Background

Hepatozoon spp. (Adeleorina: Hepatozoidae) are apicomplexan vector-borne blood parasites with a complex life cycle [1, 2]. Vertebrates serve as intermediate hosts, and hematophagous arthropods such as ticks, mites and fleas serve as both definitive hosts and vectors [3,4,5,6]. Unlike other vector-borne pathogens, Hepatozoon transmission is not achieved by arthropod blood feeding on the vertebrate host, but with the infection taking place when the intermediate host ingests the definitive host. Also, transmission routes other than vector-borne have been described. For some Hepatozoon species, such as H. americanum (canids), H. sipedon (reptiles), H. caimani (caiman crocodiles) and H. ayorgbor (snakes), transmission can take place via predation of prey [7,8,9,10,11,12] or, in case of H. canis (canids) and Hepatozoon sp. of garter snakes (Thamnophis elegans), vertical transmission has been described [13,14,15,16].

The pathogenicity of Hepatozoon in wild animals such as canids [17, 18], felids [19] and mustelids [20] seems to be low, although in the case of co-infections with for example bacteria, severe disease manifestations may occur [21]. The importance of Hepatozoon in wild animals is mainly spillover to domestic animals. Hepatozoon infection in dogs and cats is known to cause subclinical to severe disease, which can be worsened by co-infection with, e.g., bacteria or other hemoparasites as well [2, 22,23,24,25,26,27,28,29,30,31].

Until a decade ago, histological methods prevailed in the characterization of blood parasites, with few exceptions [32,33,34,35,36]. Since around the 2010s, molecular methods have been commonly used [37], more specifically amplification of the small subunit ribosomal RNA (18S rRNA) gene fragments in case of detection and species identification of Hepatozoon infections. Several Hepatozoon species have been described in European wild and domestic mammals with these methods, with most studies focusing on carnivores [38, 39], e.g. H. silvestris and/or H. felis in European wild cats (Felis silvestris silvestris) in Bosnia and Herzegovina [19, 40], Spain [41] and Hungary [42], and in domestic cats from Italy [23, 43]. Also, H. martis has been detected in mustelids from Bosnia and Herzegovina and Croatia [44], Hungary [42] and Spain [41] and in wild cats from Spain [41, 45], H. ursi in brown bears (Ursus arctos) from Turkey [46] and Hepatozoon sp. in pine martens from Scotland, UK [20]. In wild canids such as red foxes (Vulpes vulpes) [17, 18, 41, 47, 48], gray wolves (Canis lupus) [49, 50] and golden jackals (Canis aureus) [51], prevalence of H. canis can be high. In contrast, prevalence of H. canis in domestic dogs [52, 53] and cats [39, 43] is generally (much) lower. Worldwide, only few studies related to Artiodactyla included testing for Hepatozoon, finding H. canis in camels from Saudi-Arabia [54] and in ticks collected from goats in China [55] and Romania [56] and Hepatozoon sp. in ticks collected from cattle in Pakistan [57]. To our knowledge, no information is available about Hepatozoon in wild Lagomorpha, except for a Spanish study in which no Hepatozoon was detected in European hares (Lepus europaeus) [35].

In rodents, Hepatozoon has been reported in Europe in Finland, Estonia and western Russia [58], Lithuania [59], Poland [60,61,62], Hungary [63], the Czech Republic [64], Slovakia [65], Great Britain [66,67,68,69], Germany [70, 71], Austria [72], Turkey [73] and Spain [32]. In rodent-related studies in which Hepatozoon could be identified to species level, in bank voles (Myodes = Clethrionomys glareolus) H. erhardovae was often detected [58, 60, 62, 63, 65, 68, 71, 72] and, to a lesser extent, H. sciuri in red squirrels (Sciuri vulgaris) [64], H. griseisciuri in gray squirrels (Sciuris carolinensis) [69] and H. lavieri in common voles (Microtus arvalis) [61].

