Hepatozoon (Eucoccidiorida: Hepatozoidae) in wild mammals of the Americas: a systematic review

Abstract﻿ Background The study of parasites provides insight into intricate ecological relationships in ecosystem dynamics, food web structures, and evolution on multiple scales. Hepatozoon (Eucoccidiorida: Hepatozoidae) is a genus of protozoan hemoparasites with heteroxenous life cycles that switch infections between vertebrates and blood-feeding invertebrates. The most comprehensive review of the genus was published 26 years ago, and currently there are no harmonized data on the epizootiology, diagnostics, genotyping methods, evolutionary relationships, and genetic diversity of Hepatozoon in the Americas. Methods Here, we provide a comprehensive review based on the PRISMA method regarding Hepatozoon in wild mammals within the American continent, in order to generate a framework for future research. Results 11 out of the 35 countries of the Americas (31.4%) had data on Hepatozoon, with Carnivora and Rodentia orders having the most characterizations. Bats, ungulates, and shrews were the least affected groups. While Hepatozoon americanum, H. americanum-like, H. canis, H. didelphydis, H. felis, H. milleri, H. griseisciuri, and H. procyonis correspond to the identified species, a plethora of genospecies is pending for a formal description combining morphology and genetics. Most of the vectors of Hepatozoon in the Americas are unknown, but some flea, mite, and tick species have been confirmed. The detection of Hepatozoon has relied mostly on conventional polymerase chain reaction (PCR), and the implementation of specific real time PCR for the genus needs to be employed to improve its diagnosis in wild animals in the future. From a genetic perspective, the V4 region of the 18S rRNA gene has been widely sequenced for the identification of Hepatozoon in wild animals. However, mitochondrial and apicoplast markers should also be targeted to truly determine different species in the genus. A phylogenetic analysis of herein retrieved 18S ribosomal DNA (rDNA) sequences showed two main clades of Hepatozoon: Clade I associated with small mammals, birds, and herpetozoa, and Clade II associated with Carnivora. The topology of the tree is also reflected in the haplotype network. Conclusions Finally, our review emphasizes Hepatozoon as a potential disease agent in threatened wild mammals and the role of wild canids as spreaders of Hepatozoon infections in the Americas. Graphical Abstract Supplementary Information The online version contains supplementary material available at 10.1186/s13071-024-06154-3.


Background
The ongoing sixth mass extinction has driven efforts in documenting and understanding animal biodiversity for conservation and management [1].Wild vertebrates and their parasites maintain infectious agents such as viruses, bacteria, and protozoa that may eventually spill over to humans or domestic animals, leading to outbreaks or epidemics [2][3][4].As most species on earth are parasites, their role as indicators of the ecosystem's health has gained attention in recent years [5,6].To complete biological cycles, parasites rely on the trophic interactions of their hosts [7].Indeed, while some parasites co-evolve with their hosts, causing minimal damage, others are etiological agents [8][9][10].This duality in the biology of parasites is important to understanding intricate ecological relationships, ecosystem dynamics, food web structure, and evolution across multiple scales [9,10].Additionally, infectious diseases caused by parasites pose significant conservation challenges due to their impact on wild animals [11,12].Therefore, identifying the parasites that act as etiological agents is crucial in defining the threats to wildlife conservation [13,14].
Hepatozoon (Eucoccidiorida: Hepatozoidae) is a genus of protozoa that invades red and white blood cells of vertebrates [15,16].Hepatozoon spp.are commonly found in wildlife perpetuating in enzootic cycles [15,17].These hemoparasites have heteroxenous life cycles involving a vast array of intermediate hosts (vertebrates) and definitive hosts (blood-feeding invertebrates) [15,16], a fact that has shaped the remarkable biological diversity of the genus.
After its original description by Balfour in 1905, the most comprehensive review on the genus Hepatozoon was published 90 years later [15].Further reviews added data mainly on canine hepatozoonosis [16,18,19], and there are currently no harmonized data on the epizootiology of the genus Hepatozoon in the Americas.Here we provide a comprehensive review, based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) method [20], of Hepatozoon in wild mammals of the American continent, with the following objectives: (i) to catalogue the publications since the original description of the taxon, (ii) to compare the publications by country, (iii) to compile the nature of biological samples and the techniques employed to detect and genetically characterize the parasite, (iv) to assess the conservation status of wild mammals with confirmed infection, and (v) to analyze available genetic data.Collectively, the objectives of this review were to generate a framework of reference for future investigation on Hepatozoon in the Americas.

