Background

Tsetse-transmitted trypanosomes are notorious pathogens of cattle in sub-Saharan Africa, causing the disease nagana or African animal trypanosomiasis (AAT), but it is often overlooked that these trypanosomes also cause disease of varying severity in a range of other livestock species. Indeed, Trypanosoma simiae causes acute porcine trypanosomiasis and was once described as “the lightning destroyer of the domestic pig” because of the swift and fatal outcome of infection [1]. Though best known as a virulent pathogen of domestic pigs, the host range of T. simiae encompasses cattle and camels [1,2,3] and it is frequently found in tsetse across sub-Saharan Africa [4,5,6,7,8,9,10,11].

Inarguably, based on pathogenicity and prevalence, T. simiae deserves more attention, but there has been comparatively little recent research on this species compared to its better known relative, Trypanosoma congolense. Both species were assigned to subgenus Nannomonas, because they follow a similar developmental pathway in the tsetse fly vector [1]. Briefly, bloodstream form trypanosomes differentiate to procyclic forms in the tsetse midgut, multiply and then move anteriorly to colonise first the proventriculus or cardia (the valve between the mid- and foregut) and subsequently the proboscis. Infective metacyclics develop in the hypopharynx, the single duct that carries saliva from the paired salivary glands to the tip of the proboscis, and are inoculated with the saliva as the fly feeds on an animal; the complete cycle of development takes about 20 days [1].

This developmental cycle was described decades ago, before the discovery that trypanosomes may undergo sexual reproduction in the tsetse vector [12] and the description of meiotic dividing cells and gametes [13,14,15]. While these results pertain to Trypanosoma brucei, there is a strong probability that sexual stages might also be found among trypanosomes of subgenus Nannomonas, as there is evidence that T. congolense undergoes genetic exchange in nature [16, 17]. However, no obvious sexual stages were identified during the developmental cycle of T. congolense in tsetse [18], despite subsequent demonstration that hybrids are found in the tsetse proboscis in laboratory crosses (Peacock et al. unpublished). It was therefore a surprise to find putative meiotic dividing forms and gametes among trypanosomes recovered from proboscides of tsetse infected with T. simiae and here we describe these stages as well as providing a detailed account of the development of T. simiae in its tsetse vector.

Methods

Trypanosomes and tsetse flies

Procyclic forms of T. simiae strain TV008 (GMOS/GM/85/SAMIWOLO-008), originally isolated from the midgut of an infected tsetse fly (Glossina morsitans submorsitans) in The Gambia in 1985 [19], were grown in modified Cunningham’s medium (CM) [20, 21] at 27 °C. Pupae of Glossina pallidipes were kindly supplied by the tsetse-rearing facility of IAEA, Vienna; pupae and emerged flies were kept at 25 °C and 70% relative humidity. New emergents were placed in separate cages according to sex and given an infected blood meal for their first feed 24–48 h post-eclosion; thereafter, they were fed on sterile defibrinated horse blood supplemented with 1 mM dATP [22] via a silicone membrane. The infective blood meal contained approximately 107 procyclic trypanosomes ml−1 in CM with an equal volume of washed horse red blood cells resuspended in Hank’s Balanced Salt Solution, supplemented with 10 mM l-glutathione [23] to inactivate reactive oxygen species.

Fly dissection

Cold-anaesthetized flies were dissected 19 to 31 days after infection. The head of the fly was removed with a scalpel blade and retained for subsequent dissection of the proboscis. The whole alimentary tract, from the proventriculus to the rectum, was dissected in a drop of phosphate-buffered saline (PBS) and viewed as a wet mount under phase contrast (× 100 magnification) to determine whether trypanosomes were present; positive midguts were divided into mid-posterior and anterior sections and pooled separately in 500 µl CM. Proventriculi were removed from infected midguts and pooled in 50 µl CM. Proboscides from flies with a midgut infection were dissected into a separate drop of PBS and teased apart to separate the component parts, carefully removing and discarding the labium. The labrum and hypopharynx were pooled in 50 µl CM. All samples were incubated at room temperature for at least 10 min to allow trypanosomes to escape from tissue; in addition, the anterior and mid-posterior midgut samples were passed through a 100-µm filter to remove tissue and then washed twice in PBS before resuspending in 500 µl PBS. Fifty-microlitre aliquots of each sample were mixed with an equal volume of 4% w/v paraformaldehyde in PBS and incubated at room temperature for 30 min. Volume was increased to 200 µl with PBS and cells placed in a Cytospin (1350 rpm for 5 min), air dried, mounted in Vectashield (Vector Labs) containing 4’,6-diamidino-2-phenylindole (DAPI) and viewed immediately. Cibaria from flies with a midgut infection were dissected into a separate drop of PBS and viewed as a wet mount.

