Identification of tick-borne pathogens in ticks feeding on humans in Turkey.

Background The importance of tick-borne diseases is increasing all over the world, including Turkey. The tick-borne disease outbreaks reported in recent years and the abundance of tick species and the existence of suitable habitats increase the importance of studies related to the epidemiology of ticks and tick-borne pathogens in Turkey. The aim of this study was to investigate the presence of and to determine the infection rates of some tick-borne pathogens, including Babesia spp., Borrelia burgdorferi sensu lato and spotted fever group rickettsiae in the ticks removed from humans in different parts of Ankara. Methodology/Principal Findings A total of 169 ticks belonging to the genus Haemaphysalis, Hyalomma, Ixodes and Rhipicephalus were collected by removing from humans in different parts of Ankara. Ticks were molecularly screened for Babesia spp., Borrelia burgdorferi sensu lato and spotted fever group rickettsiae by PCR and sequencing analysis. We detected 4 Babesia spp.; B. crassa, B. major, B. occultans and B. rossi, one Borrelia spp.; B. burgdorferi sensu stricto and 3 spotted fever group rickettsiae; R. aeschlimannii, R. slovaca and R. hoogstraalii in the tick specimens analyzed. This is the report showing the presence of B. rossi in a region that is out of Africa and in the host species Ha. parva. In addition, B. crassa, for which limited information is available on its distribution and vector species, and B. occultans, for which no conclusive information is available on its presence in Turkey, were identified in Ha. parva and H. marginatum, respectively. Two human pathogenic rickettsia species (R. aeschlimannii and R. slovaca) were detected with a high prevalence in ticks. Additionally, B. burgdorferi sensu stricto was detected in unusual tick species (H. marginatum, H. excavatum, Hyalomma spp. (nymph) and Ha. parva). Conclusions/Significance This study investigates both the distribution of several tick-borne pathogens affecting humans and animals, and the presence of new tick-borne pathogens in Turkey. More epidemiological studies are warranted for B. rossi, which is very pathogenic for dogs, because the presented results suggest that B. rossi might have a wide distribution in Turkey. Furthermore, we recommend that tick-borne pathogens, especially R. aeschlimannii, R. slovaca, and B. burgdorferi sensu stricto, should be taken into consideration in patients who had a tick bite in Turkey.


Introduction
Ticks are very important vectors of diseases affecting both humans and animals. They transmit a broader range of viral, bacterial (including rickettsial) and protozoan pathogen microorganisms than any other arthropods worldwide, and are also the main reservoirs of these pathogens [1,2]. Tick-borne diseases (TBDs) constitute a major public health concern and they are responsible for great economic losses in terms of mortality and morbidity of livestock animals worldwide [2,3].
The importance of TBDs is increasing all over the world. We still encounter new tick-borne epidemics [4]. The best example of a recent outbreak is that the Crimean-Congo hemorrhagic fever (CCHF) outbreak that started in northern Turkey in 2002, human cases still continue to increase and the disease spreads in many regions of Turkey [5,6]. Moreover, the incidence of significant TBDs has highly increased during the last 30 years worldwide [4]. In recent years, both new tick-borne pathogens have been described and detailed epidemiological studies have been carried out owing to the commonly used molecular techniques. Therefore, molecular tools can provide a better understanding of the epidemiology of TBDs [3]. Nevertheless, we need new studies especially on the epidemiology and diagnosis of tick-borne pathogens and the ecology of these newly recognized disease agents [7].
In Turkey, 5 TBDs have remarkable importance for humans: Crimean-Congo hemorrhagic fever, Lyme borreliosis, spotted fever group rickettsiosis, babesiosis, and anaplasmosis. However, the other diseases except for Crimean-Congo hemorrhagic fever are neglected in patients, although whole of the pathogens have been commonly detected in ticks [5,[8][9][10][11][12]. In previous studies, e.g., human pathogenic spotted fever group rickettsiae; Rickettsia africae, Rickettsia aeschlimannii, and Rickettsia slovaca have been remarkable detected in ticks collected from different regions [10,12]; however, no clinical cases have been reported in Turkey so far. Nonetheless, the studies demonstrated that the tick species which belonged to the genera Dermacentor, Haemaphysalis, Hyalomma, Ixodes and Rhipicephalus are responsible for major human tick bite cases and these species have a wide distribution in Turkey [12][13][14][15][16][17].