Literature search and database
We recovered literature on Hepatozoon published between 1916 and December 31, 2022, through an exhaustive review using reputable sources including Pub-Med (https:// pubmed.ncbi.nlm.nih.gov/), SciELO [Scientific Electronic Library Online] (https:// scielo.org/ es/), Scopus (https:// www.scopus.com/ home.uri), and Web of Science (https:// webof knowl edge.com/ UA).Only peerreviewed publications were included.Data from books, gray literature such as scientific conferences, and theses were excluded.Data on Hepatozoon retrieved from domestic animals or invasive rodents were not quantified, but molecular sequences were included in genetic analyses.
To mitigate bias and maximize the number of available studies, we used targeted search strings and titleabstract-keywords unit searches, incorporating the terms "Hepatozoon, " "hepatozoonosis, " "wild mammals, " "wildlife, " "wild animals, " and "mammals, " and names of the countries in the American continent, along with names of orders and groups of mammals native to the Americas, as detailed in the Handbook of the Mammals of the World [21].To hone our search, we applied Boolean operators "AND" and "OR, " and followed the advanced search protocols suggested in the selected databases.
Compiled papers were subjected to the PRISMA protocol [20], with the following inclusion criteria: (i) reports of Hepatozoon in mammals of the Americas and (ii) studies on ectoparasites extracted from mammals of the Americas that assessed the presence of Hepatozoon.The documents lacking these criteria were excluded.The metadata matrix included the publication year, country, vertebrate host order, family, species, type of sample, diagnostic technique, conservation and population statuses of the hosts according to the International Union for Conservation of Nature (IUCN) Red List (https:// www.iucnr edlist.org/), primers used in molecular diagnostics, and study references.Scientific names and taxonomy of the hosts species referenced in this review followed the Integrative Taxonomic Information System (ITIS) (https:// www.itis.gov/).

Analysis of sequences Sequence selection and alignment
To define the final dataset included in this study, sequences of Hepatozoon and associated metadata were sourced from the GenBank database (https:// ncbi.nlm.nih.gov/ nucle otide) up to December 31, 2022, using "Hepatozoon" as the search criterion.Subsequently, the sequence filtering was carried out in four steps.First, only 18S ribosomal DNA (rDNA) sequences were included.Second, sequences annotated with a nomenclature other than Hepatozoon were excluded.Third, each sequence was individually aligned against an annotated 18S rDNA reference sequence of Hepatozoon (MH615006) to identify those sequences that showed more than 50% coverage for a given region of the 18S rRNA gene.Fourth, each sequence that satisfied the aforementioned criteria was recorded in a table, specifying the GenBank accession number, submission date, sequence length, and the flanked region of the 18S rRNA gene.Finally, all sequence datasets were aligned with MAFFT using the G-INS-i algorithm [22].
Rapid hill-climbing and stochastic disturbance methods were employed for ML phylogenies, with 1000 ultra-fast bootstrapping pseudo-replicates to evaluate the inferred tree robustness.Ultra-fast bootstrap values < 70%, 70-94%, and ≥ 95% were considered non-significant, moderate, and high statistical support, respectively [31].Regarding the BI phylogenies, two independent tests of 20 × 10 6 generations and four Markov chain Monte Carlo (MCMC) chains were implemented, sampling trees every 1000 generations and removing the first 25% as burn-in.Tracer v. 1.7.1 [32] was employed to confirm the correlation and effective sample size (ESS) of the MCMCs.Bayesian posterior probabilities (BPP) > 0.70 at nodes were considered high statistical support [33].Trees were visualized and edited with FigTree v. 1.4.1 (http:// tree.bio.ed.ac.uk/ softw are/ figtr ee/) and Inkscape v 1.1 (https:// inksc ape.org/ es/).A consensus phylogram for both ML and BI was generated following the approach outlined by Santodomingo et al. [34].

Haplotype analyses
A haplotype analysis of Hepatozoon genotypes was intended to evaluate genetic variants circulating among wild mammals in the Americas.According to the National Human Genome Research Institute (https:// www.genome.gov/), a genotype is defined as a DNA sequence at a specific locus, whereas a haplotype is a set of genetic variants, or polymorphisms, that tend to be inherited together.To calculate the number of polymorphic sites (S), haplotypes (H), haplotype diversity (Hd), nucleotide diversity (π), and haplotype frequencies, the gametic phase was inferred using the PHASE module in DNAsp v. 6.12.03 software [35].Nucleotide polymorphism analyses excluded gaps and considered invariable sites.A median-joining (MJ) haplotype network [36] was constructed using PopART v. 1.7 (http:// popart.otago.ac.nz) to display haplotype frequencies.

