The Intracellular Amastigote of Trypanosoma cruzi Maintains an Actively Beating Flagellum

ABSTRACT Throughout its complex life cycle, the uniflagellate parasitic protist, Trypanosoma cruzi, adapts to different host environments by transitioning between elongated motile extracellular stages and a nonmotile intracellular amastigote stage that replicates in the cytoplasm of mammalian host cells. Intracellular T. cruzi amastigotes retain a short flagellum that extends beyond the opening of the flagellar pocket with access to the extracellular milieu. Contrary to the long-held view that the T. cruzi amastigote flagellum is inert, we report that this organelle is motile and displays quasiperiodic beating inside mammalian host cells. Kymograph analysis determined an average flagellar beat frequency of ~0.7 Hz for intracellular amastigotes and similar beat frequencies for extracellular amastigotes following their isolation from host cells. Inhibitor studies reveal that flagellar motility in T. cruzi amastigotes is critically dependent on parasite mitochondrial oxidative phosphorylation. These novel observations reveal that flagellar motility is an intrinsic property of T. cruzi amastigotes and suggest that this organelle may play an active role in the parasite infection process.

IMPORTANCE Understanding the interplay between intracellular pathogens and their hosts is vital to the development of new treatments and preventive strategies. The intracellular "amastigote" stage of the Chagas disease parasite, Trypanosoma cruzi, is a critical but understudied parasitic life stage. Previous work established that cytosolically localized T. cruzi amastigotes engage physically and selectively with host mitochondria using their short, single flagellum. The current study was initiated to examine the dynamics of the parasite flagellum-host mitochondrial interaction through live confocal imaging and led to the unexpected discovery that the T. cruzi amastigote flagellum is motile.
KEYWORDS kinetoplastida, Trypanosoma cruzi, cilia, flagella, host-pathogen interactions, intracellular parasites, mitochondria T he kinetoplastid protozoan parasite, Trypanosoma cruzi, is the causative agent of human Chagas disease and is associated with significant morbidity and mortality in Latin America (1). The ability of T. cruzi to establish intracellular residence in mammalian host cells, and to persist in diverse host tissues, are critical factors underlying disease progression. In mammals, intracellular infection is established by the nonreplicative, motile trypomastigote stage of T. cruzi (2). Once inside a host cell, trypomastigotes transition to the nonmotile "amastigote" stage, a morphologically and metabolically distinct T. cruzi life stage that proliferates in the cytoplasm of mammalian host cells (3)(4)(5). Accompanying the loss of motility in the amastigote is a dramatic shortening of the flagellum and loss of the paraflagellar rod (6), a lattice-like structure that runs parallel to the axoneme in motile trypanosomatid life stages (7). Intracellular T. cruzi amastigotes retain a short flagellum with a 9 1 2 axonemal structure (8) (Fig. 1A) that extends beyond the opening of the flagellar pocket with access to the extracellular milieu (Fig. 1B). Dynein arms, which typically serve as drivers of flagellar motility (9) in eukaryotes are visible on most doublets (Fig. 1A). Often described as a remnant and assumed to be inert (10), the T. cruzi amastigote flagellum has received little attention. The recent discovery of intimate contact between the flagellum of intracellular T. cruzi amastigotes and host mitochondria (11) has prompted speculation about possible functions of this vestigial amastigote flagellum (10,11).
To gain insight into the spatiotemporal dynamics of the amastigote flagellum-host mitochondrial interaction, we generated transgenic T. cruzi parasites that stably express flagellar-targeted GFP ( Fig. 1C; small myristoylated protein 1-1-GFP [SMP1-1-GFP], Text S1) (12). We first confirmed that the flagellar length and width were similar to that of the parental line (Fig. S1). The SMP1-1-GFP parasites were then used to establish intracellular infection in normal human dermal fibroblasts (NHDF; mitochondrial COX8-mCherry) and live fluorescence confocal imaging was performed at 48 hours postinfection (hpi), a time point at which intracellular amastigotes are in the exponential growth phase (3). Movement of the T. cruzi amastigote flagellum, relative to the parasite body and the host mitochondria, was immediately evident in these videos ( Fig. 1D; Video S1), a time point at which intracellular amastigotes are in the exponential growth phase (3).