In this study, we aimed to investigate the presence Babesia spp., Borrelia burgdorferi sensu lato and spotted fever group rickettsiae in ticks obtained from hospitals of Ankara.

Collection of tick samples and morphological identification
This study was conducted in Ankara, the capital city of Turkey. Ankara province is located in Central Anatolia ( Fig. 1) and is the second largest city in the county. Ankara is about 24.500 km 2 and the population of the city is close to 5 million. Half of the total area (approximately 12.000 km 2 ) is used as agricultural land. Different habitats; forest, steppe, wetlands, and salty soils are encountered in Ankara. The average annual precipitation is 242-612 mm., while the average annual temperature is 10.3-14.7uC. Ankara is under the influence of semi-arid and very cold Mediterranean climate. A large part of the province is covered with steppe. The elevation is between 550 and 2000 meters [18]. In rural areas, cattle and sheep breeding are commonly made partially intensive but mostly in pasturelands. Goat breeding is limited to some villages and goats are generally pastured while mixed with sheep. Wild animals such as wild boar, hare, fox, and ground-feeding birds (partridge, crow etc.), which are also called amplifying hosts for ticks, are abundant throughout the province. From

DNA extraction and PCR amplification
Each tick was first washed in 70% alcohol, then rinsed in sterile water and dried on sterile filter paper in order to avoid contamination. Ticks were individually homogenized by crushing with liquid nitrogen in a mini-mortar. DNA was individually extracted from crushed ticks by using the Qiagen DNeasy blood and tissue kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. First, an initial PCR targeted tick 16S rDNA as internal control gene with 16S+1 and 16S-1 primers, which amplify approximately 460 base pairs (bp) fragment was performed to every tick in order to determine whether PCR inhibition [22]. Only the positive samples were further analyzed for the presence of Babesia spp., Borrelia spp., and Rickettsia spp.
A Babesia spp. genus-specific PCR was performed to each tick by using primers BJ1 and BN2, which amplify a 411-452-bp fragment of the 18S ribosomal RNA (18S rRNA) gene [23]. For the detection of B. burgdorferi sensu lato species, a nested PCR, which amplifies approximately 200-bp fragment of the 5S-23S rDNA intergenic spacer (IGS) was carried out by using two set primers RIS1 and RIS2, and RIS3 and RIS4 for the first-step and the nested-PCR, respectively [8,24]. Rickettsial DNA was detected by PCR using the primers Rp CS.409d and Rp CS.1258n, which amplify a 750-bp fragment of the citrate synthase gene (gltA) of Rickettsia spp. [25]. Additionally, each tick positive for gltA was also tested for the ompA gene of Rickettsia spp. using the primers Rr. 190.70 and Rr. 190.701, which amplify a 629-632-bp fragment [26]. DNase-RNase-free water was used as a negative control and positive controls (DNA from B. bigemina, B. divergens, R. montanensis, B. burgdorferi sensu lato) were included in all reactions. We have also performed pre-PCR with positive controls in different dilutions (1-1/100) to avoid false negative results that may occur due to low copies of bacterial and protozoan genes.

Sequencing and phylogenetic analysis
Successfully amplified product was purified using the QIAquick Extraction Kit (Qiagen GmbH). Purified DNA was sequenced using BigDye Terminator V3.1 Cycle Sequencing Kit (Applied Biosystems, Foster City, USA). Automated fluorescence sequencing was performed with an ABI PRISM 3100 Genetic Analyzer (Applied Biosystems). Nucleotide sequences were processed using nucleotide BLAST (National Center for Biotechnology Information, www.ncbi.nlmn.nih.gov/BLAST). Sequences were edited and aligned by using BioEdit software [27]. We used jModeltest version 0.1.1 [28] to determine the most appropriate model for our data. Phylogenetic and molecular evolutionary analyses were performed by using MEGA version 5.2.1 [29]. The gene sequences obtained in this study have been deposited in GenBank under the accession numbers KF791205 to KF791260.

