Borrelia burgdorferi infection prevalences in questing Ixodes ricinus ticks (Acari: Ixodidae) in urban and suburban Bonn, western Germany
- First Online:
- Cite this article as:
- Maetzel, D., Maier, W.A. & Kampen, H. Parasitol Res (2005) 95: 5. doi:10.1007/s00436-004-1240-3
- 161 Views
From March to October 2003, a total of 2,518 host-seeking Ixodes ricinus ticks (1,944 nymphs, 264 females, 310 males) were collected by blanket dragging at 45 sites all over the city area of Bonn, western Germany, to be checked for Borrelia burgdorferi infection. The collection sites included 20 private gardens, nine public recreational parks, the boundaries of 14 sylvatic suburban areas and two footpaths between suburban farmed fields. Generally, numbers of specimens collected along sylvatic suburban areas and at urban sites with dense tree populations were significantly higher than at the other collection sites. Out of 1,394 specimens (865 nymphs, 241 females, 288 males) that were randomly chosen for Borrelia analysis by a simple PCR, 250 (17.9 %) were found to be infected with B. burgdorferi sensu lato. While the infection prevalences varied significantly between females (26.6%), males (12.5%) and nymphs (17.3%), there were no striking differences between sylvatic and unwooded sites. A total of 92.8% of the ticks Borrelia-positive by the simple PCR were also positive in a diagnostic nested PCR. Using genospecies-specific oligonucleotide probes, single Borrelia genospecies infections (91.4%) could be assigned to B. afzelii (39.5%), B. garinii (27.9%), B. burgdorferi sensu stricto (15.6%) and B. valaisiana (8.6%) by DNA hybridization. Various combinations of double infections were observed in 4.3% of the infected ticks. Another 4.3% of the Borrelia infections were untypeable. The B. burgdorferi genospecies distribution in the city area was shown to be variable from site to site and, even more, it was distinct from rural collection sites near Bonn. This is ascribed to a different spectrum of reservoir hosts. Taking into account the infection prevalences of host-seeking ticks in the forested surroundings of Bonn, our study demonstrates that the risk of acquiring Lyme disease after a tick bite in urban/suburban areas is comparably as high as in woodlands outside of the city.
Lyme disease is one of the most important vector-borne diseases in North America and Europe (Steere 2001). It is caused by spirochete bacteria from the Borrelia burgdorferi genospecies complex that are principally transmitted by ticks of the Ixodes ricinus complex (Jaenson 1991). Lyme borreliosis is a multisystemic disease often provoking dermatological, neurological, cardiac and arthritic symptoms (Steere 2001). In Europe, the infection is most common in central and eastern countries where the annual incidence rates vary between 40 and 300 cases per 100,000 inhabitants (World Health Organization 1995). In Germany, 40,000–80,000 cases of Lyme disease are estimated to occur per year (Maiwald et al. 1996); however, as in most other European countries, a national reporting and surveillance system is lacking.
The B. burgdorferi genospecies complex consists of at least 13 genospecies and genomic groups (Hubálek and Halouzka 1997) that are distributed worldwide but are mainly restricted to the northern hemisphere. At least three of these are correlated with distinct clinical manifestations: B. afzelii with acrodermatitis chronica atrophicans (ACA), B. burgdorferi sensu stricto (s.s.) with arthritis, and B. garinii with neuroborreliosis (van Dam et al. 1993; Balmelli and Piffaretti 1995). The clinical relevance of B. bissettii and B. valaisiana is still unclear (Picken et al. 1996; Strle et al. 1997; Wang et al. 1997; Schaarschmidt et al. 2001; Diza et al. 2004). In Germany, four genospecies have been recorded: B. garinii, B. afzelii, B. burgdorferi s.s. and B. valaisiana (Hubálek and Halouzka 1997; Kurtenbach et al. 2001; Kampen et al. 2004).
More than 50 avian and mammalian vertebrate species are involved in the zoonotic transmission cycle as reservoir hosts for the spirochetes (Gern et al. 1998). The most important reservoir hosts are small rodents (e.g. the bank vole, Clethrionomys glareolus, and the yellow-necked mouse, Apodemus flavicollis) and passerine birds (Kurtenbach et al. 1998b, 2002a; Hanincová et al. 2003b).
