Background

Hard ticks (Ixodidae) are arthropods that suck blood from their vertebrate hosts and play an important role in the transmission and ecology of infectious diseases [1]. At least 30 ixodid tick species belonging to Amblyomma, Dermacentor, Haemaphysalis, Ixodes and Rhipicephalus genera have been documented in Ecuador [2]. These genera are recognized vectors of pathogenic bacteria with medical and veterinary relevance in neotropical regions [3].

In South America, information about the occurrence of tick-borne bacteria in wild mammals, which are frequently exposed to tick-bites, is limited [4,5]. Moreover, several severe and economically important diseases of livestock in tropical regions are caused by tick-borne pathogens (i.e. bovine anaplasmosis caused by Anaplasma marginale) that can also infect wildlife species [6].

In Ecuador, the Andean tapir (Tapirus pinchaque) is listed as endangered species. Cattle introduction into the Andean tapir refuges (i.e. Cayambe-Coca Ecological Reserve) is negatively affecting tapir populations due to loss of habitat. In this environment, pathogens of domestic animals may threaten health of wild animals and vice versa [7].

It is known that Amblyomma scalpturatum, Amblyomma latepunctatum, Amblyomma multipunctum and Amblyomma ovale tick species infest the Andean tapir in Ecuador [7]. All but A. multipunctum have been found biting humans in South America, and harboring tick-borne microorganisms [8-11]. The knowledge of bacteria transmitted by ticks (potential vectors and reservoirs of microorganisms) in a given area is useful for assessing the risk of infection in humans and animals. Therefore, the aims of our study were: 1- To investigate which tick species parasitize the Andean tapirs and cattle, and those present in the vegetation from the buffer zone of the Antisana Ecological Reserve and Cayambe-Coca National Park in Ecuador, and 2.- To detect and to identify tick-borne bacteria belonging to Rickettsia spp., Anaplasma spp., Ehrlichia spp. and Borrelia spp. genera in the collected tick specimens.

Methods

From May to October 2011 and during February 2012, an investigation was conducted in the buffer zone of the Antisana Ecological Reserve and Cayambe-Coca National Park, Napo Province, Ecuador (Figure 1). This area is located in the basin of the Papallacta River, where ‘The Andean tapir conservation project’ was developing.

Figure 1
figure 1

Study area in Ecuador (vector layers courtesy of EcoCiencia Foundation).

Full size image

Ticks were removed from 6 Andean tapirs, cattle [13 cows (Bos taurus) from 4 farms] and vegetation (10 transects of 2-Km long that were toured twice). Arthropods were kept in tubes with ethanol recording the host/sampling and date, and sent to the Center of Rickettsioses and Arthropod-Borne Diseases, located at the Center of Biomedical Research from La Rioja (CIBIR), Logroño (Spain) for further analysis.

The species were identified based on morphologic criteria following taxonomic keys from the Neotropical region [3,12,13]. DNA was individually extracted using DNeasy Blood & Tissue kit (Qiagen, Hilden, Germany). Each tick specimen was screened by PCR for both identification of tick species and detection of bacteria including Rickettsia spp., Anaplasma spp., Ehrlichia spp. and Borrelia burgdorferi sensu lato (s. l.). Tick species were confirmed by PCR targeting the tick mitochondrial 16S ribosomal RNA (rRNA) [14]. PCR assays for the tick mitochondrial 12S rRNA gene and the tick nuclear 5.8S-28S rRNA intergenic transcribed spacer 2 (ITS2) were also performed for selected samples [15,16]. For the screening of tick-borne bacteria, at least two fragment genes of each genus were tested by PCR assays. The molecular biomarkers selected to identify ticks are among the ones most widely used for the phylogenetics of ticks, being suitable to distinguish between closely related species. Biomarkers for the detection of microorganisms were selected based on our own expertise and according to previously reported usefulness and sensitivity. Target genes, specific primers and PCR conditions are listed in Table 1. Two negative controls, one of them containing water instead of template DNA and the other with template DNA but without primers, as well as positive controls of Rickettsia slovaca strain S14ab DNA (obtained from Vero cells inoculated in our facility with a Dermacentor marginatus tick from La Rioja, and known to be infected with R. slovaca), Anaplasma phagocytophilum strain Webster DNA kindly provided by Dr. Raoult (Unité de Recherche sur les Maladies Infectieuses et Tropicales Emergentes, France) and Dr. Dumler (The Johns Hopkins Hospital, USA), or Borrelia burgdorferi sensu stricto DNA kindly provided by Dr. Fingerle (German National Reference Centre for Borrelia, Germany) were included in all PCR assays. PCR products were sequenced in both directions. Sequences were compared with those available in the NCBI database using BLAST.