Data about Hepatozoon in European ticks are also scarcely reported. It has long been believed that Rhipicephalus sanguineus sensu lato is the only known vector of H. canis in Europe and of many other Hepatozoon species [48, 51, 74, 75]. Recent findings of Hepatozoon sp. in other tick species raise questions about their vectorial role. Those findings include H. canis in Ixodes ricinus, I. canisuga, I. hexagonus and Dermacentor reticulatus ticks feeding on foxes in Germany [76], in Haemaphysalis concinna ticks collected from a dog in Poland [77] and in the abovementioned ticks (I. ricinus) from goats (and also dogs, fox and cat) in Romania [56]. Also, H. canis positive questing I. ricinus ticks were found in Slovakia and the Czech Republic [65] and in a R. turanicus collected from an infected fox in Italy [78].

Here, we aim to gain more knowledge on species distribution and prevalence of Hepatozoon among wild mammals and ticks in Europe. For this, we investigated a wide range of ungulates, carnivores and small mammals and ticks collected from animals and vegetation from five European countries. Animals and ticks were tested for the presence of Hepatozoon spp. using PCR and sequencing methods.

Methods

Mammals

For this study, animal samples originating from Belgium, the Netherlands, Austria, Bosnia and Herzegovina, and Croatia were used. From each animal a piece of spleen was collected for the survey. From gray wolves (Canis lupus) one or multiple samples were collected after a complete necropsy (see Additional file 3: Table S4 for detailed information). Animals were culled during regular or sanitary shooting in a period from 2010 to 2019, and no animal was shot for the purpose of this study only. All investigated gray wolves from the Netherlands were roadkill animals. All samples were collected within the frameworks of national game management and population control programs according to national laws. Samples from other mammals besides the gray wolves from the Netherlands and Belgium were gathered for previous studies [79] and used for Hepatozoon detection in this study.

Free-living rodent adults in Croatia were captured in Sherman live traps as described in an earlier study [80]. We followed animal experimentation guidelines approved by the American Society of Mammalogists [81]. Captured live animals were anesthetised in bags containing ether-soaked cotton. Deeply anesthetized animals were killed by cervical dislocation as described in the guidelines. Dead animals were aseptically dissected, and the tissue samples for DNA extraction were frozen at − 80 °C for several days before further analysis.

Ticks

Questing ticks were collected by dragging vegetation and other environments, and ticks were collected from various animals originating from Croatia and the Netherlands. Ticks were washed and stored in 98% ethanol until further processing, after morphological identification to species level using morphological keys as described in [82, 83].

DNA extraction, amplification and sequencing

DNA from animals was extracted from 10 mg of spleen and/or other organs (gray wolves) using DNA blood and tissue kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions.

For extraction of individual ticks’ DNA NucleoSpin® DNA Insect (Macherey Nagel) was used. The forward primer HepF 5’-ATACATGAGCAAAATCTCAAC-3’ and the reverse primer HepR 5’-CTTATTATTCCATGCTGCAG-3’ were used to amplify a fragment of ~ 670 bp of the 18S rRNA gene [84].

PCR reaction mixtures of 20 µl were prepared containing 10 µl G2 GOTaq mastermix (Promega, Madison, WI, USA), 7.2 µl DNase/RNase-Free distilled water (Promega), 0.4 µl 10 pmol/µl of each primer and 2 µl of sample. Positive (DNA of H. canis confirmed with sequencing from earlier studies) and negative (water from GoTaq G2 Mastermix) controls including extraction controls were used in all amplifications. The amplification product was analyzed using capillary electrophoresis on the QIAexcel system (QIAGEN, Hilden, Germany). For the purpose of further DNA sequencing, amplified PCR product was purified using ExoSAP-IT-PCR Clean-Up Reagent, according to the manufacturer’s instructions (USB Corporation, Cleveland, OH, USA). Sequencing in both directions was performed by Macrogen Europe with the same primers used for PCRs. The sequences were assembled using the SeqMan Pro software edited with EditSeq of the Lasergene software (DNASTAR, Madison WI, USA) and compared with available sequences using BLAST (https://blast.ncbi.nlm.nih.gov/Blast.cg) system analysis.

