The life cycle of Trypanosoma (Nannomonas) congolense in the tsetse fly
© Peacock et al.; licensee BioMed Central Ltd. 2012
Received: 30 January 2012
Accepted: 7 June 2012
Published: 27 June 2012
Skip to main content
© Peacock et al.; licensee BioMed Central Ltd. 2012
Received: 30 January 2012
Accepted: 7 June 2012
Published: 27 June 2012
The tsetse-transmitted African trypanosomes cause diseases of importance to the health of both humans and livestock. The life cycles of these trypanosomes in the fly were described in the last century, but comparatively few details are available for Trypanosoma (Nannomonas) congolense, despite the fact that it is probably the most prevalent and widespread pathogenic species for livestock in tropical Africa. When the fly takes up bloodstream form trypanosomes, the initial establishment of midgut infection and invasion of the proventriculus is much the same in T. congolense and T. brucei. However, the developmental pathways subsequently diverge, with production of infective metacyclics in the proboscis for T. congolense and in the salivary glands for T. brucei. Whereas events during migration from the proventriculus are understood for T. brucei, knowledge of the corresponding developmental pathway in T. congolense is rudimentary. The recent publication of the genome sequence makes it timely to re-investigate the life cycle of T. congolense.
Experimental tsetse flies were fed an initial bloodmeal containing T. congolense strain 1/148 and dissected 2 to 78 days later. Trypanosomes recovered from the midgut, proventriculus, proboscis and cibarium were fixed and stained for digital image analysis. Trypanosomes contained in spit samples from individually caged flies were analysed similarly. Mensural data from individual trypanosomes were subjected to principal components analysis.
Flies were more susceptible to infection with T. congolense than T. brucei; a high proportion of flies infected with T. congolense established a midgut and subsequent proboscis infection, whereas many T. brucei infections were lost in the migration from foregut to salivary glands. In T. congolense, trypomastigotes ceased division in the proventriculus and became uniform in size. The trypanosomes retained trypomastigote morphology during migration via the foregut to the mouthparts and we confirmed that the trypomastigote-epimastigote transition occurred in the proboscis. We found no equivalent to the asymmetric division stage in T. brucei that mediates transition of proventricular trypomastigotes to epimastigotes. In T. congolense extremely long epimastigotes with remarkably elongated posterior ends were observed in both the proboscis and cibarium; no difference was found in the developmental stages in these two organs. Dividing trypomastigotes and epimastigotes were recovered from the proboscis, some of which were in transition from trypomastigote to epimastigote and vice versa. It remains uncertain whether these morphological transitions are mediated by cell division, since we also found non-dividing cells with a variously positioned, juxta-nuclear kinetoplast.
We have presented a detailed description of the life cycle of T. congolense in its tsetse fly vector. During development in the fly T. congolense shares a common migratory pathway with its close relative T. brucei, culminating in the production of small metacyclic trypanosomes that can be inoculated with the saliva. Despite this outward similarity in life cycle, the transitional developmental stages in the foregut and mouthparts are remarkably different in the two trypanosome species.
Trypanosomes transmitted by tsetse in Africa cause diseases of importance to the health of both humans and livestock. Of the livestock trypanosomes, Trypanosoma (Nannomonas) congolense is probably the most prevalent and widespread pathogenic trypanosome in tropical Africa, being found in ruminants, pigs, dogs and other domestic animals throughout the tsetse belt . In the mammalian bloodstream T. congolense is a small trypanosome, shorter in length than T. brucei and without a conspicuous undulating membrane. In the tsetse fly vector, both species develop and multiply in the midgut initially, before onward migration to the mouthparts; infective metacyclics develop in the proboscis for T. congolense and in the salivary glands for T. brucei. This difference in vector developmental site led to the subgeneric classification in current use for the African tsetse-transmitted trypanosomes: trypanosomes that develop in the midgut and proboscis are in subgenus Nannomonas, while those that use the midgut and salivary glands are in subgenus Trypanozoon. In phylogenetic analyses of trypanosome species, these two subgenera appear as sister groups within the clade of African tsetse-transmitted salivarian trypanosomes [3, 4].
