Species-Specific Adaptations of Trypanosome Morphology and Motility to the Mammalian Host

Abstract

African trypanosomes thrive in the bloodstream and tissue spaces of a wide range of mammalian hosts. Infections of cattle cause an enormous socio-economic burden in sub-Saharan Africa. A hallmark of the trypanosome lifestyle is the flagellate's incessant motion. This work details the cell motility behavior of the four livestock-parasites Trypanosoma vivax, T. brucei, T. evansi and T. congolense. The trypanosomes feature distinct swimming patterns, speeds and flagellar wave frequencies, although the basic mechanism of flagellar propulsion is conserved, as is shown by extended single flagellar beat analyses. Three-dimensional analyses of the trypanosomes expose a high degree of dynamic pleomorphism, typified by the 'cellular waveform'. This is a product of the flagellar oscillation, the chirality of the flagellum attachment and the stiffness of the trypanosome cell body. The waveforms are characteristic for each trypanosome species and are influenced by changes of the microenvironment, such as differences in viscosity and the presence of confining obstacles. The distinct cellular waveforms may be reflective of the actual anatomical niches the parasites populate within their mammalian host. T. vivax displays waveforms optimally aligned to the topology of the bloodstream, while the two subspecies T. brucei and T. evansi feature distinct cellular waveforms, both additionally adapted to motion in more confined environments such as tissue spaces. T. congolense reveals a small and stiff waveform, which makes these parasites weak swimmers and destined for cell adherence in low flow areas of the circulation. Thus, our experiments show that the differential dissemination and annidation of trypanosomes in their mammalian hosts may depend on the distinct swimming capabilities of the parasites.

Fig 1. Defined motility patterns characterise the behaviour of trypanosome species in the blood of different hosts.

A) Representative examples of three motility types visualised by single cell trajectories. Videos were captured with a frame rate of 500 fps. The swimming paths were traced (white lines), (S1–S5 Videos). Persistent swimmers were defined as cells that swam without a tumbling phase for the duration of a single on-chip recording period (16 s) of the high speed camera. Within this time, intermediate swimmers exhibited at least one tumbling phase. This is a period of two seconds or longer, in which the cells do not leave a circular area with a diameter of 25 μm (white circle). Tumbling cells stay within this area for the complete observation period of 16 s. B) Comparative motility analysis of T. vivax and T. congolense in blood freshly harvested from infected mouse (m), rat (r) and sheep (s). The host animals revealed a comparable parasitaemia in the range of 10⁷ trypanosomes/ml. For comparison, parasites grown in mice were incubated in sheep blood (sb). C) Comparison of the locomotion behaviour of four trypanosome species in mouse blood. All parasites were isolated from infected mice and analysed in fresh neat mouse blood. At least 300 cells were analysed per infection and the motility types were scored as shown in (A).

Fig 2. Trypanosome velocities in mouse blood.

Each bar represents the analysis of 100 high-speed videos (500 fps, 16 s recording time). The bars present the mean swimming speed +/- SD. Black lines depict the individual maximum and minimum speeds recorded. The maximum speeds for persistent swimmers are annotated. The average population swimming speed of a species was calculated from all 300 trajectories and is marked by the blue dot.

Fig 3. Representative examples of trypanosomes swimming in blood from different host animals.

The images are stills of the corresponding S6 Video. Videos were captured with a frame rate of 500 fps. The swimming paths were traced using the Fiji plugin MTrackJ [26]. The trajectories of one example each of tumbling (red), intermediate (yellow) and persistent (green) swimmers are shown. The numbers are maximum speeds (n = 20).

Fig 4. High-resolution single cell analysis of persistently swimming trypanosomes.

