Molecular detection of novel Anaplasma sp . and zoonotic hemopathogens in livestock and their hematophagous biting keds (genus Hippobosca) from Laisamis, northern Kenya

Background: Livestock are key sources of livelihood among pastoral communities. Livestock productivity is chiefly constrained by pests and diseases. Due to inadequate disease surveillance in northern Kenya, little is known about pathogens circulating within livestock and the role of livestock-associated biting keds (genus Hippobosca) in disease transmission. We aimed to identify the prevalence of selected hemopathogens in livestock and their associated blood-feeding keds. Methods: We randomly collected 389 blood samples from goats (245), sheep (108), and donkeys (36), as well as 235 keds from both goats and sheep (116), donkeys (11), and dogs (108) in Laisamis, Marsabit County, northern Kenya. We screened all samples for selected hemopathogens by high-resolution melting (HRM) analysis and sequencing of PCR products amplified using primers specific to the genera: Anaplasma, Trypanosoma, Clostridium, Ehrlichia, Brucella, Theileria, and Babesia. Results: In goats, we detected Anaplasma ovis (84.5%), a novel Anaplasma sp. (11.8%), Trypanosoma vivax (7.3%), Ehrlichia canis (66.1%), and Theileria ovis (0.8%). We also detected A. ovis (93.5%), E. canis (22.2%), and T. ovis (38.9%) in sheep. In donkeys, we detected ‘ Candidatus Anaplasma camelii’ (11.1%), T. vivax (22.2%), E. canis (25%), and Theileria equi (13.9%). In addition, keds carried the following pathogens; goat/sheep keds - T. vivax (29.3%) , Trypanosoma evansi (0.86%), Trypanosoma godfreyi (0.86%), and E. canis (51.7%); donkey keds - T. vivax (18.2%) and E. canis (63.6%); and dog keds - T. vivax (15.7%), T. evansi (0.9%), Trypanosoma simiae (0.9%) , E. canis (76%), Clostridium perfringens (46.3%), Bartonella schoenbuchensis (76%), and Brucella abortus (5.6%). Conclusions: We found that livestock and their associated ectoparasitic biting keds carry a number of infectious hemopathogens, including the zoonotic B. abortus. Dog keds harbored the most pathogens, suggesting dogs, which closely interact with livestock and humans, as key reservoirs of diseases in Laisamis. These findings can guide policy makers in disease control.