The life cycles of T. congolense and T. brucei were described in detail in the last century (reviewed by ). For T. brucei, our understanding of the developmental cycle in the fly has been steadily augmented by information emerging from molecular and cell biology analyses: e.g. differentiation of bloodstream to procyclic forms [5, 6]; major surface glycoproteins of developmental stages [7–10]; cell cycle control ; genetic exchange and meiosis [12, 13]. Much less is known about the life cycle of T. congolense, but evidence to date suggests that many aspects of the developmental pathway in tsetse are similar to that of its close relative, T. brucei. In both species bloodstream forms differentiate to procyclics in the midgut and lose the variant surface glycoprotein (VSG) coat, but the major surface molecules of T. congolense are carbohydrates rather than glycoproteins as in T. brucei[14, 15]. From the midgut both species migrate anteriorly to reach the mouthparts, using the proventriculus or cardia (the valve separating the midgut from the foregut) as a staging post. Here T. brucei undergoes an asymmetric division that yields one short and one long epimastigote, but this dividing stage has not been described in T. congolense. The short epimastigote is crucial for T. brucei, as it goes on to invade and colonise the salivary glands [11, 16, 17]. The equivalent stage that founds the proboscis infection in T. congolense is reported to be a trypomastigote and transformation of trypomastigotes to epimastigotes occurs after attachment in the labrum of the proboscis . In addition, attached trypanosomes are found in the adjacent cibarium , but it is not clear if these are an extension of the proboscis population or a separate stage of development. In both species the attached epimastigotes proliferate and subsequently differentiate into infective metacyclics that are preadapted for life in the mammalian host by their protective VSG coat [19, 20]. In T. brucei metacyclics are produced in the salivary glands, while in T. congolense they develop in the labrum and hypopharynx [21, 22].
While attempts to produce metacyclic T. brucei in vitro have met limited success [23, 24], the developmental cycle of T. congolense can be reliably reproduced in vitro and cultures yield large numbers of trypanosomes of different life cycle stages [25–28]. The crucial difference lies in the ability of T. congolense epimastigotes to attach to a plastic surface, proliferate and subsequently differentiate into metacyclics, just as they do in vivo. T. brucei epimastigotes probably need live cells for attachment, because in vivo there is intimate contact between outgrowths of the flagellar membrane with cells of the tsetse salivary gland epithelium . EM studies show that attachment of T. congolense epimastigotes is via hemidesmosomes both in vitro and in vivo[20, 22, 29]. Comparison of shaken and unshaken cultures showed that attachment is not necessary for epimastigote division but is a prerequisite for differentiation into metacyclics . The question whether the in vitro produced life cycle stages of T. congolense represent those produced in the fly has been addressed by comparison of morphology at the light and ultrastructural levels, and immunocytological analysis of the VSG coats of metacyclics [20, 26, 31]. In addition, the expected stage-specific cell surface markers were expressed by each of three life cycle stages (procyclics, epimastigotes and metacyclics) cultured in vitro.
The recent publication of the genome sequence of T. congolense, together with interest in using it as a convenient in vitro proxy to access the complete developmental cycle of the model trypanosome, T. brucei, prompts re-investigation of the T. congolense life cycle. This has been dealt with rather cursorily in the literature, probably because of its similarity to that of T. brucei, and details of transitional forms are sparse [21, 34, 35]. A key question is whether there is a form equivalent to the asymmetric divider of T. brucei[11, 16, 17] that has been overlooked. Here we have examined a detailed timecourse of the development of T. congolense in the tsetse fly to provide a comprehensive and illustrated reference to the stages in its life cycle.
Experimental tsetse flies were from the Bristol laboratory colony of Glossina morsitans morsitans originally from Zimbabwe. Flies were kept at 25°C and 70% relative humidity, and fed on sterile defibrinated horse blood supplemented with 2.5% w/v bovine serum albumen (Sigma A4503)  and 1 mM dATP  via a silicone membrane. Male and female flies were used for experiments, being given the infective bloodmeal for their first feed 24–48 hours post-eclosion. The infective bloodmeal contained approximately 8 x 106 trypanosomes ml-1 in sterile horse blood supplemented with either 60 mM N-acetyl-glucosamine (NAG)  or 10 mM L-glutathione  to increase infection rates. The infective bloodmeal for flies dissected at 2–3 days was made with sterile horse serum to aid the visualisation of trypanosomes. For examination of trypanosomes extruded in spit samples (a mixture of saliva and regurgitated foregut contents), flies were caged individually; for other experiments, flies were caged in groups of 15–25. Bloodstream form trypanosomes of T. congolense savannah 1/148 (MBOI/NG/60/1-148)  were grown in mice and used to infect flies. T. b. brucei J10 (MCRO/ZM/73/J10 [clone 1]) was used for comparison.
