Molecular prevalence of emerging Anaplasma and Ehrlichia pathogens in apparently healthy dairy cattle in peri-urban Nairobi, Kenya

Background Anaplasma and Ehrlichia species are tick-borne pathogens of both veterinary and public health importance. The current status of these pathogens, including emerging species such as Ehrlichia minasensis and Anaplasma platys, infecting cattle in Kenya, remain unclear, mainly because of limitation in the diagnostic techniques. Therefore, we investigated the Anaplasma and Ehrlichia species infecting dairy cattle in Nairobi, Kenya using molecular methods. Results A total of 306 whole blood samples were collected from apparently healthy dairy cattle. Whole blood DNA was extracted and tested for presence of Anaplasma and Ehrlichia DNA through amplification and sequencing of the 16S rDNA gene. Sequence identity was confirmed using BLASTn analysis while phylogenetic reconstruction was performed to determine the genetic relationship between the Kenyan isolates and other annotated genotypes available in GenBank. Anaplasma and Ehrlichia species were detected in 19.9 and 3.3% of all the samples analyzed, respectively. BLASTn analysis of the sequences against non-redundant GenBank nucleotide database revealed infections with A. platys (44.8%), A. marginale (31%) and A. bovis (13.8%). All four sequenced Ehrlichia spp. were similar to Ehrlichia minasensis. Nucleotide polymorphism was observed for A. platys, A. bovis and E. minasensis. The Anaplasma species clustered in four distinct phylogenetic clades including A. marginale, A. platys, A. bovis and some unidentified Anaplasma spp. The Kenyan Ehrlichia minasensis clustered in the same clade with isolates from America and Australia but distant from E. ruminantium. Conclusion This study provides the first report of infection of dairy cattle in Kenya with A. platys and E. minasensis, which are emerging pathogens. We conclude that cattle in peri-urban Nairobi are infected with various species of Anaplasma and E. minasensis. To understand the extent of these infections in other parts of the country, large-scale screening studies as well as vector identification is necessary to inform strategic control.


