Molecular detection of concurrent infections of Anaplasma sp. Omatjenne, Theileria mutans, Babesia bigemina and Anaplasma marginale in calves and yearlings in a tick endemic Guinea savannah ecosystem in Cameroon.

Background: Ticks play a major role in limiting protable livestock production in sub-Saharan Africa and the region is beleaguered by a paucity of data on diseases implicated in high morbidity and mortality in most of its livestock production systems. Methods: In a tick endemic Guinea savannah ecosystem, ticks infesting calves and yearlings during their rst eighteen months of life were collected weekly and morphologically identied. PCR, Restriction Fragment Length Polymorphism (RFLP) and sequencing were applied on DNA of tick-borne pathogens in animal blood buffy coat to amplify and characterize the 16S rRNA genes of Anaplasma , Ehrlichia and 18S rRNA gene for Babesia spp . Results: Of 31 364 adult ticks, nymphs and larvae collected from the 20 experimental animals, 6525 (20.1%), 7399 (23.6%), 990 (3.2%), 16450 (52.4%) were Amblyomma variegatum, Rhipicephalus spp ., other genera, and tick nymphs/larvae respectively . Tick infestation rates during the rainy season for A.variegatum, Rhipicephalus spp . and nymph/larvae were 50.3, 12.4 and 28.3, respectively, while in the dry season the proportions were one, 26.2 and 72.8 respectively. High Rhipicephalus spp. ( Boophilus ) infestation rates signicantly reduced haematocrit (P<0.01). All animals had mixed infections of haemoparasites. The presence of Anaplasma marginale and Anaplasma sp. Omatjenne in the blood signicantly reduced haematocrit (p<0.0001) while Babesia bigemina and Theileria mutans had no effect (P>0.05). No E. ruminantium was detected. We report for the rst time in Cameroon, the detection of Anaplasma sp. Omatjenne and T.mutans infecting all and four of these animals, respectively. Babesia bigemina , and Anaplasma sp. Omatjenne concurrently occurred in all 20 experimental animals, A.marginale in 15. The mean rst-time contact periods (in weeks) for B.bigemina , T.mutans , Anaplasma sp. Omatjenne and A.marginale were 15(3-37), 30(9-43), 21(5-55) and 25(7-55) respectively; and they were not signicantly different (P>0.05). Conclusion: The absence of disease conditions demonstrate an endemically stable situation in the region for these infections. With no clinical data on A. sp. Omatjenne and T.mutans infections in this area, further insights into their epizootiology should be of interest. monitor and conrm natural transmission by ixodid ticks of four tick-borne pathogens: Babesia, Theileria, Ehrlichia and Anaplasma in zebu Gudali (Bos indicus) calves and yearlings during their rst eighteen months of life. The study determined animal age at rst contact with the different pathogens, the effect of tick infestation burden and season on pathogen presence in blood and packed cell volume of the animals in the tick-endemic ecosystem.