Introduction
Livestock in Africa are considered as one of the most valuable agricultural assets for the rural and urban poor, and accounts for about 40% of the agricultural GDP (Malabo Montpellier Panel, 2020). In 2018, Africa's total livestock population was estimated at 2 billion poultry birds, 438 million goats, 384 million sheep, about 356 million cattle, 40.5 million pigs, almost 31 million camels, and 38 million equines (including 30 million donkeys, 6.5 million horses, and 885,000 mules) (Malabo Montpellier Panel, 2020). This represents about one-third of the world's livestock population (Otte et al., 2019). Moreover, livestock production plays a key economic role to the livelihood of pastoralists living in the marginalized arid and semi-arid regions of northern Kenya (Mburu et al., 2017). These pastoralists largely depend on livestock as a source of meat and milk, income from selling livestock, and donkeys and camels also serve as a mode of transport. Pastoralism supports about 20 million people, produces about 90% of the meat consumed in East Africa and contributes to about 13% of the GDP in Kenya (Nyariki & Amwata, 2019).
Unfortunately, livestock production is hindered by pests and diseases, which are endemic in northern Kenya (Perry & Grace, 2009). Hemopathogens of livestock, particularly those of zoonotic importance, are responsible for some of the most serious emerging infectious diseases facing sub-Saharan Africa and the rest of the world (Rosenberg et al., 2018). About 75% of newly emerging diseases currently affecting humans originated in animals (Jones et al., 2008). In Kenya, hemoparasites that cause babesiosis, theileriosis, rickettsiosis, anaplasmosis, and ehrlichiosis are a major impediment to livestock productivity and public health (Kiara et al., 2014).
Bacterial diseases in livestock include bartonellosis caused by Bartonella spp., which is mainly transmitted by biting arthropod vectors such as ticks and reported widely in both wild and domestic mammals such as dogs, cats, and cattle (Ereqat et al., 2016). In addition, brucellosis, a zoonotic disease, has been reported worldwide and mainly causes infections to the genitals of animals, abortion, and fetal death (Probert et al., 2004). Brucella species have been shown to be of high public health and socio-economic importance in northern Kenya (Kairu-Wanyoike et al., 2019). Parasitic protozoal infections, for example African animal trypanosomiasis, cause debilitating diseases in livestock and serious economic losses in Africa (Petersen et al., 2007). Etiological agents such as Clostridium perfringens cause enteric diseases such as enterotoxaemia in both humans and livestock, mostly goats and sheep (Singh et al., 2018).
Livestock act as reservoirs of infectious pathogens that can be transmitted by various vectors. Ticks and biting flies such as Stomoxys spp. and tabanids are vectors of infectious pathogens including bacteria, viruses (e.g., Rift Valley fever viruses), rickettsiae (Coxiella, Anaplasma), and protozoa (T. evansi, T. vivax, T. simiae) (Baldacchino et al., 2013;Narladkar, 2018). Hippoboscid flies, commonly known as keds and belonging to the family Hippoboscidae within the superfamily Hippoboscoidea, are obligate ectoparasites of vertebrates, both domestic and wild animals and birds (Petersen et al., 2007;Rahola et al., 2011). Members of Hippoboscidae act as vectors of many infectious agents including bacteria, viruses, and protozoans (Rahola et al., 2011). Keds cause economic losses in various ways, including annoyance and psychological disturbances produced during the act of biting and feeding, the diseases they transmit (Bargul et al., 2021), and expenditure incurred by farmers in controlling them (Narladkar, 2018). The painful bites inflicted on the bloodmeal host by keds result in skin lesions and by feeding on blood, they contribute to anaemia (Oyieke & Reid, 2003).
In northern Kenya, keds and ticks are common external pests of livestock, found on livestock all year round (Bargul et al., 2021). Keds are known to infest and blood-feed on all livestock species, domestic and wild animals. In addition, keds also feed on humans and in the process, could transmit zoonotic pathogens (Getahun et al., 2020). To date, little efforts have been put into surveillance of pathogens harbored by the livestock and the role of keds in spreading various diseases. This calls for an urgent need for research studies to catalogue livestock infectious and zoonotic pathogens circulating in livestock for a better understanding of disease prevalence, transmission routes, and for control. In this study, we screened for selected hemopathogens (Anaplasma, Trypanosoma, Clostridium, Ehrlichia, Brucella, Theileria, and Babesia spp.) in goats, sheep, and donkeys, and in keds collected from goats, sheep, dogs, and donkeys.

Study site
The study was conducted in Laisamis sub-County (1° 36' 0" N, 37° 48' 0" E) in Marsabit County, northern Kenya ( Figure 1). Marsabit County borders Ethiopia to the North, Turkana County to the West, Samburu and Isiolo Counties to the South, and Wajir County to the East. Laisamis sub-County occupies an area of 20,290 km 2 that comprises five County Assembly Wards, among which Laisamis Ward (3,885 km 2 ), the area of this study has arid and semi-arid climatic conditions (Marsabit CIDP, 2018). The main economic activity in this region is livestock rearing with limited crop production. The main livestock species kept in Marsabit County include approximately 217,360 camels, 2,029,490 goats, 1,851,452 sheep, 420,000 cattle, 81,900 donkeys, and 45,860 poultry (Marsabit CIDP, 2018).

Sample collection
Samples were collected in two field-sampling trips and each sampling site was geo-referenced with a global positioning system (GPS). Goat and sheep blood, keds on goats, sheep, and dog keds were collected in July 2019 along the Laisamis River at Tula Orbora, which is one of the main livestock watering points (1° 35' 16.4" N, 37° 48' 22.5" E). Donkey blood and donkey keds were collected at Sere-e-Sipeni (1°33'14.9" N, 37°49'32.1" E) in February 2020.

