High serological and molecular prevalence of Ehrlichia canis and other vector-borne pathogens in dogs from Boa Vista Island, Cape Verde

Despite the high global impacts of canine vector-borne diseases (CVBD) due to their wide distribution and zoonotic potential, the current epidemiological situation of CVBD in many tropical and subtropical regions remains unknown. This study examines the seroprevalence and molecular prevalence of Ehrlichia canis and other pathogens causing CVBDs (Leishmania infantum, Dirofilaria immitis, Babesia spp., Anaplasma spp. and Hepatozoon canis) in dogs living on the island of Boa Vista (Cape Verde Republic). Blood samples and infesting ticks were taken from 150 dogs across the island (stray, shelter, and pet dogs). Serum samples were tested using a rapid immunochromatographic test (Uranotest® Quattro) that detects antibodies against E. canis, L. infantum, Anaplasma spp. and D. immitis antigen. Levels of serum antibodies against E. canis were measured using the immunofluorescence antibody test (IFAT). In addition, tick-borne pathogens in blood samples (Anaplasma spp., Babesia spp., Hepatozoon spp., and Ehrlichia canis) were detected by microscopy observation and/or PCR plus sequencing. The seroprevalence of E. canis was extremely high at 82% (123/150), as revealed by both immunochromatography and IFAT. Most dogs returning a seropositive test result (82.92%; 102/123) had antibody titres > 1:1280 but showed no clinical signs or notable laboratory abnormalities. Of the 123 animals testing seropositive for E. canis, 67 (54.47%) also presented antibodies against Anaplasma spp., and 13 (10.56%) showed the presence of Hepatozoon spp. gamonts in the blood smear. Ehrlichia canis infection was detected in 17.1% (25/146) of dogs tested by direct sequencing of polymerase chain reaction (PCR) products. Co-infections were detected in seven of these dogs: four dogs tested PCR-positive for both E. canis and A. platys, two dogs tested positive for E. canis and Hepatozoon spp., and one dog tested positive for E. canis, A. platys and Hepatozoon spp. Rhipicephalus sanguineus sensu lato was the only tick species found infesting the canine study population. The high prevalence of tick-borne pathogens detected in dogs from Boa Vista Island highlights a need for improved control measures designed to prevent the transmission of these pathogens. Graphical Abstract

