Setting the Clock for Fail-Safe Early Embryogenesis

The embryogenesis of the small nematode Caenorhabditis elegans is a remarkably robust self-organization phenomenon. Cell migration trajectories in the early embryo, for example, are well explained by mechanical cues that push cells into positions where they experience the least repulsive forces. Yet, how this mechanically guided progress in development is properly timed has remained elusive so far. Here, we show that cell volumes and division times are strongly anticorrelated during the early embryogenesis of C. elegans with significant differences between somatic cells and precursors of the germline. Our experimental findings are explained by a simple model that in conjunction with mechanical guidance can account for the fail-safe early embryogenesis of C. elegans.

Figure 1

(a) Maximum intensity projection of image stacks taken with SPIM on C. elegans embryos of strain OD95 in which fluorescent markers stain the plasma membrane (red) and chromatin (green). White asterisks indicate nuclear bodies that are known to be irrelevant for embryogenesis. Scale bar: $10\mu \mathrm{m}$. (b) Corresponding individual $xy$ slices of the red channel (upper panel) and results of the membrane segmentation with cell names assigned (lower panel). (c) Early invariant cell lineage tree of C. elegans with cells of the $AB$, $P$, $MS$, and $C$ lineages being highlighted in black, red, blue, and green, respectively.

Figure 2

(a) Experimentally determined cell lifetimes ${\tau }_e$ are clearly anticorrelated with the cell’s volume $V$, e.g., in the $AB$ lineage (black circles) and in the $P$ lineage (red squares). Data were taken with untreated embryos (wt) at $22.5°\mathrm{C}$ and averaged over each generation of cells. The error bars indicate standard deviations between different embryos ($n=8$). The data are well described by Eq. (1) (full lines). (b) The cell cycle times $\tau$ obtained via Eq. (1) show an excellent agreement with experimentally determined cell division times ${\tau }_e$ for lineages $AB$, $P$, $MS$, and $C$ in untreated (wt) and RNAi-treated embyros with smaller (sm) and larger (la) size of the zygote $P_0$ (cf. the inset in the lower right corner). The data were taken at $22.5°\mathrm{C}$. Values for $\tau$ were obtained from Eq. (1) with ${\tau }_M=870\mathrm{s}$ and slight variations of $\alpha$ to account for lineage-specific behavior ($\alpha =0.66,1.26,0.72,0.84\times 10^6\mu {\mathrm{m}}^3\mathrm{s}$, for the $AB$, $P$, $MS$, and $C$ lineage, respectively, irrespective of any RNAi treatment).

Figure 3

(a) The experimentally determined cell doubling times ${\tau }_D$ (crosses) follow an Arrhenius scaling $\sim \exp (7417\mathrm{K}/T)$ (dashed gray lines). The parameter $\alpha$ [Eq. (1)] follows the same scaling (open symbols) irrespective of the lineage, whereas ${\tau }_M$ (asterisks) shows a slightly less strong temperature dependence. Deviations from the Arrhenius scaling around $25°\mathrm{C}$ reflect the thermal limit of the worm’s development [18]. (b) Projection of SPIM images perpendicular to the dorsal-ventral body axis taken on an embryo in the 24-cell stage (strain OD95). Nuclei of cells $P_4$ (red), Ea or Ep (green), and MSa or MSp (blue) are highlighted; $A$, $P$, $L$, and $R$ indicate the anterior, posterior, left, and right faces of the embryo. (c) The experimentally observed phenotype is faithfully reproduced by model simulations that consider Eq. (1) and cell-cell repulsion within the confining egg shell. (d) Aberrant phenotypes are observed when cell divisions of the $P$ lineage are enforced to follow Eq. (1) with parameters of the $AB$ lineage.