Tick infection prevalences vary depending on the region and the developmental stage of the ticks. In Europe, the highest percentages of larvae found positive for B. burgdorferi were 10.8% (Ameland, The Netherlands; Rijpkema et al. 1994), 42.9% for nymphs (Valais canton, Switzerland; Péter et al. 1995) and 58.3% for adults (south Poland; Sinski, cited in Hubálek and Halouzka 1997). However, most studies on tick infection prevalences concentrate on rural and sylvatic wildlife habitats in which an intense zoonotic transmission of borreliae between ticks and their vertebrate hosts can be expected. By contrast, few studies deal with the situation within urban and suburban areas (e.g. Guy and Farquhar 1991; Schwartz et al. 1991; Hubálek et al. 1993; Junttila et al. 1999; Michalik et al. 2003). We therefore made an attempt to assess tick distributions in the city area of Bonn, to determine the infection prevalences of the ticks with B. burgdorferi and thus to calculate the risk of acquiring ‘urban’ Lyme disease compared to the risk of becoming infected in the close-by natural park Siebengebirge, where epidemiological data had been collected for several years (Kurtenbach et al. 1995; Kampen et al. 2004).
Materials and methods
Study area and tick collection
For DNA extraction the ticks were homogenized in 100 µl 1.25% ammonia solution and then boiled at 100°C for 20 min (Guy and Stanek 1991). Afterwards the ammonia was evaporated at 100°C with the lid of the tube open, and the volume of the homogenate was finally adjusted to 50 µl by adding H2O. The DNA solutions were stored at −20°C.
B. burgdorferi detection
Using 3–5 µl of the DNA solution, two different PCR approaches were applied to detect Borrelia infected ticks: a simple PCR (Schwartz et al. 1992) for rapid sample screening and a nested PCR (Rijpkema et al. 1995) for subsequent genotyping of detected borreliae.
The simple PCR protocol was modified as follows: reaction mixtures contained 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 20 mM MgCl2, 200 µM dNTPs, 200 nM of each primer (JS1, JS2; for oligonucleotide sequences see Schwartz et al. 1992), 1 unit of Taq DNA-polymerase (Invitrogen) and H2O to adjust the volume to 50 µl. The thermoprofile included an initial denaturation step of 1.5 min at 94°C, 45 cycles of 30 s at 94°C, 20 s at 50°C and 20 s at 72°C, and a final extension step of 2 min at 72°C.
Only B. burgdorferi s.l. positive ticks were further processed by a nested PCR. The reaction mixtures for both the first and the second nested PCRs were composed of 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 20 mM MgCl2, 200 µM dNTPs and 400 nM of each primer (23SC1/23SN1 for the first nested PCR, 23SC2/23SN2 for the second nested PCR; for oligonucleotide sequences see Rijpkema et al. 1995). The first nested PCR, which was carried out in a total volume of 25 µl, also contained 0.75 units of Taq DNA-polymerase (Invitrogen) and 3 µl of DNA solution. To the second nested PCR, which was performed in 50 µl volume, 1.5 units Taq DNA-polymerase and 1.5 µl solution of the reaction mixtures that had passed the first nested PCR were added. The thermoprofiles for the two PCRs were also different from each other. Both of them started with an initial denaturation step for 1 min at 94°C. The first nested PCR was continued by 25 cycles of 30 s at 94°C, 30 s at 52°C and 1 min at 72°C, and finalized by an extension step of 2 min at 72°C. The second nested PCR comprised 40 cycles of 30 s at 94°C, 30 s at 55°C, and 1 min at 72°C, which was followed by a final extension step of 2 min at 72°C.
As positive controls, samples including 1 µl of a DNA solution prepared from a B. garinii laboratory culture were used. Within this volume, the equivalent of three Borrelia cells was added, which were shown to be detectable with both PCR approaches in preliminary experiments. For negative controls, distilled water was added.
PCR products were separated electrophoretically in 1.5% agarose gels and visualized under UV light after ethidium-bromide staining for 15 min at 0.5 µg/ml.
Typing of B. burgdorferi genospecies by DNA hybridization
Nested PCR products obtained from the ticks positive in the simple PCR were further processed by DNA hybridization. Oligonucleotide probes (for nucleotide sequences see Rijpkema et al. 1995) specific for the four genospecies of the B. burgdorferi complex known to occur in Germany (B. afzelii, B. garinii, B. burgdorferi s.s., B. valaisiana) and a complex-specific probe (B. burgdorferi s.l.) were used. For each probe, a separate gel was run followed by DNA blotting to a nylon membrane (Roti-Nylon plus transfer membrane, Roth, Germany). Hybridization was visualized by means of the DIG luminescent detection kit (Roche) according to the manufacturer’s instructions.
