Polo-like kinases play an important role in a variety of mitotic events in mammalian cells, ranging from centriole separation and chromosome congression to abscission. To fulfill these roles, Polo-like kinase homologs move to different cellular locations as the cell cycle progresses, starting at the centrosome, progressing to the spindle poles and then the midbody. In the protist parasite Trypanosoma brucei, the single polo-like kinase homolog T. brucei PLK (TbPLK) is essential for cytokinesis and is necessary for the correct duplication of a centrin-containing cytoskeletal structure known as the bilobe. We show that TbPLK has a dynamic localization pattern during the cell cycle. The kinase localizes to the basal body, which nucleates the flagellum, and then successively localizes to a series of cytoskeletal structures that regulate the position and attachment of the flagellum to the cell body. The kinase localizes to each of these structures as they are duplicating. TbPLK associates with a specialized set of microtubules, known as the microtubule quartet, which might transport the kinase during its migration. Depletion of TbPLK causes defects in basal body segregation and blocks the duplication of the regulators that position the flagellum, suggesting that its presence on these structures might be necessary for their proper biogenesis. TbPLK migrates throughout the cell in T. brucei, but the specific locations to which it targets and its functions are geared towards the inheritance of a properly positioned and attached flagellum.

Trypanosoma brucei is the causative agent of sleeping sickness in sub-Saharan Africa. It is a flagellated protozoan that requires motility in order to evade the immune response of its host (Engstler et al., 2007; Hill, 2010). The single flagellum, which is essential for motility, is nucleated by a basal body near the posterior end of the cell and extends towards the anterior (Fig. 1A; the key structures described in this manuscript are also shown). The basal body is adjacent to the kinetoplast, which is the mitochondrial DNA aggregate that is maintained separate from the nucleus (Gull, 1999; Robinson and Gull, 1991). The flagellum is attached to the plasma membrane by an underlying cytoskeletal structure known as the flagellum attachment zone (FAZ) (Gull, 1999; Vickerman, 1969). The basal body is located at the base of an invagination of the cell surface called the flagellar pocket, which is the sole site for endo- and exo-cytosis (Field and Carrington, 2009; Lacomble et al., 2009). The flagellum emerges from the membrane of the flagellar pocket and must therefore exit this compartment. The flagellar pocket collar (FPC), a structure comprising the protein BILBO1, tightly clinches the neck of the flagellar pocket against the flagellum, which limits access to the compartment (Bonhivers et al., 2008; Gadelha et al., 2009). The neck of the flagellar pocket is also partially encircled by the hook-like extension of a MORN1-containing cytoskeletal complex, the stem of which coincides with the posterior part of the FAZ (Morriswood et al., 2009). This MORN1 complex is part of a larger cytoskeletal structure termed the bilobe, which also contains small calcium-binding proteins called centrins and a leucine-rich repeat protein known as TbLRRP1 (Broadhead et al., 2006; Shi et al., 2008; Selvapandiyan et al., 2007; He et al., 2005; Zhou et al., 2010). A unique set of four microtubules, known as the microtubule quartet (MtQ), are nucleated at the basal body, wrap around the flagellar pocket, traverse the FPC and then become part of the FAZ, connecting all the structures (Lacomble et al., 2009).

Fig. 1.

Cytoskeletal structures associated with the T. brucei flagellum. (A) A schematic view of the cytoskeletal structures associated with the flagellum. The kinetoplast is present in the posterior of the cell near to the probasal body and basal body, which nucleates the single flagellum. The flagellum emerges from the flagellar pocket, the top of which is tightly clinched by the FPC and bilobe. The flagellum is adhered against the cell body by the FAZ. (B) During the cell cycle, the first structure to duplicate is the basal body. The probasal body matures to nucleate a new flagellum, followed by the formation of two new probasal bodies (I). The pocket and FPC then duplicate, which segregates the new flagellum into its own pocket (II). Subsequently, the bilobe and FAZ duplicate (III). The newly replicated bilobes begin to separate, while the new FAZ continues to extend (IV). As the basal bodies begin to segregate, the kinetoplasts resolve into two structures (V), followed by karyokinesis (VI) and cytokinesis (VII).

Fig. 1.

Cytoskeletal structures associated with the T. brucei flagellum. (A) A schematic view of the cytoskeletal structures associated with the flagellum. The kinetoplast is present in the posterior of the cell near to the probasal body and basal body, which nucleates the single flagellum. The flagellum emerges from the flagellar pocket, the top of which is tightly clinched by the FPC and bilobe. The flagellum is adhered against the cell body by the FAZ. (B) During the cell cycle, the first structure to duplicate is the basal body. The probasal body matures to nucleate a new flagellum, followed by the formation of two new probasal bodies (I). The pocket and FPC then duplicate, which segregates the new flagellum into its own pocket (II). Subsequently, the bilobe and FAZ duplicate (III). The newly replicated bilobes begin to separate, while the new FAZ continues to extend (IV). As the basal bodies begin to segregate, the kinetoplasts resolve into two structures (V), followed by karyokinesis (VI) and cytokinesis (VII).

Taken together these protein complexes regulate the position of the flagellum, near its base through the flagellar pocket collar (FPC and bilobe) and along the length of the plasma membrane (using the bilobe and the FAZ). The proper inheritance of the flagellum requires the coordinated assembly and partitioning of these complexes (Vaughan, 2010; Vaughan and Gull, 2008). This has implications for the process of flagellar biogenesis. Flagellum biogenesis is initiated by the maturation of the probasal body, which triggers growth of a new flagellum alongside the existing one (Fig. 1B, I) (Sherwin and Gull, 1989; Lacomble et al., 2010). Once it reaches the flagellar pocket neck, this needs to duplicate, so that each flagellar pocket now contains an emergent flagellum. This requires the duplication of the FPC and the bilobe (Fig. 1B, II,III). Further elongation of the new flagellum must then be coupled to growth of a new FAZ, so that the new flagellum is tethered to the plasma membrane (Fig. 1B, IV). The new flagellum is also linked to the old one by a structure known as the flagella connector (FC), which guides the positioning of the new flagellum along the cell body (Moreira-Leite et al., 2001). All this requires the coordinated duplication in sequence of those protein complexes that guide the new flagellum to its final position. What, then, organizes this process?

One of the most important features for any potential regulator of flagellum positioning is the ability to migrate from one structure to another to coordinate the assembly of new complexes at precise times as the flagellum is elongating. The polo kinases are migratory proteins that are known to regulate the biogenesis of many cytoskeletal structures in higher eukaryotes. In mammalian cells, PLK1 plays important roles in cell division, including centriole separation, chromosome congression and cytokinesis (Barr et al., 2004; Archambault and Glover, 2009). PLKs contain a phosphopeptide-binding domain that allows them to be targeted to programmed destinations by the action of upstream kinases such as Cdk1 (Elia et al., 2003a; Elia et al., 2003b; Preisinger et al., 2005). PLK1 is initially found on the centrosome, then migrates to spindle poles and the kinetochores following the breakdown of the nuclear envelope, and finally localizes to the central spindle and midbody prior to cytokinesis (Golsteyn et al., 1994; Golsteyn et al., 1995). Inhibition of PLK1 function leads to defects in spindle pole formation and chromosome congression (Lane and Nigg, 1996; Peters et al., 2006; McInnes et al., 2006).

In T. brucei, there is only a single PLK homolog, T. brucei PLK (TbPLK), which we had earlier implicated in the process of bilobe duplication (de Graffenried et al., 2008). Other evidence suggests that it can localize to the basal bodies and is involved in cytokinesis (Hammarton et al., 2007; de Graffenried et al., 2008; Umeyama and Wang, 2008). In contrast to mammalian PLKs, TbPLK does not appear to play a role in mitosis.

Here we show that the movement and functioning of TbPLK can explain the ordered sequence of duplication events of the cytoskeletal structures associated with the flagellum. The migration of the kinase with respect to the basal bodies, bilobe, FPC and FAZ has been established with high temporal resolution. TbPLK associates with the new MtQ as it initiates and is present at the tip of the microtubules as they extend during assembly of the new FAZ, which could explain how TbPLK migrates through the cell. We show that TbPLK localizes to each structure as it is duplicating and that depletion of the kinase prevents duplication, or, in the case of the basal bodies, segregation. Our data show that TbPLK is a vital regulator of key cytoskeletal duplication events in T. brucei and that interfering with these highly concerted events leads to irreparable cellular defects.

