Journal of Cell Science 113, 597-609 (2000) 597 Printed in Great Britain © The Company of Biologists Limited 2000 JCS0866

Evidence that actin and myosin are involved in the poleward flux of tubulin in metaphase kinetochore microtubules of crane-fly spermatocytes

R. V. Silverman-Gavrila and A. Forer* Biology Department, York University, Toronto, Ontario M3J 1P3, Canada *Author for correspondence (e-mail: [email protected])

Accepted 1 December 1999; published on WWW 31 January 2000

SUMMARY

We studied the effects of various drugs on the poleward flux We studied whether these drugs altered spindle actin. We of tubulin in kinetochore microtubules in metaphase-I used fluorescent phalloidin to visualize spermatocyte F- crane-fly spermatocytes. We used as a measure of tubulin actin, which was associated with kinetochore spindle fibers flux a ‘gap’ in acetylation of kinetochore microtubules as well as the cell cortex, the contractile ring and finger-like immediately poleward from the kinetochore; the ‘gap’ is protrusions at the poles. Spindle F-actin was no longer seen caused by a time lag between incorporation of new tubulin after cells were treated with cytochalasin D, swinholide A subunits at the kinetochore and subsequent acetylation of or a high concentration of latrunculin B, whereas a low those subunits as they flux to the pole. We confirmed concentration of latrunculin B, which did not completely that the ‘gap’ is due to flux by showing that the ‘gap’ remove the ‘gap’, caused reduced staining of spindle actin. disappeared when cells were treated briefly with the anti- Neither 2,3-butanedione 2-monoxime nor jasplakinolide tubulin drug nocodazole, which decreases microtubule altered spindle actin. These data suggest that an dynamics. The ‘gap’ disappeared when cells were treated actomyosin mechanism drives the metaphase poleward for 10 minutes with anti-actin drugs (cytochalasin D, tubulin flux. latrunculin B, swinholide A), or with the anti-myosin drug 2,3-butanedione 2-monoxime. The ‘gap’ did not disappear when cells were treated with the actin stabilizing drug Key words: Tubulin flux, Actin, Actomyosin, Kinetochore jasplakinolide. microtubule, Spindle, Spindle matrix, Metaphase, Acetylated tubulin