PCR products of samples with missing or unclear hybridization signals were sequenced. Amplicons were excised from agarose gels and purified using the DNA gel extraction kit (Qiagen). From each PCR product, both strands were sequenced by cycle sequencing on an ABI Prism 310 DNA sequencer (Applied Biosystems). PCR primers were used as sequencing primers.
Differences in tick infection prevalences between locations and sexes were analyzed by means of the χ2-test and Fisher’s exact test.
Numbers of Ixodes ricinus ticks collected and found infected with Borrelia burgdorferi sensu lato by simple PCR according to Schwartz et al. (1992)
No. of ticks collected
No. of nymphs positive/examined
No. of adults positive/examined
Total no. of ticks per site positive/examined
No. of ticks collected (nymphs/adults)
250/1,394 (17.9) (nymphs: 150/865 (17.3), adults: 100/529 (18.9))
Boundaries of forested outskirt areas
A total of 1,394 randomly selected ticks (865 nymphs, 241 females, 288 males) from all collection sites were examined for B. burgdorferi s.l. by a simple PCR. Of these, 250 were found to be infected (Table 1), and 233 (93.2%) were also Borrelia-positive in a nested PCR, whereas 17 (6.8%) generated no PCR product.
There were no significant differences in the tick infection prevalences between the collection sites (χ2=4.662; df=2; P=0.097), e.g. with regard to the kind of vegetation present. However, the infection prevalences varied significantly according to the developmental stage (χ2=18.26; df=2; P<0.0001). Some 26.6% (64/241) of the investigated females, 12.5% (36/288) of the males, and 17.3% (150/865) of the nymphs were Borrelia-positive.
B. burgdorferi genospecies distribution
B. burgdorferi s.l.
B. burg-dorferi s.s.
B. garinii/B. valaisiana
B. afzelii/B. burgdorferi s.s.
B. valaisiana/B. burgdorferi s.s
The prevalence of the various B. burgdorferi genospecies varied significantly according to the different tick collection sites (χ2=211.90; df=162; P<0.004). A significant accumulation of B. garinii infections (residuals=2.3) was detected in a park (site 1), of B. valaisiana (residuals=3.9) in another park (site 2), of B. afzelii (residuals=3.7) in a garden (site 10), and of B. burgdorferi s.s. (residuals=4.1 and 2.3, respectively) along two sylvatic suburban areas (sites 26, 34). By contrast, significant deficiencies of infections with B. afzelii and B. burgdorferi s.s. (residuals<2) were detected in ticks from a park (site 2) and from two outskirt areas (sites 22, 26), respectively.
While there is a considerable body of epidemiological data on the distribution of ticks and their infection prevalences with B. burgdorferi sensu lato (s.l.) in unsettled rural and sylvatic regions, few studies deal with the situation in urban and suburban areas. This gives rise to the impression that there are no or few ticks in these areas and that the infection potential for borreliosis in housing estates is negligible. However, the present study confirms the hypothesis that there is a risk of becoming infected with B. burgdorferi via a tick bite wherever there are ticks, even in urban areas. According to our study, the risk in such urban areas occurs irrespective of the kind of habitat the tick was picked up in. On the other hand, most likely due to microclimatic conditions and host availability, sites with some trees and undergrowth have higher tick abundances than others, providing more Borrelia-infected ticks.
Ticks were found in two out of nine public parks, seven out of 20 private gardens, and all 14 suburban sites examined (Fig. 1). The two tick-infested parks differed distinctly in their vegetation from the tick-free parks. One of the infested parks (site 1) covered an area of about 3 ha and was completely surrounded by housing estates. Nevertheless, the vegetation was forest-like. Rodents, birds and other small vertebrates must be considered as the main tick hosts. By contrast, the other tick-infested park (site 2) was located close to a forested area representing an urban enclave of a larger woodland area outside of Bonn. Furthermore, the park itself had some trees present. Until recently, large game such as deer and wild boar could enter this park and obviously brought along huge quantities of fed nymphs. As a fence had been built prior to the winter season in the year before the tick collections, tremendous numbers of questing adults appear to have accumulated without finding acceptable hosts (cf. Pretzmann et al. 1964; Daniel et al. 1977). The complete absence of ticks in the seven other parks in which trees and undergrowth were scarce was probably caused by unfavorable microclimatic conditions and/or a lack of suitable hosts (cf. Daniel et al. 1977; Walter and Liebisch 1980; Liebisch and Liebisch 1997).