TbPLK migrates from the posterior to the anterior end of the cell and coincides with the duplication of a series of cytoskeletal organelles

Our previous results and those of others showed that during cell division TbPLK colocalizes with the basal body, then progresses to the bilobe and finally associates with the growing tip of the new FAZ (de Graffenried et al., 2008; Kumar and Wang, 2006; Umeyama and Wang, 2008). TbPLK is present on each of these organelles during their duplication phase. To refine our understanding of the movements of TbPLK during cell division, the localization of the kinase was determined with respect to: (1) the basal body and bilobe marker TbCentrin4 (C4); (2) BILBO1, which marks the flagellar pocket collar; and (3) the FAZ marker FAZ1. The temporal order of TbPLK localization was established by the observation of external cues such as basal body duplication and the emergence of the new flagellum. The kinetoplast duplicates before the nucleus, allowing it to be used as a cell cycle marker (McKean, 2003). The relative frequency of cells at each cell cycle stage in an asynchronous culture was used to estimate the duration of each TbPLK localization pattern in temporal detail as has been described previously (Sherwin and Gull, 1989; Contois, 1959; Barford and Hall, 1976). By acquiring and analyzing 4500 images, we were able to achieve a temporal resolution of 20 minutes for this cell line.

Fig. 2 shows the temporal localization patterns of TbPLK in cells stained for the markers of the bilobe, basal body, FPC and FAZ. Additional images of intermediates of TbPLK localization with each marker, along with DIC images, are available in supplementary material Figs S1–3. In an asynchronous culture, approximately 60% of cells express TbPLK. At the earliest time point of TbPLK expression the kinase localized to a small, tendril-like projection between the basal body and the bilobe structure (supplementary material Fig. S1b). The kinase then localized to both the mature basal body and probasal body, and followed the contour of the BILBO1 stained FPC, abutting the exterior edge of the structure (Fig. 2, Ia). Basal body duplication occurs around this time. At this stage TbPLK formed thin projections that contacted the bilobe structure, identified by TbCentrin4 staining, to form an arch-like shape (Fig. 2, Ib). Subsequently, the TbPLK signal was lost from the FPC and appeared to form a bar-like shape on the bilobe structure (Fig. 2, IId,e). Localization to the posterior end of the FAZ became apparent (Fig. 2, IIe). Two discrete FPCs became visible, and then most of the TbPLK labeling moved to the anterior side of the duplicated structures (Fig. 2, IIc,IIIf). At this point the bilobe, which was normally between 1.3 µm and 1.8 µm in length, grew to 2.2 µm or longer along its major axis as TbPLK migrated towards the anterior of the cell (Fig. 2, IIIg). TbPLK staining was still visible on both pairs of basal bodies. Bilobe duplication appeared coincident with initiation of the new FAZ (Fig. 2, IIIh). TbPLK then migrated from the segregated bilobes and concentrated in a structure that preceded the growing FAZ structure (Fig. 2, IVi,j). TbPLK continued to precede the tip of the new FAZ while the kinetoplast and then the nucleus segregated (Fig. 2k,l). The signal for the kinase disappeared just prior to cytokinesis (supplementary material Fig. S3i).

Fig. 2.

TbPLK migrates across the duplicating cytoskeletal structures during the cell cycle. A series of schematics showing duplication intermediates of the flagellum-associated cytoskeletal structures and the localization of TbPLK at various stages (left) and immunofluorescence images of the migration pattern of TbPLK in cells (right). The kinase was labeled with anti-TbPLK antibodies (TbPLK; green; a-l) and the bilobe, FPC and FAZ, were labeled with anti-BILBO1 (FPC; red; a,c,f), anti-C4 (Bilobe + BB; red; b,d,g,i), anti-FAZ1 (FAZ; red; e,h,j,k,l), respectively. The black line shows the timing of important duplication events during the cell cycle. The events shown in green are represented in the schematics and immunofluorescence images.

Fig. 2.

TbPLK migrates across the duplicating cytoskeletal structures during the cell cycle. A series of schematics showing duplication intermediates of the flagellum-associated cytoskeletal structures and the localization of TbPLK at various stages (left) and immunofluorescence images of the migration pattern of TbPLK in cells (right). The kinase was labeled with anti-TbPLK antibodies (TbPLK; green; a-l) and the bilobe, FPC and FAZ, were labeled with anti-BILBO1 (FPC; red; a,c,f), anti-C4 (Bilobe + BB; red; b,d,g,i), anti-FAZ1 (FAZ; red; e,h,j,k,l), respectively. The black line shows the timing of important duplication events during the cell cycle. The events shown in green are represented in the schematics and immunofluorescence images.

The migration of TbPLK from the posterior to the anterior of the cell coincides with the track of the MtQ, which initiates at the basal body, passes through the FPC and then becomes part of the FAZ (Lacomble et al., 2009). The initiation of the MtQ occurs very early in the cell cycle and continues throughout the extension of the FAZ. The association of TbPLK with the MtQ would explain how the kinase migrates throughout the cell. At early stages of the cell cycle, TbPLK was present in a tendrillar projection next to the basal body that is very near the initiation site of the new MtQ (Fig. 3Aa). To more clearly identify this tendrillar projection, extracted cells were labeled with anti-TbPLK antibodies, negatively stained, and then visualized by electron microscopy (EM). In dividing cells, the new MtQ is initiated on the anterior side of the new basal body (Lacomble et al., 2010). At early stages of the cell cycle the new MtQ was labeled with TbPLK in a pattern similar to that seen using immunofluorescence (Fig. 3Ab–d). Once the new MtQ had extended beyond the flagellar pocket and had become part of the FAZ, TbPLK could be found on its tip (Fig. 3Ba–d). This shows that TbPLK associates with the MtQ at early and late stages of the cell cycle.

Fig. 3.

TbPLK localizes to the MtQ. (A) TbPLK labels a tendrillar structure early in the cell cycle. TbPLK labeling is shown in green, basal body and bilobe labeling in red, and DAPI in blue. To identify the source of this labeling, whole cells were extracted with detergent, stained with anti-TbPLK antibodies, and then negatively stained for examination by EM (b,c). A cell early in the cell cycle was selected (b) and a higher magnification of the flagellar pocket region was obtained (black box; c). TbPLK labeled the new MtQ (white arrowheads), which appears as a short projection next to the new basal body. (d) A model of the flagellar pocket region of the cell, with the old and new MtQ shown in turquoise. Scale bars: (a) 1 µm, (b) 5 µm, (c) 500 nm. (B) A cell at a later stage of the cell cycle. (a) The cell, labeled as in A. (b,c) A higher magnification of the tip of the new MtQ (arrowheads). Flag, flagellum; NFlag, new flagellum; NMtQ, new MtQ; NBB, new basal body; OFlag, old flagellum; OBB, old basal body; OMtQ, old MtQ. Scale bars: (a) 5 µm, (b) 500 nm, (c) 500 nm. (d) A model of the flagellar pocket region of the cell, with the old and new MtQ shown in turquoise.

Fig. 3.

TbPLK localizes to the MtQ. (A) TbPLK labels a tendrillar structure early in the cell cycle. TbPLK labeling is shown in green, basal body and bilobe labeling in red, and DAPI in blue. To identify the source of this labeling, whole cells were extracted with detergent, stained with anti-TbPLK antibodies, and then negatively stained for examination by EM (b,c). A cell early in the cell cycle was selected (b) and a higher magnification of the flagellar pocket region was obtained (black box; c). TbPLK labeled the new MtQ (white arrowheads), which appears as a short projection next to the new basal body. (d) A model of the flagellar pocket region of the cell, with the old and new MtQ shown in turquoise. Scale bars: (a) 1 µm, (b) 5 µm, (c) 500 nm. (B) A cell at a later stage of the cell cycle. (a) The cell, labeled as in A. (b,c) A higher magnification of the tip of the new MtQ (arrowheads). Flag, flagellum; NFlag, new flagellum; NMtQ, new MtQ; NBB, new basal body; OFlag, old flagellum; OBB, old basal body; OMtQ, old MtQ. Scale bars: (a) 5 µm, (b) 500 nm, (c) 500 nm. (d) A model of the flagellar pocket region of the cell, with the old and new MtQ shown in turquoise.

