Phylogenetic and histopathological characterization of putative Hepatozoon ophisauri-like lineages in rodents from Italy

Abstract

Purpose

The aim of this study was to describe the occurrence and phylogenetic diversity of Hepatozoon spp. in small mammals from the Emilia-Romagna region, northern Italy, and to assess their genetic relationships with previously described lineages.

Methods

Tissue samples (spleen and ear) from wild animals belonging to eight small mammal species were examined. Molecular screening was performed using PCR targeting the 18 S rRNA gene of Hepatozoon spp. Positive amplicons were sequenced and analyzed using phylogenetic methods to determine their taxonomic placement, and one positive fresh sample was additionally subjected to histopathological characterization.

Results

Hepatozoon DNA was detected in two rodent species, the brown rat (Rattus norvegicus) and the European edible dormouse (Glis glis). Sequences showed high identity (98.9–100%) with Hepatozoon sp. previously reported in rodents from Spain and Tanzania, and 98.32–98.9% identity with Hepatozoon ophisauri. Phylogenetic analysis placed the sequences from R. norvegicus within a rodent-associated clade, whereas those from G. glis formed a distinct lineage related to Hepatozoon detected in a great gerbil. Histopathological analysis in a European edible dormouse revealed multifocal pulmonary granulomatous inflammatory lesions associated with intracytoplasmic parasitic zoites and developmental stages consistent with type I and type II meronts.

Conclusion

This study provides molecular evidence of distinct Hepatozoon lineages in European rodents, potentially representing undescribed taxa, and highlights the need for integrative studies combining molecular, morphological, and ecological data.

Supplementary Information

The online version contains supplementary material available at 10.1007/s11686-026-01304-2.

Introduction

Hepatozoonosis is an arthropod-borne disease caused by apicomplexan protozoa belonging to the family Hepatozoidae and the suborder Adeleorina. Hundreds of Hepatozoon species have been described infecting all classes of terrestrial vertebrates, typically by arthropod vectors, and approximately 120 and 50 Hepatozoon species have been found infecting snakes and mammals, respectively [1]. Hepatozoon species have a heteroxenous life cycle, meaning they require multiple hosts to complete their development. Typically, Hepatozoon species switch between intermediate vertebrate and definitive blood-feeding invertebrate hosts [2]. Diagnosis of Hepatozoon spp. infection in vertebrates is usually through blood smear analyses, but not all infected animals show parasitemia. In addition, molecular techniques such as PCR have also been used to detect the presence of Hepatozoon DNA in mammal tissues [3, 4].

The geographic distribution of Hepatozoon species, particularly those infecting snakes and rodents, is very wide. Hepatozoon species have been identified in various wild animals in Europe [5, 6], the Americas [5, 7, 8], Asia [7], and Africa [8]. In Italy, Hepatozoon spp. are widely distributed, affecting both domestic and wild animals, but investigations have focused mainly on domestic carnivores. Hepatozoon canis is the most extensively studied species having been reported in both domestic dogs and red foxes [9, 10]. In Southern Italy, 14% of hunting dogs have tested positive for H. canis, while a prevalence of 18% was reported among domestic dogs in Sardinia [11, 12]. Higher prevalence has been observed in Central Italy, where 32.5% of dogs tested positive [13]. Other Hepatozoon species, i.e. H. felis and H. silvestris, have been detected in domestic cats in northeastern Italy, with a reported total prevalence of 16.5% [14]. Rodent-associated Hepatozoon species, including H. erhardovae and several novel genotypes, have been reported in small mammals in Europe. Fleas (e.g., Ctenophthalmus spp. and Megabothris turbidus) can act as invertebrate hosts, especially for rodent-associated Hepatozoon species, whereas other Hepatozoon spp. are hypothesized to rely on non-vectorial transmission pathways [15]. On the other hand, H. ophisauri, usually associated with reptiles, has never been reported in European countries but has been detected in rodent blood samples in geographical regions such as Malaysia [16], Pakistan [15] and Taiwan [17]. Like other Hepatozoon species, it is likely transmitted through the ingestion of infected arthropods, mainly ticks [18].

