HEF1 localization and ultrastructure of HEF1-induced cells at abscission. (A) M-phase localization of endogenous HEF1 in MCF-7 cells (left). Cells are examined at i) metaphase, ii) anaphase, iii) telophase, and iv) at the point of abscission. Red, HEF1; green, actin. (B) Normal ultrastructure at the stage of abscission in the presence or absence of overexpressed HEF1. Electron micrograph of the postmitotic bridge at low (i and iii, left) and high (ii and iv, right) magnification in CM1 cells (i and ii, top) or HEF1.M1 cells (iii and iv, bottom) in absence of tetracycline.

Inhibition of RhoA or p160ROCK kinase activity rescues the defect in persistent rounding and abscission caused by HEF1 overexpression. (A) HEF1.M1 cells in the presence (+T) and absence (–T) of tetracycline were arrested with nocodazole (0 h) and released from the taxol block for 3 h (3 h). Numerous cells trapped before abscission are seen in (–T; 3 h) cells, indicated by arrows. C3 transferase protein (C3) or Y-27632 was added to the –T, 3 h HEF1.M1 cells, and observed after 30 (C3) or 10 (Y-27632) min. Addition of either of the inhibitors allowed the cells to undergo abscission and spread. (B) FACS profile of the cells shown in A. Treatment with either C3 transferase or Y-27632 induces movement of cells from 4N stage to 2N.

siRNA depletion of HEF1. (A) Depletion of HEF1 protein by siRNA treatment. Western blot analysis of HEF1 protein levels after treatment with siRNA. scr, MCF-7 cells transfected with a scrambled siRNA. siHEF1, MCF-7 cells transfected with siRNA directed against HEF1. Cell lysates were collected 48 h after transfection and immunoblotted with anti-HEF1 and anti-p130Cas antibody to show specific depletion. Quantitation of images indicates >85% depletion of HEF1 expression in the experiment shown: the typical range of depletion for data reported in this manuscript is 80–90%. (B) Flow cytometric analysis of asynchronous cells 48 h posttransfection with control scrambled (scr, left) or HEF1-targeted (siHEF1, right) siRNAs. Asterisk indicates >4N peak. (C) Immunofluorescence of MCF-7 cells synchronized with double thymidine block, treated during block with siRNA to HEF1 or control (Scr), 8 or12 h after release, as indicated. HEF1 is shown in red; α-tubulin in green. (D) Phase micrographs of synchronized thymidine-blocked MCF-7 depleted for HEF1 or control (Scr), 8–12 h postrelease (as indicated). Arrows indicate cells with clear metaphase plates but maintaining attachment. Asterisk demonstrates HEF1-depleted cell with separated DNA, lacking cleavage furrow. Quantitation represents the percentage of cells observed rounded for mitosis at each time point as well as the total percentage observed as mitotic within 12 h after release from block. (E) Immunofluorescence of MCF-7 cells synchronized with double thymidine block, treated during block with siRNA to HEF1 or control (Scr). α-Tubulin is shown in red; actin in green. Maximal rounding of Scr-depleted cells is shown at 8 h, for HEF1-depleted at 12 h.