During the period when TbPLK is present on the bilobe and the new FAZ, a small punctate signal was seen that appeared to be beyond the tip of the new FAZ and on the opposite side of the plasma membrane (Fig. 2, IIe,IIIh,IVi). This labeling was most intense prior to basal body separation. Close inspection of DIC images showed that this signal corresponded to the contact point of the two flagella (supplementary material Fig. S4A). This interaction is mediated by the FC, which helps position the new flagellum along the cell body by using the old flagellum as a guide (Moreira-Leite et al., 2001). To test whether TbPLK localized to the FC, extracted cells were labeled with anti-TbPLK antibodies and negatively stained for examination by EM. The FC has a distinct triangular shape at the tip of the new flagellum that is clearly identifiable (Briggs et al., 2004). This structure was frequently labeled with anti-TbPLK antibodies (supplementary material Fig. S4B,C). This pool of TbPLK appears to be distinct from the one that associates with the MtQ.

TbPLK depletion does not affect basal body duplication or maturation

Once the migration of TbPLK during the cell cycle had been established, we sought to identify the effect of depleting the kinase on the various organelles to which it localizes. Previous work has shown that a doxycycline-inducible RNAi system directed against TbPLK effectively depletes the kinase and results in cells with abnormal DNA content, defects in bilobe duplication and an inability to undergo cytokinesis (de Graffenried et al., 2008). Western blotting of control and doxycycline-treated cells showed that TbPLK depletion was extremely efficient and occurred within 8 hours of RNAi induction (supplementary material Fig. S5A). TbPLK depletion led to growth arrest within 16 hours of TbPLK RNAi induction (supplementary material Fig. S5B). When DNA content was quantified after 18 hours of RNAi, the number of cells with one nucleus and one kinetoplast (1N1K) and one nucleus and two kinetoplasts (1N2K) decreased, while cells with two nuclei and two kinetoplasts (2N2K) doubled compared with controls (Fig. 4A). In 25% of the TbPLK-depleted cells the nucleus duplicated before the kinetoplast (2N1K). 2N1K cells can arise from a mispositioned cleavage furrow, which also produces a cell containing only a kinetoplast (0N1K). However, upon TbPLK depletion, 0N1K cells were rarely observed, making it unlikely that incorrect furrow ingression had taken place.

Fig. 4.

BB mature, duplicate and nucleate a new flagellum. (A) The percentage of cells with the indicated numbers of nuclei and kinetoplasts after 18 hours of RNAi induction. (B) Control cells stained with anti-TbCentrin2 antibody (C2; green), anti-TBBC antibodies (TBBC; red) and DAPI (DNA; blue). The insets are magnifications of the basal body regions. In the control condition the probasal body matures (e–g) then new probasal bodies are assembled as the new flagellum appears (i–k). The schematics on the left represent the duplication of the basal body and the flagellum. The underscore indicates the point at which the structures duplicate. Scale bar: 5 µm.

Fig. 4.

BB mature, duplicate and nucleate a new flagellum. (A) The percentage of cells with the indicated numbers of nuclei and kinetoplasts after 18 hours of RNAi induction. (B) Control cells stained with anti-TbCentrin2 antibody (C2; green), anti-TBBC antibodies (TBBC; red) and DAPI (DNA; blue). The insets are magnifications of the basal body regions. In the control condition the probasal body matures (e–g) then new probasal bodies are assembled as the new flagellum appears (i–k). The schematics on the left represent the duplication of the basal body and the flagellum. The underscore indicates the point at which the structures duplicate. Scale bar: 5 µm.

The appearance of the 2N1K cells in TbPLK-depleted samples suggested that the kinase is necessary for the proper replication of the kinetoplast. Previous work had shown that the kinetoplast DNA (kDNA) content of the single kinetoplast in 2N1K cells increases upon TbPLK depletion, which suggests that the kinetoplast defect stems from segregation of the duplicated structures (Hammarton et al., 2007). Considering the importance of the basal body in segregating the kDNA, it is possible that defects in basal body duplication and/or separation could explain the lack of segregated kinetoplasts. To test this hypothesis, a careful accounting of basal body maturation, duplication and separation was conducted.

Basal body replication occurs in 1N1K cells. The first stage of this process is maturation of the probasal body, which gains the capacity to nucleate a new flagellum (Lacomble et al., 2010). This maturation can be tracked with antibodies against trypanosome basal body component (TBBC) protein, a coiled coil protein shown to associate selectively with mature basal bodies (Dilbeck et al., 1999; Absalon et al., 2008). After maturation, the two basal bodies assemble new probasal bodies, completing the duplication cycle (Sherwin and Gull, 1989). The appearance of the probasal bodies can be monitored with an antibody to TbCentrin2 (C2), which is a component of the basal body and is recruited to the structure early in its assembly in mammalian cells (Kleylein-Sohn et al., 2007). C2 is also present in the flagellum, allowing the position and assembly of the new flagellum to be monitored. Once the new basal bodies have been established, they begin to separate, which drives the separation of the replicated kDNA, leading the cell to progress to the 1N2K state. In cells that have just emerged from division, TBBC labels the mature basal body, whereas C2 labels both the mature and probasal body, along with the flagellum (Fig. 4Ba–c). At the earliest stages of duplication the probasal body becomes TBBC positive and the formation of the new flagellum begins (Fig. 4Be–g). Soon afterwards, four C2 spots are visible, indicating the assembly of the probasal bodies (Fig. 4Bi–k). As cell division proceeds, the pairs of basal bodies begin to separate (Fig. 4Bm–o).

In TbPLK-depleted cells, basal body duplication seemed to proceed without any defects. The probasal body became TBBC positive, the C2-positive probasal bodies appeared and the newly mature basal body was capable of nucleating a new flagellum (Fig. 5Aa–c). The clearest defects were the detachment of the new flagellum (Fig. 5Ad,h,l) and the close placement of the duplicated basal bodies in 2N cells (Fig. 5Ae–g). In many 2N cells, the probasal bodies appeared to have undergone a second round of maturation (Fig. 5Ai–k), which should not occur. Other cells contained eight visible C2-positive structures, four of which were TBBC positive, along with four flagella (Fig. 5Am–o). These cells had not undergone cytokinesis but were still capable of generating new basal bodies and flagella, suggesting that maturation and duplication of these structures is independent of TbPLK.

Fig. 5.

Probasal body maturation, BB duplication and nucleation of the flagellum are independent of TbPLK. TbPLK-depleted cells were stained with anti-TbCentrin2 antibody (C2; green), anti-TBBC antibodies (TBBC; red) and DAPI (DNA; blue). The schematics on the left represent the duplication of the basal body and the flagellum. In the absence of TbPLK, probasal body maturation and BB duplication still occur (a–c, e–g). A subset of TbPLK-depleted cells underwent a second round of BB maturation, generating four TBBC-positive BBs (i–k) and subsequently eight basal bodies and four flagella (m–o). Scale bar: 5 µm. The underscores indicate the point at which the structures increase in number. (B) The ratios between the total number of BBs and the number of nuclei, the number of mature BBs and the number of nuclei, and the number of flagella and the number of nuclei in single cells in TbPLK-depleted and control cells.

Fig. 5.

Probasal body maturation, BB duplication and nucleation of the flagellum are independent of TbPLK. TbPLK-depleted cells were stained with anti-TbCentrin2 antibody (C2; green), anti-TBBC antibodies (TBBC; red) and DAPI (DNA; blue). The schematics on the left represent the duplication of the basal body and the flagellum. In the absence of TbPLK, probasal body maturation and BB duplication still occur (a–c, e–g). A subset of TbPLK-depleted cells underwent a second round of BB maturation, generating four TBBC-positive BBs (i–k) and subsequently eight basal bodies and four flagella (m–o). Scale bar: 5 µm. The underscores indicate the point at which the structures increase in number. (B) The ratios between the total number of BBs and the number of nuclei, the number of mature BBs and the number of nuclei, and the number of flagella and the number of nuclei in single cells in TbPLK-depleted and control cells.

To further confirm this result, we calculated several ratios to test whether basal body number or maturation was changed in the TbPLK-depleted cells (Fig. 5B). The ratio of total basal body number to nuclear number should be two in early 1N1K cells and all 2N2K cells because these cells have a basal body and probasal body and a single nucleus (early 1N1K) or two basal-body–probasal-body pairs and two nuclei (2N2K). In 1N1K cells that have undergone basal body duplication and in all 1N2K cells, the ratio should be four because these cells have two basal-body–probasal-body pairs and a single nucleus. Defects or a delay in basal body duplication would cause the ratio to decrease, whereas multiple rounds of duplication or early duplication would cause the ratio to increase. Similarly, changes in the ratios of TBBC-positive structures to nuclear number and of flagellar number to nuclear number would identify any defects in basal body maturation. All three ratios were indistinguishable in TbPLK-depleted and control cells. This suggests that TbPLK is not necessary for the duplication of the basal body or the flagellum. The appearance of additional flagella in a subset of 2N cells in the TbPLK-depleted sample is due to the flagellum biogenesis process continuing in the absence of cytokinesis.