The images on the left are single frames of the corresponding S8–S16 Videos, showing the cells at the beginning of an analysed flagellar beat, while two later positions of the parasites are shown as white outlines to depict the trajectory of the cell. The speed (v) is the highest average velocity reached during the video sequence analysed. The panel to the right presents the trypanosome outlines from each recorded frame for one complete flagellar beat. The successive image frames were stacked along the time axis in a three-dimensional surface representation. This allows the visualisation of the cellular waveform produced by the beating flagellum and the attached cell body. The anterior tip of the flagellum is marked in blue. The duration of the complete flagellar beat is annotated on the t-axis (ms). The right panel reveals the same 3D-surface representation, however, turned to view on to the anterior tip of the flagellum (along the x-axis of movement). This illustrates the sinusoidal oscillation of the flagellar tip, starting the travelling wave running along the body to the posterior end. The average frequency (Hz) of the flagellar beats of the analysed video is given on the right (see Materials and Methods). A) A representative slim waveform of T. vivax IL 1392 swimming forward with continuous tip-to-base beats in a wet blood film from the mouse (S8 Video). This waveform represents a small subset (1%) within the parasite population and swims significantly faster than cells exhibiting normal waveforms. B) Normal waveform of T. vivax IL 1392 in mouse blood from S9 Video, swimming with a frequency similar to that of the fast waveform in A) but reaching less than half the speed. C) T. vivax IL 2136 in mouse blood reveals slower motion when compared to the IL 1392 strain (S10 Video). D) T. brucei ILTat 1.4 swims with a lower beat frequency than normal waveform T. vivax cells in B) but reaches similar speeds (S11 Video). E) T. evansi KETRI 2479 swimming in mouse blood (S12 Video), with an intermediate beat frequency. The T. evansi waveform typically appears more elastic and curly when compared to the other trypanosomes, propelling the cells to relatively high velocities along curvy trajectories. F) T. congolense IL 1180 swimming with characteristic low frequency beats (S13 Video). The anterior tip shows far shallower oscillations due to the missing free anterior part of the flagellum and the stiff cell body causes a waveform clearly less effective for propulsion than that of the other species. Nevertheless the cell shows persistent forward swimming periods with speeds of around 20 μm/s. G) T. congolense KETRI 3827 swimming in mouse blood (S14 Video), with a lower frequency than strain IL 1180, but with an apparently more effective waveform. H) T. congolense KETRI 3827 from infected sheep (S15 Video) beat with the same frequency as in mouse blood, but with significantly reduced forward swimming speed, showing the influence of the specific host environment. I) T. vivax IL 1392 from infected sheep (S16 Video) swim slower and with decreased flagellar beat frequency when compared to cells in mouse blood (B).

Fig 5. Three-dimensional modelling of trypanosome cells reveals characteristic morphologies correlating to the species-specific motility behaviour.

The parasites were harvested from infected mice and the cell surface was fluorescently labelled with sulfo-NHS dyes. On the left of each panel representative surface-rendered models of fluorescently surface-labelled trypanosome cells are shown. The fluorescence labelling allows the simultaneous visualisation of the cell body and the flagellar membrane. The complete trace of the flagellum (yellow) attached along the cell body (grey) is shown in the three-dimensional representation. The model was orientated to show the view onto the posterior tip, allowing to evaluate the course of the flagellar attachment along the cell body, starting at the exit from the flagellar pocket at the top of the cell and tracing it towards the anterior end. A selection of 3D-models is presented to illustrate the variety of shapes the trypanosomes adopt, based on their cellular waveform. The numbers present the average length of the trypanosomes L_av (n = 100) and the average maximum width of the cells W_av (n = 100) as measured in video stills (see Fig 1). The actual length L, volume V and cells surface area A of the surface-rendered models are given below in grey letters. The first cell in each row was chosen for surface-modelling. (A) Examples of normal waveform T. vivax IL 1392 cells. The flagellum is attached in a shallow 180° right hand turn. (B) T. vivax IL 2136 normal waveforms. The flagellum exhibits a similar course to IL 1392 running along a slightly stiffer cell body. (C) The T. brucei ILTat 1.4 strain typically shows a more prominent 180° right handed turn around the posterior third of the cell body. (D) The pleomorphic T. brucei AnTat 1.1 strain slender and stumpy forms. The three cells on the right are short stumpy trypanosomes that are characterised by the absence of a free part of the flagellum. The slender forms show a similar 180° right hand turn when compared to ILTat 1.4 in C). Morphometric data are for the slender stage (E) T. evansi reveals a curlier waveform than T. brucei while the flagellum turns completely (360°) around the cell body.(F) T. congolense IL 1180 cells are small with a stiff cell body and a flagellum running along it in a relatively straight course. (G) T. congolense KETRI 3827 is larger than the IL 1180 strain, but reveals the same characteristic waveform.

