Multiple Pathogens Including Potential New Species in Tick Vectors in Côte d’Ivoire

Background Our study aimed to assess the presence of different pathogens in ticks collected in two regions in Côte d’Ivoire. Methodology/Principal Findings Real-time PCR and standard PCR assays coupled to sequencing were used. Three hundred and seventy eight (378) ticks (170 Amblyomma variegatum, 161 Rhipicepalus microplus, 3 Rhipicephalus senegalensis, 27 Hyalomma truncatum, 16 Hyalomma marginatum rufipes, and 1 Hyalomma impressum) were identified and analyzed. We identified as pathogenic bacteria, Rickettsia africae in Am. variegatum (90%), Rh. microplus (10%) and Hyalomma spp. (9%), Rickettsia aeschlimannii in Hyalomma spp. (23%), Rickettsia massiliae in Rh. senegalensis (33%) as well as Coxiella burnetii in 0.2%, Borrelia sp. in 0.2%, Anaplasma centrale in 0.2%, Anaplasma marginale in 0.5%, and Ehrlichia ruminantium in 0.5% of all ticks. Potential new species of Borrelia, Anaplasma, and Wolbachia were detected. Candidatus Borrelia africana and Candidatus Borrelia ivorensis (detected in three ticks) are phylogenetically distant from both the relapsing fever group and Lyme disease group borreliae; both were detected in Am. variegatum. Four new genotypes of bacteria from the Anaplasmataceae family were identified, namely Candidatus Anaplasma ivorensis (detected in three ticks), Candidatus Ehrlichia urmitei (in nine ticks), Candidatus Ehrlichia rustica (in four ticks), and Candidatus Wolbachia ivorensis (in one tick). Conclusions/Significance For the first time, we demonstrate the presence of different pathogens such as R. aeschlimannii, C. burnetii, Borrelia sp., A. centrale, A. marginale, and E. ruminantium in ticks in Côte d’Ivoire as well as potential new species of unknown pathogenicity.


Introduction
Ticks are important vectors of many pathogens and are considered as the second biggest vectors of human and animal diseases after mosquitoes [1,2]. Many tick-borne bacterial emerging diseases such as spotted fevers, borrelioses, anaplasmoses, ehrlichioses, and Q fever have been described worldwide [3,4,5]. It was recently shown that in many tropical countries tick-and acari-borne infections play important role in human pathology. In Senegal, for instance, arthropod-borne borreliosis and rickettsiosis were identified in 16.3% of acute fevers recorded by rural dispensaries [6]. Acari-borne tsutsugamushi fever is one of the major causes of acute febrile morbidity in South-Eastern Asia [7]. Investigations of the vectors of tick-borne diseases are one of the main keys to controlling related morbidity [8].
Rickettsia aeschlimannii is an agent of spotted fever which was first identified in a patient returning from Morocco [21]. In this country, it was first isolated from Hyalomma marginatum marginatum ticks [22]. R. aeschlimannii was also reported by PCR in other Hyalomma ticks including H. marginatum rufipes and Hyalomma truncatum ticks collected from camels and cows in Egypt, Algeria, Sudan, and Tunisia [23]. In Western Africa, R. aeschlimannii was also detected in 15% to 95% of H. marginatum rufipes from Mali, Niger, Senegal and Nigeria [12,13,24] and in 6% to 7% of H. truncatum from Senegal [12] but not in Côte d'Ivoire.