The etiological agent of canine monocytic ehrlichiosis (CME) is Ehrlichia canis, a Gram-negative intracellular bacterium classified within the group of alpha-proteobacteria, order Rickettsiales, family Anaplasmataceae, genus Ehrlichia [2].The main vector responsible for transmitting this emerging rickettsial disease is Rhipicephalus sanguineus, commonly known as "brown dog tick".Canine monocytic ehrlichiosis is a multi-systemic disease with three clinical stages: acute (occurring 8-20 days after tick transmission with non-specific clinical signs), subclinical (dogs appear healthy but can exhibit blood abnormalities) and chronic (severe haemorrhagic and ocular disorders) [3].The diagnosis of CME should be based on clinical records, laboratory findings and serological and molecular tests [4].In endemic areas, monitoring seropositive but healthy dogs is essential, as reinfection can occur due to the lack of persistent immunity.
Since the discovery in 1935 of Ehrlichia canis in dogs from Algeria, other Ehrlichia species have been described to infect dogs, including Ehrlichia ewingii, causing granulocytic ehrlichiosis, and Ehrlichia chaffeensis, causing human monocytic ehrlichiosis.Both can infect dogs and have been detected in several tick species and other vertebrate hosts in many countries.Ehrlichia ruminantium has been detected in healthy dogs and those exhibiting clinical signs using molecular diagnostic tests.Ehrlichia muris was identified in a sick dog in northern Minnesota that had tested seronegative for E. canis [5].
Canine anaplasmosis is mainly caused by two species, Anaplasma phagocytophilum, which mainly infects neutrophils causing canine granulocytic anaplasmosis, and A. platys,, which affects platelets causing infectious cyclic thrombocytopenia [4].Anaplasma phagocytophilum is transmitted by ticks of the genus Ixodes and is, therefore, more prevalent in regions with a temperate climate.This species of Anaplasma can infect a wide variety of animals, including dogs, cats, sheep, goats, cows, deer and humans [6].The arthropod vector that transmits A. platys has not been identified, although the brown dog tick is a major candidate [4].Co-infection with E. canis and A. platys is frequent in dogs where R. sanguineus s.l. is present.Also common is co-infection with other pathogens transmitted by this tick species, such as Babesia vogeli or Hepatozoon canis [7,8].
Canine babesiosis is a severe tick-borne haemoprotozoan disease caused by various Babesia species worldwide.In Africa and Europe, three large Babesia species (Babesia canis, Babesia rossi and Babesia vogeli) and two small species (Babesia gibsoni and Babesia vulpes) exist and these have been described as clinically relevant [9].The distribution of Babesia spp. is closely related to their vector's range; B. canis and B. vulpes are mainly found in Europe, while B. vogeli and B. gibsoni are distributed globally, including tropical and subtropical areas.Babesia rossi is restricted to Africa and causes severe clinical disease [10].
Canine hepatozoonosis is caused by two species of Hepatozoon.Hepatozoon canis is a tick-borne protozoan parasite that infects dogs globally.Transmitted mostly by R. sanguineus s.l., the parasite resides within the host's neutrophils and affects haemolymphoid organs, causing systemic illness.Another more pathogenic species, Hepatozoon americanum, has described in dogs from the USA [11].
Other significant zoonotic CVBDs that are not transmitted by ticks are leishmaniosis and heartworms.Canine leishmaniosis is detected in the Mediterranean Basin, North Africa, South America, and West Asia.Numerous species of Leishmania infect dogs in tropical regions.Leishmania infantum is the most widespread and is transmitted by phlebotomine sandflies [12].Infected dogs may exhibit a wide range of non-specific clinical signs, such as muscular atrophy, lymph node enlargement, skin lesions, and renal and ocular disorders, among others [13].Thus, outside endemic areas, the diagnosis of this disease can be challenging for veterinarians.Heartworm is caused by Dirofilaria immitis, a cosmopolitan filarial worm transmitted by mosquitoes (Culicidae) to various domestic and wild carnivores [14].It is also sporadically detected in humans in regions where this parasite is endemic in dogs [15].The disease typically progresses chronically in dogs due to the pathogenic effect of adult worms residing in the pulmonary arteries.
The varied clinical picture presented by CVBD makes their diagnosis extremely complex.Canine monocytic ehrlichiosis and anaplasmosis are globally distributed tick-borne diseases of major concern as they can be life-threatening for dogs if not adequately treated.Canine monocytic ehrlichiosis has been reported on all continents, mostly in tropical and subtropical regions where environmental conditions promote the survival of its vector, R. sanguineus.However, the current epidemiological situation of CVBD in many tropical and subtropical regions still needs to be defined.The emergence of CVBD is influenced by globalization, urbanization, and climate change, particularly increased temperatures and altered rainfall patterns.Because of their zoonotic potential, canine vector-borne, particularly tick-borne, diseases are now recognized as emerging global health concerns for humans and animals [16].
The present study was designed to determine the seroprevalence and molecular prevalence of infection by Ehrlichia canis and other canine vector-borne pathogens (Leishmania infantum, Dirofilaria immitis, Babesia spp., Anaplasma spp., and Hepatozoon canis) in dogs living on the island of Boa Vista.

Geographical study area
Boa Vista is an island occupying 620 km 2 that forms part of the Cape Verde archipelago in the Atlantic, off the north-west coast of Africa.It is one of Cape Verde's easternmost islands (16° 06′ 12″ N; 22° 48′ 13″ O).Its main town is Sal Rei.
Cape Verde has a tropical dry climate influenced by the north-easterly trade winds.Its rainy season typically occurs between July and October, although mean annual precipitation is relatively low.Vegetation tends to be arid, with plants adapting to the scarce water conditions.
Climate and vegetation vary among the different islands owing to differences in altitude and topography.