Ethical approval
This study was conducted in strict adherence to the experimental guidelines and procedures approved by the International Centre of Insect Physiology and Ecology (icipe) Institutional Animal Care and Use Committee (REF: IACUC/ICIPE/003/2018) and the Pwani University Ethics Review (approval number: ERC/EXT/002/2020). Goats, sheep, and donkeys were handled carefully to minimize pain and discomfort. Verbal consent was obtained from livestock owners prior to collection of samples. Written consent was not possible as the livestock keepers could neither read nor write.

Blood collection
About 5 mL of blood was obtained from the jugular vein of 245 goats (22 males and 223 females), 108 sheep (8 males and 100 females), and 36 donkeys (18 males and 18 females) of both sexes. Each sample was collected into 5 mL EDTA vacutainers (Plymouth PLG, UK), and kept under cold chain during the sampling exercise. Immediately after completion of the sampling process, all blood samples were preserved in liquid nitrogen for transportation to icipe, Nairobi, for molecular detection of pathogens.
Collection and identification of livestock keds Keds that infested goats, sheep, donkeys, and dogs were collected from their hosts by handpicking at night as previously reported (Kidambasi et al., 2020). Freshly collected keds were preserved in absolute ethanol ready for transportation to icipe for molecular screening of pathogens. Keds for use in molecular and morphological identification were sorted at icipe (Nairobi). Species identification based on morphology was done through comparison with known hippoboscid collections at icipe.

DNA extraction
Keds were surface-sterilized with 70% ethanol and left to air dry for 10 min on a paper towel in a clean hood. Each fly was then placed into a clean 1.5-mL Eppendorf tube containing 250 mg of sterile zirconia beads of 2-mm diameter (Stratech, UK). The flies were homogenized in liquid nitrogen using a Mini-Beadbeater-16 for 3 min (BioSpec, Bartlesville, OK, USA). Genomic DNA was extracted from individual flies and blood samples using DNeasy Blood & Tissue Kit (Qiagen, Hilden, Germany), following the manufacturer's instructions.
Genus-specific Anaplasmataceae primers were used for amplification of the 16S rRNA gene of Ehrlichia and Anaplasma spp. (Mwamuye et al., 2017), while Theileria and Babesia spp. were screened simultaneously using primers that target the hypervariable V4 region of the 18S rRNA gene (Gubbels et al., 1999). Clostridium perfringens was detected using specific primers targeting the 16S rRNA gene (Wu et al., 2009). A set of genus-specific primers described by (Probert et al., 2004) that targets the bcsp31 gene was used for identification of Brucella spp. A universal set of primer that targets the trypanosomal internal transcribed spacer I (ITS-I) region was used for detection of animal African trypanosomes (Njiru et al., 2005).
PCR conditions in the Rotor-Gene, Quant Studio and Mic qPCR for detection of Ehrlichia, Anaplasma and piroplasms (Theileria and Babesia) were preceded by an initial enzyme activation step at 95°C for 15 min followed by 10 cycles of denaturation at 94°C for 20 sec, touch-down annealing from 64°C with a decrease of 1°C after each cycle for 25 sec, and primer extension step at 72°C for 30 sec. Then, another 30 cycles each of: denaturation at 94°C for 20 sec, touch-down annealing from 55°C with a decrease of 1°C after every 5 cycles for 50 sec, and extension at 72°C for 30 sec, with a final elongation at 72°C for 3 min.
Specific annealing temperatures of 55°C, 53.9°C, and 63.2°C were used for detection of Trypanosome spp., Clostridium perfringens, and Brucella spp., respectively. The PCR conditions were initial enzyme activation at 95°C for 15 min, 40 cycles of: denaturation at 95°C for 30 sec, annealing for 30 sec, and extension at 72°C for 30 sec, with a final elongation at 72°C for 7 min.
HRM analysis proceeded immediately after PCR with a gradual increase in temperature from 75°C to 95°C with 2 sec increase of 0.1°C between successive fluorescence acquisitions. The melting curves were visualized based on the fluorescence signals and the change in fluorescence with time (dF/dT) plotted against change in temperature (°C).
Rotor-Gene Q Series Software 2.1.0 (Build 9), Quant Studio Design and Analysis Software version 1.5. 1 (Mwamuye et al., 2017), and micPCR Software v2.8.1 were used to assess melt profiles of the test samples in comparison with that of the known positive controls to confirm detection of pathogens. DNA sequencing of representative samples showing distinct melting curves proceeded to identify the pathogens.

