Introduction

Canine hepatozoonosis is a tick-borne disease caused by the protozoan Hepatozoon. There are two described species: Hepatozoon canis, transmitted by the dog tick Rhipicephalus sanguineus and Hepatozoon americanum transmitted by Amblyomma maculatum (Vicent-Johnson et al. 1997; Baneth et al. 2003). In most cases, the clinical presentation of H. canis infection is a mild disease but a severe illness, characterized by extreme lethargia, cachexia and anemia, that may occur in dogs with high parasitaemia. In contrast, H. americanum infection is a debilitating and often fatal disease, characterized by generalized pain, muscle atrophy, weakness and bone proliferative lesions (Baneth et al. 2003). Mathew et al. (2000) demonstrated that H. canis and H. americanum are closely related. Nevertheless, in addition to the biological and clinical differences between the two species, Baneth et al. (2000) achieved the distinction between them using genetic and antigenic analysis. Inokuma et al. (2002) determined, using the PCR, that the species from Japan might be a strain variant of H. canis.

In Brazil, the canine Hepatozoon infection has been reported and has being diagnosed during laboratory examinations (Mundim et al. 1992, 1994; Gondim et al. 1998) or during epidemiological studies in urban or rural areas (Massard 1979; O’Dwyer et al. 2001, 2004; Paludo et al. 2003). O’Dwyer et al. (2001) observed high prevalence (39.2%) in dogs from rural areas from Rio de Janeiro State and low prevalence (5.9%) in stray dogs from São Paulo State, Brazil (O’Dwyer et al. 2004). The canine Hepatozoon species from Brazil was not determined but the clinical signs and the low pathogenity of the Brazilian species indicated that we were dealing with H. canis or a close related species and not with H. americanum (O’Dwyer et al. 2001, 2004; Paludo et al. 2003).

Thus, the objective of this study was to perform the molecular characterization of canine Hepatozoon species from Brazil, by the analysis of the partial 18S rRNA gene sequences from Hepatozoon detected in naturally infected dogs.

Material and methods

Hepatozoon isolates

Investigation of dogs naturally infected by Hepatozoon sp. was performed by blood smears taken from the ear margin capillary bed, fixed with methanol and stained with Giemsa. Blood was collected by puncture of the cephalic vein to perform DNA extraction. A total of 31 dogs were examined. Positive EDTA-anticoagulated peripheral blood was kept frozen at −80°C until used.

DNA extraction

DNA extraction of Hepatozoon protozoa was performed pre-incubating the blood with proteinase K (digestion) by 4 h at 56°C. Afterwards, DNA was isolated from 200 μl aliquots of the blood using a QIAamp DNA mini Kit (QIAGEN, Hilden, Germany) according to the manufacturer’s instructions. Each DNA sample was eluted in 100 μl of TE buffer. Five microliter portions of these DNA extracts were used for PCR amplification.

PCR assay

The primers set HepF (5′ ATA-CAT-GAG-CAA-AAT-CTC-AAC 3′) and HepR (5′ CTT-ATT-CCA-TGC-TGC-AG 3′) was designed to amplify a partial 18S rRNA gene sequence of Hepatozoon spp. based upon alignment data from the H. canis from a dog in Israel (Genbank accession number AF176835), H. americanum from a dog in the US (AF176836) and H. catesbianae from a frog (AF176837) as described by Inokuma et al. (2002).

The amplification reaction was carried out in a total of 25 μl containing 1X Taq polymerase buffer (Pharmacia), 2 mM MgCl2, 0.2 mM dNTPs, 1.5 U of Taq DNA polymerase (Pharmacia) and 1 μM of each primers Hep F and Hep R (Inokuma et al. 2002).

PCR reactions were performed with the thermal profile consisted of a hot start of 3 min at 94°C and 35 repetitive cycles of 1 min at 94°C, 2 min at 57°C, and 2 min at 72°C followed by a 7 min extension at 72°C for one cycle. All amplifications were performed on thermocyler Biometra (T gradient).

Aliquots of amplified products (8 μl) were analyzed in ethidium bromide-stained 1% agarose gel by electrophoresis at 100 V for 30 min in TAE buffer and visualized under UV transluminator. The total remaining reaction products were purified by purification Kit MontageTM PCR Centrifugal Filter Devices (Millipore). The purified products were dissolved in 20 μl of TE prior to sequencing. Selected products results were confirmed by sequencing.

Sequencing

Sequencing of PCR products amplified from dog blood samples was carried out in both directions using the “ABI Prism BigDye Terminator Cycle Sequencing Ready Reaction Kit” (PE Applied Biosystems, Foster City, CA, USA). Approximately, 10 ng of purified DNA, for each sequencing reaction, was combined with 3.2 pmol of primer (sense and/or reverse) used in the amplification reaction. Nucleic acid sequence analysis was performed on an automated Applied Biosystems 377 DNA sequencer.

Nucleic acid sequence analysis

The computer analysis of nucleic acid sequence data was performed using MERGER package software. The multiple sequence alignment method and a neighbor joining phylogenetic tree (Saitou and Nei 1987) was constructed using the Clustal W program (Thompson et al. 1994). A phylogenetic tree was visualized using the TREEVIEW 1.4 program (Page 1996). The bootstrap test was applied to estimate the confidence of branching patterns of the neighbor-joining tree (Felsenstein 1985).

Divergences were estimated by the two-parameter method using the MEGA “Molecular evolutionary Genetics” software in the final documentation.

Results and discussion

From the 31 dogs examined, 7 were positive by blood smear examination (22.6%) and 21 were positive by the PCR (67.7%) (Fig. 1). All the positive animals by blood smear examination were also positive by PCR. From the 21 PCR positive blood, 8 were selected for perform the sequencing.

Fig. 1
figure 1

PCR amplification of Hepatozoon rDNA from Brazilian dog samples. Samples 1, 2, 3, 5 and 6—positive for Hepatozoon presence. Samples 4 and 7—negative amplifications. MW molecular weight marker

Full size image

The nucleotide sequences of the 625 bp PCR product excluding the primer regions revealed that Brazilian isolates were closely related with the Japanese Hepatozoon species, which was considered a strain variant of H. canis, but significantly different from H. americanum (Fig. 2).

Fig. 2
figure 2

Neighbor-Joining tree based on Hepatozoon rDNA gene. Individual sequences (HEP) were used to construct a Neighbor-Joining tree. Numbers on branches are bootstrap values (only values above 50 are shown). Fukuoka-sequence from Inokuma et al. (2002)

Full size image

Inokuma et al. (2002) obtained Hepatozoon DNA from only two dogs, and both sequences were identical to each other. The sequence of the Japanese Hepatozoon was similar to that of H. canis from Israel with 99% nucleotide identity and distantly related with H. americanum with 94% identities. Our sequences showed the presence of one polymorphic site with a transversion (T↔G).

These results indicate that the canine Hepatozoon species from Brazil is H. canis but with the possibility of the occurrence of variant strains in the same geographic region. O’Dwyer et al. (2004) already suggested that the canine Hepatozoon species from Brazil was H. canis. The infection is very prevalent in some regions of Brazil, has low pathogenicity and is frequently associated with other hemoparasites (O’Dwyer et al. 2001). Also, the tissue stages observed were similar to the ones of H. canis (O’Dwyer et al. 2004).

The clinical, epidemiological and biological significance of the Brazilian isolates should be investigated.