis Enteritis C HS2 Enteritis C HS2 Enteritis C UT Enteritis C UT GBS-NL C HS1/44c GBS-NL A HS2 Enteritis A HS2 Enteritis A HS2 Enteritis C HS2 GBS-NL C HS1/44c GBS-NL C HS1/44c Enteritis C HS8/17c Enteritis C HS8/17c Enteritis C HS8/17c Enteritis C HS8/17c Enteritis C HS8/17c Enteritis C HS8/17c C HS23/36c GBS-NL C Enteritis HS23/36c A GBS-NL HS2 B MFS-NL HS10 Enteritis B HS4c C GBS-NL HS1/44c C Enteritis HS1/44c B GBS-BD HS15/31c C HS2 GBS-NL B Enteritis HS15 HS1/44c GBS-NL A GBS-NL B HS4c HS4c Enteritis B Enteritis B HS53 A HS41 GBS-BD A HS41 GBS-BD A HS41 GBS-BD A HS41 GBS-BD A HS41 GBS-BD A HS41 GBS-BD B HS2 Enteritis B HS2 GBS-NL C Enteritis HS3c B MFS-NL HS4c B Enteritis HS4c B Enteritis HS4c Enteritis B HS4c Enteritis A HS4c Enteritis B HS4c Enteritis B HS4c Enteritis B HS4c B GBS-NL HS4c B GBS-BD HS4c B GBS-NL HS2 Enteritis C HS8/17c B GBS-NL HS4c Enteritis C HS23/36c B HS4c GBS-NL Enteritis B HS4c B HS4c GBS-NL Enteritis B HS4c Enteritis B HS4c A HS4c GBS-NL Enteritis A HS4c Enteritis A HS4c Enteritis A HS4c Enteritis A HS4c Enteritis A HS4c Enteritis A HS4c GBS-NL A HS4c GBS-NL A HS4c GBS-NL A HS4c GBS-NL A HS4c GBS-BD B HS6/7c GBS-BD B HS6/7c GBS-BD A HS6/7c GBS-NL A HS4c MFS-NL A HS4c A GBS-BD HS4c Enteritis A HS4c
M AN U
MLST CC ST-403 ST-403 ST-403 ST-403 ST-403 ST-403 ST-403 ST-403 ST-403 single single single ST-22 ST-22 ST-22 ST-22 ST-22 ST-22 ST-22 ST-22 ST-22 ST-22 ST-22 ST-22 ST-22 ST-22 ST-22 ST-658 ST-42 ST-42 ST-42 ST-42 single ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-206 ST-206 ST-21 ST-607 ST-52 single single ST-21 ST-21 ST-574 ST-21 ST-21 ST-21 ST-446 ST-362 ST-362 ST-362 ST-362 ST-362 ST-362 ST-464 ST-464 ST-353 ST-48 ST-48 ST-48 ST-48 ST-48 ST-48 ST-48 ST-48 ST-48 ST-48 ST-354 ST-607 ST-460 ST-206 ST-206 ST-206 ST-206 ST-206 ST-206 ST-206 ST-206 ST-206 ST-206 ST-206 ST-206 ST-206 ST-508 ST-508 ST-508 ST-508 single single single ST-61 ST-61 ST-61 ST-61
100
99
98
1
97
AC C
EP
TE D
MLST ST 985 985 6929 3219 3219 3219 3219 6938 83 131 131 380 149 6925 22 22 22 22 22 22 22 22 22 22 22 22 660 658 575 6928 42 42 986 53 53 53 53 53 53 53 53 148 148 21 21 21 19 50 50 50 50 50 50 50 46 46 883 607 2275 29 4292 6925 635 3015 186 185 104 875 587 587 587 587 587 6927 464 464 353 48 48 48 48 48 48 48 48 416 416 878 1707 606 4290 2104 572 572 572 572 122 122 122 122 122 122 122 508 508 64 132 3742 3742 3742 61 61 61 432
RI PT
ACCEPTED MANUSCRIPT
23
ACCEPTED MANUSCRIPT Figure3. Dendrogram of MLST genotype variation of C. jejuni strains with sialylated LOS
2
loci classes. The dendrogram was generated from MLST sequence type data of 58 GBS-, 5
3
MFS- and 51 enteritis-associated C. jejuni strains with either an LOS class A, B or C. ST,
4
sequence type; CC, clonal complex; GBS, Guillain-Barré syndrome; MFS, Miller Fisher
5
syndrome; NL, Netherlands; BD, Bangladesh.
AC C
EP
TE D
M AN U
SC
6
RI PT
1
24
S1198-743X(15)00563-7
DOI:
10.1016/j.cmi.2015.05.031
Reference:
CMI 290
To appear in:
Clinical Microbiology and Infection
Received Date: 5 November 2014 Revised Date:
5 February 2015
Accepted Date: 22 May 2015
Please cite this article as: Heikema AP, Islam Z, Horst-Kreft D, Huizinga R, Jacobs BC, Wagenaar JA, Poly F, Guerry P, van Belkum A, Parker CT, Endtz HP, Campylobacter jejuni capsular genotypes are related to Guillain-Barré syndrome, Clinical Microbiology and Infection (2015), doi: 10.1016/ j.cmi.2015.05.031. 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 1
Campylobacter jejuni capsular genotypes are related to Guillain-Barré syndrome
2 Astrid P. Heikema1*‡, Zhahirul Islam2‡, Deborah Horst-Kreft1, Ruth Huizinga3, Bart C.
4
Jacobs3,4, Jaap A. Wagenaar5,6, Frédéric Poly7, Patricia Guerry7, Alex van Belkum8, Craig T.
5
Parker9 and Hubert P. Endtz1, 2, 10
6
1
7
Medical Centre Rotterdam, The Netherlands; 2International Centre for Diarrhoeal Disease
8
Research Bangladesh, Dhaka, Bangladesh; 3Department of Immunology, Erasmus MC,
9
University Medical Centre Rotterdam, The Netherlands; Department of Neurology, Erasmus
RI PT
3
SC
Department of Medical Microbiology and Infectious Diseases, Erasmus MC, University
MC, University Medical Centre Rotterdam, The Netherlands; 5Department of Infectious
11
Diseases and Immunology, Faculty of Veterinary Medicine, Utrecht University, The
12
Netherlands; 6 Central Veterinary Institute of Wageningen UR, Lelystad, The Netherlands;
13
7
14
Microbiology, La Balme Les Grottes, France; 9Produce Safety and Microbiology Research
15
Unit, Agricultural Research Service, United States Department of Agriculture, Albany
16
California, USA; 10Fondation Mérieux, Lyon, France.