The primary vectors of Hepatozoon species are ticks, particularly Rhipicephalus sanguineus sensu lato, which are highly prevalent in Italy and are known to play a central role in the transmission of H. canis [19, 20]. Molecular detection and characterization of Hepatozoon spp. typically target the 18 S rRNA gene, a robust marker for phylogenetic and epidemiological studies across a wide range of vertebrate and invertebrate hosts [18, 21]. However, due to the highly conserved nature of this locus, the 18 S rRNA gene may lack sufficient taxonomic resolution to discriminate among closely related Hepatozoon species [4, 22]. Despite this, knowledge about the distribution and life cycles of distinct Hepatozoon genotypes in Italy – particularly those infecting wild rodents – remains limited, highlighting the need for further molecular investigations to elucidate their host associations, transmission dynamics, and potential co-evolutionary relationships.

Materials and Methods

Sample Collection and Tissue Analysis

The Emilia-Romagna region (Northern Italy) includes the intensively cultivated and urbanized Po Plain, as well as hilly and mountainous areas that create diverse habitats. Nature conservation is supported by a network of protected areas hosting a high diversity of mammalian species [23]. The Emilia-Romagna Regional Council established a wildlife health surveillance plan (Deliberation n. 1763, November 29, 2017) to monitor several infectious and parasitic diseases through the analysis of wild animal carcasses. Small mammal carcasses included in this study were obtained through passive surveillance within this framework, consisting of animals found dead and submitted for diagnostic investigation. Examinations included histopathological assessment, ectoparasite collection, and tissue sampling. Spleen samples were collected during necropsy for subsequent PCR analyses.

In the case of a freshly deceased European edible dormouse, additional tissues were preserved for histopathological assessment, as histological examination is generally not feasible in frozen carcasses. For histopathology, selected organs were fixed in 10% buffered formalin, trimmed, embedded in paraffin, sectioned at 3–4 μm, and stained with haematoxylin and eosin. This additional European edible dormouse was also subjected to PCR and sequencing, and the resulting sequence was compared with those obtained from the original dataset.

DNA Aalysis

Total DNA was extracted from 25 mg of spleen and the central hairless part of the earlobe. Tissue samples were homogenized for 3 min in TissueLyser III (Qiagen) and then incubated overnight at 56°C in 200 µL of lysis buffer solution containing 20 µL of proteinase K (10 µg/mL). Then, DNA was extracted using the NucleoSpin Tissue Kit (Macherey Nagel, Duren, Germany) following manufacturer’s instructions and eluted in 200 µL of elution buffer. Between 30 and 50 ng of DNA were used as template in PCRs targeting the 18S rRNA gene using primers amplifying Babesia, Theileria and Hepatozoon species (primer forward BJ1: 5’-GTCTTGTAATTGGAATGATGG-3’; primer reverse BN2: 5’-TAGTTTATGGTTAGGACTACG-3’), which amplify a region of approx. 420 bp [24]. PCRs were prepared in a 25 µL final volume using the HotStartTaq Master Mix Kit (Qiagen) and carried out in a BioRad CFX96 thermal cycler (Bio-Rad Laboratories, Hercules, CA, USA) with the following temperature profile: 5 min at 95 °C, then 40 cycles of 30 s at 95 °C, 30 s at 52 °C and 45 s at 72 °C, followed by final extension for 8 min at 72 °C. Positive and negative controls were included in each PCR run. The positive control consisted of positive Hepatozoon martis samples previously confirmed as for by PCR and sequencing, which were re-extracted alongside newly tested samples and amplified in parallel [25]. The negative no-template control consisted of nuclease-free water added in place of template DNA.

After amplification check via gel electrophoresis, PCR products were purified with Exonuclease I and Thermosensitive Alkaline Phosphatase (Thermoscientific, Waltham, Massachusetts, USA) and Sanger sequenced. Sequencing reaction was performed with the Big Dye^® terminator ready reaction v1.1 kit and subsequently purified with BigDye^® Xterminator Purification kit (Thermoscientific, Waltham, Massachusetts, USA). Sequencing runs were carried out on a SeqStudio platform (Thermoscientific, Waltham, Massachusetts, USA) and the resulting forward and reverse sequences were quality checked with Sequencing analysis v5.4 software, and lately assembled in a consensus sequence with the Lasergene software, using SeqMan module (DNAStar, Madison USA, v. 17).