Background
Anaplasma and Ehrlichia species of the family Anaplasmataceae are tick-borne pathogens of livestock with some species known to infect humans and therefore are of both veterinary and public health importance [1]. Anaplasma species documented to infect domestic ruminants including cattle are Anaplasma marginale (A. maginale), A. centrale, A. ovis, A. bovis, A. phagocytophilum and more recently A. platys [2,3]. Anaplasma marginale transmitted by several tick vectors including some Rhipicephalus (boophilus) species is the most common in Kenya causing a severe hemolytic disease in the ruminants. Anaplasma centrale whose only known vector is Rhipicephalus simus [4] causes mild anaplasmosis in cattle [5] but has not been reported in Kenya possibly because the tick vector is absent. Anaplasma bovis transmitted by various species of Amblyomma and Rhipicephalus ticks [6] causes sub-clinical disease in cattle and has been recently been reported in indigenous calves in Kenya [7].
Anaplasma platys transmitted by Rhipicephalus sanguineus typically infects dogs resulting in canine infectious cyclic thrombocytopenia [8]. However, A. platys has also been isolated from cattle neutrophils [9]. To date, there is no information on cattle infection with this pathogen in Kenya. Anaplasma phagocytophilum is a zoonotic species that has been documented to infect cattle resulting in fever, respiratory signs, reduced milk production and infertility [10]. Anaplasma phagocytophilum was detected recently in questing ticks in a National park in Kenya [11]. For Ehrlichia species, E. ruminantium transmitted by ticks in the genus Amblyomma and the emerging Ehrlichia minasensis (E. minasensis) are the only species in that genera known to infect cattle [12,13]. Infections of cattle with E. ruminantium and E. minasensis are mainly characterized by severe fever, anemia, thrombocytopenia and enlarged lymph nodes [14].
Infections of cattle with the pathogenic species of Anaplasma and Ehrlichia cause mortalities and morbidities with subsequent losses in production of dairy and beef products. The losses usually result in marked economic impact to dairy and beef farmers in the tropical and subtropical regions [15] contributing to poverty in the affected households.
In Kenya including the peri-urban Nairobi, anaplasmosis and theileriosis are some of the tick-borne diseases (TBDs) known to cause economic losses to dairy farmers [16,17]. Although ehrlichiosis caused by Ehrlichia ruminantium has been reported in Kenya [7] and approximately 150 million animals are at risk of infection in Africa [18], quantification of its economic impact in Kenya has not been evaluated possibly because of the difficulty in confirming the diagnosis and the fact that this infection commonly co-infects with other TBDs such as East Coast Fever and anaplasmosis [7]. For a long time, the identification of cattle infected with pathogenic Anaplasma and Ehrlichia in Kenya has been based mainly on questionnaire data, clinical signs, microscopic examination [16,19] and serological tests [20,21]. Nevertheless, few studies in Kenya have used molecular techniques such as reverse line blot, polymerase chain reaction and sequencing for the confirmation and characterization of the tick-borne pathogens infecting cattle [7,22]. Previously, there had been clinical cases reported to the University of Nairobi Veterinary Hospital, Kenya presenting with unspecific clinical signs such as unthriftness and loss of body conditions. On screening of cattle from the areas where the clinical cases had originated, E. ruminantium was identified by antigen detection using Enzyme-linked Immunosorbent Assay (ELISA) [23]. Additionally, microscopic examination of blood from these cattle revealed that some of them had Ehrlichia-like inclusion bodies in the white blood cells suggesting infections with other unknown potentially pathogenic haemoparasites [23]. Hence, there was need to identify and characterize those Ehrlichialike inclusion bodies observed in these animals.
Therefore, to further characterize the Anaplasma and Ehrlichia species infecting dairy cattle in Kenya, this study used molecular tools to determine the prevalence and genetic profiles of various species circulating in apparently healthy dairy cattle in periurban Nairobi. Sequence identities and phylogenetic analysis were used to determine the presence of emerging pathogens such as A. platys and E. minasensis. Subsequently, we provide information that will aid in the understanding of molecular epidemiology of these emerging pathogens of livestock in Kenya and possible epidemiological factors contributing to their spread.

Prevalence of Anaplasma and Ehrlichia species
Of the 306 blood DNA samples analyzed, 61 (19.9%) [95% CI 15.6-24.9] were PCR positive for Anaplasma species while 10 (3.3%) [95% CI 1.6-5.9]) were positive for Ehrlichia. The Anaplasma species yielded a specific band corresponding to 424 bp (Fig. 1a) while primers targeting the 16S rDNA gene of Ehrlichia species produced a specific band corresponding to the expected size of 838 bp (Fig. 1b). The distribution of the positive samples in different sub-counties is shown in Table 1. The highest numbers of both Anaplasma 34 (55.7) and Ehrlichia 7(70) infections were found in Kasarani-Ruai Sub-County while Lang'ata had the least number of cattle positive for Anaplasma 6(9.8) infection. Ehrlichia infections were however not detected in cattle in Dagorreti Sub-County.

Sequence identities of the Anaplasma and Ehrlichia species detected
Twenty-nine PCR amplicons for Anaplasma and four for Ehrlichia were sequenced for confirmation of the identities of the detected pathogens. BLASTn analysis revealed that majority, 13(44.8%) of the Anaplasma 16S rDNA sequences were similar to A. platys with sequence identity of between 98.72 and 100% to annotated sequences in Genbank. Nine (31%) of the sequences were similar to A. marginale with a sequence identity of between 99.07 and 100%. Other sequences matched A. bovis 4(13.8%) with sequence identity of between 99.28 and 100% and unidentified Anaplasma species 3(10.3%) sequence identity of 97.85 to 100% ( Table 2). All the four Ehrlichia sequences were similar to those of E. minasensis revealing a sequence identity of between 99.42 and 100% (Table 3).

