Cell cycle dynamics during diapause entry and exit in an annual killifish revealed by FUCCI technology

Abstract

Background: Annual killifishes are adapted to surviving and reproducing over alternating dry and wet seasons. During the dry season, all adults die and desiccation-resistant embryos remain encased in dry mud for months or years in a state of diapause where their development is halted in anticipation of the months that have to elapse before their habitats are flooded again. Embryonic development of annual killifishes deviates from canonical teleost development. Epiblast cells disperse during epiboly, and a "dispersed phase" precedes gastrulation. In addition, annual fish have the ability to enter diapause and block embryonic development at the dispersed phase (diapause I), mid-somitogenesis (diapause II) and the final phase of development (diapause III). Developmental transitions associated with diapause entry and exit can be linked with cell cycle events. Here we set to image this transition in living embryos.

Results: To visibly explore cell cycle dynamics during killifish development in depth, we created a stable transgenic line in Nothobranchius furzeri that expresses two fluorescent reporters, one for the G₁ phase and one for the S/G₂ phases of the cell cycle, respectively (Fluorescent Ubiquitination-based Cell Cycle Indicator, FUCCI). Using this tool, we observed that, during epiboly, epiblast cells progressively become quiescent and exit the cell cycle. All embryos transit through a phase where dispersed cells migrate, without showing any mitotic activity, possibly blocked in the G₁ phase (diapause I). Thereafter, exit from diapause I is synchronous and cells enter directly into the S phase without transiting through G₁. The developmental trajectories of embryos entering diapause and of those that continue to develop are different. In particular, embryos entering diapause have reduced growth along the medio-lateral axis. Finally, exit from diapause II is synchronous for all cells and is characterized by a burst of mitotic activity and growth along the medio-lateral axis such that, by the end of this phase, the morphology of the embryos is identical to that of direct-developing embryos.

Conclusions: Our study reveals surprising levels of coordination of cellular dynamics during diapause and provides a reference framework for further developmental analyses of this remarkable developmental quiescent state.

Fig. 1

FUCCI transgenic line generation. a, b Schematic representations of FUCCI green and red constructs, respectively. FUCCI constructs were injected separately in different 1-cell stage fertilized eggs. Positive eggs were raised into adult fish, bred and screened for three generations (c). F2 FUCCI green fish were finally bred with F2 FUCCI red fish to generate double FUCCI embryos, which were used for most experiments (c). d Schematic representation of how FUCCI technology works, cells are green during S/G2/M phases, colourless between M and G1, red in G1 and G0 phases and yellow during a small portion of G2 phase. e Schematic representation of zUbiquitin-EGFP construct. EGFP expression driven by zUbiquitin promoter in Nothobranchius embryos and adult fish is shown (f). g FACS analysis of double FUCCI embryo. The scatterplot on the left shows the gating used to separate the four cell populations. The numbers indicate the percentage of cells in the four populations. The middle graph represents the intensity of the Hoechst staining as measure of DNA content of the four different populations. The graph on the right is the same as the graph in the middle but population are normalized on their own cells count rather than the total cells count. h FACS analysis of adult gonads of double FUCCI fish. The scatterplot on the left shows the gating used to separate the four cell populations. The numbers indicate the percentage of cells in the four populations. The middle graph represents the intensity of the Hoechst staining as measure of DNA content of the four different populations. The graph on the right is the same as the graph in the middle but population are normalized on their own cells count rather than the total cells count

Fig. 2

FUCCI green and FUCCI red characterization. A to J show FUCCI red expression at different developmental stages. K to T show FUCCI green expression at different developmental stages. A–C, K–N dispersed phase. D–G, O–Q somitogenesis stage. H, R hatched fry. I, J, S, T adult fish. For a detailed description, see the main text

Fig. 3

Schematic representation of embryonic layers in annual killifish species. Yolk syncytial layer (YSL) is made of cell forming a syncytium and in direct contact with the yolk. On top of this layer migrate and divide the blastomeres, a discrete population of cell composing the epiblast layer. The enveloping layer (EVL) is the layer of cells on top, a thin layer that completely envelope the other two. Two representations are shown, a lateral semi-transparent view (a) and a cutaway representation (b)

Fig. 4

Cell dynamics during epiboly (Wourms’ stages 18–19). When epiboly is ongoing, both green and red cells are present in dFUCCI embryos (a). As long as epiboly proceeds (b, c) the number of green EL cells gradually and slowly decreases over time, while the number of red EVL cells increases until the reaching of a plateau (graph hours 7–12). The cells in the field of view were easily tracked and counted transforming the dots in particles with IMARIS (d–f). The images and graph refer to the acquired portion of the embryos, corresponding to the superior hemisphere

