Hepatozoon pyramidumi sp. n. (Apicomplexa: Adeleorina) from the blood of Echis pyramidum : morphology and SSU rDNA sequence

azema1@yahoo.com Abstract Hepatozoon pyramidumi sp. n. is described from the blood of the Egyptian saw-scaled viper, Echis pyramidum , captured from Saudi Arabia. Five out of ten viper specimens examined (50%) were found infected with Hepatozoon pyramidumi sp. n. with parasitaemia level ranged from 20-30%. The infection was restricted only to the erythrocytes. Two morphologically different forms of intraerythrocytic stages were observed; small and mature gamonts. The small ganomt with average size of 10.7 × 3.5 μm. Mature gamont was sausage-shaped with recurved poles measuring 16.3 × 4.2 μm in average size. Infected erythrocytes were hypertrophied; their nuclei were deformed and sometimes displaced from their central position in the normal uninfected cell. Merogonic stages were observed in the lung endothelial cell and the liver parenchyma cells. Mature meront was 17.8 × 13.6 µm and contained banana-shaped merozoites with average size of ~15 × 2 µm. Phylogenetic analysis based on the SSU rDNA sequence clustered Hepatozoon pyramidumi sp. n with previously sequenced Hepatozoon spp., most of them infected reptilian hosts without geographic consideration. The morphological and molecular comparison with closely related species proved the taxonomic uniqueness and novelty of the present haemogregarine parasites. Two of highly infected vipers were and small parts of lung, spleen and excised fixed in 10% of neutral buffered formalin. Fixed tissues were processed as usual for normal histology then, sectioned and stained with hematoxylin and eosin (H & E). The blood smears and histological sections were examined and photographed using an Olympus BX51 photomicroscope. Measurements were taken using calibrated ocular micrometer with an oil-immersion lens. All measurements are here given in micrometres; data are presented as mean ± SD (when it possible) followed by rang in parentheses.


Introduction
Genus Hepatozoon Miller 1908 is an assemblage of apicomplexan blood parasites that have been described from all members of vertebrates for more than a century. Hepatozoon has heteroxenous life cycles with merogonic and gamontogonic development within the vertebrate host, and gamogony and sporogony within the invertebrate host (Telford, 2009;Roca & Galdón, 2010). Members of the genus Hepatozoon, along with those of genera Haemogregarina, Karyolysus, Cyrilia, Hemolivia and Desseria are collectively known as "haemogregarines" (Smith, 1996). All haemogregarines shared in the characteristic of gamontogonic development in either erythrocytes or leucocytes of their intermediate vertebrate hosts. Therefore, it is very difficult to identify the generic level based exclusively on the morphological characteristics of gamont stages (Abdel-Baki & Al-Quraishy, 2012). Accordingly, Mihalca et al. (2002) suggested that the generic identification should be based on the characteristics of both gamont and sporogonic stages of its life cycle. Unfortunately, the complete life cycle and mode of transmission are lacking for the majority of haemogregarines (Abdel-Baki et al., 2014). In this regard, molecular techniques are notably helpful tools for both diagnostic and phylogenetic purposes and, as such, it has become firmly established to assign the accurate generic affiliation of haemogregarines (Abdel-Haleem et al., 2018). Using this attitude, here we tend to describe new species of Hepatozoon based on both the morphometry of gamont and merogonic stages, and the partial 18S rDNA sequence from the peripheral blood of Echis pyramidum collected from Saudi Arabia.

Materials and Methods
Ten Egyptian saw-scaled viper, Echis pyramidum (Geoffroy Saint-Hilaire, 1827) were collected from Jizan (16°53' 21" N; 42°32' 3" E), (29° 22' N 30° 37'E), Saudi Arabia in Summer 2018. Thin blood smears were made by stab the tip of the tail of each viper then, were fixed with methanol and stained with Giemsa and finally microscopically examined for haemogregarine parasites. Two of highly infected vipers were dissected and small parts of lung, liver, spleen and kidney were excised and fixed in 10% of neutral buffered formalin. Fixed tissues were processed as usual for normal histology then, sectioned and stained with hematoxylin and eosin (H & E). The blood smears and histological sections were examined and photographed using an Olympus BX51 photomicroscope. Measurements were taken using calibrated ocular micrometer with an oil-immersion lens. All measurements are here given in micrometres; data are presented as mean ± SD (when it possible) followed by rang in parentheses.