17
‡ These authors contributed equally to this work
M AN U
10
EP
TE D
Naval Medical Research Center, Silver Spring, Maryland, USA; 8bioMérieux, R&D
18
Running title: LOS classes and capsular types are correlated in GBS
AC C
19 20 21
Keywords: Campylobacter jejuni, capsular genotype, lipooligosaccharide, Guillain-Barré
22
syndrome, MLST
23 24 25
*
Corresponding author:
1
ACCEPTED MANUSCRIPT Astrid P. Heikema
2
3015 CE Rotterdam, The Netherlands
3
Erasmus MC, University Medical Centre Rotterdam
4
Department of Medical Microbiology and Infectious Diseases
5
‘s Gravendijkwal 230
6
3015 CE, Rotterdam, the Netherlands
7
Phone: +31-10-7032176
8
Fax: +31-10-4703875
9
Email: [email protected]
SC
AC C
EP
TE D
M AN U
10
RI PT
1
2
ACCEPTED MANUSCRIPT Abstract
2
In about one in thousand cases, a Campylobacter jejuni infection results in the severe
3
polyneuropathy Guillain-Barré syndrome (GBS). It is established that sialylated
4
lipooligosaccharides (LOS) of C. jejuni are a crucial virulence factor in GBS development.
5
Frequent detection of C. jejuni with sialylated LOS in stools derived from patient with
6
uncomplicated enteritis implies that additional bacterial factors should be involved. To assess
7
whether the polysaccharide capsule is a marker for GBS, the capsular genotype of two
8
geographical distinct GBS-associated C. jejuni strains collections, and an uncomplicated
9
enteritis control collection was determined. Capsule genotyping of C. jejuni strains from the
SC
RI PT
1
Netherlands revealed that three capsule types, HS1/44c, HS2 and HS4c, were dominant in
11
GBS-associated strains and capsular types HS1/44c and HS4c were significantly associated
12
with GBS (P = 0.05 and P = 0.01 respectively) when compared to uncomplicated enteritis. In
13
a GBS-associated strain collection from Bangladesh, capsular types HS23/36c, HS19, and
14
HS41 were most prevalent and the capsular types HS19 and HS41 associated with GBS (P =
15
0.008 and P = 0.02 respectively). Next, specific combinations of the LOS class and capsular
16
genotypes were identified that were related to the occurrence of GBS. Multilocus sequence
17
tying revealed restricted genetic diversity for strains populations with the capsular types HS2,
18
HS19 and HS41. We conclude that capsular types HS1/44c, HS2, HS4c, HS19, HS23/36c and
19
HS41 are markers for GBS. Besides a crucial role for sialylated LOS of C. jejuni in GBS
20
pathogenesis, the identified capsules may contribute to GBS susceptibility.
AC C
EP
TE D
M AN U
10
3
ACCEPTED MANUSCRIPT Introduction
2
Campylobacter jejuni is one of the most frequent causes of bacterial gastroenteritis in humans
3
worldwide [1, 2]. In general, C. jejuni diarrhea is self-limiting though a wide spectrum of
4
serious complications may occur, including bacteremia, reactive arthritis and peripheral
5
neuropathies e.g. Guillain-Barré syndrome (GBS) and Miller Fisher syndrome (MFS) [3].
6
GBS is a post-infectious, immune-mediated disease of the peripheral nerves, which is
7
characterized by a rapidly progressive weakness of the limb muscles [4]. About one-quarter of
8
patients develop additional respiratory failure and autonomic dysfunction and 2-10% of the
9
patients die [5, 6]. MFS is regarded a variant of GBS in which the weakness is restricted to
SC
RI PT
1
the oculomotor muscles [7].
11
Various studies have investigated which risk factors of C. jejuni predispose to the
12
development of GBS [8, 9]. Currently, sialylation of lipooligosaccharides (LOS) of C. jejuni
13
is the best characterized virulence factor associated with GBS [10, 11]. The structural
14
similarity (molecular mimicry) between sialylated C. jejuni LOS and ganglioside structures
15
present on human peripheral nerves drives a cross-reactive immune response resulting in
16
immune-mediated nerve damage [10]. In C. jejuni, genes involved in LOS sialylation are
17
located in the LOS biosynthesis locus. Variability in gene content in C. jejuni LOS loci has
18
led to the assignment of several LOS locus classes (A through S) [12]. Genes involved in the
19
synthesis of sialylated LOS and ganglioside mimics are present in the LOS classes A, B, C, M
20
and R [12]. Of these, C. jejuni strains with LOS classes A, B and C are frequently isolated
21
from stools of patients with GBS and the LOS classes A and B are associated with GBS and
22
MFS respectively [11].
23
The presence of sialic acid in the LOS, however, does not appear to be the only bacterial
24
factor involved in the pathogenesis of GBS. In patients with uncomplicated enteritis, the
25
prevalence of sialylated C. jejuni strains has been reported to be up to 63%[11]. We
AC C
EP
TE D
M AN U
10
4
ACCEPTED MANUSCRIPT hypothesize that the expression of additional C. jejuni virulence factors in combination with
2
the presence of sialylated LOS contributes to the induction of GBS. Factors that increase the
3
virulence of C. jejuni include the bipolar flagella, surface and secreted proteins involved in
4
the invasion of the intestinal epithelium [13], and cytolethal distending toxin (CDT) [14].
5
Genes that encode these virulence factors, however, are generally conserved and present in
6
strains derived from both clinical and environmental sources [15, 16].
7
The polysaccharide capsule is a virulence factor that is associated with cell invasion [17],
8
complement resistance [18] and immune modulation [19]. The capsule biosynthesis locus is a
9
highly variable region in the genome of C. jejuni. Diversity in gene content and gene function,
SC
RI PT
1
combined with the presence of phase variable genes in the capsule locus allows the
11
production of a broad repertoire of capsular structures [20]. This feature is the basis of the
12
classical Penner serotyping scheme which divides C. jejuni into 47 serotypes [21]. Previous
13
studies based on Penner serotyping showed that the capsular serotypes HS:19, HS:23/36c and
14
HS:41 are associated with GBS in Japan, Bangladesh and South Africa, respectively [22-24].
15
In contrast, no particular association between the capsule serotype and neuro-pathogenicity
16
was observed for GBS-associated strains from Europe or North America [8, 25]. This
17
discrepancy may be explained by the observation that GBS-associated C. jejuni strains in
18
Japan, Bangladesh and South Africa had a genetically related background whereas the strains
19
in Europe and the USA were more heterogeneous.
