Molecular detection and characterization of Babesia bovis, Babesia bigemina, Theileria species and Anaplasma marginale isolated from cattle in Kenya

Background Infections with Babesia bovis, Babesia bigemina, Theileria species and Anaplasma marginale are endemic in Kenya yet there is a lack of adequate information on their genotypes. This study established the genetic diversities of the above tick-borne hemoparasites infecting cattle in Kenya. Methods Nested PCR and sequencing were used to determine the prevalence and genetic diversity of the above parasites in 192 cattle blood samples collected from Ngong and Machakos farms. B. bovis spherical body protein 4, B. bigemina rhoptry-associated protein 1a, A. marginale major surface protein 5, Theileria spp. 18S rRNA, T. parva p104 and T. orientalis major piroplasm surface protein were used as the marker genes. Results B. bovis, B. bigemina, T. parva, T. velifera, T. taurotragi, T. mutans and A. marginale were prevalent in both farms, whereas T. ovis, Theileria sp. (buffalo) and T. orientalis were found only in Ngong farm. Co-infections were observed in more than 50 % of positive samples in both farms. Babesia parasites and A. marginale sequences were highly conserved while T. parva and T. orientalis were polymorphic. Cattle-derived T. parva was detected in Machakos farm. However, cattle and buffalo–derived Theileria were detected in Ngong farm suggesting interactions between cattle and wild buffaloes. Generally, the pathogens detected in Kenya were genetically related to the other African isolates but different from the isolates in other continents. Conclusions The current findings reaffirm the endemicity and co-infection of cattle with tick-borne hemoparasites, and the role of wildlife in pathogens transmission and population genetics in Kenya. Electronic supplementary material The online version of this article (doi:10.1186/s13071-015-1106-9) contains supplementary material, which is available to authorized users.


Background
Babesiosis, theileriosis and anaplasmosis are important tick-borne diseases of cattle worldwide including Kenya. Babesiosis and theileriosis are caused by the protozoan parasites, Babesia and Theileria respectively while anaplasmosis is caused by rickettsial organism known as Anaplasma. In Africa including Kenya, bovine babesiosis is caused by Babesia bovis and B. bigemina and the disease is transmitted by Rhipicephalus ticks [1]. Though, B. bigemina is more widespread, B. bovis infection is the most critical and fatal because of its neurological symptoms [2].
Bovine theileriosis is another tick-borne disease found in many African countries. Tropical theileriosis and East Coast fever (ECF) are the most severe forms of the disease. Tropical theileriosis, caused by T. annulata and transmitted by Hyalomma ticks is distributed in Northern Africa. ECF is probably the most important tick-borne disease in Eastern, Central and Southern Africa. The disease is caused by Theileria parva and transmitted mainly by Rhipicephalus appendiculatus. T. parva natural host is the African Cape buffalo (Syncerus caffer), which serves as source of infection for cattle. Some variants of the parasite are transmitted solely from buffalo to cattle whereas others can spread from cattle to cattle. The other Theileria species reported in Africa are T. mutans, T. taurotragi, T. sergenti/ buffeli/orientalis (referred to as T. orientalis complex) and T. velifera. These species are considered to be either less pathogenic or not pathogenic at all, and only cause benign, moderate to asymptomatic theileriosis [3]. For bovine anaplasmosis, Anaplasma marginale is one of the causative agents and this pathogen is transmitted biologically, by approximately twenty tick species, and mechanically by biting flies and blood-contaminated fomite [4].
B. bovis, B. bigemina, T. parva, T. annulata and A. marginale cause mortalities and morbidities leading to losses in production of milk, meat, and other livestock by-products. Consequently, they cause severe economic losses to livestock farmers involved in dairy and beef production in tropical and sub-tropical regions [5]. In Kenya, theileriosis, anaplasmosis and babesiosis are ranked among the most common causes of economic losses in dairy and beef industry [6][7][8]. The diagnosis of these diseases in Kenyan cattle have relied mostly on clinical signs [6], microscopic examination of blood smears [6,[9][10][11][12] and antibody detection [12][13][14]. A few studies employing molecular methods (polymerase chain reaction (PCR), reverse line blot hybridization (RLB), real time PCR) and genome sequencing have been exploited in Kenya [15][16][17][18]. However, these studies have been limited to a few tick-borne diseases with most of the studies lacking adequate information on their genotypes or even knowledge on their molecular epidemiology, which is critical for the control and prevention of these diseases.
Therefore, this study was done with the main objective of determining and understanding the genetic diversities and molecular epidemiology of some species of Babesia, Theileria and Anaplasma pathogens infecting cattle in Kenya. In particular, blood samples of cattle raised in farms located in Machakos and Ngong districts of Kenya were screened for specific target genes of B. bovis, B. bigemina, Theileria spp. and A. marginale. The sequences generated from these target genes were used to confirm the identity of the pathogens and establish phylogenies to aid in the understanding of their molecular epidemiology in Kenya in relation to other regions of the world.  [19]. Cattle in both farms are kept under a semi-extensive system, characterized by free grazing on pastures. The cattle were kept under semi-enclosed system at night and allowed to graze on pastures where the animals mixed with Masaai cattle grazing in the same area. The grazing together with Masaai cattle was particularly seen in case of cattle kept in Ngong farm.