Transfection

Plasmid constructs previously designed for integration into the PFR1 or HOP1 genomic loci of T. brucei [13] were used for transfection of T. simiae. In brief, the PFR1 construct introduced a C-terminal YFP fusion of PFR1 (Tb927.3.4290), while the HOP1 construct introduced an N-terminal YFP fusion of HOP1 (Tb10.70.1530). Successful integration of these constructs thus relied on sufficient homology between the PFR1 and HOP1 genes of T. simiae and T. brucei. Transgenes were expressed via read-through transcription from the endogenous promotors and regulated via the 3′ UTR; while PFR1 is expressed constitutively, HOP1 is expressed only during meiosis and therefore the 3′ UTR of YFP::HOP1 was left unaltered as this region is likely to determine developmentally regulated expression. The antibiotic resistance genes used for selection of transfectants were expressed constitutively. Procyclic trypanosomes were transfected by electroporation (Amaxa program X-01) with 10–20 µg of the linearised plasmids and were selected with either hygromycin (25 µg/ml) or puromycin (0.25 µg/ml). The PFR1::YFP transfected line had a fluorescent flagellum and was successfully fly transmitted. Procyclics of the YFP::HOP1 line were non-fluorescent as expected but failed to develop beyond the midgut during fly transmission and the line died out in culture.

In vitro development of proventricular trypanosomes

Flies were dissected 27 days after infection with either wild-type T. simiae TV008 or a PFR1::YFP transfected line. The proventriculus was separated from the foregut and midgut using forceps and hypodermic needles and placed in a tube containing CM with 1 × anti-contamination cocktail (ACC) [24]. Aliquots of this trypanosome suspension were then added to 1-ml wells of CM with 1 × ACC, each well containing a sterile 10–13-mm-diameter round glass coverslip. Plates were incubated at 27 °C and coverslips removed at time intervals and processed as previously described [25].

Immunofluorescence

Fly proboscides were dissected directly into CM as above and the supernatant fixed for 30 min in 4% w/v paraformaldehyde in PBS at room temperature. The immunofluorescence protocol of [15] was followed to identify the basal body using YL1/2 antibody which recognises tyrosinated α-tubulin [26].

Imaging and measurements

Stained slides were viewed using a DMRB microscope (Leica) equipped with a Retiga Exi camera and Volocity version 4.1 software (Improvision). Each image was photographed under phase contrast and UV fluorescence at 400 × magnification. Digital images were collated and mensural data were collected and analysed using ImageJ software (Version 1.41) (http://rsb.info.nih.gov/ij/); measurements were as previously described [18, 25], with the addition of cell and nuclear area, as shown in Fig. 1A. Data are compiled in Additional file 3: Tables S1 and S3.

Fig. 1
figure 1

Trypomastigote morphology. A The dimensions measured are as indicated: length = total cell length; width = maximum cell width; KNuc = distance between the kinetoplast and nucleus; KPost, NPost = distances from the cell posterior to the kinetoplast or nucleus respectively; KAnt and NAnt, the distances from the cell anterior to the kinetoplast or nucleus, respectively, were derived by subtracting KPost or NPost from the total cell length; NL, NW, NArea = nucleus length, width and area respectively; CellArea = area of cell body. In addition, the length of the paraflagellar rod of the flagellum (FL) was measured in the PFR1::YFP transfected line of T. simiae. CellArea and NArea were determined by drawing around the cell body or nucleus, respectively. B Schematic of changes in trypomastigote morphology in midgut and proventricular trypanosomes. Dimensions of each cell (total = 327) based on mean measurements (µm) for N trypanosomes (Table 1); y axis length, scale bar width

Full size image

Statistical analysis

Mensural data (11 variables, see Fig. 1A and Additional file 3: Table S3) for 1044 trypanosomes with a single nucleus (excluding cells with two or more nuclei but including cells with one or two kinetoplasts) from midgut, proventriculus and proboscis were subjected to principal components analysis (PCA) using the statistical package IBM SPSS Statistics for Windows, version 26. Three uncorrelated factors, PC1, PC2 and PC3, accounted for 78% of the variance (PC1 36%, PC2 27%, PC3 15%; Additional file 3: Table S2). Extracted scores for these three factors were plotted for each trypanosome (PC1 vs PC2, PC1 vs PC3). A subset consisting of 327 midgut and proventricular cells was subjected to a multivariate analysis using Bonferroni correction to reduce the chances of obtaining false-positive results for the comparison between tsetse tissue of the measurements of the 11 dimensions of cells.