Introduction
In Sub-Saharan Africa, ticks and tick-borne diseases are to a large extent responsible for the stagnation in the livestock industry and their ine cient management constitutes an important limiting factor to beef and dairy livestock production, particularly in smallholder farms with low input options [1].The sub-Saharan African region is beleaguered by a paucity of data with respect to what extent ticks and tickborne diseases are implicated in the high morbidity and mortalities recorded in most livestock production systems. In Wakwa, located in the High Guinea Savannah zone of Cameroon, tick burdens are in uenced by season and infestation varies greatly between breeds as well as amongst individuals of the same breed [2,3,4]. About 63% of cattle mortalities at IRAD Wakwa, have been associated with ticks and tick-borne diseases [5]. The situation is thought to be similar in livestock production systems where de cient tick control programmes persist [1]. Exotic taurines and cross-bred animals between local cattle and exotic ones, with the hope of improving their productivity, are unable to adapt in the local production systems plagued with ine cient tick control options [unpublished data]. The resulting heavy losses through high morbidity and mortality rates [5] contribute to poverty among local breeders with low input systems, frustrates and hinders them from developing the beef and milk industries.
Apart from causing disease, tick infestation also has a negative impact on zebu cattle live weight gain [2].
Factors favouring tick-borne diseases in low input traditional extensive production systems include the high cost of acaricides, irregularity in supply and lack of knowledge concerning the risks from various tick species and management of acaricides [6]. Hand tick removal from animals is still very popular as a means of tick control in the High Guinea Savannah of Cameroon [1]. The abundance of different genera of ixodid ticks in most grazing pastures of Cameroon suggest that tick-borne pathogens (babesiosis, ehrlichiosis, anaplasmosis and theileriosis), transmitted by these ticks, are important factors limiting livestock productivity, yet studies on their relative importance with respect to early calf morbidity and mortalities in this area are almost inexistent.
The general perception amongst local livestock owners is that Ehrlichia ruminantium causes huge economic losses to the livestock industry in the region [1]. Adult A. variegatum are the main vectors of E.ruminantium in cattle [7,8] although other tick species in the region are possible vectors of the disease. Taken together, the role of these four tick-borne pathogens in constraining ruminant production (usually occurring in de nitive hosts as high morbidity and mortality) in the pastoral regions in sub-Sahara Africa is only faintly understood by pastoralists and the available information is obsolete because most of the available results are based on assumptions [5], serology and microscopy with obvious limitations [9,10,11,12].
In this longitudinal study we used molecular tools (PCR-RFLP analysis and sequencing) to identify, monitor and con rm natural transmission by ixodid ticks of four tick-borne pathogens: Babesia, Theileria, Ehrlichia and Anaplasma in zebu Gudali (Bos indicus) calves and yearlings during their rst eighteen months of life. The study determined animal age at rst contact with the different pathogens, the effect of tick infestation burden and season on pathogen presence in blood and packed cell volume of the animals in the tick-endemic ecosystem.

Animals and production system
Twenty naive zebu Gudali female and male calves born on the experimental animal farm of IRAD, Wakwa, located in the High Guinea Savannah (Adamawa plateau) of Cameroon were successively recruited into the study soon after birth (between October 2005 and May 2006) and were identi ed by ear tags. The animals were raised in a sedentary low input traditional cattle production system with grazing in fenced paddocks. Salt supplementation was all year round and cotton seed cake was fed only during the dry season. From birth, the experimental calves were left to graze with their mothers in the same paddocks in both pre-and post-weaning management. The calves were weaned at the age of nine months.

Local conditions including altitude, seasons, temperature and hygrometry
The experimental farm is located at E07", N013"35.895' and at an altitude of 1217 metres. The high altitude of this region (ranging between 1000 and 1300 meters) provides a relatively cool climate but about two decades ago the effect of climate change saw increasing temperatures ranging between 22-25°C and the average annual rainfall decreasing to between 900 mm to 1500 mm [13]. However, the farm's weather station recorded 1600 mm of rainfall during the year of the study. The climate is of the Sudanese tropical type with two seasons: the dry season that occurs from November to March, followed by the wet season (April to October). The area is covered by discontinuous wooded vegetation consisting of savanna grasses such as Hyparrheenia, Panicum and Sporobolus.

The Gudali breed
The Gudali is a short-horned and short-legged zebu cattle (Bos indicus) mainly occurring in Nigeria and Cameroon. A small population of the Gudali breed have also been seen in Burkina Faso (personal communication). They have a high potential for beef and consist of two major sub-types: the Sokoto and Adamawa Gudali. The latter comprises three regional variants, namely the Ngaoundere, Banyo and Yola Gudali [14]. Only the Ngaoundere Gudali were used without spraying with any acaricide during the entire study period.

Collection of blood and identi cation of ticks
Once a week, the animals were restrained and all ticks were removed using a forceps. The ticks were stored in labelled vials containing 70% alcohol and as soon as possible they were counted and identi ed morphologically using a low-power stereomicroscope with the aid of standard identi cation keys [15]. All 16450 Amblyomma spp. collected, 1101 nymphs and larvae and 573 ticks of the Boophilus (Rhipicephalus) genera were examined for spp. identi cation ( Figure 1).
Blood was collected from the jugular vein of each animal fortnightly for the rst year and thereafter once a month until the animal attained the age of 18 months when it was removed from the study. The blood was used for measurement of packed cell volume (PCV) and DNA extracted from its buffy coat. Each buffy coat sample was stored at +4°C until it was used for DNA isolation.
DNA extraction, ampli cation and analysis DNA extraction from the buffy coat was undertaken using the PureGene Genomic DNA Isolation kit (Biozym, Belgium) following the instructions of the manufacturer. The DNA samples were ampli ed by semi-nested polymerase chain reaction (PCR) and gel electrophoresis was used to visualise the amplicons. The analysis was based on the ampli cation of the 18S rRNA gene for the detection of all species of Babesia and was similar to that described by Devos & Geysen [16]. To amplify the 16S rRNA gene for the detection of Anaplasma spp. and Ehrlichia spp. the primers used were developed at IMT (Antwerp, Belgium).
The isolated DNA was used for species-speci c detection of the pathogens by restriction fragment length polymorphism (RFLP).

