Undoubtedly, there are fundamental processes driving the structural mechanics of cell division in eukaryotic organisms that have been conserved throughout evolution and are being revealed by studies on organisms such as yeast and mammalian cells. Precision of structural mechanics of cytokinesis is however probably no better illustrated than in the protozoa. A dramatic example of this is the protozoan parasite Trypanosoma brucei, a unicellular flagellated parasite that causes a devastating disease (African sleeping sickness) across Sub-Saharan Africa in both man and animals. As trypanosomes migrate between and within a mammalian host and the tsetse vector, there are periods of cell proliferation and cell differentiation involving at least five morphologically distinct cell types. Much of the existing cytoskeleton remains intact during these processes, necessitating a very precise temporal and spatial duplication and segregation of the many single-copy organelles. This structural precision is aiding progress in understanding these processes as we apply the excellent reverse genetics and post-genomic technologies available in this system. Here we outline our current understanding of some of the structural aspects of cell division in this fascinating organism.
The African trypanosome has a long slender shape with a single flagellum laterally attached to the cell body in a left-handed helix from close to the posterior end along to the anterior  (Figure 1A). This is a major defining characteristic, yet the specific cell body shape is defined by a highly stable, highly cross-linked subpellicular microtubule cytoskeleton that underlies the plasma membrane  (Figure 1B). The microtubules of this array are equally spaced and have an identical polarity with their plus ends at the posterior end of the cell . The only exception to this polarity is a specific microtubule quartet (see later). The length and number of microtubules therefore varies in order to accommodate the change in cell diameter from the posterior to anterior end. Hence, shorter microtubules are inserted between longer microtubules .
The procyclic form of T. brucei
Organelles such as the mitochondrion and Golgi are present as single copies in specific positions resulting in a highly reproducible, organized and polarized cellular structure (Figure 2A). The kinetoplast (containing the mitochondrial DNA) is physically connected to the proximal end of the two basal bodies [4,5]. A mature basal body subtends the single flagellum with the pro-basal body connected alongside it. The pro-basal body will mature and form a new flagellum in the next cell cycle . In fact, basal bodies are both morphologically and functionally analogous to centrioles in mammalian cells and exhibit the same conserved maturation and inheritance pattern (for a review see ). The single flagellum exits the cell body via the flagellar pocket, the only reported site of endocytosis and exocytosis (for a review see ). The flagellum is composed of the canonical 9+2 microtubule axoneme plus a specialized structure called the PFR (paraflagellar rod) that runs along the length of the flagellum once it exits the flagellar pocket and is essential for motility [8–11]. The flagellum is connected along the length of the cell body by the FAZ (flagellum attachment zone). The FAZ is composed of a filament structure that connects the cell body with the PFR in the flagellum plus a set of four specialized microtubules that originate close to the basal bodies and so have the opposite polarity to the subpellicular microtubules [1,3,12]. The FAZ therefore results in a ‘seam’ in the subpellicular corset with ideal properties for defining axis and polarity for the mechanics of cytokinesis. Since the flagellum is tethered to the cell body via the FAZ, one can envisage a highly connected set of internal structures (kinetoplast, mitochondrion, basal bodies, flagellar pocket) that are physically tethered to each other and the external flagellum/cell body via the FAZ region. The single Golgi is also specifically positioned and, although there is no direct evidence for a physical connection, it certainly follows the same temporal and spatial duplication and segregation patterning with the other interconnected cytoskeletal structures [13,14].
Diagrammatic illustration of the cell cycle of the procyclic form of T. brucei
Structural mechanics of cell division
The trypanosome cell division cycle is the same as other eukaryotic cells, except that there two distinct S-phases which must be co-ordinated; one for the single mass of mitochondrial DNA contained within the kinetoplast and one for nuclear DNA [1,15]. Most of the single-copy organelles are duplicated via an apparent templated mechanism, whereby a new organelle is built next to an old one. The process begins with S-phase of the mitochondrial DNA followed by basal body maturation and duplication (Figure 2B) . A filament system connecting the duplicated DNA and a specific membrane region of the mitochondrion is duplicated and connected to the duplicated basal bodies . The new flagellum invades the flagellar pocket and the tip is physically connected to the old flagellum by a transmembrane mobile junction called the flagella connector (Figure 2B) [16,17]. The discovery of the flagella connector has been fundamental to our understanding of how these procyclic trypanosomes divide. We now understand that growth of the new flagellum is crucial in cell division. Perturbing flagellum growth or attachment to the cell body results in two unequal daughter cells, improper segregation of organelles, and is ultimately lethal [18–22]. The flagella connector provides a cytotaxis mechanism whereby an existing cellular structure transmits spatial organization to the new structure. In Trypanosoma brucei, transmission of cell polarity, axis, alignment and helicity occurs as the new flagellum extends along the old flagellum, guided by the flagella connector [16,17,23,24]. We do not yet know the mechanism that drives this mobile junction, but assembly and growth of the new flagellar axoneme via the IFT (intraflagellar transport) mechanism appears not to be necessary .
