Mechanisms underlying Myosin 10′s contribution to the maintenance of mitotic spindle bipolarity

Figures

• 

  FIGURE 1:. Metaphase Myo10 depleted cells exhibit a significant increase in the frequency of multipolar spindles. (A–D) Westerns blots of whole cell extracts prepared from WT, KO-1, and KO-2 HeLa cells (A), HeLa cells treated twice for 48 h each with a control, nontargeting siRNA or a Myo10 Smartpool siRNA, and collected 12, 24, and 48 h after starting the second treatment (B), nontransduced cKO MEFs and cKO MEFs 0, 6, 12, 24, and 48 h after transduction with lentiviral cre (C), and MEFs from a WT B6 mouse and from the tm1d Myo10 KO mouse (D), all probed for Myo10 and β-actin as a loading control. (E1–E4) A metaphase HeLa cell harboring a bipolar spindle that was stained for γ-tubulin, α-tubulin, and DNA (DAPI), along with the overlay (see Z-Stack Movie 1). (F1–F4) Same as E1–E4 except that this HeLa cell harbored a semibipolar spindle (see Z-Stack Movie 2). The two γ-tubulin-positive poles with asterisks indicate they were in different z planes, resulting in chromosome distortion (arrow in F3). (G1–G4) Same as E1–E4 except that this HeLa cell harbored a multipolar spindle (see Z-Stack Movie 3). (H) Quantitation at metaphase of the percent of bipolar, semibipolar, and multipolar spindles in…

• 

  FIGURE 2:. Metaphase Myo10 depleted cells exhibit increased frequencies of spindles that are not parallel to the substratum, misaligned chromosomes, and lagging chromosomes. (A–C) Average distance in Z between the two poles of bipolar WT HeLa, KO-1, and KO-2 at metaphase (A), anaphase (B) and telophase (C; 180 WT, 234 KO-1 and 228 KO-2 cells from four experiments; see Z-Stack Movies 4–6). (D) The distributions of the values in (A) in 1-µm intervals. (E1–E6) Shown are representative examples of misaligned chromosomes at metaphase in KO-1 cells undergoing bipolar mitosis (E1) or multipolar mitosis (E2), and lagging chromosomes at metaphase and telophase in KO-1 cells undergoing bipolar mitosis (E3 and E5, respectively) or multipolar mitosis (E4 and E6, respectively; arrows point to lagging chromosomes). (F) Percent of mitotic WT HeLa, KO-1 and KO-2 that possess misaligned and/or lagging chromosomes (258 WT, 308 KO-1 and 330 KO-2 cells from four experiments). All mag bars are 10 µm.

• 

  FIGURE 3:. Myo10 localizes exclusively in metaphase cells to the spindle and to the tips of retraction fibers and dorsal filopodia. (A1–A6) An interphase Myo10 KO-1 cell expressing mScarlet-Myo10 stained with Alexa488-labeled Phalloidin and DAPI to show the accumulation of mScarlet-Myo10 at the tips of ventral filopodia. (A7 and A8) Same as in A1–A6 except an apical section to show the accumulation of mScarlet-Myo10 at the tips of dorsal filopodia. (B1–B4) A metaphase Myo10 KO-1 cell expressing mScarlet-Myo10 stained as in A1–A8 to show the dorsal, mScarlet-Myo10-positive filopodia at the cell’s equator. (B5–B8) Same as B1–B4 except a ventral section to show the accumulation of mScarlet-Myo10 at the tips of retraction fibers (see Z-Stack Movie 7). (C1 and C2) A metaphase Myo10 KO-1 cell expressing mScarlet-Myo10 stained with anti-γ-tubulin-AF488 and DAPI (equatorial section; see Z-Stack Movie 8). (D1–D4) An interphase HeLa cell stained with anti-Myo10, Alexa488-labeled Phalloidin and DAPI to show the accumulation of endogenous Myo10 at the tips of ventral (D1 and D2) and dorsal (D3 and D4) filopodia. (E1–E4) A metaphase HeLa cell stained as in D1–D4 to show the accumulation of endogenous…

• 

  FIGURE 4:. Causes of spindle multipolarity in Myo10-depleted HeLa cells and Myo10 KO MEFs. (A1–A4). Representative images of cells stained for γ-tubulin, centrin-1, and DNA that exhibit centriole disengagement (A1), PCM fragmentation (A2; shown are three different focal planes in one cell), centrosome declustering (A3), and partial centrosome declustering (A4). (B) Percent of multipolar WT HeLa, KO-1, and KO-2 that have more than four centrioles/two centrosomes. (C) Percent of multipolar WT HeLa, KO-1, and KO-2 with more than four centrioles/two centrosomes that exhibit centrosome declustering only (blue), centrosome declustering plus acentriolar foci (purple), centrosome declustering plus centriole disengagement (green), or centrosome declustering plus acentriolar foci and centriole disengagement (yellow). (D) Percent of multipolar HeLa cells exhibiting the mitotic defects described in Panel C that had been treated with control siRNA or Myo10 siRNA, and that had more than four centrioles/two centrosomes. (E) Percent of multipolar KO-1 and KO-2 with four centrioles/two centrosomes. (F) Percent of multipolar WT HeLa, KO-1, and KO-2 with four centrioles/two centrosomes that exhibit…

