Content uploaded by Thomas Kesteman
Author content
All content in this area was uploaded by Thomas Kesteman on Feb 10, 2014
Content may be subject to copyright.
Acta Clinica Belgica, 2010; 65-5
319BORRELIA BURGDORFERI IN IXODES TICKS IN BELGIUM
PREVALENCE AND GENETIC HETEROGENEITY OF
BORRELIA BURGDORFERI SENSU LATO IN IXODES
TICKS IN BELGIUM
T. Kesteman1, C. Rossi2, P. Bastien3, J. Brouillard3, V. Avesani1, N. Olive1, P. Martin4, M.
Delmée1
1Microbiology Department, Université Cathololique de Louvain, Brussels, Belgium, 2Internal medicine
and Infectious diseases, Hôpital Ambroise Paré, Mons, Belgium, 3Rheumatology and Clinical Biology
Departments, Réseau Hospitalier de Médecine Sociale Ath-Baudour-Tournai, Ath, Belgium and
4Microbiology Department, Clinique Saint-Luc, Bouge, Belgium
Correspondence and offprint requests to: Thomas Kesteman, E-mail: kesteman@agora.eu.org
Original Article
ABSTRACT
Borrelia burgdorferi sensu lato is a genetically diverse
group of spirochetes that includes the agent of Lyme
borreliosis in which genospecies tend to be associated
with specifi c clinical features. The aim of the study was to
determine the prevalence and genetic diversity of Borrelia
burgdorferi sensu lato in 524 ticks collected in woods of a
western province of Belgium. Presence of spirochetes in ticks
was determined by phase contrast microscopy. The mean
infection rate of ticks was 12.0%. Variability was observed in
the prevalence of infection among the fi ve sites examined,
ranging from 2.8 to 21.6%. Identifi cation to genospecies
was determined by PCR and sequencing. The most common
genomospecies were Borrelia afzelii (55%) and Borrelia
garinii (21%). For the fi rst time in Belgium, we detected
Borrelia valaisiana and Borrelia spielmanii, representing
14% and 2%, respectively. Borrelia burgdorferi sensu stricto
counted only for 2%. Co-infections were present in 8% of
ticks. We emphasize the need for clinical studies to assess
the prevalence of specific genospecies-related clinical
manifestations of Lyme borreliosis in Belgium.
Key words : Borrelia, Lyme disease, Belgium, tick-borne disease,
Ixodes
INTRODUCTION
Borrelia burgdorferi sensu lato (s.l.), is a genetically diverse
species complex of, at least, ten genospecies that include the
tick-borne agent of Lyme borreliosis (LB), and is transmitted
mainly by the tick Ixodes ricinus in Europe. There is evidence
that, on one side, infections with different genospecies of Bor-
relia correlate with different clinical symptoms of LB, and that,
on the other hand, the risk of LB depends on the density of
the tick populations and on the infection rate of ticks by Bor-
relia (1,2). The geographic distribution of different genospecies
of Borrelia burgdorferi s.l. is thus of ecological, epidemiological
and clinical relevance. In south-western Belgium, clinicians are
concerned by an increasing trend in incidence of LB during the
last decade. The aim of the study was to determine the preva-
lence and the genetic diversity of Borrelia burgdorferi s.l. in ticks
in this region.
MATERIALS AND METHODS
From April to May 2004, 524 ticks were collected for analy-
sis in fi ve locations, within three woods of the western part
of the Belgian province of Hainaut. The woods of Bon-Secours,
Baudour and Stambruges were selected for their high frequen-
tation by hikers. Local vegetation is described in Table 1. Quest-
ing ticks were collected by dragging a ±1m2 white blanket in
lower vegetation.
Ticks were individually identifi ed at species level, dead ticks
were discarded and living ones were dissected under micro-
scope. Abdominal content was suspended in a drop of physi-
ological saline water and examined by Phase contrast Micros-
copy (PM), 400x for 3 to 5 minutes, to assess the presence of
spirochetes in the midgut. The rest of the abdominal content,
the chitinous parts of the animal and larger tissue particles
were kept frozen at –20°C for subsequent PCR analysis.