Results

Tsetse infection with Trypanosoma simiae

Tsetse infections were initiated with cultured T. simiae procyclics rather than bloodstream forms to avoid the use of experimental animals. Flies were dissected from 19 days after infection, allowing time for completion of development to infective metacyclics, which typically takes a minimum of 2 weeks. Of 45 Glossina pallidipes fed an infected blood meal containing wild-type T. simiae and dissected on days 19–28 post infection, 42 (93%) had trypanosomes in the midgut and in 14 of these flies, and trypanosomes were also present in the region of the cibarium (14/42, 33%), though only three had trypanosomes within the cibarium itself. Twenty-three flies with an infected midgut also had an infected proboscis (transmission index 23/42 = 55%). A variety of trypanosome cell types were seen in and around the labrum of the proboscis, both attached and unattached (Additional file 1: Movie S1), but trypanosomes did not usually form the discrete rosettes observed for T. congolense [18]. Trypanosomes were also found in the hypopharynx (3 of 7 proboscides examined; Additional file 2: Movie S2), but it was difficult to ascertain cell type in the live preparations and whether the trypanosomes were free or attached.

Trypanosome cell types in the midgut and proventriculus

To further investigate the cell types present in different fly organs, trypanosomes from the midgut (anterior and mid/posterior sections), proventriculus and proboscis were fixed and stained with DAPI to visualize the nucleus and kinetoplast. In addition, cell structures containing tyrosinated tubulin were visualized by immunofluorescence using the YL1/2 antibody, which serves as a marker for the basal body and newly formed microtubules [26,27,28,29]. As well as wild-type trypanosomes, a genetically modified line of T. simiae TV008 carrying a PFR1::YFP fusion allowed the paraflagellar rod (PFR) inside the flagellum to be visualised by fluorescence microscopy.

The majority of trypanosomes in the midgut and proventriculus were of trypomastigote morphology, where the kinetoplast is posterior to the nucleus, while trypanosomes from the proboscis were much more variable, with epimastigotes (kinetoplast anterior to the nucleus) also present. Measurements of 11 dimensions for a total of 327 midgut and proventricular cells were made (Fig. 1A; Table 1) and subjected to multivariate analysis with Bonferroni correction. Surprisingly, the majority of measurements were significantly different between trypanosomes from the mid/posterior versus anterior midgut (9 of 11 dimensions) and anterior midgut versus proventriculus (8 of 11 dimensions); mid/posterior versus proventricular cells were significantly different for 10 of the 11 dimensions (Table 1). Thus trypanosomes from the mid/posterior midgut, anterior midgut and proventriculus were significantly different in morphology (Fig. 1B), as described in more detail in the next paragraph.

Table 1 Morphometry of Trypanosoma simiae trypomastigotes from the midgut and proventriculus
Full size table

There was a progressive increase in the length of the cell body and nucleus (Length, NL; P < 0.001) among trypanosomes from the mid/posterior to anterior midgut through to the proventriculus (Fig. 1B; Table 1); the distance between the kinetoplast and cell posterior (KPost) and cell anterior (KAnt) also progressively increased. The distance between the nucleus and kinetoplast (KNuc) and nuclear width (NW) were similar between cells from the mid/posterior and anterior midgut but were significantly reduced (P < 0.001) in proventricular cells as the nucleus became elongated and sausage-shaped. Cell and nuclear areas, as well as the distance of the nucleus to the posterior of the cell (NPost), were significantly larger (P < 0.001) in anterior midgut and proventricular trypanosomes compared to mid/posterior midgut trypanosomes. The proventricular cell population was uniform and consisted of long trypomastigotes with an elongated nucleus. No dividing cells were found in this population whereas occasional dividing cells were encountered in the midgut population (6 dividing trypanosomes in mid/posterior midgut and 7 in the anterior midgut).