Phylogenetic Analyses

Sequences were manually cleaned and inspected for double bases observed in the electropherograms. Sequences derived from different tissues of the same individual were identical and collapsed into a single consensus sequence. The resulting consensus sequences were compared with those available in the NCBI GenBank database using BLAST tool (http://blast.ncbi.nlm.nih.gov/), through the ‘megablast’ algorithm. Cleaned sequences were aligned with other Hepatozoon spp. available in GenBank^® with the MUSCLE algorithm [26] within the MEGA7 software [27]. Then, the Hasegawa-Kishino Yano nucleotide substitution model was estimated with jModelTest 2 as the best nucleotide substitution model in the dataset [28]. A Bayesian Inference phylogenetic tree was constructed with the BEAST 2.6 package [29]. Sequence alignments were uploaded into BEAUti to define 10⁸ Markov chain Monte Carlo with 10% burning and sampling every 10³ trees. Chain convergence and effective sample sizes larger than 300 were inspected for each tree prior in Tracer. Trees were summarized in TreeAnnotator and visualized in FigTree. Then, a Templeton Crandall Sing haplotype network with 95% connection limit was traced with PopArt [30].

Results

During 2023 and 2025, we analysed a total of 132 tissue samples from 74 small-sized mammals belonging to 8 different species and collected from nineteen municipalities of the Emilia-Romagna region (Tables S1, S2).

Histopathological examination of a European edible dormouse lung specimen revealed pulmonary lesions and developmental stages consistent with Hepatozoon spp. infection. At low-magnification examination, the histological examination of the pulmonary parenchyma exhibited multifocal areas of atelectasis associated with increased cellularity (Fig. 1a). Higher-magnification analysis of the collapsed and thickened parenchymal regions revealed multifocal, irregularly distributed, intracytoplasmic cyst-like structures within reactive, phagocytic pulmonary histiocytes, containing elongated, falciform, mononucleated cells consistent with developmental stages of apicomplexan protozoa (Fig. 1b).

Fig. 1.

Pulmonary histopathological alterations at low magnification. a Overview of an affected parenchymal region showing multifocal atelectasis and increased cellularity (4× objective; scale bar = 200 μm). b Higher magnification of the affected area illustrating the interstitial inflammatory infiltrate, composed predominantly of histiocytes with a minor lymphoplasmacytic component, and intracytoplasmic cyst-like structures containing parasitic zoites (arrows) (20× objective; scale bar = 20 μm). Hematoxylin and eosin stain

The atelectatic changes were attributed to diffuse interstitial inflammatory infiltration, predominantly composed of histiocytes, with a minor, scattered lymphoplasmacytic component, occasionally organizing into discrete follicular aggregates. Reactive hyperplasia of type II pneumocytes was additionally observed lining the affected alveolar walls.

The cyst-like structures, measuring approximately 18–20 μm in diameter and enclosed by a cyst wall of uniform thickness, exhibited two distinct morphological phenotypes consistent with two sequential merogonic developmental stages. The first phenotype comprised structures harbouring one to six large, broad-bodied, falciform cells with a conspicuous centrally to subterminally positioned nucleus, morphologically consistent with type I meronts containing macromerozoites (Fig. 2a, b,c). The second phenotype consisted of structures enclosing 12 to 20 slender micromerozoites with terminally positioned nuclei arranged in a palisade pattern, consistent with type II meronts (Fig. 2d, e). In some type II meronts, nuclei with condensed chromatin were peripherally aligned, surrounding a large, central, amorphous eosinophilic mass, consistent with an incompletely differentiated residual cytoplasm.

Fig. 2.