Author Summary
Ticks are widespread in over all Turkey. Primary tick-borne diseases (TBDs), such as theileriosis, babesiosis, and anaplasmosis, affecting animals have been known for a long time in Turkey. However, TBDs have become a major concern in humans in recent years due to the recent Crimean-Congo hemorrhagic fever outbreak in Turkey. We know that some TBDs like CCHF, Lyme borreliosis, spotted fever group rickettsiosis, babesiosis and anaplasmosis exist in this geography. However, the other diseases except for CCHF are neglected in patients. In this study, we aimed to investigate Babesia spp., Borrelia burgdorferi sensu lato and spotted fever group rickettsiae in ticks removed from humans in different parts of Ankara by using PCR and sequencing. The result of this study showed that 4 Babesia species, one B. burgdorferi sensu lato, and 3 spotted fever group rickettsia are detected in ticks. The most striking result of this study is that B. rossi, which is very pathogenic for dogs, was reported for the first time from a region that is out of Africa and in Ha. parva. Therefore, B. rossi should be considered in dogs in Turkey. Furthermore, we propose that R. aeschlimannii, R. slovaca, and B. burgdorferi sensu stricto should be taken into consideration in patients who had a tick bite in Ankara.  Hyalomma marginatum (30) Haemaphysalis parva (35)  H. marginatum were obtained as engorged nymph and were then allowed to molt to the adult stage (as unfed ticks) under suitable conditions. Ticks collected in this study have been partially fed except the unfed 10 ticks. The number of the ticks found per patient is one. 55 ticks were found to be infected with at least one of the pathogens in the genera Babesia, Borrelia, or Rickettsia. Most of the ticks were infected with a single pathogen; however, one tick (Ha. parva) contained the DNA of both Babesia spp. and Rickettsia spp. in a mixed infection. The regarding information is summarized in Table 1.
Every DNA sample was found to be positive for the tick 16S rDNA and was subjected to PCR assay to detect tick-borne pathogens. Additionally, pre-PCR tests, which are performed to determine the sensitivity of PCR assay, showed that even the lowest copies of bacterial and protozoan genes are yielded visible positive bands. As a result of the PCR analyses, Babesia spp. was detected in 4 tick specimens (2.3%). 18S rRNA nucleotide sequence indicated that B. crassa, B. major and B. rossi were found in Ha. parva, Ha. punctata and Ha. parva tick individuals, respectively, attached to humans in central Ankara. Among these, one H. parva infected with B. crassa was also infected with R. hoogstraalii. Additionally, B. occultans was detected in 1 H. marginatum to a person attached in Kalecik. 5S-23S rDNA IGS nucleotide sequences derived from Borrelia spp. positive ticks displayed that B. burgdorferi sensu stricto was found in 6 tick specimens (3.5%): 1 H. marginatum attached in Ç ubuk, 1 H. excavatum, 2 Hyalomma spp. (nymph) and 2 Ha. parva attached in Central Ankara. Furthermore, rickettsial DNA was detected in 46 ticks (27.2%). Rickettsia spp. was determined as the most prevalent tick-borne pathogen in this study. OmpA gene sequence analysis showed that R. aeschlimannii was detected in 5 H. marginatum attached in Bala, Ç amlıdere, Kızılcahamam and central Ankara, 1 H. aegyptium attached in Kazan and 1 unfed H. aegyptium, which was originally obtained as engorged nymph from a person and then allowed to molt to the adult stage as described above in central Ankara and 1 H. excavatum attached in central Ankara. 16 D. marginatus, which makes 64 percent of the total D. marginatus ticks, attached in Akyurt, Elmadag, Ç ubuk, Gölbaşı, Kızılcahamam, Polatlı and central Ankara were found to be infected with R. slovaca. According to the gltA nucleotide sequences (we could not obtain PCR products from the ompA gene), R. hoogstraalii was detected in 22 of 35 Ha. parva ticks attached in Ayaş, Akyurt, Ç ubuk, Gölbaşı, and central Ankara. We could not detect DNA of Babesia spp., Borrelia spp., or Rickettsia spp. in Haemaphysalis spp. nymph, I. ricinus, Rh. bursa, Rh sanguineus or Rh. turanicus specimens. Detailed information about the tested ticks and the nucleotide similarities of the obtained Babesia, Borrelia, and Rickettsia sequences in this study are given in Table 1 and 2, respectively. In addition, Phylogenetic trees were constructed separately by using 18S rRNA gene of Babesia spp., 5S-23S rDNA IGS of B. burgdorferi sensu lato and ompA and gltA genes of Rickettsia spp. are illustrated in Figs. 2-5.