All ticks carrying spirochetes (n=63) were analyzed by both
end-point and real-time PCR. We also applied both PCR tech-
niques on a randomized control group of PM negative ticks
(n=63), matched in location, Ixodes species, stage and sex with
the 63 PCR positive ticks. DNA was fi rst extracted from the tick,
using the Dneasy Tissue Kit (QIAGEN, Hilden, Germany) fol-
320
Acta Clinica Belgica, 2010; 65-5
BORRELIA BURGDORFERI IN IXODES TICKS IN BELGIUM
lowing manufacturer’s instructions. DNA was diluted with AE
buffer (QIAGEN) to avoid effects of inhibitors.
End-point PCR was performed by amplifying Borrelia
DNA using primers 5’-AAT-AGG-TCT-AAT-AAT-AGC-CTT-AAT-
AGC-3’ and 5’-CTA-GTG-TTT-TGC-CAT-CTT-CTT-TGA-AAA-3’,
targeting a 308 bp segment of the ospA gene of common hu-
man pathogen Borrelia burgdorferi s.l. The reaction mixtures of
a total volume of 40 μL contained 5 μL Perkin Elmer Buffer II
(Applied BioSystems, Foster City, CA, USA), a 200 μM fi nal con-
centration of dNTP, a 2 mM fi nal concentration of MgCl2, a 0.2
μM fi nal concentration of primers and a 2.5 U fi nal concentra-
tion of Taq Gold PE (Applied BioSystems). A volume of 10 μl of
undiluted extracted DNA or 1:2 dilution of the DNA was then
added to the reaction mixture. DNA amplifi cation consisted of
incubation at 94°C for 10 min, fol lowed by 50 cycles of 94°C
for 1 min, 62°C for 1 min and 72°C for 1 min, and a fi nal in-
cubation at 72°C for 7 minutes, on a Gene Amp – PCR system
2400 (Applied BioSystems). PCR products were separated by
electrophoresis on 2% agarose gel followed by ethidium bro-
mide staining, and visualised under UV light.
Real-time PCR was performed using primers 5’-AGA-GTT-
CAT-GTG-GGA-GCA-AAT-CAA-3’ and 5’-TGA-GCA-CCC-TCT-
TGA-ACA-GGT-G-3’ selected from fl agellin gene of Borrelia
burgdorferi s.l. The expected product size was 128 bp. All reac-
tions were set up in a total volume of 50 μl containing 12.5 μL
of SYBR Green PCR Master Mix (Applied BioSystems), a 0.2 μL
fi nal concentration of primers and 5 μL of undiluted or 1:10
Table 1: Frequencies and percentage of ticks infected by spirochetes collected in Belgium
Geographical location Local vegetation Stage No. of ticks
collected and
analysed by PM
No. of ticks
positive
by PM (%)
Frequency of
nymphs
(/1m2 fl ag/hour)
Frequency of
infected nymphs
(/1m2 fl ag/hour)
Wood of Bon-Secours,
Location A
Beech stand with eagle
fern
Nymph 96 21 104.0 22,8
Adult 6 1
Total 102 22 (21.6)
Wood of Bon-Secours,
Location B
Forest ecotone Nymph 90 6 77.6 5,2
Adult 9 1
Total 99 7 (7.1)
Wood of Carnoy,
Stambruges
Mixed acidophilous oak
stand
Nymph 90 19 66.2 14,0
Adult 17 3
Total 107 22 (20.6)
Wood of Baudour,
Site of Croix Caillaux
Birch stand with eagle
fern
Nymph 76 5 48.0 3,2
Adult 33 4
Total 109 9 (8.3)
Wood of Baudour,
Site of Croix Martin
Acidophilous oak stand Nymph 96 3 66,2 2,1
Adult 11 0
Total 107 3 (2.8)
All sites Nymph 448 54 67.1 8.1
Adult 76 9
Total 524 63 (12.0)
dilution of extracted DNA. The real-time PCR was carried out
with a Gene Amp – 5700 Sequence Detection system (Applied
BioSystems) by using the following PCR parameters: initial de-
naturation step at 95°C for 10 min followed by 40 cycles of 15
s at 95°C and 1 min at 60°C. At the end of PCR, melting curves
were obtained by slowly increasing the temperature from 60°C
to 95°C and were used to check the specifi city of the signal.