Table 1 PCR primer pairs used in this study
Full size table

Results

Identification of ticks

A total of 161 ticks (75 removed from Andean tapirs, 66 from cattle and 20 collected over vegetation) were included in the study. Ten specimens (one of each stage and gender in case of adult ticks) were deposited in the Museum of Zoology of Pontificia Universidad Católica from Ecuador.

Morphologically, 84 specimens (12 nymphs, 47 male and 25 female ticks) corresponded to A. multipunctum, 4 specimens to A. scalpturatum and 7 were classified as Amblyomma spp. For all but 4 specimens, the mitochondrial 16S rRNA sequences (409 bp) were identical to the 16S rRNA gene from A. multipunctum (GenBank accession no. KC677673), or differed by 0.2-1.7% (1–7 bp) when compared to this species. No 12S rRNA sequence from A. multipunctum was available in GenBank. Therefore, ours (from a specimen whose 16S rRNA sequence was identical to A. multipunctum KC677673) was deposited in GenBank under no. KM077433. It differed in sequence by 10% when compared to those available, and showed the highest identity (90%) with the 12S rRNA gene from Amblyomma sp. (GenBank accession no. AY342251). For the 4 tick specimens morphologically classified as A. scalpturatum, sequences of the 16S rRNA showed maximum identity (90%; 370/410 bp) with A. multipunctum, whereas 12S rRNA and ITS2 sequences were closest to Amblyomma varium (90.6% identity; 309/341 bp and 93.6% identity; 836/893 bp, respectively). Obtained sequences showed lower percentages of identity when compared to those from A. scalpturatum: 87% for 12S rRNA (GenBank accession no. AY342276), and 90% for ITS2 (GenBank accession no. AY619574). Therefore, these 4 ticks were classified as Amblyomma spp. and these three fragment genes were deposited in GenBank under nos. KM077434-6.

A total of 60 specimens were morphologically classified as Rhipicephalus microplus (formerly, Boophilus microplus) (6 nymphs, 16 male and 38 female ticks). In all these cases, the 16S rRNA sequences were identical to the 16S rRNA gene from R. microplus (GenBank accession no. EU918187).

According to morphological features, 5 female ticks were classified as Ixodes lasallei. The 16S rRNA sequences did not match with those from I. lasallei (GenBank accession no. AF549850) but were closest to this tick species (90% identity). Due to this discrepancy, they were classified as Ixodes spp. and deposited in GenBank under no. KM077438.

Lastly, one specimen morphologically corresponded to Ixodes boliviensis. The 16S rRNA sequences showed the highest identity (94%) with the 16S rRNA gene from Ixodes sp. (GenBank accession no. KF702351). It was deposited in GenBank since no sequences for I. boliviensis were available (KM077437).

According to morphological and genetic classifications, our ticks included 91 A. multipunctum, 4 Amblyomma spp., 60 R. microplus, 5 Ixodes spp. and 1 Ixodes boliviesis (Table 2).

Table 2 Ticks included in this study
Full size table

Detection and identification of tick-borne bacteria

Tick-borne bacteria were tested for 151/161 specimens, excluding those deposited in the museum.

The presence of rickettsiae was screened by PCR assays targeting 2 fragments of the gltA rickettsial gene (1019 and 401 pb, respectively). Positive amplicons were obtained for 3 R. microplus (2 male and 1 female specimens) removed from 2 cows from different farms. There were no differences in the sequences of gltA for amplicons derived from the DNA of the 3 rickettsial-infected R. microplus, and showed maximum identities (99.7% -99.2%) with gltA gene from Rickettsia monacensis and Rickettsia tamurae as validated species (Table 3).

Table 3 Maximum identities of rickettsial sequences detected in 3 Rhipicephalus microplus from Ecuador with validated Rickettsia species
Full size table

Subsequently, fragments of ompA (532 bp), ompB (420 bp), sca4 (928 bp), 16S rRNA gene (426 bp and 1500 bp, respectively), and 17 kDa-antigen gene (334 bp) were amplified to classify the Rickettsia at the species level.

The sequences of ompA (also identical each other) were closest to R. tamurae (95.9% identity) and R. monacensis (95.7% identity) (Table 3).

For ompB, the DNA sequences of the 3 rickettsiae-positive R. microplus were identical to each other and showed 99.2% identity with R. monacensis and 97.1% identity with R. tamurae (Table 3).