TbPLK depletion leads to a basal body segregation defect

Although TbPLK does not seem to play a role in basal body duplication or maturation, closer inspection of the data suggested that separation of the duplicated structures might be failing. Many of the 2N cells in TbPLK-depleted samples had pairs of basal bodies that were close together (Fig. 5Ae–l). To quantify this effect, negative stain EM was conducted on extracted whole-mount cytoskeletons of TbPLK-depleted and control cells. The distance between the basal body pairs was measured and plotted against the ratio of the length of the new and old flagella, which serves as a measure of cell cycle progression (Robinson et al., 1995). In control cells, the distance between basal body pairs had a bimodal distribution, with slow separation occurring before a flagellar length ratio of 0.4 (Fig. 6A,Ca, arrowheads). At this point, the distance between the basal body pairs increased more rapidly, with a maximal distance of 7 µm observed prior to cytokinesis (Fig. 6Cb–d, arrowheads). The shortest distance between basal body pairs in a 2N cell was 3 µm. We set this distance as the lower limit for basal body separation in 2N cells and compared it with 2N cells in TbPLK-depleted samples. Although no 2N cells had a basal body pair distance below 3 µm in the control sample, over 70% of the 2N cells in the TbPLK-depleted sample did (Fig. 6B,Cg,h, arrowheads; supplementary material Table S1). In 1N cells with two basal body pairs the inter-basal body distance was less (Fig. 6Ce,f, arrowheads). This suggests that the separation defect occurs before karyokinesis. The formation of a new flagellum was not impeded in terms of ratio to the old flagellum or in absolute length. This supports the hypothesis that the nucleation of the new basal body and the elongation of the new flagellum are independent of TbPLK.

Fig. 6.

TbPLK is necessary for BB segregation. Whole-mount cytoskeletons were generated and visualized by TEM. (A,B) The ratio of the length of the new to old flagellum and inter-basal body distance determined in control (A) and TbPLK-depleted cells (B). (C) Representative images of control and TbPLK-depleted samples depicting 1N cells with comparable new flagellum length (a,e), cells undergoing mitosis (b,f), cells with visible microtubule bundles (c,g), and cells that had completed karyokinesis (d,h) are shown. Arrowheads indicate the basal bodies. Scale bar: 5 µm.

Fig. 6.

TbPLK is necessary for BB segregation. Whole-mount cytoskeletons were generated and visualized by TEM. (A,B) The ratio of the length of the new to old flagellum and inter-basal body distance determined in control (A) and TbPLK-depleted cells (B). (C) Representative images of control and TbPLK-depleted samples depicting 1N cells with comparable new flagellum length (a,e), cells undergoing mitosis (b,f), cells with visible microtubule bundles (c,g), and cells that had completed karyokinesis (d,h) are shown. Arrowheads indicate the basal bodies. Scale bar: 5 µm.

TbPLK depletion causes defects in the formation of a new FPC

While observing the lack of basal body segregation in TbPLK-depleted cells, we noted that many of the duplicated flagella seemed to be emerging from the cell body at the same site; this defect was especially noticeable in 2N1K cells (Fig. 5Ag). Under normal conditions a 2N cell should have two spatially separated flagellar pockets, each containing a single flagellum. The proximity of the two flagella suggested that flagellar pocket duplication could be impacted by the absence of TbPLK, forcing the cell to thread the new flagellum through the old pocket. The FPC demarcates the pocket and links it to the cytoskeleton, and TbPLK was found to partially localize in this region during its duplication (Fig. 2a). Difficulties in duplicating the FPC could lead to two flagella emerging from the same pocket because of the linkage of the flagella by the FC. To test this, extracted cytoskeletons were prepared from TbPLK-depleted and control cells and stained with an antibody against the FPC component BILBO1. Kinetoplast preservation is poor in extracted cytoskeletons, making it difficult to distinguish between 1N1K and 1N2K cells. To simplify the analysis, the DNA states of the cells were categorized only in terms of the number of nuclei.

The formation of a new FPC is impeded in cells depleted of TbPLK. In control cells, FPC duplication occurs at the 1N state (Fig. 7Ab). A third of 1N cells have two FPCs, whereas almost all 2N cells had two FPCs (Fig. 7Ac,d, B). In TbPLK-depleted cells, the number of 1N cells with two FPCs dropped to 12% (Fig. 7B). Of cells with two nuclei, 55% had only a single FPC (Fig. 7Aj, B). In many of the 2N cells that had two FPCs the replicated structures were positioned closely together between the two nuclei, suggesting that basal body separation had failed (Fig. 7Ak).

Fig. 7.

TbPLK depletion causes defects in FPC duplication. TbPLK-depleted and control cells were labeled with anti-BILBO1 antibody (FPC; green) and DAPI (DNA; blue). (A) FPC duplication stages in relation to FAZ duplication. The new FPCs are indicated by arrowheads. In the control condition, FPC duplication (b) occurs at the 1N stage. Upon TbPLK depletion, cells with 1 FPC in the 2N state are seen (j). Scale bars: 5 µm. (B) Quantification of the FPC duplication defect in TbPLK-depleted cells. The cells are categorized according to the number of nuclei (1N or 2N). Each nuclear state in each treatment sums to 100%. (C) Whole-mount cytoskeletons of TbPLK-depleted and control cells were negatively stained for visualization by EM. In the control sample, a cell with two emergent flagella (a) had two separate FPCs (arrowhead; b,c). In the TbPLK-depleted sample, cells at a similar cell cycle stage frequently had flagella that seemed to emerge from the same point (d). (e) Magnified view of the boxed region in d showing that the flagella were encircled by a single FPC (arrowheads). NFlag, new flagellum; NBB, new basal body; OFlag, old flagellum; OBB, old basal body. Scale bars: (a) 5 µm, (b) 500 nm, (c) 500 nm, (d) 5 µm, (e) 500 nm.

Fig. 7.

TbPLK depletion causes defects in FPC duplication. TbPLK-depleted and control cells were labeled with anti-BILBO1 antibody (FPC; green) and DAPI (DNA; blue). (A) FPC duplication stages in relation to FAZ duplication. The new FPCs are indicated by arrowheads. In the control condition, FPC duplication (b) occurs at the 1N stage. Upon TbPLK depletion, cells with 1 FPC in the 2N state are seen (j). Scale bars: 5 µm. (B) Quantification of the FPC duplication defect in TbPLK-depleted cells. The cells are categorized according to the number of nuclei (1N or 2N). Each nuclear state in each treatment sums to 100%. (C) Whole-mount cytoskeletons of TbPLK-depleted and control cells were negatively stained for visualization by EM. In the control sample, a cell with two emergent flagella (a) had two separate FPCs (arrowhead; b,c). In the TbPLK-depleted sample, cells at a similar cell cycle stage frequently had flagella that seemed to emerge from the same point (d). (e) Magnified view of the boxed region in d showing that the flagella were encircled by a single FPC (arrowheads). NFlag, new flagellum; NBB, new basal body; OFlag, old flagellum; OBB, old basal body. Scale bars: (a) 5 µm, (b) 500 nm, (c) 500 nm, (d) 5 µm, (e) 500 nm.

In the TbPLK-depleted sample, many 2N cells with a single FPC appeared to have a new flagellum that emerged from the same position as the old flagellum (Fig. 7Aj). The failure to form a new FPC could force the old FPC to accommodate two flagella. Cytoskeletons were extracted from TbPLK-depleted and control cells and then subjected to negative stain EM to test this hypothesis. Under normal conditions the new flagellum is confined within its own pocket and FPC by the time it emerges from the cell body (Fig. 7Ca–c). In the TbPLK-depleted sample, the old and new flagella emerged from a single FPC, which seemed enlarged (Fig. 7Cd,e). This is consistent with our images at the light microscope level that appeared to show the new flagellum emerging from the same position as the old.

TbPLK depletion causes defects in new FAZ assembly

All our TbPLK-depleted samples contained large numbers of cells that had detached new flagella. Considering that TbPLK is present at the growing tip of the new FAZ, which is essential for the adherence of the new flagellum, it seemed likely that depleting the kinase might lead to a FAZ defect. To test this, we depleted TbPLK and stained the cells with an antibody that detects the FAZ protein FAZ1 and DAPI to detect DNA (Kohl et al., 1999; Vaughan et al., 2008).

