Introduction

Hepatozoonosis is a parasitic infection caused by apicomplexan protozoans of the genus Hepatozoon (family Hepatozoidae). More than 340 species of this genus parasitize a wide range of vertebrate hosts, including amphibians, reptiles, birds, and mammals (Baneth 2011; Smith 1996). The life cycle of Hepatozoon spp. is heteroxenous. The sporogonic development and oocyst formation occur in a hematophagous definitive invertebrate host, whereas merogonic and gamontogonic development occur in intermediate vertebrate hosts. Ixodid and Argasid ticks, mites, mosquitoes, sandflies, flies, fleas, sucking lice, and triatomid bugs have been described as vectors and definitive hosts (Smith 1996). Vertebrates usually become infected through the ingestion of an infected arthropod, but transplacental transmission and predation are also described as alternative pathways of transmission (Baneth et al. 2013; Johnson et al. 2009).

Two species, Hepatozoon canis and Hepatozoon americanum, have been described infecting domestic and wild canids. Hepatozoon canis was first described in dogs from India in 1905, and Rhipicephalus sanguineus was identified as the main vector (Christophers 1907; James 1905). Hepatozoon canis is widespread in tropical, sub-tropical, and temperate climate regions of the globe (Baneth et al. 2003; Ewing et al. 2000). Hepatozoon americanum was initially found in southern United States and is transmitted by the Gulf coast tick Amblyomma maculatum (Ewing et al. 2002; Vincent-Johnson et al. 1997). Hepatozoon canis infection is often subclinical but may vary from asymptomatic to severe. Potentially fatal disease develops in dogs with high parasitemia, causing extreme lethargy, cachexia, and anemia (Baneth et al. 2003). However, H. americanum infection causes a severe disease in dogs, leading to debilitation and death. The clinical signs are characterized by muscular pain induced by myositis, severe lameness, and subsequent atrophy (Baneth 2011; Vincent-Johnson et al. 1997). In wild animals, Hepatozoon infections are usually subclinical (André et al. 2010; East et al. 2008; Kocan et al. 2000; Metzger et al. 2008). However, coyotes (Canis latrans) experimentally infected with H. americanum developed disease with lesions like those seen in dogs (Garrett et al. 2005; Kocan et al. 2000). Moreover, infection by protozoans of the genus was associated with mortality in young spotted hyena (Crocuta crocuta) (East et al. 2008).

The presence of Hepatozoon spp. in wild canids has been reported worldwide. For instance, H. canis was found in red foxes (Vulpes vulpes) from Spain, Germany, Portugal, Hungary, and Israel (Cardoso et al. 2014; Criado-Fornelio et al. 2003; Criado-Fornelio et al. 2006; Farkas et al. 2014; Margalit Levi et al. 2018; Najm et al. 2014), in golden jackals (Canis aureus) from Hungary and Israel (Farkas et al. 2014; Margalit Levi et al. 2018), and in crab-eating foxes (Cerdocyon thous), grey pampean foxes (Lycalopex gymnocercus), and maned wolves (Chrysocyon brachyurus) from Brazil (Arrais et al. 2021; Criado-Fornelio et al. 2006). On the other hand, while H. americanum was reported in coyotes from the USA (Mercer et al. 1988), an H. americanum-like protozoan, an Hepatozoon closely related to H. americanum, was detected in the South American grey fox (Lycalopex griseus) from Argentina, and C. thous and C. brachyurus from Brazil (Almeida et al. 2013; Andre et al. 2010; Arrais et al. 2021; Criado-Fornelio et al. 2006; de Sousa et al. 2017; Millan et al. 2019). Furthermore, Hepatozoon spp. were also found in hyenas from Tanzania (East et al. 2008), African wild dogs (Lycaon pictus) from South Africa (Matjila et al. 2008), and in bush dogs (Speothos venaticus), C. brachyurus, and a fox of undetermined identity from Brazil (Andre et al. 2010; Perles et al. 2019). Moreover, an Hepatozoon sp. closely related to Hepatozoon felis was reported in L. gymnocercus and L. griseus from Argentina (Giannitti et al. 2012; Millan et al. 2019).

In Uruguay, although Hepatozoon spp. was previously confirmed in snakes, Philodryas patagoniensis, and in domestic cats (Bazzano et al. 2020, 2021), the occurrence of this protozoan in native mammals is still obscure. The aim of this work was to assess the presence of Hepatozoon spp. in wild canids from Uruguay.