Study population
This was a cross-sectional study in which 150 dogs from Boa Vista Island were enrolled.Participating dogs were divided into four groups on the basis of their lifestyle: companion dogs (n = 23), stray dogs that were homeless yet cared for and fed by a responsible person (n = 74), stray dogs with no person responsible for them (n = 11), and dogs living at a shelter run by the NERINA Animal Protection Association (n = 31) (Fig. 1).

Samples and data collection
Blood and ticks were collected from dogs in April 2022 during a sterilization programme carried out by NERINA to mitigate overpopulation.
All participating dogs were subjected to a thorough physical exam in which the data-date, location, age, breed, sex, weight, and lifestyle-were collected along with clinical findings.Three millilitres of blood were drawn by cephalic or jugular venipuncture and immediately transferred to two EDTA tubes (0.5 ml each) and a tube without anticoagulant (2 ml).Serum was then separated by centrifugation.Serum and EDTA-mixed blood samples were initially kept at 4 °C for complete blood counts, blood smears and a rapid immunochromatographic test.Subsequently, all samples were stored at −20 °C until serological and molecular processing in the laboratory.

Microscopy observation of stained blood smears
Thin blood smears (n = 139) were Diff-Quick stained and examined under a light microscope to detect Ehrlichia spp., Anaplasma spp., Hepatozoon spp.and piroplasm species or microfilariae.

Immunological diagnosis
Serum samples (n = 150) were subjected to a rapid immunochromatographic test (ICT) (Uranotest ® Quattro, Spain) to detect antibodies against E. canis, L. infantum, Anaplasma spp., and D. immitis antigen.In addition, an immunofluorescence antibody test (IFAT) was used to determine anti-Ehrlichia canis serum IgG antibodies using a commercially available antigen kit (Anti-Ehrlichia Antibodies, Biosystems, Barcelona, Spain).The IFAT test was performed according to the manufacturer's instructions.A cut-off of 1:80 was taken to indicate seropositivity.Positive sera were further tested in a serial dilution starting at 1:20.

Molecular diagnosis
DNA from whole blood was isolated using the QIAamp ® DNA mini kit (Qiagen, Germany) following the manufacturer's instructions.In the last step, DNA was eluted in 200 µl and stored at −20 °C until further use.
Blood DNA samples (n = 146) were analysed using different polymerase chain reaction (PCR) methods to detect tick-borne pathogens including piroplasm species, Ehrlichia canis and Anaplasma spp.
The 146 DNA samples were tested using an Ehrlichia genus-specific PCR assay targeting approximately 838 bp of 16S ribosomal DNA (rDNA).This PCR method was performed as described by Peter et al. [17].Briefly, a final volume of 20 µl reaction mixture was prepared containing 5 µl of genomic DNA, 12.4 µl Master mix (Biotools, Spain) and a 10 µM final concentration of each primer (EHRF and EHRR).The amplification conditions were an initial denaturation step of 95 °C for 5 min, followed by 35 cycles of denaturation at 95 °C for 45 s, annealing at 62 °C for 45 s and extension at 72 °C for 45 s, and a final extension at 72 °C for 7 min.
In addition, a conventional PCR was performed to detect Anaplasma spp.only in blood samples from sick dogs with a PCR-positive result for E. canis (n = 25) or a positive serology result for Anaplasma spp.(n = 5).A total of 30 DNA samples were tested using an Anaplasma genus-specific PCR method following the protocol described by Peter et al. [17], using the primers ANAF and ANAR, which targeted an 424 bp of Anaplasma 16S rDNA.The PCR conditions were the same as those described for the Ehrlichia genus-specific PCR assay.The amplification conditions were an initial denaturation step of 95 °C for 5 min, followed by 40 cycles of denaturation at 95 °C for 45 s, annealing at 57 °C for 45 s and extension at 72 °C for 45 s, and a final extension at 72 °C for 7 min.
To estimate the incidence of Babesia spp. on the island, a piroplasm genus-specific nested PCR was run on 12.6% of randomly selected dog DNA samples (n = 19) using primer sets BTF1 and BTF2 F/R, which target the 18S ribosomal RNA (rRNA) gene, as described elsewhere [18].