General findings
A total of 1406 studies were identified as potentially useful for the review.After thoroughly reviewing the papers meeting the selection criteria, 84 studies (5.97%) were selected, as they contained information on Hepatozoon in wild mammals in the Americas and their ectoparasites (Fig. 1).Eleven out of the 35 countries of the Americas (31.4%), including Argentina, Brazil, Canada, Chile, Colombia, Costa Rica, French Guiana, Panama, Uruguay, the USA, and Venezuela, had data on Hepatozoon (Additional file 1: Table S1).No data were found for the other 24 American countries.
The majority of studies on Hepatozoon in wild mammals were concentrated in Brazil and the USA, with 36 (42.9%) and 30 (35.7%) papers, respectively (Additional file 1: Table S1).In contrast, other countries in the Americas yielded less than five studies (~ 6%), underscoring a significant gap in the literature.In fact, research in the Americas has largely focused on domestic animals [16,19,37,38].This disparity not only underscores the need for a broader geographical scope in future studies, but also calls for assessing the occurrence of Hepatozoon in wild mammals in regions that currently lack data.In this regard, Chile displays a recent uptick in research on Hepatozoon infections in wildlife (Additional file 1: Table S1).
Regarding the number of studies on Hepatozoon among mammalian orders, Carnivora and Rodentia had the most, with 43 (51.2%) and 35 (41.7%) reports, respectively.Didelphimorphia followed, with 10 studies (11.9%),Chiroptera with two studies (2.4%), and Artiodactyla with two studies (2.4%).Lagomorpha, Microbiotheria, Perissodactyla, and Soricomorpha each had one study (1.2%) (Fig. 2).The disproportional distribution of studies by country-predominantly in Brazil and the USAindicates a geographical bias.This fact not only skews the distribution of Hepatozoon spp., but also highlights the lack of a comprehensive understanding of infection by this protozoan in other mammalian orders.Furthermore, it denotes the limited research efforts on these hemoparasites in other American countries.
Overall, the number of studies on Hepatozoon in mammals of the Americas has displayed a notable upswing in recent decades (Fig. 2).Undoubtedly, the availability of molecular diagnostic tools contributed to this increase [39].Indeed, molecular tools enhance sensitivity in detecting Hepatozoon from different types of biological samples, thus facilitating the identification of new species and previously unrecognized infections in wild mammals.The potential impact of these infections on wildlife health, coupled with the growing interest in arthropodborne diseases, particularly those transmitted by ticks, has also contributed to the increase in research regarding Hepatozoon in wild mammals.

Epizootiology of Hepatozoon in mammals of the Americas
Hepatozoon spp.thrive in enzootic cycles [15,17], which are ecological systems involving vertebrate hosts, vectors, environmental components, and the critical community size of hosts required to maintain the infectious agent indefinitely [2,40].Hepatozoon spp.are sustained in both vertebrate and invertebrate hosts [15], with transmission pathways that shift depending on the specific parasitehost interactions.
To date, 106 species of mammals from nine orders have been screened for Hepatozoon.Of these, two species of carnivores were found infected with H. americanum (1.89%) and four species with H. americanum-like (3.77%) (Additional file 1: Table S1).The former species is known for its virulence in canids [66].The latter was considered an emerging South American variant of H. americanum [67]; however, H. americanum-like exploits a niche with different vectors and hosts, so it should be considered a different species.Moreover, pathogenic effects of H. americanum-like still need to be assessed in detail.In contrast, H. canis, which appears well adapted to its canine hosts [66,67], has also been found infecting eight out of nine canid species in the continent (88.9%).Notably, some species of Artiodactyla, Chiroptera, Didelphimorphia, and Rodentia orders may also be susceptible to H. canis (Additional file 1: Table S1), a fact that underlines the generalist and opportunistic nature of this species.
A total of 60 mammal species (56.60%) were found to be infected with Hepatozoon spp.such as Hepatozoon didelphydis, H. griseisciuri, H. milleri, and H. procyonis, each one exhibiting a different IF and infecting specific mammal groups (as shown in Table S1).Furthermore, lineages of Hepatozoon have been documented, covering species associated with carnivores, herpetozoa and small mammals, opossums, reptiles, and rodents (Additional file 1: Table S1).According to Dupré [70], a "lineage" is defined as an independent evolutionary line that extends back in time from a current species to its ancestors.In this sense, Hepatozoon lineages reflect unique evolutionary histories and specific adaptations of Hepatozoon spp. to different hosts, suggesting putatively novel species associated with a diverse array of wild mammals in the Americas.
According to the data gathered (Additional file 1: Table S1
Concerning mammal species in the Americas, bloodsucking ectoparasites that could be related to the epizootiology of the Hepatozoon species include ticks such as Amblyomma dubitatum, A. maculatum, A. tigrinum, and A. sculptum in carnivores [13,71,[87][88][89]; Ixodes neuquenensis in Microbiotheria [75]; and Amblyomma fuscum and species of the Ixodes sigelos group in rodents [79,90].To better understand the implications of these ticks in the Hepatozoon epizootiology among the Americas' ecosystems, experimental transmission studies are necessary to clarify these associations.Transstadial detection of oocysts in tick hemolymph would also contribute to elucidate vector roles.
While fleas, lice, and mites are also considered vectors or definitive hosts of Hepatozoon spp.associated with rodents [64,65], only one flea species (Megabothris abantis) and two mite species (Euhaemogamasus ambulans and Echinolaelaps echidninus) have been reported as vectors for Hepatozoon spp. in rodents of the Americas [64].Moreover, the fleas Amalaraeus dissimilis and Peromyscopsylla ostsibirica are possible vectors of Hepatozoon in rodents of genus Microtus in Alaska [91].The limited knowledge on Hepatozoon vectors stems from the challenges that the detection of Hepatozoon oocysts in bloodsucking arthropods pose [64,92].These may include finding the ectoparasites on their hosts, maintaining the vectors alive in laboratory conditions, and submitting their hemolymph to microscopical analyses for the detection of oocysts and sporocysts, and eventually genetic identification.