Each positive PCR amplicon was purifi ed by the QIAquick
purifi cation Kit (QIAGEN), following the manufacturer’s in-
structions. The ABI Prism Big Dye Terminator Cycle Sequencing
Ready Reaction Kit (Applied BioSystems) was used for sequenc-
ing the PCR product. The sequencing reaction and template
preparation were performed in accordance with the instruc-
tions of the manufacturer (Applied BioSystems). Sequencing
reactions were performed with corresponding primers specifi c
for ospA and fl agellin genes used for the previous amplifi ca-
tion. The sequencing product was purifi ed by Ethanol/Sodium
Acetate Precipitation method (Applied BioSystems) and se-
quence analysis was realized with a 3100 Automatic Sequenc-
er (Applied BioSystems). Each nucleotide sequence was then
compared to those available on genome data base (Genbank,
www.ncbi.nlm.nih.gov) by multiple sequence alignment using
the Blast software, to confi rm the identifi cation and to deter-
mine the genospecies. Ticks were considered as PCR positive if
sequencing confi rmed presence of a known Borrelia genome.
Acta Clinica Belgica, 2010; 65-5
321BORRELIA BURGDORFERI IN IXODES TICKS IN BELGIUM
Differences in tick infection prevalence and in prevalence
of genospecies between sexes, stage, localisation or between
studies were statistically analysed by chi-square test with a
threshold of 0.05. A comparison of detection methods was per-
formed using McNemar and Cohen’s kappa tests. The thresh-
old for signifi cance of symmetry in the McNemar test was
0.05; Cohen’s ko <0.4 was considered as non concordant, 0.4
≤Cohen’s ko <0.75 was considered as good concordance and
Cohen’s ko ≥0.75 was considered as excellent concordance. A
linear relation between median collecting time per tick and in-
fection prevalence was performed by Spearman test.
RESULTS
A total of 524 questing ticks were collected and analysed
by PM, and 63 (12.0%) were found to be infected with spiro-
chetes. Signifi cant variability was observed in the prevalence
of infection among the fi ve sites examined (Table 1), ranging
from 2.8 to 21.6 % (p<0.0001). We identifi ed 520 Ixodes ricinus
and four Ixodes (Pholeoixodes) hexagonus. No I. hexagonus was
found to contain Borrelia.
Collect frequency of nymphs (F) and infected nymphs (Fi),
as calculated by Hubálek et al., were determined for our tick
population (3). Average F was 67.1 nymphs/1m2 fl ag/hour and
average Fi was 8.1 infected nymphs/1m2 fl ag/hour. F varied
from 48.0 to 104.0 nymphs/1m2 fl ag/hour and Fi varied from
2.1 to 22.8 infected nymphs/1m2 fl ag/hour; those differences
are statistically signifi cant (p<0.0001). The difference between
rate of infection in adults (11.8 %,) and in nymphs’ ticks (12 %)
was not statistically signifi cant, as well as difference between
rate of infection in male (15%) and female (10%) adults (both
p>0.1).
Comparison between PM and both PCR techniques showed,
on the one hand, that six of the 63 (9.5%) ticks carrying spiro-
chetes as observed by PM, were negative by PCR. On the other
hand, nine of the 63 (14.3%) selected ticks which were nega-
tive by PM were shown to contain Borrelia DNA by PCR.
By DNA sequencing of the 66 PCR positive ticks, we identi-
fi ed 36 Borrelia afzelii (55%), 14 Borrelia garinii (21%), 9 Borrelia
valaisiana (14%), 1 Borrelia spielmanii (2%), 1 Borrelia burgdor-
feri sensu stricto (2%) and 5 co-infections (8%).