Phylogenetic analysis
Multiple sequence alignment was done to assess the genetic similarity of the Kenyan isolates. The nucleotide sequences of three A.bovis isolates were conserved while one (MT160357) had three nucleotide polymorphisms at position 267, 268 and 332 (Table 4). Anaplasma platys sequences showed divergence of upto 4% ( Table 5) with regions of nucleotide polmorphism (Fig. 2). The A. platys sequences MT163377 and MT163388 indicated multiple single nucleotide polymorphism while the other five isolates showed a single nucleotide polymorphism (SNP) ( Table 4). All the Anaplasma marginale sequences from this study were however highly conserved sharing 97.6 to 100% nucleotide similarity (< 2.5% divergence) ( Table 6). For E.minasensis, two isolates had conserved sequences while isolates MT163430 and MT163431 appeared to be genetically different showing multiple SNPs (Table 4). The multiple sequence nucleotide polymorphisms observed in the Kenyan isolates of A.bovis, A.platys and E.minasensis indicate that various strains of the pathogens may exist in the cattle in periurban Nairobi.
Phylogenetic analysis was done to understand genetic relatedness of the Kenyan isolates of the two genera with    pathogen, E. ruminantium. The E. minasensis isolated in this study grouped in one clade with other isolates from USA, Australia and Brazil. These isolates were however closely related to E. canis than E. ruminantium (Fig. 4).