Sequencing
All PCR products corresponding to the expected length were sequenced at the Genome Sequencing Service (Universidad Complutense de Madrid, Spain) using an ABI Prism 3730 instrument (Applied Biosystem, Foster City, USA).The DNA sequence chromatogram files obtained were aligned with Chromas software, edited sequences were compared with sequence databases, and significance was determined using the BLAST ® programme (https:// blast.ncbi.nlm.nih.gov/ Blast.cgi).
In addition, they were sexed and the tick stage (larvae, nymph or adult) was determined.

Statistical analysis
Statistical tests were performed using the IBM SPSS Statistics package (version 28.0).A descriptive analysis was performed using absolute and relative frequencies for qualitative variables, and the mean, standard deviation and percentiles for quantitative variables.The chisquared test was used to examine relationships between E. canis and Anaplasma spp.seropositivity and the remaining categorical variables.The non-parametric Kruskal-Wallis test was used to compare blood counts between infected dogs with E. canis and/or A. platys (PCR positive), dogs seropositive and PCR negative for E. canis and/or A. platys, and dogs seronegative and PCR negative for E. canis and A. platys.Significance was set up at P < 0.05.
The epidemiological variables recorded in dogs testing seropositive for E. canis and Anaplasma spp.are shown in Tables 1 and 2, respectively.No significant differences were found in the factors study area, age, size or reproductive status and seropositivity for E. canis or Anaplasma spp.However, the factor breed did differ significantly among the groups established, and a relationship was found between exposure to E. canis infection and mixed breed (P = 0.002).According to lifestyle, stray and shelter dogs showed the highest seroprevalences, 85.3% (99/116) for E. canis and 52.6% for Anaplasma spp.(61/116), varying significantly with respect to companion animals which had a seropositivity of 65.2% (15/23) for E. canis (P = 0.02) and 21.7% (5/23) for A. platys (P = 0.007).

Molecular diagnosis
PCR was able to detect Ehrlichia spp.DNA in 17.1% of the blood samples tested (25/146 dogs).Subsequent sequencing of all PCR-positive samples served to identify only the species E. canis.
The 25 dogs testing PCR-positive for E. canis showed high anti-E.canis antibodies as determined by IFAT (between 1/1280 and 1/5120).Additionally, five of these dogs (5/25) scored PCR positive for A. platys (Table 3).Further, among the sick dogs with positive serology results for Anaplasma spp.but negative results for E. canis by PCR (n = 5), two tested positive for A. platys.Babesia spp.were undetected by PCR in all dogs tested.

Co-infections
Seven dogs were simultaneously infected by more than one pathogen detected by PCR and/or blood smear microscopy examination: four dogs tested positive for both E. canis and A. platys, two dogs tested positive for both E. canis and Hepatozoon spp., and one dog tested positive for E. canis, A. platys and Hepatozoon spp.(Table 3).
All dogs showing laboratory abnormalities were found to be seropositive or infected by at least one pathogen.A total of 11 dogs were infected by E. canis, 2 were coinfected by E. canis and A. platys, and 24 dogs were seropositive (14 for E. canis only and 10 for E. canis and Anaplasma spp.).
Of 123 dogs found seropositive for E. canis, 25 returned a positive result by PCR while 98 were negative by PCR.  4. Mean haemoglobin concentrations, red blood cell and platelet counts, haematocrit, and MCV and MCHC values were lower in the infected group than in the seropositive and negative groups (P < 0.01).However, no differences were found in blood counts between the groups of seropositive and negative dogs (Table 4).