Trypanosome cell types in the proboscis

Trypanosomes were extracted from pooled tsetse proboscides from flies dissected on days 27, 28 and 31 post infection and stained preparations examined by microscopy to identify the cell types present (Table 2). Among 1463 proboscis trypanosomes examined, the 1K1N epimastigote was the most numerous cell configuration, comprising over 40% (602 of 1463 trypanosomes); such cells had the kinetoplast anterior or juxtaposed to the nucleus and were classified into two types based on the length of the posterior, which was extremely elongated in about one third of cells (204 of 602, 34%; Table 2; Fig. 2A). The posterior of these cells stained intensely with the YL1/2 antibody (Fig. 2A, B), indicating large quantities of tyrosinated tubulin present in the subpellicular microtubules that form the supporting scaffold, and therefore recent growth of the cell posterior. Large numbers of dividing epimastigotes were also evident, having two kinetoplasts and sometimes also two nuclei in the configuration NKNK from the posterior of the cell (90 of 1463, 6.2%) (Table 2; Fig. 2C).

Table 2 Trypanosoma simiae cell types from the tsetse proboscis
Full size table
Fig. 2
figure 2

Images of trypanosomes (Trypanosoma simiae TV 008 PFR1::YFP) from proboscis. From L to R: phase contrast, DAPI false coloured cyan for clarity, YFP, TRITC immunofluorescence with YL1/2 antibody, merge. Arrows: kinetoplasts; arrowheads, basal bodies. A Epimastigote with an extremely long posterior. B Epimastigote with a short posterior. C 2K2N dividing epimastigote. D Two trypomastigotes. E Metacyclic. F 2K1N dividing trypomastigote. G 2K2N dividing trypomastigote-epimastigote. H 3K2N dividing epimastigote-epimastigote-trypomastigote. Scale bar = 10 µm

Full size image

A further ~ 20% (291/1463) of proboscis trypanosomes were trypomastigotes, with the kinetoplast posterior to the nucleus (Fig. 2D). Only eight of these were classified as metacyclics, as judged by their morphology (small non-dividing trypanosome with short body, lack of free flagellum and kinetoplast positioned close to the posterior end of the cell; Fig. 2E). As T. simiae metacyclics develop in the hypopharynx [1], it is probable that they were poorly represented because the hypopharynx was not disrupted during dissection and was often only lightly infected (Additional file 2: Movie S2). As trypanosomes had ample time for full development to metacyclics in the hypopharynx, a possible reason for light infection is that the T. simiae strain used, which originates from The Gambia in West Africa, was not well adapted to the fly species used, G. pallidipes, which is of East African origin, but we have limited experience of the development of other strains of T. simiae.

Trypomastigotes from the proboscis differed from those in the midgut and proventriculus as the distance between the kinetoplast and nucleus (KNuc) in proboscis trypomastigotes (0.99 ± 0.05 µm, N = 116) was less than that for trypomastigotes from the midgut and proventriculus (3.67 ± 0.14 µm, N = 219; 2.31 ± 0.08 µm, N = 108, respectively), and these organelles were closer to the posterior end of the body (KPost: 3.22 ± 0.18 µm; NPost: 4.74 ± 0.17 µm) compared with trypomastigotes from the midgut (KPost: 4.95 ± 0.19 µm; NPost: 9.49 ± 0.22 µm) and proventriculus (KPost: 7.19 ± 0.20 µm; NPost: 10.22 ± 0.24 µm) (Additional file 3: Table S1). To analyse all 11 dimensions simultaneously, we used principal components analysis (PCA), which demonstrated some clustering of trypomastigotes by their tissue location (Fig. 3A). The clusters are not entirely separate and show some overlap, as the four trypomastigote populations are not confined to particular regions, but move from the mid/posterior midgut to the anterior midgut and proventriculus, and thence to the proboscis. Three principal components accounted for 79% of the variance between cells (PC1 37%, PC2 27%, PC3 15%); PC1 was correlated with cell area (CellArea), nucleus area (NArea), nucleus length (NL) and the distance of the nucleus from the cell anterior (NAnt), while PC2 was correlated with KNuc and KAnt and PC3 with KPost and NPost (Additional file 3: Table S2).

Fig. 3
figure 3

Principal components analysis (PCA): 1044 cells with a single nucleus from all tsetse tissues (midgut, proventriculus and proboscis) were analysed by PCA. A Comparison of 443 trypomastigotes from the mid/posterior midgut (104, MidPostMG), anterior midgut (115, AntMG), proventriculus (108, Prov) and proboscis (116, Trypo). For mensural data see Additional file 3: Table S1. B Comparison of 717 trypanosomes from the proboscis. Cell types: trypomastigotes (116, Trypo, gametes (93, Gamete), epimastigotes (315, Epi), epimastigote-trypomastigote dividers 2K1N (18, EpiTrypo), epimastigote dividers 2K1N (36, DivEpi), meiotic dividers 2K1N (129, Meiotic) and meiotic intermediates (10, 3K1N). For mensural data see Additional file 3: Tables S1 and S3; for PCA weightings see Additional file 3: Table S2

Full size image

Some proboscis trypomastigotes were undergoing cell division as they had two kinetoplasts and sometimes also two nuclei (7 of 1463 cells; Table 2). Some of these were trypomastigote-trypomastigote dividers (Fig. 2F), while others were asymmetric epimastigote-trypomastigote dividers (Fig. 2G, H). In addition, numerous other trypanosomes with multiple kinetoplasts and/or nuclei were observed, some of which resembled sexual stages previously observed in T. brucei; these are described in the next section.

