Pan-American Trypanosoma (Megatrypanum) perronei sp. n. in white-tailed deer (Odocoileus virginianus) and its deer ked Lipoptena mazamae: morphological, developmental and phylogeographical characterisation CURRENT STATUS:

Background: The subgenus Megatrypanum comprises trypanosomes of cervids and bovids from around the world. Here, Odocoileus virginianus (white-tailed deer = WTD) and its ectoparasite, the deer ked Lipoptena mazamae (hippoboscid fly), were surveyed for trypanosomes in Venezuela. Results: Haemoculturing unveiled 20% infected WTD, while 47% (7/15) of blood samples and 38% (11/29) of ked guts tested positive for the Megatrypanum- specific TthCATL-PCR. CATL and SSU rRNA sequences uncovered a single species of trypanosome. Phylogeny based on SSU rRNA and gGAPDH sequences tightly cluster WTD trypanosomes from Venezuela and the USA, which were strongly supported as geographical variants of the herein described Trypanosoma ( Megatrypanum ) perronei sp. n. In our analyses, T. perronei was closest to T . sp. D30 of fallow deer (Germany), both nested into TthII alongside other trypanosomes of cervids (North American elks and European fallow, red and sika deer), and bovids (cattle, antelopes and sheep). Insights into T. perronei life cycle were obtained from early haemocultures of deer blood and co-culture with mammalian and insect cells showing flagellates resembling Megatrypanum trypanosomes previously reported in deer blood, and deer ked guts. For the first time, a trypanosome from a cervid was cultured and phylogenetically and morphologically (light and electron microscopy) characterised . Conclusions: In the analyses based on SSU rRNA, gGAPDH, CATL and ITS rDNA sequences, neither cervids nor bovids trypanosomes were monophyletic but intertwined within TthI and TthII major phylogenetic lineages. One host species can harbour more than one species/genotype of trypanosomes, but each trypanosome species/genotype was found in a single host species or in phylogenetically related hosts. Molecular evidence that L. mazamae may transmit phylogenetic relationships, range of vertebrate hosts, vectors, and geographical distribution. In the present study, we describe a new species of Megatrypanum in WTD and its deer ked and hypothesise its probable life cycle and evolutionary history by integrating morphological, culture behaviour, biological, and phylogeographical data. rRNA and gGAPDH genes support the independent species status for the new trypanosome from Venezuelan white-tailed deer in the subgenus Megatrypanum trypanosomes sitatunga,

livestock breeding area at north-eastern Venezuela (N10°07′08.95'', W64°38′23.80'') ( Fig. 1), where the mean annual temperature is 25°C (21.8°C -32.2°C) and rainfall is below 1000 mm, with a dry season from December to April. The captured WTDs (Fig. 2) were kept in quarantine prior to their introduction into a protected reserve. Chemical immobilization and anaesthesia were performed with a combination of 10 mg/kg ketamine and 0.6 mg/kg xylazine intramuscularly and reverted with 0.2 mg/kg of yohimbine intravenously. Anesthetized deer were examined to record heart and breathing frequency rates and temperature. Institutional and National guidelines for the care and use of wild animals were followed. Deer handling were performed in accordance with the protocol approved by the MINEC from Venezuela 2013-154125.
Blood samples were obtained via jugular vein using tubes with EDTA and aliquots (~ 500 µl) of blood were preserved in 99.5% ethanol (v/v), incubated overnight in 250-500 μl of lysis buffer (20mM EDTA; 50mM Tris-HCL; 117mM NaCl; 1% SDS; 10 mg/ml Proteinase K), precipitated with 400-800 μl ammonium acetate (4M), and centrifuged (10 min at 20,000 g). Then, DNA was precipitated with ethanol, dried at room temperature and resuspended in TE (Tris-EDTA). Deer keds (Fig. 2) were removed from the WTDs, preserved in ethanol, and identified by morphological parameters [39], and COI barcoding using DNA obtained from the gut contents of keds as described previously for tsetse flies [40]. Due to limited field facilities, the guts of deer keds were not examined by microscopy or culturing. decomposition using the Neighbor-Net method with Kimura 2-parameter implemented in Splits Tree4 V4.10 [48]. Internode support was estimated by 500 bootstrap replicates using the same parameters optimized for network inferences.
Trypanosomes included in our analyses, and their host species, geographical origins and GenBank accession numbers of DNA sequences are detailed in the Supplementary Tables S1-S4 (S1, CATL sequences; S2, SSU rRNA; S3, gGAPDH; S4, ITS1 rDNA). All DNA sequences here determined were deposited in GenBank.

