Figure 3
Comparison of experimental findings and model predictions. (A) Modeling cells as spherical entities and allowing for an off-center division into spherical caps (cf. inset upper left) predicts well the experimentally observed cell volumes Vexp (main plot). Using the volume of P1 as sole input, volumes of EMS and P2, and from this volumes of P3 and C, and finally P4 and D were deduced assuming a constant shift Δx of the division plane. Predicted values Vtheo matched best the experimentally found ones, when choosing Δx = 1.75 µm (see inset lower right for the relative deviations when varying Δx). (B) Spindle displacements measured relative to the cell center in somatic cells (black dots) follow a normal distribution with a mean of −0.1 µm and standard deviation 0.56 µm (gray dashed line). In contrast, spindle displacements in germline cells (red dots) are grouped around a constant shift of 1.36 µm (dashed vertical line). A Kolmogorov-Smirnov test yielded a significance level of 0.38% that data from somatic and germline cells are from the same distribution, i.e. they can be regarded as different with a high significance. (C) Division asymmetries VR, predicted for P1, P2, and P3 on the basis of the last image stack that shows an unambiguous metaphase (grey bars), follow the experimental results for the daughter cells (black bars) but consistently underestimate the asymmetry. In fact, using these image stacks and dissecting segmented cells into two caps via a division plane through the metaphase plate is consistent with a median spindle shift of only 1.36 µm (see Fig. 3B and main text). Accounting for an additional, unmonitored spindle shift by approximately 450 nm during the lag period between consecutive image stacks (see main text) and repeating the dissection scheme the extrapolated asymmetries (red bars) show a favorable agreement with our experimental data. (D) Spindle displacement by a constant offset Δx can be rationalized by assuming constant forces FA < FP that pull the spindle towards the anterior and posterior end of the cell, respectively. As a result, the spindle is stretched and its center of mass moves into the posterior direction. Stress resistance of the spindle is modeled via a passive Hookean element (spring constant k, resting length L0) until a maximum extension is reached and the spindle ruptures at the onset of anaphase. See main text for details.