Flies were dissected 2 to 78 days after infection. Alimentary tracts, from the proventriculus to the rectum, were dissected in a drop of phosphate buffered saline (PBS) and viewed as wet mounts under phase contrast (x100 magnification) to search for trypanosomes. Proventriculi were removed from infected midguts and viewed separately. Proboscides from flies with midgut infection were dissected into a separate drop of PBS and teased apart, gently rubbing a fine needle down the length of the proboscis to dislodge the trypanosomes. Cibaria from flies with midgut infection were dissected into a separate drop of PBS. Trypanosomes were fixed in 2% w/v paraformaldehyde in PBS for 15 minutes in a humid chamber, followed by three PBS washes. After brief drying, preparations were stained using 4’,6-diamidino-2-phenylindole (DAPI) in Vectashield (Vector labs) mounting medium and viewed within 1 hour. Bloodstream form trypanosomes obtained from a mouse infected with T. congolense 1/148 (3 day infection) were fixed and stained in the same way.
Spit samples were obtained from individually caged flies essentially as described by . Flies were starved for approximately 48 hours before being allowed to probe onto an alcohol-cleaned microscope slide on a heating plate held at approximately 37°C; flies were fed once they had probed, or after a maximum of 30 minutes. Flies were probed on alternate days commencing 10 days after the infected feed. Saliva samples dried immediately on contact with the microscope slide and slides were stored in the dark at ambient temperature before examination. The samples were checked for the presence of trypanosomes under phase contrast (100x magnification); positive slides were fixed for 30 s in methanol, then stained and mounted with DAPI in Vectashield as above.
Images were recorded 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 400x magnification. Measurements were made on the digital images using Image J software (Version 1.41) (http://rsb.info.nih.gov/ij/). Dimensions measured were those used by  and shown in Additional file 1: Figure S1.
Since many of the dimensions measured were likely to be internally correlated to some extent, the entire dataset was subjected to principal components analysis using the princomp procedure from the statistical package R (http://www.r-project.org/) to extract underlying latent variables. Only three uncorrelated factors were identified with eigenvalues greater than 1, accounting for 46.3 (factor 1), 26.1 (factor 2) and 13.4% of the observed variance. Extracted scores for factors 1 and 2 for each trypanosome were plotted. Loadings were extracted and the absolute values plotted (Additional file 2: Figure S2) to determine the extent to which each of the individual measurements contributed to the first two factors. See legend to Additional file 2: Figure S2 for details.
Trypanosome infection rates in Glossina morsitans morsitans
T. congolense (Tc)
T. b. brucei (Tbb)
No. of flies dissected
No. of infected midguts
No. of infected proboscides (Tc) or salivary glands (Tbb)
Transmission index (TI)
No. of midguts with positive spit samples
No. of proboscides (Tc) or salivary glands (Tbb) with positive spit samples
Midgut procyclic trypanosomes gradually lengthened over time, from a mean length of 20.0 ± 0.4 μm on day 2 to 40. 9 ± 0.4 μm on day 17, i.e. doubling in length (Figures 1B to E; Additional file 3: Table S1). At the same time, the trypanosomes became more slender, halving in width from 2.6 ± 0.1 μm on day 2 to 1.3 ± 0.1 μm on day 17. Corresponding changes were seen in nuclear length and width, the nucleus becoming longer and thinner (Figures 1B to E; Additional file 3: Table S1). Keeping the kinetoplast as the point of reference, most of the increase in cell length appeared to occur at the anterior end of the cell: the kinetoplast-anterior distance increased from 17.0 ± 0.3 μm on day 2 to 35.7 ± 0.3 μm by day 17, compared to a modest increase in the kinetoplast-posterior distance of <3 μm. The relative distance between the nucleus and kinetoplast remained constant (Additional file 3: Table S1). Assuming that the increase in cell length results from posterior extension of microtubules as in T. brucei, then the apparent growth of the anterior end of the cell in T. congolense procyclics actually results from gradual movement of the kinetoplast and nucleus towards the posterior. It is important to note that at any one timepoint the trypanosome population was not of uniform morphology, but showed large variability both in length and shape (Additional file 3: Table S1). For example, in some trypanosomes the posterior tapered to a point, while in others the posterior was blunt.