An intriguing observation was the presence of 1K0N trypanosomes (zoids) (Fig. 4A) among trypanosomes in the proboscis at relatively high frequency (1.3%; Table 2). Zoids may have arisen from 2K1N cells with one extremely anterior kinetoplast (Fig. 4B, C). Zoids were not previously noticed among T. brucei trypanosomes developing in the salivary glands [30] or T. congolense developing in the proboscis [18], and thus this is the first time we have recorded zoids in tsetse infections. In previous studies we have searched through thousands of trypanosome images and applied a threshold of 1% in assuming that any particular cell type occurs frequently enough to be a biologically relevant stage. The biological significance of zoids in T. simiae is unknown.

Fig. 4
figure 4

Zoid trypanosomes (T. simiae TV008 wild type) from proboscis. From L to R: phase contrast, DAPI false coloured cyan for clarity, merge. Arrows: kinetoplast of zoid; arrowheads, kinetoplast of 1K1N trypanosome. A Zoid. B, C Zoids apparently arising from 1K1N trypanosomes with an extra, extremely anterior kinetoplast. Scale bar = 5 µm

Full size image

Identification of putative sexual stages of T. simiae

A prominent cell type, comprising 11.1% (162/1463) of proboscis trypanosomes, resembled the meiotic dividing cells previously identified in T. brucei [13], having a large nucleus that was positioned close to, and usually filling, the posterior end of the cell, and two kinetoplasts (Figs. 5 and 6A, B). Both kinetoplasts were adjacent to basal bodies stained by the YL1/2 antibody, which recognises tyrosinated α-tubulin (Fig. 5A, B). In addition, the presence of two flagella was verified in the genetically modified line T. simiae TV008 PFR1::YFP (Fig. 5C, D); the PFR of the posterior (daughter) flagellum was usually less fluorescent than that of the anterior flagellum, as also observed for T. brucei [13]. Trypanosoma brucei cells with such morphology were shown to express three different meiosis-specific proteins in the nucleus [13], verifying their identity as meiotic dividers, but here our attempt to fly transmit a genetically modified line of T. simiae carrying a fusion of YFP with the meiosis-specific gene HOP1 was unsuccessful, as the trypanosome clone failed to develop beyond the midgut during fly transmission. Thus, the identification of these cells as meiotic dividers currently rests on morphology alone.

Fig. 5
figure 5

Images of putative meiotic dividers. A, B Wild-type T. simiae. From L to R: Phase contrast, DAPI false coloured cyan for clarity, TRITC immunofluorescence with YL1/2 antibody, merge. Arrows: kinetoplasts; arrowheads: basal bodies. C, D PFR1::YFP line of T. simiae. From L to R: phase contrast, DAPI false coloured cyan for clarity, YFP, merge. The PFR of the posterior (daughter) flagellum (double chevrons) was usually less fluorescent than that of the anterior flagellum. Scale bar = 10 µm

Full size image
Fig. 6
figure 6

Putative meiotic intermediates and gametes. A–E Wild-type T. simiae stained with DAPI. From L to R: phase contrast, DAPI false coloured cyan for clarity, merge. Arrows: kinetoplasts. A, B Putative meiotic dividers with large posterior nucleus. C Putative meiotic intermediate with three nuclei and four kinetoplasts. D, E Putative meiotic intermediates with two nuclei of different fluorescence intensities, i.e. of different sizes; in D the posterior kinetoplast is more intensely fluorescent than the anterior and elongated, suggesting that it may be about to divide; in E it is uncertain whether the most posterior fluorescent spot is truly a kinetoplast. F, G PFR1::YFP line of T. simiae. Putative meiotic intermediates producing daughter cells, which may be gametes. From L to R: phase contrast, DAPI false coloured cyan for clarity, YFP, merge. Double chevrons indicate the PFR of the daughter flagellum, which usually has lower intensity fluorescence compared to the original flagellum. In F the identity of the nucleus and kinetoplast(s) of the daughter cell is uncertain, though a new flagellum is evident, judging from the length of visible PFR; in G the original cell is already replicating again, judging by the elongation of the nucleus and presence of a new flagellum, though only a single kinetoplast is visible. H, I Putative 1K1N and 2K1N gametes, respectively. Wild-type T. simiae stained with DAPI. From L to R: phase contrast, DAPI false coloured cyan for clarity, merge. Arrows: kinetoplasts. Scale bar = 10 µm