Discussion
Ticks transmit a great variety of viral, bacterial (including rickettsial) and protozoan disease agents so that they play a major role in the epidemiology of tick-borne diseases affecting both human and animals. The incidence and recognition of ticks and tick-borne diseases increases steadily year by year worldwide [1,2,7]. The increasing outbreaks in recent years are the evidence of this situation [4]. In Turkey, TBDs such as Lyme-borreliosis and spotted fever group rickettsiosis for humans and babesiosis for both animals and humans have a remarkable importance [8][9][10]12,[30][31][32][33][34][35].
In the present study, we have retrieved new data related to both some tick-borne pathogens and its vectors. The most striking of these is that we have detected B. rossi in Ha. parva in Turkey. Babesia rossi, the most pathogenic Babesia species of dogs, is transmitted by Haemaphysalis ticks (mainly Ha. leachi) and prevalent in South Africa and other African countries, such as Nigeria and Sudan [36]. Interestingly, B. rossi was detected in one Ha. parva removed from humans in central Ankara in this study. According to 18S rRNA sequence, our Turkish B. rossi sequence has a 98.9% similarity with the B. rossi isolate Dog-44 (accession no. DQ111760) obtained from a dog in Sudan (Table 2). This is the first report of the presence of the native-B. rossi in a country that is outside of the African continent and in a Ha. parva tick. Furthermore, we detected B. crassa, a large Babesia spp. of small ruminants, in a Ha. parva tick. Babesia crassa was first detected from a sheep in Iran [37], but no detailed information about the distribution and vectors of this species was given. According to the sequence analysis, the B. crassa sequence detected in this study was 100% similar to the B. crassa (accession no. AY260177) obtained from a sheep in Turkey (Table 2). Although the other B. crassa was detected from a sheep previously in Turkey, the detected area information was not provided [38]. However, it is clear that B. crassa exists in both ticks and sheep in Turkey. Our new findings also point out the existence of B. crassa in Ha. parva. Interestingly, the B. crassa sequence detected in this study has 95% similarity with B. crassa (accession no. AY260176) that was detected from the Turkish border of Iran. Schnittger et al. (2004) stressed that it is necessary to design strain-specific oligonucleotides allowing the discrimination between B. crassa (Iran) and B. crassa (Turkey) [39]. More detailed studies towards B. crassa and a large number of sequences are required to determine the difference between both Babesia and to come to the conclusion whether we come across a variant of B. crassa. Babesia occultans, which is moderately pathogenic bovine Babesia species, was detected in H. marginatum. According to the sequence analysis, the Turkish B. occultans sequence was 100% similar to the B. occultans isolate Italy-366/12-20 (accession no. KC157568) detected from the cattle in Italy and B. occultans isolate 58 (accession no. HQ331478) detected from H. marginatum in Tunisia (Table 2). Babesia occultans, which is transmitted by mainly Hyalomma spp., is prevalent in Africa. It is thought to be a non-pathogen or of low pathogenicity [36,40]. On other hand, a clinical outbreak caused by B. occultans has recently been