From the analysis of 240 18S rDNA sequence-structure pairs (Fig. 1) and selection of 43 different species (Fig. 2), we obtained trees supported by high bootstraps values (> 75) on sister groups displaying the following Trypanosoma clades: the Trypanosoma pestanai Bettencourt & Franca, 1905 clade, represented in our tree by this species found in the Eurasian badger [13]; the T. brucei clade, consisting of trypanosome species naturally transmitted by tsetse flies, such as Trypanosoma vivax Ziemann, 1905, Trypanosoma congolense Broden, 1904, Trypanosoma godfrey McNamara et al., 1994, Trypanosoma simiae Bruce et al., 1913, Trypanosoma equiperdum Doflein, 1901, Trypanosoma evansi Steel, 1885, and T. brucei [13,17,19,21]; the T. cruzi clade, comprising mammalian trypanosomes with worldwide distribution, such as T. cruzi, Trypanosomarangeli Tejera, 1920, and Trypanosomawauwau Teixeira & Camargo, 2016, endemic of Latin America, Trypanosomaconorhini (Donovan, 1909) Shortt & Swaminath, 1928 found in Europe, South America and Africa, and Trypanosoma dionisii Bettencourt & Franca, 1905 distributted in Latin America, Africa, Asia and Europe [19–21,35]; the Trypanosoma lewisi (Kent, 1880) Laveran & Mesnil, 1901 clade, including the rodent parasites Trypanosoma microti Laveran & Pettit, 1910, Trypanosoma grosi Laveran & Pettit, 1909 and T. lewisi [36]; the Crocodilian clade, harboring Trypanosoma grayi Novy, 1906 from Africa and Trypanosoma ralphi Teixeira & Camargo, 2013 from South America [15,37]; the Avian clade, with Trypanosoma corvi Stephens & Christophers, 1908, Trypanosoma avium Danilewsky, 1885 and Trypanosoma thomasbancrofti Slapeta, 2016 [38]; the Trypanosoma theileri Laveran, 1902 clade, with T. theileri, a worldwide distributted cattle parasite, and the subclade representant Trypanosoma cyclops Weinman, 1972 [13,21]; and the Aquatic clade, harboring trypanosomes from fish, anurans and platypus [14,39,40]. Interestingly, the lizard/snake clade is also represented in our tree with Trypanosoma varani Wenyon, 1908, a snake trypanosome, branching together with the mammal parasite Trypanosoma freitasi Rego et al., 1957. The branching of marsupial and rodent trypanosomes inside this clade has been previously observed [36,41]. Thus, our analysis corroborates the existence of the lizard-snake/marsupial-rodent clade composed by trypanosomes transmitted by sandflies [36].
The phylogenetic analyses using sequence-structure data of 18S rDNA (Fig. 2) supports the monophyly of Trypanosoma as previously observed in trees constructed with partial/ complete sequences of 18S rDNA and/or gGAPDH sequences [15,17–19]. Intriguingly, in the tree obtained using a greater number of sequences (Fig. 1) Strigomonas culicis (Wallace & Johnson, 1961) Teixeira & Camargo, 2011 (U05679-1 and HQ659564-1) appear as a basal group of African trypanosomes (Fig. 1). To date, studies on Trypanosomatida showed a basal position of Trypanosoma in relation to Strigomonas Lwoff & Lwoff, 1931 [2,22,42]. Of notice, S. culicis is a monoxenic parasite of the order Diptera [22], the same order of the well-known vector of T. brucei clade trypanosomes, the tsetse fly. Although it may be tempting to speculate the derivation of the tsetse lifestyle of African trypanosomes from Strigomonas, we cannot exclude the possibility of poor sequence assembly, which could interfere with the topology observed. Finally, one sequence of the bat parasite T. dionisii clustered within the T. brucei clade (Fig. 1). This species is distributed worldwide, with its origin in Africa, and presents a high phyletic diversity [35]. However, its branching inside T. cruzi clade is strongly supported [13,17–19,21].
The first branching of Trypanosoma (Fig. 2) forms two major groups: one lineage composed by T. brucei and T. pestanai clades, and another with trypanosomes from Terrestial (T. cruzi, T. lewisi, T. theileri, snake-lizard/marsupial-rodent, avian and crocodilian clades) and Aquatic lineages. Thus, our tree corroborates the hypothesis of the independent evolutionary history of both human pathogens, T. brucei and T. cruzi [13]. The topology of our tree shows the Aquatic clade as a solid lineage, in accordance with previous observations [15,18,19,21]. However, the origin of this clade is still under debate. Many studies using different DNA markers, such as long (> 1.4 kb) 18S rDNA sequences, v7v8 hypervariable region of 18S rDNA and/or partial sequences of gGAPDH, showed either an early division between Aquatic and Terrestrial lineages as a single event [15,18,19,21] or in subsequent events with amphibian trypanosomes and Trypanosoma therezieni Brygoo, 1963 at the basis of Trypanosoma [17,43,44]. Interestingly, our tree suggests a later evolution of the Aquatic clade from Terrestrial trypanosomes (Fig. 2), which agrees with the insect-first hypothesis [13,36]. This hypothesis assumes that trypanosomes were originated from a monogenetic insect parasite that adapted to live inside terrestrial vertebrates and later spread to leeches and other aquatic animals, most likely through amphibians [13].