Rickettsia massiliae is another SFG rickettsia. Since its description in 2005, R. massiliae infections in humans have been confirmed in Europe and South America [25,26,27]. It is associated with Rhipicephalus ticks. R. massiliae was found by PCR in Rhipicephalus spp. ticks including Rhipicephalus spp. from Côte d'Ivoire [28], Rhipicephalus guilhoni from Senegal [12], Rhipicephalus senegalensis from Guinea [16], and Rhipicephalus eversti from Nigeria [18]. Different borrelioses are caused by bacteria from the Borrelia genus. They are traditionally classified into the Lyme disease group and the relapsing fever group. The former is ecologically associated with hard ticks and is mostly found in the temperate northern hemisphere [29]. Relapsing fever group borreliae are mostly associated with soft ticks and found in subtropical regions worldwide [30,31]. In endemic regions, borrelioses may play an important role, for example in Slovakia [32]. Relapsing fever is one of the most common diseases in several African regions including Senegal [33,34] and east African countries [35]. It is caused by different Borrelia species such as Borrelia hispanica, Borrelia duttonii, and Borrelia crocidurae. B. hispanica was recently detected in 11.6% to 20% of Ornithodoros ticks from northern Africa [31,36]. B. crocidurae is responsible for tick-borne relapsing fever in West Africa. Its distribution in the south is thought to be limited by the 750 mm isohyets [37]. Neither this borrelia nor any other from the relapsing group has been reported in Côte d'Ivoire. A controversial study, based on molecular data, reported 30 cases of borreliosis in Togo but its epidemiology was not identified [38] and studies in neighboring countries did not confirm the presence of borreliosis in west tropical sub-Saharan Africa. In Ethiopia, Borrelia sp. was recently identified by PCR in 7.3% of Amblyomma cohaerens [39]. Phylogenetically, this Borrelia sp. was placed in an intermediate position between Lyme disease and relapsing fever groups.
All bacteria from the Anaplasmataceae family are intracellular mammal parasites, arthropods nematodes, and trematodes [40]. Anaplasma centrale and Anaplasma marginale are two etiological agents of bovine anaplasmosis in ruminants [41]. These species are distributed in tropical and subtropical regions of Africa and naturally infect cattle [42]. They were previously found by molecular biology in ticks in neighboring Mali [43]. These bacteria are often found in Dermacentor, Rhipicephalus, and Amblyomma ticks throughout the world [44]. Ehrlichia ruminantium is responsible for cowdriosis in ruminants with the Amblyomma genus ticks as a vector [45]. Cowdriosis induces mortality in ruminants in sub-Saharan Africa and in islands in the Caribbean where it causes serious losses to animal production [40]. E. ruminantium was previously identified in Am. variegatum in Burkina Faso but not in Côte d'Ivoire. No cases of human ehrlichiosis or anaplasmosis have been reported in Africa, but recently human pathogens such as Anaplasma phagocytophilum have been reported in Senegal and Algeria [46,47]. Bacteria from the Wolbachia genus of the Anaplasmatacae family are associated with arthropods and filarial nematodes. They are responsible for reproductive alterations in arthropods which are indirectly (via nematodes) associated with human pathogenesis [40].
Finally, Q fever is a zoonotic disease caused by Coxiella burnetii. This bacterium may cause severe infections such as chronic endocarditis and abortion [48,49]. It infects humans usually by a direct contact with domestic animals such as cattle, sheep, goats, and dogs [50]. It was previously reported in Amblyomma, Rhipicephalus, and Dermacentor ticks [43]. In Senegal, C. burnetii was detected in 0.8% to 14.2% of ticks including Am. variegatum, Rhipicephalus spp., Hyalomma spp., and Ornithodoros sonrai [51] and may play a role in Q fever epidemiology. In Côte d'Ivoire, the seroprevalence was estimated at 3.4% [52].
Although these diseases have emerged in many African countries, they remain neglected. In Côte d'Ivoire, little information is available about these diseases and their epidemiology. To date, the existence and/or prevalence of tick-borne associated pathogens remain poorly understood. Our study provides the first data screening for multiple tick-borne associated pathogens in Côte d'Ivoire.

Ethics statement
To perform this study, an approval of Cote d'Ivoire Ethics committee was received under the number N°86/MSLS/CNERN-dkn.

Period, study area and tick collection
The tick collection was conducted over a period ranging from October 30 to November 8, 2014. Ticks were manually collected from cattle in two regions of Côte d'Ivoire: Savannah and Bandama Valley (Fig 1, Table 1). In total, 378 ticks (304 adults and 74 nymphs) were collected   from three cities and 12 villages in the Savannah region and three cities and seven villages in the Bandama Valley region (Table 1). Ticks were stored in 70% ethanol until morphological and molecular analyses in laboratory of URMITE, Marseille (France). The species and sex of the ticks were identified according to standard taxonomic keys for adult ticks [2].