Purification of PCR amplicons and gene sequencing
Representative samples with expected and distinct melting curves relative to the known positive controls were amplified in larger PCR reaction volumes of 15 μL. Five μL of the PCR amplicons were resolved through 2% ethidium bromide-stained agarose gel electrophoresis followed by visualization of the DNA under ultraviolet light using Kodak Gel Logic 200 Imaging System (SPW Industrial, Laguna Hills, CA, USA). About 10 μL of each sample with clear bands was purified using ExoSAP-IT PCR Product Cleanup kit (Affymetrix, Santa Clara, CA, USA) following the manufacturer's protocol. Purified samples were then incubated at 37°C for 15 min and 85°C for 15 min in a Proflex thermocycler (Applied Biosystems) prior to Sanger sequencing by Macrogen, Inc. (Amsterdam, Netherlands). All sequences generated by this study were deposited in the GenBank (NCBI) database and assigned accession numbers.  (2) by PCR-HRM ( Figure 2).
Alignment of the edited Anaplasma 16S rRNA sequences with closely related sequences queried on NCBI GenBank nr database, showed that most samples were 100% identical to A. ovis (GenBank accession MG869525). However, some sequences were distinctly different from the queried A. ovis among other sequences with an identity of 96.8% and below ( Figure 3). In addition, alignment of the edited Theileria 18S    The morphological identification of the keds matched with the molecular identification of two ked species: Hippobosca variegata (GenBank accession MW128366) and Hippobosca longipennis (GenBank accession MW128365).

Phylogenetic analysis of Anaplasma spp. 16S rRNA sequences
The phylogenetic tree comparing sequences of 16S rRNA gene fragments of Anaplasma (900-1000 bp) from this study to other sequences of the same gene available in GenBank is presented in (Figure 4). Phylogenetic relationships and molecular evolution were inferred using the maximum likelihood method. Tree Topologies were estimated using nearest neighbor interchange improvements over 1,000 bootstrap replicates. The tree was drawn to scale representing a 2% evolutionary change in nucleotides per site.

Discussion
There is little information about infectious pathogens, particularly zoonotic, circulating in livestock of northern Kenya, due to lack of proper disease surveillance. In the area of study in Laisamis, northern Kenya, we aimed to determine the occurrence and prevalence of selected hemopathogens in livestock (goats, sheep and donkeys) and their predominant ectoparasitic keds (collected from goats, sheep, donkeys and dogs).
Our findings revealed T. vivax as the predominant trypanosome species in livestock and keds, outside the tsetse belts, with an infection rate of 6.7% (26/389) and 22.6% (53/235), respectively. This agrees with a previous report that showed T. vivax as the major case of trypanosome infection outside tsetse-infested areas in western Kenya (Thumbi et al., 2010). This species is known to be pathogenic to goats, sheep and equids (Galiza et al., 2011 et al., 2018). Detection of the canine pathogen, E. canis, in donkeys is not surprising because the pastoralist farmers rear mixed livestock species together with other domestic animals including dogs; thus, the donkeys could have acquired this pathogen from infected dogs through insect or tick bites. Ehrlichiosis is an emerging disease of domestic animals mainly transmitted by ticks and has previously been reported to infect dogs, cattle, humans, and goats (Zhang et al., 2017).
Further, donkeys were found to be infected with the camelassociated bacteria, 'Ca. Anaplasma camelii', previously reported in Kenyan camels (Kidambasi et al., 2020). The presence of this camel pathogen in donkeys could be attributed to co-herding of donkeys with camels, and the donkeys could have acquired the pathogen from infected camels. Keds are common ectoparasites infesting livestock in the study area and have been reported as mechanical vectors of this bacterial pathogen (Bargul et al., 2021). Further research is needed to determine the zoonotic potential as well as the pathogenic role of this pathogen in donkeys and other livestock species. T. equi (13.9%) was also detected in donkeys. Pathogens associated with this genus are among the causative agents of equine piroplasmosis and have previously been shown to infect donkeys in some parts of Kenya including Mwingi (Oduori et al., 2015).