Proliferation of midgut and proventricular trypomastigotes
1 K 1 N
2 K 1 N
2 K 2 N
Proportion of 2 K trypanosomes
Early midgut (days 2–7)
Late midgut (days 12–26)
Early proventriculus (days 6–12)
Late proventriculus (days 13–26)
T. congolense proboscis infections were detected as early as day 13 by dissection, before the first appearance of epimastigotes in spit samples on day 15. This accords with previous findings that the epimastigotes arise in the proboscis, not the proventriculus or foregut [21, 34]. Relatively few epimastigotes were observed in spit samples and these were highly variable in length (n = 17; Additional file 4: Table S2). Epimastigotes in the process of division (2 K 1 N) were also seen (Figure 5C). The morphology of epimastigotes in the spit and proboscis was similar. The posterior was sometimes extremely elongated, such that the nucleus was positioned equidistant between the posterior and anterior poles of the cell (Figure 5A and B); the elongated posterior sometimes twisted during fixation (Figure 5B and C). In T. brucei, procyclics with the so-called “nozzle phenotype”, which have a similarly elongated posterior end, have been produced in vitro by perturbing expression of single genes, e.g. by overexpression of the zinc finger CCCH motif protein tbZFP2  or by knockdown of cyclin CYC2 . Whether the nozzle phenotype is analogous in T. brucei procyclics and T. congolense epimastigotes, these experiments serve to demonstrate that relatively minor changes in gene expression are sufficient to produce the gross phenotypic changes in trypanosome length and morphology seen here.
The close proximity of the kinetoplast to the nucleus in the epimastigotes was associated with a bulge in the cell near the nucleus, particularly in dividing cells (Figures 5A-C); this was also observed for some proventricular trypomastigotes where the kinetoplast and nucleus were juxtaposed, but in that case the kinetoplast was posterior not anterior to the nucleus (Figure 3B).
Metacyclics first appeared in spit samples on day 21, identified by their very short length (12.8 ± 1.3 μm; Additional file 4: Table S2) and characteristic S-shape  (Figure 5D). The metacyclics were significantly shorter and thinner than BSF (Additional file 3: Tables S1, Additional file 4: Table S2); this can be clearly seen in the PCA plot in which metacyclics and BSF cluster separately (Figure 2). In metacyclics, the kinetoplast was very close to the posterior pole of the cell (Figure 5D; Additional file 4: Table S2).
The cibarium is a widening of the alimentary tract that lies between the proboscis and foregut; the chitin-lined walls act as a pump allowing the fly to suck blood . Both T. congolense and T. vivax are found attached to the cibarial walls . Here the cibarium was examined for trypanosomes when flies were dissected at 77 days. Trypanosomes were found widely distributed across the dorsal wall as described  (Additional file 8: Movie 3). The majority of cibarial trypanosomes were epimastigotes (144/158, 91%), the remainder being trypomastigotes (Additional file 5: Table S3). The morphology of these cells was similar to that of trypomastigotes and epimastigotes in the proboscis, suggesting that the cibarium harbours an extension of the proboscis population rather than a separate morphological stage. This is clearly seen in the PCA comparing trypanosomes from the cibarium and proboscis; there is considerable overlap between these two groups (Figure 2).
Measurement of proboscis epimastigotes
23.0 ± 1.4 a
23.4 ± 1.1 a
29.0 ± 1.5 ab
30.8 ± 1.0 b
2.2 ± 0.1 a
2.2 ± 0.1 a
2.1 ± 0.6 a
2.1 ± 0.1 a
Kinetoplast to posterior
6.8 ± 0.7 a
8.6 ± 0.6 a
14.4 ± 1.2 b
17.5 ± 0.8 b
Nucleus to posterior
6.6 ± 0.7 a
7.4 ± 0.5 a
12.0 ± 1.2 b
14.0 ± 0.8 b
2.9 ± 0.2 a
2.7 ± 0.1 a
2.8 ± 0.1 a
2.9 ± 0.1 a
1.2 ± 0.1 a
1.2 ± 0.1 a
1.3 ± 0.1 a
1.2 ± 0.1 a
Nucleus to anterior
16.4 ± 1.3 a
16.0 ± 0.9 a
16.9 ± 0.6 a
16.9 ± 0.4 a
Kinetoplast to anterior
16.3 ± 1.2 a
14.8 ± 0.9 a
14.6 ± 0.6 a
13.3 ± 0.4 a
Thus the process of differentiating from trypomastigote to epimastigote is radically different in the two species. In T. brucei this is achieved by an asymmetric division of the proventricular trypomastigote; the kinetoplast migrates round the nucleus towards the anterior before dividing, so that two epimastigote daughter cells are formed when the nucleus divides . We found no evidence of such a well orchestrated transition in T. congolense. A range of long trypanosomes with the kinetoplast positioned adjacent to the nucleus were recovered from the proboscis; by subdividing these according to the relative positions of the kinetoplast and nucleus, we infer that the transition from trypomastigote to epimastigote involves both migration of the kinetoplast around the nucleus to an anterior position and elongation of the posterior of the trypanosome. In the longest epimastigotes, the anterior kinetoplast was well separated from the nucleus. The re-positioning of the kinetoplast relative to the nucleus and remodelling of the cell posterior may occur without cell division, but we also found examples of dividing forms in transition from epimastigotes to trypomastigotes and vice versa.
