Molecular evidence of Rickettsia spp. in ixodid ticks and rodents in suburban, natural and rural habitats in Slovakia

Natural foci of tick-borne spotted fever group (SFG) rickettsiae of public health concern have been found in Slovakia, but the role of rodents in their circulation is unclear. Ticks (Ixodes ricinus, Ixodes trianguliceps, Dermacentor marginatus, Dermacentor reticulatus, Haemaphysalis concinna and Haemaphysalis inermis) and tissues of rodents (Apodemus flavicollis, Apodemus sylvaticus, Myodes glareolus, Microtus arvalis, Microtus subterraneus and Micromys minutus) were examined for the presence of SFG rickettsiae and Coxiella burnetii by molecular methods. Suburban, natural and rural habitats were monitored to acquire information on the role of ticks and rodents in the agents’ maintenance in various habitat types of Slovakia. The overall prevalence of rickettsial infection in questing I. ricinus and D. marginatus was 6.6% and 21.4%, respectively. Rickettsia helvetica, R. monacensis and non-identified rickettsial species were detected in I. ricinus, whereas R. slovaca and R. raoultii were identified in D. marginatus. Rickettsia spp.-infected I. ricinus occurred during the whole tick questing period. Rickettsia helvetica dominated (80.5%) followed by R. monacensis (6.5%). The species were present in all studied habitats. Rickettsia slovaca (66.7%) and R. raoultii (33.3%) were identified in D. marginatus from the rural habitat. Apodemus flavicollis was the most infested rodent species with I. ricinus, but My. glareolus carried the highest proportion of Rickettsia-positive I. ricinus larvae. Only 0.5% of rodents (A. flavicollis) and 5.2% of engorged I. ricinus removed from My. glareolus, A. flavicollis and M. arvalis were R. helvetica- and R. monacensis-positive. Coxiella burnetii was not detected in any of the tested samples. We hypothesize that rodents could play a role as carriers of infected ticks and contribute to the maintenance of rickettsial pathogens in natural foci. Long-term presence of SFG Rickettsia spp. was confirmed in questing ticks from different habitat types of Slovakia. The results suggest a human risk for infection with the pathogenic R. helvetica, R. monacensis, R. slovaca and R. raoultii.

Small rodents are important hosts for immature stages of ixodid ticks and are considered, e.g., as reservoirs of tick-borne encephalitis virus [11], Borrelia miyamotoi and "Ca. N. mikurensis" [12] in Europe. Species such as Microtus agrestis, Myodes glareolus, Apodemus sylvaticus, Apodemus flavicolis as well as insectivores, e.g., Sorex araneus, have been found to be susceptible to R. helvetica infection [13]. On the other hand, R. helvetica did not show any pathogenic effect on Swiss mice, guinea pigs or domestic rabbits in the laboratory [14]. However, following the bite of infected D. marginatus, R. slovaca and specific antibodies against this bacterium were found in A. flavicollis and My. glareolus blood [15,16].
Humans are only occasional hosts for ticks and rarely play a role in the subsequent transmission of bacteria. Rickettsia helvetica is the etiological agent of uneruptive tick bite fever in humans which is characterized by fever, headeache, and myalgia. It may also cause myocarditis and meningitis [17,18]. Rickettsia raoultii is probably less pathogenic than R. slovaca [19]; they were described as etiological agents of TIBOLA/DEBONEL/SENLAT disease in humans, which is associated with a tick bite, an inoculum eschar, and cervical lymphadenopathies.
Coxiella burnetii, the causative agent of Q fever in humans, is an intracellular bacterium. The most common reservoirs are ruminants, primarily cattle, sheep and goats, although some documented outbreaks have been associated with parturient cats or birds. Humans typically acquire an infection from inhaling infected aerosols or dust generated by infected animals or animal products [20]. In Slovakia, C. burnetii has been isolated and identified from I. ricinus, D. marginatus, D. reticulatus, H. concinna and H. inermis, but the role of these ticks in its maintenance is unclear [21,22].
In the present study, we aimed to reveal the presence of SFG rickettsiae in questing ticks, wild-living rodents and rodent-attached ticks in different habitat types in order to identify potential reservoirs and amplifying hosts. The results will be a basis for future epidemiological studies and risk assessment of human tick-borne rickettsioses in Slovakia.