Phylogenetic analysis

The 18S rRNA sequences of Hepatozoon obtained in this research and the deposited sequences of other Hepatozoon species and isolates available in the GenBank® were analyzed for phylogenetic relationships. The phylogenetic tree was subjected to an unweighted pair group method with arithmetic mean (UPGMA) clustering analysis using the tree builder tool incorporated in Geneious Prime (HKY). Hepatozoon sequences generated in this study were deposited in the NCBI GenBank® database under the accession numbers MH656727-MH656732 and KT274177-KT274186.

Results

Descriptive results

For this study, 2801 mammals and 754 ticks were tested for the presence of Hepatozoon with PCR. The mammals were order Artiodactyla (n = 1233), further divided in the Families Bovidae (n = 181), Suidae (n = 289) and Cervidae (n = 763); order Carnivora (n = 865), further divided in the Families Canidae (n = 336), Ursidae (n = 79), Mustelidae (n = 446) and Procyonidae (n = 4); order Eulipotyphia (n = 1), of which only the family Erinaceidae (n = 1); order Lagomorpha (n = 171), of which only the family Leporidae (n = 171) and order Rodentia (n = 531), further divided into the families Sciuridae (n = 53), Cricetidae (n = 167) and Muridae (n = 311). Of these 2801 animals, 36 originated from Bosnia and Herzegovina, 1186 from Croatia, 208 from Austria and 1371 from Belgium/the Netherlands. From Austria and Bosnia and Herzegovina, no Eulipotyphia, Lagomorpha or Rodentia were tested, and from the Netherlands/Belgium, no Eulipotyphia and Lagomorpha. Precise numbers of each mammal species and origin are given in Table 1 (Artiodactyla), Table 2 (Carnivora), Table 3 (small mammals; Rodentia, Eulipotyphia and Lagomorpha) and Additional file 1: Table S1 (all mammals).

Table 1 Presence of Hepatozoon spp. detected in samples of Artiodactyla
Full size table
Table 2 Presence of Hepatozoon spp. detected in samples from Carnivora
Full size table
Table 3 Presence of Hepatozoon spp. detected in samples from small mammals (Rodentia, Eulipotyphia and Lagomorpha)
Full size table

In total 754 ticks of 15 tick species (Table 4) were used for this study. Of these, 287 originated from the Netherlands (all collected from animals), and the rest, 467 ticks, originated from Croatia (collected from animals n = 376, collected from the environment n = 91). Table 5 specifies the results of 38 ticks that were collected from three foxes, which tested negative for the presence of Hepatozoon DNA in their spleen samples.

Table 4 Prevalence and species of Hepatozoon in ticks collected from animals and environment (questing ticks)
Full size table
Table 5 Hepatozoon in ticks collected from negative foxes
Full size table

Prevalence of Hepatozoon in mammals

See Tables 1, 2, 3 and Additional file 1: Table S1 for Hepatozoon prevalence results of 2801 tested mammals. Overall, 370 (13.2%) mammal samples tested positive for Hepatozoon. The Carnivora showed highest prevalence (34.8%), followed by Rodentia (12.1%). Artiodactyla and Lagomorpha showed the lowest Hepatozoon prevalence (0.5% and 0.6%, respectively). Within the Artiodactyla, only chamois (Bovidae) and roe deer (Cervidae) tested positive. The only animal that was tested within the order Eulipotyphia (an European hedgehog, Erinaceus europeaus) tested positive. Within the Carnivora, the four families showed differences in prevalence: the Canidae showed the highest prevalence (49.7%), and within that family especially the golden jackals (80.8%), followed by gray wolves (over half of the tested wolves were positive) and red foxes (less than half of the tested foxes were positive). Within the Mustelidae, pine martens were most frequently infected, followed by stone martens and European polecats. In contrast, all samples of the Families Ursidae and Procyonidae of the Carnivora tested negative. Within the Rodentia, the Muridae showed much lower prevalence (3.2%) than Sciuridae (24.5%) and Cricetidae (24.6%).

