Extended Data Figure 9: Dynamics and apical-basal localization of the actomyosin flows in the neighbours.
From: Transmission of cytokinesis forces via E-cadherin dilution and actomyosin flows
a, b, Lifeact–Ruby expressing cells neighbouring a wild-type dividing cell in a MyoII–3 × GFP tissue. Insets highlight progressive accumulation of MyoII–3 × GFP and Lifeact–Ruby in the neighbours. Kymograph in b along the yellow box. n = 25 (18 pupae). c, Velocity of F-actin and MyoII flows in wild-type neighbours facing wild-type dividing cells (7 and 15 pupae, respectively). d, F-actin speckle probability density in wild-type neighbours facing wild-type dividing cells (4 pupae). Error bars denote s.e.m. of a multinomial distribution. e–g, Lifeact–Ruby expressing neighbour facing a dividing cell in a Dlg–GFP tissue. Top panel corresponds to apical plane; bottom panel corresponds to a more basal plane, at the level of the septate junctions (schematically represented in e). Kymograph in g along the yellow boxes. Insets and kymographs show that the Lifeact–Ruby flows observed in the ingressing junction during cytokinesis are restricted to the apical ___domain. Note that the bright dots visible in the basal kymograph correspond to Lifeact–Ruby aggregates, and not directional speckles (see Supplementary Video 7e). n = 8 cells (5 pupae). h, i, Rok–GFP and MyoII–3 × mKate2 distribution during cytokinesis. Kymograph in i generated along the yellow box. During cytokinesis, Rok–GFP is localized at the membrane/cortex and also flows with MyoII–3 × mKate2 during the ingression of the future daughter cell’s membranes (insets and kymograph). The localization of Rok is therefore in full agreement with previous findings reporting that this kinase can bind to the membrane in both vertebrates and Drosophila47,48. Moreover, Rok flows were already reported in germband cells and shown to depend on MyoII activity49. Overall, the constitutive nature of the ROCK kinase47, its known binding to the membrane and MyoII fully support the notion that self-organized actomyosin flows are triggered by the local decrease in E-cad concentration. n = 25 (3 pupae). j, k, Ratio of F-actin intensity (labelled by Lifeact–GFP) at the ingressing region versus the remaining AJs (j) and the ratio of F-actin intensity at the medial pool versus the AJs (k) in wild-type or rok cells neighbouring a wild-type dividing cell (7 or 7 pupae, respectively). Note that F-actin intensity within the ingressing region is similar to its intensity at the remaining AJs for both wild-type and rok neighbours, indicating that loss of Rok activity does not reduce the amount of F-actin within the ingressing region. l, E-cad–GFP and MyoII–mChFP localization in wild-type (n = 38 cells, 2 pupae) and Shroom (shr; n = 33 cells, 4 pupae) dividing cells, as well as its neighbour. Arrowheads denote MyoII–mChFP accumulation in the neighbours. To test further the role of Rok in F-actin polymerization, we analysed whether Shroom and fhos, two Rok effectors known to regulate F-actin polymerization48,50, would be required for MyoII accumulation in the neighbours. Neither the loss of fhos (Extended Data Fig. 7o–q) nor Shroom activity affects MyoII accumulation (see n), arguing against a role of Rok in F-actin polymerization. m, Rate of contractile ring constriction in wild-type and shr dividing cells (4 and 4 pupae, respectively). n, Normalized MyoII accumulation at 80% of the initial cell diameter for wild-type and shr neighbours (2 and 4 pupae, respectively). n denotes number of cells throughout. In c and d, n/n = number of speckles quantified/corresponding number of cells quantified. Mann–Whitney U-test (c, j, m) and Student’s t-test (k, n). Data are mean ± s.e.m. Scale bars, 5 μm.