Full size image

The putative meiotic dividers formed a separate cluster from other 2K1N epimastigotes in PCA (Fig. 3B). These cell types differed in terms of the area and position of the nucleus: mean NArea meiotic dividers 7.31 ± 0.13 µm2 compared to 3.23 ± 0.20 µm2 in 2K1N epimastigotes; mean NPost meiotic dividers 0.84 ± 0.11 µm compared to 6.20 ± 0.91 µm in 2K1N epimastigotes (Additional file 3: Table S3).

In trypanosomes as cell division progresses, the distance between the two kinetoplasts K1 and K2 increases [28], and therefore K1-K2 distance can be used to monitor the progress of meiotic division and investigate associated morphological changes. Correlation analysis showed that as the K1-K2 distance increased, there was a significant lengthening of the cell, yet no significant increase in width or cell area; at the same time the nucleus became longer and slightly narrower without a noticeable change in area (Additional file 4: Fig. S1).

In our previous analysis of meiotic division in T. brucei [15], we assumed that the 4C (4 × haploid DNA content) posterior nucleus of the meiotic dividing cell went on to divide into two 2C (2 × haploid DNA content) nuclei, but as the morphology of this cell would resemble that of a dividing epimastigote, it would be difficult to identify unequivocally. We inferred that subsequent division of these nuclei was asynchronous, because trypanosomes containing either one 2C and two 1C nuclei or one 2C and one 1C nucleus were observed; such trypanosomes sometimes had a well-advanced cell cleavage furrow, splitting off a small haploid daughter cell [15]; a further unusual feature of these putative multi-nucleate meiotic intermediates was that they had variable numbers of kinetoplasts. Cells of similar morphology were identified here in T. simiae (Table 2; Fig. 6C–G), potentially representing meiotic intermediates. These trypanosomes were observed alongside abundant meiotic dividers in the same microscope fields (Additional file 5: Fig. S2).

We also identified potential gametes among proboscis T. simiae trypanosomes by their resemblance to gametes of T. brucei, which have a small pear-shaped body and a relatively long flagellum [14, 15]. In T. simiae the putative gametes had a small cell body that was wide near the posterior to accommodate the single nucleus and sometimes tapered almost to the anterior end of the cell (Fig. 6H, I); both 1K1N and 2K1N putative gametes were observed, but the majority were 1K1N (100 1K1N and 6 2K1N, Table 2). These putative 1K1N gametes were distinct from other 1K1N epimastigotes in terms of the area of the cell and nucleus and position of nucleus and kinetoplast: mean CellArea, mean NArea 1K1N gametes 22.22 ± 6.93, 1.58 ± 0.57 µm2 (respectively) compared to 33.77 ± 0.50, 2.41 ± 0.05 µm2 (respectively) in 1K1N epimastigotes; mean NPost, mean KPost 1K1N gametes 2.59 ± 1.46, 4.53 ± 1.66 µm (respectively) compared to 10.04 ± 0.33, 11.93 ± 0.34 µm (respectively) in 1K1N epimastigotes (Additional file 3: Table S3). The putative gametes formed a distinct cluster in PCA (Fig. 3B).

In summary, the presence of putative meiotic intermediates and gametes suggests that meiosis is part of the T. simiae life cycle in the tsetse proboscis, just as it is part of the T. brucei life cycle in the tsetse salivary glands [13, 14].

In vitro development of proventricular trypanosomes

As it would be difficult to identify newly arrived, migratory trypanosomes in the proboscis of a live tsetse fly, we studied the development of proventricular trypanosomes in vitro, an approach successfully applied to T. congolense previously [25]. Dissected proventriculi were pooled to release proventricular trypanosomes, which were then incubated overnight in wells of culture medium containing a sterile glass coverslip. The long proventricular trypomastigotes attached to the coverslip and gradually became shorter and stouter, as previously observed for T. congolense [25] (Fig. 7A). However, attachment was not as rapid as for T. congolense, where the cells attached almost as soon as they touched the coverslip; for T. simiae, trypanosomes moved around and appeared to “probe” with the tip of their flagellum for 30 – 60 min before finally starting to attach. After 8 h, some cells initiated cell division, again as previously observed for T. congolense [25], but between 24 and 48 h increasing numbers of cells became misshapen and bloated and cultures died out by 72 h.