Growth behaviour and light and electron microscopy of trypanosomes co-cultivated with insect and mammalian cells
The isolate TCC2268 was co-cultivated with a monolayer of Hi-5 insect cells (TC100 medium) and mean of 37.4 °C). Inspection of WTDs (Fig. 2) for ectoparasites revealed an abundance of deer keds iñ 73% (11/15) of them, mainly in the ventral, inner legs, and inguinal regions of the WTDs. No skin or fur damage was observed in ked-infested WTDs. Twenty-nine keds preserved in ethanol were identified morphologically as Lipoptena mazamae (Fig. 2) [39] and confirmed by COI DNA barcoding (GenBank: MN756795).
Microscopy of Giemsa-stained blood smears and the microhematocrit technique were unable to detect trypanosomes in blood of the 15 WTDs examined. However, haemoculturing (Fig. 2)  Megatrypanum-specific assay, TthCATL-PCR, was employed aiming at the detection of trypanosomes in blood samples from WTD and in gut contents from the deer keds. PCRs were positive for 7 out of 15 WTDs (~47%), and 11 out of 29 deer keds (~38%), including keds from the three WTD positive by haemoculturing.
Characterisation of trypanosomes from blood and keds of white-tailed deer using CATL sequences CATL DNA sequences (274 bp) obtained by TthCATL-PCR from blood samples and culture (isolate TCC2268) of WTD and deer keds share highest similarity with sequences of Megatrypanum trypanosomes. Sequences herein determined were aligned with those of other Megatrypanum trypanosomes (GenBank) and used to infer their genetic relatedness. Trypanosome sequences obtained from deer keds were virtually identical to those obtained from WTD blood (TCC2268), indicating that they belong to a single trypanosome species.
Network inferences of 100 CATL sequences of 274 bp, amplified by TthCATL-PCR (Fig. 3) or 53 larger CATL sequences of 477 bp (data not shown) generated highly similar networks. A divergence of 2.0% in the small fragment of CATL sequences separated TCC2268 from T. sp. D30. Even though originating from different continents, these two deer isolates were much more related to each other than to trypanosomes from bovids (cattle, buffalo, sheep and antelopes), even those from Venezuelan cattle and buffalo [25]. TCC2268 and T. sp. D30 were assigned to close but different genotypes, TthII H (which includes CATL sequences from the deer keds) and TthII C, respectively. TCC2268 diverged by 5.3% from T. melophagium (TthII D) and by 3.3% from T. theileri of cattle nested into TthII (Fig.   3B).
SSU rRNA sequences of the Venezuelan isolate WTD TCC2268 was highly similar (99.7%) to that of the isolate WTD A3 from the USA, both sharing 99.5% similarity with T. sp. D30 of fallow deer from Germany. Besides these three deer trypanosomes, TthII comprised Trypanosoma cf. cervi from North American elk (elk328), and isolates of red deer (Cel34), fallow deer (DdP18) and sika deer (Cn1) from Poland. In addition, TthII included trypanosomes from African antelopes and T. melophagium of sheep, all clustering close to the trypanosomes of deer, whereas trypanosomes of cattle from North and South America, Europe and Asia formed a more separated cluster (Fig. 4). Trypanosomes from these European deer nested into TthII diverged from TCC2268 by 0.3%-0.6% in highly conserved SSU rRNA sequences. For comparison, TCC2268 diverged by just 0.6% from the reference T. theileri TREU124 of cattle, thus reinforcing that these sequences are too conserved to clearly resolve the relationships of The lineage TthI also harboured trypanosomes (misclassified) referred to as Trypanosoma cf. cervi identified in WTDs (WTD A1, A5, A21, A148, NL15) and elks (elk142, elk328, elk416, elk421) from the USA [12], and