DNA extraction and real-time PCR
Total DNA from half of each tick was extracted using the EZ1 DNA tissue kit (Qiagen, Hilden, Germany) following the manufacturer's instructions. DNA extracts were stored at +4°C until use. Bacterial DNA was initially detected using bacterial genus-specific or species-specific quantitative real-time PCRs (qPCRs) targeting: Rickettsia spp., R. africae, R. aeschlimannii, R. massiliae, Borrelia spp., Anaplasmataceae spp., A. phagocytophilum, Bartonella spp., C. burnetii, and Spiroplasma spp. (Table 2). Samples with a high discordance in the cycle threshold number (Ct) for Rickettsia spp. and R. africae (in all cases, low Ct for Rickettsia spp. and high Ct for R. africae) were subjected to specific qPCRs for two other rickettsial species: R. aeschlimannii and R. massiliae in order to identify possible co-infection. qPCRs were performed using a CFX 96 Real Time System (Bio-Rad, Marnes-la-Coquette, France) and the Eurogentec MasterMix Probe PCR kit (Eurogentec, Liège, Belgium). PCR tests were considered to be positive when the Ct was lower than 35 Ct [22]. In addition, two different specific qPCRs targeting two different sequences had to be positive in order to confirm the presence of a bacterium in the ticks. Positive controls (bacterial DNA) and negative controls (master mix or water) were used to validate the PCR runs.

Standard PCR and sequencing
Most of samples which were considered positive by qPCRs were subsequently subjected to standard PCR. All samples which were positive using Rickettsia genus-specific but negative with R. africae qPCR were subjected to standard PCR to amplify a portion of the ompA gene. We also chose two positive ticks for R. africae by species to confirm the presence of R. africae by standard PCR. The primers used (190.70, 190.180, and 190.701) amplified a 632-bp fragment of the Rickettsia ompA gene [60]. For the identification of Borrelia species, primers targeting a portion of the flaB gene were used [33]. Anaplasmataceae spp. (Anaplasma spp., Ehrlichia spp., and Wolbachia spp.) were identified using Ana 212f and Ana 753r primers targeting a 500 bp portion of the 23S rRNA gene [47]. Standard PCR was performed on a ThermalCycler (Applied Biosystem, Paris, France). The reactions were carried out using the Hotstar Taq-polymerase (Qiagen), in accordance with the manufacturer's instructions. The amplicons were visualized using electrophoresis on a 1.5% agarose gel stained with ethidium bromide and examined using an ultraviolet transilluminator. The PCR products were purified using a PCR filter plate Millipore NucleoFast 96 PCR kit following the manufacturer's recommendations (Macherey-Nagel, Düren, Germany). The amplicons were sequenced using the BigDye Terminator Cycle Sequencing Kit (Applied Biosystems) with an ABI automated sequencer (Applied Biosystems).The sequences which were obtained were assembled using ChromasPro software (ChromasPro 1.7, Technelysium Pty Ltd.,Tewantin, Australia) and compared with those available in GenBank by NCBI BLAST (http://blast. ncbi.nlm.nih.gov/Blast.cgi).

Phylogenetic analysis
DNA sequences alignment was carried out using MEGA 6 (http://www.megasoftware.net/ mega.php). We selected the Bayesian method [61] using TOPALi 2.5 software (Biomathematics and Statistics Scotland) to construct phylogenetic trees.  (Table 3). To confirm the presence of R. africae, we performed standard PCR using two positive ticks per species. The BLAST search of the ompA gene sequences from ticks revealed 100% nucleotide identity with the ompA gene of R. africae detected in Am. variegatum collected in Antigua (GenBank EU622980). We amplified the ompA fragment in all ticks positive for Rickettsia spp. but negative for R. africae qPCR. The BLAST analyses showed that ompA sequences of R. aeschlimannii were detected in 7/16 (44%) H. marginatum rupifes and 3/27 (11%) H. truncatum. The sequences were 99% identical to those of R. aeschlimannii, previously detected in H. impeltatum collected in Egypt (GenBank HQ335157) and 100% identical to those detected in H. marginatum in Turkey (GenBank KF791251). R. massiliae was observed in 1/3 (33%) Rh. senegalensis with 100% similarity R. massiliae, previously detected in Rh. senegalensis in Guinea (GenBank JN043508). Finally, these results were confirmed by a specific qPCR for R. aeschlimannii and R. massiliae (Table 2). We also performed these species-specific qPCR on three samples (two H. marginatum rufipes and one Rh. senegalensis) where we observed a high discordance (more than 5 Cts) between Rickettsia genus-specific qPCR (low Ct) and R. africae species-specific qPCR (higher Ct). We found that in all three cases, a coinfection by two rickettsia species: R. massiliae plus R. africae in Rh. senegalensis and R. aeschlimannii plus R. africae in H. marginatum rufipes.