Goats harbored T. vivax, E. canis, T. ovis, A. ovis and a novel
Anaplasma species. Sheep were also found to harbor E. canis, T. ovis, and A. ovis. T. ovis was more prevalent in sheep (38.9%) than in goats (0.8%). This finding is consistent with a study done on livestock in Palestine (Azmi et al., 2019). Ovine theileriosis caused by T. ovis is among the most important infectious diseases affecting small ruminants, leading to significant economic losses to farmers (Al-Hosary et al., 2021). E. canis and T. vivax were detected in most of the samples analyzed, including keds, suggesting these two are the common pathogens circulating in livestock herds in the study area. Further research will be needed to understand the vectors of these pathogens and whether keds are competent vectors of the pathogens.
Additionally, a high prevalence of A. ovis was detected in goats (84.5%) and sheep (93.5%), which is consistent with findings in a study done in Tunisia, North Africa, in goats and sheep by PCR (Said et al., 2015). This high prevalence of A. ovis could also be attributed to ticks that were present in most of the domestic animals. A. ovis is distributed worldwide and considered a major cause of small ruminant anaplasmosis in tropical and subtropical regions of the world, with general clinical effects ranging from fever, fatigue, low milk production and abortion but with a low mortality rate (Stuen & Longbottom, 2011). A previous study in Corsica, France, reported the presence of A. ovis in dairy goats after an extensive survey due to health and production problems encountered in the goat flocks (Cabezas-Cruz et al., 2019b).
We discovered an unidentified Anaplasma-like species in 11.8% (29/245) of goats analysed. The Anaplasma sp. shared 96.77% sequence identity with A. ovis sequenced from goat blood in China (GenBank accession MG869525). The GC content of this novel Anaplasma sp. was 52.9% with clear observation of differences in the bases with the queried Anaplasma spp. from NCBI GenBank nr database (Figure 3). The pathogenic role of this Anaplasma-like species in goats is not understood. However, it is related to A. ovis, which is known to be pathogenic in sheep, goats, and some wild ruminants (Said et al., 2015).
More hemopathogens were detected in dog keds than in other keds, and they include T. vivax, T. evansi, T. simiae, E. canis, C. perfringens, B. schoenbuchensis, and B. abortus. Due to the fact that most dogs had a free-roaming lifestyle in the study region, it is possible that dogs were at a higher risk of being infected with a wide range of pathogens. Therefore, keds collected from dogs acquired these pathogens from infected dogs during their bloodmeal feeding. We were unable to collect blood from dogs during sampling because we lacked proper protective gear against dog bites, and the dogs were not vaccinated from rabies; thus, collecting blood from dogs was considered too high-risk.
Among all the pathogens detected in keds obtained from dogs, E. canis (76%) and B. schoenbuchensis (76%) had the highest prevalence rates. The high prevalence rate of these pathogens could be attributed to the high competition of pathogens circulating in the same host population, considering the possible interactions between the pathogens and, host immune system, and host life cycle as well (Poletto et al., 2015). In previous studies, B. schoenbuchensis was also detected in deer ked by PCR test with a prevalence rate of more than 60% (Szewczyk et al., 2017). B. schoenbuchensis is one of the most important species that cause bartonellosis and has been reported to cause infections in humans, cattle and wild animals such as the cervids in Asia, North America and Europe (Rolain et al., 2003;Vayssier-Taussat et al., 2016). Additionally, Bartonella infection often manifests as various cardiovascular, neurological and rheumatologic conditions, making it a public health concern since pastoralists, farmers and veterinarians who interact with domestic animals are at a high risk of infection (Maggi et al., 2012). There is little information on the presence of E. canis in keds, but this pathogen has been detected in Rhipicephalus sanguineus (the brown dog tick), the biological vector of the pathogen (Cabezas-Cruz et al., 2019a).