Genetic characterization
To genetically characterize the present haemogregarine, polymerase chain reaction (PCR) amplification using both the BTF1 (Megía-Palma et al., 2014) and the HEP1600R primers (Ujvari et al., 2004) was performed. The PCR protocol was as previously described by Abdel-Haleem et al. (2018). Briefly, 3ul were prepared by adding 50-100 ng of genomic DNA, 1 mM of each primer, 1.5 mM MgCl2, 2 mM of dNTPs and 0.5U of iProof™ High-Fidelity DNA (Bio-Rad, Hercules, CA, USA). For PCR amplification, after initial denaturation at 95 °C for 4 mn, 35 cycles of three steps were performed (96 °C for 20 s, 56 °C for 20 s and 72 °C 90 s). The program ends after final extension at 72 °C for 10 mn). A negative control was used for each reaction to check for possible DNA contamination. All PCR products were check by electrophoresis in 1% gel agarose in TBE buffer 0.5X and stained with Ethidium bromide 0.5 μg/mL and visualized in UV-transilluminator workstation Gel Doc System (Biorad). The amplified PCR products were purified using A ExoSAP-IT™ (USB, Afymetrix) treatment and sequenced by an external commercial sequencing company (Macrogen Inc., Seoul, South Korea). The obtained sequence was assembled, and edited manually to resolve the ambiguous sites with aid of BioEdit version 7.2.5 (Hall, 1999). The generated sequence was compared with related published 18S rDNA sequences available in GenBank using the BLASTn program to confirm the generic identification and sequence was then submitted to GenBank (accession number: MT025290). For phylogenetic analysis, thirty sequences of Hepatozoon species in addition to sequences of Sarcocystis lacertae (used as an outgroup) were retrieved from GenBank to be used for multiple sequence alignment using ClustalX 2.1.0.12 software (Larkin et al., 2007) with default parameters. Both of Maximum likelihood (ML) and Bayesian inference (BI) were utilized for phylogenetic analyses. For generating of ML trees; the General Time Reversible model with gamma-distributed rates was employed and invariant sites (GTR + I + G) estimated using jModelTest 2.1.1 (Darriba et al., 2012). Constructing the consensus tree was carried out in the software MEGA version 7 (Kumar et al., 2016). Bootstrap technique, with 1,000 replicates was employed to estimate the support for nodes. The evolutionary relationships between sequences were assessed using Kimura 2-parameter model (Kimura, 1980). The BI was accomplished using Mr-Bayes v3.2.5 program (Ronquist et al., 2012). The analysis was run for 2 ×10 6 generations with one tree saved every 100 generations. The posterior probability of the trees was determined utilizing Metropolis-Coupled Markov Chain Monte Carlo (MCMC) algorithm running four chains (nchains = 4). In every Bayesian run, the first 25% of the saved trees were discarded as burn-in prior to reaching stationary. The consensus tree was constructed with an average standard deviation of split frequencies of 0.008.

Results
Out of 10 snakes examined, 5 were found positive for haemogregarines with an overall prevalence of (50%). The parasitaemia level (number of infected erythrocytes per 1000 examined) in the infected snakes ranged from 20-30%.

Molecular and phylogenetic study
A consensus partial sequence of 1389 bp of the SSU rRNA was generated and deposited in GenBank (accession number MT025290). The BLAST results of the obtained sequence displayed nucleotide sequence similarities ranged from 93.34 to 99.7% with other known Hepatozoon sequences from various geographic localities. The highest sequence identity of 99.7% was to an unnamed Hepatozoon sp. (GenBank entry KJ574012) parasite of Cerastes cerastes from Egypt and Hepatozoon sp. (GenBank entry HQ734790) parasite of Ptyodactylus oudrii from Algeria. Phylogenetic based on maximum likelihood and Bayesian inference, yielded a similar topology and cluster H. pyramidumi tree with other Hepatozoon from various geographic area but most of them infect reptilian hosts. Outside this main cluster, two other clusters were observed, one grouping Hepatozoon spp. from mammals and the second one grouping Hepatozoon spp. infecting amphibian host (Figure 3).

Site of infection:
Gamonts in the erythrocytes and merogonic stages in the lung endothelial cells and the liver parenchyma cells.
Type-material: Two slides with blood smears are deposited at the Museum of the Zoology Department, College of Science, King Saud University, Saudi Arabia (Hep 01-2020).

Sequence accession numbers: MT025290.
Etymology: The specific epithet is given after the host name.