Sample collection
Questing ticks, rodents, and rodent-attached ticks were collected from three study sites in Slovakia. The locations were described in detail in Kazimírová et al. [23] and Berthová et al. [24]. Two study sites (Bratislava and Fúgelka) are located in the Small Carpathian Mountains (southwestern Slovakia). The mountains are partly densely forested with sessile oak (Quercus petraea) and European hornbeam (Carpinus betulus) dominating at lower altitudes and European beech (Fagus sylvatica) dominating at higher altitudes [23]. The recreational forest city park Železná studnička and the campus of the Slovak Academy of Sciences (SAS) at Bratislava (202-334 m above sea level, asl) represent a suburban habitat with significant anthropogenic impact. The natural habitat at Fúgelka (336-436 m asl) is set in a deciduous forest. The third study site located at Prievidza (central Slovakia; 289 m asl), is set nearby a settlement and represents a forest-steppe rural area with Carpathian oak-hornbeam woods [24].
Questing ticks were collected by dragging a 1 m 2 white wool blanket over the vegetation in monthly intervals during the peak tick questing activity periods, i.e., April -October 2011-2013 in the suburban and natural habitat, and during March -September 2012-2013 in the rural habitat. In each site three 100 m long transects were selected, the blanket was checked for ticks each 5 m. All collected ticks were stored in 1.5 ml tubes filled with 70% ethanol. The species and developmental stage of ticks were identified using the key by Siuda [25]. The impacts of environmental factors, including food availability for rodents, on the abundance and life history of ticks and population dynamics of rodents have not been analysed in this study.
Live-trapping of small rodents was carried out during April -June and September -October 2012-2014 in the suburban and natural habitats [26], and in May -July, and September -October 2012-2013 in the rural habitat. Sherman live traps baited with oat flakes were placed 5 m apart; 100 traps were set in each capture site and trapping event for a total of 1800 trap nights in each study site. Traps were exposed during two consecutive nights and were checked regularly every morning. Captured rodents were transported to the laboratory, anaesthetized and sacrificed humanely according to current laws of the Slovak Republic. This procedure was approved by the Ministry of Environment of the Slovak Republic, Regional Environmental Office in Bratislava (licence ZPO-594/2012-SAB). The rodents were identified to species, sex and age classes (juvenile, subadult, adult). Each rodent individual was screened for the presence of ticks. Collected ticks were stored in 70% ethanol until identification. Blood samples obtained from sinus orbitalis and spleen biopsies were stored in 70% ethanol, skin samples from ear were stored at -80°C.

Molecular analysis
Genomic DNA of ticks and rodent tissues were extracted using NucleoSpin Tissue kit (Macherey-Nagel, Düren, Germany) according to manufacturer instruction. Presence of Rickettsia spp. and C. burnetii DNA in the samples was evaluated by PCR assays targeting 381 bp and 401 bp fragments of the citrate synthase gene gltA [27,28] of rickettsiae and a 493 bp fragment of the outer membrane protein com1 gene of C. burnetii [22]. In each reaction positive (DNA of R. slovaca and R. helvetica, and C. burnetii strain Nine Mile originated from the deposit of the Institute of Virology BMC SAS Bratislava) and negative controls (miliQ water) were applied. The primers and cycling conditions are shown in Additional file 1. PCR amplifications were followed by electrophoresis in 1.0% agarose gel stained with GelRed TM (Biotium, Hayward, California). Rickettsia-positive tick samples were screened for the presence of R. helvetica using TaqMan PCR assay targeting a 65-bp fragment of the 23S rRNA gene as previously described [29], details are shown in Additional file 1. Alternatively, Rickettsia-positive amplicons were purified and analysed by sequencing both DNA strands by Macrogen Inc. (Amsterdam, The Netherlands). In the samples with non-identified rickettsial species fragments of 16S rRNA, sca4, ompA and ompB genes were amplified [30][31][32][33] (details are shown in Additional file 1). The DNA sequences were compared with those available in the GenBank database using the Basic Local Alignment Search Tool (Blast) on http://blast.ncbi.nlm.nih.gov/. The new sequences generated in this study were submitted to GenBank under accession numbers KY073144 for gltA, KY081649 for sca4 and KY628821-KY628825, KY628827 for 16S rRNA genes. A phylogenetic analysis was further performed using MEGA5 software [34]. The phylogenetic analysis of gltA gene of the samples was carried out using the Neighbor-Joining method [35]. The evolutionary distances were computed using the Maximum Composite Likelihood method [36] and are in the units of the number of base substitutions per site. The rickettsial genes gltA and 16S rRNA were concatenated and subjected to analysis by the Neighbor-Joining method. The evolutionary distances were computed using the p-distance method [34].
All PCR-positive questing D. marginatus ticks were subsequently analysed for the presence of sca4 gene [31] (see primers and cycling conditions in Additional file 1). To distinguish R. slovaca from other rickettsial species PCR-RLFP analysis was employed. Amplified sca4 gene was digested using HaeIII restriction endonuclease (digested fragments for R. slovaca are 477 bp and 146 bp [37]). After 3 h of incubation at 37°C, samples were loaded in 1% agarose gel and separated using electrophoresis. Undigested samples were purified and both strands were sequenced by Macrogen Inc. (Amsterdam, The Netherlands).