High-magnification photomicrographs of meront structures. In all panels, the inflammatory infiltrate is visible, consisting of a predominant histiocytic component and a less represented lymphoplasmacytic component. a Type I meront containing two well-formed macromerozoites. b Two adjacent type I meronts: the left one is monozoic, harboring a single piriform, broad-bodied merozoite; the right one contains six well-defined merozoites with a centrally positioned nucleus. c Type I meront containing six well-defined macromerozoites. d Type II meront within an interstitial septum, amid moderate inflammatory infiltration, displaying approximately fifteen micromerozoites arranged in a palisade pattern with terminally positioned nuclei. e Type II meront exhibiting incompletely differentiated micromerozoites, characterized by a central cytoplasmic residual body surrounded by peripherally arranged nuclei with condensed chromatin. f Type II meront containing approximately twelve irregularly arranged micromerozoites with subterminally positioned nuclei. Hematoxylin and eosin stain; all scale bars = 20 μm; 40× objective

Regarding the molecular investigation, we obtained a total of five consensus sequences from five positive individuals, two brown rats (Rattus norvegicus) and three European edible dormice (Glis glis). The histopathologically examined European edible dormouse yielded a sequence showing 100% identity to another dormouse isolate (i.e. isolate 1, PX232825); it was therefore considered confirmatory and not included as a separate sequence in the phylogenetic tree or haplotype network.

Among these, two animals (one brown rat and one European edible dormouse) tested positive for Hepatozoon sp. in both ear and spleen tissues. Isolates 1 and 2 obtained from Glis glis were 98.70% and 98.86%, respectively, similar to several uncharacterized Hepatozoon sp. (including PP980736 from Crycetomys ansorgei from Tanzania), but also 98.70% similar to H. ophisauri (MN723845 from the blood of an unknown animal of Iran) and H. bashtari (MN497412 from the snake Echis coloratus from Saudi Arabia). In addition, isolates 1 and 2 from R. norvegicus were 100% and 99.44% similar to an uncharacterized Hepatozoon sp. (MT919389 obtained from the liver of another brown rat from Spain) and 98.32–98.37% similar to H. ophisauri (MN723845 obtained from undefined host in Iran).

Bayesian inference phylogenetic analysis included different Hepatozoon spp. associated with rodents and reptiles, such as H. ophisauri, H. ayorgbor, H. bashtari, H. griseisciuri, H. annularis, H. domerguei, H. caimani, while Neospora caninum was used as an outgroup (Fig. 3a). Isolates 1 and 2 obtained from G. glis and R. norvegicus clustered in a group containing two H. ophisauri sequences, one obtained from the spleen of Natal multimammate mouse (Mastomys natalensis) from Tanzania (OL982745) and one from the spleen of great gerbil (Rhombomys opimus) from China (MW256822). Within this group, the sequences from R. norvegicus and G. glis obtained in this study clustered separately with high posterior probabilities. Other H. ophisauri sequences reported from Kashmir rock agama (Laudakia tuberculata) from Pakistan and common muskrat (Ondatra zibethicus) from the USA were placed close to the Hepatozoon sp. from rodents of Italy of this study, as well as the Hepatozoon sp. sequences to which our sequences showed the closest BLAST similarities, i.e. MT919389 from Spain and PP98070 from Tanzania. In addition, two sequences of H. annularis and H. ayorgbor were placed within the H. ophisauri group, whereas H. domerguei, H. caimani and H. felis were grouped in a basal cluster.

Fig. 3.

Phylogenetic analyses based on 18 S rRNA fragments (546–549 bp) of Hepatozoon spp. detected in wild rodents of Italy. a Bayesian inference tree including Hepatozoon spp. sequences from this study (marked with black diamonds) and reference sequences. Posterior probability (PP) values lower than 0.6 are not shown; node size and colour are proportional to PP values. b Haplotype network of Hepatozoon spp. Circle size is proportional to the number of sequences sharing each haplotype, black circles denote hypothetical haplotypes, and hatch marks represent mutational steps between them. Neospora caninum was used as outgroup, while the scale bar represents the number of substitutions per site

The TCS haplotype network showed that all sequences were highly interconnected among them, with a vast majority represented by uncharacterized Hepatozoon sp. (Fig. 3b). A main haplotype composed by Hepatozoon sp., H. ophisauri and H. bashtari sequences was central to the other haplotypes, including the sequences obtained herein. Isolates from R. norvegicus formed a separate haplotype that was separated by five mutations from a haplotype composed by H. ophisauri and Hepatozoon sp. Moreover, the two isolated from G. glis formed another haplotype closely related to the main haplotype.