isolates referred to as T. cervi from Poland red deer (Cel14St), fallow deer (DdP287) and tabanids (Fig. 4). The trypanosome (TSD1) from the Japanese sika deer [16] is the deer trypanosome more genetically distant (~1.4% sequence divergence) from the isolate TCC2268 (Fig.   4). In addition, this lineage also comprises T. theileri of cattle from South America, Japan and the USA, It was previously demonstrated that gGAPDH sequences are generally more variable than SSU rRNA sequences, thus being more suitable for a better differentiation between closely related trypanosomes such as those of the subgenus Megatrypanum [15,24,25]. Corroborating results using SSU rRNA sequences, gGAPDH sequence from the isolate WTD TCC2268 was closest to T. sp. D30 (~2.2% sequence divergence), and more similar to T. melophagium and trypanosomes of antelopes, all clustering together. Cattle isolates formed other group within TthII, separated from WTD TCC2268 by an average of ~ 4.4% sequence divergence.
Here, gGAPDH sequences of TCC2268 were compared with available sequences from species/genotypes of Megatrypanum and concatenated with SSU rRNA sequences. The inferred phylogeny (Fig. 5) displayed a highly congruent topology compared with those of independent SSU rRNA ( Fig. 4) or gGAPDH sequences (data not shown). Together, phylogenetic positioning and genetic distances of WTD TCC2268 from other phylogenetically validated species of Megatrypanum allowed for the description of this new isolate as a novel species herein designated Trypanosoma (Megatrypanum) perronei sp. n. Unfortunately, most trypanosomes from cervids have been known merely by partial SSU rRNA sequences such as those reported from elk and WTD from the USA, and deer isolates from Japan and Poland included in our previous SSU rRNA analysis (Fig. 4). Results obtained in the present study are congruent with previous studies using SSU rRNA, gGAPDH, CATL and ITS rDNA sequences [15,24,25,42,50,53].
Confirming previous studies, ITS1 rDNA sequences of Megatrypanum trypanosomes were distributed in the TthI and TthII lineages (Fig. 6) regardless of originating from bovids or cervids; thus, corroborating data from this (Figs 3-5) and previous studies [12,15,24,25,53]. Our analysis supported 6 genotypes of TthI (IA-IF) and 9 of TthII (IIA-II I) (Fig. 6). ITS1 rDNA sequences of T. perronei from Venezuela and the USA shared highly similar sequences (~0.6% divergence), and were tightly clustered together supporting the genotype TthII H. Trypanosomes from European cervids assigned to the lineage TthII C were the German T. sp. D30 of fallow deer [12,25] that clustered with the highly similar Croatian trypanosome of red deer [15], and they diverged by ~17% from T. perronei.
Regarding the relationship between cervid and bovid trypanosomes, ITS1 rDNA sequences of T. melophagium of sheep diverged only by ~18% from T. perronei while divergences above 24% separated T. perronei from trypanosomes of cattle (TthII A and TthII B) and African antelopes (TthII F, Interestingly, ITS1 rDNA sequence of a trypanosome obtained from the gut of an Italian sand fly [54] was virtually identical to those of T. sp. D30 from Germany [22]. As expected, deer trypanosomes of TthI diverged from T. perronei by remarkable divergences in ITS rDNA (Fig. 6): ~46% from both TthI D (WTD and elk) and TthI E (elk), which are genotypes identified in deer sympatric with the WTD from which T. perronei sp. n. (isolate WTD A3) was obtained in the USA [12]. As we showed with other molecular markers, the greatest distances of ITS rDNA (~49%) among deer trypanosomes was between T. perronei and the trypanosome from Japanese sika deer (TthI F).