Detection and characterization of Hepatozoon in mammals of the Americas Biological samples used to detect Hepatozoon
Blood was the main sample employed to detect and characterize Hepatozoon, with 56 studies (66.7%).The liver was the second most frequently used sample, accounting for 13 studies (15.5%), followed by the spleen with 12 studies (14.3%).Other samples included lung (10 studies; 11.9%), heart (eight studies; 9.5%), and muscle (seven studies; 8.3%).Less commonly employed samples were bone marrow, tail, skeletal muscle, kidney, and synovial fluid (ranging from 1.2% to 3.6%) (Additional file 1: Table S1).Moreover, as ticks feed on vertebrate blood, they were used as sentinels to assess the presence of Hepatozoon in wild mammals in three studies (3.6%) [78,79,90].Therefore, the variety of tissues that Hepatozoon spp.may attain in the vertebrate host, including their ectoparasites, maximizes the chance of detection.
Detecting gamonts in blood is frequently used as a quick diagnosis of Hepatozoon infection in vertebrates [39].However, the spleen is more sensitive, because it harbors meronts (groups of meront cells in multiple division) and subsequently higher loads of parasites; therefore, it is considered the best target for detection [42,109].Meronts have also been observed in bone marrow of red foxes (Vulpes vulpes) [48,110].Depending on the Hepatozoon species, alternative samples for detection and histopathology include biopsies of bone marrow or skeletal muscle for H. canis [66,83,111], skeletal muscle for H. americanum and H. felis [66,68], and cardiac muscle for H. felis [68].
The effectiveness of each technique varies depending on the intensity of infection.For instance, while observing blood smears is sensitive in animals with high parasitemia, histology is valuable in detecting subclinical infections, in which Hepatozoon encysts in different organs and fewer gamonts circulate in blood [39].This is particularly relevant in H. americanum and H. americanum-like infection, two species that produce very low parasitemia during the clinical disease [16,39,84].PCR stands out as the gold standard for diagnosing and characterizing Hepatozoon spp., particularly in cases of subclinical infection with low parasitemia [39,112,113], where nPCR can significantly improve the specificity [39,89].Although real-time PCR is highly sensitive, none of the reviewed studies implemented this technique.Applying real-time PCR to surveil Hepatozoon infections in wildlife could significantly expand diagnostic capabilities.This approach not only will enable the rapid processing of large sample volumes, reducing the time and costs of diagnosis, but also will improve the sensitivity of cPCR in detecting Hepatozoon infections [114,115].

Hepatozoon genotyping
Molecular analyses of the phylum Apicomplexa have provided valuable insight into the genomic composition and genetic structure of the group, revealing the existence of three kinds of genomes: nuclear, mitochondrial, and the apicoplast [116][117][118].The first genetic characterizations of the genus Hepatozoon relied on partial sequences of the apicoplast-encoded 16S rRNA gene [119], the nuclear ribosomal DNA internal transcribed spacer 1 (ITS-1) [120,121], and the 18S rRNA gene [58,59,122,123].Since then, 18S rDNA sequences have been widely used for genotyping and molecular systematics of the genus [124][125][126].
Extrachromosomal genomes are often detected in significant quantities, with two or even 15 copies per cell [127,128].For the genus Hepatozoon, the first assembled mitochondrial genome (mitogenome) belonged to Hepatozoon catesbianae, a species that infects the frog Lithobates catesbeianus [129].Moreover, the apicoplast genome of H. canis [128] and the mitogenome of Hepatozoon spp.associated with rodents were recently sequenced [125,126].It was not unexpected that the mitochondrial cytochrome c oxidase I (cox1) locus had a higher nucleotide divergence (≥ 0.08) if compared with the 18S rRNA gene (≥ 0.012), becoming the marker of choice to distinguish Hepatozoon species [125,126,128].Nevertheless, the paucity of mitochondrial and apicoplast sequences in online databases represents a significant gap that needs to be addressed in order to unravel the intricate diversity and systematics of the genus.In this sense, the works of Léveillé et al. [125,128,129] and Hrazdilová et al. [126] are groundbreaking on a global scale.
Only one out of 84 reviewed articles used primers targeting the mitogenome of Hepatozoon in wild rodents [125].In contrast, the remaining studies employed primers for the 18S rRNA gene to detect and characterize Hepatozoon.This approach has been the primary method applied in mammals of the Americas since 2006 (Additional file 1: Tables S1, S3).Overall, the 18S rRNA gene is conserved and frequently used in Hepatozoon phylogenies because it yields good resolution to the genus or even species level in the Apicomplexa order [123,124,126,130].Furthermore, the most abundant collection of Hepatozoon sequences in public database derives from this marker (Additional file 1: Tables S2, S3).
It is important to keep in mind that an accurate molecular detection of Hepatozoon spp.relies mostly in the selection of primers and an optimal thermal cycling condition [39].Reported sequences in the reviewed studies were obtained through different PCR protocols, using 23 distinct primer sets (refer to Fig. 4 and Additional file 1: Table S2 for details).These protocols yielded sequences ranging from 300 to 1,816 base pairs (bp), with most primers targeting the V4 region of the 18S rRNA gene (Fig. 4).
A correct choice of primers for amplifying specific DNA fragments is crucial to achieving the objectives in a given study [39,131].A key aspect in primer selection is using those primers with the lowest potential to form secondary structures, such as hairpins or dimers (Additional file 1: Tables S2).For molecular diagnostics, it is preferable to opt for primers that amplify shorter fragments to maximize the sensitivity of the results [39].Conversely, when conducting phylogenetic inferences, it is recommended to choose primers that yield longer fragments or even a complete gene sequence (Fig. 4, Additional file 1: Table S2) [39].Nevertheless, some studies constructed phylogenies of the genus Hepatozoon incorporating short and long sequences, and the trees yielded congruent topologies consistent with the evolutionary history of the genus [124,130,132,133].It is important to note that these studies implemented short sequences above 387 bp flanking the IV region of the 18S rRNA gene.