Statistical analysis
For rodents, we determined tick infestation prevalence (P, the proportion between the number of rodents with ticks and the number of captured rodents) and intensity of parasitization (I, the number of ticks per infested rodent). Statistical analyses to test the differences in the prevalences of infection with rickettsiae between years and sites were carried out with the Fisher's exact test using Past version 2.17b software [38]. A P-value < 0.05 was considered significant. 95% confidence intervals (CI) were calculated for prevalences. Differences in the prevalence of infection with Rickettsia spp. in questing and feeding ticks were analyzed between sexes and habitats applying Fisher's exact test, supplemented with Mantel-Haenszel common odds ratio estimate (OR) and its 95% confidence interval in cases when two prevalences were compared.
Altogether 262 Rickettsia-positive questing I. ricinus samples were analysed by sequencing or by R. helvetica-specific real time PCR. Rickettsia helvetica was the dominant species (80.5%) present in all tick developmental stages. Rickettsia monacensis was identified in 17 (6.5%) ticks of all developmental stages. In 7 (2.7%) ticks, two different unidentified Rickettsia spp. were detected ( Table 2).
Ixodes ricinus infestation prevalence and intensity of parasitization on A. flavicollis were higher than on My. glareolus (Table 4), but the prevalence of rickettsial infection was higher in ticks collected from My. glareolus than from Apodemus spp. (Table 5). The highest prevalence of rickettsial infection was determined in ticks collected from M. arvalis, but the total number of ticks collected from these rodents was only 28.
A collection of randomly selected 640 (282 from suburban, 266 from natural, 92 from rural habitat) rodent-attached ticks were screened for the presence of C. burnetii, with negative results.
Ear biopsies and/or spleen and/or blood samples from rodents were analysed for the presence of rickettsiae and C. burnetii. DNA of Rickettsia spp. was found in 0.5% (95% CI: 0.1-1.0; 3/663) rodents. All were A. flavicollis. None of the rodent samples was found to be PCR-positive for C. burnetii.
The total Rickettsia infection rate in questing I. ricinus was 6.6% (Table 1). No significant differences between years were found for females, but significant differences in overall prevalence and in prevalence in males and nymphs were revealed between 2011 and 2013 (P = 0.021, P = 0.025 and P = 0.001, respectively). The difference in infection rate in females and males was not significant (13.2 vs 8.4%; P = 0.075; OR = 1.7; 95% CI: 1.0-2.9), but it was significant between adults and nymphs (11.9 vs 5%; P < 0.001; OR = 2.3; 95% CI: 1.6-3.3). Analysis of seasonal distribution of Rickettsia spp.-infected I. ricinus during 2011-2013 showed that infected ticks occurred during the whole questing period (Fig. 1a). The highest total infection rate (10.3%, 95% CI: 3.9-16.7) was observed in July and the lowest one (4.3%,  Fig. 2). An unidentified Rickettsia sp. was found in one male I. ricinus (sample BA2 M12) (0.8%). Partial sequencing of the gltA gene showed 100% identity with Rickettsia japonica (GenBank DQ909073) previously found in patient blood in Thailand [39], and 99% similarity with the Rickettsia sp. from an I. ricinus male (FT2 M17, GenBank KY073144) from the natural habitat as described below. The quality of the sca4 gene sequences of sample BA2 M12 from this study was low. DNA of 25 Rickettsia-positive ticks, but negative by R. helvetica-specific real-time PCR, were not sequenced.  Rickettsia helvetica and R. monacensis were identified in rodent-attached I. ricinus (Table 5) and in ear biopsies from two A. flavicollis females, which were not infested by ticks.