Study areas and blood samples collection
A total of 154 and 38 blood samples were collected in EDTA-vacutainer tubes from cattle in Ngong and Machakos farms, respectively during a cross-sectional survey done in August 2011. The samples were collected from randomly selected male and female crossbred cattle including adults and yearlings; all of which were apparently healthy. The samples were then transported on ice to the Central Veterinary Laboratory in Nairobi, Kenya and stored at −20°C prior to DNA extraction.

Ethical statement
The managers of surveyed farms were informed about the study and gave their approval for the sampling of cattle. All the procedures were carried out according to ethical guidelines for the use of animal samples permitted by Obihiro University of Agriculture and Veterinary Medicine (Permit for animal experiment: 26-73, 26-72; DNA experiment: 1219-3; Pathogen: 201210-5; 201206-5).

Extraction of hemoparasites DNA
The genomic DNA was extracted at the Central Veterinary Laboratory in Nairobi, Kenya using a commercial DNA extraction kit according to the manufacturer's instructions (QIAamp DNA Blood Mini-Kit, Germany). The extracted DNA samples were transported to the National Research Center for Protozoan diseases in Obihiro, Japan where they were stored at −30°C pending further genetic analysis.
To detect co-infection with T. parva and T. orientalis complex, samples positive for Theileria spp. 18S rRNA were further amplified by nPCR using primers targeting T. parva p104 (p104) gene [23] for T. parva and those targeting T. orientalis major piroplasm surface protein (MPSP) gene [24] for T. orientalis.

Sequencing of the hemoparasites DNAs
All Theileria spp. positive samples and randomly selected positive samples of B. bovis, B. bigemina, T. parva, T. orientalis complex and A. marginale (three samples for each parasite per farm) were used as templates for genetic characterization of the hemoparasites. Nested PCR amplicons of Theileria spp. 18S rRNA, B. bovis SBP-4 and T. parva p104; PCRs amplicons of A. marginale Msp5 and T.orientalis MPSP; and the products of B. bigemina RAP-1a-semi-nPCR (Table 1) were purified by using QIAquick Gel Extraction Kit (QIAGEN GmbH, Germany). The Theileria spp. 18S rRNA amplicons were sequenced with the nPCR primers. DNA sequences with heterozygous base-calling were analyzed using Mixed Sequence Reader web-based program [25] and identified as two distinct sequences. The other purified DNA templates were first sequenced with the amplification primers to identify heterozygous base-calling positions and then cloned in pGEM-T Easy Vector (Promega, USA). Initially, two The primers sets forSBP-4, 18S rRNA, p104 and MPSP genes were used for detection of corresponding pathogens and the products of the last amplification served as template for genetic characterisation. With regard to BbigRAP-1a and Msp5, PCR and nPCR primers were used in pathogen detection, however for genetic characterization, amplicons from a semi-nPCR (BbigRAP-1) and from the first PCR (Msp5) were used positive clones per template were randomly selected and sequenced with pGEM-T Easy Vector-primers (pUC/ M13). When the sequences of the clones did not include all the genotypes identified during direct sequencing, other two clones from the same template were sequenced. All sequencing analysis assays were performed using the Dye Terminator Cycle Sequencing Kit (Applied Biosystems, USA) and an ABI PRISM 3100 genetic analyzer (Applied Biosystems, USA).