Regarding origin, the mammals from Croatia (n = 1186) and the Netherlands/Belgium (n = 1371) showed higher prevalence (14.6% and 13.9%, respectively) than animals from Austria (n = 208) and Bosnia (n = 36) (2.9% and 2.8%, respectively). The chamois and roe deer (Artiodactyla) that tested positive, were all from Austria. It is interesting to point out the difference in prevalence of Hepatozoon among bank voles (Croatia 81.8% and the Netherlands/Belgium 10.5%) and wood mice (Croatia 14.6% and the Netherlands/Belgium 0.0%).

Prevalence of Hepatozoon in ticks

Overall, 31 (4.1%) of the 754 collected ticks tested positive for Hepatozoon (Table 2). Ticks of the genera Dermacentor showed the highest prevalence (16.0%), followed by Ixodes (5.3%) and Rhipicephalus (1.4%). None of the ticks of the genera Hyalomma (n = 35) and Haemaphysalis (n = 23) and none of the ticks from the Netherlands (n = 287) tested positive. Ticks that were collected from animals (29/663, 4.4%) tested positive more often than ticks collected from the environment (2/91, 2.2%).

Sequence results

Results of the 18S sequence molecular analysis (Tables 1, 2, 3, 4, Figs. 1, 2, Additional file 1: Table S1, Additional file 2: Table S2) showed that five different Hepatozoon species were detected in the tested mammals and ticks. Also, Hepatozoon that could not be further specified to species level (Hepatozoon sp.) was found. Within all tested animals, H. canis was most prevalent (6.2%; 173/2801 mammals and 3.7%; 28/754 ticks). Interestingly, H. canis isolate MH656730 was most prevalent in mammals (3.8%; 107/2801) from Austria, Bosnia, Croatia and the Netherlands/Belgium, while H. canis isolate MH656729 was most often detected in ticks (2.5%; 19/754), and isolated solely from Croatian mammals (Tables 1, 2, 5 and Fig. 1). Furthermore, H. canis MH656729 was detected within the family Canidae (golden jackals and gray wolves), Mustelidae (badger), Erinaceaidae (European hedgehog) and Leporidae (European hare), while H. canis MH656730 was detected in carnivores, Canidae (golden jackals, gray wolves and red foxes), Mustelidae (badger), but also in Artiodactyla Cervidae (roe deer). Of the 38 ticks that were collected from Hepatozoon-negative foxes, in 16 (42.1%) ticks H. canis MH656729 was detected (Table 5). In Mustelidae, H. martis (MH656728) was most prevalent (96.1%; 122/127). In Sciuridae (squirrels), only H. sciuri (MH656732) was detected. In Rodentia, Hepatozoon sp. vole isolate (MH656731) was detected in bank voles (Cricetidae) from the Netherlands and Croatia and H. ayorgbor (EF157822) in yellow necked mice and wood mice (Fig. 2). Also, in bank voles, yellow-necked mice and wood mice from Croatia and the Netherlands, Hepatozoon sp. was detected (Table 1, Table 2, Fig. 2).

Fig. 1
figure 1

Phylogenetic tree (HKY, UPGMA with Babesia canis as outgroup) of Hepatozoon sequences from Carnivora. The sequences derived in this study are in bold

Full size image
Fig. 2
figure 2

Phylogenetic tree (HKY, UPGMA with Babesia canis as outgroup) of Hepatozoon sequences from Rodentia. The sequences derived in this study are in bold