Polymerase Chain Reactions
In each reaction microtube, 5 µl of template and 20 µl of the master mix were added. The latter contained 1µl of Yellow Sub TM (Geneo BioTech Hamburg, Germany), 11.2 µl distilled water, 5µl of buffer (20 mMTris-HCl, pH 8.4; 100mM KCl), 1.6µl of MgCl (25 mM), 0.2µl of dNTP (100mM each), 0.4 µl each of Babesia, Theileria, Anaplasma and Ehrlichia primers ( Table 1) (25µmol/µl) and 0.2µl of Taq polymerase (5U/µl). A thin layer (2 drops) of mineral oil was placed over the content of all microtubes. The whole tube was put in a preheated (84°C) thermocycler (PTC-100 TM or T3 Biometra R , Westburg, Germany) programmed as follows: denaturing at 92°C for 30 seconds, annealing at 62°C for 45 seconds, extension at72°C for 1 minute and this whole cycle was undertaken 39 times. A last extension of 8 minutes was added before further analysis. Double distilled water was used as a negative control and a known Babesia bigemina, T.mutans, Anaplasma or Ehrlichia sample was used as a positive control in each case.
In the second round reactions, microtubes containing 24.5µl of the master mix consisting of 1µl ofYellow Sub TM (Geneo BioTech Hamburg, Germany), 15.7µl of distilled water, 5µl of buffer solution, 1.7 µl of MgCl 2 (25 mM), 0.4 µl of each primer, 0.13µl ofTaq polymerase and 0.5µl of the rst round PCR product (template) were each covered with 2 drops of mineral oil. The negative and positive control samples were similarly prepared. The reaction tubes were then preheated and maintained at 84°Cand the thermocycler was programmed as follows: denaturation at 92°C for 30 seconds, annealing at 62°C for 45 seconds, extension at 72°C for 1 minute and this whole cycle was undertaken 24 times. A last extension of 8 minutes was added before further analysis. A total of 548 samples were examined using this nested PCR technique.
Five microliters of each PCR product was mixed with 2 µl of loading buffer and loaded into a 2% agarose gel wells and allowed to migrate in the agarose gel in TAE (0.5%) buffer at 100 Volts for 20 minutes. Thereafter, the gel was dipped in an ethidium bromide solution (1µg/ml) for 30 minutes to permit visualisation of the DNA using UV light. A photograph of each gel was made (Figure 2 and 3).

Restriction Fragment Length Polymorphism (RFLP) and Sybr Green digestion by enzymes
Positive PCR samples were further analysed by RFLP to determine the pathogen species implicated. Four to six microliters of each positive PCR product was digested by restriction enzymes (Hind61 incubation temperature at 37°C and BseDI incubation temperature at 55°C overnight) in a mixture of RO-DI water, the enzyme and corresponding buffer in a nal volume of 15 µl consisting of 11 µl of reaction mix and 4-6 µl of PCR product.
Four to 6µl of enzyme-digested product was mixed with 2µl of loading buffer and loaded into marked wells of a polyacrilamide (PAGE) gel that was totally immersed in TBE buffer. The power unit was set at 100Volts and allowed to run for 2hours 30 minutes. After migration, the gel was transferred into Sybr Green and maintained therein for 40 minutes. All positive PCR samples were submitted to RFLP where enzymatic digestion and their different fragment pro les were used to identify most of the parasite species.