20
Recently, capsular genotyping was introduced as a fast, readily available and reliable method
21
to assess the capsular type in C. jejuni [26]. Compared to capsule based serotyping, capsular
22
genotyping is not sensitive to variations in capsule gene expression or influenced by genes or
23
gene products outside the capsule locus. With the availability of two large and geographically
24
distinct GBS-associated strain collections, we determined whether variation in C. jejuni
25
capsule genotype is involved in the risk to develop GBS. The capsule genotype defined by
AC C
EP
TE D
M AN U
10
5
ACCEPTED MANUSCRIPT multiplex PCR was determined in collections of C. jejuni strains isolated from Dutch
2
GBS/MFS patients, Dutch enteritis control patients and Bangladeshi GBS/MFS patients. The
3
capsular type was related to the LOS class in order to determine whether particular LOS
4
class/capsular type combinations were more prevalent amongst GBS/MFS-associated strains
5
compared to enteritis-associates strains. In addition, multilocus sequence typing (MLST) was
6
performed to assess the genetic relation between strains.
RI PT
1
7 Materials and methods
9
Strain collection, culture conditions and DNA isolation
SC
8
In this study, a Dutch- and a Bangladeshi GBS-associated C. jejuni strain collections and a
11
Dutch enteritis-associated control strain collections were used. The Dutch GBS collection
12
comprises 32 GBS- and 5 MFS-associated strains which were isolated from stools of
13
primarily Dutch patients. Among these are three GBS-associated strains from patients from
14
the Netherlands Antilles and one from Belgium. The GBS-associated strain collection from
15
Bangladesh consists of 30 strains.The Dutch enteritis-associated strains were isolated from
16
patients with diarrhoea in The Netherlands and were from two different provenances. The
17
strains in the first enteritis collection (n = 21) were isolated between 1990 and 1999 and have
18
been previously described [11]. A second enteritis collection originates from a Dutch study
19
performed by the Rijksinstituut voor Volksgezondheid en Milieu in 2002-2003, on behalf of a
20
national Campylobacter and Salmonella surveillance study [27, 28]. From this large
21
collection, we selected 122 enteritis strains that matched to our GBS/MFS-associated strains
22
based on age (decade) and sex. Thus, we selected 3-4 age and sex matched enteritis control
23
strains per GBS/MFS case. There were no significant differences in the distribution of the
24
various capsule types between the two enteritis collections. Therefore, the two enteritis strain
25
collections were combined in this study and referred to as the uncomplicated enteritis-
AC C
EP
TE D
M AN U
10
6
ACCEPTED MANUSCRIPT associated strain collection (n = 143). For the Dutch GBS-, MFS- and enteritis-associated
2
strains, LOS class typing (A, B, C, D and E) was performed previously [11] (Heikema et al.,
3
manuscript in preparation). The LOS class of 10 of the GBS-associated C. jejuni strains
4
derived from Bangladesh was determined previously [24]. For the remaining 20 strains, the
5
LOS class was newly assessed. C. jejuni strains were cultured from -80°C stocks and
6
maintained on Colombia blood agar (BA) plates (Becton Dickinson BV, Alphen aan den Rijn,
7
The Netherlands) supplemented with 10 µg/ml vancomycin in a micro-aerobic atmosphere at
8
37°C. DNA was isolated from overnight grown C. jejuni, using a DNeasy blood & tissue kit
9
(Qiagen, Venlo, The Netherlands) as described by the manufacturer’s protocol.
SC
RI PT
1
M AN U
10 LOS locus class PCR
12
Genotyping was performed for LOS classes A, B, C, D and E using standard PCR. Primers
13
used are based on the DNA sequence of unique genes or regions of the LOS locus classes
14
involved and were previously described[11].
15
TE D
11
Multiplex PCR-based capsule analysis
17
Capsule genotyping was performed as described [26]. In short, based on the sequence of 17
18
CPS loci, unique regions were identified and specific primers pairs that recognize each
19
serotype were designed. The primer pairs were divided in two mixes in such a way that the
20
expected amplicons differed at least 20 bp from one another. Standard PCR was performed
21
and the PCR products were separated on 2% agarose gels using a 50-bp DNA ladder
22
(Fermentas, Landsmeer, The Netherland) to assess the size of each amplicon.
AC C
EP
16
23 24
Multi Locus Sequence Typing (MLST)
7
ACCEPTED MANUSCRIPT 1
MLST, which relies on the partial sequencing of seven housekeeping genes, was performed as
2
described [8]. A dendrogram was constructed using BioNumerics software. Pairwise
3
alignment of the sequences of the separate MLST alleles was performed. The dendrogram
4
was generated using UPGMA clustering.
RI PT
5 Statistical analysis
7
Fisher exact and Chi square tests were performed using Prism software (GraphPad, La Jolla,
8
CA, USA). Two-sided P ≤ 0.05 was considered to be statistically significant.
SC
6
AC C
EP
TE D
M AN U
9
8
ACCEPTED MANUSCRIPT Results
2
Validation of the capsular genotyping method
3
The capsular genotype determined by PCR was compared to the serotype for our C. jejuni
4
strains with known Penner serotype (n = 40). The serotype did not match the genotype in six
5
cases and in one case the serotype was not included in the multiplex capsule PCR (data not
6
shown). Full genome sequence data were available for 4/6 strains with a mismatch between
7
capsular serotype and genotype. Blast analysis of the capsule locus region revealed homology
8
to strains with serotypes that matched the identified genotype and not the serotype in all four
9
cases (data not shown). Overall, we found a specificity of 95% and a sensitivity of 100% for
SC
RI PT
1
strains (n = 39) with serotypes covered in the genotyping method. Therefore, we concluded
11
that the genotyping method is a reliable method to determine the capsular type in C. jejuni.