Evolutionary relationships and haplotype diversity of Hepatozoon spp. in mammals of the Americas
The small subunit of the ribosomal RNA gene (SSU rRNA or 18S rRNA) encompasses nine regions (V1-V9), with the regions V2, V4, and V9 the most hypervariable [131].As of December 31, 2022, the GenBank database lists over 2916 18S rDNA sequences (≥ 387 bp) associated with Hepatozoon, flanking one or more 18S rRNA gene variable regions (Additional file 1: Table S3).For the V4 region, 2596 sequences (390-1816 bp) showed over 50% coverage, and 247 have been recovered from 48 wild mammal species across the Americas (Additional file 1: Table S3).Conversely, Hepatozoon sequences corresponding to other regions of the gene were less represented (Additional file 1: Table S3).Therefore, analyses of genotypes incorporated a final curated dataset of 115 sequences obtained from mammals in the Americas,  S3 along with 126 Hepatozoon sequences recovered from birds, small mammals, herpetozoa, and carnivores worldwide.In contrast, haplotype analyses were confined to the aforementioned 115 sequences.All sequences overlapped in the V4 region of the 18S rRNA gene, each with ≥ 500-bp length, similar to the approach applied by Vásquez-Aguilar et al. [38].
Before detailing the results of genotype and haplotype analyses, a brief recall on the definitions of these terms is given for clarity of interpretation.A genotype refers to the specific DNA sequence of a given locus, while a haplotype encapsulates a collection of genetic variants, typically co-inherited.These definitions are instrumental to analyze and understand the phylogenetic relationships and genetic variability within the Hepatozoon genus, enabling a deeper comprehension of the evolutionary dynamics shaping the species among diverse ecological niches.

Phylogenetic inferences and haplotype diversity
Both ML and BI trees revealed two main clades with strong node support: (i) Hepatozoon spp.associated with small mammals, birds, and herpetozoa (Fig. 5, Clade I), and (ii) Hepatozoon spp.related to carnivores (Fig. 6, Clade II).However, Hemolivia and Karyolysus render the genus Hepatozoon non-monophyletic (Figs. 5, 6).Previous studies also support these evolutionary relationships [120,130,134].In the meantime, the paraphyly of the genus Hepatozoon remains unsolved, and a denser taxon sampling and additional molecular markers are required [124,126,134].
Our results collectively reveal that the Hepatozoon genotypes found in Pampas foxes in Chile, crab-eating fox, and gray short-tailed opossum in Brazil could have been acquired by preying on infected snakes [137][138][139][140].However, rodents that frequently constitute the primary diet of both Pampas foxes [137,141] and crab-eating foxes [140] may act as paratenic hosts for Hepatozoon species infecting snakes [142].Moreover, crab-eating foxes also prey on frogs [143], which are also considered paratenic hosts for Hepatozoon spp. of reptiles [15].However, infective cystic stages have been found in rodents from Brazil, suggesting that they are paratenic hosts in the transmission of Hepatozoon towards predators [144,145].
A Hepatozoon genotype (KX776354) recovered from a Paraguayan fat-tailed mouse opossum (Thylamys macrurus) in Brazil [145] clustered into the clade of Hepatozoon spp.detected in rodents (Fig. 5, subclade of Brazilian rodents).This opossum species has been reported to be in syntopy with several rodent species, such as Oecomys mamorae [146], in which Hepatozoon has been also documented [136,145], a fact that would explain a common haplotype (H17) among rodents and opossums (Fig. 7,  Clade I).Therefore, it is likely that these small mammal species may share ectoparasites that could facilitate accidental cross-species infections [145].
Hepatozoon lineages found in South American cricetids (Rodentia: Cricetidae) and echimyids (Rodentia: Echimyidae) formed two clearly separated clades (Fig. 5, subclade sigmodontine rodents and subclade H).In this context, the Hepatozoon sp.infecting the Paraguayan punaré (Thrichomys pachyurus) appears to be a lineage specific to echimyids, as suggested for South American cricetid rodents [79,102].Likewise, the haplotype network supports the separation of Hepatozoon spp.from echimyids (Fig. 7, haplotypes H27 and H26).However, the inclusion of more Hepatozoon genotypes associated with South American echimyid rodents within this phylogenetic framework is needed to substantiate this phylogenetic hypothesis.
Regarding opossums, although the phylogeny does not relate Hepatozoon genotypes associated with the monito del monte (Dromiciops gliroides) and the big-eared opossum (Didelphis aurita) (Fig. 5, subclades G and I), the haplotype network places the haplotypes characterized from these opossum species at the same origin node (Fig. 7, haplotypes H29, H30, H31, H28, and H87).These findings suggest Hepatozoon lineages with a genetic structure associated with South American opossums.However, to further support this hypothesis, additional data are needed.
On the other hand, experimental studies have shown that rabbits (Lagomorpha) serve as paratenic hosts for Hepatozoon spp. that infect carnivores [44].But Allen et al. [59] proposed that some undescribed species of Hepatozoon may cycle in lagomorphs.However, the initial claim is supported by the close evolutionary relationships observed in Hepatozoon genotypes found in rabbits and felids (Fig. 6, subclade P), as well as the shared node of origin among rabbit and felid haplotypes (Fig. 7, Clade II, haplotypes H50, H48, H49, H33, H32, and H47).These associations could suggest that rabbits might serve as paratenic hosts for Hepatozoon spp.related to felids.
Regarding Hepatozoon genotypes and haplotypes of Didelphimorphia and Rodentia, their clustering into Clade II suggests that the sampled mammals came from regions endemic for canine hepatozoonosis; moreover, it Fig.7 Haplotype network inferred for a subset of Hepatozoon 18S rDNA sequences obtained from wild mammals in the Americas using 115 sequences and an alignment of 621 bp.For details of haplotypes of Hepatozoon circulating among wild hosts refer to Table S4 demonstrates the potential of these mammalian groups to act as paratenic hosts for Hepatozoon species that infect canids [148][149][150].Particularly, rodents have been implicated as vertebrate reservoirs of Hepatozoon in wildlife [15,147].Indeed, infective cystic stages in rodents facilitate the persistence of Hepatozoon spp. in the ecosystem [144,147], including H. americanum [45].This fact might account for the shared genotypes or haplotypes among canid and rodents, for both H. americanum (Fig. 6, subclade K; and Fig. 7, haplotype H62) and H. canis (Fig. 6, subclade L; and Fig. 7, haplotypes H53 and H55).Furthermore, it could suggest that rodents might also act as paratenic hosts for H. canis in Brazil [148].However, this hypothesis remains unclear [109,144].
Overall, our results reveal previously unreported associations between Hepatozoon genotypes in both distantly related taxa (Fig. 5, subclade D; and Fig. 6, subclades K, L, and P) and closely related mammalian groups (Fig. 5, subclades Sigmodontine and North American rodents), as well as in predator-prey relationships (Fig. 5, subclade D; and Fig. 6, subclades K, L, and P).This wide host spectrum suggests that the diversity and biogeographical patterns of Hepatozoon spp. in wild mammals across the Americas are more complex than currently understood.In this sense, the studies of Di Cataldo et al. [88] and Weck et al. [136] may not accurately reflect the relationships among Hepatozoon genotypes found in foxes in Chile and fat-tailed mouse opossum in Brazil due to lack of data in their phylogenetic framework.Therefore, incorporating more Hepatozoon sequences into an alignment could support with more confidence the monophyly of new genotypes and clarify their evolutionary relationships and ecological associations, as our analyses demonstrate.

