Tastin is required for bipolar spindle assembly and centrosome ...

The FASEB Journal • Research Communication

Tastin is required for bipolar spindle assembly and centrosome integrity during mitosis Shuo Yang,* Xuan Liu,* Yanqing Yin,* Michiko N. Fukuda,† and Jiawei Zhou*,1 *Laboratory of Molecular Cell Biology, Institute of Biochemistry and Cell Biology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Shanghai, China; and †Cancer Research Center, Burnham Institute, La Jolla, California, USA Tastin was previously characterized as an accessory protein for cell adhesion that participates in early embryo implantation. Here, we report that tastin is also required for spindle assembly during mitosis. Tastin protein levels peaked in the G2/M phase and abruptly declined after cell division. Microscopy showed that tastin is primarily localized on the microtubules, centrosomes, and the mitotic spindle during the cell cycle. Tastin interacted with the dynein intermediate chain, p150Glued, and ␥-tubulin in addition to Tctex-1 (the light chain of dynein). Overexpression of tastin led to monopolar spindle formation, whereas loss of tastin expression caused profound mitotic block and preferentially induced multipolar spindles. These multipolar spindles were generated through a loss of cohesion in mitotic centrosomes; specifically, tastin depletion caused the fragmentation of pericentrosomal material and the splitting of the centrioles at the spindle poles. Tastin depletion induced centrosome abnormalities exclusively during mitosis and required both microtubule integrity and Eg5 activity. However, tastin depletion did not disrupt the organization of spindle poles, as revealed by localization of nuclear mitotic apparatus protein (NuMA) and the p150Glued component of dynactin. These data indicate that the major function of tastin during mitosis is to maintain the structural and dynamic features of centrosomes, thereby contributing to spindle bipolarity.—Yang, S., Liu, X., Yin, Y., Fukuda, M. N., Zhou, J. Tastin is required for bipolar spindle assembly and centrosome integrity during mitosis. FASEB J. 22, 1960 –1972 (2008) ABSTRACT

Key Words: microtubules 䡠 mitotic block

During mitosis, the genetic material is equally distributed to two daughter cells. Errors during mitosis result in the uneven segregation of chromosomes, generating aneuploid or polyploid cells. Such genomic imbalances are among the most common characteristics of cancer and are considered to be crucial for tumorigenesis. Chromosome segregation during mitosis is mediated by a complex microtubule (MT)-based structure, the spindle. The proper assembly and function of the mitotic spindle are therefore essential to ensure genomic stability during cell division (1– 4). The 1960

spindle is assembled in a spatially and temporally regulated manner during the cell cycle, and its assembly and function are intimately associated with MT dynamics (5). In animal cells, the duplication of centrosomes and their subsequent separation are critical initial steps in establishing a bipolar MT array (6). Centrosomes act as the major organizing centers for MT nucleation in somatic cells, and hence their function is intimately connected with the organization of spindle poles. The organization of spindle poles is also governed by the interaction of MTs with accessory proteins that localize primarily at the MT minus ends (centrosomes and the pericentrosomal region in somatic cells). In the past several decades, tremendous efforts have been made to identify factors that function in mitotic spindle formation. An array of candidates coordinating early mitotic events has been identified, including centrosomal factors such as pericentrin (7), noncentrosomal proteins such as NuMA (8), kinesin-like proteins, dynein/dynactin (9, 10), and the small GTPase Ran. However, timely assembly and disassembly of the spindle are an extremely complex processes involving numerous molecules. The identification of new molecules involved in this process and elucidation of their activities are crucial for understanding spindle formation. Tastin, first described by Fukuda and colleagues (11–13), was identified as a protein that forms a complex with trophinin and bystin. This complex mediates early embryo implantation, which is accompanied by rapid cellular invasion and proliferation. Human tastin is a cytoplasmic protein composed of 778 amino acid residues, and it is rich in proline residues (16% of the total amino acids). The tastin N terminus is highly basic, whereas the C terminus is acidic (14). Motif analysis based on the eukaryotic linear motif (ELM) database shows that tastin contains one tyrosine and several serines and threonines, many of which are in contexts favorable to phosphorylation by multiple protein kinases, including protein kinase C, CDK1, pololike kinase, and mitogen-activated protein kinase. Tas1

Correspondence: Institute of Neuroscience, Chinese Academy of Sciences, 320 Yueyang Rd., Shanghai, 200031, China. E-mail: [email protected] doi: 10.1096/fj.07-081463 0892-6638/08/0022-1960 © FASEB

tin also contains five putative cyclin recognition sites that could target the mitotic kinase to appropriate substrates. Tastin mRNA levels are high in testis, bone marrow, and thymus and absent in most adult human tissues (14). Interestingly, tastin is also highly expressed in human cancer cell lines such as HeLa and Jurkat cells (Genomics Institute of the Novaris Research Foundation [GNF] database). Considering that tastin is expressed in multiple tissues and cells unrelated to embryo implantation, it is possible that tastin has additional functions. Human tastin has also been identified as a MT-associated protein (MAP) in mammalian cells (14). Moreover, it binds to Tctex-1, a light chain of dynein, and colocalizes with interphase MTs; during mitosis, it overlaps with the mitotic spindle, particularly in the spindle pole region (14). These data strongly suggest that tastin could be involved in spindle assembly and mitotic integrity. However, the exact role of this protein in mitosis has not been previously studied, and little is known about its underlying function in spindle assembly. Here, we report the subcellular localization of tastin and its cell cycle-dependent expression. Tastin localized to MTs, to centrosomes during interphase, and to the spindle during mitosis. Tastin interacted with the dynein/dynactin complex by binding to Tctex-1, a light chain of dynein, and also with ␥-tubulin. Overexpression of tastin led to monopolar spindles, whereas depletion of tastin caused mitotic delay and multipolar spindle formation. Loss of tastin also resulted in the fragmentation of pericentrosomal materials and centriole splitting at spindle poles. This process was specifically induced in an MT- and Eg5-dependent manner. Thus, we have identified tastin as a novel regulator of bipolar spindle assembly and centrosome integrity during mitosis.