M AN U
10
12
Six capsular genotypes are dominant in GBS/MFS-associated C. jejuni strains
14
In the Dutch GBS/MFS-associated strains, a genocapsular type could be assigned for all
15
strains whereas in the Dutch enteritis-associated strains a capsular genotype could be assigned
16
for 120/143 (83%) of the strains.. Twenty-two strains in the enteritis group failed to produce
17
amplicons and in one case a mix of two capsular types (HS1/HS41) was found. In the Dutch
18
GBS/MFS collection, nine different capsular types were identified of which three capsular
19
types; HS4c, HS2 and HS1/44c were dominant (Fig. 1). In the Dutch enteritis strain
20
collection, a broader panel of 15 different capsular types was identified of which capsules
21
HS4c, HS8/17 and HS2 were most prevalent (Fig. 1). When GBS/MFS-associated C. jejuni
22
strains and enteritis-associated strains were compared, capsular type HS4c (P = 0.01; Chi
23
square test) and HS1/44c (P = 0.05; Chi square test) significantly associated with GBS/MFS
24
(Fig. 1). No association with a specific capsular type was observed for the MFS strains. In the
25
five MFS associated strains, four different capsular types were detected; HS2, HS4c (in two
AC C
EP
TE D
13
9
ACCEPTED MANUSCRIPT cases), HS10 and HS23/36c. When combined, 28/37 (76%) of the GBS/MFS-associated
2
strains from The Netherlands either had capsular type HS1/44c, HS2 or HS4c, compared to
3
50/143 (35%) in the enteritis control strains (P < 0.0001; Chi square test).
4
In order to determine whether the dominant capsular types identified in the GBS/MFS-
5
associated strain collection from the Netherlands were also prominent in GBS-associated
6
strains from another geographical area, capsule genotyping was performed on 30 GBS-
7
associated strains isolated from patients from Bangladesh. In these strains, a capsular type
8
could be assigned for 28/30 (93%) of the strains and capsular types HS23/36c, HS19 and
9
HS41 were dominant (Fig. 1). In a randomly selected uncomplicated enteritis control group
SC
RI PT
1
from Bangladesh, 0/39 (0%) of the C. jejuni strains had the capsular type HS19, 11/39 (28%)
11
had HS23/36 and in 1/39 (3%) capsular type HS41 was identified (Z. Islam, personal
12
communications). The capsular types HS19 and HS41 significantly associated with GBS in
13
Bangladesh when compared to enteritis (P = 0.008 and P = 0.02 respectively; Chi square test).
14
Capsular type HS41 was only found in strains from Bangladesh whereas capsular types
15
HS23/36c and HS19 were also observed in the Dutch strain collections. When combined,
16
56/67 (84%) of the GBS/MFS-associated C. jejuni strains either had capsular type HS1/44c,
17
HS2, HS4c, HS19, HS23/36c or HS41 (from here on referred to as GBS/MFS-related capsular
18
types).
TE D
EP
AC C
19
M AN U
10
20
Sialylated LOS classes and specific capsular genotypes are correlated
21
Upon correlation of LOS class and capsular genotype in the GBS/MFS strain collection from
22
The Netherlands, we observed that all but one strain with the LOS class A, B or C had the
23
GBS/MFS-related capsular types (Fig. 2A). A particular distribution pattern was observed
24
when LOS class and capsular type were combined. In the GBS/MFS collection, strains with
25
LOS class A were typed HS1/44c, HS2, HS4c or HS19, in strains with LOS class B, capsular
10
ACCEPTED MANUSCRIPT types HS2 and HS4c were the most prominent and LOS class C strains, mostly had capsular
2
type HS1/41 or HS2 (Fig. 2A). In the enteritis control collection, approximately the same
3
capsule distribution was observed for LOS class A strains. In contrast, for the class B and C
4
strains, similar capsular types were only found for a subset of the enteritis strains when
5
compared to the GBS/MFS strains. A distinct capsular type was found in 9/28 (32%) of the
6
LOS class B strains and 18/34; 53% of the class C strains had a capsular type that did not
7
belong to the GBS/MFS-related capsule group (P = 0.01; GBS/MFS vs enteritis; Chi square).
8
In enteritis strains with LOS class B, there was diversity in the remaining capsular types but in
9
the LOS class C strains, capsular type HS8/17 was detected in 9/18 (50%) of the strains which
SC
RI PT
1
did not belong to the GBS/MFS-related capsule group.
11
In enteritis strains with LOS class D, capsular type HS53 was dominant (9/31; 29%; HS53 is
12
not present in class D strains of the GBS/MFS collection) whereas capsular types within class
13
E strains were diverse (data not shown). It should be noted that LOS class D and E loci do not
14
possess the sialylation genes necessary to produce ganglioside mimics.
15
LOS genotyping of 20 newly isolated GBS-associated C. jejuni strains from Bangladeshi
16
patients resulted in the identification of 10 strains with LOS class A and 10 strains with LOS
17
class B. When LOS class and capsular type of the GBS-associated strains derived from
18
Bangladeshi patients (n = 30) were correlated, all strains with capsular type HS41 or HS19
19
correlated with LOS class A, whereas all the strains with the HS23/36c capsular type
20
correlated with LOS class B (Fig. 2B).
TE D
EP
AC C
21
M AN U
10
22
A genetic linkage between LOS/capsule combinations
23
MLST was performed on all (n = 67) GBS- and MFS-associated strains and on 84/143 (59%)
24
of the enteritis-associated strains. The 84 enteritis control strains on which MLST was
25
performed comprise 60% sialylated strains and 40% non-sialylated strains. These percentages
11
ACCEPTED MANUSCRIPT are representative for the whole enteritis control collection of 143 strains. A total of 76
2
different sequence types were identified which were classified into 21 clonal complexes. C.
3
jejuni clonal complex ST-21 was most prominent (29/151; 19%), followed by ST-206
4
(16/151; 11%) and ST-22 (15/151; 10%). In 13 cases, the MLST sequence type represented a
5
singleton. In the GBS collection, the clonal complexes ST-21 and ST-22 were the most
6
common (both 10/67; 15%) followed by ST-403 (9/67; 13%). A minimum spanning tree
7
(MST), in which our MLST data were combined with MLST data of C. jejuni isolates from
8
sporadic cases of human enteritis present in the Oxford C. jejuni MLST database (n = 4066),
9
showed that our GBS/MFS- associated and enteritis-associated strains were widely spread
10
within the total C. jejuni MLST population (Supplemental figure 1). GBS/MFS- as well as
11
enteritis-associated control strains were present within the same clonal complexes.