Trypanosomes of the T. brucei clade are virtually restricted to Africa, having an exception in T. vivax [45,46]. The early divergence of T. vivax inside the T. brucei clade (Fig. 1, 2) is in accordance to previous results showing a higher evolutionary rate of this species among the Salivarian trypanosomes [19,21,47]. It is interesting to consider that a previous analysis of 18S rDNA sequences revealed that members of the T. brucei clade show an evolutionary rate higher than other trypanosomes [47]. However, this high divergence has proven not to alter the topology of sequence-based trees [17]. Inside this clade, a low bootstrap value (ML = 53) is observed in the differentiation between T. brucei and T. evansi (Fig. 2), suggesting an unresolved positioning. In fact, the relationship between these species is controversial, with results supporting either T. evansi as a subspecies of T. brucei or showing great diversity between both depending on the T. evansi strains [48,49].
Regarding the other major group of our analysis (Terrestrial/Aquatic lineages), the first branching inside this group suggests the differentiation of the snake-lizard/marsupial-rodent clade as a basal group of other trypanosomes (Fig. 2). However, other studies have suggested avian trypanosomes as a basal group among terrestrial lineages [13,17]. This can be associated with the low bootstrap values of our tree in either three events: the snake-lizard/marsupial-rodent clade (ML = 54) differentiation, the divergence of crocodilian trypanosomes (ML = 55), and the internal branch of avian trypanosomes (ML = 61), which will be further explored in our discussion.
The T. cruzi and T. lewisi clades appear as sister groups in our analysis (Fig. 2), as has been previously demonstrated [17–19,21]. The T. cruzi clade can be subdivided into three subclades: Schyzotrypanum, T.wauwau (and other Neotropical bat trypanosomes, Trypanosomanoyesi Botero & Cooper, 2016, and Trypanosomalivingstonei Teixeira & Camargo, 2013) and T. rangeli/T.conorhini [21,35]. The specific sequences of Trypanosoma minasense Chagas, 1908 and Trypanosoma leeuwenhoeki Shaw, 1969 grouped with T. rangeli in our tree were previously considered synonyms of this species by 18S rDNA sequence analysis, explaining their positioning [13,50]. Regarding the T. lewisi clade, our results suggest the existence of two subclades inside the group (ML = 100), one harboring T. microti, and the other with T. lewisi and T. grosi. This finding is in accordance with a recent analysis of long fragments of 18S rDNA which demonstrated this subdivision despite the similarities in the v7v8 hypervariable region [51].
Concerning avian trypanosomes, our tree reflects previous findings in both the divergence between T. avium and T. corvi and the highly supported (ML = 100) proximity between T. avium and T. thomasbancrofti [38,50]. A low bootstrap value (ML = 61) is observed in the divergence of T. corvi and T. avium/ T. thomasbancrofti. It is interesting to note that we currently have two topologies known in literature, with the possibility of paraphyly demonstrated by analysis of long sequences of 18S rDNA [19,38]. This, however, was not observed in trees constructed with v7v8 hypervariable region of 18S rDNA, gGAPDH sequences, or concatenated trees using both [15,37,52]. Thus, our tree indicates the need for a better resolution on avian trypanosome positioning. Considering that we used only three species of avian and two species of crocodilian trypanosomes in our reconstruction, our approach represents an interesting method to be applied in further studies.
In our analysis we see the crocodilian/alligator trypanosomes (T. grayi and T. ralphi) branching together with a high support value (ML = 100) (Fig. 2). Although T. grayi is found in Africa and T. ralphi in South America [15,37,52], in a tree with distant external groups like ours, this topology is expected due to their proximity inside the Crocodilian clade [15,37]. Our tree reflects the proximity of the crocodilian trypanosomes with T. cruzi clade, as previously observed through full genome analysis [53]. Interestingly, crocodilian trypanosomes, such as T. grayi and Trypanosomakaiowa Teixeira & Camargo, 2019 are tsetse-transmitted species that are not restricted to the sub-Saharan belt [15,37,53], suggesting higher adaptive plasticity of crocodilian trypanosomes.
The trypanosomes of the Aquatic lineage branched together (Fig. 2). The subgroups observed are anuran trypanosomes (Trypanosoma rotatorium (Mayer, 1843) Laveran, 1901, Trypanosoma mega Dutton & Todd, 1903, Trypanosoma fallisi Martin & Desser, 1990, Trypanosoma ranarum (Lankester, 1871) Danilewsky, 1885, and Trypanosoma neveulemairei Brumpt, 1928) and fish trypanosomes (Trypanosoma siniperca Chang, 1964, Trypanosoma ophiocephali Chen, 1964, Trypanosoma cobitis Mitrophanow, 1884, Trypanosoma granulosum Laveran & Mesnil, 1902, Trypanosoma pleuronectidium Robertson, 1906) along with the platypus parasite Trypanosoma binneyi Mackerras, 1959, which is in accordance to the literature [14,15,39,40]. Interestingly, the anuran parasite, Trypanosoma chattoni Mathis & Leger, 1911, appears in our analysis more related to fish and platypus trypanosomes than to the anuran clade. This positioning of T. chattoni was shown in a previous study using complete 18S rDNA sequences and non-trypanosomes as the outgroup [54]. However, recent trees using complete 18S rDNA sequences and concatenated analysis of v7v8 hypervariable region and gGAPDH rooted by other trypanosomes sustained a monophyletic anuran clade [39,40]. Thus, T. chattoni positioning in our tree can be related to the use of non-trypanosomes as the outgroup.