Fig. 7
figure 7

In vitro development of T. simiae proventricular trypanosomes. A Representative images of T. simiae cells from a 24-h time course of development of wild-type proventricular trypanosomes in vitro. At T = 0 h proventricular trypanosomes were put into wells of culture medium and sampled every 4 h. By T = 16 h some cells had almost completed cell division (Div). Kinetoplasts are indicated by arrows. Scale bar = 10 µm. B Morphological changes of 1K1N cells recorded over the 24-h time course; drawings based on mean measurements of N cells; y axis length, scale bar width. C Representative images of T. simiae TV008 PFR1::YFP cells from a 36-h time course of development of proventricular trypanosomes in vitro. Each image is a merge of phase contrast, DAPI (false coloured cyan for clarity) and YFP images. Kinetoplasts are indicated by arrows. At T = 0 proventricular trypanosomes were put into wells of culture medium and sampled at intervals. The progression of this time course with genetically modified trypanosomes was slightly different to the wild type and no dividing cells were seen, perhaps because trypanosome numbers were smaller. Scale bar = 10 µm. D Morphological changes of 1K1N cells recorded over the 36-h time course; drawings based on mean measurements of N cells. Dimensions of each cell (total = 327) based on mean measurements (µm) for N trypanosomes (Table 1); y axis length, scale bar width. For mensural data see Additional file 3: Table S4

Full size image

Image and morphometric data were collected from an overlapping sequential time course from T = 0 to T = 24 h. At T = 0 proventricular trypanosomes formed a uniform population of long trypomastigotes with a single nucleus and kinetoplast (1K1N). At T = 8 h 2K1N cells started to appear and by T = 16 h some cells had already completed cell division, producing a smaller daughter cell (Fig. 7A); approximately equal numbers of epimastigote and trypomastigote morphologies were observed among the daughter cells. Figure 7B shows the morphological changes in 1K1N trypanosomes from T = 0 to T = 24 h. Over the first 4 h, the distance of the cell posterior from the nucleus and kinetoplast decreased, while the distance between these organelles and the cell anterior increased; the cells became noticeably wider. From 8 to 12 h, the cells shortened, with the reduction in length reflected in decreased distance between the nucleus and cell anterior; the nucleus, at first elongated, became rounded. The morphology of the cells remained relatively constant from 12 to 24 h (Fig. 7B).

Previous analysis of a similar time course of development of T. congolense proventricular trypanosomes showed that the paraflagellar rod (PFR) of the flagellum shortened by about 30%, accompanied by a noticeable accumulation of the PFR protein, PFR1, visualised by tagging PFR1 with YFP [24]. The same approach was used here to tag PFR1 of T. simiae TV008 and follow remodelling of the flagellum during development of proventricular forms in vitro (Fig. 7C, D). For unknown reasons, the progression of this time course with genetically modified trypanosomes was slightly different to wild type; the initial proventricular trypanosomes were substantially longer, ~ 45 µm compared to 35 µm for wild type, and no dividing cells were seen, perhaps because trypanosome numbers were smaller. As with the wild-type trypanosomes, the genetically modified trypanosomes showed a similar pattern of morphological changes, with cells becoming shorter and stouter over the time course (Fig. 7C, D). In contrast to T. congolense [25], the maximum reduction in PFR length was 10%, which occurred within the first 12 h, and no further reduction was observed up to the end of the 36 h time course; the accumulation of PFR1::YFP was not observed at any timepoint, perhaps because flagellar length reduction was less pronounced in T. simiae compared to T. congolense.

Discussion

Paradoxically the virulent pig pathogen Trypanosoma simiae is widespread and common in sub-Saharan Africa but little studied compared to its relatives T. brucei and T. congolense. Here we analysed image and mensural data from both wild-type and genetically modified T. simiae during its development in the tsetse midgut, proventriculus and proboscis, confirming that the life cycles of T. simiae and T. congolense are closely similar as expected, since both species were placed in the same subgenus based on their developmental characteristics in the tsetse vector [1]. The surprising finding was the presence of abundant (putative) sexual stages among T. simiae trypanosomes in the proboscis. To date, T. brucei is the only tsetse-transmitted trypanosome with experimentally proven capability to undergo sexual reproduction, which occurs in the fly salivary glands [31]. By analogy, sexual stages of T. simiae or T. congolense are predicted to occur in the proboscis, where the corresponding portion of the developmental cycle, involving proliferation of epimastigotes and the generation of infective metacyclics, takes place. However, no obvious sexual stages were observed in a previous detailed investigation of the T. congolense life cycle in tsetse [18], though population genetics evidence points to sexuality in T. congolense [16] and hybrids have been identified in the tsetse proboscis in laboratory crosses (Peacock et al., unpublished).