Factors driving the diversity and transmission of Hepatozoon spp. in mammals of the Americas
The genetic diversity of a given species is modulated by multiple processes that include mutation, recombination, and biodemography [151,152].For Hepatozoon spp., the life cycle, transmission dynamics, and dispersion capacity are factors that shape their diversity as well [38].Moreover, the transmission of Hepatozoon species may be facilitated by syntopy of hosts sharing ectoparasites [145], the distribution of suitable vectors [52,62,63,153], and through ingestion of infective cystozoites by carnivorism [44,45,142] or ectoparasites attached to prey [18,45].However, it seems that host specificity and food webs play a crucial role in the transmission of Hepatozoon species, as corroborated by previous studies [147,154].
Small mammals and herpetozoa are typical prey of carnivores, and those harboring cysts are likely to transmit Hepatozoon spp. to their predators.This ecological pattern is shown in Fig. 5, where occasional infections of canids with snake-related genotypes of Hepatozoon can be observed (subclade D), suggesting a low host specificity in certain Hepatozoon spp.However, given that vector capacity-defined as the daily rate of effective infections spread by a specific arthropod population and influenced by vector density, behavior, longevity, vectorhost encounter rates, and vector competence-varies among invertebrate hosts [155][156][157][158], predators can be considered dead-end hosts for these Hepatozoon spp., thus affecting the transmission dynamics.In this context, predators may contribute to reducing the transmission of the parasite (dilution effect) [41,147].Nevertheless, these infections may affect the immune response of predators, potentially increasing their susceptibility to other infectious agents [93,159,160].
In addition, Hepatozoon spp.exhibit greater host specificity in invertebrates (definitive host) than in vertebrates (intermediate host) [15,17,123].Therefore, the degree of specificity that an invertebrate parasite shows to its vertebrate host will define the chance for a given Hepatozoon sp. to find and colonize a suitable vertebrate host [157,161].Colonization may be affected by the immune response of the vertebrate hosts [8].Thus, considering that the composition of parasite communities is primarily structured by host species (both intermediated and definitive), phylogenetically related hosts are more likely to share parasite species since they exhibit similar immunological pathways and ecological and evolutionary processes [8,162].These interactions with invertebrate and vertebrate hosts play a significant role in the diversification and dissemination of Hepatozoon spp.
Some studies have proposed that Hepatozoon genotypes exhibit close phylogenetic relationships and a genetic structure according to the vertebrate groups that they parasitize [109,144,163].Notably, rodent-associated Hepatozoon spp.seem to be specific, in contrast to those species that infect reptiles [164].Likewise, studies have confirmed a degree of genetic diversity in Hepatozoon spp.infecting rodents [144,145].Although some genotypes of Hepatozoon found in rodents are shared with reptiles, they are considered to be Hepatozoon spp. of reptiles using rodents as paratenic hosts within their life cycle [142].
In particular, the findings in Chile suggest discrete lineages of Hepatozoon spp.associated with the native rodent genera Abrothrix, Oligoryzomys, and Phyllotis [75,79,102,132], and with the ancient marsupial monito del monte (D. gliroides) [75], suggesting that Hepatozoon co-evolved with these mammals [102,165,166].Thus, the evolutionary history and diversification dynamics of these hosts could be shaping the phylogenetic relationships and genetic structure of the Hepatozoon lineages characterized in Chile [166].However, a denser sampling across hosts and the inclusion of both 18Sr DNA and cox1 sequences in the genetic analyses are needed to support this hypothesis.
The knowledge of the Hepatozoon genotypes and haplotypes circulating among wildlife mammals in the Americas provides valuable insight into the epidemiology of this hemoparasite, shedding light on exposed or susceptible hosts.Also, it facilitates tracking the spread and occurrence of unique haplotypes of Hepatozoon among mammalian groups (Fig. 7), especially in threatened host species.Although a greater diversity of Hepatozoon spp. is found in canids and rodents, this may reflect a sampling bias.Overall, these results reveal a significant lack of data and highlight the need for a comprehensive sampling of less prospected mammals in the Americas.