INTRODUCTION kinetochore microtubule stubs moved poleward, often accompanied by their associated metaphase chromosomes Poleward flux of tubulin molecules in kinetochore (Spurck et al., 1997). microtubules of metaphase cells was demonstrated as poleward What might generate the forces that act on kinetochore movement of locally uncaged fluorescent tubulin in vivo microtubules? Mitchison and Sawin (1990) suggested that (Mitchison, 1989; Mitchison and Salmon, 1992; Zhai et al., forces on kinetochore microtubules might be generated within 1995) and in reconstituted spindles in vitro (Sawin and the spindle from motor molecules (kinesins) acting on Mitchison, 1991; Desai et al., 1998). Thus, tubulin kinetochore microtubules. Alternatively, Wilson and Forer polymerization occurs at the kinetochore (plus) ends of (1994) and Forer and Pickett-Heaps (1998a) proposed that kinetochore microtubules and depolymerization occurs at the actin and associated motor proteins apply force on kinetochore pole (minus) ends (Mitchison et al., 1986). microtubules; their arguments are based on the effects of Poleward flux in kinetochore microtubules might be due to ultraviolet microbeam irradiations on chromosome movement ‘treadmilling’, a natural occurrence for microtubules in (Forer and Wilson, 1994) and on evidence that anti-actin equilibrium in vitro, where the flux is driven by a higher drugs interfere with anaphase chromosome motion (Forer subunit-lattice affinity at the microtubule plus end relative to and Pickett-Heaps, 1998a). Consistent with the latter the minus end (Margolis and Wilson, 1981). But it also is possibility, Waterman-Storer and Salmon (1997) have shown possible that the flux is caused by forces acting on kinetochore experimentally that microtubule translocation in lamellipodia microtubules, as suggested by Mitchison and Sawin (1990), is inhibited by drugs that alter actin and myosin. Moreover, similar to other cell systems where lengthwise forces on recent work indicates that microtubule dynamics are modulated microtubules can promote microtubule polymerization (e.g. by actin-microtubule interactions (Hely and Willshaw, 1998) Putnam et al., 1998; Heidemann, 1990; Zheng et al., 1993). and that actin filaments bind to astral microtubules in extracts Direct evidence that forces act on kinetochore microtubules of Xenopus eggs (Sider et al., 1999). was obtained by experiments in which ultraviolet microbeam We have tested the idea that forces on kinetochore irradiations severed kinetochore microtubules: the remnant microtubules applied by actomyosin promote tubulin flux of 598 R. V. Silverman-Gavrila and A. Forer kinetochore microtubules in metaphase; we measured Optiphot microscope using Nikon or Zeiss 100× oil-immersion kinetochore microtubule flux in the presence of inhibitors of objective lenses (NA=1.3), were recorded in real time on videotape. actin and of myosin. We used as a measure of flux the acetylation of kinetochore microtubules. Kinetochore Cell treatments microtubules are not acetylated at their kinetochore end, but For every treated cell we first followed control cells from the same are acetylated closer to the pole (Wilson and Forer, 1989). preparations or from the contralateral testis, perfused only with IR or with the maximal DMSO concentration (0.3%) used in different drug Since acetylation of newly polymerized tubulin in solutions, to ensure that cells proceeded normally from metaphase to microtubules occurs with a 5-15 minute time lag after tubulin telophase. All experiments were performed at room temperature. incorporation (Wilson et al., 1994), this suggests that the ‘gap’ The dose and duration of treatments were chosen from the literature in acetylation at the kinetochore end is due to microtubule flux (Forer and Pickett-Heaps, 1998a) for latrunculin B (LATB) and from kinetochore to pole (Wilson and Forer, 1989; Wilson et cytochalasin D (CD), from our own work (unpublished) for al., 1994). Further experiments confirmed that the ‘gap’ can be nocodazole (NOC), and from analysis of effects on anaphase chromo- used to visualise tubulin flux since taxol, known to stabilize some movement for swinholide A (SWA), jasplakinolide (JAS) and microtubule dynamics (Schiff and Horowitz, 1980), caused the 2,3-butanedione 2-monoxime (BDM). We chose the minimum dose ‘gap’ in acetylation to disappear (Wilson and Forer, 1997); i.e. that affected anaphase chromosome movement. in the presence of nanomolar concentrations of taxol the Experimental cells were treated for 10 minutes with CD, LATB, SWA, JAS or BDM, or for 4 minutes with NOC. In some experiments kinetochore microtubules were acetylated at the kinetochore. cells treated with CD for 10 minutes were allowed to recover for 10 Thus the ‘gap’ is due to flux of tubulin from kinetochore to minutes by washing out the drug with fresh IR; some cells treated pole. with NOC for 4 minutes were allowed to recover for 20 minutes. At We studied effects of inhibitors of actin and of myosin on the end of the treatments (or of the recovery period) the cells were the ‘gap’ in acetylation at the kinetochore end of microtubules, processed for fluorescence microscopy, and either lysed directly as using the presence of the ‘gap’ to indicate that microtubule flux described by Wilson et al. (1994), for studying spindle microtubules, continues and its absence to indicate the absence of flux. Our or for treatment with one of a variety of methods described below, for results, that actin and myosin inhibitors cause the ‘gap’ to studying spindle actin. disappear, suggest that actomyosin is involved in driving the Fluorescence staining flux of tubulin in kinetochore microtubules in metaphase. For every session of staining we stained at least one control coverslip as a positive control for the procedures. MATERIALS AND METHODS Double staining for acetylated and tyrosinated α-tubulin We stained for acetylated and tyrosinated α-tubulin as described by Solutions Wilson et al. (1994). Briefly, the cells were lysed for 10 minutes in a The following were used for cell preparations: halocarbon oil lysis buffer (100 mM Pipes, final pH 6.9, 10 mM EGTA, 5 mM (Halocarbon Products Corp., River Edge, NJ, USA); insect Ringer’s MgSO4, 5% DMSO, 1% Nonidet P-40), fixed for 9 minutes in 0.25% solution (IR; 0.13 M NaCl, 0.005 M KCl, 0.001 M CaCl2, 6 mM glutaraldehyde in PBS (phosphate-buffered saline: 0.13 M NaCl, Sørensen’s phosphate buffer (3 mM KH2PO4, 3 mM Na2HPO4), final 6 mM Sørensen’s phosphate buffer, final pH 7), and then rinsed with pH 6.9); fibrinogen (Calbiochem, La Jolla, CA, USA), 10 mg/ml in PBS and incubated for 20 minutes in 1 mg/ml sodium borohydride IR; thrombin (Sigma Chemical Co., St Louis, MO, USA), 50 units/ml (NaBH4) in PBS. Then the coverslips were rinsed with PBS and stored IR; and valap (vaseline: lanolin: paraffin, 1:1:1). in 1:1 PBS/glycerol at 4¡C until they were processed further. The Various drugs were dissolved in dimethylsulfoxide (DMSO); PBS/glycerol was removed by rinsing the coverslips with PBS prior measured portions were stored as stock solutions at −80¡C until use, to staining. The cells were dual stained for acetylated α-tubulin by when they were thawed and diluted with IR. The concentrations after incubation with 6-11B-1, a mouse monoclonal antibody specific for dilution were 20 µM for Cytochalasin D and 0.75-1.5 µM for acetylated α-tubulin (Sigma-Aldrich) diluted 1:300, followed by rat- latrunculin B (Calbiochem, La Jolla, CA, USA), 50 nM for Swinholide adsorbed fluorescein-isothiocyanate (FITC)-conjugated goat anti- A (Kamiya Biomedical, Seattle, WA, USA), 0.3 µM for Jasplakinolide mouse IgG (Caltag Laboratories, San Francisco, CA, USA), diluted (Molecular Probes, Eugene, OR, USA) and 10 µM for Nocodazole 1:50, and for tyrosinated α-tubulin by incubation with YL1/2 rat (Sigma, St Louis, MO, USA). The final concentrations of DMSO in monoclonal antibody specific for tyrosinated α-tubulin (a gift from Dr the diluted drug solutions ranged from 0.06% to 0.3%. 2,3-butanedione J. Kilmartin) diluted 1:500, followed by mouse-adsorbed Texas Red- 2-monoxime (Sigma) was prepared as a 10× stock solution in IR and conjugated goat anti-rat IgG (Caltag Laboratories, San Francisco, CA, kept at −80°C. Frozen portions were diluted in IR to a final USA), diluted 1:100. All antibodies were diluted in PBS. concentration of 20 mM on the day of the experiment. Double staining for filamentuous actin and α-tubulin Crane-fly spermatocyte preparations We tried different fixatives and lysing agents, with different Living crane-fly Nephrotoma suturalis (Loew) spermatocytes were concentrations, different durations and different orders of application. prepared from IV-instar larvae as described by Forer and Pickett- Some cells were fixed in 0.03-0.125% glutaraldehyde in PBS for 1-3 Heaps (1998a). Briefly, testes, dissected under halocarbon oil and minutes, followed by incubation for 15 minutes in NaBH4 (1 mg/ml), rinsed in IR, were placed in a drop of fibrinogen on a previously and then in lysis buffer (100 mM Pipes, final pH 6.9, 10 mM EGTA, flamed coverslip and pierced to allow the cells to enter the fibrinogen. 