This study also reports the first occurrence of C. perfringens in dog keds. This pathogen is an important cause of enteric diseases in humans and domestic animals and is responsible for several forms of enterotoxaemia, which differs in clinical manifestation and severity according to the toxigenic type involved and specific toxins produced (Singh et al., 2018). It affects small ruminants worldwide, causing heavy mortality and significant economic impact (Sumithra et al., 2013). In previous reports, this pathogen has been shown to cause death in dogs due to hemorrhagic gastroenteritis of the gastrointestinal tract; thus further research is needed to better understand the role of this bacterium in enteric diseases of dogs (Schlegel et al., 2012). Among trypanosomes detected in keds obtained from dogs, T. vivax (15.7%) was more prevalent than T. evansi (0.9%) and T. simiae (0.9%), which had low prevalence rates. The presence of these trypanosome species in dog keds suggests that the keds were infected during their bloodmeal acquisition from dogs that were initially infected with trypanosomes from ticks and possibly from other biting flies like Stomoxys. Additionally, the dogs had a free-roaming lifestyle and thus, it is also possible that they were infected when moving into neighboring tsetse-infested regions. Similarly, T. vivax (29.3%) was also more prevalent than T. evansi (0.86%) and T. godfreyi (0.86%) in keds obtained from goats. This low infection rate could be attributed to disease stability in the area, change in climate and seasonal outbreaks (Gutierrez et al., 2006). Both T. vivax and T. evansi have recently been detected in camel keds in northern Kenya (Kidambasi et al., 2020). Further, previous studies have shown T. godfreyi and T. simiae to infect a wide range of domestic animals including pigs, cattle, camels, dogs and goats with T. simiae being highly pathogenic to domestic pigs (Hamill et al., 2013;Simwango et al., 2017).
T. simiae and T. godfreyi are among the Trypanosome spp. that cause African animal trypanosomiasis with tsetse flies being their main vector (Isaac et al., 2016). This is the first report of occurrence of these two Trypanosome spp. in keds.
The molecular data from keds collected from donkeys showed detection of T. vivax (18.2%) and E. canis (63.6%) as the common pathogens. Detection of these pathogens in donkeys and goats, as well as in their associated ectoparasitic keds, shows the xenodiagnostic potential of using keds to indirectly screen for pathogens occurring in their associated hosts. Similarly, a recent report demonstrated the occurrence in keds of pathogens that were similarly present in their camel host from which they were collected, and further proposed the potential use of keds in xenodiagnosis (Kidambasi et al., 2020). However, detection of pathogens in keds does not incriminate them as vectors, but studies should be carried out to determine the vector competence of these keds. Keds are known to transmit mammalian trypanosomatidae of the genus Megatrypanum and are suspected to be vectors of T. avium and T. corvi in birds (Svobodová et al., 2015).
The impact of zoonotic pathogens is often underestimated due to limited surveillance and insufficient data of disease burden in most developing countries (Munyua et al., 2016). This study reveals that the domesticated animals, as well as keds collected from them, carried infectious pathogens of veterinary and public health concern. Notably, we sequenced multiple hemopathogens in dog keds, including zoonotic ones (B. abortus, B. schoenbuchensis, and C. perfringens). Close association of humans with domestic animals infested by keds and other disease vectors increases chances of pathogen transmission. It is therefore crucial to conduct further studies to map out circulating livestock diseases in northern Kenya and establish the role of keds in disease transmission.

Conclusions
We detected various selected infectious hemopathogens present in livestock and their associated ectoparasitic biting keds in northern Kenya, which calls for further surveillance studies to increase the understanding of the epidemiology of livestock diseases and the transmission of zoonotic ones by insect vectors such as keds that also occasionally feed on humans. This will guide the policy makers and livestock farmers in disease control.