Given that flagellar motility had never been described in the intracellular T. cruzi amastigote stage, we performed time-lapse imaging of the parental T. cruzi (Tulahuén) parasites by bright field microscopy (Video S2). These videos confirm that amastigote flagellar motility is independent of flagellar SMP1-1-GFP expression. We also demonstrate that flagellar beating occurs in intracellular T. cruzi amastigotes from genetically diverse backgrounds (Videos S3 to S5) and when resident in human (Fig. 1) or mouse T. cruzi Amastigote Flagellum Beats Actively mBio cells (Video S6). Combined, these findings suggest that the previously unrecognized capacity for flagellar movement may be a universal property of intracellular T. cruzi amastigotes.
To determine the characteristics of flagellar motility in T. cruzi amastigotes, we analyzed a series of time lapse movies (n = 24) from each of three independent infection and imaging experiments. Amira software was employed to create kymographs as three-dimensional (3D) surface models of the fluorescence signal and to track the course of the flagellar tip for a duration of 60 s. This allowed the recording of absolute coordinates for the amplitude of tip movement (Fig. 2A). These data were then used to perform a peak analysis with OriginPro (Fig. 2B). The distance between successive peaks was calculated and plotted to generate precise and normalized displacement measurements for each flagellum independent of the orientation or position of the parasite cell body. The frequency, temporal distribution, and displacement are plotted for each flagellar beat over the observation period of 1 min for each parasite. Data obtained from eight individual parasites are shown (Fig. 2C, all parasites Fig. S2). The aggregated data for all parasites in each biological replicate is shown in Fig. 2D. Although the flagellar beat frequency is highly variable on an individual parasite level, the average flagellar beat frequency is 0.69 6 0.30 Hz. Such quasiperiodic behavior can be expected in flagella shorter than 4 mm (13). Note that all measurements are necessarily confined to the imaging plane and thus quantify the apparent microscopic behavior routinely captured in 2D. We observed clear indications of three-dimensional, rotational movement of the flagella but presume that the current detailed description is sufficient for an initial analysis. Flagellar movement was consistently documented in intracellular T. cruzi amastigotes in the context of mammalian host cell infection. To uncouple amastigote flagellar motility from the potential influences of host cytoskeletal/organellar dynamics, SMP-1-1-GFP amastigotes were imaged following their release from mechanically disrupted host cells. Isolated T. cruzi amastigotes continue to beat their flagellum after separation from host cells, with beat frequencies like those measured for host cell resident amastigotes, albeit slightly faster (;0.2 Hz), likely from decreased viscosity of the medium versus the host cell cytoplasm (Fig. 2E). We also demonstrate that amastigote flagellar motility is severely impaired upon inhibition of amastigote mitochondrial respiration. Exposure of isolated amastigotes to GNF7686, a small molecule inhibitor that targets trypanosomatid cytochrome b (14) and blocks oxidative phosphorylation in amastigotes (15), resulted in severe impairment of flagellar movement with a significant increase in the proportion of parasites with no measurable flagellar movement (Fig. 2F). In line with the demonstration that GNF7686 lacks cytotoxicity toward T. cruzi amastigotes (3) flagellar beat resumed once the compound was washed out (Fig. 2F). Combined, these results demonstrate that flagellar beating is an intrinsic property of T. cruzi amastigotes and point to mitochondrial energy metabolism as a key driver of amastigote flagellar motility.
In summary, this work has provided the first description of flagellar beating in the nonmotile intracellular amastigote stage of T. cruzi. Moreover, we now appreciate that the interaction between the T. cruzi amastigote flagellum and host mitochondria is dynamic, not (semi) stable as originally concluded from studies with fixed cells (11). While the biological role of amastigote flagellar motility has yet to be determined, the intracellular beating flagellum could alter the physical properties of the host cell cytoplasm, a shear-thinning fluid (16), to increase cytoplasmic fluidity and significantly influence diffusion dynamics locally (17). It could also facilitate metabolite sensing and/or uptake of nutrients via the flagellar pocket (18,19) or the cytostome-cytopharynx complex (20). Future studies will be aimed at defining the role of amastigote flagellar beating in the context of intracellular infection, including the dynamics of the unique amastigote flagellum-host mitochondrial interaction (11).

SUPPLEMENTAL MATERIAL
Supplemental material is available online only. VIDEO S1, MP4 file, 0.