MATERIALS AND METHODS Plasmid construction Myc- and enhanced green fluorescent protein EGFP-tagged forms of tastin (pTastin-myc and pEGFP-Tastin) were constructed by ligating tastin cDNA (from M.N.F.) into the pcDNA3.1 (Invitrogen, Carlsbad, CA, USA) and pEGFP-C1 (Clontech, Palo Alto, CA, USA) vectors, respectively. The primer sequences used in the present study were pcDNA3.1-tastin: forward, 5⬘-CGTGAATTCATCATGACCACCCGGCAAGCCACGAA-3⬘ and reverse 5⬘-ATTGATATCTTGGAGAGCCCTGGGGGGCAGCAGAA-3⬘; and pEGFP-tastin: forward, 5⬘-CGTGAATTCTATGACCACCCGGCAAGCCACGAA-3⬘ and reverse 5⬘- ATTGTCGACTGGAGAGCCCTGGGGGGCAGCAGAA-3⬘. Cell culture, transfection, and synchronization All cell lines used in the present study were grown in Dulbecco’s modified Eagle’s medium (DMEM)/F12 (1:1, Invitrogen) containing 10% fetal bovine serum (FBS), 2 mM glutamine, 100 U/ml penicillin, and 0.1 ␮g/ml streptomycin at 37°C in an atmosphere of 5% CO2. Transient transfection TASTIN IN SPINDLE ORGANIZATION

was carried out with calcium phosphate, FuGENE 6 Transfection Reagent (Roche Diagnostics, Mannheim, Germany), or electroporation (Eppendorf, Germany). HeLa cell synchronization was achieved as described previously (15). Briefly, synchronization at the G1/S boundary was achieved by the thymidine (2.5 mM, Sigma, St. Louis, MO, USA) double block technique; an S-phase sample was collected 3.5 h after release from the second thymidine block. For G2/M synchronization, cells were released from thymidine block for 7 h followed by nocodazole (NOC, 400 ng/ml, Sigma) blocking for 4 h and then were collected by mitotic shake-off; G0 cells were collected 48 h after serum starvation. All cell synchronization was monitored by flow cytometry. For treatment with NOC or monastrol (MA, Sigma), tastin-depleted cells were synchronized in the G1/S phase by thymidine at ⬃55 h after transfection, followed by release from thymidine arrest for 7 h and subsequent exposure to 5 ␮g/ml NOC or 100 ␮M MA for 4 h before harvesting. The cells were then fixed or released in drug-free medium for 1 h. Antibodies Specific monoclonal and polyclonal antibodies (mAbs and pAbs, respectively) against amino acids (aa) 41– 49 of tastin were verified previously (12). We also generated anti-tastin pAbs by immunizing rabbits with a polypeptide (aa 584 – 603) of human tastin and affinity purified the pAbs. The anti-tastin pAb against aa 584 – 603 was also affinity purified. Antibodies to the following proteins were also used: anti-myc (clone 9E10, Berkeley Antibody, Richmond, CA, USA); anti-EGFP mAb (Santa Cruz Biotechnology, Santa Cruz, CA, USA); anti-␤-actin mAb (clone AC-15, Sigma); anti-p150Glued mAb and anti-p50 mAb (BD Biosciences, San Jose, CA, USA); anti-␣-tubulin mAb (clone DM1A), anti-␣-tubulin pAb, anti␥-tubulin mAb (clone GTU-88, Sigma), and anti-␥-tubulin pAb (Sigma); anti-NuMA and hTPX2 pAbs (gifts from Dr. Duane A. Compton, Dartmouth Medical School, Hanover, NH, USA); anti-centrin 2 pAb (a gift from Dr. Michel Bornens, CNRS-Institute Curie, Paris, France); anti-Tctex-1 pAb (a gift from Dr. Stephen M. King, University of Connecticut Health Center, Farmington, CT, USA); and rootletin pAb (a gift from Dr. Jun Yang, Harvard Medical School, Boston, MA, USA). Peroxidase-conjugated secondary antibodies to mouse IgG or rabbit IgG and tetramethylrhodamine isothiocyanate (TRITC) -conjugated goat anti-mouse IgG and fluorescein isothiocyanate (FITC) -conjugated goat anti-mouse IgG were purchased from Jackson Immunoresearch (West Grove, PA, USA); Cy5-conjugated goat anti-rabbit IgG and Texas Red-conjugated goat anti-rabbit IgG were purchased from Sigma. Immunoblotting and coimmunoprecipitation HeLa cells were lysed with ice-cold TNE buffer (50 mM Tris/HCl, pH 7.5, containing 150 mM NaCl, 1 mM EGTA, and 1% Nonidet P40) plus protease inhibitors for 20 min at 4°C. Immunoblots probed with peroxidase-linked secondary antibodies were visualized by SuperSignal West Pico Chemiluminescent substrate (Pierce, Rockford, IL, USA). In some cases, bands were digitized and optical densities were analyzed using ImageMaster 2D Platinum (v.5.0, Amersham Biosciences Piscataway, NJ, USA). For coimmunoprecipitation experiments, transfected HeLa cells were synchronized in the G2/M phase and lysed. The supernatant was pretreated with control IgG and protein A/G-Sepharose (Santa Cruz Biotechnology) and then centrifuged. The supernatant was incubated with appropriate antibodies followed by protein A/G gel. After four washes with PBS buffer, the protein 1961