Results
C. burnetii was detected in one tick ( Table 3). Screening of all ticks for Borrelia spp. using qPCR, detected 16/378 (4%) positive ticks. We succeeded in amplifying a fragment of flaB gene and 16S rRNA sequence only in 4/378 (1%) ticks. A BLAST search showed that these sequences probably belong to an undescribed species, because only 87% (288/329 bp), 87% (287/328 bp), 97% (319/328 bp), and 87% (288/329 bp) similarities were observed with, respectively, the flaB gene of Borrelia duttonii (GenBank AB105132), Borrelia sp. IA-1 (GenBank EU492387), Borrelia sp. BrFlab (GenBank EF141022), and Borrelia sp. IA-1 (GenBank EU492387). The phylogenetic position of this Borrelia is shown in Fig 2. Because these potentially new species had not previously been isolated, we propose the provisional names Candidatus Borrelia africana for the genotype TCI22 and Candidatus Borrelia ivorensis for the genotypes TCI140 and TCI351. In a phylogenetic tree based on a 344 bp fragment of the Borreliae flaB gene, the sequences of Candidatus Borrelia africana and Candidatus Borrelia ivorensis are situated in the Borrelia genus near Borrelia sp. from Ethiopian Amblyomma cohaerens (GenBank JX089967) and are closer to the relapsing fever group than to that of Lyme disease. As previously shown, Ethiopian Borrelia group together with these new genotypes to form a separate and well-supported (bootraps 100) branch on the phylogenetic tree situated between Lyme disease and relapsing fever clusters, albeit closer to the latter. We also identified Borrelia sp. (genotype TCI301) in Rh. microplus which was almost identical to Borrelia sp. previously identified in the same ticks in Brazil (GenBank EF141022). Sixty-three ticks were positive using qPCR targeting the 23S rRNA of Anaplasmataceae. Only 39 DNA samples were positive using qPCR were successfully amplified in standard PCR. A possible explanation may consist of the lower sensitivity of standard PCR compared to qPCR. After sequencing, we obtained good quality sequences for only 22 samples (22/378; 6%). We suggest that the poor sequence quality may be explained by co-infection by two or more species belonging to the Anaplasmataceae family. We have identified one case of A. centrale in Am. variegatum (100% identity with the A. centrale strain Israel, NR_076686), and two cases of E. ruminantium in Am. variegatum (100% identity with the E. ruminantium strain Welgevonden, NR_077000). We have identified A. marginale in two Rh. microplus (100% of homology with A. marginale strain Florida, NR_0765879). Finally, for all remaining sequences, Blast analysis shows a homology score of under 92% which means that these sequences are likely to correspond to new species. After the construction of a phylogenetic tree (Fig 3), we propose that the status of Candidatus is applied to an uncultured species but not formally recognized by the International Code of Nomenclature of Bacteria [62]. The result shows three cases of Anaplasma: Candidatus Anaplasma ivorensis related to A. phagocytophilum identified in ticks, two in Am. variegatum, and one in Rh. microplus. The three sequences have one to two SNP (single nucleotide polymorphism) between them. In one Rh. microplus, a potential new Wolbachia sp., Candidatus Wolbachia ivorensis, was identified, closely related to the Wolbachia endosymbiont of Cimex lectularius (GenBank AP013028). We also identified two groups of sequences corresponding to new Ehrlichia spp. which cluster in two clades. Indeed, in four cases (one Am. variegatum, two Rh.microplus, and one H. truncatum), we identified Candidatus Ehrlichia rustica in the subgroup of Ehrlichia chaffeensis. In nine ticks (five Am. variegatum, three Rh. microplus and one H. truncatum), we detected Candidatus Ehrlichia urmitei that was previously observed by our team in Rh. bursa ticks collected in the Bacque area of France (M. Dahmani, personal communication) (Fig 3). Candidatus Ehrlichia urmitei forms an independent and well-supported clade situated between the E. ruminantium clade and that of Ehrlichia muris (Fig 3).