Discussion
This study aimed at detecting and characterizing Anaplasma and Ehrlichia species infecting dairy cattle in peri-urban Nairobi. This information is important in guiding development of control and preventive measures against the infections caused by these tick-borne pathogens.
Overall, we observed more Anaplasma infections than Ehrlichia in the study cattle population. Previous studies in Kenya [7], Sudan [24] and Ethiopia [25] also reported more Anaplasma pathogens than Ehlichia. This could be explained by the wide diversity of Anaplasma compared to Ehrlichia species that can potentially infect cattle [3]. Moreover, spatial occurrence of tick-borne pathogens has been associated with the presence of their tick vectors [26], so it could be that tick vectors which transmit Anaplasma species are more widespread in the study area than those of Ehrlichia pathogens.
The distribution of the infections varied across sample sub-counties. A higher proportion of cattle were infected with Anaplasma and Ehrlichia spp. in Kasarani-Ruai Sub-County compared to the other three study areas of Dagoretti, Westlands and Lang'ata. A possible explanation of this is that dairy farmers in Kasarani-Ruai practice mixed production system involving free and zero grazing unlike the other areas where farmers practiced exclusive zero-grazing. Various studies have shown that free grazing cattle have higher risk of tick-borne infections than zero grazed cattle because of high exposure to tick vectors [22].   Anaplasma platys pathogens were detected in the study cattle. Anaplasma platys has been considered an emerging Anaplasma species whose clinical disease is yet to be described [9,27]. Previous studies in Algeria [2], Senegal [27] and Tunisia [28] similarly reported this pathogen in cattle. Yang et al. [29] suggested a possibility of domestic ruminants acting as alternative hosts or reservoirs for A. platys which is typically a canine pathogen [30]. Therefore, the detection of this pathogen in cattle raises questions of host specificity as earlier speculated [31].
Zobba et al. [9] noted that several domestic ruminants can harbor a number of strains of A. platys although these strains have different cell tropism compared to those infecting dogs. The ruminant strains infect neutrophils and are thought to be the ancestral pathogens that evolved to adopt to the canine platelets instead [9,30].
We think that infection with A. platys may be associated with co-existence of dogs and cattle in the same households, a common practice observed in the dairy farms in peri-urban Nairobi. It is possible that tick-bites from Rhipicephalus sanguineus which are the main vectors of A. platys [32] may have played a role in the transmission. Screening of dogs for this pathogen can reveal if they are acting as maintenance hosts of the parasite.
Previous studies have documented the zoonotic potential of A. platys causing human disease characterized by headaches, intermittent edema and muscle pains [33]. In this regard, detection of A. platys pathogens in this study would indicate a possible zoonotic health risk to cattle owners who are in constant contact with their cattle.
In this study, cattle were also found to be infected with A. marginale, a common pathogen of cattle that has   [35]. However, cattle in this study were apparently healthy suggesting a possible endemic situation or the animals had a persistent infection (PI) state which is known to occur in A. marginale infections [36]. Thus, infected animals can appear apparently healthy despite harboring the pathogen. Anaplasma bovis which is a monocytic pathogen of ruminants was also detected in this study. Different tick species in the genera Amblyomma and Rhipicephalus have been documented to transmit this pathogen [6]. Similar studies have detected this pathogen in Kenya [7], China [37], South Korea [3] Tunisia [38] and Algeria [39]. Despite it causing a mild disease in cattle, some infected animals have been shown to manifest with decreased weight gain, fever and lymphadenopathy [40]. Other studies have documented sub-clinical infection with this parasite where animals don't show clinical signs of the disease despite the infection [41] and this may have been the case in our study.
To date, bovine anaplasmosis in Kenya is mainly known to be caused by A. marginale and to some extent A. bovis [7,16,22]. However, we have detected A. platys pathogens for the first time in cattle in Kenya possibly Fig. 3 Maximum Likelihood tree of Anaplasma spp. constructed using partial sequences of 16S rDNA gene. The tree is drawn to scale with branch lengths measured in the number of substitutions per site. The analysis involved 29 nucleotide sequences from this study and 12 others obtained from Genbank. The tree shows the phylogenetic relatedness of Anaplasma isolates obtained from cattle blood in Kenya marked with dark box and sequences from other countries. Neorickettsia risticii was used as the outgroup. Sequence accession numbers are given at the end of each isolate contributing equally to the disease burden in dairy cattle. Further studies to investigate this pathogen using more specific genes such as membrane surface proteins (Msps) [42] are justified.
An emerging pathogen E. minasensis in the Ehrlichia genera was also detected in this study. This novel pathogen was initially reported in Canada [43] and Brazil [44] but has since been isolated in Ethiopia [45], South Africa [46], Pakistan [47] and China [48]. The clinical disease due to E. minasensis is variable with some reports of severe disease [14] and at times sub-clinical disease as observed in this study [45]. Screening of ticks collected from animals may confirm if the pathogen is common in cattle.
Although the specific ticks that transmit E. minasensis have not been well studied [49], its detection and transstadial transmission by Rhipicephalus microplus ticks has been documented [50]. Other tick species may still transmit the pathogen in areas where R. microplus is absent [45]. Indeed, Iweriebor et al. [46] detected this pathogen from R. appendiculatus, R. evertsi eversi, R. sanguineus and Ambylomma hebraeum ticks. Some of these tick species were observed to be infesting some animals in this study. It is possible that these ticks could be involved in transmission of this pathogen.
Similar to other studies [38,51], phylogenetic analysis based on 16S rDNA was used to infer genetic diversity of Anaplasma and Ehrlichia species. However, other authors have used 16S rDNA gene in combination with other genes such as heat shock protein (groEL), citrate synthase (gltA), 23S rDNA and major surface protein 4 gene (msp4) [2,52]. Combined gene assays have been used to enhance sensitivity since genes with multiple copies such as membrane surface proteins (MSP) are more sensitive for detection of Anaplasmataceae while more conserved genes (16S rDNA) are useful for database cross matching and sequence comparisons [53]. In this study, A. marginale isolates were found to be highly conserved indicating sequence divergence of less than 2.5% and clustered together with those from USA, Uganda, Iran and Australia similar to findings by Rjeibi et al. [39]. Anaplasma bovis, A. platys pathogens and E. minasensis strains detected in this study indicated certain levels of nucleotide polymorphism suggesting various strains of the pathogens may exist in the study cattle. This may be related to the increased cattle movement from other regions of the country for slaughter at the country's major export abbatoirs located in Nairobi County. Extensive animal movement has been associated with development of new strains and introduction of the tick-borne pathogens to new geographic areas [38,54].
In agreement with previous studies, phylogenetic analysis also indicated that E. minasensis is closely related to E. canis and distantly related to E. ruminantium despite infecting similar hosts [12,55]. The clinical presentation of E. minasensis in cattle has been observed to be similar to the acute form of disease by E. canis in dogs [44,56]. Cabezas-Cruz et al. [57] links the close relatedness of the two pathogens to possible evolution of E. Fig. 4 Maximum Likelihood tree of Ehrlichia spp. constructed using partial sequences of 16S rDNA gene. The tree is drawn to scale, with branch lengths measured in the number of substitutions per site. The analysis involved 4 nucleotide sequences from this study and 10 others obtained from Genbank. The phylogeny shows the relatedness of E. minasensis isolated from this study marked with blue dot with other isolates from USA, Brazil and Australia and its relation to E. canis and E. ruminantium. Anaplasma phagocytophilum was used as the outgroup minasensis from highly variable strains of E. canis. The detection of this novel Ehrlichia species suggests that it could be circulating in cattle in Kenya and its pathogenicity in the affected animals needs to be determined.
Despite the animals in this study not presenting with the clinical signs of the diseases caused by the pathogens they harbor, poor animal husbandry practices which are commonly practiced in smallholder dairy farms in periurban areas of Nairobi [58], causes stress to the animals consequently predisposing them to possible flaring up of clinical disease and mortalities [59]. The detection of these pathogen therefore highlights the importance of continued investigation into tick-borne diseases and the need for effective diagnosis and prevention.