Host-parasite-vector relationships and evolutionary history of cervid trypanosomes
Although host specificity of Megatrypanum trypanosomes remains to be clearly demonstrated, our findings provide additional support for relevant host-parasite-vector association in the evolution of these trypanosomes. Species diversification was likely shaped by evolutionary constraints exerted by ruminant hosts. In addition, vectors may be also involved in trypanosome host-restriction because deer flies (tabanids) and deer keds (hippoboscids) are highly associated to their cervid hosts. In agreement with this hypothetical evolutionary scenario, we demonstrated that deer keds taken from WTD exclusively harboured T. perronei sp. n., corroborating a previous suggestion that these flies can transmit, cyclically and/or mechanically, trypanosomes to cervids [38]. L. mazamae occurs from south-eastern USA to South America [39,55], and is tightly linked to WTD, although this ked can eventually jump to phylogenetically close deer species [39]. Deer flies, which have been experimentally proven to transmit deer trypanosomes, and deer keds have been implicated as vectors of cervid trypanosomes [21,29,31,36,38,56,57]. Recent studies report on DNA from deer trypanosomes in guts of sand flies and culicids [54,58], but their roles as vectors remains to be investigated.
perronei sp. n. while sympatric cattle and water buffalo were found infected with T. theileri and T.
Deer, cattle and sheep harboured host-specific trypanosomes in Croatia [15]. All these findings, coupled with data herein reported, provide strong evidence that Megatrypanum trypanosomes exhibit a narrow host range or even host specificity. Each trypanosome species/genotypes were found in a single host species or in closely phylogenetically related hosts, those found in cervids were never detected in bovids although one host species can harbour trypanosomes of more than one species or It has been demonstrated by isoenzyme and karyotype analyses that the trypanosome found in Swedish reindeer differ from those found in moose, and both differed from cattle isolates, despite all these animals living in sympatry [13]. Similarly, data from zymodemes suggested the existence of different species of Megatrypanum infecting distinct species of deer and cattle in Germany [59]. The isolates of T. perronei sp. n. from Venezuela and the USA are closely related, but not identical.
Our findings agreed with multiple and relatively recent crossings of the Bering Strait by cervids infected with Megatrypanum trypanosomes reaching North America from Eurasia, and from these regions dispersing through the world. Altogether, deer-trypanosome-vector associations and phylogeography support a plausible evolutionary scenario where WTD infected with the ancestor of T.
perronei sp. n., likely infested by its tightly linked ectoparasite L. mazamae, were introduced from North America into South America through the Panama Isthmus, reaching this continent at the Pliocene-Pleistocene boundary [1,60]. Cervidae originated in Asia between 7.7 and 9.6 mya, and according to fossil records, deer did not cross the Bering Land Bridge to North America before 4.2 to 5.7 mya [60]. South American cervids are thought to have originated from at least two invasion events by North American deer: firstly, by the common ancestor of all deer species endemic of South America during the Great American Interchange at the Early Pliocene (~3 mya); and more recently (~1.5 mya) only by WTD at the Pliocene-Pleistocene boundary [1,60]. Concordant with our data on Megatrypanum trypanosomes, host-helminth assemblages were also associated with an early dispersion of cervids and bovids from Eurasia into North America and then into the Neotropics [61].
Also supporting the recent dispersion of cervids and their parasites, Plasmodium sp. from the South American pampas deer (Ozotoceros bezoarticus) is closely related to Plasmodium odocoilei of North American WTD, and these two species are estimated to have diverged just by 0.3-0.9 mya [62].