More recently, we have speculated on a further role for the flagella connector in these procyclic cells. The flagella connector stops at ∼0.6 of the length of the old flagellum and we refer to this as the ‘stop point’. The timing of the flagella connector stop point coincides with basal body and kinetoplast segregation. The position of the flagella connector at this stop point may actually aid in the segregation of the two flagella with their interconnected organelles by effectively acting as an anchor point as the two structures segregate [3,4,21] (Figure 2C). The flagella connector has only been characterized in the procyclic life cycle stage, but a similar structure has been identified in differentiation divisions (, and see later).
Of course, during cell division, the cell body increases in length and diameter while the new flagellum continues to extend. Extension of the subpellicular microtubules at the posterior end of the cell appears to be important in driving increases in cell body length. Remarkably, increase in cell diameter is achieved by insertion of new microtubules between old microtubules. This effectively results in a semi-conservative inheritance of the subpellicular microtubules to each daughter cell [26,27]. We do not yet understand this spatially at the individual microtubule level. Nor do we understand how the duplication and segregation of the organelles and the subpellicular microtubules are co-ordinated in the formation of the two daughter cells.
Mitosis occurs leaving one of the two nuclei positioned between the two kinetoplasts (Figure 2C), so ensuring two ‘clusters’ of cytoplasmic organelles ready for cleavage. Cleavage furrow ingression is unidirectional along the cell body from anterior to posterior between the old and new flagella. The exact selection site for furrow ingression is not known, but we have suggested that the position and end of the FAZ is implicated since it marks a unique seam in the cytoskeleton coincident with growth of the new flagellum to the anterior end of the cell where ingression initiates  (Figure 2D). The mechanism of furrow ingression is unknown and a functional characterization of actin in T. brucei failed to find a role in cytokinesis . As in mammalian cells, the two daughter cells remain attached to each other for some time before final cell abscission.
Structural mechanics of differentiation
In this overview we have outlined how two essentially identical daughter cells are formed within the confines of the existing cytoskeleton in T. brucei. However, when these cells receive certain cues, cell proliferation stops and cell differentiation occurs, giving rise to a new cell type that is both morphologically and biochemically distinct. This cell type is often pre-adapted for invasion of a new tissue or host and there are at least five morphologically distinct developmental forms in the life cycle of T. brucei. They are classified by the relative positions of the kinetoplast, nucleus and flagellum along the posterior–anterior axis of the cell . The changes in the position of these structures are again performed within the confines of the existing cytoskeleton.
An extreme example of these positional changes and dramatic asymmetric cell division occurs in the differentiation event from the procyclic to epimastigote form that takes place in the tsetse fly vector. Procyclic cells undergo a period of cell proliferation in the tsetse fly midgut and after 2–3 days they start migrating towards the salivary glands, during which differentiation to the epimastigote life cycle form occurs [30–32]. As they arrive at the proventriculus, an almost doubling of cell body length has occurred. Growth of a short new flagellum takes place, and at mitotic anaphase kinetoplast segregation occurs. After mitosis, furrow ingression initiates at the anterior end of the short new flagellum, resulting in one very long and one very short epimastigote cell (Figure 3). The kinetoplast of the short cell is now anterior to the nucleus, rather than posterior to the nucleus in procyclic cells. It is thought that only the short epimastigote cells migrate to the salivary glands where cell proliferation initiates once more . This highly asymmetric division illustrates the ability of the trypanosome to orchestrate its cytokinetic mechanics to fit both the proliferative and differentiation type of divisions required of its life cycle.
A scanning electron micrograph of an asymmetrical division giving rise to one short and one long epimastigote cell
Although we have good descriptions of the positional changes and cell shape alterations during differentiation, we do not understand how these are achieved within the confines of the existing cytoskeleton. We also lack an understanding of how the cell extends in length and how flagellum length is regulated in order to produce cells with shorter or longer flagella.
Conclusions and future perspectives
The mechanisms and control of cell proliferation and differentiation is essential to the parasite life cycle and as such this is a fascinating organism to study. Understanding these processes is important in trying to find targets to combat this disease. Sequencing of the genome of T. brucei and other related parasites are complete  and we have many biochemical and genetic tools (see [34–37]). These tools have enabled a molecular dissection of genes involved in cell proliferation and differentiation (for reviews see [38,39]). However, in order to better understand the phenotypes that have arisen from these experiments our next challenge is to understand the three-dimensional spatial organization of these cytoskeletal structures, how co-ordination of assembly and segregation is performed and, no doubt, the dependency relationships that exist between the events.
Mechanics and Control of Cytokinesis: Biochemical Society Focused Meeting held at Royal College of Surgeons, Edinburgh, U.K., 9–12 January 2008. Organized and Edited by Gwyn Gould (Glasgow, U.K.) and Iain Hagan (Manchester, U.K.).
Work in our laboratory is funded by The Wellcome Trust and the EP Abraham Trust. K.G. is a Wellcome Trust Principal Research Fellow. We thank members of our group for generous discussions.