5 mM MgSO4, 5-10 % DMSO, 1-4% Nonidet P-40) from 10 minutes Thrombin was then added so that the cells, spread and flattened to overnight. We also tried to lyse and fix cells in one step, followed against the glass, were attached to the coverslips. In order to perfuse by treatment with NaBH4. Other cells were lysed for 10 minutes in the cells with different solutions, the coverslips were sealed with valap the same lysis buffer as for staining acetylated and tyrosinated α- onto perfusion chamber slides (Pickett-Heaps and Spurck, 1982). tubulin, fixed in 0.1-0.25% glutaraldehyde in PBS for 9 minutes and incubated in NaBH4 (1 mg/ml) for 15-20 minutes After the last step, Phase-contrast microscopy the coverslips were sometimes fixed for 6 minutes at −20¡C in Phase-contrast microscope images, taken with a Nikon inverted absolute methanol. In another procedure cells were fixed and treated Actin and myosin in tubulin flux 599 further using the solutions and timings described by La Fountain et treatment to after treatment. The effects of CD and LATB were al. (1992). Following each of these procedures coverslips were rinsed the same as described previously (Forer and Pickett-Heaps, in PBS and then stored in 1:1 PBS/glycerol at 4¡C until processed 1998a), and are similar to the effects of DMSO, SWA, JAS and further. BDM, namely that in prometaphase and metaphase there are Labelled phalloidins were stored at −80¡C in methanol in measured µ no drastic effects on spindle shape or appearance, or on portions from which 2 M phalloidin solutions were prepared by chromosome appearances or movements. Cells entered evaporating the methanol and resuspending the dried powder in PBS. We removed PBS/glycerol from the coverslips by rinsing them with anaphase in the presence of all the drugs; we have no data on PBS, then we incubated the coverslips with labelled phalloidins and timing of anaphase but previous experiments showed that antibodies in a moist chamber at room temperature, for 45 minutes neither CD nor LATB altered the timing of anaphase onset each. The cells were first stained for actin with Alexa 594 Phalloidin (Forer and Pickett-Heaps, 1998b). Anaphase chromosome or Alexa 488 Phalloidin (Molecular Probes), and then for α-tubulin movements were altered in the continued presence of CD or with a mouse monoclonal anti-α-tubulin antibody (Cedarlane, ON, LATB, as described previously (Forer and Pickett-Heaps, USA) diluted 1:1000. The cells that were stained with Alexa 594 1998a): the movements were blocked or slowed and, for CD, Phalloidin were incubated with FITC-conjugated goat anti-mouse IgG they returned to normal when the drugs were washed out. (Caltag) antibody diluted 1:50 to detect the mouse IgG, while the cells Similar reversible stoppages of movement were seen in cells that were stained with Alexa 488 Phalloidin were incubated with rat- treated with SWA (2 cells), BDM (2 cells) or NOC (4 cells). adsorbed Texas Red-conjugated goat anti-mouse IgG (Caltag) antibody diluted 1:50. All antibodies were diluted with PBS. As a DMSO had no effect on anaphase movements; the actin control for nonspecific binding by phalloidin some slides were stabiliser, JAS, had no effect on chromosome movement when incubated for 45 minutes with the same concentration of unlabelled added in anaphase (2 cells), and doubled chromosome velocity phalloidin prior to incubation with labelled phalloidin. The coverslips when added in metaphase (1 cell). were mounted on slides in Mowiol 488 (Polysciences, Inc., We measured distances between the two kinetochores of Warrington, PA, USA) containing as antifading agent 0.2 g/l individual bivalents to see if the drugs cause shortening of the paraphenylene-diamine (Sigma Chemical Co., St Louis, MO, USA) interkinetochore distances and thus reduction in tension, as as described by Wilson et al. (1994). occurs after ultraviolet-microbeam irradiation of individual Confocal microscopy kinetochore fibres (Spurck et al., 1997). We measured individual interkinetochore distances from the videotapes, Preparations were examined with a mixed gas krypton-argon laser BioRad MRC-600 confocal scanner (Mississauga, Canada) attached before and after addition of drug. The graphs of distance versus to a Nikon Optiphot 2 microscope. We used a Nikon Fluor oil- time generally extended from 4-5 minutes before treatment to immersion 60× objective lens (NA=1.4) for our observations. The two 4-5 minutes after treatment; we averaged the distances before fluorochromes were sometimes visualised at once using simultaneous treatment and we averaged the distances after treatment, each dual detection mode. To confirm that there was no ‘bleed through’ we being generally 10-20 values, and compared the averages using also studied the slides with filters that isolated only one laser line. Student’s t-test. Interkinetochore distances for monochiasmatic From every slide, cells in metaphase were identified using phase- bivalents were 6-7 µm and for bichiasmatic bivalents 4-5 µm. contrast microscopy; optical sections of fluorescence images were We considered differences due to the treatment as significant µ taken 0.8-1.2 m apart and were saved on optical disks. The images when P<0.001. As summarised in Table 1, control treatments used for illustration represent the superposition of 6-14 sections and (DMSO, NOC, JAS) had no effect on interkinetochore were printed using a Sony videographic printer. For phase-contrast images of the stained cells, a Nikon Fluor 40× distances, whereas treatments with the anti-actin and anti- (NA=1.3) oil immersion objective and a Panasonic videocamera myosin drugs did. Shortening on average was 9.7±4.6% (± s.d., attached to a TV screen were used. n=19). In three replicates on the same sets of bivalents (in three cells) there were no significant differences in the Fluorescence intensity analysis interkinetochore distances between replicate determinations. We analysed the ‘gap’ in acetylation at the kinetochore end of Thus, CD, LATB, SWA and BDM cause a reduction in kinetochore microtubules from confocal images of metaphase cells interkinetochore distances. double stained for acetylated and tyrosinated α-tubulin. The ‘gap’ was estimated either by eye or by comparing the fluorescence intensity Effects of treatments on tubulin flux in metaphase profiles of lines drawn along kinetochore fibres as described by kinetochore microtubules Wilson et al. (1994). We confirmed that our visual estimations were valid by choosing at random kinetochore fibres from cells that were To test the hypothesis that actin and myosin are involved in judged by eye to have ‘normal gap’, ‘no gap’ or ‘reduced gap’ tubulin flux we used as a measure of tubulin ‘flux’ a specific (considered reduced if the ‘gap’ length was less than half of the characteristic of kinetochore microtubules in metaphase crane- control ‘gap’) and by measuring their ‘gap’ with computer programs fly spermatocytes: a ‘gap ‘ in acetylation at their kinetochore as described by Wilson et al., (1994). We chose 12 kinetochore fibres end. We treated the cells with the same inhibitors used by of cells from each of the three categories, corresponding to 12 Waterman-Storer and Salmon (1997) to inhibit actomyosin- different sessions of staining. We chose from each cell a fibre that was based retrograde flow of microtubules in lamella of migrating found in only one optical section. In almost all cases the measured epithelial cells, namely CD to inhibit actin and BDM to inhibit values agreed with our visual estimates (e.g. Figs 1, 2). actomyosin. In addition we treated cells with two other actin inhibitors (SWA and LATB), an actin stabilizer (JAS), and an inhibitor of microtubule dynamics (NOC): we studied the RESULTS effects of these drugs on the acetylation of kinetochore microtubules. Effects of treatments on spindles in vivo In metaphase control cells rinsed only with insect Ringer’s For each drug we followed individual living cells from before solution, flagella and kinetochore microtubules are acetylated, 600 R. V. Silverman-Gavrila and A. Forer