Conclusion
The dairy cattle from peri-urban Nairobi, Kenya are infected with a range of Anaplasma species and Ehrlichia minasensis even though clinical disease was not evident. There is need for accurate diagnosis and effective tick control so as to reduce infection of cattle with these pathogens. To the best of our knowledge, this study provided first reports of cattle infected with A. platys and E. minasensis pathogens in Kenya. Extensive epidemiological studies would be necessary to determine the extent and pathogenicity of the newly detected A. platys and E. minasensis in cattle. Moreover, investigation into tick vectors involved in their transmission will be needed to inform strategic disease management and control. Moreover, there is need for further investigation of the unidentified Anaplasma species using more specific genes such as membrane surface proteins (Msps) [42].

Study area and design
The study area and design have been described in detail previously [23]. This was a cross-sectional study undertaken between January and May 2017. For purposes of data collection, Nairobi County was divided into four quadrants and for each quadrant purposive sampling was used to select the sub-county with high cattle population. Dairy farms in the four sub-counties in periurban Nairobi, namely; Dagoretti, Lang'ata, Kasarani-Ruai and Westlands were included to this study. These sub-counties are part of the peri-urban areas of the Nairobi County where dairy production has been established to meet the high milk demands of the urban population. Animals of different age-groups ranging from 3 months to 8 years were randomly selected and sampled for whole blood.

Collection of cattle blood samples
Three milliliter of whole blood were collected in ethylenediaminetetraacetic acid (EDTA)-coated vacutainers from the coccygeal vein of 306 apparently healthy dairy cattle. The samples were collected as follows: Dagoretti (n = 116), Lang'ata (n = 55), Kasarani-Ruai (n = 110) and Westlands (n = 25). The samples were then transported in an ice-box to the Molecular Laboratory at the Department of Public Health, Pharmacology and Toxicology of the University of Nairobi and stored at -20°C pending subsequent analysis.