A/G-bound immune complexes were separated with SDSPAGE and then immunoblotted. Immunofluorescence, confocal microscopy, and image analysis For immunostaining of endogenous tastin, ␣-tubulin, and ␥-tubulin, HeLa cells growing on glass coverslips were permeabilized with 0.1% Triton X-100 in PEM buffer (100 mM PIPES pH 6.9, 5 mM EGTA, 1 mM MgCl2, and 30% glycol) for 1 min and then fixed with 2% paraformaldehyde in PEM for 20 min. Transfected HEK293T cells were permeabilized for 30 s followed by fixation in similar manner. HeLa cells transfected with pSUPER RNA interference (RNAi) were fixed in 2% paraformaldehyde for 3 min followed by permeabilization with 0.1% Triton X-100 in PEM for 3 min and then incubated with methanol at ⫺20°C for 15 min. Fixed cells were then blocked in phosphate buffered saline containing Triton X-100 containing 5% goat serum. After washes, cells were first incubated for 1 h with the primary antibodies, followed by incubation with secondary antibodies and DAPI (Sigma). Cells were imaged using either a cooled CCD SPOT II (Diagnostic Instruments, Sterling Heights, MI, USA) on a microscope (BX51; Olympus) or a Leica TCS SP2 laser confocal microscope. Optical sections were in some cases scanned at 0.2 to 0.5 ␮m intervals. Z stack images were then formed by maximal projection. Data were obtained and processed using Adobe Photoshop 7.0 software (Adobe Systems). Anti-tubulin-stained cells were used to determine the mitotic index. At least 300 cells were counted per coverslip in each of four separate experiments. RNAi The pSUPER vector-based construct encoding tastin smallinterfering (si)RNA and EGFP (designated pSUPER-Tas), which allows for tracking of transfected cells, was prepared according to the manufacturer’s protocols (OligoEngine, Seattle, WA, USA). Briefly, a tastin target sequence 2122GCCTGATCTTCTCTTCCCA2140 was selected by GeneChem (Shanghai, China). Two complementary oligonucleotides, 5⬘-GATCCCCGCCTGATCTTCTCTTCCCATTCAAGAGATGGGAAGAGAAGATCAGGCTTTTTA-3⬘ and 5⬘-AGCTTAAAAAGCCTGATCTTCTCTTCCCATCTCTTGAATGGGAAGAGAAGATCAGGCGGG-3⬘, were synthesized. After annealing, the oligonucleotides were cloned into the BglII and HindIII sites of pSUPER to generate the siRNA duplex. An empty vector and a mock construct (target sequence: AGGAGAACGCGAGGTTGTC) that showed no effects on tastin expression were used as controls. Live cell imaging HeLa cells were cotransfected with pHistone 2B-red fluorescent protein (pH2B-RFP, a gift from Xueliang Zhu, SIBCB, Shanghai, China) and pSUPER-Tas at a ratio of 1:3 using electroporation. Seventy-two hours after transfection, the synchronized cells growing on coverslips were mounted in a chamber with DMEM/F12 medium containing 10% FBS at 37°C in an atmosphere containing 5% CO2. At this point, cells released for 10 h from a double thymidine block were imaged at 37°C for up to 400 min with exposure intervals of 2–3 min using an ⫻63 objective. Cells were not exposed to light between exposure intervals. Images were captured using the Leica AS MDW system with a CoolSNAP HQ Monochrome camera (Roper Scientific, Duluth, GA, USA) and processed using the U.S. National Institutes of Health ImageJ program and QuickTime v5.02. Normal cell divisions were observed repeatedly at the end stages of the monitoring 1962

Vol. 22

June 2008

period in control cells, indicating that the growth and recording conditions were suitable for time-lapse microscopy, and normal mitosis of neighboring untransfected cells was seen during experiments (see Supplemental Video 4). Fluorescence-activated cell sorter (FACS) analysis HEK293T and HeLa cells were fixed and stained with propidium iodide and analyzed for DNA content on a FACSca flow cytometer using CellQuest software (BD Biosciences). Statistical analysis Statistical analysis was carried out using statistical software (GraphPad Prism v4.0, GraphPad Software, San Diego, CA, USA). The control group was compared with the different treatment groups using one- or two-way ANOVA followed by the Dunnet test. Differences were considered significant when P values were ⬍0.05.

RESULTS Characterization of tastin expression during the cell cycle A full-length human tastin cDNA containing a single open reading frame of 2334 bp was previously shown to encode a protein with a calculated molecular mass of 83 kDa (14). Here, we analyzed tastin expression in HEK293T cells with a Western blot analysis using a pAb against human tastin aa 41– 49 and observed a band corresponding to ⬃110 kDa (Fig. 1A). When we overexpressed full-length myc-tagged versions of tastin in HEK293T cells, we also obtained a ⬃110 kDa band in Western blots (Fig. 1A). For further analysis, we generated pAbs against a second tastin polypeptide (aa 584 – 603). These pAbs also recognized both the 110 kDa tastin and the full-length myc-tagged version of the protein (Fig. 1A). Thus, the 110 kDa band likely represents the full-length tastin protein. The appearance of a protein of higher molecular mass than the calculated mass could be due either to conformational differences or to post-translational modifications. Since neither dephosphorylation nor deglycosylation (data not shown) resulted in the appearance of the 83 kDa form on SDS-PAGE gels, we favor the hypothesis that conformational differences account for the difference in size. We found that tastin was widely expressed in tumor cell lines, particularly in those that proliferate rapidly, such as HeLa, Ho8910, Skov3, and Pc3 (Fig. 1B), suggesting that tastin expression may be related to cell proliferation. We next determined whether tastin expression depends on the cell cycle. HeLa cells were synchronized specifically in the G1/S, S, G2, M, and G1 phases, after which we analyzed the expression levels of tastin protein. Western blot analysis showed that tastin protein levels were ⬃3- to 4-fold higher in the G2/M than in the G1/S phase (Fig. 1C). Furthermore, the tastin protein from G2/M lysates appeared to be larger than those

The FASEB Journal

YANG ET AL.