HEF1 positively regulates RhoA activation and associates with ECT2 during M phase. (A) Overexpression of HEF1 causes persistence of ECT2 in the late stages of M phase. Cell lysates from HEF1.M1 in the presence (+Tet) and absence (–Tet) of tetracycline were prepared as described in Figure 4A, and immunoblotted with α-ECT2 antibody. Overexpression of HEF1 leads to persistence of ECT2 in mitotic cells released from either nocodazole (left) or taxol (right). (B) HEF1 overexpression leads to persistence of ECT2 (see arrow) at the midbody during anaphase. HEF1.M1 cell line, in the presence (top) and absence (bottom) of tetracycline. Red, ECT2; green, α-tubulin. (C) Epitope-tagged HEF1 and ECT2 reciprocally coimmunoprecipitate. Cell lysates from cells cotransfected with epitope-tagged HEF1 and ECT2, or negative vector controls, were either directly analyzed by Western blot as WCL, or subjected to IP using the antibodies as indicated, followed by immunoblotting (IB) with the specified antibodies. For i and ii, left two panels represent WCL directly probed by Western; extreme right panel shows immunoprecipitates probed with antibody against the coimmunoprecipitated protein; i, α-HA antibody was used to immunoprecipitate HA-ECT2 containing complexes, and immunoprecipitates were probed with α-HEF1; WCL were probed with α-HA and α-GFP. Lane 1, pCDNA6HA and pEGFP-HEF1; lane 2, pCDNA6HA-pEGFP-HEF1; lane 3, pCDNA6HA-ECT2 and pEGFP; and lane 4, pCDNA6HA-ECT2 and pEGFP-HEF1. ii, α-Flag antibody was used to immunoprecipitate Flag-HEF1 containing complexes, and immunoprecipitates were probed with α-ECT2; WCL were probed with α-Flag and α-HA. Lane 1, pCDNA6HA and pCatchFlag; lane 2, pCDNA6HA-pCatchFlag-HEF1; lane 3, pCDNA6HA-ECT2 and pCatchFlag; and lane 4, pCDNA6HA-ECT2 and pCatchFlag-HEF1. Asterisk (*) indicates immunoprecipitated protein. (D) Endogenous ECT2 preferentially immunoprecipitates with hyperphosphorylated HEF1. Whole cell lysates from asynchronous or nocodazole arrested MCF-7 cells were directly loaded to an SDS-PAGE gel (WCL), or 1 mg of lysate was used for immunoprecipitation with antibody to ECT2 or a negative control sera from nonimmunized rabbits (control). Blots were visualized with antibody to HEF1 or to ECT2.

Depletion of HEF1 inhibits RhoA activation and ECT2 expression. (A) HEF1 depletion causes hypoactivation of RhoA in mitotic cells. Cell lysates were prepared from nocodazole-blocked cells (M) that have been treated with a scrambled siRNA (scr) or with siRNA targeted to HEF1 (siHEF1) or ECT2 (siECT2). Cell lysates were probed for presence of activated RhoA or total RhoA (assessed as in Figure 4A). (B) FACS profiles of nocodazole-arrested cells treated with scrambled oligo (left), siHEF1 (middle), and siECT2 (right) and indicate equivalent cell synchronization of all the treated cells. (C) HEF1 depletion reduces steady-state levels of ECT2 in mitotic cells. Cells were prepared as in A after treatment with scrambled siRNA, or siRNA to HEF1, and lysates were probed with antibody to HEF1 (top) or ECT2 (bottom). Fifty (50), 52, 5, and 1 denote micrograms of lysate loaded per lane, to allow a visual measure of degree of depletion. (D) HEF1 or ECT2 depletion reduces activation of RhoA in interphase cells. Synchronized thymidine-blocked MCF7 depleted for HEF1, ECT2 or control (Scr) were released from block, and RhoA activation assessed for 0–12 h after release. Degree of HEF1 and ECT2 depletion was confirmed by parallel Western blots. (E) Reduction of HEF1 or ECT2 levels in double thymidine blocked cells 8 or 12 h after release, in experiment described in D. (F) Depletion of HEF1 causes a reduction in ECT2 levels in untreated asynchronous cells. Lysates of cells treated with siRNA as described in A were resolved by SDS-PAGE and probed with antibodies indicated to HEF1 or ECT2: a nonspecifically hybridizing band provides a loading control (REF). (G) Phase micrographs of synchronized thymidine-blocked MCF-7 depleted for ECT2 or control (Scr), 8–12 h postrelease, as indicated. Arrow and asterisk indicate a mitotic cell that fails in cleavage furrow ingression.