• 

  FIGURE 5:. PCM/pole fragmentation is a force-dependent event that is not associated with a defect in TPX2 localization. (A) Shown is the percent of unsynchronized KO-1 and KO-2 cells containing only two centrosomes that exhibited y-tubulin-positive, centriole-negative PCM fragments at prophase, prometaphase and metaphase. (B–E) Representative images of WT and KO-1 cells at prophase (B), prometaphase (C), and metaphase (D and E) that were stained for α-tubulin, TPX2 and DNA (plus overlay) while undergoing bipolar divisions (B–D) or a multipolar division (E). All the mag bars are 10 µm.

• 

  FIGURE 6:. The defect in PCM/pole integrity is not due to an obvious defect in spindle pole maturation. (A–C) Representative images of unsynchronized KO-1 cells stained at prophase (A), prometaphase (B), and metaphase (C) for centrin-1, DNA, and the pole protein CDK5Rap2. The enlarged insets show the signal for CDK5Rap2 only (the white arrowheads in the insets for (A) point to CDK5Rap2-postive pericentriolar satellites). (D) Quantitation of the amount of CDK5Rap2 at poles in WT HeLa, KO-1, and KO-2 at prophase, prometaphase and metaphase (125 to 150 cells per condition from three independent experiments). The fact that the pole content of CDK5Rap2 is higher at prophase than at prometaphase and metaphase may be due to the presence of CDK5Rap2-positive pericentriolar satellites surrounding the poles at prophase but not at prometaphase and metaphase. All mag bars are 5 µm.

• 

  FIGURE 7:. Complementation experiments reveal the contributions made by Myo10’s FERM and MyTH4 domains to spindle pole integrity and supernumerary centrosome clustering. (A) Westerns blots of whole cell extracts prepared from WT HeLa, KO-1, and KO-1 rescued with mScarlet tagged WT Myo10, mut-FERM Myo10 or mut-MyTH4 Myo10 plus/minus doxycycline addition, and probed for Myo10 and for β-actin as a loading control. (B) Quantitation at metaphase of the percent of multipolar spindles in WT HeLa, KO-1, and KO-1 rescued with WT Myo10 (338 cells), mut-FERM Myo10 (386 cells) or mut-MyTH4 Myo10 (335 cells) (from three experiments). (C) The efficiency of supernumerary centrosome clustering in KO-1 rescued with WT Myo10 (76 cells), mut-FERM Myo10 (69 cells) or mut-MyTH4 Myo10 (75 cells; from three experiments). (D) The percent of multipolar cells with two centrosomes in KO-1 rescued with WT Myo10 (24 cells), mut-FERM Myo10 (44 cells), or mut-MyTH4 Myo10 (45 cells; from three experiments). (E) The percent of multipolar KO-1 cells with two centrosomes rescued with mut-FERM Myo10 or mut-MyTH4 Myo10 that exhibited acentriolar foci only or centriolar disengagement only (from three experiments).

• 

  FIGURE 8:. Mechanisms underlying Myo10’s contribution to the maintenance of mitotic spindle bipolarity. Myo10 depleted cells exhibit two defects that lead to spindle multipolarity: PCM fragmentation (left) and impaired clustering of supernumerary centrosomes (right). PCM fragmentation creates y-tubulin-positive acentriolar foci that serve as extra spindle poles (left purple box), and is the major cause of spindle multipolarity in Myo10 KO MEFs. Impaired supernumerary centrosome clustering also results in extra spindle poles (right purple box) and is the major cause of spindle multipolarity in Myo10-depleted HeLa cells. Imaging of Halo-Myo10 KI cells shows that the myosin localizes exclusively during metaphase to the spindle and the tips of adhesive retraction fibers. The spindle signal, while faint, appears real as it lost when Myo10’s microtubule-binding MyTH4 domain is mutated. Complementation shows that Myo10 must interact with both microtubules and integrins to promote PCM/pole integrity (upper left, shaded gray), arguing that Myo10 accomplishes this task at least in part by interacting with spindle microtubules. We suggest that spindle microtubule-associated Myo10, presumably together…