Blast analysis, sequence alignment and phylogenetic analysis
The sequenced DNAs were analyzed by BLASTn tool of NCBI GenBank database. The correct species identity was established by comparing the query sequences with those available in the GenBank database. Species confirmation was done when the closest BLASTn match has a 98 %-100 % identity to the homologues found in the Gen-Bank. DNA sequences identities were also computed using the pairwise alignment by EMBOSS NEEDLE software [26]. Multiple sequence alignments were performed using MUSCLE and GUIDANCE algorithms [27]. Phylogenetic analyses were inferred by the maximum likelihood method using MEGA version 6 software [28].

Statistical analysis
Proportions of DNA samples positive for respective hemoparasites per farm were computed and a comparison of pathogen prevalence was done using the chi-square test employing the EPI INFO™ software (CDC, USA, version 7.1.1) and VassarStats [29]. Statistically significant differences were determined at P < 0.05.

Nucleotide sequences accession numbers
The nucleotide sequences of all the genes sequenced are available in the GenBank of the NCBI database under the accession numbers outlined in Table 2.

Identification of Theileria species
Genetic analysis of the sequenced Theilera spp.-18S rRNAs revealed homologues that belonged to seven distinct Theileria species including T. parva, T. taurotragi, T. mutans, T. velifera, T.orientalis complex, T. ovis and Theileria sp. (buffalo) ( Table 3). In particular, T. parva, T. taurotragi, T. mutans and T. velifera were detected in samples from cattle in both farms while T. orientalis complex, T. ovis and Theileria sp. (buffalo) were detected in Ngong farm-cattle only. Co-infections with T. parva and T. taurotragi, T.mutans and T. taurotragi were observed in some blood samples from Ngong farm-cattle. Other samples with T. parva and T. velifera co-infections were seen in the same farm. In contrast, co-infections with T. taurotragi and T.

Mixed infections with Babesia, Theileria and Anaplasma
More than half of the positive samples were infected with at least two hemoparasites, which generally belonged to different genus. Twenty nine different types of mixed infections were seen in Ngong farm with some cattle having up to five pathogens coinfecting the same cattle (Table 4). In Machakos farm, 11 different types of mixed infections were seen with up to three pathogens simultaneously being detected in some samples (Table 5). Single infections with B. bovis, T. velifera and A. marginale were observed in both farms, while single infections with B. bigemina and T. parva were seen in Ngong farm only. In most cases, mixed infections rather than single infection were detected for all hemoparasites in both farms except for B. bovis for which most of positive samples in Machakos farm were single infections (Table 5).

Phylogenetic analysis
Phylogenetic analyses were done to determine whether the tick-borne pathogens are genetically diverse within different geographical regions of the world. Analysis based on SBP-4 gene grouped the Kenyan B. bovis isolates in the same clade (Clade 1) as Egyptian, Ghanaian and South African isolates (Fig. 1). The other B. bovis isolates from Thailand, Syria, Mexico, Brazil, Mongolia and United States of America (USA) were grouped in a separate clade. The Kenyan B. bigemina isolate belonged to the same clade as the isolates from Egypt, Thailand, Syria and Mexico (Fig. 2). A further phylogeny using the Msp5 gene grouped the Kenyan A. marginale isolates in the same clade as the isolates from China, Australia, Brazil and The Philippines (Fig. 3) (Fig. 4). The polymorphism of T. parva p104 nucleotide sequences is shown in an additional file [see Additional file 3]. The phylogenetic analysis based on MPSP gene of T. orientalis/sergenti/ buffeli grouped three of the isolates of this study [KP347560, KP347562 and KP347563] in the same clade and these isolates were classified as MPSP type 3. The divergent isolate [KP347561] belonged to a separate clade and was identified as MPSP type 5 (Fig. 5).