Discussion
Domestic animal resources supply some 30% of total human food and agricultural production requirements. They are particularly vital to subsistence and economic development in developing countries as they continually provide essential food products, draught power and manure for crop production and generate income as well as employment for most of the rural poor [63]. However, livestock-associated ticks are often reservoirs or vectors of human vector-borne diseases [18]. Intensification of livestock farming is one cause of the abundance of various vectors and tick-borne diseases. In recent years, the spectrum of tick-borne diseases infecting animals has increased; many of these diseases, such as rickettsioses, borrelioses, Q fever, anaplasmoses, and ehrlichioses, are gaining increasing attention from clinicians and veterinarian [4]. Advances in the development of molecular biology tools facilitate the detection of new bacteria [4,64].
Rickettsioses have been identified in humans, animals and ticks which are considered to be the main vectors of such pathogens as R. africae, R. aeschlimannii, and R. massiliae in sub-Saharan Africa [9]. In our study, rickettsial DNA was found in 49% of ticks collected from cattle. For the first time, the presence of R. aeschlimannii in ticks in Côte d'Ivoire is shown. This study provides evidence of R. aeschlimannii infection in 23% of Hyalomma ticks including H. marginatum rufipes (44%) and H. truncatum (11%). R. aeschlimannii has not been observed in other tick species. These data support the theory that the Hyalomma genus is a main vector and reservoir of R. aeschlimannii. It was previously reported in 45% to 51% of H. marginatum rufipes and 6% to 7% in H. truncatum collected from cows, donkeys, sheep, goats and horses in Senegal [12]. These data are comparable to those of our study. The high prevalence of R. africae (90%) in Am. variegatum can be explained by the high transovarial and trans-stadial transmission rates (100%) and a filial infection rate (93%) that was previously demonstrated in Am. variegatum [20]. This result shows that this tick species acts as a vector but also as a reservoir for R. africae in Côte d'Ivoire. R. africae was recently detected in other tick genera including Rhipicephalus and Hyalomma [12,18,19,57]. In our study, the prevalence of R. africae is 10% in Rh. microplus and 9% in Hyalomma spp., which is lower than in co-fed Am. variegatum, suggesting that these ticks are probably not the competent vectors for R. africae. This bacterium likely infects Rh. microplus and Hyalomma spp. during co-feeding. The first report of the presence of R. massiliae in Côte d'Ivoire was in Rhipicephalus spp. [28]; this is comparable to the detection of R. massiliae in a Rh. senegalensis tick found in our study.
C. burnetii infections have been also reported as being between 0.7% and 6.8% in ticks from cattle in western African countries [51] but not in Côte d'Ivoire where the seroprevalence of C. burnetii was estimated to be 3% [52]. Here, we show for the first time the presence of C. burnetii in Côte d'Ivoire, although only in one tick. Most Borrelia species such as B. hispanica, B. duttonii, and B. crocidurae detected in Africa, are related to soft ticks. Their main vectors are Ornithodoros spp. [65]. To date, Borrelia sp. was identified only once in an African hard tick, Am. cohaerens, in Ethiopia [39]. It has been also reported that Rhipicephalus spp. transmits Borrelia theileri to cattle, causing bovine borreliosis [18]. In Côte d'Ivoire, we show that Am. variegatum were infected by three potential new Borrelia and Rh. microplus by one potential new Borrelia. The sequences of Borrelia sp. (genotype TCI301) were identical to 99% of those of Borrelia sp. found in engorged Rhipicephalus sp. ticks collected from horse in Brazil (EF141022). Phylogenetic analysis showed that Borrelia sp. TCI301 is classified in the relapsing fever group, close to B. crocidurae and B. hispanica, two etiological agents of relapsing fever in Africa [26,31]. Blast analysis of the flaB gene showed that three new Candidatus Borrelia ivorensis and Candidatus Borrelia africana borreliae were significantly different to all other borreliae, except for this Borrelia sp in Am. cohaerens in Ethiopia [39]. These potential new Borrelia form a new clade between the clades of Lyme disease borreliae and relapsing fever borreliae. Thus, this is the first time that Borrelia species have been detected in Côte d'Ivoire and the first time their presence has been confirmed in hard ticks in Africa.