Control cells showed normal FAZ morphology and number, with the new FAZ emerging in a subset of 1N1K cells and elongating during the 1N2K and 2N2K stages (Fig. 8Aa–e, arrowheads). Both the old and new flagella were attached to the cell body adjacent to their FAZ filament. In the doxycycline-treated cells, defects in the formation of the new FAZ were evident (Fig. 8Ak–o, arrowheads). These defects will subsequently be described in more detail. Numerous cells that had detached new flagella were seen in DIC images (Fig. 8Ap–t, arrows).

Fig. 8.

TbPLK is necessary for FAZ elongation and flagellar attachment. TbPLK-depleted and control cells were stained with anti-FAZ1 (FAZ; red) and DAPI (DNA; blue), and were then visualized by fluorescence and DIC microscopy. (A) FAZ elongation during the cell cycle in control cells. Arrowheads indicate duplicating FAZs. The new FAZ is initiated in the 1N1K stage (b), and elongated thereafter (c–e). In TbPLK-depleted cells, the new FAZs (k–o, arrowheads) were frequently defective, and detached flagella (p–t, arrows) were observed. Scale bars: 5 µm. The term snFAZ refers to new FAZ in 2N cells that is under 5 µm in length. (B) FAZ phenotypes observed in control cells and TbPLK-depleted cells at different points of the cell cycle. The aberrant 2N1K DNA state is only observed in TbPLK-depleted cells. (C) TbPLK depletion leads to detachment of the new flagellum. In both B and C, each condition in each cell cycle stage sums to 100%. (D) Percentage of 2N cells with abnormal FAZs (either 1 FAZ or snFAZ) or normal FAZs that have detached new flagella. Only data for the TbPLK-depleted condition are shown.

Fig. 8.

TbPLK is necessary for FAZ elongation and flagellar attachment. TbPLK-depleted and control cells were stained with anti-FAZ1 (FAZ; red) and DAPI (DNA; blue), and were then visualized by fluorescence and DIC microscopy. (A) FAZ elongation during the cell cycle in control cells. Arrowheads indicate duplicating FAZs. The new FAZ is initiated in the 1N1K stage (b), and elongated thereafter (c–e). In TbPLK-depleted cells, the new FAZs (k–o, arrowheads) were frequently defective, and detached flagella (p–t, arrows) were observed. Scale bars: 5 µm. The term snFAZ refers to new FAZ in 2N cells that is under 5 µm in length. (B) FAZ phenotypes observed in control cells and TbPLK-depleted cells at different points of the cell cycle. The aberrant 2N1K DNA state is only observed in TbPLK-depleted cells. (C) TbPLK depletion leads to detachment of the new flagellum. In both B and C, each condition in each cell cycle stage sums to 100%. (D) Percentage of 2N cells with abnormal FAZs (either 1 FAZ or snFAZ) or normal FAZs that have detached new flagella. Only data for the TbPLK-depleted condition are shown.

Cells in the TbPLK-depleted sample had defects in the formation of a new FAZ, which is located at the posterior end of the cell. These cells should have two FAZs because duplication of this structure begins when the cells are still in the 1N1K state. In 1N2K cells, 36% had only one FAZ, whereas 45% of 2N2K cells showed defects [Fig. 8B. Although the results are statistically significant (P<0.05), it should be noted that 1N2K cells constituted only 2.5% of the population, or 41 out of 1649 total cells counted.]. In 2N cells, defects could be separated into two categories: cells with only one FAZ (Fig. 8An), and cells that made a short new FAZ (snFAZ) that was under 5 µm in length (Fig. 8Al,o). All 2N2K cells in the control had a new FAZ that was larger than 5 µm. This argues that the 2N2K cells with snFAZs in the TbPLK-depleted sample were not able to complete the nascent structure. A total of 16% of 2N2K cells completely lacked a new FAZ, whereas 28% had an snFAZ (Fig. 8B). The aberrant 2N1K cells showed the strongest FAZ defects, with 92% of the cells containing either an snFAZ (50%) or a single FAZ (42%; Fig. 8B). We were unable to establish criteria to distinguish normal from abnormal FAZs in 1N1K cells because of the small size of the emerging structure. However, 15% of TbPLK-depleted 1N1K cells had detached new flagella, presumably because they had difficulty assembling a new FAZ (Fig. 8Ak,p). Quantification showed that 40% of the TbPLK-depleted cells had detached flagella, compared with less than 5% in the control (Fig. 8C). Cells with detached flagella had additional defects in FAZ formation (Fig. 8D).

The cytoskeletal elements that TbPLK interacts with duplicate in a prescribed order, which may reflect a hierarchy among them. If a hierarchy does exist, then some aspects of the TbPLK-depletion phenotype might be the indirect. The FAZ, which is assembled last, could be influenced by defects in the upstream structures. The bilobe and FAZ are linked and depletion of the bilobe protein C2 blocks the formation of a new FAZ, whereas blocking the formation of a new FPC can also affect FAZ biogenesis (Shi et al., 2008; Bonhivers et al., 2008). If TbPLK directly affects the assembly of a new FAZ, it should be possible to observe cells that have only FAZ defects upon TbPLK depletion. To test this, we stained TbPLK-depleted samples for BILBO1 and FAZ1 simultaneously, and we then looked for examples of 2N cells with two FPCs and a single FAZ or an snFAZ. If these cells were present, it would suggest that TbPLK depletion could cause FAZ defects even if FPC duplication proceeded normally. These two FPC cells with FAZ defects constituted 18% of the 2N population (supplementary material Fig. S6).

In this article, we have charted in detail the movement of TbPLK throughout the cell cycle and determined the effect of its depletion on several important cytoskeletal structures that function to position the flagellum. The kinase initially localizes to the MtQ, then the basal bodies, subsequently progresses to the FPC, bilobe, FC and, finally, to the growing tip of the new FAZ. This pattern closely matches the order in which these structures duplicate, suggesting that TbPLK triggers the replication or inheritance of these cytoskeletal elements. To test this theory, TbPLK was depleted by RNAi, which caused defects in these structures: the basal bodies fail to segregate, and the FPC and FAZ fail to duplicate properly. These defects are especially damaging because flagellar duplication and karyokinesis proceed normally in TbPLK-depleted cells, leading to the production of multinucleate cells with detached flagella. Perturbing the ordered duplication of these structures prohibits the highly ordered series of cytoskeletal events necessary for the inheritance of an attached, functional flagellum that resides within its own pocket.

The migration of TbPLK during the cell cycle closely resembles the track of the MtQ. The kinase is present on the MtQ as it first assembles next to the basal body, and is on its tip once the microtubules becomes part of the FAZ. The MtQ comes into contact with all the structures to which TbPLK localizes, making it possible that the kinase selectively interacts with the MtQ and uses its extension to move throughout the cell. PLK1 binds directly to a subset of microtubules and has been shown to interact with kinesins, which could facilitate the migration of the kinase (Feng et al., 1999; Neef et al., 2003). The positive ends of the MtQ microtubules nucleate from the basal body at the posterior end of the cell, as opposed to the microtubules that constitute the rest of the subpellicular array that underlies the surface of the cell (Robinson et al., 1995; Lacomble et al., 2009). This arrangement could facilitate the migration of TbPLK by association with positive-end-directed kinesins.

At early phases in the cell cycle, TbPLK localizes to the basal bodies and remains associated with them as they duplicate. TbPLK depletion has no effect on either basal body duplication or maturation, which is consistent with the function of the homologous kinase in mammalian cells and in budding yeast. PLK1 depletion in mammalian cells causes defects in the separation of the centrosomes and the positioning of the spindle poles, while centrosome duplication is unaffected (Hanisch et al., 2006; Lane and Nigg, 1996; Tsou et al., 2009). Yeast lacking the PLK homolog cdc5 can replicate their centrosome equivalents, known as the spindle pole bodies, but cannot move them apart to form an appropriate spindle (Snead et al., 2007; Kitada et al., 1993). TbPLK conserves this function of partitioning its centriolar equivalents. However, TbPLK appears to be dispensible for mitosis, which is not the case for cdc5 and PLK1. TbPLK has become specialized towards controlling the duplication of cytoskeletal structures such as the FPC and FAZ that assure the proper positioning of the flagellum. This change in function may reflect a change in the upstream kinases that generate binding sites for TbPLK (Barr et al., 2004; Archambault and Glover, 2009).