Unfortunately, no amplicons were obtained in PCR assays targeting sca4 gene. Attempts to sequence the rickettsial 16S rRNA and pan-bacterial 16S rRNA amplicons for the 3 R. microplus were inconclusive for Rickettsia. Curiously, A. marginale was amplified in 1 out of these 3 specimens using pan-bacterial 16S rRNA primers (see below). In addition, the sequences of 17 kDa antigen gene did not match with those available in GenBank.

In 2005, Raoult et al. established the criteria for the taxonomic classification of potential new Rickettsia species [32]. They proposed the ‘Candidatus’ status for a bacterium not established in pure culture that did not exhibit more than one of the following percentages of nucleotide identity: >99.8, >99.9, >98.8, >99.2, and >99.3 for rrs (16S rRNA), gltA, ompA, ompB, and sca4, respectively, with a validated Rickettsia species. According to our results, only amplicons for the gltA, ompA, ompB and 17KDa were obtained. Therefore, based on the recommended nomenclature [32], a Candidatus status could not be assigned to this microorganism. We designated this bacterium as Rickettsia sp. 12G1.

The presence of Anaplasma spp. was detected in 15 out of 151 samples, including 2 A. multipunctum and 13 R. microplus. On the one hand, the partial sequences of msp2 and 16S rRNA gene from Anaplasma spp. for A. multipunctum (a female tick from Andean tapir and a male tick from vegetation) and 8 R. microplus (all female ticks from one cow) were, when available, closest (96.6-100% identity) to A. phagocytophilum (Table 4). On the other hand, the 16S rRNA sequences (EHR and GEP regions) for 5 R. microplus (3 female and 2 male specimens) removed from 3 cows in two farms, were respectively identical each other, and matched (100% identity) with more than one Anaplasma species (assigned to A. marginale, Anaplasma ovis, A. phagocytophilum and Anaplasma centrale) for both PCR targets. Maximum identity with validated species of Ehrlichia genus did not exceed 95% with any of these 16S rRNA target genes (Table 4). Since these fragment genes were highly conserved for these species, in an attempt to identify the Anaplasma species, DNA extracts of these 5 samples were used as templates of pan-bacterial 16S rRNA PCR assays. The sequences corresponding to 2 out of 5 R. microplus were identical to each other and homologous (100% identity) to A. marginale. In these 2 cases, percentages of identity were 99.6, 99.5 and 97.2% when compared to A. ovis, A. centrale and human pathogenic A. phagocytophilum, respectively. Sequencing results for the 3 remaining ticks matched (99.4% identity) with a Coxiella endosymbiont of Rhipicephalus turanicus (GenBank accession no. J480818) (Table 4).

Table 4 Anaplasmataceae species detected in ticks from Ecuador
Full size table

Table 5 summarizes the detection rates for Rickettsia spp. and Anaplasma spp. Ehrlichia species were not amplified in any of the 151 ticks analyzed in this study. Lastly, B. burgdorferi s.l. was not detected in any ticks when flaB gene and 5S-23S rRNA intergenic spacer region were tested by PCR.

Table 5 Detection rates for Rickettsia spp. and Anaplasma spp.
Full size table

Co-infections

Out of 18 positive ticks, one of them (5.6%) was found co-infected with 2 bacteria. The co-infection detected was A. marginale with Rickettsia sp. 12G1 in one R. microplus tick collected from a cow.

GenBank accession numbers

Sequences obtained in this study have been deposited in the GenBank database under the following accession numbers: KM077433-8 (identification of ticks) and KF831358-62 (rickettsial genes).

Discussion

A total of 161 ticks (nymphs or adult specimens) removed from Andean tapirs, cattle and vegetation, and belonging to Amblyomma, Rhipicephalus and Ixodes genera, was included in the present study. These tick genera had been previously reported to occur in Ecuador [3,33,34]. Based on morphological and genetic criteria, arthropods were classified as 91 A. multipunctum, 4 Amblyomma spp., 60 R. microplus, 5 Ixodes spp. and 1 I. boliviesis. On the one hand, A. multipunctum was collected from vegetation and found attached to Andean tapirs. This tick species was originally described from a Tapirus sp. in North America, and it has been reported in Venezuela, Colombia and Ecuador [35,36]. Partial sequences of the mitochondrial 16S rRNA gene of A. multipunctum specimens from Ecuador had been previously generated [37]. Our group has completed this molecular description with sequences of the 12S rRNA fragment gene (GenBank accession no. KM077433). On the other hand, R. microplus and Ixodes spp. were removed from cows, as well as one specimen of I. boliviensis that was genetically characterized herein using mitochondrial 16S rRNA gene as PCR target (GenBank accession no. KM077437). R. microplus, known as the cattle tick, is widely distributed in cattle from tropical regions [3]. This is the first description of I. boliviensis in Ecuador, although it has been found in cattle from Costa Rica [38].

As far as we know, this is the first report where ticks from Ecuador were evaluated for the presence of Rickettsia spp., Anaplasma spp., Ehrlichia spp. and Borrelia spp.

The circulation of a potential Candidatus Rickettsia species (designated Rickettsia sp. 12G1) in R. microplus ticks removed from cattle in Ecuador is reported. According to our data, this novel Rickettsia was closest to R. monacensis and R. tamurae, as validated species. R. monacensis has been so far reported from Ixodes ricinus, and R. tamurae from Amblyomma testudinarium [39]. The human pathogenic role of R. monacensis was first reported in Spain [18], and one case of R. tamurae infection has been detected in Japan [40]. Nevertheless, no evidence of human pathogenicity is presented herein for Rickettsia sp. 12G1, and there is no evidence to suggest that this Rickettsia is transmissible to humans. Other new genotypes with unknown pathogenicity that also belong to the same lineage of R. tamurae and R. monacensis, such as Rickettsia sp. strain Colombianensi or Rickettsia sp. strain IbR/CRC, have been documented in R. microplus or I. boliviensis from the New World [41,42].

In our study, A. phagocytophilum or closely related genotypes have been detected in ticks removed from Andean tapirs, cows and vegetation. It is known that the high intraspecific variability observed in the msp2 gene of A. phagocytophilum promotes the adaptation of the bacterium to different hosts and could justify its distribution in various environments [43]. As expected, the msp2 sequences obtained in this study (corresponding to 10 ticks) showed high genetic variability. Whereas the 16S rRNA sequences matched, when available (n = 2), with A. phagocytophilum pathogenic for humans (GenBank accession no. CP000235), msp2 sequences for 5 specimens (1 A. multipunctum from an Andean tapir and 4 R. microplus from cows) demonstrated relatedness with human pathogenic A. phagocytophilum but differed by 0.7-3.4% [44,45]. In addition, msp2 sequences obtained from 4 R. microplus were closest (96.6-98.3% identity) to A. phagocytophilum from Japanese Ixodes persulcatus [46]. Lastly, the msp2 sequence for 1 A. multipunctum from vegetation was 99% identical to one A. phagocytophilum strain from rodents in Florida (also highly similar to human pathogenic reference strain) [47].

As far as we know, the occurrence of A. phagocytophilum or closely related genotypes had not been previously detected neither in Ecuador nor in ticks removed from tapirs. Nevertheless, A. phagocytophilum or closely related Anaplasma spp. have been found in blood samples from domestic (dogs and cats) and wild animals (deer) in Brazil [48-50]. This is the first evidence of A. phagocytophilum in R. microplus in the New World. Nevertheless, this bacterium had been previously found in R. microplus from China [51].

Based on the sequencing results of the 16S rRNA gene, 2 R. microplus specimens removed from cows tested positive for A. marginale and 3 harbored Anaplasma spp. (assigned to A. marginale, A. ovis, A. phagocytophilum and A. centrale).

A. marginale, which is transmitted by R. microplus, has a worldwide occurrence and is considered as one of the most prevalent pathogens causing cattle morbidity and mortality in subtropical and tropical countries, including Latin America [52,53]. Our study evidences the first molecular detection of A. marginale in R. microplus from Ecuador. This bacterium had been previously detected in Ecuadorian blood samples from cattle by PCR [54] and also in R. microplus ticks in Philipinnes [55].

Moreover, no evidence of Ehrlichia spp. or B. burgdorferi s.l.-infected ticks has been found in Ecuador. Nevertheless, in South American countries, new members of the Ehrlichia genus and the B. burgdorferi s.l. complex have been recently described in Brazil, Uruguay and Chile [56-59].

Conclusions

In summary, this is the first description of Rickettsia sp. in ticks from Ecuador, and the analyses of sequences suggest the presence of a potential novel Rickettsia species. The complete characterization and distribution of the novel Rickettsia sp. 12G1, as well as its possible pathogenic role for animals and humans, needs to be determined.

Our data also showed that ticks from Andean tapirs, cattle and vegetation in Ecuador (Amblyomma and Rhipicephalus) were naturally infected with Anaplasmataceae and that co-infection (A. marginale and Rickettsia sp.) occurred.