Full size image

Distribution of Hepatozoon canis in organs of gray wolves

Of the 119 (Croatia n = 114, the Netherlands n = 5) wolves of which different numbers of organs (range 1 to 9) were screened to assess the distribution of H. canis, 71 (Croatia n = 66, the Netherlands n = 5) had one or more positive organ(s) (Additional File 3; Table S4). Hepatozoon canis MH656729 was detected in organs of 50 wolves and H. canis MH656730 in organs of 19 wolves. Also, in two wolves, both H. canis isolates were detected. In most of the wolves, one (n = 18) or two (n = 34) organs were tested. Three and four organs were tested in six wolves each. Six organs were tested in four wolves, eight organs in one wolf and finally nine organs in two wolves. The organs that were most tested were spleen (n = 64), lymph node (n = 36) and skeletal muscle (n = 20). Bone marrow (n = 3) and blood (n = 1) were tested the least and brain (n = 7), myocardium (n = 10), lungs (n = 12), kidneys (n = 12) and liver (n = 13) in between.

Spleen samples tested positive most often (93.8%; 60/64), followed by samples of lungs (83.3%; 10/12), lymph nodes (75.0%; 27/36) and bone marrow (66.7%; 2/3). The seven brain samples and one blood sample tested negative. Also, myocardium (60.0%; 6/10), liver (53.9%; 7/13), kidney (41.7%; 5/12) and skeletal muscle (25.0%; 5/20) samples tested positive.

Discussion

We investigated a wide range of mammal and tick species originating from five southeastern, central and western European countries for Hepatozoon prevalence and species diversity. Hepatozoon was detected in mammals from all five countries, with differences in prevalence. Hepatozoon prevalence of mammals in Austria and Bosnia and Herzogovina was ~ 3% compared to a prevalence in Croatia and Belgium/the Netherlands of 14–15%. The main reason for this difference in prevalence seems to be sampling bias. The animal species with the highest prevalence (Canis aureus, C. lupus, Vulpes vulpes, Martes foina, M. martes and Myodes glareolus; Tables 1, 2, 3, Additional file 1: Table S1) originated mainly from Croatia and Belgium/the Netherlands. From Austria and Bosnia and Herzegovina, fewer and different animal species were sampled (mainly Artilodactyla, a few Carnivora and no Rodentia). Since Hepatozoon is known to be present in Austrian [85] and Bosnian [40, 44, 86] carnivores and in Austrian rodents [72] and is most likely also present in Bosnian rodents, including samples of Carnivora and Rodentia from those countries, it probably would have increased prevalence.

High overall Hepatozoon prevalence was found in Carnivora, especially the Canidae and Mustelidae, and in Rodentia, especially the Cricetidae and Sciuridae, which is in accordance to other studies [41, 49, 63, 70, 87].

In this study, 670-bp fragments of the 18S rRNA gene were used for phylogenetic analysis and species determination. Even though for this purpose amplifications of longer fragments [88] or next-generation sequencing of nuclear, apicoplast and mitochondrial genes [89] are currently advised, we identified five different Hepatozoon species in the mammals and ticks: H. canis [90], H. felis [91] and the more recently named H. martis [44], H. sciuri [64] and H. ayorgbor [92]. Surprisingly, H. martis was not only detected in Martes foina, M. martes and other Mustelidae, but also in Artiodactyla (R. rupicapra and C. capreolus) from Austria. Also, in Austrian roe deer (C. capreolus), H. canis was detected. Broader host specificity is known for H. martis and H. canis, although the species were detected in other Carnivora, Canidae and Mustelidae [41]. The presence of H. canis in the spleen samples of roe deer and chamois therefore represents an unexpected finding. Although prevalence was low, current detection could suggest lack of host specificity as seen in other tick-borne apicomplexans, e.g. Theileria capreoli infecting gray wolves [93].

In the 31 positive (4.1%) ticks, only Hepatozoon species that are associated with Carnivores were detected (mainly H. canis and, to a lesser extent, H. martis and H. felis). This is not surprising, since Hepatozoon transmission is known to take place by carnivores (intermediate hosts) ingesting infected ticks (definitive hosts) [2]. We found H. canis not only in R. sanguineus, but also in D. reticulatus, I. hexagonus, I. ricinus, I. canisuga and I. ventalloi, which adds to the observations that other tick species than R. sanguineus may also be definitive tick hosts in the life cycle of H. canis [76, 94]. More research is necessary to confirm vector competence and capacity of these tick species for Hepatozoon since merely detection of the parasites in a tick is insufficient to designate the tick species as a vector.