Materials and methods

Sample collection

Between May 2015 and July 2020, we collected blood samples and ticks from fresh road-killed foxes encountered along Uruguayan highways. Blood samples were collected in EDTA-tubes, identified, and kept with cool packs until arrival to the laboratory. Ticks retrieved from foxes were stored in tubes with 95% ethanol and identified using keys provided by Nava et al. (2017). At the laboratory, samples were stored at − 20 °C until DNA extraction.

DNA extraction and PCR amplification

For molecular analysis, DNA was extracted from 200 µl of whole blood using the commercial kit PureLink™ Genomic DNA Mini Kit (Invitrogen, Germany), following the manufacturer’s instructions. DNA of Hepatozoon was amplified using two conventional PCR protocols. First, we used primers HEMO1 (5′-TAT TGG TTT TAA GAA CTA ATT TTA TGA TTG-3′) and HEMO2 (5′-CTT CTC CTT CCT TTA AGT GAT AAG GTT CAC-3′) that amplify a region of approximately 900 base pairs (bp) of the 18S rRNA gene (Perkins and Keller 2001). PCR thermal conditions followed Harris et al. (2011). A second PCR targeting approximately 670 bp of the same gene was performed using primers HEP1mod (5′-CGC GAA ATT ACC CAA TTC TA-3′) and HEP4 (5′-TAA GGT GCT GAA GGA GTC GTT TAT-3′) as described by Spolidorio et al. (2009). Both pairs of primers were selected because the retrieved sequences overlap into a fragment of near 1300 bp. DNA of Hepatozoon sp. obtained from P. patagoniensis and DNAse-free water were included in each reaction as positive and negative controls, respectively. PCR products were visualized under UV transillumination in 1.5% agarose gels stained with GoodView™ Nucleic Acid Stain (Beijing SBS Genetech Co., LTd). Positive PCR products were purified using the PureLink™ Quick PCR Purification kit (Invitrogen, Germany) and sent to a commercial sequencing company (Macrogen Inc., Seoul, Korea).

Analyses of sequences and phylogenies

Each sequence was carefully checked, and manual corrections were done, when necessary, with Geneious (Kearse et al. 2012). The two overlapping fragments of each sample were assembled into consensus sequences. Nucleotide identities of obtained sequences were calculated using the Sequence Identity and Similarity (SIAS) calculator (http://imed.med.ucm.es/Tools/sias.html). Alignments with herein obtained sequences and 27 homologues retrieved from GenBank were performed with MUSCLE algorithm (Edgar 2004) implemented in MEGA 7 (Kumar et al. 2016).

Two phylogenetic trees were constructed. A maximum-likelihood (ML) inference was employed to get one tree using PhyML (Guindon and Gascuel 2003). Best fitting evolutionary model for 18S rRNA gene was calculated with MEGA 7 and the Tamura 3 parameters with gamma distribution selected. The support of the internal branching was assessed using 1000 bootstrap replicates. A second phylogeny was inferred using Bayesian statistics (Ronquist et al. 2012), using the general time reversible model. Four independent Markov chains run for 1,000,000 metropolis-couples MCMC generations, begun with random seeds, and ran four times sampling a tree every 100th generations. The first 25% of the trees were considered as burn-in. The remaining trees were employed to calculate the Bayesian posterior probability. Adelina grylli (DQ096836) rooted each tree.

Results

Overall, we collected blood samples from 45 C. thous and 32 L. gymnocercus in roads from 14 departments of Uruguay. Although eight Amblyomma tigrinum (four females and four males) and 54 Amblyomma aureolatum (eight females and 46 males) were collected from nine foxes (seven C. thous and two L. gymnocercus) (Supplementary Table 1), no PCR analyses were performed with the specimens.

Out of 77 tested blood samples, eight (10.4%) were positive for Hepatozoon 18S rRNA gene PCR. Sequences were obtained from six (6/45, 13.3%) C. thous (four females and two males) and two (2/32, 6.25%) L. gymnocercus (2 males). Four of these samples, F2, F5, F7, and F72, were amplified with one primer pair (HEP1mod, HEP4) obtaining fragments of 618, 630, 642, and 659 bp, respectively. The remaining positive samples, F18, F33, F52, and F59, were amplified by both sets of primers, whose amplified fragments overlapped, obtaining larger sequences (1322, 1322, 1318, and 1322 bp, respectively). The phylogenetic analyses using both ML and Bayesian methods showed the same topology for main branches (Fig. 1a, b). With high support values, Hepatozoon sequences obtained from C. thous and L. gymnocercus from Uruguay clustered with H. americanum from USA (AF176836), and with H. americanum-like species from dog and foxes from Brazil (KU729739, AY461377, KC127679) and Argentina (MK049949) (Fig. 1a, b). As 18S rRNA gene is conserved, we opted to retain the nomenclature H. americanum-like for the sequences retrieved in this study (GenBank accession numbers: MZ230031-038). Due to length differences between the sequences we obtained, pairwise comparisons were done considering a 571 bp overlapping fragment and showed 97.32–99.82% identity between them. Moreover, 96.63–100% of nucleotide identity was observed in comparisons with H. americanum and H. americanum-like included in the phylogenetic trees (Supplementary Table 2).