Figure 1. Characterization of tastin expression during the cell cycle. A) Identification of tastin. Lysates of HEK293T cells transiently transfected with empty vector or pcDNA3.1-tastin were analyzed by immunoblotting using pAbs against tastin aa 41–49 or 584–603. Expression and integrity of the exogenous tastin-myc protein were verified by Western blot analysis using an anti-myc antibody. B) Detection of tastin in a variety of cancer cell lines. The blot was also probed for ␣-tubulin expression as a loading control. C) Western blot analysis for tastin expression in synchronized HeLa cells. Cells were synchronized and their tastin protein content was determined at various stages of mitosis. ␤-Actin expression was used as a loading control. Protein bands were digitized.

from other lysates (Fig. 1C). Treatment of the G2/M lysate with ␭ protein phosphatase markedly reduced the mobility shift of the mitotic protein, returning it to the interphase mobility (data not shown). Thus, tastin is not only upregulated during G2/M phase, it is also phosphorylated, indicating that it may play an important role in mitosis. Tastin localizes to interphase MTs, centrosomes, and the mitotic spindle over the course of the cell cycle The localization of exogenous tastin to interphase MTs and spindle poles has been previously demonstrated in COS-1 cells (14). We further characterized the subcellular localization of endogenous tastin during the cell cycle in HeLa cells using immunofluorescence microscopy. The specificity of the tastin mAb was examined by preincubation of the antibody with 20 ␮g of the immunogen, which completely abolished immunostaining (Fig. 2Ab). Endogenous tastin localized primarily to centrosomes and MTs during interphase (Fig. 2Aa, C). To determine whether polymerized MTs are required for the centrosomal localization of tastin, we used NOC to disrupt MT polymerization. The location of the centrosome was revealed using ␥-tubulin staining. After treatment with NOC, tastin remained at the centrosome, suggesting that its localization to the centrosome is MT independent (Fig. 2C). During prophase and prometaphase, tastin was concentrated at the center of MT asters and was associated with MTs growing between the two asters. It was then concentrated at spindle poles and extended along spindle MTs during the metaphase and anaphase. During anaphase and telophase, tastin was redistributed to the spindle midzone (Fig. 2A). Consistent with these observations, in HEK293T cells, overexpressed EGFP-tagged tastin proteins also localized to interphase MTs, centrosomes, the mitotic spindle pole, and midzone MTs over the course TASTIN IN SPINDLE ORGANIZATION

of the cell cycle (Fig. 2B, D). Thus, these findings strongly suggest that tastin colocalizes with tubulin in the cytoplasm of interphase cells and in the spindle during mitosis. Tastin interacts with the dynein complex and centrosomal material Using a yeast two-hybrid assay, others have demonstrated that tastin directly interacts with Tctex-1, one of the light chains of cytoplasmic dynein (14, 16, 17). Tctex-1 is also immunoprecipitated by a mAb raised against the 74 kDa cytoplasmic dynein intermediate chain, and it is redistributed onto mitotic spindles along with other dynein/dynactin subunits (16, 17). To verify whether tastin interacts with the dynein/dynactin complex, we assayed extracts from cells by coimmunoprecipitation using tastin antibodies (against aa 584 – 603). Endogenous tastin coprecipitated with endogenous Tctex-1, the dynein intermediate chain, p150Glued (subunit of dynactin), and p50 in HeLa cells (Fig. 3A). In contrast, tastin antibodies failed to precipitate actin (Fig. 3A). The inability of control IgG to coimmunoprecipitate dynein suggests the high specificity of the tastin-dynein interaction. Moreover, we found that tastin also precipitated pericentrosomal material ␥-tubulin but not centrin 2 (Fig. 3B). Dynein/dynactin and ␥-tubulin are widely redistributed on interphase centrosomes and mitotic spindles. Given that tastin localizes to the spindle during mitosis and to the centrosome during interphase (Fig. 2), it is likely that the interaction of tastin with dynein and ␥-tubulin contributes to the localization of tastin in the spindle and centrosome. Overexpression of tastin causes mitotic block and monopolar spindle formation To investigate the role of tastin in mitosis, we first examined the effect of gain of tastin function. We 1963

Figure 2. Dynamic localization of tastin during the cell cycle. A) Double-labeling immunofluorescence photomicrographs showing subcellular localization of endogenous tastin. HeLa cells were double immunostained for endogenous tastin (green) and ␣-tubulin (red). DNA was visualized with DAPI staining (blue). Interphase localization of tastin is shown (a). Cells in a variety of mitotic phases are shown in the panels at the right. An immunogen block served as a negative control (b). Scale bars ⫽ 10 ␮m. B) Interphase and mitotic localization of EGFP-tastin fusion proteins. HEK293T cells were transiently transfected with EGFP-tastin and labeled with anti-␣-tubulin mAbs. Scale bars ⫽ 5 ␮m. C, D) Localization of tastin at the centrosome was visualized with indirect immunofluorescence (C) and EGFP autofluorescence (D). Centrosomes, stained with anti-␥-tubulin antibody (indicated with arrows), remain after treatment with nocodazole. Scale bars ⫽ 10 ␮m (C); 5 ␮m (D).