Would Hepatozoon spp. pose a risk to mammals of the Americas?
Infectious diseases pose a significant threat to wildlife, leading to population decline, biodiversity loss, ecological disruptions, and an increased risk of disease transmission [11,12,167].Therefore, studying infectious diseases in wildlife is crucial for the conservation of animal biodiversity and human health, particularly in the context of global climate change [3,168].In the Americas between 1916 and 2022, Hepatozoon spp.have been reported in 106 species of mammals, belonging to 62 genera, 18 families, and nine orders (Table S1).This fact highlights the broad range of susceptible mammals and the potential impact on health that Hepatozoon spp. in the wildlife of the Americas could pose.
Based on our findings, H. canis and H. felis emerge as the most widespread species among wild mammals in the Americas.Hepatozoon canis was identified in 13 wild mammal species, infecting 133/657 individuals (20.24%), while H. felis was found in 12 species, infecting 132/475 individuals (27.97%) (Additional file 1: Table S1).It is recognized that H. canis infection in canids might persist sub-clinically, and its severity can range from mild to life-threatening [16,53,66,83], while H. felis seems well suited to felid hosts [68].However, IFs of H. canis and H. felis among non-canid and non-felid hosts (Additional file 1: Table S1) raise concerns and warrant further investigation because of the potential risks they might pose to other mammalian groups, particularly those with conservation threats [169].
Remarkably, H. americanum exhibited IFs of 44.09% (41/93 individuals) across three wild mammal species, while H. americanum-like showed IFs of 11.43% (44/385 individuals) in four canid species (Additional file 1: Table S1).The first species is considered the primary cause of American canine hepatozoonosis (ACH) in North America [16,38,66,170].Meanwhile, the latter is a species closely related to H. americanum, and is emerging in South American canids [13,67,171].Although H. americanum was found infecting only two mammal species in North America, it showed higher IFs.
While H. americanum and H. americanum-like were found in one North American and four South American canid species, respectively, H. canis was found in nine species of canids, including three in North America and six in South America (Additional file 1: Table S1).Additionally, H. felis was found in three canid species and six felid species across the Americas (Additional file 1: Table S1).Further assessment is necessary to understand the spread and potential impact of these infections among canid and felid populations, given that carnivores play an essential role in maintaining balance of the ecosystems as predators within the food webs, and are threatened species, [172][173][174][175][176][177].
The high occurrence of Hepatozoon infection in both North American and South American canids can be linked to their presence in endemic areas for canine hepatozoonosis [18,58,59,144,153].It is worth mentioning that some canids, such as the coyote (Canis latrans) and crab-eating fox, are expanding beyond their natural ranges [178].This fact, coupled with crossbreeding events with domestic dogs [178,179], poses a significant risk for spillover of parasites between domestic animals and wildlife, as well as the emergence of new endemic canine hepatozoonosis foci [180].
Indeed, coyotes are commonly reported to be infected with H. americanum in North America [87,153,[181][182][183][184], while crab-eating foxes are associated with H. canis and H. americanum-like infections in South American ecosystems [67,171,185,186].Both H. americanum and H. canis are recognized for their virulence in canids [16,66,84], as well as for their diverse modes of transmission between hosts and ticks, that are common ectoparasites of canids in the Americas [52,63,187].
Notably, Hepatozoon was primarily observed in Carnivora (42.12%; 730/1733); however, Rodentia exhibited the highest number of species (62.26%; 66/106) infected with Hepatozoon (Additional file 1: Table S1).This can be attributed to the diverse ecological traits of rodents, such as their fast life pace, terrestriality, high population densities, various activity cycles, and diet breadth [188].Likewise, being one of the most geographically widespread, diverse, and abundant mammalian orders, Rodentia influences parasite richness and transmission [3,21,189].Our findings highlight the importance of both carnivores and rodents to understanding the epizootiology and transmission of Hepatozoon spp.Moreover, they suggest that rodents are key in maintaining Hepatozoon spp. in the ecosystems of the American continent.
Despite the fact that certain etiological agents in wildlife may not cause diseases in their hosts due to a harmonious parasite-host relationship (co-evolution) [190,191], it is worth mentioning that while Hepatozoon infections in wildlife are often subclinical, they may vary from mild to severe [67,182,183,192].Indeed, these hemoparasites become pathogenic and opportunistic in immunocompromised individuals, exhibiting high virulence in concurrent infections, thereby increasing susceptibility to other vector-borne agents [66,181,[193][194][195].Furthermore, spillover of Hepatozoon spp. in atypical hosts may lead to infections more virulent than those observed in natural hosts [196][197][198].
Hepatozoon transmission in vertebrates primarily occurs through the ingestion of an infected bloodsucking invertebrate.However, the second merogony-a phase of asexual reproduction producing cystozoites-seems pivotal for some Hepatozoon spp., as cystozoites are the infectious forms leveraging transmission through predation, thereby facilitating its spread across food webs [15,42].Rodents often act as paratenic hosts in this mode of infection [15,147,207,208].In fact, studies indicate that carnivores may acquire Hepatozoon infection by preying on rodents [147,207].Nevertheless, the role of rodents in the epidemiology of Hepatozoon species detected in carnivores is still unclear [145,209].
A pre-existing infection can aid the establishment of a new infectious agent which otherwise might have been cleared by the host's immune system; consequently, the acquisition of a novel infectious agent may promote the dissemination of an existing latent or dormant infection in the host, acting synergistically to enhance pathogenic processes, parasite transmission, and disease severity [93,159,160].
Hepatozoon infection in mammals of the Americas requires comprehensive attention, considering factors such as the availability of suitable arthropod vectors, the prevalence of infection in intermediate hosts, and the susceptibility of different mammal species.In addition, it is important to recognize other threats to wildlife in the Americas, including habitat loss (by urbanization, livestock, and farming), pollution, climate change, and tick-borne diseases [156,211].The conjunction of these factors may influence the spillover of Hepatozoon infections among wild and domestic animals.To date, only one study has monitored the fluctuation and assessed the impact of Hepatozoon infection on health in wildlife of the Americas [89], for which additional studies of this kind are necessary.
Given that parasites pose a risk to species susceptible to extinction [212], detecting and monitoring Hepatozoon spp. in wildlife becomes necessary.Genetic screening can reveal the patterns of dispersion of Hepatozoon lineages among hosts, enriching our knowledge of their roles in the Hepatozoon epizootiology [169,212,213].It is crucial for further research and surveillance to comprehensively understand the dynamics of Hepatozoon infections in mammal populations of the Americas and assess their potential as a significant threat [88].