12
To assess whether the LOS class/capsular type co-occurrence within one strain was due to a
13
closely related genetic background, in particular for the strains that carry genes involved in
14
LOS sialylation, a dendrogram was generated using strains from both, the Netherlands and
15
Bangladesh, which either had LOS loci classes A, B or C (Fig. 3 ). The same genetic
16
background, ST-21, was frequently observed for strains with capsular types HS1/44c and HS2
17
and LOS class C (18/29; 62%). Most HS8/17 strains with LOS class C (6/7; 86%) also
18
belonged to ST-21. The HS4c strains with LOS class A or B were predominantly present in
19
four genetic backgrounds: ST-206, ST-48, ST-61 and ST-508 (30/35; 86%). A strong
20
individual genetic relationship was observed for LOS class A strains with capsular type HS41
21
(n = 6) and HS19 (n = 13). All HS19 strains belonged to ST-22 whereas all HS41 strains
22
belonged to ST-362. The strains with capsular type HS23/36c were more widely spread
23
within the dendrogram. Most (11/14; 78%), however, belonged to two MLST CCs, ST403
24
and ST-42 (Fig 3). One MLST clonal complex, ST-508 contained GBS-associated strains
AC C
EP
TE D
M AN U
SC
RI PT
1
12
ACCEPTED MANUSCRIPT 1
only but overall MLST typing did not discriminate strains carrying genes involved in LOS
2
sialylation when GBS/MFS- and enteritis-associated strains were compared.
3 Discussion
5
We identified a group of six main capsular types HS1/44c, HS2, HS4c, HS19, HS23/36c and
6
HS41 in two large, geographically distinct GBS/MFS-associated C. jejuni strain collections.
7
In a predominantly Dutch collection (n = 37), capsular genotypes HS1/44c, HS2 and HS4c
8
were dominant. Of these, capsular type HS4c and HS1 associated with the neuro-
9
pathogenicity. In a GBS-associated strain collection from Bangladesh (n = 30), three other
10
capsule genotypes, HS19, HS23/36c and HS41 were abundant and capsular type HS19 and
11
HS41 associated with GBS
12
In the GBS/MFS group, a capsular type was assigned to all but two strains whereas twenty-
13
two strains in the enteritis group failed to produce amplicons. The twenty-four strains in total
14
that did not produce amplicons likely have a capsular type distinct from the 17 major capsular
15
types that were tested. Multiple amplification fragments found in one case might be due to a
16
mix of two strains, gene translocations between particular capsular types or multiple capsule
17
operons in one genome.Earlier, we reported that no association was observed between GBS
18
and a particular capsule serotype in a collection of strains from GBS patients in The
19
Netherlands [8]. The discrepancy with the current results may be explained by the fact that
20
serotyping is sensitive to capsule expression whereas genotyping is not. For example, phase
21
variation and intragenic mutations that lead to premature stop codons result in more diversity
22
in capsule serotypes compared to the capsule genotypes [26]. Furthermore, for some
23
serotypes, the LOS plays a role in sero-diversity [29]. Reduction of the number of capsular
24
genotypes compared to serotypes, enhanced sensitivity of the genotyping method compared
25
to serotyping and, in particular, more power due to an expanded GBS/MFS-associated C.
AC C
EP
TE D
M AN U
SC
RI PT
4
13
ACCEPTED MANUSCRIPT jejuni strain collection are factors that explain why an association between capsular types and
2
GBS was found in the current study and not in the study performed by Dingle et al.[8]. The
3
association with GBS and the capsular types HS1/44c and HS4c reported in our study could
4
be biased due to a suboptimal enteritis control group in the case it would underrepresent
5
HS1/44c and HS4c and strains. Geographical differences in capsule type distribution has been
6
reported [30] but with 9% of HS1/44c and 26% of HS4c strains, our Dutch enteritis collection
7
certainly does not underrepresent the capsule proportion in Europe (9.1 % HS1/44c, 17.3%
8
HS4c) and North America (9.3% HS1/44c, 23.4% HS4c) [30].
9
As great diversity between the capsular loci and multiple rearrangements between capsular
10
genes has been reported for the capsular types HS1/44c, HS2, HS4c, HS19, HS23/26c and
11
HS41 [20, 26], it remains unclear whether particular genes or mutations in particular genes
12
within GBS-associated capsular types may promote the development of GBS.
13
Our current data, derived from a three-fold expanded Bangladeshi strain collection, confirm
14
that HS23/36c is a prevalent capsular type amongst GBS-associated strains from Bangladesh
15
[24, 31]. This capsular type, however, was also frequently observed in enteritis-associated
16
strains from Bangladesh and did not associated with GBS when compared to enteritis. When
17
we combined the results of the LOS classification with the capsule genotyping, a striking
18
correlation was observed. Capsular types HS19 and HS41 correlated solely with LOS class A,
19
HS23/36c often with LOS class B, HS1/44c was predominantly found in strains with LOS
20
class A and C and capsular type HS4c mostly associated with LOS class A and B strains. In
21
agreement with our data, strains with the HS19 (ST-22), HS23/36c (ST-403) and HS41 (ST-
22
362) capsular type represent genetically related populations [24, 32, 33]. A correlation
23
between capsular type and LOS class in these strains is therefore to be expected. The frequent
24
sharing of capsular types in GBS-associated strains that differ in MLST, observed in this
25
study, argues more strongly for a role of the capsule in GBS induction. The bi-functionality of
AC C
EP
TE D
M AN U
SC
RI PT
1
14
ACCEPTED MANUSCRIPT some of the enzymes involved in polysaccharide biosynthesis may additionally explain why
2
particular LOS loci and capsules types are linked. In C. jejuni, a phosphatase encoded by the
3
LOS locus gene cj1152 is active in the biosynthesis of both, LOS and capsule [20].
4
Additionally, gmhA in the LOS locus and gmhA2 in the capsule locus have duplicate functions
5
[20]. More in depth studies on gene functionality of LOS and capsule biosynthesis genes are
6
needed to explain why certain LOS classes and capsules are genetically linked together.
7
MLST analysis showed that the Dutch and Bangladeshi HS19 (ST-22) strains and the
8
Bangladeshi HS41 (ST-362) strains originate from a similar genetic lineage as HS19 strains
9
frequently isolated in Japan [34] and HS41 strains that were associated with GBS in South
10
Africa, respectively [35]. Moreover, the Japanese and South African GBS strains were all
11
LOS class A. The preservation of stable lineages in HS19 and HS41 strains is likely do to
12
DNase activity or defects in the DNA transformation mechanism resulting in nonnaturally
13
transformable strains [36]. The described feature of C. jejuni to exchange genes and even
14
entire loci due to horizontal gene transfer [20, 37] explains the observed variations in LOS
15
class and capsular type within identical MLST clonal complexes. Overall, MLST failed to
16
segregate strains associated with GBS from strains associated with gastroenteritis. Does
17
capsule discriminate between neuro-pathogenic and uncomplicated diarrhoeal strains of C.