Fig. 1. (A,B) Fluorescence intensity profiles of a kinetochore fibre in a control metaphase primary spermatocyte judged visually to have a normal ‘gap’ in acetylation. The points represent fluorescence intensity at the pixels along the lines drawn on the fibres in C and D; the scans proceed in the direction from kinetochore to pole. (A) Tyrosinated α-tubulin: I indicates the kinetochore position. (B) Acetylated α-tubulin: II indicates the position of maximum acetylation. The difference in distance between I and II along the abscissa gives the length of the ‘gap’: 1.6 µm, a normal length (see Wilson et al., 1994). (C,D) Confocal images of the control cell from which the graphs in A and B were obtained, illustrating staining with antibodies for tyrosinated (total) (C) and acetylated α-tubulin (D). These images are the summation of several optical sections in the series, whereas the graphs were obtained from single optical sections. The staining for total α-tubulin ends at the kinetochore, while the staining for acetylated α-tubulin ends in front of the kinetochore (the arrow in D indicates kinetochore position). a, flagellum; b, astral microtubules; c, kinetochore microtubules; d, nonkinetochore spindle microtubules. Bar, 10 µm. asters and nonkinetochore spindle microtubules are not (Fig. 1976; Vasquez et al., 1997; Mikhailov et al., 1998). NOC 1D). Kinetochore fibre acetylation is incomplete, tapering treatment caused the disappearance of nonkinetochore gradually toward the kinetochore, leaving a ‘gap’ in staining spindle microtubules and astral microtubules in 32 out of 37 immediately poleward from the kinetochore (compare Fig. metaphase cells; only kinetochore and flagella microtubules 1B,D). In computer assisted measurements (e.g. Fig. 1A,C), remained. The ‘gap’ disappeared in all NOC treated cells the average ‘gap’ length in control cells (rinsed only with insect (Fig. 4A), similar to experiments using taxol that showed that Ringer’s solution) was 1.62±0.8 µm (± s.d., n=12), near the the ‘gap’ is due to tubulin flux in kinetochore microtubules value of 1.7±0.7 µm (± s.d.) reported previously (Wilson et al., (Wilson and Forer, 1997). When the cells treated with NOC 1994). 95.5% of the control cells rinsed only with IR had ‘gaps’ were allowed to recover for 20 minutes, a normal ‘gap’ in acetylation at the kinetochore. 94% of control cells treated reappeared in 66.6% of the cells (Fig. 3C). Our data on effects with 0.3% DMSO had ‘gaps’ (Fig. 3A) similar in size and of NOC confirm that the ‘gap’ in acetylation is a measure of appearance to those in control cells (Fig. 3B). tubulin flux. In assessing the effects of treatments on tubulin flux we Having confirmed that the ‘gap’ can be used as a measure studied first the effect of NOC, a tubulin depolymerizing drug of poleward tubulin flux, we assessed whether anti-actin drugs, that also affects microtubule dynamics (e.g. Hoebeke et al., an actin stabilizing drug, and an anti-myosin drug affected the Actin and myosin in tubulin flux 601

Fig. 2. (A,B) Fluorescence intensity profiles of a kinetochore fibre in a metaphase primary spermatocyte treated with cytochalasin D and judged visually not to have a ‘gap’ in acetylation. The points represent fluorescence intensity at the pixels along the lines drawn on the fibres in C and D; the scans proceed in the direction from kinetochore to pole. (A) Tyrosinated α-tubulin: I indicates the kinetochore position. (B) Acetylated α-tubulin: II indicates the position of maximum acetylation. The difference in distance between I and II along the abscissa gives the length of the ‘gap’; 0.1 µm, essentially zero. (C,D) Confocal images of the cytochalasin D-treated cell from which the graphs in A and B were obtained, illustrating staining with antibodies for tyrosinated (total) (C) and acetylated α-tubulin (D). These images are the summation of several optical sections in the series, whereas the graphs were obtained from single optical sections. Both total α-tubulin staining and acetylated α-tubulin staining end at the kinetochores (the arrow in D indicates kinetochore position). Bar, 10 µm.