HEF1 controls the RhoA activation cycle. (A) Overexpression of HEF1 induces persistent activation of RhoA. RhoA activation was assessed in cells collected by mitotic shake-off after synchronization with thymidine and subsequent nocodazole or taxol treatment at indicated time points 0, 30, 60, and 90 min after release from both nocodazole (Noc., left) or taxol (Tax., middle) or maintained in double thymidine block (G₁, right). HEF1.M1 grown in the presence (+Tet) and absence (–Tet) of tetracycline are shown. RhoA* represents activated RhoA pulled down by GST-RBD, and RhoA (WCL) represents total RhoA in whole cell lysate: each is visualized by antibody to RhoA. HEF1-overexpressing cells showed persistent activation of RhoA at 60 and 90 min when released from either nocodazole (left) or taxol (middle). Similar results were obtained in three independent experiments, and a representative blot is shown. (B) Flow cytometry analysis of HEF1.M1 cells prepared as described in Fig. 4A, after release from nocodazole (left) or taxol (middle) or maintained in double thymidine block (right, G₁). (C) The activation of RhoA effectors in nocodazole-synchronized cells with uninduced (+Tet) or induced (–Tet) HEF1 was examined after release from nocodazole block. Left, p160ROCK was immunoprecipitated, total levels were determined by Western (p160ROCK), and used for kinase assay with a Histone H2B substrate, followed by visualization with a phospho-H2B-specific antibody, to determine activity. Middle, whole cell lysates were probed with antibody to phosphorylated (phospho-MLC) or total MLC. Below each set of Western blots, quantitation of relative levels of phosphorylated H2B and MLC in Tet+ and Tet–cells. Right, expression of Cyclin B1 in cells at comparable time points after release from nocodazole. Blots were stripped and reprobed with actin to confirm equal load.

HEF1 overexpression delays mitotic progression and induces abscission defects. (A) Induced expression of HEF1. Western blot analysis of HEF1 expression in stable MCF-7 cells expressing empty vector (CM1) or HEF1 (HEF1.M1 and HEF1.M2) from a tetracycline-repressed promoter in the presence (+T) and absence (–T) of tetracycline, 24 h after induction; quantitation reflects fold-expression relative to MCF-7 cells. Typically, the cell lines maintained with tetracycline present have levels of HEF1 reduced versus the MCF-7 parental line. Note, HEF1 occurs as a doublet migrating at 105 and 115 kDa, reflecting differentially phosphorylated forms. Details of this phosphorylation and antibody characterization have been published in Law et al. (1996). (B) Timing of progression through M phase and M-phase defects induced by HEF1 overexpression. (i) Events occurring during normal M-phase progression, time denoted in hours:minutes after initial cell rounding. Cleavage furrow formation and ingression is evident at ∼00:12, formation of midbody at ∼00:24, and abscission and cell separation complete at 1:09. Three different phenotypes observed on overexpression of HEF1 include (ii) cleavage furrow regression and formation of a binucleate cell (seen at 8:18), (iii) delay between formation of a midbody structure and abscission, and (iv) failure to undergo abscission and cell separation within the period of observation (>12 h). In this last case, daughter cells remain joined by a postmitotic bridge (white arrow). (C) Time required for each stage of M phase, calculated for 15 cells for each sample in three different experiments, using solely cells that had been shown to ultimately undergo separation (e.g., corresponding to the Figure 1B, iii phenotype described above). Results are expressed in minutes ± SE of mean required for formation of cleavage furrow after initial rounding of cell, time from cleavage furrow to formation of a midbody, and time from midbody formation to final abscission. *p < 0.001 and **p < 0.05. (D) HEF1 overexpression inhibits progress of cells through M phase. HEF1.M1 cells cultured in medium with tetracycline were synchronized at S phase with a double thymidine block, released for 5 h, blocked with nocodazole for 7 h in the presence or absence of tetracycline to induce HEF1 expression, and collected by mitotic shake-off. After nocodazole release, cells were followed for 3 h. Within panels, +/–T, with/without tetracycline; 0 h, 1 h, and 3 h, time after release from nocodazole. (E) Mitotic spindles of HEF1.M1 cells in the presence (+T) or absence (–T) of tetracycline. HEF1 is shown in red; actin in green.