DISCUSSION
The prevalence of Borrelia infection in ticks is one of the
most essential components of risk assessment for human LB
(2). Observation by PM of the midgut of each of the 524 ticks
collected in western Belgium showed that 12.0 % of them car-
ried spirochetes. Compared with other available Belgian studies,
our value is statistically different from one undertaken in the
Philippeville region in 1996 which showed 23% of ticks carry-
ing Borrelia (p<0.0001), but not from two others undertaken in
the Meuse Valley before 1990, a region known for high LB preva-
lence, where 11% and 12% of ticks were found to contain Bor-
relia (both p>0.1) (4,5). Our results are in accordance with the
overall mean infection rate of 13.6% reported in a recent meta-
analysis of Borrelia infection of ticks in Europe (6), and compari-
son of infection rates show no signifi cant difference (p>0.1).
Hubálek et al. described that the frequency of nymphs (F)
and the frequency of infected nymphs (Fi) is much more corre-
lated with LB incidence than infected ticks’ percentage (3). Un-
fortunately, our values cannot be compared to any other study;
to our knowledge, no other author published any results of F
and Fi obtained with a 1m2 fl ag. In our study, F and Fi statisti-
cally vary from one place to another neighbouring place. This
means that for a whole region, F and Fi must not be defi ned
from a single location.
The study of Rauter et al. reports a signifi cantly higher in-
fection rate in adult ticks than in nymphs, with an increasing
trend from Western to Eastern Europe (6). In our study, the in-
fection rate in adult ticks was not signifi cantly different than
in nymphs. This could be explained by the location of sampling
sites in the Western part of Europe.
There is strong evidence that infections with different geno-
species of Borrelia correlate with different clinical symptoms of
LB (1). Therefore, knowledge of the geographic distribution of
different genospecies of Borrelia burgdorferi sensu lato within
their tick vector is of ecological, epidemiological and clinical
relevance. This study is the second epidemiological report of
the distribution of genospecies of Borrelia burgdorferi s.l. in
ticks in Belgium, and the fi rst one to identify Borrelia burgdor-
feri s.l. species with broad-spectrum PCR.
Our genospecies distribution is radically different from the
single previous Belgian study, which showed that Borrelia garinii
was predominant (53%), Borrelia afzelii much rarer (9%) and
Borrelia burgdorferi sensu stricto much more frequent (38%)
(4). The differences in genospecies distribution between our
study and the one by Misonne et al. could be due to changes in
ecological conditions or diversity in reservoir hosts. Neverthe-
less, our genospecies distribution was statistically similar to a
study conducted in France, close to the Belgian border, in 2002
(p>0.1) (7). Finally, no statistical difference could be found
when comparing our results with those of a large-scale Euro-
pean study (p>0.1) (8).
This is the fi rst description of the presence of Borrelia val-
aisiana and Borrelia spielmanii in Belgium. In contrast with a re-
stricted number of clinical descriptions, Borrelia valaisiana and,
in a more limited way, Borrelia spielmanii have been found in a
large number of ticks in various countries. This phenomenon
may be associated to a lower pathogenicity but, contrarily to
Borrelia afzelii, Borrelia garinii, and Borrelia burgdorferi sensu
stricto, pathogenicity of Borrelia valaisiana or Borrelia spielma-
nii remains to be characterised. Another hypothesis could be
that screening serology tests may lack sensitivity for detection
of anti-Borrelia valaisiana or anti-Borrelia spielmanii antibodies,
but the performances of those tests in infections by Borrelia
valaisiana or Borrelia spielmanii have not yet been studied.
Five of the 524 examined ticks were infected by more than
one Borrelia burgdorferi s.l. genospecies. Among those co-in-
fections, we identifi ed two associations of Borrelia garinii with
Borrelia valaisiana. This fi nding is consistent with various stud-
ies showing that there are specifi c, although non-exclusive, as-
sociations between Borrelia species and reservoir hosts (6,9).