Fig 6. Extended single flagellar beat analysis of trypanosome motility in wet mouse blood.

The graphs plot the velocities produced by single, consecutive flagellar beats (red dots), together with the beat frequency (black dots). The velocities, which were derived from measurement of the translocation of the posterior tip after each flagellar beat (white marks in S8–S16 Videos), are averaged over 5 beats to generate the average directional speeds (blue dots). The red number is the overall observation period (ms). A) Fast type of a T. vivax IL 1392 persistent swimmer with a slim waveform (S8 Video), showing a 30 ms—stop upon mechanical hindrance. The flagellar tip directly hit an erythrocyte, whereupon the trypanosome stopped without flagellar beating being interrupted. Translocation resumed with the next flagellar beat. Note that neither beat frequency nor average velocity (blue line) were markedly changed by this stop. B) Example of a T. vivax IL 1392 normal waveform swimmer revealing a short (< 1 second) period characterised by several base-to-tip beats, resulting in backward movement, i.e. negative speed (S9 Video). C) T. vivax IL 2136 exhibited lower beat frequencies than the IL 1392 strain (S10 Video). D) An intermediate T. brucei ILTat 1.4 swimmer exhibited two persistent swimming stretches interrupted by a short tumbling phase. This was followed by a period of beat reversal and backward swimming (S11 Video). E) A persistently swimming T. evansi parasite (S12 Video) reducing velocity by half, while the flagellar beat frequency remained in a constant range. (F) T. congolense IL 1180 showed a short period of rather fast backward motion (S13 Video). (G) T. congolense KETRI 3827 switching between fast forward and slow backward movement (S14 Video). H) T. congolense KETRI 3827 isolated from infected sheep revealing persistent slow forward motion (S15 Video).

Fig 7. Trypanosomes show species-specific changes in motile behaviour in response to viscosity changes.

A) The effect of viscosity on the proportion of swimming trypanosomes. The parasites were isolated from infected mice and resuspended in fresh trypanosome dilution buffer (TDB) in the absence or presence of methylcellulose (0.2, 0.4, 0.6 and 0.8%), generating medium viscosities between 1 mPa·s (TDB) and 35 mPa·s (0.8% methylcellulose). For each condition the trajectories of 300 parasites were analysed. B) The viscosity of the micro-environment influences the average persistent swimming speed of trypanosomes in a species-dependent manner (n ≥ 100; data are means ± SD). The trypanosomes were incubated in TDB-buffer supplemented with or without methylcellulose. While T. vivax motility was negatively affected by conditions above blood viscosity (i.e. 4 mPa·s), T. brucei and T. evansi parasites swam faster even in very viscous medium. T. congolense motion was not consistently influenced by viscosity. C) The influence of increasing viscosity on trypanosome swimming speed in mouse blood (n ≥ 100; data are means ± SD). Increased viscosity in the presence of blood cells led to reduction of speeds in T. vivax. In contrast, average speeds of T. brucei and T. evansi increased with increasing viscosity. Average speeds of T. congolense parasites did not vary significantly with increased media viscosity.

Fig 8. Motile trypanosomes are accelerated by arrays of obstacles mimicking the cellular environment of the bloodstream.

The parasites were harvested from infected mice and incubated in PDMS pillar arrays of varying geometries (n ≥ 200/ species). Swimming velocities of persistently swimming cells were measured as before (see Fig 6) and the maximum speeds determined. The grey shades indicate the range of swimming speeds measured in other experiments, i.e. in high viscosity media or wet blood films. The comparison allows the visualisation of maximal swimming speed increase as a function of defined pillar parameters. T. brucei and T. evansi swim most efficiently in pillars of 8–10 μm diameter with 4μm spacing, while T. vivax reveals very high maximum speeds in even narrower pillar arrays. T. congolense was also included in the experiment, however, did not reveal any persistent motion in the pillar arrays.