‘gap’. The results are summarized in Table 2. Anti-actin drugs, kinetochore end of kinetochore microtubules to disappear. In which disrupt actin filaments, i.e. CD, LATB and SWA, caused contrast with the effect of these anti-actin drugs, the actin the ‘gap’ in acetylation to disappear. CD caused the disappearance of the ‘gap’ in 96% of the cells (Fig. 4B). The Table 1. Influence of drugs on interkinetochore distances CD effect was reversible: normal ‘gaps’ reappeared in 96% of in individual bivalents the cells allowed to recover for 10 minutes after washing out Number of interkinetochore the CD (Fig. 3D). SWA caused the disappearance of the ‘gap’ Drug distances that shortened in 77.7% of the cells (Fig. 4C); the remaining cells had either DMSO 0 (2) a normal or reduced ‘gap’. LATB had more variable effects. NOC 0 (2) We studied two different concentrations of LATB. 0.75 µM JAS 0 (2) LATB caused the disappearance of the ‘gap’ in 50% of the CD 8 (9) analyzed cells, the rest having normal (Fig. 3E) or reduced gaps LATB 3 (9) µ SWA 4 (4) (Fig. 4D) in equal proportion. 1.5 M LATB caused BDM 4 (5) disappearance of the ‘gap’ in 76% of the cells (Fig. 4E), while 20% of the remaining cells had a reduced ‘gap’. Thus, all three Numbers in parentheses are the total numbers of bivalents measured for anti-actin drugs caused the ‘gap’ in acetylation at the each drug treatment. 602 R. V. Silverman-Gavrila and A. Forer stabilizing drug JAS had no effect on tubulin flux: all the spindle actin we used fluorescent phalloidin to study the F- studied cells had normal ‘gaps’ (Fig. 3F). The anti-myosin actin distribution in control and treated spermatocytes. ATPase drug BDM caused the disappearance of the ‘gap’ in all treated cells (Fig. 4F). Actin in metaphase control cells Our data show that the ‘gap’ in acetylation at the kinetochore We tried a variety of fixation/lysis protocols for phalloidin end of kinetochore microtubules disappears when actin staining of F-actin in spermatocytes. We were guided in filaments are disrupted or actomyosin interactions are our assessment by previous studies that reported the presence prevented. To obtain information on the effects of the drugs on of actin filaments in the cell cortex (oriented parallel to the

Fig. 3. Confocal images of metaphase primary spermatocytes double stained with antibodies for acetylated (left panels) and tyrosinated (total) α-tubulin (right panels) in which there is a ‘gap’ in acetylation at the kinetochore. (A) Cell treated with DMSO for 10 minutes. (B) Control cell: the same cell as in Fig. 1. (C) Cell treated with NOC for 4 minutes, rinsed with IR and lysed 20 minutes later. Non-kinetochore spindle microtubules are recovered, but not astral microtubules. (D) Cell treated with CD for 10 minutes, rinsed with IR and lysed 10 minutes later. (E) Cell treated with 0.75 µM LATB for 10 minutes. (F) Cell treated with JAS for 10 minutes. Bars, 10 µm. Actin and myosin in tubulin flux 603

Fig. 4. Confocal images of metaphase primary spermatocytes double stained with antibodies for acetylated (left panels) and tyrosinated (total) α-tubulin (right panels) in which there is no ‘gap’ in acetylation at the kinetochore end of kinetochore microtubules. (A) Cell treated with NOC for 10 minutes; there are no astral and nonkinetochore spindle microtubules. (B) Cell treated with CD for 10 minutes: the same cell as in Fig. 2. (C) Cell treated with SWA for 10 minutes. (D) Cell treated with 0.75 µM LATB for 10 minutes; the ‘gap’ is not completely reduced. (E) Cell treated with 1.5 µM LATB for 10 minutes. (F) Cell treated with BDM for 10 minutes. Bars, 10 µm. cell surface), within the cellular projections that extend and stained with labelled phalloidin, have all the above actin radially from the poles, in the contractile ring, and between the filament subpopulations (Fig. 5). Additional methanol fixation chromosomes and the poles, amongst the microtubules, caused loss of the cortical and the polar actin, so we omitted in metaphase and anaphase spindles (Forer and Behnke, post-fixation with cold methanol in our experiments. In both 1972a,b). non-treated (Fig. 5A) and DMSO-treated (Fig. 5B) control cells Our confocal images of control cells lysed for 9 minutes in filamentous actin at the poles was in arrays that coincided with lysis buffer, then fixed with 0.25% glutaraldehyde for 9 minutes the polar projections seen in living cells; there were actin fila- 604 R. V. Silverman-Gavrila and A. Forer

Table 2. Effects of various drugs on the ‘gap’ in acetylation at the kinetochore end of kinetochore microtubules and on F-actin distribution Effects on ‘gap’ Number of ‘gap’ present ‘gap’ reduced ‘gap’ absent Spindle Treatments cells (% of cells) (% of cells) (% of cells) actin Insect Ringer’s controls 45 95.5% 4.5% Present DMSO controls 17 94.1% 5.9% - Present Nocodazole 37 - - 100% Not assessed Nocodazole reversed 6 66.6% - 33.3% Not assessed Cytochalasin D 32 4% - 96% Absent Cytochalasin D reversed 25 96% - 4% Not assessed Latrunculin B (0.75 µM) 20 25% 25% 50% Reduced Latrunculin B (1.5 µM) 25 4% 20% 76% Absent Swinholide A 9 11.1% 11.1% 77.7% Generally absent Jasplakinolide 7 100% - - Normal 2,3-butanedione 2-monoxime 21 - - 100% Normal ments in the spindle extending from chromosomes to the poles, Fragments appeared aggregated in brightly stained triangular in close association with kinetochore microtubules, and there spikes and round patches connected by randomly distributed were actin filaments under the cell membrane with a short actin filaments (Fig. 6A). The polar actin protrusions also longitudinal distribution. F-actin was also seen in the contractile were disrupted and seemed to agglomerate near the cell ring of cells in cytokinesis (Fig. 5C). surface, as described in living cells treated with CD by Forer The staining with fluorescent phalloidin is due to binding by and Pickett-Heaps (1998a) and as described in fluorescence the phalloidin moiety and not the fluorescent moiety, because images by La Fountain et al. (1992). there was no staining of cortex or spindle when non-labelled SWA generally brought about a total absence of actin phalloidin was added prior to the fluorescent phalloidin: there filaments (Fig. 6B); only the nonspecific fluorescence of was at most only weak staining of chromosomes and of chromosomes and centrioles could be seen in the phalloidin centrioles in some cells (Fig. 5D). channel. In a few SWA-treated cells there was some greatly While our usual procedure did indeed stain actin in the cell reduced spindle actin and no cortical or polar actin. cortex and in the spindle, we were concerned about the possible 0.75 µM LATB caused varying degrees of depolymerization relocation of actin into the spindle during the initial lysis step, of the actin array. Spindle actin remained associated with kine- as suggested by La Fountain et al. (1992), who saw actin tochore microtubules, but the staining intensity was reduced, staining in the cell cortex and polar extrusions, but none in the ranging from slightly reduced fluorescence to almost complete spindle. We applied the protocol of La Fountain et al. (1992) disappearance of actin staining, the majority of the cells having to cells on coverslips and found the same distribution of actin roughly half the spindle actin fluorescence (Fig. 6C), as judged they described, i.e. cortical and polar actin, but no spindle actin by quantitative comparison of confocal images with (Fig. 5E, left panel). However, the spindle did not stain with corresponding images in control cells. The cortical actin was antibodies against tubulin, though the flagella at the poles did disrupted; fragments seemed to gather in amorphous (Fig. 5E, right panel); the flagella staining acts as a ‘positive’ aggregates with reduced phalloidin labeling (Fig. 6C). 1.5 µM control for the procedures, suggesting that spindle actin was LATB caused the absence of detectable actin filaments, and not accessible to labelled phalloidin. only the nonspecific labelling of chromosomes and centrioles To further test the possibility that actin might have moved was seen (Fig. 6D). into the spindle during our lysis procedure, we pre-fixed JAS treatment did not result in significant modifications in cells with glutaraldehyde prior to lysis: we reasoned that spindle actin or polar actin staining. Kinetochore actin fibers glutaraldehyde would fix actin in place and prevent were not disrupted; moreover, actin filaments outside the ‘relocation’. In our best procedure (fixation with 0.05% spindle area appeared thicker than in controls (Fig. 6E), due glutaraldehyde for 1 minute followed by lysis for 1 hour), there probably to actin polymerization (Bubb, 1994). was a high level of cytoplasmic background staining, but BDM did not disrupt spindle actin filaments, polar actin nonetheless spindle actin was clearly visible, as were cortical projections or cortical actin (Fig. 6F). and polar actin (Fig. 5F). Thus we are confident that our usual To sum up, all the anti-actin drugs caused loss of spindle (‘pre-lysis’) procedure visualises the distribution of actin actin to varying degrees. Neither JAS nor BDM altered filaments representative of living cells. spindle actin. The effects on actin filaments seem to match the effects on tubulin flux in that the low concentration of Effects of drugs on actin in metaphase cells LATB, which had the least effect on spindle actin, also had For all experiments we used the procedure described above, the least effect on tubulin flux. Taken together with the effects which preserved the cellular distribution of both actin and on the ‘gap’ in acetylation, the data suggest that forces from tubulin (10 minutes lysis followed by 9 minutes 0.25% spindle actomyosin induce tubulin flux from kinetochore to glutaraldehyde fixation). The effects of drugs on spindle actin pole. are summarized in Table 2. CD treatment caused the disappearance of spindle actin and Effects of treatments on spindle morphology fragmentation of the cortical and polar actin filaments. Some treatments caused modifications of normal spindle Actin and myosin in tubulin flux 605