Natural habitat
In the natural habitat, questing I. ricinus clearly dominated over H. concinna (for details see [23]). Three hundred and six rodent specimens were captured (for details on species composition see [26]) and 445 rodentattached ticks (I. ricinus, H. concinna and D. reticulatus) were collected ( Table 3).
Rickettsia helvetica and R. monacensis were identified in rodent-attached I. ricinus ticks (Table 5). Rickettsia helvetica was found in spleen from one A. flavicollis female, which was not infested by ticks.

Rural habitat
The highest diversity of tick species was recorded in this habitat: I. ricinus, D. marginatus, H. concinna and H. inermis. Ixodes ricinus dominated (n = 327), followed by  Table 3).

Discussion
In the present study, we screened questing ticks, rodentattached ticks and rodents for rickettsial and C. burnetii infections in three different habitat types in Slovakia. Rickettsial infection was detected in questing and rodentattached ticks, and in skin and spleen samples derived from A. flavicollis. Rickettsia helvetica and R. monacensis were identified in questing and rodent-attached I. ricinus ticks and in rodent samples. Non-identified Rickettsia spp. were found in seven I. ricinus ticks, but their subsequent analysis failed because of low quality sequenced data or lack of DNA. Rickettsia slovaca and R. raoultii were identified only in questing D. marginatus ticks. No evidence of C. burnetii infections was found in ticks or rodents in any of the studied habitats.
Ixodes ricinus is the most common tick in Slovakia and also dominated in all study sites. The highest diversity of tick species was observed in the rural habitat, where in addition to the dominant I. ricinus, D. marginatus, H. concinna and H. inermis were found. Sympatric occurrence of different tick species was previously recorded in a few other sites of southwestern Slovakia [5,41]. Similarly as in our study, I. ricinus was the dominant species in Martinský forest [41], whereas D. reticulatus dominated in the Vojka nad Dunajom flood-plain forest [5]. Thus, the differences in tick abundance and species spectrum in the studied habitats probably The presence of rickettsial DNA was confirmed in 6.9% of questing ticks. The overall prevalence of rickettsiae was 21.4 and 6.6% in D. marginatus and I. ricinus, respectively. These findings are in accordance with the results of previous studies from urban, suburban, natural and agricultural sites of Slovakia [5,[8][9][10].
During 2011-2013, the annual prevalence of rickettsiae in I. ricinus ranged from 5.2 to 8.2% in the suburban habitat and from 5.7 to 7.9% in the natural habitat. Similar data (from 6.7 to 8.7%) were reported from an agricultural site of eastern Slovakia [9] and in the Netherlands where the annual prevalence of R. helvetica was practically constant during the years 2000-2008 [42].
Analysis of the seasonal distribution of Rickettsia-infected I. ricinus showed that infected ticks were present during the whole tick questing period in the suburban and the natural habitat, similarly as in an agricultural site of Slovakia [9]. Seasonal/monthly variation in prevalence of Rickettsia-infected I. ricinus was observed not only in the suburban and natural habitat included in this study, but also in the agricultural site in eastern Slovakia [9], in Germany [43][44][45][46], Denmark [47] and the Netherlands [42]. The observed seasonal differences in the prevalence of rickettsial infections in questing ticks could be due to many factors such as tick phenology, sex ratio, habitat type, microclimate, presence and diversity of tick host species.
The initial infection of ticks with rickettsiae occurs when uninfected ticks feed on rickettsemic hosts. The probability to be infected by rickettsiae rises depending on the quantity of blood the tick ingests and the duration of its attachment. The exact process of natural infection of ticks is unrevealed for the majority of tick-borne rickettsiae, but it was observed that all subsequent life stages of the tick would be infected by transstadial transmission [7] and the ticks might acquire rickettsial infection even by co-feeding [48].