The sequences obtained from the common rat and the dormouse corresponding to Hepatozoon ophisauri-like genotypes have been deposited in GenBank^® under accession numbers PX232825 and PX232826 for sequences obtained from European edible dormice, and PX232827 and PX232828 for sequences obtained from brown rats.

Discussion

This study represents the first report of natural infection with Hepatozoon ophisauri-like hemoparasites in wild rodents from Italy, expanding the known host range and geographical distribution of this parasite. The detection of Hepatozoon DNA in two rodent species, R. norvegicus and G. glis, suggests that these mammals may act as intermediate or paratenic hosts within a complex transmission network involving both mammals and reptiles. These findings contribute to growing evidence that Hepatozoon spp. circulation is shaped by ecological interactions among vertebrate hosts, in agreement with recent coevolutionary analyses [2]. Hepatozoon spp. parasites detected in this work formed a distinct cluster relative to several other Hepatozoon lineages reported from different geographical areas and were closely related to H. ophisauri sequences previously obtained from rodents in China, Tanzania, and the USA. The recurrent detection of H. ophisauri-like lineages in potential prey species supports the hypothesis that rodents may act as first intermediate or paratenic hosts of snake-associated Hepatozoon species, with infection acquired through trophic interactions.

Evidence for prey–predator transmission has been repeatedly reported for hemogregarines, although direct evidence for this mechanism was not obtained in the present study. In saurophagous snakes of the genus Psammophis, three unrelated Hepatozoon lineages were each most closely related to lineages previously detected in lacertids and gekkonids on which these snakes feed, indicating that diet may represent a potential determinant of parasite acquisition [31, 32]. Importantly, prey–predator phylogenetic signatures have also been observed across other vertebrate groups: Allen et al. [33] reported that Hepatozoon sequences from wild European rabbits (Oryctolagus cuniculus) clustered most closely with carnivore-associated isolates, while a sequence from a boa constrictor (Boa constrictor) was most similar to rodent-associated lineages, supporting the view that trophic links may contribute to structuring Hepatozoon diversity across hosts. These patterns are consistent with the hypothesis that close host–parasite interactions, including trophic relationships and host switching, contribute to shaping Hepatozoon spp. evolutionary history [2]. In this context, predator–prey interactions documented in Italy further support the ecological plausibility of rodent-mediated circulation of H. ophisauri-like parasites. Rodents represent common prey for mesocarnivores such as domestic cats (Felis catus) and red foxes (Vulpes vulpes), which could potentially contribute to parasite dissemination within synanthropic and peri-urban food webs [34, 35], and local predator–prey interactions may facilitate contacts among potential intermediate hosts and predators in anthropogenic landscapes [36, 37]. In addition, available H. ophisauri sequences deposited in GenBank^® have frequently been derived from potential snake prey, including rodents and lizards from geographically distant regions (e.g., Pakistan, Taiwan, Tanzania, China, the USA, and Iran) [15], supporting this hypothesis but not demonstrating the transmission pathway directly. Nevertheless, experimental elucidation of complete life cycles will be necessary to confirm the specific role of rodents and lizards as intermediate or paratenic hosts of snake-associated Hepatozoon lineages. In the case of the Hepatozoon ophisauri-like lineages detected here, prey–predator transmission therefore remains a hypothesis requiring further investigation, particularly through the study of sympatric reptiles potentially involved in the local transmission cycle.

The snake fauna of Emilia-Romagna includes several species whose diet comprises small mammals and/or reptiles, making trophic transmission biologically plausible in the study area. Generalist feeders such as the green whip snake (Hierophis viridiflavus), as well as species primarily preying on small mammals [e.g., the Aesculapian snake (Zamenis longissimus) and the four-lined snake (Elaphe quatuorlineata)], could potentially acquire infection through ingestion of infected rodents. In contrast, species with predominantly amphibian or fish-based diets (e.g., Natrix spp.) are less likely to participate in rodent-associated transmission cycles [21]. While this ecological context supports the feasibility of multi-host transmission pathways, direct evidence from sympatric reptile populations is still lacking.