Fig. 5. Confocal images of primary spermatocytes stained for actin (left panels) and for total α-tubulin (right panels). Unless otherwise indicated, cells were lysed, fixed with glutaraldehyde and then stained as described in the text. a, spindle actin; b, polar actin; c, cortical actin. (A) Control metaphase cell. (B) DMSO-treated metaphase cell. (C) Control cell in cytokinesis; the arrow points to the cleavage furrow. (D) Metaphase cell preincubated with unlabelled phalloidin before staining with labelled phalloidin. (E) Metaphase cell fixed and lysed using procedures as in La Fountain et al. (1992). (F) Metaphase cell fixed with glutaraldehyde and then lysed. Bars, 10 µm. morphology. In control cells, rinsed with IR or treated only concave (Fig. 4A). The fact that not all treated cells with no with DMSO, spindle microtubules and the two flagella at each ‘gap’ had flat or concave poles, suggests that modifications of pole converge to the poles from opposite directions, while the ‘gap’ are independent of modifications of spindle pole astral microtubules extend radially from the poles (Fig. 3A,B). morphology. In a few cells some nonkinetochore spindle NOC treatment resulted in depolymerization of astral microtubules were not depolymerized and were acetylated. microtubules and generally caused the poles to appear flat or When NOC treated cells were rinsed and remained in IR for 606 R. V. Silverman-Gavrila and A. Forer

Fig. 6. Confocal images of metaphase primary spermatocyte treated for 10 minutes with various drugs and stained for actin (left panels) and for total α-tubulin (right panels). All cells were lysed and fixed with glutaraldehyde. (A) Cell treated with CD. (B) Cell treated with SWA. (C) Cell treated with 0.75 µM LATB. (D) Cell treated with 1.5 µM LATB. (E) Cell treated with JAS. (F) Cell treated with BDM. Bars, 10 µm.