Although rodents have been confirmed as reservoirs and amplifying hosts for a number of tick-borne microorganisms [11,12], their reservoir role has not been resolved in the case of rickettsiae. In our study, we found only 0.5% R. helvetica-infected A. flavicollis, whereas 4.9% of ticks attached to A. flavicollis, My. glareolus and M. arvalis were Rickettsia-positive. Our results support previous findings from Slovakia where rickettsial infection was confirmed only in a low proportion of wild-living rodents [49], or no infection was found [50]. Similarly, in a sylvatic habitat of west-central Poland, rickettsial DNA was detected in 8% of rodent-attached I. ricinus nymphs and in 10.7% of larvae, but none of the tested A. flavicollis and My. glareolus blood samples were positive [51]. However, R. helvetica has frequently been detected in ticks feeding on non-rickettsiemic rodents. For example, in central Germany, SFG rickettsiae (mainly R. helvetica) were found in 1.8% of I. ricinus collected from A. flavicollis and My. glareolus [52], but in southern Germany ticks feeding on M. arvalis were C. burnetiiand Rickettsia spp.-negative [53]. In Switzerland, Rickettsia spp. were detected in 6.8% of immature I. ricinus ticks attached to A. flavicollis and My. glareolus, whereby 20.7% of larvae were infected with R. helvetica, but none of the xenodiagnostic ticks fed on rodents were found infected, suggesting that rodents are not amplifying hosts for SFG rickettsieae [12]. In contrast, as much as 29% blood samples obtained from A. sylvaticus and My. glareolus in the Netherlands were found to be infected with R. helvetica, 14% with Rickettsia spp. and 1.4% with Rickettsia conorii [42]. In south-eastern Germany, rickettsial DNA was found in 7.6% of ear skin samples from rodents [54] and in Hungary, R. helvetica was found in 1.9% of Mus musculus and in 20% of Leptopsylla segnis fleas [55]. During field studies in Italy, 22% of ear biopsies from Apodemus spp., but none from My. glareolus were R. slovaca-infected and the prevalence of R. slovaca in host-seeking D. marginatus larvae was 42% [56], suggesting that Apodemus spp. might play a role as amplifiers of the R. slovaca infection. However, based on the published data on rickettsial infections in wild-living rodents it is difficult to define their exact role in the maintenance of Rickettsia spp. in natural foci. We can only speculate that rickettsial infections of rodent-attached ticks are the result of either a vertical route of infection (transovarial and/or transstadial transmission) in ticks or a very short rickettsiaemia in rodents. Nevertheless, the presence of pathogenic R. helvetica, R. monacensis, R. slovaca and R. raoultii in ticks indicate their potential epidemiological and epizootological significance in the studied sites.
Although C. burnetii was previously detected in an A. flavicollis specimen from Central Slovakia [49], in this study, the presence of C. burnetii was not confirmed in any of the studied samples from ticks or rodents. Similarly, no evidence of C. burnetii infections could be found in ticks or rodents in southern Germany [53], suggesting that rodents do not play any essential role in the epidemiology of Q fever. On the other hand, in central Italy C. burnetii was detected in blood samples of Apodemus spp. and in rodent-attached I. ricinus and Ixodes acuminatus ticks [57]. These data may suggest that Apodemus spp. play a role in the maintenance of Q fever in the focus in the studied area. Moreover, wild-living rodents in Heilongjiang Province (border of China and Russia) [58] and in a Canadian Natural Environment Park [59] were found to be naturally infected with C. burnetii and thus are suggested as reservoirs of this patogen.
The presence of different pathogenic tick-borne Rickettsia spp. such as R. helvetica, R. monacensis, R. raoultii and R. slovaca was confirmed in suburban, natural and rural habitats of Slovakia, which are used by humans for relax (cycling, dog walking, jogging), or hunting. Thus, our data about the presence of various pathogenic rickettsiae in ticks and rodents indicate their potential epidemiological as well as epizootological significance in Slovakia. Therefore, humans visiting the study sites alone or with their pets should be aware of the possible risk of tick infestation and contracting tick-borne rickettsiae. The knowledge and early detection of infections in ticks and animal sources can prevent the transmission of pathogenic rickettsiae to humans.

Conclusions
The presence of Rickettsia spp. was confirmed in questing and rodent-attached I. ricinus and D. marginatus ticks in south-western and central Slovakia. The infection rates and presence of Rickettsia species depended on the habitat type, year, season, tick species and developmental stage. The presence of the pathogenic R. helvetica, R. monacensis, R. slovaca and R. raoultii in ticks indicate their potential epidemiological and epizootological significance and the risk of acquiring human rickettsial infections in different habitat types in Slovakia. Wild-living rodents could play a role as carriers of infected ticks and contribute to maintenance of rickettsial pathogens in natural foci.

Additional file
Additional file 1: PCR protocols for the detection of Rickettsia spp. and Coxiella burnetii. (DOCX 21 kb) Abbreviations CI: 95% confidence interval; OR: Mantel-Haenszel common odds ratio estimate; SFG: Spotted fever group