18
jejuni? In particular for the LOS classes B and C the answer is affirmative. LOS class B and C
19
strains with capsular types other than HS1/44c, HS2, HS4c or HS23/36c were predominantly
20
detected in the enteritis-associated control strain collection. Capsular type HS8/17 and HS31
21
for example were observed in enteritis-associated strains with LOS class A, B or C but not in
22
GBS-associated strains with similar LOS classes. Such strains may therefore be considered to
23
be less potent with regard to GBS induction. Certainly, more comparative capsule studies,
24
including functional studies, are needed to determine if and how a particular capsular type
25
might contribute to GBS development. In conclusion, we demonstrate that the combination of
AC C
EP
TE D
M AN U
SC
RI PT
1
15
ACCEPTED MANUSCRIPT LOS loci that contain genes involved in LOS sialylation with the capsular types HS1/44c,
2
HS2, HS4c, HS19, HS23/36c and HS41 are associated with the development of GBS and
3
MFS and are therefore markers for these neurological diseases. The observation that similar
4
capsular types are prevalent in GBS-associated strains with genetically unrelated lineages
5
strongly argues for a role of the capsule in GBS inductions. A distinct capsular type other than
6
the six capsular types identified in this study may in part explain why not all infections with
7
C. jejuni that carry ganglioside mimics lead to GBS. The geographical differences in C. jejuni
8
capsular type distribution in GBS patients reported here needs to be confirmed in other
9
comparative studies.
AC C
EP
TE D
M AN U
SC
RI PT
1
16
ACCEPTED MANUSCRIPT
6 7 8
9
10
11
12
13
14
15
16
RI PT
4 5
SC
3
M AN U
2
Havelaar AH, Ivarsson S, Lofdahl M, Nauta MJ. Estimating the true incidence of campylobacteriosis and salmonellosis in the European Union, 2009. Epidemiol Infect. 2013; 141: 293-302. Scallan E, Hoekstra RM, Angulo FJ, et al. Foodborne illness acquired in the United States--major pathogens. Emerg Infect Dis. 2011; 17: 7-15. Allos BM. Campylobacter jejuni infections: Update on emerging issues and trends. Clin Infect Dis. 2001; 32: 1201-1206. Hughes RA, Cornblath DR. Guillain-Barré syndrome. Lancet. 2005; 366: 1653-1666. Winer JB, Hughes RAC, Anderson MJ, Jones DM, Kangro H, Watkins RP. A prospective study of acute idiopathic neuropathy. Ii. Antecedent events. J Neurol Neurosurg Psychiatry. 1988; 51: 613-618. Alshekhlee A, Hussain Z, Sultan B, Katirji B. Guillain-Barré syndrome: Incidence and mortality rates in US hospitals. Neurology. 2008; 70: 1608-1613. Fisher M. An unusual variant of acute idiopathic polyneuritis: Syndrome of ophthalmoplegia, ataxia and areflexia. N Engl J Med. 1956; 255: 57-65. Dingle KE, van den Braak N, Colles FM, et al. Sequence typing confirms that Campylobacter jejuni strains associated with Guillain-Barré and Miller Fisher syndromes are of diverse genetic lineage, serotype and flagella type. J Clin Microbiol. 2001; 39: 3346-3349. Louwen R, Horst-Kreft D, de Boer AG, et al. A novel link between Campylobacter jejuni bacteriophage defence, virulence and Guillain-Barre syndromé. Eur J Clin Microbiol Infect Dis. 2012. Yuki N, Susuki K, Koga M, et al. Carbohydrate mimicry between human ganglioside GM1 and Campylobacter jejuni lipooligosaccharide causes Guillain-Barré syndrome. PNAS. 2004; 101: 11404-11409. Godschalk PC, Heikema AP, Gilbert M, et al. The crucial role of Campylobacter jejuni genes in anti-ganglioside antibody induction in Guillain-Barré syndrome. J Clin Invest. 2004; 114: 1659-1665. Parker CT, Gilbert M, Yuki N, Endtz HP, Mandrell RE. Characterization of lipooligosaccharide-biosynthetic loci of Campylobacter jejuni reveals new lipooligosaccharide classes: Evidence of mosaic organizations. J Bacteriol. 2008; 190: 5681-5689. Eucker TP, Konkel ME. The cooperative action of bacterial fibronectin-binding proteins and secreted proteins promote maximal Campylobacter jejuni invasion of host cells by stimulating membrane ruffling. Cellular microbiology. 2012; 14: 226238. Jain D, Prasad KN, Sinha S, Vishwakarma AL. Cell cycle arrest & apoptosis of epithelial cell line by cytolethal distending toxin positive Campylobacter jejuni. The Indian journal of medical research. 2009; 129: 418-423. Talukder KA, Aslam M, Islam Z, et al. Prevalence of virulence genes and cytolethal distending toxin production in Campylobacter jejuni isolates from diarrheal patients in Bangladesh. J Clin Microbiol. 2008; 46: 1485-1488. Datta S, Niwa H, Itoh K. Prevalence of 11 pathogenic genes of Campylobacter jejuni by PCR in strains isolated from humans, poultry meat and broiler and bovine faeces. J Med Microbiol. 2003; 52: 345-348.