Ectoparasite identification
A total of 26 animals were found to have, at least, an ectoparasite infestation during physical examination and parasites were collected for subsequent identification in the laboratory.In total, 349 ectoparasites were identified morphologically, of which 73.07% (255/349) were ticks and 26.93% (94/349) were fleas.The number of ectoparasites collected from the same individual ranged from 1 to 75 for ticks, and 1 to 73 for fleas.
The taxonomic classification of Rhipicephalus, was based on the hexagonal shape of the basis capituli and the anal groove positioned posterior to the anus.Once classified to the genus level, species identification relied on examining specific parts of the ticks, such as the spiracles.Rhipicephalus sanguineus is characterized by spiracle plates with narrow tails, which are less wide than the adjacent festoon.In females of R. sanguineus, the separation of porose areas is broad and palp pedicels are short.In males, the shape of the adanal plates is narrow and trapezoidal, the posterior grooves of the scutum are distinct and the caudal appendage is broad in fed males [21,22].
For the taxonomic classification of fleas, we first checked for the presence of genal and pronotal combs.If both combs were present, the head and genal comb morphology were examined.A head that was twice as long as it was wide, and a genal comb with spines I and II of the same length corresponded to Ctenocephalides felis; whilst a head of equal length and width, with spine I of the genal comb shorter than spine II, corresponded to Ctenocephalides canis.In the absence of combs, the thorax and head were carefully examined.Echidnophaga gallinacea has a contracted thorax, and the anterior margin of its head was angled [20].

Discussion
Our results indicate the presence of three tick-borne pathogens (E.canis, A. platys and Hepatozoon sp.) in dogs from Boa Vista, each with a different prevalence.Ehrlichia canis was detected with a frequency of 82% by both ICT and IFAT, and with one of 17.1% by PCR.This study is the first to examine the seroprevalence of this pathogen and highlights the existence of a hyperendemic area of canine ehrlichiosis.The second most prevalent pathogen was A. platys, with an overall seroprevalence of 47.2% (44.6% together with E. canis and 2.6% alone).This pathogen was PCR detected simultaneously with E. canis in five dogs.Hepatozoon sp. was also detected in dogs from the island at a prevalence of 11.5% (16/139) and found co-infecting three dogs with E. canis.Co-infections are common as different pathogens are transmitted by the same arthropod vector through the tick's bite or via their ingestion [23].This makes it difficult to interpret   the role of these pathogens in the clinical and pathological manifestations of the animals examined.
Various studies have confirmed the presence of H. canis, A. platys and E. canis on the different islands of Cape Verde.These studies reveal significant differences in CVBD prevalences depending on the study area.On Maio Island, PCR-confirmed prevalences of 3.3% for E. canis, 35.9% for H. canis and 34.6% for A. platys were reported [8], while in Praia, prevalences found for these pathogens were 26.2%, 63.8% and 7.7%, respectively [24].
The signs of CME are non-specific and difficult to assess.Moreover, because of coinfections, some conditions may be worsened or masked.The clinical status of the animals could not be related to the seroprevalence of E. canis, as this infection only shows measurable clinical signs during early or severe disease stages.This, along with the mostly high antibody titres obtained > 1:1280 (82.92%; 102/123), indicates that most of the dog population examined had had contact with the agent and was in a sub-acute disease phase.At this stage, dogs show no clinical signs, but do have high serological titres which are maintained over time due to the prolonged duration of E. canis antibodies or possible reinfections owing to a lack of persistent and effective immunity towards the pathogen [25].While in some animals, haematological findings of thrombocytopenia with or without anaemia were detected together with clinical signs compatible with CME (mucosal pallor), the remaining participants showed no findings suggestive of active disease.
The different groups established for our epidemiological analysis of variables that could affect the risk of E. canis and Anaplasma spp.infection failed to vary in terms of these factors.The lack of a strong link found here between seroprevalence and the origin of the dogs suggests that the prevalence of infection is homogeneous throughout the island, probably reflecting the extension of the R. sanguineus vector [26].Only the factors dog breed and lifestyle were found significantly associated with seroprevalence (P < 0.05), with higher rates observed in the mongrel group and stray or shelter dogs.This may be because most dogs living and bred on the island are not pure breeds and being so prevalent, CME will inevitably mostly affect the largest population group.In addition, we must consider that most pure-bred dogs had their owners, similar to companion animals, who could provide individualised care such as tick prophylaxis, which the street animals did not have.
In our microscopy analysis of blood smears, E. canis morulae were detected in 7 out of 139 samples and PCR revealed a higher active infection rate (25/146).This proportion of positive blood smears could be higher in cases of acute disease when there are more circulating organisms in infected cells.Microscopy also revealed the presence of co-infections with Hepatozoon sp., which was not unexpected considering that this protozoan is transmitted by ingesting the tick R. sanguineus, which is regarded as its definitive invertebrate host.Furthermore, other non-vectorial routes, such as the vertical transmission, can be implicated in the spread of H. canis in dogs [27].The presence of H. canis was previously described in Cape Verde with its smear identification in 4 of the 57  animals examined [28].In our study, A. platys was also identified by PCR in seven dogs, which is transmitted by the same vector and mentioned in the studies of Lauzi et al. [8] and Götsch et al. [24].Finally, through morphological identification of ectoparasites collected from dogs using morphological keys, we found that the only tick species on dogs in Boa Vista was R. sanguineus s.l.Indeed, this is the only hard tick species found in Cape Verde in recent studies [8,24], although other species such as Amblyomma variegatum, Margaropus decoloratus (Syn.R. decoloratus) and Hyalomma sp. had been previously reported [26].The higher percentage of males (48.63%) than females (43.14%) of R. sanguineus s.l. in hosts may be attributed to tick males remaining on the host after mating while females detach for oviposition [16].The habits of street dogs may explain the presence of fleas on these animals.The finding of C. felis and E. gallinacea is understandable, considering that some animals cohabited with species such as cats, chickens and horses.Although their role as vectors is not very important in domestic animals, they may be involved in Dipylidium caninum, Acanthocheilonema reconditum, Bartonella spp.and Rickettsia spp.Therefore, while they mainly cause flea allergy dermatitis in dogs via their bite (haematophagous), for the prevention and control of CVBD we should still consider their vector role.
The data emerging from this study will help design measures to prevent the risk of transmission of these pathogens between dogs and humans who live on or travel to this island.Some limitations of our study include a need for knowledge of the animals' history to make a more accurate diagnosis of how the different parasites affect Cape Verdean individuals.Also, difficulties in handling and transporting samples made it difficult to perform a comprehensive analysis of all blood samples.Knowing the prevalence of tick-borne diseases is important not only to improve current knowledge, but also to help prevent their spread and thereby maintain the health of animal and human populations under a One Health approach [29].