Sequencing and Blast
On RFLP we could not distinguish the fragment pro le of Theileria spp. from Ehrlichia spp. For that reason, we had to sequence our PCR products to con rm the diagnosis of infection by Theileria spp. and/or Ehrlichia spp. In each case one animal with a respective positive RFLP pro le was randomly selected for sequencing.
Puri cation of the PCR products was done using QIAquick® puri cation kit (QIAGEN®) according to the manufacturer's protocol. Cloning was then carried out using the TOPO TA Cloning® kits 2006 (Invitrogen™, California USA) according to the manufacturer's protocol. Positive colonies were selected for sequencing and further analysed by comparing the sequences to the NCBI nucleotide database using Blast and Jalview 2.8.2 [17].

Analysis of collected data
Data was collected from each animal only during its rst 18 months of life. Adjustments were made to permit data collected from animals with the same age and during the same season to be grouped together for analysis. Analysis of covariance to look for the effect of season and tick infestation rate on PCV was undertaken. A logistic regression was undertaken to determine the effect of the different tick genera infestation rates on animal blood parasite/bacteria incidence. The logit model employed the maximum likelihood function using the Newton-Raphson algorithm. Threshold signi cance value was 5%.

Results
Of 31

Mortalities
Two of the experimental animals died of intestinal blockage due to Toxocara vitulorum infestation: one at 9 and the other at 11 months age. Of these, one tested positive for Babesia bigemina and Ehrlichia spp. while the other was positive for the four pathogens. Both of them presented neither clinical nor postmortem signs for these infections.
Effect of tick infestation rate on PCV and the incidence of blood parasites Mean PCV were 33.3 and 33.5 in the dry and rainy season respectively. Season had no signi cant effect (P>0.05) on haematocrit (normal value range 32-45%). From a logistic regression that was run to determine the effect of tick infestation rate on the incidence of blood parasites (

Molecular analysis of blood samples
Animals positive for Babesia and Theileria gave a PCR product at 939bp (Figure 2) while animals positive for Anaplasma and Ehrlichia gave a product at 700bp (Figure 3). The PCR product SN404 of calf G144/9 was further processed for sequencing and gave on Blast alignment 99,68% similarity with a T. mutans isolate [18] from Uganda (GenBank: KU206320.1). The sequence alignment of calf G92/67 (PCR nr SN119) with the RFLP pro le typical of Ehrlichia spp. was identi ed to be Ehrlichia sp. Bom Pastor having almost 100% similarity with one that was previously reported in goats in Mozambique [19]. This pathogen had been renamed Anaplasma sp. Omatjenne. The two sequences were submitted to GenBank (MN719893.1for SN404_ and MN719091.1for SN119). Although more than 20% of adult ticks and all nymphs and larvae infesting the animals were A. variegatum only A. sp. Omatjenne but no E. ruminantium infection was detected in this group of animals.
During the study period the following number of animals indicated in brackets were found positive for any of the following pathogens: B. bigemina (20), Theileria mutans (4) (Fig 4 and  5).