transiently transfected HEK293T cells with EGFPtagged tastin and assayed the cell cycle profiles using flow cytometry at 48 h after transfection. EGFP-tastin transfectants markedly accumulated in the G2/M phase (45% of tastin-transfected cells vs. 24% of control cells; Fig. 4A). In HEK293T cells, except for the mitotic block, we observed an increase in the appearance of monopolar spindle at ⬃36 h after transfection (Fig. 4B). Monopolar spindle formation was the main consequence of the overexpression of tastin. Among the mitotic cells, ⬃53% displayed one large MT aster (monopolar spindle) emanating from a single organizing center (Fig. 4B, C). Tastin appeared to be more restricted to the center of the astral MTs, toward the minus end of the MTs (Fig. 4B, bottom). A few mitotic cells containing multipolar spindles were also observed (data not shown), indicating that formation of multinucleated cells resulted from monopolar spindle in the last mitosis. We then immunostained ␥-tubulin and centrin 2 (well-characterized markers of pericentrosomal materials and centrioles, respectively) to observe the effect of tastin overexpression on centrosomes in spindle poles. ␥-Tubulin staining revealed the presence of two foci, and monopolar spindles also frequently contained four centrin 2 foci (Fig. 4D). These results 1964

Vol. 22

June 2008

indicate that the centrosome had clearly duplicated but had failed to separate sufficiently to form a bipolar spindle. Thus, overexpression of tastin inhibited centrosome separation, leading to monopolar spindle formation. These data provide evidence that overexpression of tastin can interfere with mitotic progression in human cells, similar to what is generally observed with any factor that disrupts normal spindle assembly. Depletion of tastin causes mitotic block To further study the role of tastin during mitosis, we next evaluated the effect of loss of tastin function in HeLa cells using a pSUPER-based RNAi plasmid expressing GFP. This marker allows us to verify tastin depletion and to localize another spindle protein in the same cell by double immunostaining. Immunoblot analysis indicated that endogenous tastin levels were reduced by 60% 48 h after transfection with pSUPERTas and by at least 90% at 72 h, whereas the expression of p150Glued and ␣-tubulin was not affected (Fig. 5A). This suggests that the RNAi was specific to tastin and did not affect cell health or mRNA stability. Neither the vector pSUPER nor a construct for mock RNAi affected

The FASEB Journal

YANG ET AL.

Figure 3. Tastin interacts with the dynein complex and centrosomal material. A) Coimmunoprecipitation of endogenous tastin with the dynein complex. Lysates of HeLa cells were incubated with antibodies to tastin (aa 584 – 603) or control IgG for immunoprecipitations. Immunoprecipitates were then immunoblotted for Tctex-1, dynein intermediate chain (DIC), p150Glued, p50, and actin (negative control). The cell extracts were used as a positive control (left lane). B) Coimmunoprecipitation of tastin with centrosomal material. HeLa cells were lysed and immunoprecipitated with pAb tastin (aa 584 – 603) or control IgG as indicated. Immunoprecipitates were then immunoblotted for ␥-tubulin and centrin 2.

tastin expression in HeLa cells (Fig. 5A). In a FACS analysis of DNA content 48 h after transfection, tastin siRNA treatment had no significant effect on the DNA content compared with controls. In contrast, at 72 h after transfection tastin depletion resulted in the accumulation of a significant population of cells in the G2/M phase (50.3% of RNAi-treated cells vs. 20.6% of controls; Fig. 5B). Consistent with this, closer examinations of the same samples using microscopy 72 h after transfection indicated that depletion of tastin resulted in a mitotic block, with more prometaphase-like cells that had clearly undergone chromosome condensation and nuclear envelope breakdown (NEB) but not metaphase chromosome alignment (⬃35.33%; Fig. 5C, D) than untransfected (4.88%) and pSUPER-Tas mocktransfected (5.67%) cells. Metaphase and anaphase/ telophase configurations were extremely rare in tastindepleted cells (Fig. 5D). To better understand how depletion of tastin affects cell proliferation, we performed time-lapse imaging on synchronized live HeLa cells that were cotransfected with tastin RNAi and pH2B-RFP (a marker of chromosome dynamics) at a 3:1 ratio for 72 h. At this point, TASTIN IN SPINDLE ORGANIZATION