Conclusions
The seroprevalence of Ehrlichia canis in dogs on the island of Boa Vista (Cape Verde) was found to be extremely high.Our results indicate that most dogs had subclinical or chronic infections with no clinical signs or laboratory abnormalities.However, with high antibody titres, which could mean frequent reinfections, these dogs could act as important carriers of the infection.Moreover, as dogs showing clinicopathological abnormalities were infected mostly by E. canis,, this pathogen should be considered the most relevant in this geographical area.Further, Rhipicephalus sanguineus s.l. was here consolidated as the main vector responsible for the transmission of Ehrlichia canis and potentially of Anaplasma platys on the island, thus allowing for co-infections with other tick-borne pathogens such as Hepatozoon spp.

Fig. 1
Fig. 1 Sampling sites (red dots) and distribution areas of dogs testing seropositive for Ehrlichia canis and Anaplasma platys infection in Boa Vista.The size of the red dot indicates the number of dog samples analysed per sampling site.Pie charts show the proportions of seropositive dogs by sampling site

Table 1
Epidemiological variables recorded in dogs testing seropositive (n = 123) or seronegative (n = 27) for E. canis Clinicopathological findings in 16 E. canis-and/or A. platys-infected dogs, 74 dogs seropositive for E. canis and/or A. platys (PCR-negative), and 12 dogs seronegative and PCR negative are provided in Table

Table 3
Results obtained for PCR tests (Ehrlichia spp.and Anaplasma spp.) and microscopy observations of blood smears according to the serological status of the dogs

Table 4
Clinicopathological findings in E. canis-and/or A. platys-infected dogs (PCR positive) compared with E. canis and/or A. platys seropositive (ICT or IFAT positive and PCR negative) and seronegative (ICT or IFAT and PCR negative) dogs Negative: ICT or IFAT (−) and PCR (−) b Seropositive: ICT or IFAT (+) and PCR (−) c