Fig. 1
figure 1figure 1

Phylogenetic trees for a subset of 18S rDNA sequences of Hepatozoon spp. a Bayesian phylogeny. Bayesian posterior probabilities are indicated upon or arrowing each branch. b Maximum-likelihood tree. Numbers represent bootstrap supports. Sequences generated in this study are annotated in red bold letters. The clade of Hepatozoon americanum is highlighted within yellow box. GenBank accession numbers are indicated in brackets. Scale bars indicate the number of substitutions per nucleotide positions, respectively

Full size image

Discussion

The eight Hepatozoon spp. 18S rRNA gene partial sequences amplified in samples from C. thous and L. gymnocercus are closely related with H. americanum (Mathew et al. 2000), and H. americanum-like previously reported in Argentina and Brazil (Almeida et al. 2013; Criado-Fornelio et al. 2006; Gomes et al. 2016; Millan et al. 2019). Maximum-likelihood and Bayesian phylogenetic analyses yielded a highly similar tree topology, and a high support value suggests that sequences related to H. americanum conform a monophyletic group. Within this clade, the sample F2 grouped with Hepatozoon sp. F3 (KC127679) and Hepatozoon sp. Curupira 2 (AY461377) detected in C. thous from Brazil. Further phylogenetic analysis using less conserved genes will be necessary to clarify if H. americanum is one species or a species complex distributed along the American continent.

The prevalence of Hepatozoon sp. infection in wild canids from South America varies between studies. The 10.4% prevalence of Hepatozoon spp. infection found in foxes from Uruguay is much more lower than the 77.6% reported by Criado-Fornelio et al. (2006) in foxes from the State of Rio Grande do Sul (Southern Brazil). However, we found C. thous and L. gymnocercus infected with H. americanum-like protozoan, whereas Criado-Fornelio et al. (2006) detected an H. americanum-like species only in C. thous (7.6% prevalence), and both C. thous and L. gymnocercus infected with H. canis (69.2% prevalence). Our study represents the first report of L. gymnocercus infected with an H. americanum-like protozoan. Another study with wild canids from Brazil reported C. thous infected with H. americanum-like protozoan (2.5% prevalence) and C. brachyurus with Hepatozoon sp. closely related with H. canis genotype “Spain 1” (5% prevalence) (Andre et al. 2010). Almeida et al. (2013) reported 28 C. thous infected with H. americanum-like protozoan (48.3% prevalence) and 1.7% infected by an organism closely related to reptile associated Hepatozoon. Furthermore, Hepatozoon sp. closely related with Hepatozoon felis was found in a L. gymnocercus from the north-central Patagonia region of Argentina (Giannitti et al. 2012).

Although we found ticks only in 10 foxes, it is noteworthy that we worked with dead animals encountered in different levels of decay, so it is also possible that more animals were infested at the moment of death. The tick species collected from foxes in this study (i.e., A. aureolatum and A. tigrinum) have been reported parasitizing domestic dogs in Uruguay (Lado et al. 2014; Martins et al. 2014; Venzal et al. 2003). The presence of A. aureolatum in foxes was previously reported by Criado-Fornelio et al. (2006), Nava et al. (2017), and Labruna et al. (2005). While herein collected ticks were not analyzed, previous work in South America have suggested A. tigrinum as putative vector of Hepatozoon (Arrais et al. 2021; Giannitti et al. 2012; Millan et al. 2019).

The expansion of human activities has forced the coexistence of domestic and wild canids into the same habitat. Although H. americanum infection is of minor veterinary concern for wild canids, habitat overlap between dogs and foxes sharing the same tick species in Uruguay makes the presence of H. americanum-like protozoan in foxes a probable threat to domestic canids. This work represents the first report of wild canids infected with Hepatozoon in Uruguay and reaffirms the importance of further studies to elucidate the vectors of Hepatozoon spp. in South America.