cells were ready to enter G2/M phase with reduced tastin levels. Eleven live cells cotransfected with pH2BRFP and empty vector served as controls. In the control cells, after NEB, all chromosomes moved quickly toward the metaphase plate to achieve full alignment in 20 ⫾ 2 min. They all started anaphase within 50 ⫾ 3 min and completed mitosis within 95 ⫾ 5 min (Fig. 5E; see Supplemental Video 1). By contrast, in tastindepleted cells, 12 of 13 cells divided abnormally. These abnormal cells, which contained various lagging chromosomes, did not align their chromosomes on the metaphase plate as normally occurs, resulting in a profound delay in mitosis (see Supplemental Videos 2– 4). Moreover, in 6 of the 12 cells, mitosis was eventually aborted, leading to cell death within 320 min accompanied by multiple membrane protrusions after chromosome deformation (Fig. 5E; see Supplemental Video 2), suggesting apoptosis (18). The remaining 4 of 12 cells did not divide even after 350 min (compared with 95 min in control cells; see Supplemental Video 3). Although the remaining 2 of the 12 abnormal cells exhibited chromosome alignment and eventually completed mitosis after 180 min, they produced multiple daughter cells after division (see Supplemental Video 4). In addition, normal anaphase onset was not seen in any of the 12 cells, and all of the cells exhibited defective chromosome segregation. Thus, these data further indicate that tastin is necessary to guarantee successful mitosis. Tastin depletion results in multipolar spindle formation Since tastin overlaps with spindle MTs and depletion of tastin expression causes mitotic block, it is likely that tastin plays an important role in spindle assembly. To test this hypothesis, we examined the organization of the mitotic spindle and chromosome in tastin-depleted HeLa cells at ⬃65 h after transfection, a time sufficient for a cell to complete one cell cycle with a suppressed level of tastin. In tastin-suppressed cells, the spindle MTs were of normal length and density, but the spindle morphology was often severely disrupted compared with control cells (⬃73.9% vs. 6.5%, Fig. 6A, B). As shown in Fig. 6A, B, the most frequently observed abnormality after siRNA treatment was the formation of multipolar spindles. Up to 70% of abnormal mitotic cells (Fig. 6B) had asymmetrical multipolar spindle architecture. These cells showed multipoles, ranging from tripolar and tetrapolar to octapolar and occasionally even nonapolar, with misaligned chromosomes scattered around MT arrays. In these tastin-depleted spindles, one aster was often observed to form two half-spindles. Each pole in these multipolar spindles had several distorted MT bundles that connected with at least one other pole, but these MTs appeared to focus in spindle poles. Moreover, fewer abnormal bipolar cells with distorted MT arrays were observed in tastin RNAi cells (Fig. 6Am– o). In tastin-depleted cells, condensed chromosomes were associated with spindle MTs. However, chromosomes often failed to align at a 1965

Figure 4. Overexpression of tastin results in monopolar spindle formation. A) Cell cycle profiles of tastin-overexpressing cells. The DNA content of HEK293T cells was analyzed by flow cytometry 48 h after transfection. B–D) Overexpression of tastin results in monopolar spindle formation. B) HEK293T cells were transiently transfected with pEGFP-tastin and stained with DAPI and ␣-tubulin at 36 h post-transfection. Arrows indicate EGFP-tastin-expressing mitotic cells, whereas arrowheads indicate interphase cells. C) Quantitation of monopolar mitotic cells. At least 350 mitotic cells in 3 independent experiments were examined by microscopy. GFP-transfected cells were used as controls. D) Cells were labeled with ␥-tubulin and centrin 2 (centrioles). Insets show a higher magnification of centrin 2 foci. Scale bars ⫽10 ␮m (all panels except insets).

single metaphase plate, and lagging chromosomes not positioned between any two poles were observed in both bipolar and multipolar spindles. Consistent with this, migration of chromosomes to the metaphase plate

during congression appeared to be blocked in most of the cells analyzed by time-lapse microscopy. Multipolar spindles were observed 65 h after tastin siRNA transfection. Because cells should have divided

Figure 5. Depletion of tastin causes mitotic block. A) efficiency of RNAi. HeLa cells transfected with pSUPER (vector), pSUPER-Tas mock, or pSUPER-Tas (RNAi) for 48 or 72 h were lysed and subjected to immunoblot analysis to visualize the indicated proteins. ␣-Tubulin served as loading control. B) Cell cycle profiles of tastin-depleted cells. The DNA content was analyzed by flow cytometry at 48 and 72 h post-transfection. C) Tastin RNAi led to the accumulation of mitotic cells in a prometaphase-like state, with unaligned chromosomes. At ⬃72 h post-transfection, HeLa cells were fixed for staining. GFP fluorescence was expressed by pSUPER. DNA was visualized using DAPI staining. D) Graph showing the cell cycle phases of control untransfected cells (open bars), pSUPER-Tas mock cells (shaded bars), and pSUPER-Tas cells (solid bars) at 72 h post-transfection. Cell cycle phases were identified as interphase, prophase/prometaphase, metaphase, anaphase/ telophase, and dead based on chromosome configurations. All values are means from 3 independent experiments. A total of 480 –720 cells were scored. E) Mitotic progression of control vs. tastin-suppressed cells. HeLa cells cotransfected with pSUPER-tastin and pH2B-RFP were randomly selected for time-lapse microscopy in prophase or early prometaphase. GFP expressed by pSUPER was used as a transfection marker. Top rows: representative images showing the normal M-phase progression of a control cell (top panels: RFP; leftmost bottom panel: GFP; other bottom panels: phase contrast). Bottom rows: representative images for a typical tastin-depleted cell. Scale bars ⫽ 10 ␮m. 1966

Vol. 22

June 2008

The FASEB Journal

YANG ET AL.

Figure 6. Tastin depletion results in multipolar spindle formation, and Eg5 activity is required for the multipolar spindle formation induced by tastin RNAi. A) Immunofluorescence micrographs showing spindle morphology in tastin-depleted cells. The cells were fixed and stained with ␣-tubulin ⬃65 h after tastin RNAi. DNA was visualized with DAPI. Insets show GFP fluorescence expressed by the pSUPER vector. Tastin depletion mainly results in multipolar spindle formation (d–l). B) Quantitative data of the abnormal mitotic cells. Values are means ⫾ se; *P ⬍ 0.01 (t test, n⫽3) vs. control. At least 480 mitotic cells in 3 independent experiments were scored. Among the abnormal mitotic cells, multipolar spindles were quantified (inset). C, D) Eg5 activity is required for multipolar spindle formation in tastin-depleted cells. C) Immunofluorescence micrographs of MA-treated cells. Tastin-depleted cells were synchronized in the G2 phase by thymidine block and release as described in the Materials and Methods. Cells were then treated for an additional 4 h with 100 ␮M MA and stained for ␣-tubulin and DNA. Inset shows GFP fluorescence expressed by the pSUPER vector. D) Percentages of mitotic cells with monopolar (solid bars), bipolar (open bars), or multipolar (shaded bars) spindles were quantified in tastin RNAi and control RNAi cells either just before MA washout or 1 h after washout (2 independent experiments). Scale bars ⫽ 10 ␮m.