TE D
1
EP
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47
Reference List
AC C
1
17
ACCEPTED MANUSCRIPT
22
23 24
25
26 27
28
29 30 31 32
33
34
RI PT
21
SC
20
M AN U
19
TE D
18
Bacon DJ, Szymanski CM, Burr DH, Silver RP, Alm RA, Guerry P. A phase-variable capsule is involved in virulence of Campylobacter jejuni 81-176. Molecular microbiology. 2001; 40: 769-777. Keo T, Collins J, Kunwar P, Blaser MJ, Iovine NM. Campylobacter capsule and lipooligosaccharide confer resistance to serum and cationic antimicrobials. Virulence. 2011; 2: 30-40. Rose A, Kay E, Wren BW, Dallman MJ. The Campylobacter jejuni NCTC11168 capsule prevents excessive cytokine production by dendritic cells. Medical microbiology and immunology. 2012; 201: 137-144. Karlyshev AV, Champion OL, Churcher C, et al. Analysis of Campylobacter jejuni capsular loci reveals multiple mechanisms for the generation of structural diversity and the ability to form complex heptoses. Mol Microbiol. 2005; 55: 90-103. Penner JL, Hennessy JN, Congi RV. Serotyping of Campylobacter jejuni and Campylobacter coli on the basis of thermostable antigens. Eur J Clin Microbiol. 1983; 2: 378-383. Kuroki S, Saida T, Nukina M, et al. Campylobacter jejuni strains from patients with Guillain-Barré syndrome belong mostly to Penner serogroup 19 and contain beta-nacetylglucosamine residues. Ann Neurol. 1993; 33: 243-247. Lastovica AJ, Goddard EA, Argent AC. Guillain-Barré syndrome in South Africa associated with Campylobacter jejuni O:41 strains. J Infect Dis. 1997; 176: S139-143. Islam Z, van Belkum A, Wagenaar JA, et al. Comparative genotyping of Campylobacter jejuni strains from patients with Guillain-Barré syndrome in Bangladesh. PLoS ONE. 2009; 4: e7257. Allos BM, Lippy FT, Carlsen A, Washburn RG, Blaser MJ. Campylobacter jejuni strains from patients with Guillain-Barré Syndrome. Emerg Infect Dis. 1998; 4: 263268. Poly F, Serichatalergs O, Schulman M, et al. Discrimination of major capsular types of Campylobacter jejuni by multiplex PCR. J Clin Microbiol. 2011; 49: 1750-1757. Doorduyn Y, Van Den Brandhof WE, Van Duynhoven YT, Breukink BJ, Wagenaar JA, Van Pelt W. Risk factors for indigenous Campylobacter jejuni and Campylobacter coli infections in The Netherlands: A case-control study. Epidemiol Infect. 2010; 138: 1391-1404. Doorduyn Y, Van Den Brandhof WE, Van Duynhoven YT, Wannet WJ, Van Pelt W. R Guillain-Barré syndrome (DT104 and non-DT104) infections in The Netherlands: Predominant roles for raw eggs in enteritidis and sandboxes in typhimurium infections. Epidemiol Infect. 2006; 134: 617-626. Preston MA, Penner JL. Characterization of cross-reacting serotypes of Campylobacter jejuni. Can J Microbiol. 1989; 35: 265-273. Pike BL, Guerry P, Poly F. Global distribution of penner serotypes: A systematic review. PLoS ONE. 2013; 8: e67375. Islam Z, van Belkum A, Cody AJ, et al. Campylobacter jejuni HS:23 and GuillainBarré syndrome, bangladesh. Emerg Infect Dis. 2009; 15: 1315-1317. Nachamkin I, Engberg J, Gutacker M, et al. Molecular population genetic analysis of Campylobacter jejuni HS:19 associated with Guillain-Barré syndrome and gastroenteritis. J Infect Dis. 2001; 184: 221-226. Wassenaar TM, Fry BN, Lastovica AJ, Wagenaar JA, Coloe PJ, Duim B. Genetic characterization of Campylobacter jejuni O:41 isolates in relation with Guillain-Barré syndrome. J Clin Microbiol. 2000; 38: 874-876. Yabe S, Higuchi W, Iwao Y, et al. Molecular typing of Campylobacter jejuni and C. Coli from chickens and patients with gastritis or Guillain-Barré syndrome based on
EP
17
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
18
ACCEPTED MANUSCRIPT
36
37
RI PT
35
multilocus sequence types and pulsed-field gel electrophoresis patterns. Microbiology and immunology. 2010; 54: 362-367. Quinones B, Guilhabert MR, Miller WG, Mandrell RE, Lastovica AJ, Parker CT. Comparative genomic analysis of clinical strains of Campylobacter jejuni from South Africa. PLoS ONE. 2008; 3: e2015. Gaasbeek EJ, Wagenaar JA, Guilhabert MR, et al. A DNase encoded by integrated element CJE1 inhibits natural transformation of Campylobacter jejuni. J Bacteriol. 2009; 191: 2296-2306. Gilbert M, Godschalk PCR, Karwaski MF, et al. Evidence for acquisition of the lipooligosaccharide biosynthesis locus in Campylobacter jejuni GB11, a strain isolated from a patient with Guillain-Barré syndrome, by horizontal exchange. Infect Immun. 2004; 72: 1162-1165.
SC
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Acknowledgements
17
This project was supported in part by Navy Work Unit 6000.RAD1.DA3.A0308 and the
18
United States Department of Agriculture, Agricultural Research Service CRIS projects 5325-
19
42000-047.
AC C
EP
TE D
M AN U
16
19
ACCEPTED MANUSCRIPT 1
Figures
P = 0.01
40
30
20
P = 0.05
RI PT
Percentage of total (%)
50
10
0 HS15/31 HS6/7 HS41 HS23/36c HS19
HS4c
HS2 HS1/44c HS31
2
HS5
HS37
HS10 HS8/17 HS53
Capsule genotype
HS3
HS15
HS42
HS22
mix
UT
Figure 1. Capsule genotypes in GBS/MFS- and enteritis associated strains. A multiplex PCR
4
was used to determine the capsular type of 37 GBS/MFS- (black bars) and 143 enteritis-
5
associated (white bars) C. jejuni strains derived from Dutch patients and 30 GBS-associated
6
strains derived from patients with GBS from Bangladesh (grey bars). With this PCR, 17 main
7
capsular types can be distinguished. Chi square and Fisher exact tests were used to compare
8
the frequency of specific capsular types between the Dutch GBS/MFS- patients and the
9
enteritis group. UT, untypable; mix, more than one capsular type detected. P ≤ 0.05 was
TE D
M AN U
SC
3
10
considered as statistically significant.
11
A
GBS/MFS NL (n = 37)
B
Enteritis NL (n = 143)
GBS BD (n = 30)
EP
P = 0.01
HS41
AC C
Capsule type
Others
UT
Capsule type
UT
HS23/36c
HS19
Others HS41 HS23/36c HS19
HS4c
HS4c
HS2
HS2
HS1/44c
HS1/44 A
B
C
D
E
A
B
LOS loci class
C
D
E
A
B
C
D
E
LOS loci class
12
20
ACCEPTED MANUSCRIPT Figure 2. Correlation of LOS class and capsular type. For each strain, the LOS class was
2
correlated to the respective capsular type. The six main capsular types observed in the
3
GBS/MFS-strain collection are depicted. A. C. jejuni strains from The Netherlands. The Chi
4
square tests were used to assess whether there were significant differences in LOS
5
class/capsular type combinations between the GBS/MFS- and enteritis strains. Hereto, the six
6
main capsular types were compered to others and UT. B. C. jejuni strains from BD. Others,
7
capsular types that did not belong to the group of six dominant capsular types observed in the
8
GBS/MFS-associates strain collections but that could be determined with the multiplex
9
capsule PCR; NL, The Netherlands; BD, Bangladesh; UT, untypable; P ≤ 0.05 was
12
AC C
EP
TE D
11
SC
considered as statistically significant.