Bacteria from the Anaplasmataceae family were previously known to be pathogens of veterinary importance. However, in the three last decades, many human pathogens have been identified in this family [66]. Recently, based on the rrl gene, our team developed new tools to identify most bacteria belonging to Anaplasmataceae family [47]. These tools combine a qPCR followed by a standard PCR then sequencing, and have been used successfully to amplify DNA from bacteria belonging to Anaplasma spp., Ehrlichia spp., Neorickettsia spp., and Wolbachia spp. available in our laboratory [67]. We have successfully amplified Anaplasmataceae DNA in 6% of our ticks. For the first time, we have demonstrated the presence of A. marginale, A. centrale, E. ruminantium, and potential novel Ehrlichia, Anaplasma, and Wolbachia spp. in ticks in Côte d'Ivoire. A. marginale was observed in Am. variegatum and Rh. microplus. To the best of our knowledge, A. marginale has never been reported in Africa in Rh. microplus. The first report of its presence in Côte d'Ivoire was in 2007, but the exact route of its introduction into this region has not yet been determined [68]. A recent study indicated that the majority of the Rhipicephalus (ex-Boophilus) spp. collected and identified from farms around Azaguié (Côte d'Ivoire) are Rh. microplus (96%) [69]. A. centrale is a species closely related to A. marginale; this naturally attenuated strain has been used as a live vaccine to prevent severe diseases due to A. marginale senso stricto strains for 100 years [70]. We identified this species in Am. variegatum. To the best of our knowledge, A. centrale has never previously been detected in these ticks. The potential of Am. variegatum to transmit A. centrale needs further investigation. E. ruminantium was previously described in Am. variegatum which is invasive to cattle attaches to the hooves and cattle remain standing, particularly in the rainy season [71,72]. Recent phylogenetic analyses of Am. variegatum from Kenya, Mali, Burkina Faso, Ethiopia and the Caribbean show low genetic diversity within this population, suggesting an westward expansion of these ticks and supporting east-west genetic separation, with Caribbean genetic sequences being associated with and often identical to West African haplotypes. The data suggest that Am. variegatum reached West Africa from Zambia [73]. We have also identified three potential new species, Candidatus Anaplasma ivorensis, Candidatus Ehrlichia urmitei, and Candidatus Ehrlichia rustica. The detection of these potential new species has limitations, as not all previously described species are already molecularly characterized. Indeed, such species as Anaplasma caudatum, Anaplasma bovis, and Anaplasma mesaeterum [74] are incompletely characterized with no strain available and no or few genes sequenced, so the detection of a 'new' genotype may, in fact, be the re-discovery of an old, incompletely characterized species. Further studies are required to clarify whether these new genetic variants represent a new species. The other potential new ehrlichiae was closely related to Ehrlichia sp. amplified from Rh. bursa in France. Interestingly, this new species was amplified from two different regions in the world and from different species of ticks (Rhipicephalus, Amblyomma, and Hyalomma spp).
Finally, it is reported that ticks are often co-infected following a blood meal from a coinfected host [75,76]. Recently, mixed infections were reported for the first time in West Africa in feeding ticks and caused mainly by Rickettsia spp. and C. burnetii [18]. In Côte d'Ivoire, for the first time we show multiple co-infections in ticks. These co-infections systematically involved R. africae. To date, no human cases of anaplasmoses, ehrlichioses, borrelioses, rickettsioses or co-infections have been reported in Côte d'Ivoire. However, these diseases are still little known by clinicians and laboratory diagnostic is lacking in most cases. It is important to continue to study the epidemiological data of such emerging pathogens which may be the source of disease complications in both animals and humans. We provide evidence and demonstrate the endemicity of these different bacteria in the studied regions that have the same characteristics agro-ecological and climatic. Furthermore, these diseases could be a cause of death of unknown origin in rural areas in Côte d'Ivoire [77].