many times after transfection, the observed multipolar spindles could have resulted from previous failed divisions during the 65-h incubation time. It is unlikely that this was the case because an examination of cells fixed 48 h after transfection showed that only ⬃13% of mitotic cells assembled multipolar spindles, and multinucleated cells were rare (our unpublished results). At this point, most cells also had ⬃40% of the endogenous tastin expression of controls (Fig. 5A). We did not observe a large percentage of cells with multipolar spindles until 60 h after transfection. Taken together, these results indicate that the multipolar spindles assembled in tastindepleted cells did not arise from previous failed divisions. TASTIN IN SPINDLE ORGANIZATION

Eg5 is required for multipolar spindle formation in tastin-depleted cells The kinesin Eg5 is crucial for the assembly of bipolar spindles in vertebrate cells (19). To examine whether Eg5 is also required for multipolar spindle formation in mitotic cells with suppressed tastin, MA, a cell-permeable small-molecule inhibitor of Eg5 (20), was used to inhibit Eg5 activity. Sixty-five hours after treatment with tastin siRNA, cells synchronized in the G2 phase were either cultured for additional 4 h and stained to verify the appearance of multipolar spindles in arrested mitotic cells or treated with 100 ␮M MA for an additional 1967

4 h and then stained for MTs and DNA. In the absence of tastin, MA-treated cells had mitotic indices similar to those of untreated cells. However, unlike cells lacking tastin alone, ⬃90% of tastin-depleted cells treated with MA contained monopolar spindles (Fig. 6C, D). The inhibitory effect of MA is reversible, so to examine whether the elimination of Eg5 inhibition would lead to the restoration of multipolar spindles from monopolar arrays in tastin-siRNA-treated cells, cells were switched to MA-free cell culture medium after the incubation in MA. One hour after MA washout, ⬃52% of mitotic cells treated with tastin-siRNA contained multipolar spindles (Fig. 6D). Taken together, these data suggest that in the absence of tastin Eg5 activity induces multipolar spindle formation. Tastin suppression results in MT-dependent fragmentation of pericentrosomal material during mitosis We showed that depletion of tastin in HeLa cells induces the formation of multipolar spindles (Fig. 6). Such supernumerary poles might originate through mechanisms involving the disruption of the centrosomal structure, abnormal centrosome duplication, or centrosome missegregation to daughter cells at previous rounds of abnormal cytokinesis (21, 22). It was unlikely that the defects resulted from a previous failed division because abnormal cytokinesis and multinucleated cells were rare at 48 h after tastin siRNA treatment. As a first step to identify the processes targeted by tastin depletion, we analyzed the pattern of pericentrosomal material in tastin-depleted cells. About 60 h after treatment with tastin siRNA, the cells synchronized in the G2/M phases were then processed for double immunofluorescent staining to visualize ␣-tubulin and ␥-tubulin, the major pericentrosomal matrix proteins required for MT nucleation. More than half of tastindepleted mitotic cells (⬃56%) showed multipolar spindles with ␥-tubulin foci at each supernumerary pole (Fig. 7). In contrast, control cells often had two ␥-tubulin foci at a bipolar spindle (Fig. 7A). Except for a great number of ␥-tubulin foci, the average size of each ␥-tubulin foci in tastin-depleted cells was also uneven. However, ␥-tubulin foci in interphase tastin-depleted cells were of normal number and structure (Supplemental Fig. S1Ai), suggesting that tastin depletion results in pericentrosomal fragmentation during mitosis. MT disruption by NOC inhibits the separation of parental centrosomes in mitosis (23). To test whether the pericentrosomal fragmentation resulting in localization of centrosomal components to superabundant poles was dependent on MTs in the tastin-depleted cells, ␥-tubulin localization was examined in tastindepleted cells that entered mitosis in the absence of polymerized MTs. About 60 h after tastin RNAi, cells synchronized in the G2 phase were treated for an additional 4 h with 5 ␮g/ml NOC and then stained for MTs, ␥-tubulin, and DNA. Immunostaining revealed 1968