M AN U
10
RI PT
1
13
Supplemental figure 1. Minimal spanning tree of C. jejuni isolates derived from sporadic
14
cases of human enteritis present in the Oxford MLST data base together with our GBS/MFS-
15
and enteritis-associated strains. Each circle presents one sequence type (ST). Circles of
21
ACCEPTED MANUSCRIPT increased diameter represent higher strain numbers within on ST. Dominant STs are
2
indicated. NL, The Netherlands; BD, Bangladesh.
AC C
EP
TE D
M AN U
SC
RI PT
1
22
SC
Disease- LOS Capsule Country class genotype GBS-BD B HS23/36c GBS-BD B HS23/36c B HS23/36c GBS-BD B HS23/36c GBS-BD B HS23/36c GBS-BD B HS23/36c GBS-BD B HS23/36c GBS-BD B UT GBS-BD B HS4c MFS-NL B HS2 MFS-NL GBS-NL B HS2 B HS23/36c Enteritis A HS19 GBS-NL A HS19 GBS-BD A HS1/44c GBS-NL A GBS-BD HS15/31c A HS19 GBS-NL A HS19 GBS-NL A HS19 Enteritis A HS19 Enteritis A HS19 Enteritis A HS19 Enteritis A HS19 Enteritis A HS19 GBS-BD A HS19 GBS-BD GBS-BD A HS19 GBS-BD A HS19 A HS31 Enteritis GBS-BD B HS23/36c GBS-BD B HS23/36c B MFS-NL HS23/36c GBS-BD B HS23/36c Enteritis C UT C GBS-NL HS2 Enteritis C HS2 Enteritis A HS2 C C HS2 Enteritis Enteritis C HS2 Enteritis C HS2 Enteritis C UT Enteritis C UT GBS-NL C HS1/44c GBS-NL A HS2 Enteritis A HS2 Enteritis A HS2 Enteritis C HS2 GBS-NL C HS1/44c GBS-NL C HS1/44c Enteritis C HS8/17c Enteritis C HS8/17c Enteritis C HS8/17c Enteritis C HS8/17c Enteritis C HS8/17c Enteritis C HS8/17c C HS23/36c GBS-NL C Enteritis HS23/36c A GBS-NL HS2 B MFS-NL HS10 Enteritis B HS4c C GBS-NL HS1/44c C Enteritis HS1/44c B GBS-BD HS15/31c C HS2 GBS-NL B Enteritis HS15 HS1/44c GBS-NL A GBS-NL B HS4c HS4c Enteritis B Enteritis B HS53 A HS41 GBS-BD A HS41 GBS-BD A HS41 GBS-BD A HS41 GBS-BD A HS41 GBS-BD A HS41 GBS-BD B HS2 Enteritis B HS2 GBS-NL C Enteritis HS3c B MFS-NL HS4c B Enteritis HS4c B Enteritis HS4c Enteritis B HS4c Enteritis A HS4c Enteritis B HS4c Enteritis B HS4c Enteritis B HS4c B GBS-NL HS4c B GBS-BD HS4c B GBS-NL HS2 Enteritis C HS8/17c B GBS-NL HS4c Enteritis C HS23/36c B HS4c GBS-NL Enteritis B HS4c B HS4c GBS-NL Enteritis B HS4c Enteritis B HS4c A HS4c GBS-NL Enteritis A HS4c Enteritis A HS4c Enteritis A HS4c Enteritis A HS4c Enteritis A HS4c Enteritis A HS4c GBS-NL A HS4c GBS-NL A HS4c GBS-NL A HS4c GBS-NL A HS4c GBS-BD B HS6/7c GBS-BD B HS6/7c GBS-BD A HS6/7c GBS-NL A HS4c MFS-NL A HS4c A GBS-BD HS4c Enteritis A HS4c
M AN U
MLST CC ST-403 ST-403 ST-403 ST-403 ST-403 ST-403 ST-403 ST-403 ST-403 single single single ST-22 ST-22 ST-22 ST-22 ST-22 ST-22 ST-22 ST-22 ST-22 ST-22 ST-22 ST-22 ST-22 ST-22 ST-22 ST-658 ST-42 ST-42 ST-42 ST-42 single ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-21 ST-206 ST-206 ST-21 ST-607 ST-52 single single ST-21 ST-21 ST-574 ST-21 ST-21 ST-21 ST-446 ST-362 ST-362 ST-362 ST-362 ST-362 ST-362 ST-464 ST-464 ST-353 ST-48 ST-48 ST-48 ST-48 ST-48 ST-48 ST-48 ST-48 ST-48 ST-48 ST-354 ST-607 ST-460 ST-206 ST-206 ST-206 ST-206 ST-206 ST-206 ST-206 ST-206 ST-206 ST-206 ST-206 ST-206 ST-206 ST-508 ST-508 ST-508 ST-508 single single single ST-61 ST-61 ST-61 ST-61
100
99
98
1
97
AC C
EP
TE D
MLST ST 985 985 6929 3219 3219 3219 3219 6938 83 131 131 380 149 6925 22 22 22 22 22 22 22 22 22 22 22 22 660 658 575 6928 42 42 986 53 53 53 53 53 53 53 53 148 148 21 21 21 19 50 50 50 50 50 50 50 46 46 883 607 2275 29 4292 6925 635 3015 186 185 104 875 587 587 587 587 587 6927 464 464 353 48 48 48 48 48 48 48 48 416 416 878 1707 606 4290 2104 572 572 572 572 122 122 122 122 122 122 122 508 508 64 132 3742 3742 3742 61 61 61 432
RI PT
ACCEPTED MANUSCRIPT
23
ACCEPTED MANUSCRIPT Figure3. Dendrogram of MLST genotype variation of C. jejuni strains with sialylated LOS
2
loci classes. The dendrogram was generated from MLST sequence type data of 58 GBS-, 5
3
MFS- and 51 enteritis-associated C. jejuni strains with either an LOS class A, B or C. ST,
4
sequence type; CC, clonal complex; GBS, Guillain-Barré syndrome; MFS, Miller Fisher
5
syndrome; NL, Netherlands; BD, Bangladesh.
AC C
EP
TE D
M AN U
SC
6
RI PT
1
24