Discussion
This study was done to determine the genetic diversities as well as molecular epidemiology of B. bovis, B. bigemina, Theileria spp. and A. marginale isolates of cattle raised in Machakos and Ngong districts of Kenya. Such information is critical for controlling and preventing infections caused by these pathogens, which leads to loss of livelihoods of many livestock owners [30]. The causative agents of babesiosis, theileriosis, and anaplasmosis were prevalent in the two farms surveyed. These findings are consistent with previous reports ranking tick-borne pathogens as important causes of diseases in Kenyan cattle [6-8, 31, 32]. Though Theileria spp. has always been reported to be the most prevalent hemoparasite in Kenya [8,11,12], this was not the case in Ngong farm as Babesia species were the most prevalent pathogens. However, Theileria species were more prevalent in Machakos farm consistent with the previous studies in Kenyan cattle [8,11,12]. A. marginale was the least prevalent in both farms suggesting that it may not be one of the most prevalent tick-borne diseases in these regions.
The prevalence of B. bovis was higher than that of B. bigemina in Machakos and vice-versa in Ngong farm indicating that the epidemiology of babesiosis in the two farms may be different. The prevalence of B. bigemina reflects the distribution of its vectors, Rhipicephalus evertsi and R. decoloratus [1] which are both present in Machakos and Ngong districts [8,33]. B. bigemina is the  [8,34] and its predominance in Ngong farm is not surprising. The low prevalence of B. bigemina in Machakos farm could be explained by lower exposure to tick vectors. The detection of B. bovis was unexpected since none of its vectors namely, R. microplus, R. geigyi and R. annulatus [1] have been reported in the two districts. Our results, therefore indicate that B. bovis has been overlooked or is becoming endemic in Machakos and Ngong districts. The competition between R. microplus and R. decoloratus may explain why B. bigemina is more widespread than B. bovis in Africa [1]. Hence, the higher prevalence of B. bovis in Machakos farm suggests changes in tick distribution. Uncontrolled animal movement that is common in Kenya [34] or changes in ecological pattern may have contributed to the thriving of B. bovis tick-vectors.
This study revealed higher prevalences of B. bovis and B. bigemina than those previously observed in Western Kenya [16,17]. This may not be surprising as the studies were done in different ecological areas. B. bovis, B. bigemina isolated in the two farms were genetically conserved and closely related to isolates from other African countries.
B. bovis SBP-4 sequences from Kenya were consistent with the findings of [35] which indicated the existence of a "fingerprint" discriminating B. bovis isolates of African cattle from others. However, B. bovis SBP-4 phenogram showed the existence of an "intermediate" clade (Clade 2, Fig. 1) hosting sequences from South Africa which despite being isolated from cattle were not bearing the discriminatory "fingerprint". Hence, further studies on B. bovis SBP-4 gene in other African countries are needed for clarification of the scope, the origin and impact on the parasite of this "fingerprint".
The detection of A. marginale DNA in both farms is consistent with previous studies which reported antibodies to A. marginale in Kenyan cattle [8,[12][13][14]. The sequence identity of the two A. marginale genotypes detected in this study suggests that the parasite isolates circulating in Kenya may be genetically conserved. However, further studies are required to unravel the genetic diversity of the parasite isolates.
Although this study detected Theileria species in both farms, the samples analyzed here were not large enough and elaborate studies will be needed to determine the R.appendiculatus, which transmits T. parva and T. taurotragi as well as Amblyomma variegatum, the vector for T. mutans and T. velifera are known to exist in the districts surveyed [8,33]. Therefore, the presence of these tick vectors may explain the occurrences of the hemoparasites in the farms. T. ovis is known to infect small ruminants [36] while T. buffeli and Theileria sp. (buffalo) infect African Cape buffalo (Syncerus caffer) [37]. The detection of buffalo-Theileria isolates in Ngong farm-cattle may be attributed to their transmission by ticks and interaction of these cattle with nomadic Maasai cattle. These Maasai cattle usually graze together with wildlife including buffaloes. Likewise, Maasai cattle