reported in cows that displayed fever, anemia, and severe alteration in the hematological parameters in Italy [41]. Mean while, there are no clear reports about the existence of B. occultans in Turkey, although the similar sequences, Babesia sp. Kayseri 1 (accession no. EF434786) [42] and Babesia sp. H4 (accession no. JF923655) [9] obtained from H. marginatum ticks collected from cattle and Babesia sp. CS58 (accession no. EU622824) obtained from cattle [32], were reported in previous studies. These sequences are not enough for a definitive identification; therefore, we could not compare our B. occultans sequence with the others. However, our study clearly shows the presence of B. occultans in Turkey. Additionally, we detected B. major, a low-pathogenic bovine babesia species, in Ha. puctata. According to the sequence analysis, the Turkish B. major has 100% similarity with the B. major isolate France 1 (accession no. EU622907) which was obtained from cattle in France ( Table 2). The presence of B. major was reported from the cattle previously in East Black Sea Region of Turkey [32]. However, this is the first report of the presence of  Table 2 and Fig. 2, respectively.
Borrelia burgdorferi sensu stricto, which is the primary pathogenic genospecies of Lyme disease [43], was detected in one H. marginatum, 2 Hyalomma spp. (nymph), 2 Ha. parva and one H. excavatum. According to 5S-23S rDNA IGS sequence analysis, B. burgdorferi sensu stricto sequences obtained in this study are 100% similar to the reference strains (B. burgdorferi N40, accession no. CP002228, isolate TN19, accession no. DQ860271 and genotype I-181, accession no. AF497980). More detailed data and a phylogenetic tree are given in Table 2 and Fig. 3, respectively. Lyme borreliosis, transmitted mainly by ticks belonging to Ixodes genus, is the most common tick-borne zoonosis in the temperate zone of the northern hemisphere [44]. In Turkey, B. burgdorferi sensu stricto was hitherto isolated from only one unfed I. ricinus collected from Istanbul in 2003 [31]. However, non-Ixodes spp. ticks were to be infected with B. burgdorferi sensu stricto in this study. This indicated that either the ticks ingested infected body fluids or may be capable to transmit this bacterium (potential vectors). Yet, we cannot clearly say that these tick species are the vector for B. burgdorferi sensu stricto with the current data. It is possible that the humans had bacteremia of B. burgdorferi sensu stricto and this pathogen passed to the ticks during the feeding process or the ticks might have been infected in previous life stages, as well. As noted by Kahl et al. (2002): ''to be considered a vector, a tick species must: (1) feed on infectious vertebrates; (2) be able to acquire the pathogen during the blood meal; (3) maintain it through one ore more life stages (transstadial passage); and (4) pass it on to other host when feeding again. Otherwise it is a non-vector tick'' [45]. Therefore, more detailed experimental studies are required to determine the vector competence of these tick species. Nevertheless, it is already evident that B. burgdorferi sensu stricto is circulating in Ankara, because we did not record any history of a tick bite outside of Ankara from the patients. Hence, Lyme disease should be taken into consideration in patients who had a tick bite in Ankara.
Spotted fever group rickettsiae have recently been found with remarkable infection rates in ticks in Turkey [10,12]. Likewise, spotted fever group rickettsiae were the most commonly observed agents with 27.2% in this study (Table 1). Two human pathogenic rickettsial species (R. aeschlimannii and R. slovaca) and one species (R. hoogstraalii) with unknown pathogenicity were detected in ticks. Among them, R. aeschlimannii is transmitted mainly by Hyalomma ticks [46]. This bacterium was found in 5 H. marginatum, 2 H. aegyptium, and 1 H. excavatum in this study. As a result of the ompA gene sequence, R. aeschlimannii sequences obtained in this study are by 99.8-100% similar to the reference strains (R. aeschlimannii strain EgyRickHimp-El-Arish-17, accession no. HQ335158 and strain TR/Orkun-H.agyp86/Ankara accession no. JQ691728). In Turkey, R. aeschlimannii was found in 5 H. aegyptium, 2 H. marginatum, and one Rh. bursa collected from humans in Istanbul [10]. In our previous study, this pathogen was detected (in 32% of Rickettsia-positive ticks) in 3 H. marginatum, 2 H. aegyptium (unfed), 2 Hyalomma spp. (nymph), 1 Rh. turanicus collected from humans and animals in Kırşehir and Ankara [12]. These results show that our findings are in parallel with the previous findings. Rickettsia slovaca, the etiological agent of Tick-borne lymphadenopathy/Dermacentor-borne necrosis erythema and lymphadenopathy (TIBOLA/DEBONEL) [47], was detected in 16 D. marginatus, which are 64 percent of the total D. marginatus ticks (Table 1). According to ompA sequence, R. slovaca sequences detected in this study have 99.8-100% similarity with the reference strains (R. slovaca 13-B, accession no. CP002428, and strain TR/Orkun-D.marg79/Ankara accession no. JQ691724). Previously, we had detected this bacterium in 8 D. marginatus (80% of total D. marginatus ticks) collected from human and cattle in Ankara [12]. As a result, we again stress that R. aeschlimannii and R. slovaca are significant disease agents for humans and should not be neglected in this area. The clinical signs caused by these pathogens should be considered in the differential diagnosis of the tick-borne diseases in patients. Additionally, we detected R. hoogstraalii, a rickettsia with an unknown pathogenicity [48], in 22 out of 35 Ha. parva ticks collected (Table 1). According to the gltA gene sequence, the sequences obtained from this study are 100% similar to R. hoogstraalii strain TR/Orkun-Ha.parva164/Ankara (accession no JQ691712). In Turkey, R. hoogstraalii had been detected in 4 Ha. parva collected from humans in Ankara in our previous study [12]. The pathogenicity of R. hoogstraalii is unknown, but this rickettsia is widespread in Ha. parva ticks in this area. The nucleotide similarities of rickettsial sequences detected in this study and phylogenetic relationships are shown in detail in Table 2 and Figs. 4-5, respectively.
In conclusion, we have studied a broad range of tick-borne pathogens and detected several pathogens in ticks removed from humans in Ankara. Moreover, this study shows the presence of new tick-borne pathogens in Turkey. Among them, B. rossi, especially, was only known to be prevalent in Africa. However, this pathogen was detected in a Turkish tick and this finding reported here is the first evidence of the existence of B. rossi in a country outside of Africa. This disease agent, which is very pathogenic for dogs, must be considered in this area, because our results tell us that it is likely that B. rossi has a wide distribution in Turkey. Hence, more detailed epidemiologic studies are required in the future. Additionally, we detected B. crassa, of which we have very limited information on its distribution and vector, in a Ha. parva tick. Moreover, there is no clear information about the presence of B. occultans in Turkey, which we detected again in our study. Babesia major was also detected in ticks. Although we could not detect human pathogenic Babesia species, very significant epidemiological data related to animal babesiosis were collected in this study. Furthermore, we detected two human pathogenic rickettsia species (R. aeschlimannii and R. slovaca) which were previously detected in the same area with a high prevalence. We recommend that these pathogens should be taken into consideration in patients who had a tick bite in Turkey, especially in Ankara. Additionally, B. bugdorferi sensu stricto was detected in different tick species (non-Ixodes spp.). This pathogen is transmitted by mainly I. ricinus, but this tick is not common among tick species in Ankara due to the existence of very restricted suitable habitat for I. ricinus ticks [16]. This situation also opens up the discussion of whether the other vector or carrier tick species exist in the region. However, we do not know the vector competence of these tick species at the moment without detailed experimental studies. It is important to point out that this pathogen caused Lyme disease exists in the region and should not be overlooked in patients exposed to tick bites.