Discussion
The practice of tick control by hand removal of ticks from their predilection sites on animals has remained the tick control method of choice for most local farmers even though it is labour-intensive, cannot be 100% e cient and is considered in the local context to permit disease transmission [1].This implies that the livestock owners allow ticks to infest their animals before they follow up to remove them. Thus, the time lapse between tick xation and its removal by the herdsman is enough for the ticks to inoculate pathogen. However, this depends on two factors: the periodicity of tick removal and the pathogen species that eventually get inoculated. For example, if it is a Babesia or a Theileria, one needs three days approximately to get the pathogen transmitted. On the other hand, various forms of acaricides are available for the effective control of ticks in Cameroon but these drugs are well out of the reach of most small-scale livestock breeders mainly because of their high cost and irregularity in supply [6]. Other constraints include the supply of fake and dangerous products by charlatans and the lack of technical knowhow in the appropriate use of the products [6].
The naive calves in this study quickly acquired three pathogens (B. bigemina, A. sp. Omatjenne and A. marginale) at a very early point in life (within the rst two months (3-7weeks) while earliest contact with T. mutans pathogens was much longer (nine weeks). The age at rst contact with the pathogens then increased gradually from ve months onwards, recording the longest periods as 43-55 weeks in a few animals for B. bigemina and A. sp. Omatjenne and 55 weeks for A. marginale and T. mutans. This observation probably depicts the similarity in animal contact with vectors of the pathogens, the homogeneity in both the animal production system and ecosystem used for the study. The trends may also be associated with the less attractiveness of some calves to tick vectors in their early life and/or inherited maternal resistance to ticks that may be associated with breed difference to tick infestation and or transfer of maternal immunity through colostrum. It has been previously established in the same ecosystem that A. variegatum ticks infest lactating animals and older calves to a higher degree than nonlactating and younger calves [20]. The established immunity may wane in old animals and disease state resurgence occurring in stressed animals could cause mortalities.
In the present study, the detection rate of A.marginale pathogen (in 15 out of 20 animals) was lower than that of B. bigemina and Ehrlichia spp. (A. sp. Omatjenne) both of occurred in 20 out of 20 animals exposed. This nding coroborates with that of Chollet [11] who reported a seroprevalence of antibodies to Anaplasma of between 40 and 76% in a cross-sectional study in the same region in which our study was conducted. In these highly endemic areas it would be logical to expect the development of a "pre-immune state" that is due to constant exposure of the animals to the pathogen causing clinical cases to occur mainly in extremely stressed animals. A common stress that occurs in the area is the dry season, characterized by fodder and water scarcity and frequent high cattle mortalities even for the local zebu (Bos indicus) breeds. The detection of T. mutans pathogens in four out of 20 animals in the present longitudinal study had not been previously reported in the area. Given our small sample size, the presence of other pathogenic strains in the region cannot be excluded and ought to be the subject of future investigations.
Despite the high density of A. variegatum ticks, their nymphs and larvae that infested the animals, no clinical case of cowdriosis due to E. ruminantiumalso commonly called heartwater disease) infection was recorded during the rst 18 months of life of the animals. Heartwater disease has been previously diagnosed between 1988 and 1994 (unpublished data) using crushed brain smears stained with Giemsa in adult Gudali cattle and in Gudali X Holstein cross yearlings in the same experimental farm. The observation on Ehrlichia sp. detection rate is in line with the ndings of Merlin et al. [21] who used microscopic examination of crushed brain tissue stained with Giemsa in the adjacent highland plateau of the north Western Cameroon and concluded that the rate of E. ruminantium infection was very low in village herds although A. variegatum infestations were very high. That all animals (20) under two years of age were infected with Ehrlichia sp. (A. sp. Omatjenne) indicates the endemicity of this pathogen in the region and one is poised to believe that such early infections without clinical disease occurrence may confer cross protecting immunity in the local animals against the other Ehrlichia spp. including E. ruminantium. Interestingly, a recent molecular screening of over 1500 adult cattle [22] in the three northern regions of Cameroon (including Adamawa where our study was undertaken) detected only one case of E. ruminantium in Faro et Deo division of the Adamawa region. In other studies in Cameroon E. ruminantium DNA was detected in 142 (28.4%) of 500 un-engorged A. variegatum ticks collected from 182 cattle at, Société de Développement et d'Exploitation des Productions Animales (SODEPA) Dumbo ranch (SDR) in the North West Region and Upper Farms ranch (UFR) in the South West Region. A higher infection rate of E. ruminantium (40.9%) was observed in ticks from SDR than in the ticks (24.7%) collected from cattle at UFR [23]. A serological study reported a high (61-67%) seroprevalence of E. ruminantium and demonstrated the association of heartwater with the presence of A. variegatum in the north region of Cameroon [10]. In