20 minutes, the astral microtubules often did not return and the ends of kinetochore microtubules disappears after treatment poles remained flat or concave (Fig. 3C). All BDM (Fig. 4F) with anti-actin drugs or an anti-myosin drug, and that the anti- and the majority of 1.5 µM LATB (Fig. 4E) treated cells had actin drugs depolymerise the actin that is associated with the a greatly reduced spindle width as compared to control cells. kinetochore spindle fibres. We interpret these results to mean In about half the BDM treated cells nonkinetochore spindle that actomyosin produces force on kinetochore microtubules microtubules were acetylated (Fig. 4F). that causes poleward flux of tubulin. This interpretation relies on two main assumptions: (1) that the drugs used specifically affect actin or myosin, and (2) that the ‘gap’ in acetylation at DISCUSSION the kinetochore end of kinetochore microtubules represents tubulin flux. We discuss these assumptions in turn. Our data show that the ‘gap’ in acetylation at the kinetochore The anti-actin drugs that caused the disappearance or a Actin and myosin in tubulin flux 607 reduction in length of the ‘gap’ (Table 2) are reported to inhibit Consider, for example, the suggestion that the ‘gap’ might be or depolymerize actin; we confirmed that each of them greatly due to oscillations of chromosomes from pole to pole, that the disrupted spindle actin by staining for F-actin with ‘gap’ is formed not because of tubulin flux to the pole but rather fluorescently labelled phalloidin. The effects ranged from because the non-acetylated newly polymerised microtubules are complete disappearance of spindle actin to reduced levels of added onto older sections of kinetochore microtubules to actin associated with kinetochore fibres (Table 2). Our staining replace microtubules depolymerised during the oscillations. of spindle actin is not an artefact due to the fluorescent label First, in crane-fly spermatocytes ‘gaps’ are symmetric on the itself because pre-incubation with unlabelled phalloidin blocks two sides of each bivalent (Wilson and Forer, 1989; Wilson et binding of fluorescent phalloidin (Fig. 5D). Staining of spindle al., 1994). Asymmetric ‘gaps’ indeed can be detected: in actin is not an artefact due to actin relocating from the cell anaphase, each sex chromosome has an elongating fibre and a cortex, as suggested by La Fountain et al. (1992), because with shortening fibre, and the elongating fibre has ‘gaps’ that are our procedures we see actin in the cortex and in the polar longer than usual while the shortening fibre has ‘gaps’ that are protrusions, as well as in the spindle (Fig. 5A), and we see shorter than usual (Wilson and Forer, 1989; Wilson et al., 1994). similar actin in cells pre-fixed with glutaraldehyde (Fig. 5F). Thus any putative oscillations must be identical (or similar) to The absence of actin in spindles reported by La Fountain et al. both poles. Second, the ‘gaps’ extend about 2 µm from each (1992) may be due to lack of penetration after the procedure kinetochore (Wilson et al., 1994). Thus, if the ‘gaps’ are to they used, since we were unable to label spindle microtubules represent oscillations, both kinetochores in each bivalent must when we followed their fixation protocol. Thus, the anti-actin oscillate equally and regularly, in excursions that extend about drugs did affect spindle actin, and apparently the lesser the 2 µm. Regular oscillations of about this distance are seen in effect on spindle actin the lesser the effect on the ‘gap’ (Table bivalents in Mesostoma ehrenbergii (Fuge, 1989), but they are 2) and therefore on flux. not seen in video images of crane-fly spermatocytes. Even if The effect of drugs on flux would seem to be due to direct oscillations were rapid, regular and ‘invisible’, in order to effects on actin rather than secondary ‘side effects’ of the explain the effects of taxol and nocodazole one might argue that drugs: the structures and mechanisms of action are different for these drugs stabilise microtubules and thereby prevent the the three anti-actin drugs, CD (Cooper, 1987; Spector et al., oscillations, but to explain the effects of the anti-actin and anti- 1989; Urbanik and Ware, 1989), LATB (Ayscough, 1980; Coue myosin drugs one would need to say either that these drugs et al., 1987; Spector et al., 1989) and SWA (Bubb et al., 1995; affect kinetochore extensibility, invoking interactions between Bubb and Spector, 1998; Saito et al., 1998), and so would be kinetochores and actin and myosin, or that they affect expected to be their side effects. While the myosin ATPase kinetochore microtubules, invoking interactions between inhibitor BDM (Cramer and Mitchison, 1995) may have non- microtubules and actin and myosin. It seems far simpler to specific ‘side effects’, it did not affect kinetochore fibre actin interpret ‘gaps’ as representing flux, and to conclude that actin and did affect flux, implicating actin and myosin in producing and myosin produce force on the microtubules to drive the flux, force on kinetochore microtubules, a result identical to that especially since the interkinetochore distances decreased in the of Waterman-Storer and Salmon (1997), who implicated presence of the anti-actin and anti-myosin drugs (Table 1), actomyosin in movement of microtubules in lamellipodia. implying that the drugs cause a reduction in the forces acting We interpret the ‘gap’ in acetylation to be due to flux of on the kinetochore microtubules, and since there is evidence tubulin along kinetochore microtubules, from kinetochore to from completely different sets of experiments that points to pole. Direct measure of flux involves following labelled tubulin external forces acting on kinetochore microtubules (discussed in vivo; our method is indirect, studying localisation of in Forer and Wilson, 1994; Pickett-Heaps et al., 1996; Spurck acetylated tubulin in fixed/stained cells. Evidence that the ‘gap’ et al., 1997). represents flux was given and discussed by Wilson and How might actin and myosin interact with kinetochore Forer (1989, 1997) and Wilson et al. (1994). Acetylation microtubules? Spindle actomyosin might produce force on of microtubules occurs on older tubulin subunits, so it is kinetochore microtubules in much the same way Waterman- reasonable to assume that the absence of acetylation at the Storer and Salmon (1997) suggested that actomyosin produces kinetochore is because that is where new subunits enter, and that force on lamellipodial microtubules: the ‘motor’ protein the newly polymerized subunits become acetylated only later, myosin in a spindle matrix might react with actin that is bound after they move polewards (Wilson et al., 1994). Experiments to kinetochore microtubules to propel the actin (and attached with taxol confirmed this assumption and ruled out alternative kinetochore microtubules) poleward (e.g. Pickett-Heaps et al., interpretations that the localised ‘gap’ is due to other processes 1996, 1997; Forer and Pickett-Heaps, 1998a). There is (discussed in Wilson and Forer, 1989; Wilson et al., 1994), such evidence in other systems for tight association between spindle as asymmetric localisation of deacetylating enzymes; (astral) microtubules and F- actin (Sider et al., 1999) and nanomolar concentrations of taxol, known to inhibit exchange between actin and microtubules in the pre-prophase band in of tubulin at microtubule ends, caused disappearance of the plant cells (e.g. Pickett-Heaps et al., 1999), and there is ‘gap’ (Wilson and Forer, 1997). We confirmed this result using evidence that the drugs that we used block poleward force: CD NOC, another drug known to inhibit tubulin dynamics: NOC and LATB block anaphase chromosome movement (Forer and caused the disappearance of the ‘gap’ (Fig. 4A). Thus, all these Pickett-Heaps, 1998a), as do SWA and BDM. Further, UV experiments indicate that the presence of the ‘gap’ represents microbeam experiments in which kinetochore microtubules are poleward flux of tubulin and, consequently, the absence of a severed also suggest that forces external to kinetochore micro- ‘gap’ represents the absence of tubulin flux. This is the simplest tubules propel the microtubules poleward (Spurck et al., 1997; interpretation of the data, especially when considered together Forer and Wilson, 1994). Thus it is not unreasonable to expect with our data on the effects of anti-actin and anti-myosin drugs. that actin is bound to kinetochore microtubules and that 608 R. V. Silverman-Gavrila and A. Forer poleward force on the actin produces poleward force on the R.V.S.G.). Two referees offered constructive criticisms, which helped kinetochore microtubules. Nor is it unreasonable that external to strengthen the manuscript. forces might influence microtubule dynamics, since in other systems externally applied forces can cause microtubule polymerization (Putnam et al., 1998; Heidemann, 1990; Zheng REFERENCES et al., 1993). Furthermore, others report that the dynamics, distribution and length of microtubules are modulated by the Ayscough, K. (1998). Use of latrunculin A Ð an actin monomer binding drug. actin cytoskeleton (Krendel et al., 1998) and that there is Meth. Enzymol. 298, 18-25. interaction between microtubules and actomyosin in the cell Bednara, J., Willemse, M. T. M. and van Lammeren, A. A. M. (1990). Organization of the actin cytoskeleton during megasporogenesis in Gasteria cortex (Canman and Bement, 1997). verrucosa visualized with fluorescent-labelled phalloidin. Acta Bot. Neerl. Other reports are consistent with this interpretation. Mitotic 39, 43-48. spindles have been shown to contain actin (Forer and Behnke, Bubb, M. R., Spector, I., Bershadsky, A. D. and Korn, E. D. (1995). 1972a; Gawadi, 1974; Forer et al., 1979; Seagull et al., 1987; Swinholide A is a microfilament disrupting marine toxin that stabilizes actin dimers and severs actin filaments. J. Biol. Chem. 270, 3463-3466. Van Lammeren, 1989; Traas et al., 1987, 1989; Bednara et al., Bubb, M. R. and Spector, I. (1998). Use of F-actin binding drugs, 1990; Schmit and Lambert, 1990; Staiger and Cande, 1991; misakinolide A and swinholide A. Meth. Enzymol. 298, 26-32. Czaban and Forer, 1992, 1994; Panteris et al., 1992; Dinis and Bubb, M. R. (1994). Jasplakinolide, a cytotoxic natural product induces actin Mesquita, 1993; Cleary, 1995; Kengen et al., 1995) and myosin polymerization and competitively inhibits the binding of phalloidin to F- (Fujiwara and Pollard, 1978; Wu et al., 1998), and inhibitors actin. J. Biol. Chem. 269, 14869-14871. Canman, J. C. and Bement, W. M. (1997). Microtubules suppress of myosin light-chain kinase inhibit spindle formation actomyosin-based flow in Xenopus oocytes. J. Cell Sci. 110, 1907-1917. (Fishkind et al., 1991). In other systems interaction between Cleary, A. L. (1995). F-actin redistributions at the division site in living actin and microtubules is mediated through microtubule- Tradescantia stomatal complexes as revealed by microinjection of associated proteins (Correas et al., 1990; Fujii et al., 1993; rhodamine-phalloidin. Protoplasma 185, 152-165. Cooper, J. (1987). Effects of cytochalasin and phalloidin on actin. J. Cell Biol. Tögel et al., 1998; Sider et al., 1999) and through actin- 105, 1473-1478. associated proteins (Fach et al., 1985); it is likely that similar Correas, I., Padilla, R. and Avila, J. (1990). The tubulin-binding sequence interactions occur in kinetochore fibres. Further, in other of brain microtubule-associated proteins, tau and MAP-2, is also involved systems there is coordination between three kinds of motor in actin binding. Biochem. J. 269, 61-64. molecules, myosins, kinesins and dyneins, which enable Coue, M., Brenner, S., Spector, I. and Korn, E. (1987). Inhibition of actin polymerization by latrunculin A. FEBS Lett. 213, 316-318. ‘cargo’ to switch between microtubule-based and actin-based Cramer, L. P. and Mitchison, T. J. (1995). Myosin is involved in postmitotic cellular transport (Euteneur et al., 1987; Tabb et al., 1998; cell spreading. J. Cell Biol. 131, 179-189. Huang et al., 1999), so it is not unreasonable to expect that Czaban, B. and Forer, A. (1992). Rhodamine-labelled phalloidin stains similar interactions occur in spindles. Finally, we have components in the chromosomal spindle fibres of crane-fly spermatocytes and Haemanthus endosperm cells. Biochem. Cell Biol. 70, 664-676. assumed that the spindle actin and myosin have not only a Czaban, B. and Forer, A. (1994). Rhodamine-phalloidin and anti-tubulin structural role, but also a functional role in generating force on antibody staining of spindle fibres that were irradiated with an ultraviolet kinetochore microtubules, as argued above. In support of this microbeam; Protoplasma 178, 18-27. idea we note the apparent consistent polarity of actin in spindle Desai, A., Maddox, P. S., Mitchison, T. J., Salmon, E. D. (1998). Anaphase fibres (Gawadi, 1974; Forer and Jackson, 1976; Forer et al., A chromosome movement and poleward spindle microtubule flux occur at similar rates in Xenopus extract spindles. J. Cell Biol. 141, 703-713. 1979) and the different distribution of actin filaments in Dinis, A. M. and Mesquita, J. F. (1993). The F-actin distribution during different division stages (Fishkind and Wang, 1993; Schmit microsporogenesis in Magnolia soulangeana Soul. (Magnoliaceae). Sex. and Lambert, 1990; Kengen et al., 1995). Plant. Reprod. 6, 57-63. We have suggested one way in which actin-myosin- Euteneuer, U., Cooper, M. C., Pryzwavsky, K. B. and Schliwa, M. (1987). Dynamic aspects of the interaction between microtubules and actin microtubules might interact to produce poleward force on filaments. J. Electron Microsc. Tech. 6, 167-173. kinetochore microtubules and thereby induce poleward flux. That Fach, B. L., Graham, S. F. and Keates, R. A. B. (1985). Association of fodrin is not the only possible model that could explain our data; it is with brain microtubules. Can. J. Biochem. Cell. Biol. 63, 372-381. conceivable, for example, that actin, myosin and tubulin are part Fishkind, D. J., Cao, L.-G. and Wang, Y.-L. (1991). Microinjection of the of an interacting ‘tensegral’ network, as suggested by Pickett- catalytic fragment of myosin light chain kinase into dividing cells: effects on mitosis and cytokinesis. J. Cell Biol. 5, 967-975. Heaps et al. (1997), or that actin and myosin produce force on Fishkind, D. J. and Wang, Y. (1993). Orientation and three-dimensional kinetochore microtubules in other ways. The data discussed organization of actin filaments in dividing cultured cells. J. Cell Biol. 123, above, however, all point to some role for actin and myosin in 837-848. mitotic spindles; such possible roles need to be considered. Forer, A. and Behnke, O. (1972a). An actin-like component in spermatocytes of a crane-fly (Nephrotoma suturalis Loew) I. The spindle. Chromosoma In summary, our data show that actin inhibitors (CD, LATB, (Berl.) 39, 145-173. SWA) and the myosin inhibitor BDM block the poleward Forer, A. and Behnke, O. (1972b). An actin-like component in spermatocytes flux of kinetochore microtubules, as indicated by the of a crane-fly (Nephrotoma suturalis Loew) II. The cell cortex. Chromosoma disappearance of the ‘gap’ in acetylation adjacent to the (Berl.) 39, 175-190. metaphase kinetochore. These data, which, for actin inhibitors, Forer, A. and Jackson, W. T. (1976). Actin filaments in the endosperm mitotic spindles in a higher plant, Haemanthus katherinae Baker. Cytobiol. 12, 199- are correlated with loss of spindle actin, suggest that 214. actomyosin is involved in driving the poleward flux of tubulin Forer, A., Jackson, W. T. and Engberg, A. (1979). Actin in spindles of in metaphase kinetochore microtubules. Haemanthus katherinae endosperm II. Distribution of actin in chromosomal spindle fibres, determined by analysis of serial sections. J. Cell Sci. 37, 349- 371. We are grateful for the support of grants from the Natural Sciences Forer, A. and Wilson, P. J. (1994). A model for chromosome movement and Engineering Research Council of Canada (to A.F.) and to an during mitosis. Protoplasma 179, 95-105. Ontario Graduate Scholarship in Science and Technology (to Forer, A. and Pickett-Heaps, J. D. (1998a). Cytochalasin D and latrunculin Actin and myosin in tubulin flux 609

