Non-equivalence in old- and new-flagellum daughter cells of a proliferative division in Trypanosoma brucei

Abstract

Differentiation of Trypanosoma brucei, a flagellated protozoan parasite, between life cycle stages typically occurs through an asymmetric cell division process, producing two morphologically distinct daughter cells. Conversely, proliferative cell divisions produce two daughter cells, which look similar but are not identical. To examine in detail differences between the daughter cells of a proliferative division of procyclic T. brucei we used the recently identified constituents of the flagella connector. These segregate asymmetrically during cytokinesis allowing the new-flagellum and the old-flagellum daughters to be distinguished. We discovered that there are distinct morphological differences between the two daughters, with the new-flagellum daughter in particular re-modelling rapidly and extensively in early G1. This re-modelling process involves an increase in cell body, flagellum and flagellum attachment zone length and is accompanied by architectural changes to the anterior cell end. The old-flagellum daughter undergoes a different G1 re-modelling, however, despite this there was no difference in G1 duration of their respective cell cycles. This work demonstrates that the two daughters of a proliferative division of T. brucei are non-equivalent and enables more refined morphological analysis of mutant phenotypes. We suggest all proliferative divisions in T. brucei and related organisms will involve non-equivalence.

Figure 1

mAb62, a monoclonal antibody that recognizes the flagella connector and the flagella connector remnant in new‐flagellum daughter cells. Scale bars represent 5 µm. A. Micrographs of SMOxP9 detergent‐extracted cytoskeletons stained with mAb62 (magenta) and with DAPI stained DNA (blue). NF‐new flagellum constructed in the current cell cycle, OF‐old flagellum constructed in one of previous cell cycles. B. Micrographs of cytoskeletons of SMOxP9 cells expressing eYFP::FCP4/TbKin15 (green) stained with Ab62 (magenta) and the DNA stained with DAPI (blue). C. mAb62 (magenta) recognizes the flagella connector also in fixed non‐detergent‐treated SMOXP9 cells expressing eYFP::FCP4/TbKin15 (green). DNA was visualized with DAPI (blue). D. Micrographs of parental and ΔFCP4/TbKin15 cytoskeletons stained with mAb62 (magenta) and with DAPI stained DNA (blue). The arrows indicate the flagella connectors with the mAb62 signal and the arrowheads the flagella connectors without the signal. The additional signals from mAb62 are particular noticeable in D) as the contrast has been increased to show that no flagella connector signal remains in the case of ΔFCP4/TbKin15.

Figure 2

Combination of mAb62 and eYFP::FCP3 enables identification of new‐flagellum daughters and old‐flagellum daughters. Cytoskeletons of cells expressing eYFP::FCP3 (green) stained with mAb62 (magenta) and DAPI (blue) at different stages during the cell cycle. Scale bars represent 5 µm. Insets in 2E, F and I show a higher magnification view of the flagella connector region; in these cells the side view of the flagella connector allows for an unambiguous determination of the position of signals along its new flagellum—old flagellum axis. The cells in 2H and I demonstrate that the flagella connector severing happens at variable points in cytokinesis (Briggs et al., 2004). Arrows in 2I indicate the anterior (A–N and A–O) and posterior (P–N and P–O) cell ends of the emerging new‐flagellum and old‐flagellum daughter cells, respectively.

Figure 3

Morphometric analysis of the three cohorts of 1F1K1N cells. Data were obtained on cytoskeletons of cells expressing eYFP::FCP3 and stained with mAb62 and DOT1—for images see Fig. S1. A. Mean lengths/distances (± s.d.) of different cell parameters for the three 1F1K1N cell cohorts. Horizontal bars with stars indicate a significant difference at P < 0.05 as determined by t‐test. N = 50 for each cell cohort. Cartoons below indicate the parameter measured (in red). B. Histograms of different parameters shown in A for the three cohorts of 1F1K1N cells.

Figure 4

Morphometric analysis of cytoskeletons of different cell cycle stages. Cells expressing eYFP::FCP3 were stained with mAb62 and DOT1—for images see Fig. S1. Mean lengths/distances (± s.d.) of different cell parameters for different cell cycle stages were plotted. Timing of major cell cycle events, such as initiation of new flagellum construction, kinetoplast segregation and nuclear mitosis are indicated. N = 24 for 2F2K2N old and new flagellum portions and N = 50 for the remaining categories. The data for the three cohorts of 1F1K1N cells are reproduced from Fig. 3. Cartoons indicate parameters measured for different cell cycle stages (in red). For 2F1K1N cells the parameters were measured in the same way as for 2F2K1N cells. Note that in the case of the old flagellum portion of 2F2K2N cells the whole body length was measured; the posterior cell end of the old‐flagellum daughter is not yet readily discernible. The cartoon at the bottom indicates the new and old flagellum portion of a 2K2N cell in cytokinesis. The red dashed line indicates how the cell will be resolved by the cytokinetic furrow into two daughters.

Figure 5

Re‐modelling of the anterior cell end. A. Measurement of the angle at which the cell body meets the flagellum in new‐ and old‐flagellum daughter cells in phase contrast images of fixed cells. B. Example SBF‐SEM acquired model of a cell in cytokinesis with the tapered end of the old flagellum portion of the cell indicated with an asterisk and the non‐tapered end of the new flagellum portion indicated with an arrow. C. Measurement of the radius of the cell body 2 µm from the anterior cell end for 1F cells, 2F pre‐cytokinesis cells and for the new flagellum and old flagellum portion of cells in cytokinesis that were modelled using the SBF‐SEM data. A cartoon at the top highlights the position at which the measurement was taken. D. Example models of 1F cells built from the SBF‐SEM data highlighting the tapered (an asterisk) and non‐tapered (an arrow) anterior cell ends.

Figure 6

G1 phase duration is equal in both daughter cells. A. Schematic showing the types of eYFP::TAX2 expressing cells generated after induction of RNAi against TAX‐2. The first round of new flagella produced after RNAi induction will have eYFP signal in the proximal part of the flagellum, flagella produced thereafter will have no eYFP signal. B. Micrographs showing an example cell with eYFP::TAX2 signal along the full length of its flagellum (left) and a cell with eYFP::TAX2 signal in the proximal part of the flagellum (right). DNA stained with DAPI is shown in magenta. C. Plots comparing the modelled and experimental results following induction of RNAi. Cells were categorized by cell cycle stage and flagellar signal type. Initially all cells have a full signal flagellum (blue, green, turquoise), then cells appear with partially labelled flagella, first in the 1K1N category (red) followed by 2K1N (purple) and 2K2N (orange). Partially labelled flagella are slower to appear in the observed results than in the model. N = 500 cells at each timepoint. D. To determine relative duration of G1 phase of the cell cycle of old‐flagellum versus new‐flagellum daughter cells, the value at which the proportion of each type within the 1K1N population reaches a plateau is critical. Experimental results show that this occurs at 50% for both types; this is the same as in the modelled results where all cells were dividing at the same rate. To characterize the sensitivity of our approach, we also modelled how would proportions of both cell types change if the new‐flagellum daughter cell cycle lasted 10% longer; in that case the plateau would be reached at 48% partially labelled and 52% fully labelled flagella. If the old‐flagellum daughter cell cycle lasted 10% longer the proportions would be reversed.

Figure 7

Schematics of changes to the new‐flagellum and old‐flagellum daughters in G1. Major changes to morphology of new‐flagellum daughter cells and old‐flagellum daughter cells detected in this study and in Wheeler et al. (2013) during G1 phase are indicated with arrows. Areas of major cytoskeletal re‐modelling are highlighted in red. See Fig. 4 for measurements.