Benin, a study undertaken in 4 localities [24] showed that on the overall 10.8% of tested A. variegatum were infected by E. ruminantium (ranging: 8.9-15.6%), while in Burkina Faso the infection rate of ticks by this pathogen was higher, ranging from 9 to 20% [25].Taken together, these previous studies con rm the endemicity of E. ruminantiumin West and Central Africa in the tick vector and indicates that the tick also serves as the reservoir since the pathogen is maintained transstadially. On the contrary our study on Ehrlichia spp. corroborates with those of Vanegas et al. [26] who collected ticks at least ve years later after our study from the same High Guinea Savannah zone of Cameroon from which they sequenced the ompB gene and three intergenic spacers (dksA-xerC, mppA-purC and rpmE-tRNAfMet) and detected Rickettsia africae in A. variegatum ticks, R. aeschlimannii in H. truncatum and H. ru pes ticks, R. massiliae in R. lunulatus ticks, R. sibrica in H. truncatum ticks and Candidatus R. barbariae in R. lunulatus ticks. That E. ruminantium was not found in our samples and those of Vanegas et al. [26] and only one case reported recently in the same ecosystem [22], should be a matter of great concern for epizootiologists. Put together, this data disagrees with the general perception of the local animal breeders that heartwater caused by E. ruminantiumis one of the greatest killer of indigenous young and adult cattle in the High Guinea Savannah zone of Cameroon [1]. This perception of the local cattle breeders may be unfounded. Observed mortalities may be due to other pathogens. It has been indicated that Ehrlichia spp. are maintained in nature through subclinical infections of ruminants (carriers) as well as ticks and have evolved mechanisms to persistently infect mammalian hosts by subverting the innate and adaptive immune responses [27]. In the present study we found for rst time in central Africa that the Ehrlichia sp. infecting calves and yearling cattle was A.sp. Omatjenne previously reported in goats in Mozambique as Ehrlichia sp. Bom Pastor [19] and later renamed E. sp. (Omatjenne). Ehrlichia sp. (Omatjenne), which is apparently apathogenic has also been detected in several ruminants in South Africa including Boer goats [28], in Uganda [18] and in Ethiopia [29,30].The observed epizootiological picture with respect to Ehrlichia spp [E. sp. (Omatjenne)] and E. ruminantium from all the cited studies and ours suggests that other rickettsia or pathogens may be causing pathological or clinical signs similar to those of heartwater in the region since both show crossreactivity [18]. Also, such immune cross-reactivity under natural conditions hinder, through cross-protection the expression of pathology by those otherwise pathogenic Ehrlichia spp. concurrently infecting the animals. We have the tendency to speculate on the possibility that the overwhelming presence of Ehrlichia sp. (Omatjenne) in our study area may protect the animals from infection by a pathogenic species such as E. ruminantium which did not occur in any of our experimental animals during the entire study period. Therefore, there is need for extensive genetic characterization, transmission, clinical and pathogenicity studies of Ehrlichia spp. from the major livestock raising regions of Cameroon and neighbouring countries to generate more data that could clarify the situation. Although A. sp. Omatjenne which we found in our study is apparently apathogenic [28,31] in ruminants, a strain (E. sp. (Omatjenne) 1) of this parasite has been used experimentally to produce disease indistinguishable from cowdriosis in sheep [31] thereby not ruling out its pathogenicity in natural conditions. Anaplasma sp. Omatjenne may contribute to seropositivity to E. ruminantium with which it greatly cross reacts [18]. More so, Anaplasma and Ehrlichia spp. are reported to cause signi cant economic losses to the livestock sector in Uganda [18].
Babesia bigemina was detected in all exposed animals implying that it is endemic in the study area. Semi-nested polymerase chain reaction (PCR) was undertaken with the same primers described by Devos & Geysen [16] based on the ampli cation of the 18S rRNAgene for the detection of Babesia spp. Thus Babesia species were perfectly differentiated based on their RFLP pro les. This technique has been described as being more sensitive than microscopic and serologic examinations [32,33]. It has also been reported that in B. bigemina infection, calves from non-immune cows are as sensitive as adult animals [34,35]. In the present study, calves that had their rst contact with the pathogen within the rst 18 months of life did not reveal any clinical disease condition during their rst 18 months of life. This corroborates with ndings from other studies which reported that cattle between 3 and 9 months of age have higher innate resistance to most tick-borne diseases and consequently disease incidence and corresponding mortality are typically lower for this stock class [36].
It has been reported that amongst many other farm management practices, the reduction of the tick vector population by use of acaricides or pasture management [37], increasing aridity [38] and intensive and prolonged tick control may be associated with breakdown in endemic instability to tick-borne diseases, which generally establishes in indigenous cattle following continuous contact with tick-borne pathogens [39,40] A situation of endemic stability has been reported more than two decades ago in an A. variegatum-infested ecosystem inhabited by indigenous cattle [7]. Maintaining the equilibrium state or balance between host, vector and pathogen is therefore crucial in ensuring endemic stability and thus avoiding economic loses.