Many of the aberrant 2N1K cells that are present in TbPLK-depleted samples are 2N2K cells that have failed to effectively segregate their kinetoplast DNA. Previous work has shown that the kinetoplast in 2N1K cells generated by TbPLK depletion contain two equivalents of kDNA (Hammarton et al., 2007). The division of the duplicated kinetoplast into two discrete structures is directly coupled to the separation of the basal body pairs (Robinson and Gull, 1991). In TbPLK-depleted cells, probasal body maturation and the formation of two new probasal bodies proceed normally, but the newly formed basal body pairs do not separate. This lack of separation leaves a single kinetoplast that contains two copies of the kDNA. 2N1K cells could also arise if cytokinesis occurs prior to karyokinesis or segregation of the duplicated nuclei (Ploubidou et al., 1999; Shi et al., 2008). This would generate a cell containing only a kinetoplast, known as a zoid. Our DNA counts show a limited number of zoid cells in our TbPLK-depleted samples. 2N1K cells that arise from the production of a zoid should also have only one flagellum because one copy of the kinetoplast and flagellum are necessary to produce the zoid. Almost all the 2N1K cells in TbPLK-depleted samples contain two flagella, suggesting that they are the product of a basal body segregation defect.

TbPLK then moves to the flagellar pocket region, which partially coincides with the bilobe structure and the FPC. It is possible to categorize the flagellar pocket localization of TbPLK into four discrete types: a tendril-like structure that marks the initiation of a new MtQ, an arch-like structure with protrusions that contact the basal bodies, a bar-like structure that traverses the duplicating and segregating bilobes, and a punctate structure that marks the emergence of the new FAZ. The arch-like structure might represent a nascent FPC, suggesting that TbPLK is present prior to new BILBO1 recruitment. The kinase might also depart before the deposition of BILBO1 on the new FPC. The identification of additional FPC-resident proteins would allow this hypothesis to be tested.

TbPLK has previously been implicated in bilobe duplication, causing defects in the assembly and segregation of a new structure (de Graffenried et al., 2008). Failure of FPC duplication has been observed upon depletion of BILBO1, but this led to the formation of a flagellum that lacks a flagellar pocket (Bonhivers et al., 2008). In the case of TbPLK depletion new FPC formation is blocked, but the new flagellum appears to be of normal length and remains associated with the old pocket. The old and new flagellum appear to exit through the same FPC in some cells, suggesting that the new flagellum is able to thread through the old FPC or that the FPC is caught in some intermediate duplication state that can accommodate both flagella. The passage of the new flagellum through the old FPC might be facilitated by its attachment to the old flagellum through the FC. The absence of a new FPC upon TbPLK depletion suggests that the flagellar pocket fails to duplicate or segregate properly, as it does in the absence of BILBO1, although this remains to be tested directly.

Once duplication of the FPC is complete, TbPLK migrates to the bilobe and tip of the new FAZ. Before departing the structure entirely, TbPLK concentrates at the anterior tip of the duplicated bilobe structure. At this point, TbPLK appears to reside at the tip of the growing FAZ. This suggests that the bilobe and FAZ are segments of a larger continuous cytoskeletal structure. Once the new FAZ is visible, TbPLK is present on the growing tip of this structure, which overlaps with the end of the new FAZ. The TbPLK signal on the FAZ is present until just prior to cytokinesis, when it disappears from the cell altogether.

At this point of the cell cycle TbPLK also localizes to the FC (Moreira-Leite et al., 2001). This signal diminishes as the new flagellum extends, but is still visible at later stages. TbPLK is the first component of the FC to be identified. Although the precise role that the kinase plays on the FC is unknown, the proposed role of this structure in templating the position and orientation of the new flagellum is similar to the function of the other organelles whose inheritance are controlled by TbPLK. This reinforces the idea that the main function of TbPLK is to assure the proper inheritance of the flagellum.

TbPLK depletion causes strong defects in the extension of a new FAZ, which leads to detached flagella. The production of a detached flagellum cannot be reversed and ensures that one of the progeny of any subsequent cell division will lack proper motility. Proper motility is also essential for completing cytokinesis (Ralston et al., 2006). Ingression of the cleavage furrow seems to occur at a position dictated by the new FAZ, and defects in FAZ formation frequently lead to misplaced furrows, which can cause the unequal partitioning of the daughter cells (Vaughan, 2010; Kohl et al., 2003; Vaughan and Gull, 2008). Although TbPLK-depleted cells have major FAZ defects, they appear to fail at cytokinesis rather than trigger it prematurely. This is shown by the accumulation of 2N2K cells and the fact that the 2N1K cells produced by TbPLK depletion do not arise from the production of zoids.

Inheritance of an attached and positioned flagellum is an essential part of cell division in T. brucei. The structures that control these flagellar properties, such as the basal body, bilobe, FPC, FC and FAZ must therefore be assembled and partitioned in a specific order. TbPLK is present on each of these structures as they duplicate; depletion of the kinase causes a failure in their inheritance, which leads to detached and misplaced flagella. PLK homologs are programmed to localize at specific times by the action of upstream kinases and phosphatases, which act in concert to create and destroy binding sites for the kinases. TbPLK shares this high degree of mobility with its homologs in other organisms, but its localization pattern during the cell cycle, and therefore the organelles whose duplication it licenses, has been configured for the unique features of the trypanosome cytoskeleton.

Cell culture

The 29.13 cell line was supplied by G. Cross (The Rockefeller University) (Wirtz et al., 1999). The 29.13 cell line with a tetracycline-inducible, stably integrated RNAi construct against TbPLK has been described previously (de Graffenried et al., 2008). Cells were cultured as previously described (de Graffenried et al., 2008).

Antibodies

The antibodies were obtained from the following sources: anti-FAZ1 (L3B2) from Keith Gull (University of Oxford, UK), anti-BILBO1 from Derrick Robinson (University of Bordeaux, France), anti-TBBC from Etienne Pays (University of Brussels, Belgium). Anti-C4 and anti-C2 mouse IgG1 monoclonal antibodies were raised against recombinant TbCentrin4 and TbCentrin2, respectively. The specificity of these antibodies is shown in supplementary material Fig. S7. Rabbit anti-TbPLK has been described previously (de Graffenried et al., 2008).

Immunofluorescence

Cells were adhered to coverslips and fixed in methanol at −20°C for 15 minutes, or 20 minutes for detergent extracted cytoskeletons. In the latter case, cells were incubated for 5 minutes at room temperature or on wet ice in PEME buffer (2 mM EGTA, 1 mM MgSO4, 0.1 mM EDTA, 0.1 M PIPES pH 6.9) containing different concentrations of NP-40 (0.25% at 0°C for anti-TbPLK with anti-BILBO1, 0.5% for anti-BILBO1 with anti-FAZ1, 1% for anti-TBBC with anti-C2) and washed in PBS prior to fixation. The cells were rehydrated in PBS and blocked overnight at 4°C with 3% BSA in PBS, except for anti-BILBO1 (no BSA). The cells were stained with primary antibody (anti-TbPLK diluted 1:50, anti-FAZ1 1:200, anti-BILBO1 1:8, anti-C4 1:80, anti-C2 1:30 or anti-TBBC 1:800) washed in PBS, then stained with secondary antibodies. The cells were washed in PBS, mounted in DAPI Fluormount G (Southern Biotech), and visualized as previously described (Morriswood et al., 2009).

Electron microscopy

Cells were washed and resuspended at 1×106 per ml in PBS, then mounted on glow-discharged formvar-coated copper grids. The cells were detergent extracted with 1% NP-40 in PEME for 3 min at room temperature and then fixed in 2.5% glutaraldehyde in PEME buffer, rinsed twice in water, and stained with 0.5% aurothioglucose (Sigma-Aldrich) for 15 seconds (Höög et al., 2010). Cells were visualized with a JOEL JEM 1210 TEM 80 KeV transmission electron microscope (Joel Ltd. Tokyo, Japan), a Morada digital camera (Olympus) and analySIS FIVE software, iTEM (Soft Imaging System Germany). Image quantification was conducted with ImageJ 1.42q (http://rsbweb.nih.gov/ij/), and figures were prepared for publication using Photoshop CS4 software (Adobe Systems, San Jose, CA).

Immuno-gold electron microscopy

Cells were settled onto glow-discharged, formvar- and carbon-coated nickel grids (200 mesh) and detergent extracted (0.25% NP40, 0.1 M PIPES pH 6.9, 1 mM MgCl2). They were then blocked in 3% BSA in PBS and labeled with rabbit polyclonal anti-PLK antibodies (1:50 in 3% BSA in PBS), followed by 10 nm colloidal gold goat anti-rabbit antibodies (BBInternational) diluted 1:10 in PBS with 3% BSA. The grids were then washed in PBS, fixed in 2.5% glutaraldehyde and negatively stained with 0.35% aurothioglucose.