Conclusions and future perspectives
This review provides valuable insight into the distribution of Hepatozoon among mammalian hosts and potential vectors in the Americas, establishing a foundation for subsequent research.Notably, this review represents the first comprehensive summary of Hepatozoon infection in wild mammals in the region.However, numerous questions remain unanswered, particularly regarding the impact of hemoparasites on the health of wild mammals in the Americas.
Given the expanding distribution of certain canids, combined with hybridization events between some species and the diverse transmission modes of Hepatozoon, there is a risk for spillover between wildlife and domestic animals.This could also lead to the emergence of new areas with endemic foci for canine and felid hepatozoonosis.Canid species such as coyotes, crab-eating foxes, and Pampas foxes may serve as effective sentinels to track the expansion of H. americanum, H. canis, H. americanum-like, and H. felis in ecosystems across the American continent.Finally, genomic data and new molecular markers are urgently needed for the implementation of effective strategies to detect, control, and manage Hepatozoon infections.

Fig. 1
Fig. 1 Flowchart of the current systematic review on Hepatozoon infections in wild mammals and their ectoparasites across the Americas (constructed with draw.iov. 22.0.3;https:// www.drawio.com/)

Fig. 2
Fig. 2 Temporal patterns of published articles regarding Hepatozoon organized by mammal orders in the American Continent

Fig. 3
Fig. 3 Map of the American Continent showing the infection rates (IRs) of Hepatozoon by mammalian order (A), and occurrence of Hepatozoon spp. in American country (B).Maps were constructed with Quantum Geographic Information System (QGIS) v. 3.18.1-Zürich(https:// www.gnu.org/ licen ses)

Fig. 4
Fig.4 Primers and flanked regions (V1 to V9) used for the molecular diagnosis and genotyping of Hepatozoon in wild mammals in the Americas.The position of each primer was aligned with the 18S rDNA complete reference sequence of Hepatozoon canis detected in a domestic dog from Israel (MH615006, 1,816 bp; shown in black).For details on the primers, refer to TableS3