Fig. 5

ESL nuclei as reference point for drift correction. Embryos continuously move throughout development (a–d), but EVL nuclei do not move once they reach their final position at the end of epiboly. These nuclei can therefore be used as reference system and drifts and rotations that occur during development can be corrected using their position (e–h). Correcting the drifts allows a more precise and reliable cell tracking and data analysis

Fig. 6

Cell dynamics during the early dispersed phase (Wourms’ stages 19–20). The first part of the dispersed phase is characterized by the presence of few EL green or red cells (a). ESL cells are pseudocoloured as large white dots and EL green and red cells as small green and red dots, respectively (b). Tracking of the three cell types over time. The number of each type of cell was constant for more than 10 h (c) and EL cells continuously moved during the whole time. Tracks of individual cells are shown in b, and speed average values in c, right panel. The images and graph refer to the acquired portion of the embryos, equivalent to the superior hemisphere

Fig. 7

Transition from early to late dispersed phase. a–h Time lapse showing the transition from a stage with few EL red or green cells could be detected through multiple reactivation and division events up to a relatively stable condition where the number of green EL cells is ~ 5 times their original number. g Quantification of green EL cell numbers over time. The letters a–h correspond to the pictures shown in a–h. Multiple peaks of synchronized proliferation are clearly visible (c, e, g) and divided by phases when cells synchronously enter into G1 phase (d, f, h). The images and graph refer to the acquired portion of the embryos, equivalent to the superior hemisphere

Fig. 8

Cell dynamics during late dispersed phase (Wourms’ stage 20). The second part of the dispersed phase is characterized by the presence of a larger number of EL green or red cells as opposed to early dispersed phase (Fig. 4). (a, b) Cell tracking. EVL cells are visualized as large white dots, while EL green and red cells as small green and red dots, respectively. All detected cells were tracked over time. c Quantification of the numbers of each type of cell over time. Note that the number is relatively constant for more than 10 h, while still moving. The images and graph refer to the acquired portion of the embryos, equivalent to the top superior hemisphere

Fig. 9

Reaggregation phase (Wourms’ stages 21–25). Green EL cells converge initially onto a single region of the embryo (a–c), forming a circular structure that over time becomes more compact with a reduced radius

Fig. 10

Axis extension phase (Wourms’ stage 26). The initial circular formation (a) lengthens to form the primordial axis (b, c)

Fig. 11

Direct-developing embryos late somitogenesis (Wourms’ stages 29–32). Green cells populate the whole embryonic axis for all the somitogenesis in embryos that do not enter diapause II. Three subsequent stages of the same embryo are reproduced to illustrate how green cells number and density slowly drops as long as development proceeds. a The embryo is composed mainly of green cells, b red cells increased with time and populated the first the somite pairs c somites contain red cells along almost the entire antero-posterior axis of the embryo

Fig. 12

Diapause II arrested embryo. An embryos arrested in diapause II is illustrated in brightfield in (a). b Overlay of green and red signal. c Green cells are concentrated in the medial part of the embryo between the somites. d Red cells clearly define the already formed somites and diffusely populate the whole axis region

Fig. 13

Time lapse of red and green fluorescence. Release from diapause II (A) as opposed to direct development (B). Upon release from diapause II, green cells greatly increase their number and density in less than 10 h (A). After the initial proliferation burst (c, d), green cells slowly decrease in number and density (e–f). In the lower panel, a quantification of green signal intensity as an indirect estimate of the number of green cells is reported. The lettering indicates the correspondence of the curve with the pictures (a–f). B An example of direct development. The number of green cells decreases gradually, while the number of red cells is roughly constant (a’–f’). Note that at the end of the processes the direct developed embryo (f’) is very similar to the embryo that escaped diapause (f). In the lower panel, a quantification of green signal intensity as an indirect estimate of the number of green cells is reported. The lettering indicates the correspondence of the curve with the pictures (a’–f’). Note that at the end of the processes, the ratio between red fluorescence and green fluorescence in the direct developed embryo (f’ in the right panel) is similar to that of the embryo that escaped diapause (f in the left panel)

Fig. 14

Graphical abstract representing FUCCI N. furzeri fish development. Images and developmental times refer to 26 °C incubation conditions, except purple arrows that refer to low temperature (18–21 °C) incubation conditions. Stages refers to developmental stages described by Wourms [6]