The seven private gardens in which ticks were collected were suburban and also adjacent to sylvatic areas or close to such areas. In two gardens and their direct neighborhood (sites 10, 13) deer was observed which may have contributed to the higher tick densities determined at these sites in comparison to the other gardens. The seven gardens in which no ticks were detected were both urban and suburban but had no contact with sylvatic areas.
Ticks were detected at high densities in all 14 study sites along the forested outskirts, which function as local recreation areas for citizens. The continuous availability of numerous tick hosts (deer, wild boar, rodents, hedgehogs, lagomorphs, etc.) appear to be the major reason for the high tick densities there (Liebisch and Liebisch 1997).
In general, in addition to microclimate, the tick-host biocenosis plays an important role in local tick density. Due to their size and radius of activity, larger animals usually collect far more ticks than smaller ones. However, for physiological reasons, some vertebrate species are better tick hosts than others. Voles, for example, may acquire resistance to ticks in a density-dependent way (Kurtenbach et al. 1994; Randolph 1994; Dizij and Kurtenbach 1995). This, in turn, reduces the survival rates of the ticks feeding on them. Sites with similar vegetation and climatic conditions that differ in their tick abundances must, therefore, be assumed to be characterized by the vertebrate biocenoses diverging from each other.
The ticks collected in urban and suburban Bonn were examined for B. burgdorferi infections by two different PCR approaches, a simple one and a nested one followed by DNA hybridization as suggested by Oehme et al. (2002) and Kampen et al. (2004). The two PCR methods applied obviously differed in their sensitivities. Out of the 250 samples positive by the simple PCR, only 233 (93.2%) were also positive by the nested PCR. In a multicenter study with different simple and nested PCR approaches, 87.5% concordance was reached (Kahl et al. 1998), and in a study comparable to this present one the degree of correspondence between the simple and nested PCRs was 94.4% (Kampen et al. 2004). However, while the latter study made use of the same nested PCR, the simple PCR was different from the one used here. Moreover, in that study, most samples were tested both by the simple and the nested PCR, whereas in this study only samples positive by the simple PCR were subsequently examined by the nested PCR. It must be assumed that some samples negative by the simple PCR would have been positive by the nested PCR had this been applied. Although nested PCR approaches are generally considered to be more sensitive than simple PCR approaches, they are particularly susceptible to contamination and thus prone to produce more false positives than simple PCR techniques (Kahl et al. 1998).
On average, the tick infection prevalence obtained for urban and suburban Bonn amounted to 17.9%. This percentage is distinctly lower than that calculated for Helsinki (32%, Junttila et al. 1999) or for Brno (27%, Hubálek et al. 1993), but higher than for recreational areas in the city of Poznan (16.2%, Michalik et al 2003). In two parks in Richmond (a borough of London) the infection prevalences were 5% and 12%, respectively (Guy and Farqhar 1991). Surprisingly, the infection prevalence in Bonn is even higher than that determined in 2001 for the close, densely forested Siebengebirge park (14%, Kampen et al. 2004), where a far more intense contact between ticks and potential B. burgdorferi reservoir hosts must be expected. However, when comparing such figures it is necessary to know the numbers of the various developmental stages included. In the present study, for example, the proportion of nymphs (62.1%) was lower than in the Siebengebirge study (67.2%), and the ratio of females to males was 1:1.2 for Bonn and 1:1.5 for the Siebengebirge (H. Kampen unpublished data). As a higher percentage of females are infected than nymphs, due to their additional blood meal and coeval chance of taking up borreliae, and male infection prevalences often only exceed those of nymphs by a small amount, it becomes obvious that studies mainly including females do not lead to the same result as studies mainly including nymphs or males, even when the infection prevalences of the different developmental stages are exactly the same in the areas compared. Therefore, pure average population infection prevalences are difficult to assess, and it is preferable to refer to a certain developmental stage.
This is supported by analysis of the infection prevalences which showed no significant differences when nymphs and adults (χ2=0.128, df=1, P=0.720) were compared. However, significant differences occurred when the adult stages were considered separately (simple PCR: χ2=18.256, df=2, P<0.0001) due to a considerably lower infection prevalence of males (12.5%) compared to females (26.6%). Although this phenomenon has been observed before (Matuschka et al. 1992; Hubálek et al. 1993; Kampen et al. 2004), a satisfying explanation has never been found. Only Matuschka et al. (1992) speculated that the smaller bloodmeals taken by subadults destined to become males often do not contain enough borreliae to establish an infection in the tick.