Unexpectedly, Borrelia spielmanii was implicated in the three
other identifi ed co-infections, one with Borrelia afzelii, and two
with undetermined species of Borrelia burgdorferi s.l. Since Bor-
relia afzelii is preferentially hosted by small rodents, and Bor-
relia spielmanii by the garden dormouse, this association might
322
Acta Clinica Belgica, 2010; 65-5
BORRELIA BURGDORFERI IN IXODES TICKS IN BELGIUM
be explained by a common resistance to the serum comple-
ment of a rodent host (9,10).This study also compared PM and
PCR abilities to detect Borrelia in ticks. On the one hand, six of
the 63 ticks shown to carry spirochetes by PM were constantly
negative by both techniques of PCR, despite typical observa-
tion by PM (high amount of spirochetes, mobility). This can-
not be explained by the presence of inhibitors, since we could
amplify tick genome from DNA extracts (data not shown). This
means that the primers we used failed to recognise Borrelia
species – or another Borrelia-shaped mobile spirochete. On the
other hand, nine of the 63 ticks in which no spirochete had
been observed by PM were shown to contain Borrelia DNA. As
no reference technique has ever been described, neither sen-
sitivity nor specifi city could be calculated for PM and PCR.
Nonetheless, Cohen and McNemar’s statistical method allows
a reliable comparison of those techniques. In the group of ticks
studied by both techniques, PM and PCR show an excellent
concordance and a symmetry in their respective results (Co-
hen’s ko=0.76 and McNemar’s p>0,1). Thus, the comparison of
PM and PCR techniques is biased in our study, but shows similar
characteristics of optical and molecular methods.
In conclusion, the results enhance the genetic heterogene-
ity of Borrelia described in Belgium and emphasize the need for
clinical studies to assess the prevalence and specifi c genospe-
cies-related clinical manifestations of LB in Belgium. Our ap-
proach, using PM followed by PCR and sequencing of positive
PM ticks, associates a low cost screening of Borrelia carriage by
ticks with a high precision in genotyping.
ACKNOWLEDGEMENTS
We thank Mr Caulier J.M. for his invaluable help in collect-
ing ticks. We also thank Dr Beguin C. for her precious statisti-
cal advice and assistance and Pr Glupczynski Y. for his helpful
redaction advice.
REFERENCES
1. van Dam AP, Kuiper H, Vos K, et al. Different genospecies of Borrelia burg-
dorferi are associated with distinct clinical manifestations of Lyme borrelio-
sis. Clin Infect Dis 1993; 17: 708-17
2. Gray J. Risk assessment in Lyme borreliosis. Wien Klin Wochenschr 1999;
111: 990-993
3. Hubálek Z, Halouzka J, Juricova Z. Longitudinal surveillance of the tick Ixo-
des ricinus for borreliae. Med Vet Entomol 2003; 17: 46-51
4. Misonne MC, Van Impe G, Hoet PP. Genetic heterogeneity of Borrelia burg-
dorferi sensu lato in Ixodes ricinus ticks collected in Belgium. J Clin Microbiol
1998; 36: 3352-4
5. Martin P, Bigaignon G, Gillion P, Thirion A, Fain A. Fréquence de Borrelia
burgdorferi (maladie de Lyme) et répartition de son vecteur Ixodes ricinus
(Acari:Ixodidae) dans le district mosan en Belgique. Bull Soc Fr Parasitol
1990; 8: 331-8
6. Rauter C, Hartung T. Prevalence of Borrelia burgdorferi sensu lato genospe-
cies in Ixodes ricinus ticks in Europe: a metanalysis. Appl Environ Microbiol
2005; 71: 7203-16
7. George JC, Chastel C. Tick-borne diseases and changes in the ecosystem in
Lorraine. Bull Soc Pathol Exot 2002; 95: 95-9
8. Kurtenbach K, De Michelis S, Sewell HS, et al. Distinct combinations of Bor-
relia burgdorferi sensu lato genospecies found in individual questing ticks
from Europe. Appl Environ Microbiol 2001; 67: 4926-9
9. Kurtenbach K, Sewell HS, Ogden NH, Randolph SE, Nuttall PA. Serum com-
plement sensitivity as a key factor in Lyme disease ecology. Infect Immun
1998; 66: 1248-51
10. Richter D, Schlee DB, Allgöwer R, Matuschka FR. Relationships of a novel
Lyme disease spirochete, Borrelia spielmani sp. nov., with its hosts in central
Europe. Appl Environ Microbiol 2004; 70: 6414–9