Amongst the various trypanosome morphotypes present in the proboscis were cells with a large posterior nucleus and two anterior kinetoplasts, the characteristic morphological features of trypanosomes in Meiosis I in T. brucei, as determined by expression of three meiosis-specific proteins in the nucleus [13]. Our attempt to verify the identity of these putative meiotic dividers in T. simiae in a similar way by expression of a YFP-tagged version of HOP1 unfortunately failed, so the identification currently rests on morphology alone. However, some supporting evidence is provided by the co-occurrence, alongside the meiotic dividers in the proboscis, of putative gametes and other meiotic intermediates, again identified by morphological similarity to these sexual stages in T. brucei [14, 15].

Experimental verification that the T. simiae morphotypes observed are truly sexual stages is essential. Previous experiments to elucidate the sexual cycle of T. brucei have relied on genetically modified parental lines to facilitate the identification of hybrid progeny [31,32,33] rather than the needle-in-haystack approach used initially [12]. The demonstration that T. simiae can be genetically manipulated using tools designed for T. brucei will facilitate similar experiments involving genetically modified lines.

The development of the subgenus Nannomonas trypanosomes T. simiae and T. congolense in tsetse differs profoundly from that of T. brucei in several respects. In all three species, procyclics proliferate in the midgut and colonise the proventriculus, but the population of proventricular forms in subgenus Nannomonas is uniform and non-dividing [18], whereas in T. brucei proventricular forms undergo an asymmetric division into one short and one long epimastigote [30, 34]. Thus, the differentiation to epimastigote forms occurs after migration to the mouthparts in T. simiae and T. congolense, rather than before and during migration as in T. brucei. This difference allows the differentiation of proventricular forms of T. simiae and T. congolense to be followed in vitro. In our previous investigation of this process in T. congolense, we found that almost as soon as the long proventricular forms were put into culture medium, they attached to the substrate and then underwent dramatic shortening before starting an asymmetric division to produce a small daughter cell [25]. A similar pattern of development was observed here for T. simiae, though initial attachment took slightly longer and shortening of the cell body and flagellum did not result in formation of a depot of paraflagellar rod protein, thought to result from the rapid deconstruction of the flagellum [25].

The developmental cycles of all three trypanosome species culminate in the production of infective metacyclics that can be inoculated into the next host with saliva as the fly feeds. In T. brucei this is achieved by direct colonisation of the salivary glands via the hypopharynx (the tube that carries saliva from the glands to the tip of the proboscis), whereas in T. simiae and T. congolense the internal lumen of the proboscis is colonised first before invasion of the hypopharynx. Thus, trypanosomes of all three species need to get into the hypopharynx, presumably via its open end, which is surrounded by finger-like projections [35]. The biological function of the finger-like projections is unknown, but could be to guard against passive entry of microorganisms. In contrast, trypanosomes are motile and in T. brucei it is the migratory life cycle stages from the proventriculus—asymmetrically dividing epimastigotes, and long and short epimastigotes—that invade the hypopharynx; these stages have been recovered from both the proventriculus and salivary exudate during early colonisation of the salivary glands [30, 36]. For T. congolense and T. simiae, it is more difficult to identify the invasive life cycle stages, because salivary exudate will contain a heterogeneous mixture of trypanosomes from the hypopharynx and food canal. Trypanosomes within the hypopharynx can be viewed in live dissected material, and here mostly small trypanosomes, possibly metacyclics, were observed (Additional file 2: Movie S2). Determining the identity of the invasive stage awaits the development of appropriate stage-specific markers.

Conclusions

To date, T. brucei is the only tsetse-transmitted trypanosome with experimentally proven capability to undergo sexual reproduction, which occurs in the fly salivary glands. By analogy, sexual stages of T. simiae or T. congolense are predicted to occur in the proboscis, where the corresponding portion of the developmental cycle takes place. While no such stages have been observed in T. congolense, for T. simiae putative sexual stages were abundant in the tsetse proboscis. Although our initial attempt to demonstrate expression of a YFP-tagged, meiosis-specific protein was unsuccessful, the future application of transgenic approaches will facilitate the identification of meiotic stages and hybrids in T. simiae.