Mathematical analysis

For quantification of immunofluorescence, 500–550 cells for each condition were analyzed. For EM, 85–120 cells were analyzed. In all histograms, tables and the time line, the averages of three independent experiments are represented. When shown, the sample standard deviation (s.d.) is represented either as error bars or as numerical values.

The estimation of the cumulative time x spent by the cells to complete a specific stage was calculated according to the equation:
formula
where y is the cumulative percentage of cells including the stage in question, and µ is the relative growth rate (Sherwin and Gull, 1989). The duration of each stage was obtained by subtraction (Δx = xn+1xn, where xn+1 is a later stage with respect to xn). µ was obtained from:
formula
where t1 is a time point later than t0, N1 and N0 are their respective total population estimated from the culture density (Contois, 1959).

We thank Graham Warren (Center for Molecular Biology, Max F. Perutz Laboratories, Austria) for support and critical comments and the members of the Warren laboratory for critical reading of the manuscript. We thank Catarina Gadelha and Harald Kotisch for advice on immuno-EM. We would also like to thank all the laboratories that contributed essential reagents.

Funding

This work was supported by the Austrian Science Fund [grant number P21550-B12 to Graham Warren and C.L.d.G.]; and Austrian Science Fund doctoral program Molecular Mechanisms of Cell Signaling [grant number W1220-B09 to Graham Warren].