Although analysis of the infection prevalences showed no overall significant differences in terms of individual tick collection sites (i.e. infection prevalences at sylvatic sites were not significantly different from those at unwooded sites), it turned out that the infection prevalences of some unwooded sites were in fact significantly different from other unwooded sites (χ2=66.46, df=6, P<0.0001), in contrast to the group of sylvatic sites and sites with dense tree populations (χ2=22.95, df=16, P<0.115). Just as with tick abundance, this finding may also be due to variations in the tick-host biocenosis. In fact, it is well known that the tick hosts have an impact on infection prevalences with B. burgdorferi (Kurtenbach et al. 1994, 1998a, 1998b; Dizij and Kurtenbach 1995; Matuschka et al. 1993, 1996) as they differ in their reservoir host competences. For example, voles transmit spirochetes to ticks much more efficiently than mice (Kurtenbach et al. 1994, 1998b), and the presence of rats positively influences the urban tick infection prevalences as they are much more infective to ticks than mice (Matuschka et al. 1996). Tälleklint and Jaenson (1993) stressed the role of lagomorphs in maintaining Lyme disease spirochetes at insular sites where rodents are absent and argued that fluctuations in lagomorph populations may influence the number of B. burgdorferi-infected ticks.
The observed distribution of the various B. burgdorferi complex genospecies is in agreement with the results of earlier studies conducted in Germany and other European countries (Hubálek and Halouzka 1998; Kurtenbach et al. 2001; Schaarschmidt et al. 2001) with B. afzelii as the predominant genospecies and B. valaisiana as the genospecies least represented. By contrast, the percentages of the genospecies distributed in urban/suburban Bonn differ clearly from that in the Siebengebirge. There, B. valaisana was, surprisingly, by far the most frequently found genospecies (60.7% of all infections, Kampen et al. 2004). Again, the composition of the reservoir host population is to be held responsible. For example, there is strong evidence that rodents are transmission-competent for B. afzelii, but not for B. garinii or B. valaisiana, and that the uptake of rodent blood even results in the elimination of the latter genospecies in the tick (Kurtenbach et al. 1998a, 1998b, 2002a, 2002b; Hanincová et al. 2003a, 2003b). Further studies suggest avian reservoirs for B. garinii and B. valaisiana (Kirstein et al. 1997; Kurtenbach et al. 1998b, Hanincová et al. 2003b), whereas deer and other ruminants are considered reservoir-incompetent for all B. burgdorferi genospecies (Telford et al. 1988; Jaenson and Tälleklint 1992; Matuschka et al. 1993). Accordingly, the significant correlation between collection site and genospecies in Bonn (χ2=210.56; df=144; P<0.0001) suggests differences in the composition of the tick host populations. Birds are presumably the main hosts at the sites where B. valaisiana and B. garinii are dominant (e.g. sites 1, 2), and rodents at the sites where B. afzelii and B. burgdorferi s.s. are most commonly detected (e.g. sites 10, 34; Kurtenbach et al. 1998a, 1998b). While deer may influence the tick infection prevalences in suburban areas, rabbits and rats, both of which reach huge urban population densities, may be the main contributors to urban infection prevalences.
We have demonstrated that the tick population of the city area of Bonn, irrespective of the tick density, is infected with B. burgdorferi s.l. to a degree comparable to rural woodland areas. Given an average infection prevalence of 26.6% for females and 17.3% for nymphs, and based on the premise that only every fourth bite of a B. burgdorferi-infected tick leads to an infection in humans (Zappe et al. 2002), the potential risk of contracting Lyme borreliosis in urban/suburban Bonn is about 6.7% after a bite by a female I. ricinus tick and 4.3% after a bite by a nymph.
Our results suggest that even in urban/suburban areas it is necessary to have a closer look at the local vertebrate biocenosis in epidemiological studies on Lyme borreliosis, as has exemplarily been done by Michalik et al. (2003). Since the vertebrate biocenosis provides the tick hosts, and at the same time the reservoir hosts for B. burgdorferi, it has a major impact on three epidemiological factors: (1) tick abundance, (2) the general infection prevalence of tick populations with B. burgdorferi, and (3) the spatial distribution of certain B. burgdorferi genospecies.
The authors are indebted to Rainer Oehme and Kathrin Hartelt (Landesgesundheitsamt Baden-Württemberg, Stuttgart, Germany) for providing control aliquots of the different B. burgdorferi genospecies and for making available DNA sequencing facilities.