Extraction of Anaplasma and Ehrlichia DNAs
Whole blood genomic DNA (gDNA) was extracted from aliquots of 200 μl using QIAamp DNA Blood Mini Kit (Qiagen, Hilden, Germany) following manufacturer's instructions. The DNA concentration and quality were assessed using QIAxpert (Qiagen, Hilden, Germany). DNA samples were then stored at − 20 0 C.
A forward primer EHRF 5′-AGCTGGTCTGAGAG GACGAT-3′ and a reverse primer EHRR 5′-GAGTGC CCAGCATTACCTGT-3′ targeting an approximately 838 bp of Ehrlichia 16S rDNA were also designed. PCR amplifications were performed using a thermal cycler (Applied biosystems Veriti 96 well, ThermoFisher). The Anaplasma and Ehrlichia 16S rDNA were amplified in a final volume of 20 μl reaction, each containing 3 μl of genomic DNA, 10 μl Master-mix (Taq PCR 2x mastermix, Qiagen, Germany) and 10 μM final concentration of each primer. The thermocycling conditions for Anaplasma involved a pre-denaturation at 95 o C for 5 min followed by 40 cycles of denaturation at 95 o C for 45 s, annealing at 57 o C for 45 s and extension at 72 o C for 45 s. A final cycle of extension at 72°C for 7 min was performed. The amplification conditions for Ehrlichia 16S rDNA involved an initial denaturation cycle at 95 o C for 5 min followed by 35 cycles of denaturation at 95°C for 45 s, annealing at 62 o C for 45 s and extension at 72 o C for 45 s. The amplification cycles were followed by a final cycle of extension at 72 o C for 7 min. Double distilled water was used as negative control for both assays. The amplified products were electrophoresed using 1.5% agarose gel in Tris-Borate-EDTA (TBE) buffer, pH 8, stained with Ethidium Bromide and visualized using UV-illuminator (GelMax® Imager, UK). The sizes of the amplicons were determined using molecular ladder (Gelpilot 1 kb plus ladder (100), Qiagen, Germany).
Purification and sequencing of Anaplasma and Ehrlichia DNA PCR amplicons in gels were excised and purified using QIAquick Gel Extraction Kit (Qiagen, GmbH, Germany) following the manufacturer's protocol. The purified DNA were sequenced at Macrogen Europe Laboratories (Amsterdam, The Netherlands). Sequencing was done using the same forward and reverse primers as for the PCR reactions. The obtained sequences were viewed and manually verified using chromatogram peaks, edited and assembled using CLC Main Workbench 6.8.3 software (CLC bio, Qiagen GmbH, Germany).

Data analysis
Bioinformatics analysis of the parasite 16S rDNA sequences was done by using Basic Local Alignment Search Tooln (BLASTn), multiple sequence alignment and phylogenetic analyses. Sequence identities of the Anaplasma and Ehrlichia species were confirmed by BLASTn analysis [61] at https://blast.ncbi.nlm.nih.gov/ Blast.cgi. Multiple sequence alignment was done using Log-Expectation (MUSCLE) v3.8.31 [62]. Sequence similarity was calculated using Clustal Omega to obtain identity matrixs [63]. A phylogenetic reconstruction was done using MEGA 6.0 [64]. The evolutionary history was inferred by using the Maximum Likelihood method based on the Tamura-Nei model [65]. Initial trees for the heuristic search were obtained automatically by applying Neighbor-Join and BioNJ algorithms to a matrix of pairwise distances estimated using the Maximum Composite Likelihood (MCL) approach and then selecting the topology with superior log likelihood value. All positions containing gaps and missing data were eliminated. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) was shown next to the branches.

Nucleotide sequence accession numbers
The partial 16S rDNA sequences obtained in this study were deposited in the GenBank under the following accession numbers; MT163376 to MT163388 for A. platys, MT160355 to MT160358 for A. bovis, MT163438 to MT163446 for A. marginale, MT163683 to MT163685 for unidentified Anaplasma species and MT163429 to MT163432 for E. minasensis.