Discussion
Haemogregarines are diversified group of intracellular apicomplexan parasites that are known to infect a wide variety of vertebrate hosts. For a long time, the identification of haemogregarine species was principally based solely on morphological and/or biological features of gamont stages in erythrocytes of the host (Perkins & Keller, 2001). However, morphological comparison alone is not quite enough for distinguishing the closely related species, because of the low host specificity and pliability of gamonts morphology (Telford, 2010;Harris et al., 2011). More recently, molecular approach together with morphological data are strongly suggested as prerequisite for differential diagnoses and taxonomy of haemogregarines (O'Donoghue, 2017). Within hemogregarines, Hepatozoon is the most commonly described and the most speciose genus (Harris et al., 2018). Snakes have been proven worldwide as hosts of more than 145 Hepatozoon species (Úngari et al., 2018). However most of these species descriptions were primarily based alone on the morphology of the intraerythrocytic gamont, as the life cycle and mode of transmission of these parasites were unidentified (Tomé et al., 2016;Abdel-Haleem et al., 2018). Currently, molecular attitudes based on 18S rRNA gene sequence variations became extremely helpful complementary tool for differentiating the closely related species of the genus Hepatozoon (Maia et al., 2016;Tomé et al., 2016). Consequently, it is become widely accepted that the morphological characteristics of intraerythrocytic stages combined with the molecular are quite enough to establish the novelty the Hepatozoon species (Abdel-Haleem et al., 2018). In present study, therefore, H. pyramidumi sp. n. was identified through morphological comparison to Hepatozoon spp. of similar host genus and locality and to species that are phylogenetically assembled in the same clade. So far, two species of Hepatozoon have been described from snakes of the genus Echis. These species are: H. echisi described from saw-scaled viper Echis carinatus in West Pakistan (Mohiuddin et al., 1967) and H. mehlhorni described from the same host (E. carinatus) in western Egypt (Bashtar et al., 1991). H. echisi easily differentiate from the present species in having shorter and thinner gamonts (13.97 × 2.8 vs. 16.3 × 4.2). As well as the Merogony of H. echisi was restricted to lung endothelial cells while the present species occurred in both lung endothelial cells and liver parenchyma. Meanwhile, H. mehlhorni differs in having larger gamont (17.2 × 5.4 vs 16.3 × 4.2) and dimorphic meronts. The merogonic pattern of H. mehlhorni is also different, with meronts apparently dimorphic producing up to 38 merozoites measuring 6 × 1.5 µm in size compared to monomorphic meronts producing up to 15 merozoites 15 × 2 µm in the present species. Phylogenetic tree analysis revealed that the 18S rRNA sequences of E. pyramidum clustered together with Hepatozoon spp. from reptiles ( Figure 3). Hepatozoon pyramidumi n. sp. showed highest sequence identity (99.7%) to Hepatozoon sp. (KJ574012) and Hepatozoon sp. (HQ734790) which still unidentified to the species level and with no available morphological data that precluded the further morphological comparison. Of the other clustered species; the morphological and morphometric data are available only for H. domerguei Landau, Chabaud, Michel, Brygoo, 1970 from Madagascarophis colubrinus in Madagascar, H. garnhami Abdel-Baki, Al-Quraishy, Zhang, 2014 from Psammophis schokari in Saudi Arabia, H. chinensis Han, Wu, Dong, Zhu, Li, Zhao, Wu, Pei, Wang, Huang, 2015 from Elaphe carinata in China and H. aegypti Bashtar, Boulos, Mehlhorn, 1984 from Spalerosophis diadema in Egypt which give the opportunity for further comparison (Maia et al., 2014;Abdel-Baki et al., 2014;Han et al., 2015;Abdel-Haleem et al., 2018). In this regard, H. domerguei differs in having shorter and thinner encapsulated gamonts (13.8 × 3 vs 16.3 × 4.2). In addition, the merogonic cycle of H. domerguei produce dimorphic meronts compared to monomorphic ones in the present species. H. garnhami can be easily differentiated from the present species by its quite thinner gamonts (2 vs 4.2) and its merogonic cycle that was restricted to lung with dimorphic meronts. In the same way, the thinner gamonts that was folded back in a hook-wise at one end separated H. chinensis from the present species. Finally, H. aegypti differs in having shorter gamonts (14 vs 16.3) with merogonic cycle apparently restricted to the endothelial cells of lung.
These morphometric comparisons clearly distinguished the species investigated here from the closely related ones. As well as molecular data confirmed the classification of the present Hepatozoon as a new species.