Absalon
S.
,
Blisnick
T.
,
Bonhivers
M.
,
Kohl
L.
,
Cayet
N.
,
Toutirais
G.
,
Buisson
J.
,
Robinson
D.
,
Bastin
P.
(
2008
).
Flagellum elongation is required for correct structure, orientation and function of the flagellar pocket in Trypanosoma brucei.
J. Cell Sci.
121
,
3704
3716
.
Archambault
V.
,
Glover
D. M.
(
2009
).
Polo-like kinases: conservation and divergence in their functions and regulation.
Nat. Rev. Mol. Cell Biol.
10
,
265
275
.
Barford
J. P.
,
Hall
R. J.
(
1976
).
Estimation of the length of cell cycle phases from asynchronous cultures of Saccharomyces cerevisiae.
Exp. Cell Res.
102
,
276
284
.
Barr
F. A.
,
Silljé
H. H. W.
,
Nigg
E. A.
(
2004
).
Polo-like kinases and the orchestration of cell division.
Nat. Rev. Mol. Cell Biol.
5
,
429
441
.
Bonhivers
M.
,
Nowacki
S.
,
Landrein
N.
,
Robinson
D. R.
(
2008
).
Biogenesis of the trypanosome endo-exocytotic organelle is cytoskeleton mediated.
PLoS Biol.
6
,
e105
.
Briggs
L. J.
,
McKean
P. G.
,
Baines
A.
,
Moreira–Leite
F.
,
Davidge
J.
,
Vaughan
S.
,
Gull
K.
(
2004
).
The flagella connector of Trypanosoma brucei: an unusual mobile transmembrane junction.
J. Cell Sci.
117
,
1641
1651
.
Broadhead
R.
,
Dawe
H. R.
,
Farr
H.
,
Griffiths
S.
,
Hart
S. R.
,
Portman
N.
,
Shaw
M. K.
,
Ginger
M. L.
,
Gaskell
S. J.
,
McKean
P. G.
et al.  (
2006
).
Flagellar motility is required for the viability of the bloodstream trypanosome.
Nature
440
,
224
227
.
Contois
D. E.
(
1959
).
Kinetics of bacterial growth: relationship between population density and specific growth rate of continuous cultures.
J. Gen. Microbiol.
21
,
40
50
.
de Graffenried
C. L.
,
Ho
H. H.
,
Warren
G.
(
2008
).
Polo-like kinase is required for Golgi and bilobe biogenesis in Trypanosoma brucei.
J. Cell Biol.
181
,
431
438
.
Dilbeck
V.
,
Berberof
M.
,
Van Cauwenberge
A.
,
Alexandre
H.
,
Pays
E.
(
1999
).
Characterization of a coiled coil protein present in the basal body of Trypanosoma brucei.
J. Cell Sci.
112
,
4687
4694
.
Elia
A. E. H.
,
Cantley
L. C.
,
Yaffe
M. B.
(
2003a
).
Proteomic screen finds pSer/pThr-binding domain localizing Plk1 to mitotic substrates.
Science
299
,
1228
1231
.
Elia
A. E. H.
,
Rellos
P.
,
Haire
L. F.
,
Chao
J. W.
,
Ivins
F. J.
,
Hoepker
K.
,
Mohammad
D.
,
Cantley
L. C.
,
Smerdon
S. J.
,
Yaffe
M. B.
(
2003b
).
The molecular basis for phosphodependent substrate targeting and regulation of Plks by the Polo-box domain.
Cell
115
,
83
95
.
Engstler
M.
,
Pfohl
T.
,
Herminghaus
S.
,
Boshart
M.
,
Wiegertjes
G.
,
Heddergott
N.
,
Overath
P.
(
2007
).
Hydrodynamic flow-mediated protein sorting on the cell surface of trypanosomes.
Cell
131
,
505
515
.
Feng
Y.
,
Hodge
D. R.
,
Palmieri
G.
,
Chase
D. L.
,
Longo
D. L.
,
Ferris
D. K.
(
1999
).
Association of polo-like kinase with alpha-, beta- and gamma-tubulins in a stable complex.
Biochem. J.
339
,
435
442
.
Field
M. C.
,
Carrington
M.
(
2009
).
The trypanosome flagellar pocket.
Nat. Rev. Microbiol.
7
,
775
786
.
Gadelha
C.
,
Rothery
S.
,
Morphew
M.
,
McIntosh
J. R.
,
Severs
N. J.
,
Gull
K.
(
2009
).
Membrane domains and flagellar pocket boundaries are influenced by the cytoskeleton in African trypanosomes.
Proc. Natl. Acad. Sci. USA
106
,
17425
17430
.
Golsteyn
R. M.
,
Schultz
S. J.
,
Bartek
J.
,
Ziemiecki
A.
,
Ried
T.
,
Nigg
E. A.
(
1994
).
Cell cycle analysis and chromosomal localization of human Plk1, a putative homologue of the mitotic kinases Drosophila polo and Saccharomyces cerevisiae Cdc5.
J. Cell Sci.
107
,
1509
1517
.
Golsteyn
R. M.
,
Mundt
K. E.
,
Fry
A. M.
,
Nigg
E. A.
(
1995
).
Cell cycle regulation of the activity and subcellular localization of Plk1, a human protein kinase implicated in mitotic spindle function.
J. Cell Biol.
129
,
1617
1628
.
Gull
K.
(
1999
).
The cytoskeleton of trypanosomatid parasites.
Annu. Rev. Microbiol.
53
,
629
655
.
Hammarton
T. C.
,
Kramer
S.
,
Tetley
L.
,
Boshart
M.
,
Mottram
J. C.
(
2007
).
Trypanosoma brucei Polo-like kinase is essential for basal body duplication, kDNA segregation and cytokinesis.
Mol. Microbiol.
65
,
1229
1248
.
Hanisch
A.
,
Wehner
A.
,
Nigg
E. A.
,
Silljé
H. H. W.
(
2006
).
Different Plk1 functions show distinct dependencies on Polo-Box domain-mediated targeting.
Mol. Biol. Cell
17
,
448
459
.
He
C. Y.
,
Pypaert
M.
,
Warren
G.
(
2005
).
Golgi duplication in Trypanosoma brucei requires Centrin2.
Science
310
,
1196
1198
.
Hill
K. L.
(
2010
).
Parasites in motion: flagellum-driven cell motility in African trypanosomes.
Curr. Opin. Microbiol.
13
,
459
465
.
Höög
J. L.
,
Gluenz
E.
,
Vaughan
S.
,
Gull
K.
(
2010
).
Ultrastructural investigation methods for Trypanosoma brucei.
Methods Cell Biol.
96
,
175
196
.
Kitada
K.
,
Johnson
A. L.
,
Johnston
L. H.
,
Sugino
A.
(
1993
).
A multicopy suppressor gene of the Saccharomyces cerevisiae G1 cell cycle mutant gene dbf4 encodes a protein kinase and is identified as CDC5.
Mol. Cell. Biol.
13
,
4445
4457
.
Kleylein–Sohn
J.
,
Westendorf
J.
,
Le Clech
M.
,
Habedanck
R.
,
Stierhof
Y-D.
,
Nigg
E. A.
(
2007
).
Plk4-induced centriole biogenesis in human cells.
Dev. Cell
13
,
190
202
.
Kohl
L.
,
Sherwin
T.
,
Gull
K.
(
1999
).
Assembly of the paraflagellar rod and the flagellum attachment zone complex during the Trypanosoma brucei cell cycle.
J. Eukaryot. Microbiol.
46
,
105
109
.
Kohl
L.
,
Robinson
D.
,
Bastin
P.
(
2003
).
Novel roles for the flagellum in cell morphogenesis and cytokinesis of trypanosomes.
EMBO J.
22
,
5336
5346
.
Kumar
P.
,
Wang
C. C.
(
2006
).
Dissociation of cytokinesis initiation from mitotic control in a eukaryote.
Eukaryot. Cell
5
,
92
102
.
Lacomble
S.
,
Vaughan
S.
,
Gadelha
C.
,
Morphew
M. K.
,
Shaw
M. K.
,
McIntosh
J. R.
,
Gull
K.
(
2009
).
Three-dimensional cellular architecture of the flagellar pocket and associated cytoskeleton in trypanosomes revealed by electron microscope tomography.
J. Cell Sci.
122
,
1081
1090
.
Lacomble
S.
,
Vaughan
S.
,
Gadelha
C.
,
Morphew
M. K.
,
Shaw
M. K.
,
McIntosh
J. R.
,
Gull
K.
(
2010
).
Basal body movements orchestrate membrane organelle division and cell morphogenesis in Trypanosoma brucei.
J. Cell Sci.
123
,
2884
2891
.
Lane
H. A.
,
Nigg
E. A.
(
1996
).
Antibody microinjection reveals an essential role for human polo-like kinase 1 (Plk1) in the functional maturation of mitotic centrosomes.
J. Cell Biol.
135
,
1701
1713
.
McInnes
C.
,
Mazumdar
A.
,
Mezna
M.
,
Meades
C.
,
Midgley
C.
,
Scaerou
F.
,
Carpenter
L.
,
Mackenzie
M.
,
Taylor
P.
,
Walkinshaw
M.
et al.  (
2006
).
Inhibitors of Polo-like kinase reveal roles in spindle-pole maintenance.
Nat. Chem. Biol.
2
,
608
617
.
McKean
P. G.
(
2003
).
Coordination of cell cycle and cytokinesis in Trypanosoma brucei.
Curr. Opin. Microbiol.
6
,
600
607
.
Moreira–Leite
F. F.
,
Sherwin
T.
,
Kohl
L.
,
Gull
K.
(
2001
).
A trypanosome structure involved in transmitting cytoplasmic information during cell division.
Science
294
,
610
612
.
Morriswood
B.
,
He
C. Y.
,
Sealey–Cardona
M.
,
Yelinek
J.
,
Pypaert
M.
,
Warren
G.
(
2009
).
The bilobe structure of Trypanosoma brucei contains a MORN-repeat protein.
Mol. Biochem. Parasitol.
167
,
95
103
.
Neef
R.
,
Preisinger
C.
,
Sutcliffe
J.
,
Kopajtich
R.
,
Nigg
E. A.
,
Mayer
T. U.
,
Barr
F. A.
(
2003
).
Phosphorylation of mitotic kinesin-like protein 2 by polo-like kinase 1 is required for cytokinesis.
J. Cell Biol.
162
,
863
876
.
Peters
U.
,
Cherian
J.
,
Kim
J. H.
,
Kwok
B. H.
,
Kapoor
T. M.
(
2006
).
Probing cell-division phenotype space and Polo-like kinase function using small molecules.
Nat. Chem. Biol.
2
,
618
626
.
Ploubidou
A.
,
Robinson
D. R.
,
Docherty
R. C.
,
Ogbadoyi
E. O.
,
Gull
K.
(
1999
).
Evidence for novel cell cycle checkpoints in trypanosomes: kinetoplast segregation and cytokinesis in the absence of mitosis.
J. Cell Sci.
112
,
4641
4650
.
Preisinger
C.
,
Körner
R.
,
Wind
M.
,
Lehmann
W. D.
,
Kopajtich
R.
,
Barr
F. A.
(
2005
).
Plk1 docking to GRASP65 phosphorylated by Cdk1 suggests a mechanism for Golgi checkpoint signalling.
EMBO J.
24
,
753
765
.
Ralston
K. S.
,
Lerner
A. G.
,
Diener
D. R.
,
Hill
K. L.
(
2006
).
Flagellar motility contributes to cytokinesis in Trypanosoma brucei and is modulated by an evolutionarily conserved dynein regulatory system.
Eukaryot. Cell
5
,
696
711
.
Robinson
D. R.
,
Gull
K.
(
1991
).
Basal body movements as a mechanism for mitochondrial genome segregation in the trypanosome cell cycle.
Nature
352
,
731
733
.
Robinson
D. R.
,
Sherwin
T.
,
Ploubidou
A.
,
Byard
E. H.
,
Gull
K.
(
1995
).
Microtubule polarity and dynamics in the control of organelle positioning, segregation, and cytokinesis in the trypanosome cell cycle.
J. Cell Biol.
128
,
1163
1172
.
Selvapandiyan
A.
,
Kumar
P.
,
Morris
J. C.
,
Salisbury
J. L.
,
Wang
C. C.
,
Nakhasi
H. L.
(
2007
).
Centrin1 is required for organelle segregation and cytokinesis in Trypanosoma brucei.
Mol. Biol. Cell
18
,
3290
3301
.
Sherwin
T.
,
Gull
K.
(
1989
).
The cell division cycle of Trypanosoma brucei brucei: timing of event markers and cytoskeletal modulations.
Philos. Trans. R. Soc. Lond. B Biol. Sci.
323
,
573
588
.
Shi
J.
,
Franklin
J. B.
,
Yelinek
J. T.
,
Ebersberger
I.
,
Warren
G.
,
He
C. Y.
(
2008
).
Centrin4 coordinates cell and nuclear division in T. brucei.
J. Cell Sci.
121
,
3062
3070
.
Snead
J. L.
,
Sullivan
M.
,
Lowery
D. M.
,
Cohen
M. S.
,
Zhang
C.
,
Randle
D. H.
,
Taunton
J.
,
Yaffe
M. B.
,
Morgan
D. O.
,
Shokat
K. M.
(
2007
).
A coupled chemical-genetic and bioinformatic approach to Polo-like kinase pathway exploration.
Chem. Biol.
14
,
1261
1272
.
Tsou
M-F. B.
,
Wang
W-J.
,
George
K. A.
,
Uryu
K.
,
Stearns
T.
,
Jallepalli
P. V.
(
2009
).
Polo kinase and separase regulate the mitotic licensing of centriole duplication in human cells.
Dev. Cell
17
,
344
354
.
Umeyama
T.
,
Wang
C. C.
(
2008
).
Polo-like kinase is expressed in S/G2/M phase and associated with the flagellum attachment zone in both procyclic and bloodstream forms of Trypanosoma brucei.
Eukaryot. Cell
7
,
1582
1590
.
Vaughan
S.
(
2010
).
Assembly of the flagellum and its role in cell morphogenesis in Trypanosoma brucei.
Curr. Opin. Microbiol.
13
,
453
458
.
Vaughan
S.
,
Gull
K.
(
2008
).
The structural mechanics of cell division in Trypanosoma brucei.
Biochem. Soc. Trans.
36
,
421
424
.
Vaughan
S.
,
Kohl
L.
,
Ngai
I.
,
Wheeler
R. J.
,
Gull
K.
(
2008
).
A repetitive protein essential for the flagellum attachment zone filament structure and function in Trypanosoma brucei.
Protist
159
,
127
136
.
Vickerman
K.
(
1969
).
On the surface coat and flagellar adhesion in trypanosomes.
J. Cell Sci.
5
,
163
193
.
Wirtz
E.
,
Leal
S.
,
Ochatt
C.
,
Cross
G. A.
(
1999
).
A tightly regulated inducible expression system for conditional gene knock-outs and dominant-negative genetics in Trypanosoma brucei.
Mol. Biochem. Parasitol.
99
,
89
101
.
Zhou
Q.
,
Gheiratmand
L.
,
Chen
Y.
,
Lim
T. K.
,
Zhang
J.
,
Li
S.
,
Xia
N.
,
Liu
B.
,
Lin
Q.
,
He
C. Y.
(
2010
).
A comparative proteomic analysis reveals a new bi-lobe protein required for bi-lobe duplication and cell division in Trypanosoma brucei.
PLoS ONE
,
5
,
e9660
.

Supplementary information