Asymmetric cortical extension shifts cleavage furrow position in Drosophila neuroblasts

Abstract

The cytokinetic cleavage furrow is typically positioned symmetrically relative to the cortical cell boundaries, but it can also be asymmetric. The mechanisms that control furrow site specification have been intensively studied, but how polar cortex movements influence ultimate furrow position remains poorly understood. We measured the position of the apical and the basal cortex in asymmetrically dividing Drosophila neuroblasts and observed preferential displacement of the apical cortex that becomes the larger daughter cell during anaphase, effectively shifting the cleavage furrow toward the smaller daughter cell. Asymmetric cortical extension is correlated with the presence of cortical myosin II, which is polarized in neuroblasts. Loss of myosin II asymmetry by perturbing heterotrimeric G-protein signaling results in symmetric extension and equal-sized daughter cells. We propose a model in which contraction-driven asymmetric polar extension of the neuroblast cortex during anaphase contributes to asymmetric furrow position and daughter cell size.

FIGURE 1:

Drosophila neuroblasts undergo asymmetric polar elongation during anaphase. (A) Schematic of a symmetric division in which cortical extension is equal at both poles during anaphase. Myosin II is localized uniformly early in mitosis but becomes restricted to the equatorial region during anaphase. (B) Neuroblast cortical dynamics during mitosis using Dlg-GFP as a cortical marker. Selected frames from the movie are shown along with a kymograph of the entire division at 6-s intervals. The lines in the movie frames denote the section of the frame used for the kymograph. Cortical extension during anaphase is denoted by yellow brackets in the kymograph. The signal is enriched at the basal cortex because of contact with GMCs from previous divisions that also express GFP-Dlg (Supplemental Figure S1). (C) Mean anaphase polar extension in cultured Drosophila S2 cells transiently expressing Cherry-Zeus. The edge of the cell was marked at the point at which cytoplasmic fluorescence was no longer observed. Error bars, 1 SD. (D) Quantification of anaphase cortical extension in wild-type neuroblasts. The mean cortical extension from metaphase to the end of anaphase is shown for the apical (top) and basal (bottom) cortexes (NB, neuroblast). (E) The surface area of dividing neuroblasts measured using three-dimensional reconstruction normalized to that at the end of metaphase. The time points for measurements were early anaphase (completion of cortical extension), telophase (initiation of furrowing), and cytokinesis (completion of furrowing). (F) Mean relative surface areas of the daughter neuroblast (NB) and GMC that results from a neuroblast asymmetric cell division measured as in E at the completion of furrowing.

FIGURE 2:

Basal myosin II is required for asymmetric polar elongation. (A) Mean anaphase polar extension in sas4 mutant neuroblasts. Error bars, 1 SD. (B) Kymograph of myosin II (Sqh-GFP) in a wild-type larval neuroblast imaged at 6-s intervals. Top, selected frames with time relative to nuclear envelope breakdown. A line marks the section of the frame used for the kymograph. Anaphase cortical extension is denoted by brackets in the kymograph. (C) Quantification of apical and basal cortical extension in sqh^ax3; sqh-GFP (larval neuroblasts), sqh-GFP, and worniu-Gal4; UAS-Dlg-GFP neuroblasts. Error bars, 1 SD. Anaphase onset was determined using spindle (Jupiter-cherry) or chromosome (His2A-mRFP) markers. (D) Time dependence of cortical myosin signal and cortical position for wild-type neuroblasts. Dashed lines indicate the cortical position at each pole relative to the position at anaphase start. Solid lines indicate the intensity at each pole relative to the apical cortical intensity at anaphase start (determined as in C). Equatorial contraction indicates the time point at which the initiation of furrow ingression was observed. (E) Kymograph of Sqh-GFP in pins^P89 mutant neuroblasts. Brackets denote polar extension during anaphase. (F) Quantification of anaphase cortical extension in pins mutant neuroblasts. Error bars, 1 SD. (G) Time dependence of cortical myosin signal and cortical position for pins^P89 mutant neuroblasts. Annotations as in D.

FIGURE 3:

G-protein signaling regulates the basal furrow domain. (A) Myosin II (Sqh-GFP) localization in larval brain neuroblasts expressing Gαi using worniu-Gal4; UAS-Gαi. Images shown were taken 12 s apart. Scale bar, 5 μm. (B) Kymograph of Sqh-GFP signal across the poles from movie in A. Cortical extension is marked by the white lines. (C) Mean polar elongation for neuroblasts expressing Gαi or Gαi Q205L, a constitutively active variant that does not bind Gβγ or Pins. Cortical extension for Gαi is shown with two different cortical markers (Sqh-GFP or Dlg-GFP). Error bars, 1 SD. (D) Comparison of daughter cell size ratio for various cell types examined here. For asymmetrically dividing cells, this ratio was determined by dividing the diameter of the apical cell by the diameter of the basal cell. (E) Time dependence of cortical myosin signal and cortical position for neuroblasts expressing Gαi. Dashed lines indicate the cortical extension at each pole, whereas solid lines indicate the normalized intensity at each pole (as in Figure 2D). (F) Anaphase cortical extension for mud⁴ mutants where the spindle was aligned with basal myosin domain or orthogonal to it.

FIGURE 4:

Asymmetric cortical extension does not require spindle-induced equatorial contraction. (A) Sequence of a Colcemid treated rod mutant neuroblast expressing Jupiter-cherry to ensure the absence of a spindle. Scale bar, 10 μm. (B) Quantification of cortical extension in Colcemid-treated rod mutants. Error bars, 1 SD. (C) Model for the role of the basal furrow domain in daughter cell size asymmetry.