are generally kept with flocks of small ruminants and therefore can be a source of T. ovis-infected ticks for Ngong farm-cattle. Such accidental Theileria infections in cattle have previously been reported in wildlife-domestic animals interface and in areas where there is high animal movement in Kenya [15][16][17]. Before this study, only one MPSP allele of T. sergenti/ buffeli/orientalis had been isolated in Kenya [38,39]. The polymorphism of the T. orientalis MPSP type 3 isolates identified and the description for the first time of T. orientalis MPSP type 5 indicate that at least two strains of this benign Theileria parasite are present in Kenya. None of the T. orientalis MPSP alleles identified in Kenya have ever been associated to disease outbreaks. However, in Ethiopia [40], Burundi [41], India [42], Australia [43][44][45] and New Zealand [46], some T. orientalis strains have caused disease outbreaks. Hence, T. orientalis complex should not be ruled out as a probable cause of disease, particularly for crossbred and exotic breeds dairy cattle reared in Kenya.
T. parva was the most frequent Theileria in the two farms. Although they were positive for both Theileria 18S rRNA and T. parva p104 nPCRs, sequencing of the Theileria 18S rRNA amplicons of 35 samples from Ngong farm and 11 samples from Machakos farm, did not show a sequence specific to T. parva. This is probably due to low levels of T. parva-parasitemia and to several Theileria species co-occurring in the same animal. In the current study, genus specific primers were used to amplify Theileria 18S rRNA and obtained products therefore contained amplicons from several species. Only amplicons derived from species with abundant DNA were reflected in the sequencing chromatogram. Hence, T. parva 18S rRNA was not identified in some samples because it was outnumbered by the 18S rRNA gene of other Theilera species. East Coast fever (ECF) caused by T. parva is the most important tick-borne disease in Kenya, and "immunization" against it is common [6,47,48]. The carrier state is particularly important for this parasite as it contributes to and may be necessary for maintenance of immunity against overt disease [18,36]. Most of T. parva positive animals in this study seemed to be carriers and previous studies [49][50][51] suggested that such status could have been induced by previous "immunization" as well as natural infection. T. parva prevalences obtained were lower than the 67 % observed in Marula, Rift Valley [15], similar to the values recorded by [18] but higher than the RLB data from Western Kenya [16,17]. The detection of T. parva isolate in Machakos farm and not the buffalo-derived genotypes suggested that these cattle are only exposed to the cattle-derived T.
parva. The identification of buffalo-derived genotypes in Ngong farm corroborates previous reports of the occurrence of buffalo-derived T. parva in cattle in Kenya [33,34].
All the cattle in this study appeared healthy, although pathogenic hemoparasites (B. bovis, B. bigemina, A. marginale and T. parva) were detected in their blood. This absence of clinical disease in infected cattle may be attributed to a state of enzootic stability as described in previous reports [1,3,4]. The high rate of multiple infections in both farms, sometimes involving hemoparasites belonging to different genus may be explained by the presence of a range of tick-vectors that exist in the same ecosystem. Benign T. taurotragi, and T. mutans detected in this study have been previously associated with morbidity and mortality in calves in Kenya [52,53]. Perhaps further studies should explore the importance of these hemoparasites in Kenya with regard to Theileria infections in calves. SBP-4, RAP-1a, Msp5 and MPSP genes, and the corresponding PCR assays were exploited for the first time in Kenya. Our findings confirm the value of these assays [21,22,24,35,45] and suggest that they can be used to improve hemoparasites detection in Kenya.

Conclusions
This study has confirmed the occurrences of a range of genetically diverse tick-borne hemoparasites in farms located in two districts of Kenya. The detection, prevention and control of these hemoparasites in cattle should consider their genotypes, the co-infective nature of these pathogens and the role of wildlife in the transmissions of the tick-borne parasites. Therefore, this study will provide a basis for further research on tick-borne hemoparasitic diseases and their molecular epidemiology in Kenya and other regions of the world.

Additional files
Additional file 1: