Molecular Detection of Theileria spp. in Livestock on Five Caribbean Islands

Theileria spp. are tick-transmitted, intracellular apicomplexan protozoan parasites infecting a wide range of animals. As there is very limited information on the prevalence of Theileria spp. in the Caribbean we used the recently described genus-specific pan-Theileria FRET-qPCR to identify infected animals in the region and a standard 18S rRNA gene PCR and sequencing to determine the species involved. We found Theileria spp. in 9% of the convenience samples of animals (n = 752) studied from five Caribbean islands. Donkeys (20.0%: 5/25) were most commonly infected, followed by sheep (17.4%, 25/144), cattle (6.8%; 22/325), goats (5.0%; 12/238), and horses (5.0%; 1/20). Six species of Theileria were identified: T. equi (donkeys, cattle, goats, and sheep), Theileria sp. OT3 (sheep and goats), Theileria sp. NG-2013a (cattle), Theileria sp. YW-2014 (donkeys), Theileria sp. B15a (goats), and Babesia vulpes or a closely related organism (sheep and goats). Only T. equi has been previously reported in the Caribbean. Our findings expand the known host ranges of Theileria spp. and the known distribution of the organisms around the world.


Background
Theileria spp. are tick-transmitted, intracellular apicomplexan protozoan parasites infecting leukocytes and erythrocytes of a wide range of animals [1,2]. The organisms have been described in all livestock species and can cause significant economic losses to farmers. They are transmitted by a variety of ixodid ticks of the genera Rhipicephalus, Hyalomma, Amblyomma, and Haemaphysalis [3]. Infections with some Theileria spp. can result in fever, anemia, hemoglobinuria, and death in severe cases, but many species are benign and cause minor or no signs. Animals that recover from acute or primary infections usually remain persistently infected and may act as reservoirs for tick vectors [4,5]. Infected animals are found particularly in tropical and subtropical regions in Africa, the Middle East, Southern Europe, and Asia [6][7][8][9][10][11].
There is little information on infectious agents in livestock in the Caribbean although animal production is an important source of income for many people in the region. In the case of Theileria spp., morphological and serological evidence has been presented that T. mutans and T. velifera, both benign species transmitted by Amblyomma spp., occur in cattle on Guadeloupe [12]. Also, an organism with the morphology of T. mutans was seen in a blood smear from a bovine on Martinique [13]. In Trinidad, T. equi (previously Babesia equi) has been demonstrated in horses with a specific nested 18S rRNA PCR [14,15] and a serosurvey has provided supporting evidence for its presence [16].
While there are many tests to detect Theileria spp. in animals, their specificity varies, as does their usefulness in finding the full spectrum of organisms present in an area. Microscopic detection of parasites can be difficult with low parasitemia and does not readily allow differentiation of species [2]. Serological studies, although sensitive and relatively easy to perform, are not specific as there is crossreactivity between Theileria spp. [17]. Although molecular techniques have been described, many are for specific species which limits their usefulness in surveys. Reverse line blotting (RLB) assays enable the simultaneous identification of multiple species [18,19] Figure 1: Alignment of the sequences in the polymorphic region of the 18S RNA gene of Theileria spp. and related organisms that was targeted in the standard PCR used in the study. Nucleotides that are identical for all species are highlighted in blue while those that vary between species and can be used for differentiation are not highlighted. species-specific oligonucleotide sequences can be challenging [20]. Recently, a sensitive genus-specific pan-Theileria FRET-qPCR has been described that detects the recognized Theileria spp. of domestic animals in a single reaction (Table 1) [21]. To provide further data on Theileria spp. in the Caribbean, we used the pan-Theileria FRET-qPCR to screen livestock from five islands for evidence of infection. Further, we used a standard 18S rRNA PCR and gene sequencing on positive reactors to identify the Theileria spp. involved. The results of this survey are described below.

Samples Collection.
Jugular venipuncture was used to collect blood in EDTA from convenience samples of apparently healthy livestock (cattle, goats, sheep, donkeys, and horses) on five Caribbean islands [22]. This study was reviewed and approved by the Institutional Animal Care and Use Committee of the Ross University School of Veterinary Medicine (RUSVM), St. Kitts. Owners of animals gave permission for the blood samples to be collected.

PCRs for Theileria Detection and Species Determination.
All the PCRs were performed on a Roche Light-Cycler 480-II platform with the HMBS gene as an endogenous control [23]. Samples found positive in the pan-Theileria FRET-qPCR were tested in a conventional PCR with primers targeting a highly polymorphic 584-610-nucleotide region of the 18S rRNA gene of Theileria spp. (Table 2, Figure 1) [21]. The amplicons from positive conventional PCRs were sequenced directly with forward and reverse primers to determine the Theileria spp. present (BGI, Shanghai, China) [21] as has been done with 18S rRNA sequences in a number of previous studies [11,14,17,21,24].

Results and Discussion
The sensitive and specific pan-Theileria FRET-qPCR [21] we used in our study demonstrated that a substantial proportion of livestock (8.6%; 65/752) on the five Caribbean islands we studied were infected with Theileria (Table 3). Each island had positive animals and each livestock species we studied was found to be infected with Theileria spp. Sequencing of representative samples of positive 18S rRNA PCR amplicons we obtained ( = 43) showed that there was one recognized Theileria spp. (T. equi) present in the Caribbean, along with five less well characterized Theileria sp. (Tables 3 and 4, Figure 2). The average copy number of 18S rRNA per mL whole blood was relatively low at 116.6 ± 440.8, indicating that the animals we studied were chronically infected.
Theileria equi and the Theileria sp. YW-2014 were the species we identified in equids. Theileria sp. YW-2014 has been described in a Sika deer (Cervus nippon) from Japan but there is little sequence data on the organism with only a 552 bp sequence of the 18S rRNA gene reported in GenBank (AB981984). On the other hand, T. equi is a well-recognized cause of equine piroplasmosis [25], an important disease of horses which has been recognized and studied in the Caribbean [14][15][16]. The organism has also been found in dogs in Spain [26], South Africa [27], and Nigeria [28] with some having clinical signs that responded to appropriate treatment [29]. Our finding that T. equi also occurs in domestic ruminants further expands the host range of the organism. The significance, extent, and consequences of infections with T. equi in domestic ruminants require further investigation.      Numerous species of Amblyomma, Dermacentor, Hyalomma, Ixodes, and Rhipicephalus are confirmed or suspected vectors of T. equi [29]. Of these, only A. cajennense [30], R. microplus [30], R. sanguineus [31], and R. turanicus (unpublished observation) occur in the Caribbean. There is conflicting data that Amblyomma cajennense is a competent vector of T. equi [32] but the tick is localized to Jamaica, Trinidad, and Cuba in the Caribbean [30] and it appears, then, not to be the vector of T. equi we found on Nevis, St. Kitts, and Dominica. Dermacentor nitens, the tropical horse tick, is very common in the Caribbean and the tropical Americas [33]. Although there is no data on the competence of D. nitens as a vector of T. equi and there is contradictory epidemiological evidence [32], PCR positive D. nitens have been found [32] and Asgarali et al. [16] have suggested that this tick is a vector in Trinidad. They also suggested that R. microplus [16], which is very common on cattle throughout the Caribbean, might also be a vector. While there is some evidence that R. microplus is a competent vector [34] and our PCR identified T. equi in cattle on two islands, it seems unlikely that R. microplus is an important natural vector as it is a one host species and transovarial transmission has not been demonstrated [32]. Although R. sanguineus and R. turanicus have been implicated as vectors of T. equi, more recent studies have failed to confirm their role [29]. Further studies are needed to establish the epidemiology of T. equi and its vectors in the Caribbean and the neighboring Americas [32].
Theileria sp. OT3 was first described in sheep, deer, and chamois in Spain [35][36][37] and later in sheep in Italy [38], China [39], and Turkey [40]. Its pathogenicity and vectors have yet to be determined. Ours is the first report of the organism in the Caribbean and also the first report of the Theileria sp. OT3 in goats which further demonstrates that the organism has a wide distribution and host range. Although we used only convenience samples which were not representative of the islands, it is of note that Theileria sp. OT3 was the most prevalent species we detected in ruminants. Similar high prevalences of infections have also been reported in the other countries where the organism has been described. The sequences of our Theileria sp. OT3 were identical to one another and to that of an isolate from China (KF470868) [39]. A phylogenetic relationship tree established for the Chinese isolate showed that the Theileria sp. OT3 forms a separated cluster and that the organism is closely related to T. uilenbergi, T. luwenshuni, and T. ovis. Although the organism has only been found in apparently healthy animals, studies on its pathogenicity seem indicated as, with its high prevalence and wide distribution, it might be causing substantial economic losses for livestock farmers. Theileria mutans and T. velifera are African species that have been reported on Guadeloupe [12] and appear most likely to have been imported with cattle from West Africa in the 18th Century. This might also have been the case with the other African Theileria sp. we identified. Seven of the Theileria we found had an identical sequence to that of Theileria sp. NG-2013a (KF597076) described from waterbuck (Kobus defassa) in Kenya [41]. This organism has been found to cluster with T. equi but might represent a novel taxon. Its vectors and pathogenicity are unknown. One goat we studied had a Theileria spp. with a sequence identical to that of Theileria sp. B15a (JN572700) from an African buffalo (Syncerus caffer) in South Africa [42]. This organism is close to T. mutans which is widespread in Africa where it is transmitted by Amblyomma spp. and causes benign theileriosis. Previously T. mutans has been reported on Guadeloupe based on serological test results [12] and, since we found no confirmatory molecular evidence for the presence of T. mutans, it might be that the cross-reacting antibodies detected in Guadeloupe were against this closely related Theileria sp. B15.
The only non-Theileria sp., the pan-Theileria FRET-qPCR identified in our study, was one that appeared to be Babesia vulpes [43] or a closely related species. This organism was previously known as "T. annae" or the "Babesia microtilike organism" but was reclassified based on 18S rRNA and tubulin-beta gene sequence data [43]. The sequences of the short amplicons we obtained in the pan-Theileria FRET-qPCR were identical to those of B. vulpes (JX454779; KF773740) and had one mismatch with B. microti (AB219802, HQ629933, and LC005772). We were unable, however, to obtain longer amplicons with the standard PCR, probably because there were only very low copy numbers present in the positive animals. We were, then, unable to use this additional sequencing data to confirm that the organism we detected was B. vulpes or determine if it was a closely related species or strain. The pan-Theileria FRET-qPCR we used in our study was designed to detect seventeen recognized Theileria spp., which did not include "T. annae" or B. vulpes (Table 1) [21]. These seventeen recognized Theileria spp. differed from one another by only a maximum of 4 nucleotides in the regions of the primers and the probes used in the pan-Theileria FRET-qPCR. It is of note, then, that these primers and probes enabled the multiplication and detection of B. vulpes, or a closely related species, with 8 nucleotide differences and it therefore appears that the pan-Theileria FRET-qPCR might not be as genus specific as first thought. Further work is currently underway in our laboratory to more clearly characterize the B. vulpes or closely related organism found in the Caribbean.

Conclusions
Our study has confirmed the sensitivity of the pan-Theileria PCR in the rapid detection of a wide range of Theileria spp. but has also shown it might detect B. vulpes or closely 7 related organisms. We found livestock infected with Theileria spp. on each of the five islands we studied. While we could not confirm previous reports of T. mutans and T. velifera in cattle, we found that one recognized species, T. equi, four poorly characterized Theileria spp., and B. vulpes or a closely related organism are present in the region. Further studies are indicated to more precisely determine the phylogenetic relationships of these organisms in the Caribbean with closely related organisms from other parts of the world. Also, the prevalences of infections on the different islands should be determined as well as the impact these poorly characterized organisms might have on livestock production, both in the Caribbean and around the world where they are found.