Balance between RhoA activation via RhoGEF2 and inactivation via C-GAP is required for organizing pulse behavior. (A) Mean area response within pulse behavior classes (ratcheted, unratcheted, or unconstricting) from C-GAP overexpression pulses (three embryos, 145 cells, 447 pulses). (B) Summary of pulsing phenotypes from WT, twist RNAi, and C-GAP OE embryos. (C and D) Live images of control (H2O-injected) or twist RNAi injected embryos expressing GFP:RhoGEF2 and membrane::RFP. RhoGEF2 accumulates across the apical domain (arrows) in control embryos, which requires Twist. However, subapical/junctional RhoGEF2 (brackets) does not depend on Twist but may decrease in twist RNAi. (E) Images from fixed control (twist heterozygous) or twist homozygous mutant embryos, stained for RhoA and Diaphanous (subapical Dia marks membranes). RhoA is radially polarized in control embryos, but in twist mutants, RhoA is diffuse across the entire apical domain and lacks structure or organization. (F) Fraction of pulses from each behavior for WT and C-GAP OE embryos. C-GAP OE leads to more unratcheted pulses and fewer ratcheted pulses. (G) Probability density functions of the timing of different pulse behaviors. WT pulses transition from unconstricting and unratcheted behaviors to ratcheted behavior, whereas all pulse behaviors in C-GAP OE pulses are co-occurring. Developmental time of 0 s corresponds to beginning of tissue contraction. (H) The probability of a cell transitioning from having a pulse of a behavior class (left columns) into having a subsequent pulse of another behavior class (right columns). WT cells show biased transitions to the ratcheted state, whereas C-GAP OE cells show a much weaker transition direction. The colors and the widths of the arrows represent the probability of transition. (I) The time interval between consecutive pulses within a cell is shown with respect to developmental time. Although WT pulses become more frequent, C-GAP OE pulses occur at a constant frequency. Line shows the best fit. Bars, 5 µm.