Seen by scanning electron microscopy (SEM), attached epimastigotes of T. brucei are of fairly uniform size and possess an elongated thin posterior that protrudes into the lumen of the salivary gland . In contrast, T. congolense epimastigotes are highly variable in morphology. Some are spectacularly long with elongated twisted or transparent posterior extensions, much broader than the thin, pointed “nozzles” of T. brucei epimastigotes. Others appear truncated, the posterior of the cell with a blunt rather than tapering end, almost as if the end has been broken off or folded back on itself. Such truncated forms have been described by several workers [20, 21, 31]; both truncated and elongated forms are visible in SEM images of in vitro cultivated epimastigotes . Here truncated forms were observed as live, motile cells and they are clearly not an artefact of fixation.
The extreme length of T. congolense epimastigotes prompts questions of how they differentiate into metacyclics, which involves a gross reduction in cell length from about 30 μm to 13 μm. In cultures of long attached epimastigotes, metacyclics only appeared if short, attached trypomastigotes were present, suggesting that these are a transitional stage . In vivo, VSG-coated trypanosomes were found in both the hypopharynx and occasionally the labrum by transmission EM , supporting the view that differentiation to metacyclics occurs predominantly in the hypopharynx [2, 21]. Here, the hypopharynx trypanosomes were distinguishable in live dissected material, but were mixed with those from the labrum during fixation and staining; attached cells were probably under-represented. In such preparations both long and truncated epimastigotes were observed in division, as well as trypomastigotes. As far as could be discerned, these divisions rarely resulted in large reduction in length of the daughter cell, although in the early stage of division before the cleavage furrow develops, it is not possible to judge the length of the posterior of the daughter cell. The short dividing trypomastigotes of about 20 μm in length were assumed to be pre-metacyclics and fit the description of the transitional stage observed in vitro. As far as can be ascertained, in vitro production of T. congolense metacyclics follows the same developmental pathway as in the fly [26, 28, 30, 31]; however, there may be subtle differences in terms of gene expression, for example of surface molecules, that will only be evident from detailed investigation.
Both T. congolense and T. brucei share the same migration route in the tsetse fly from the midgut to the mouthparts via the proventriculus and foregut, but the two strains used here evidently differ in their ability to complete the developmental cycle in G. m. morsitans. Established midgut infections led to far more proboscis infections in T. congolense than salivary gland infections in T. brucei (transmission indices of 93% and 25% respectively). Invasion of the foregut was equally efficient in both species, as judged by the percentage of spit-positive/midgut-positive flies (81% v. 83%). While most of these foregut infections led to the successful invasion and colonisation of the proboscis by T. congolense, relatively few salivary gland infections were established by T. brucei (100% v. 30%). Surface coat proteins and carbohydrates are thought to play a protective role against insect innate immune responses , so although these molecules differ between the two species, they evidently offer equivalent levels of protection while the parasites are in the midgut, proventriculus and foregut. The problem comes when the parasites reach the mouthparts: for T. congolense it is clearly quite easy to settle down and proliferate in the proboscis, but the migratory forms of T. brucei pass through this region to enter the salivary glands via the hypopharynx. The attrition rate is high and it seems that only a few trypanosomes initiate the infection [41, 49]. The evolutionary driver of this strategy in T. brucei was presumably competition for space and nutrients, because several different trypanosome species attach in the proboscis (T. congolense T. vivax T. simiae and T. godfreyi). By-passing this region opens the greater surface area of the salivary glands for colonisation.
The trypanosome strain used here represents the savannah subgroup of T. congolense. Within subgenus Nannomonas, there are an additional two subgroups of T. congolense (forest and Kenya Coast or kilifi), plus T. simiae and T. godfreyi and related trypanosomes . Comparative analysis will show how far this description of the life cycle of T. congolense savannah generalizes to the whole subgenus.
We have presented a detailed description of the life cycle of T. congolense in its tsetse fly vector. This comprehensive account has allowed comparison with the better known life cycle of T. brucei. These related trypanosomes share a common migration pathway during development in the fly, involving the establishment of infection in the ectoperitrophic space of the midgut and invasion of the proventriculus. After this, the transitional developmental stages in the foregut and mouthparts are remarkably different, before the life cycles converge again to culminate in the production of infective metacyclics.
We thank the International Atomic Energy Agency, Vienna for experimental tsetse flies. This work was funded by grants from The Wellcome Trust (064436, 079375).
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.