Growth behaviour and developmental forms of T. perronei sp. n. in early haemocultures and co-cultivated with insect and mammalian cells examined by light and scanning electron microscopy
In early (7-10 days) haemocultures of WTD blood, live flagellates (phase microscopy) exhibited a few trypomastigotes with large body length, pointed posterior ends and noticeable undulating membrane, alongside large transition forms between trypo-and epimastigote forms, and dividing epimastigotes ( Fig. 2A). Flagellates of T. perronei sp. n. from early haemocultures seeded on monolayers of insect cell (Hi-5), at 25 °C, initially formed clumps of small and rounded forms attached to insect cell membranes; these forms increased in length to became epimastigotes that remained adhered by their flagella forming rosettes until released into the supernatant of cultures (Fig. 2B). The developmental forms of T. perronei sp. n. co-cultivated with Hi-5 insect cells very much resembled those reported for T. (Megatrypanum) spp. in the guts of the ked L. cervi taken from red deer [38] and T. melophagium adhered to the cells of gut walls of the sheep keds [15,21].
Epimastigotes of T. perronei sp. n. in log phase Hi-5 cultures (5 days) multiply intensively attached by their flagella forming large rosettes (Fig. 7a), which initially remained adhered to the insect cells and afterwards are released in the supernatant, where free epimastigotes became progressively abundant ( Fig. 7a,b). Giemsa-stained epimastigotes showed the rounded kinetoplast adjacent and lateral to the central nucleus with an almost imperceptible undulant membrane, and a long free flagellum (Fig.   7a,b). In mid-log cultures (7 days), most epimastigotes became longer and thinner with a pointed posterior extremity (Fig. 7b). Stationary phase cultures (10 days) of T. perronei sp. n. exhibited variable forms, all with a long free flagellum, including some wider epimastigotes exhibiting more preeminent undulant membranes (Fig. 7c). Some forms became progressively shortened in their posterior ends giving origin to blunted forms (indicated by arrow heads) during the differentiation of epi-to trypomastigotes (Fig. 7 b,c) and, finally, to 'rounded' forms with a long flagellum (Fig. 7b,c), which most likely represented metacyclic trypomastigote forms (Fig. 7d). In contrast with the slow movement of long epimastigotes, these "rounded" forms are highly mobile, and resemble metacyclic trypomastigotes of T. theileri described previously in the guts of tabanid flies and stationary cultures [57]. Overall, initial co-cultivation of T. perronei sp. n. with insect cells, at 25 ºC, showed flagellates resembling those of Trypanosoma (Megatrypanum) spp. present in the guts of L. cervi, the Old-World deer ked taken from red deer [38].
Scanning electron microscopy (SEM) of the mid-log cultures (7 days) of T. perronei sp. n. in Hi-5 cultures showed flagellates of variable length and shape (Fig. 7e-j): slender epimastigotes without a noticeable undulant membrane (Fig. 7e) become broader epimastigotes exhibiting a conspicuous undulant membrane easily detectable by SEM (Fig. 7f,i,j). Following the differentiation from epi-to metacyclic trypomastigotes, a range of transition forms (indicated by arrows heads) were observed, including flagellates with a pointed posterior end and swollen central region (Fig. 7f-h), which progressively turn into forms with a blunted posterior extremity until whole differentiation into bellshaped flagellates with long free flagella (Fig. 7f,g,k), which correspond to the apparently 'rounded' metacyclic trypomastigotes observed by light microscopy (Fig. 7c,d).
Log phase epimastigotes from Hi5-cultures were seeded into monolayers of mammalian LLC-MK2 cells, incubated at 37 °C with 5% CO 2 , and after one to 5 days, cultures were examined by light microscopy of Giemsa-stained flagellates (Fig. 8a-f) and SEM (Fig. 8 g-j). In the supernatant of these cultures, slender epimastigotes gradually became wider (Fig. 8a,g -one day culture) and gave origin to large and wide transition forms (indicated by arrow heads) between epi-and trypomastigotes initially exhibiting wide bodies (Fig. 8b,c,h), and then becoming long and slender showing welldeveloped undulant membranes and pointed posterior ends (Fig. 8 e,f,i,j). Both large epi-and trypomastigotes are multiplicative forms (Fig. 8b,d). Long and slender forms with sharpened posterior ends and prominent undulant membranes (Fig. 8 e,f, (Fig. 9a,b) distributed throughout the cell body; a kinetoplast exhibiting long and weakly compacted DNA fibrils and, consequently, wide thickness ( Fig. 9a-d), a noticeable spongiome comprising a network of tubules and contractile vacuoles near the flagellar pocket (Fig. 9c,e,f), and the absence of cytostome. This is the first time that a deer trypanosome is characterized by TEM, and the ultra-structural arrangement was similar to that showed previously for T. theileri [23,65].

New species description
Trypanosoma perronei sp. n. Teixeira, Camargo and García Type material: Hapantotype, culture of the isolate TCC2268. The isolate WTD A3, known just by partial ITS rDNA and SSU rRNA sequences obtained from the blood of WTD from USA, and deposited in GenBank as Trypanosoma sp. PJH-2013a isolate WTD A3, was herein designed as a genotype of T. perronei sp. n.

Authors' contributions
HAG, MMGT and EPC conceived and supported the study. HAG, CMFR and APBM designed and coordinated field and laboratory work. HAG and APBM performed the fieldwork in Venezuela and participated in parasite isolation and morphological analysis. HAG, CMFR and ACR were responsible for the molecular diagnosis and phylogenetic analyses. MC carried out culturing, behavioural and morphological analyses. CSAT performed electron microscopic analyses. HAG, MMGT and EPC wrote the manuscript. All authors contributed to the revisions of the manuscript, read and approved the final manuscript.

Ethics approval and consent to participate
All field procedures and deer handling were performed in accordance with the protocol approved by the MINEC (the Venezuelan Ministerio del Poder Popular para el Ecosocialismo) permit Number 2013-154125.

Consent for publication
Not applicable.