affect chromosome behaviour during meiosis in crane-fly spermatocytes. dimeric macrolides, bistheonellide A and swinholide A. J. Biochem. 123, Chromosome Res. 6, 533-549. 571-578. Forer, A. and Pickett-Heaps, J. D. (1998b). Checkpoint control in crane-fly Sawin, K. E. and Mitchison, T. J. (1991). Poleward microtubule flux in spermatocytes: unattached chromosomes induced by cytochalasin D or spindles assembled in vitro. J. Cell Biol. 112, 941-954. latrunculin treatment do not prevent or delay the start of anaphase. Schiff, P. B. and Horowitz, S. B. (1980). Promotion of microtubule assembly Protoplasma 203, 100-111. in vitro by taxol. Nature 227, 665-667. Fuge, H. (1989). Rapid kinetochore movements in Mesostoma ehrenbergii Seagull, R. W., Falconet, M. M. and Werdenburg, C. A. S. (1987). spermatocytes: action of antagonistic chromosome fibres. Cell Motil. Microfilaments: dynamic arrays in higher plant cells. J. Cell Biol. 104, 995- Cytoskel. 13, 212-220. 1004. Fujii, T., Watanabe, M., Ogoma, Y., Kondo, Y. and Aray, T. (1993). Schmit, A. C., and Lambert, A. M. (1990). Microinjected fluorescent Microtubule associated proteins MAP1A and MAP1B interact with F-actin phalloidin in vivo reveals the F-actin dynamics and assembly in higher plant in vitro. J. Biochem. 114, 827-829. mitotic cells. The Plant Cell 2, 121-138. Fujiwara, K. and Pollard, T. D. (1978). Simultaneous localization of myosin Sider, J. R., Mandato, C. A., Weber, K. L., Zandy, A. J., Beach, D., Finst, and tubulin in human tissue culture cells by double antibody staining. J. Cell R. J., Skoble, J. and Bement, W. M. (1999). Direct observation of Biol. 77, 182-195. microtubule-f-actin interaction in cell free lysates. J. Cell Sci. 112, 1947-1956. Gawadi, N. (1974). Characterization and distribution of microfilaments in Spector, I., Shochet, M., Blasberger, D. and Kushman, I. (1989). dividing locust testis cells. Cytobios 10, 17-35. Latrunculins: novel marine macrolides that disrupt microfilament Heidemann, S. (1990). Neuronal tip growth. In Tip Growth in Plant and organization and affect cell growth. I Comparison with cytochalasin D. Cell Fungal Cells (ed. I. B. Heath), pp. 284-316. New-York: Academic Press Inc. Motil. Cytoskel. 13, 127-144. Hely, T. A. and Willshaw, D. J. (1998). Short-term interactions between Spurck, T., Forer, A. and Pickett-Heaps, J. (1997). Ultraviolet microbeam microtubules and actin filaments underlie long-term behaviour in neuronal irradiations of epithelial and spermatocyte spindles suggest that forces act growth cones. Proc. R. Soc. Lond. B 265, 1801-1807. on the kinetochore fibre and are not generated by its disassembly. Cell Motil. Hoebecke, J., Van Nijen, G., and De Brabander, M. (1976). Interactions of Cytoskel. 36, 136-148. oncodazole (R17934), a new anti-tumoral drug, with rat brain tubulin. Staiger, C. J. and Cande, W. Z. (1991). Microfilament distribution in maize Biochem. Biophys. Res. Commun. 69, 319-324. meiotic mutants correlates with microtubule organization. The Plant Cell 3, Huang, J.-D., Brady, S., Richards, B., Stenoien, D., Resan, J., Copeland, 637-644. N. and Jenkins, N. (1999). Direct interaction of microtubule-and actin- Tabb, J., Molyneaux, B., Cohen, D., Kuznetsov, S. and Langford, G. based transport motors. Nature 397, 267-270. (1998). Transport of ER vesicles on actin filaments in neurons by myosin Kengen, H., Eygensteyn, J. and Van Aster, T. (1995). F-actin in mitotic V. J. Cell Sci. 111, 3221-3234. spindles of synchronized suspension culture cells of Tobacco visualized by Tögel, M., Wiche, G and Propst, F. (1998). Novel features of the light chain confocal laser scanning microscopy. Cell Biol. Internatl. 7, 585-592. of microtubule associated protein MAP1B: microtubule stabilization, self Krendel, M., Sgourdas, G. and Bonder, E. (1998). Dissasembly of actin interaction, actin filament binding, and regulation by the heavy chain. J. Cell filaments leads to increased rate and frequency of mitochondrial movement Biol. 143, 695-707. along microtubules. Cell Motil. Cytoskel. 40, 368-387. Traas, J. A., Doonan, J. H., Rawlins, D. J., Shaw, P. J., Watts, J. and Lloyd, La Fountain, J. R. Jr., Janicke, M. A., Balczon, R. and Rickards, G. K. C. W. (1987). An actin network is present in the cytoplasm throughout the (1992). Cytochalasin induces abnormal anaphase in crane-fly spermatocytes cell cycle of carrot cells and associates with the dividing nucleus. J. Cell and causes altered distribution of actin and centromeric antigens. Biol. 105, 387-395. Chromosoma 101, 425-441. Traas, J., Burgain, S. and Dumas de Vaulx, H. (1989). The organization of Margolis, R. L. and Wilson, L. (1981). Microtubule treadmills: possible the cytoskeleton during meiosis in eggplant (Solanum melongena (L.)): molecular machinery. Nature 293, 705-711. microtubules and F-actin are both necessary for coordinated meiotic Mikhailov, A. and Gundersen, G. (1998). Relationship between microtubule division. J. Cell Sci. 92, 541-550. dynamics and lamellipodium formation revealed by direct imaging of Urbanik, E. and Ware, B. (1989). Actin filament capping and cleaving microtubules in cells treated with nocodazole or taxol. Cell Motil. Cytoskel. activity of cytochalasins B, D, E and H. Arch. Biochem. Biophys. 269, 181- 41, 325-340. 187. Mitchison, T. J. (1989). Polewards microtubule flux in the mitotic spindle: Van Lammeren, A. A. M., Bednara, J. and Willemse, M. T. M. (1989). evidence from photoactivation of fluorescence. J. Cell Biol. 109, 637-652. Organization of the actin cytoskeleton during pollen development in Mitchison, T. J., Evans, L., Schulze, E. and Kirschner, M. (1986). Sites of Gasteria verrucosa (Mill.) H. Duval visualized with rhodamine-phalloidin. microtubules assembly and disassembly in the mitotic spindle. Cell, 45, 515- Planta 178, 531-539. 527. Vasquez, R. J., Hovell, B., Yvon, A.-M., Wadsworth, P. and Cassimeris, L. Mitchison, T. J. and Sawin, K. E. (1990). Tubulin flux in the mitotic spindle: (1997). Nanomolar concentrations of nocodazole alter microtubule dynamic where does it come from, where is it going? Cell Motil. Cytoskel. 16, 93-98. instability. Mol. Biol. Cell. 8, 973-985. Mitchison, T. J. and Salmon, E. D. (1992). Poleward kinetochore fiber Waterman-Storer, C. M. and Salmon, E. D. (1997). Actomyosin-based movement occurs during both metaphase and anaphase-A in newt lung cell retrograde flow of microtubules in the lamella of migrating epithelial cells mitosis. J. Cell Biol. 119, 569-582. influences microtubule dynamic instability and turnover and is associated Panteris, E., Apostolakos, P. and Galatis, B. (1992). The organization of F- with microtubule breakage and treadmilling. J. Cell Biol. 139, 417-434. actin in root tip cells of Adiantum capillus veneris throughout the cell cycle. Wilson, P. J. and Forer, A. (1989). Acetylated alpha tubulin in spermatogenic Protoplasma 170, 128-137. cells of the crane fly Nephrotoma suturalis: kinetochore microtubules are Pickett-Heaps, J. D. and Spurck, T. (1982). Studies on kinetochore function selectively acetylated. Cell Motil. Cytoskel. 14, 237-250. in mitosis I: effects of colchicine and cytochalasin on mitosis in the diatom Wilson P. J., Forer, A. and Leggiardo, C. (1994). Evidence that kinetochore Hantzschia amphioxys. Eur. J. Cell Biol. 28, 77-82. microtubules in crane-fly spermatocytes disassemble during anaphase Pickett-Heaps, J. D., Forer, A. and Spurck, T. (1996). Rethinking anaphase: primarily at the poleward end. J. Cell Sci. 107, 3015-3027. where ‘Pac-Man’ fails and why a role for the spindle matrix is likely. Wilson, P. J. and Forer, A. (1997). Effects of nanomolar taxol on crane-fly Protoplasma 192, 1-10. spermatocyte spindles indicate that acetylation of kinetochore microtubules Pickett-Heaps, J. D., Forer, A. and Spurck, T. (1997). Traction fibre: towards can be used as a marker of poleward tubulin flux. Cell Motil. Cytoskel. 37, a ‘tensegral’ model of the spindle. Cell Motil. Cytoskel. 37, 1-6. 20-32. Pickett-Heaps, J. D., Gunning, B. E. S., Brown, R. C., Lemmon, B. E. and Wu, X., Kocher, B., Wei, Q. and Hammer, J. A. (1998). Myosin Va associates Cleary, A. L. (1999). The cytoplast concept in dividing plant cells: with microtubule-rich domains in both interphase and dividing cells. Cell cytoplasmic domains and the evolution of spatially organized cell division. Motil. Cytoskel. 40, 286-303. Am. J. Bot. 86, 153-172. Zhai, Y., Kronebusch, P. J. and Borisy, G. G. (1995). Kinetochore Putnam, A., Cunningham, J., Dennis, R., Linderman, J. and Mooney, D. microtubule dynamics and the metaphase-anaphase transition. J. Cell Biol. (1998). Microtubule assembly is regulated by externally applied strain in 131, 721-734. cultured smooth muscle cells. J. Cell Sci. 111, 3379-3387. Zheng, J., Buxbaum, R. E. and Heidemann, S. R. (1993). Investigation of Saito, S., Watabe, S., Ozaki, H., Kobayashi, M., Suzuki, T., Kobayashi, H., microtubule assembly and organization accompanying tension-induced Fusetani, N. and Karaki, H.-J. (1998). Actin depolymerizing effect of neurite elongation. J. Cell Sci. 104, 1239-1250.