In the Rodentia that were tested in our study, two Hepatozoon species could be identified: H. sciuri (in squirrels) and H. ayorgbor (in A. flavicollis and A. sylvaticus). The Hepatozoon sp. detected in bank voles (M. glareolus) from Croatia and the Netherlands could not be identified to species level. All Hepatozoon spp. sequences of bank voles from this study were (nearly) identical to each other and to Hepatozoon sequences from voles in GenBank® (Additional file 2: Table S3). As far as we know, these Hepatozoon sp. sequences were not found in any other mammalian species. Hence, we refer to this group as Hepatozoon sp. Vole isolate.

The species that were detected in rodents were not detected in the tested ticks. This could mean that ticks are not involved in the life cycle of these Hepatozoon species or that other tick species (e.g. I. triangucileps, which feeds only on small mammals [95]) and/or tick stages are involved. Hepatozoon ayorgbor has been described in snakes, ectoparasites and rodents, even though our finding in Croatian rodents (A. flavicollis and A. sylvaticus) is the first reported in European mammals. In the life cycle of H. ayorgbor, snakes can be infected via predation of rodents, with rodents serving as paratenic or intermediate hosts and mosquitoes as definitive invertebrate hosts [9], but findings in ticks and mites [96] suggest that other arthropods may be serving as definitive hosts as well. Whether the Portuguese tick (species unknown) in which H. ayorgbor was detected and of which the sequence from GenBank® (MZ475989) was used in our sequence analysis was truly infected or merely contaminated via blood feeding on an infected host is therefore not clear. Hepatozoon ayorgbor-like sequence was detected in three spleen samples of great gerbils in northwestern China [97], sharing 98.2% similarity to H. ayorgbor in the blood and liver of a ball python (Python regius) fed with tissues of mice experimentally infected with H. ayorgbor [9, 92]. Together with our findings, this confirms that rodents play a role in the life cycle of H. ayorgbor- or H. ayorgbor-related genotypes.

Our findings of Hepatozoon in carnivores and rodents from the Netherlands and Belgium are the first reported, but our study is also the first performed regarding detection of Hepatozoon in those countries. The species that were identified in mammals from the Netherlands and Belgium were H. canis in gray wolves and red foxes, H. martis in stone martens, pine martens and European polecats, and H. sciuri in squirrels. Even though a high prevalence of Hepatozoon was found in wild carnivores from the Netherlands/Belgium, to our knowledge, in The Netherlands no autochthonous Hepatozoon spillover from wild to domestic carnivores has been reported so far. No Hepatozoon was detected in the investigated Dutch and Belgian ticks. This could be because transmission of H. canis in for example foxes takes place via ticks in fox burrows (such as I. canisuga [98]) or via vertical transmission [18], which makes spillover to dogs less likely. To detect H. canis presence in ticks from the Netherlands and Belgium, other ticks than I. hexagonus and I. ricinus collected from hedgehogs would be advisable (Table 4).

The results of the investigated gray wolves show that spleen samples are most likely to test positive in case of a positive animal, which is in accordance with other reports [13].

Conclusion

Our results show that Hepatozoon is widely present in wild mammals and ticks originating from several countries in West, Central and Southeast Europe. Presence of Hepatozoon was confirmed in ticks other than the ‘usual suspect’ R. sanguineus. Besides confirming presence of Hepatozoon in wild mammals and ticks in countries in which Hepatozoon was previously detected, presence of this tick-borne parasite in the Netherlands/Belgium was demonstrated for the first time, even though circulation in ticks could not be confirmed. Since spillover from wildlife to domestic animals in countries where Hepatozoon is endemic occurs, veterinary clinicians in the Netherlands/Belgium should be aware of the presence of this tick-borne disease.