Fig. 3. PON takes a different route to the basal crescent than Miranda, and requires Myosin II but not Myosin VI. Time-lapse analysis to compare the localization of Miranda-GFP (A,C) with PON-GFP (B,D) shows that PON localizes mainly along the cortex at pro/metaphase in NBs (B), as well as in NE cells (D). PON-GFP does not accumulate in the cytoplasm (white arrowheads). Miranda-GFP consistently showed strong cytoplasmic localization in NBs (A), as well as in NE cells (C, white arrowheads). At metaphase, Miranda (A,C) and PON (B,D) form an overlapping basal crescent. (E,F) Time-lapse microscopy of PON-GFP localization in NBs from untreated, control embryos (E), or NBs from embryos lacking fully functional Myosin II (F) because of injection of a Rho kinase inhibitor. In the absence of Myosin II activity, PON-GFP does not form a basal crescent but is mislocalized to the cortex in metaphase and anaphase, and concentrates at the cleavage furrow in telophase (arrowheads). (G) Downregulation of Myosin VI by RNA interference (MyoVI RNAi) does not affect crescent formation at metaphase, or the asymmetric segregation of PON-GFP at ana- and telophase. The mitotic spindle and, hence, the cleavage plane are rotated by 45° owing to downregulation of Myosin VI. The white circle in telophase depicts the position of the NB. Embryos heterozygous (H) and homozygous (I) for the jar¹ allele were fixed and stained with a PON antibody (green). DNA, blue. In agreement with the live imaging data, PON still formed a basal crescent at metaphase in control (H) and mutant (I) embryos. The mitotic spindle is misoriented by 90° owing to the lack of Myosin VI activity.