While the trophic transmission hypothesis is supported by phylogenetic and ecological considerations, the life cycle of Hepatozoon species typically involves hematophagous arthropods as definitive hosts. Although ticks are recognized as primary definitive hosts of many Hepatozoon species, no ectoparasites were detected on the examined rodents, preventing assessment of their infection status in potential invertebrate hosts. However, the absence of ectoparasites on carcasses does not exclude vector-mediated transmission. Ectoparasites may abandon hosts after death [8], and carcass condition, seasonal factors, or host-specific ectoparasite preferences may further reduce detectability. In addition to ticks, other arthropods such as fleas have been proposed as potential hosts of certain Hepatozoon species [38], suggesting that transmission routes may be more diverse than previously assumed. Targeted live-trapping surveys coupled with systematic ectoparasite collection will therefore be necessary to clarify the role of different arthropod taxa in the local transmission cycle. Although ecological and phylogenetic evidence supports a trophic transmission framework, our conclusions are currently based solely on molecular data. The use of previously frozen carcasses may have compromised host tissue and parasite preservation, limiting detectability of intracellular developmental stages [39]. In vertebrate hosts, Hepatozoon meronts are typically located within muscle or reticuloendothelial tissues and can be identified histologically when adequately preserved [1]. Although this limitation applied to most carcasses included in the study, histopathological examination of one well-preserved European edible dormouse provided morphological confirmation of the parasite and evidence of stages consistent with merogonic development in lung tissue. Although this anatomical site is atypical for mammal-associated Hepatozoon species with respect to merogonic development, it has been frequently reported in reptile-associated species [40, 41]. Nevertheless, future studies based on additional well-preserved samples will be necessary to fully characterize tissue tropism and developmental stages of this lineage [25, 42]. Future studies should also consider applying additional PCR protocols targeting longer fragments of the 18 S rRNA gene, which may improve species delimitation and provide greater resolution for assessing genetic diversity and phylogenetic relationships within Hepatozoon lineages [25].

In conclusion, this study provides the first histopathological and molecular evidence of H. ophisauri-like parasites in wild rodents from Italy, expanding the known host range and geographical distribution of this lineage. Phylogenetic relationships, together with ecological considerations, support the hypothesis that rodents may participate in trophic transmission involving reptiles and possibly mammalian predators. However, the complete life cycle of this parasite in the study area remains unresolved. Broader molecular screening of sympatric snakes and potential invertebrate hosts, combined with histological examination of fresh vertebrate tissues, will be essential to clarify transmission routes, characterize developmental stages, and assess the genetic diversity of the detected lineage within the genus Hepatozoon.

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contributions

G.M.: Conceptualization, Data curation, Writing, Project administration, Funding acquisition; D.S.: Conceptualization, Methodology, Writing and editing; M.C.: Conceptualization, Writing, review and editing; M.S., D.T., V.C, P.B.: Methodology; F.M.D.; Methodology, Writing and editing; A.R.: Conceptualization, Formal analysis, Visualization, Writing.All authors have read and approved the final version of the manuscript.D.S. and M.S. share first authorship.

Funding

This work was supported by the Italian Ministry of Health, Directorate General for animal health and veterinary medicinal products (DGSAF), grant number PRC2022011. D.S. was supported by EU funding within the NextGeneration EU-MUR PNRR Extended Partnership initiative on Emerging Infectious Diseases (project no. PE00000007, INF-ACT).

None of the funding sources had roles in study design, collection, analysis, and interpretation of data, in the writing of the report, and in the decision to submit this article for publication.

Data Availability

Sanger sequences from this study have been deposited to GenBank^® under the accession numbers PX232825–PX232828.

Declarations

Conflict of interest

The authors declare no competing interests.

Ethical Approval

This article does not contain any studies with alive animals performed by any of the authors. Samples collected for genetic analysis were obtained from dead animals.