Vol. 22

June 2008

that in the presence of NOC, only 5.8% of mitotic cells treated with tastin siRNA contained more than two ␥-tubulin foci, and 94.2% of mitotic cells were observed with two distinct ␥-tubulin foci (n⫽300, 3 independent experiments; Fig. 7A, B). Thus, the induction of multipolar spindles during mitosis in tastin-depleted cells involves the fragmentation of pericentrosomal material in an MT-dependent manner. Tastin depletion disrupts cohesion of sister centrioles in mitotic diplosomes but not in interphase Normal cells undergo only one round of centrosome duplication, during which each of the two centrioles composing the centrosome duplicate in a semiconservative manner. Loss of spindle bipolarity is often related to abnormal centrosome duplication (24). We therefore examined the effect of tastin depletion on centrosome duplication using immunostaining for centrin 2 (a centriole marker). Centriole abnormalities were quantified as described previously (25): cells with two pairs of dots (corresponding to 2 centrosomes) were taken to be normal, whereas cells with more than two pairs of dots or with scattered dots were assumed to reflect overduplication and abnormal splitting of centrioles, respectively, and were therefore considered to be abnormal. After thymidine synchronization and release, mitotic cells were fixed and stained at ⬃65 h after tastin siRNA treatment. In control mitoses, chromosomes were correctly aligned, and centrioles were arranged in typical diplosomes at each pole (Fig. 7Ca– d, see magnification in b). Tastin depletion did not significantly affect centrosome duplication, but specifically induced sister centrioles from single diplosomes to move apart from one another (Fig. 7Ce– h, see magnification in f). Up to 50% of tastin-depleted mitotic cells showed split diplosomes, compared with ⬃10% in control cells (Fig. 7D). In the absence of polymerized MTs, the effects of tastin depletion in the G2/M cultures were diminished, and the frequency of mitoses with split centrioles was reduced in tastin siRNA– treated cells (Fig. 7Ci–l, D), indicating that MT integrity is required for the induction of diplosome splitting. We next analyzed the association of pericentrosomal materials and centrioles in tastin-depleted cells by ␥-tubulin and anti-centrin 2 immunostaining. In most abnormal mitoses, all analyzed combinations of markers showed multiple foci of anti-␥-tubulin and four foci derived from a single split centriole (Fig. 7E), which further indicates that induction of multipolar spindles in tastindepleted mitoses involves the fragmentation of pericentrosomal material and centriole splitting rather than an alteration in centrosome number. We also examined the distribution of centrosomal protein rootletin after tastin depletion. It has been demonstrated that rootletin and cNap1 serves as a physical linker between the pair of basal bodies/centrioles in cells thereby contribute to centriole cohesion (26). Our results showed that rootletin was localized in

The FASEB Journal

YANG ET AL.

Figure 7. Centrosomal abnormalities are induced in tastin-depleted cells during mitosis. A, B) Tastin RNAi resulted in MT-dependent fragmentation of pericentrosomal material during mitosis. A) Immunofluorescence micrographs of pericentrosomal material in tastin-depleted cells. Tastin-depleted cells synchronized in the G2/M phase were either fixed and stained or treated for an additional 4 h with 5 ␮g/ml nocodazole and then stained with a mAb for ␣-tubulin (red) and a pAb for ␥-tubulin (pericentrosomal material marker; blue). DNA was visualized with DAPI. Insets show the GFP fluorescence expressed by the pSUPER vector. B) Quantitation of the number of cells with 2 ␥-tubulin foci or more than 2 ␥-tubulin foci in cells treated with control RNAi, tastin RNAi, or both nocodazole and tastin RNAi. Values are means ⫾ se; *P ⬍ 0.01 vs. control (ANOVA with Dunnet test). At least 300 mitotic cells were scored in 3 independent experiments. C, D) Tastin RNAi induces centriole splitting during mitosis in an MT-dependent manner. C) Immunofluorescence micrographs of centrioles in tastin-depleted cells with or without nocodazole treatment. Tastin-depleted cells were stained with a mAb for ␣-tubulin and a pAb for centrin 2 (centriole marker, blue). Insets in a, e, i show the GFP fluorescence expressed by pSUPER vector. Normal centriole pairs in each centrosome are shown in the magnified insets in b, j. A single split centriole is visible in the magnified inset in f. D) Quantitation of centriole abnormalities (overduplication and splitting) in cells treated with control RNAi, tastin RNAi, or both nocodazole and tastin RNAi. Values are means ⫾ se; *P ⬍ 0.01, compared with control. At least 300 mitotic cells were scored in 3 independent experiments. E) Immunolocalization of ␥-tubulin and centrin 2 in a tastin-depleted cell. The magnified insets show an overlay of ␥-tubulin (red) and centrin 2 (blue). Multiple foci of pericentrosomal material (␥-tubulin) contained 4 foci from a single split centriole (centrin 2). Inset shows GFP fluorescence expressed by the pSUPER vector. F) Centrosomal abnormalities induced by tastin RNAi require Eg5 activity. MA-arrested mitoses from tastin-depleted cells were stained with an antibody to ␥-tubulin (a, b) or an antibody to centrin 2 (c, d). No fragmentation of pericentrosomal material or centriole splitting are observed. Insets at lower left show GFP fluorescence expressed by the pSUPER vector. Scale bars ⫽ 10 ␮m.

centrosome in interphase of control cells (see Supplemental Fig. S1Ab, j) as evidenced that it was colocalized with ␥-tubulin (see Supplemental Fig. S1Al). Conversely, rootletin immunoreactivity became undetectable in mitotic phase (Supplemental Fig. S1Af, n). After tastin siRNA treatment, the centrosomal localization of rootletin in interphase remained unaffected (see Supplemental Fig. S1Ab, j). However, rootletin was not expressed in spindle poles and anywhere else at all in the cell (see Supplemental Fig. S1Af, n), despite of multipolar spindles formation in tastin siRNA-treated cells (see Supplemental Fig. S1Ae, m). Thus these results suggest that tastin depletion does not increase the frequency of abnormal centriole numbers or splitting in interphase cells and centrosomal protein rootletin is not involved in tastin-mediated centriole cohesion. TASTIN IN SPINDLE ORGANIZATION

Centrosomal abnormalities caused by tastin depletion require Eg5 activity Kinesin Eg5 is required for multipolar spindle formation in tastin-depleted cells (Fig. 6). Given that Eg5 controls the establishment of spindle bipolarity by causing parental centrosome separation at the onset of mitosis (27), we examined whether Eg5 activity influences the tastin depletion–induced centrosomal abnormalities within spindle poles. Tastin-depleted cell cultures synchronized in the G2/M phase were incubated with MA for 4 h. Cells were then fixed and stained for ␥-tubulin. Mitoses in MA-treated cells with unseparated centrosomes had normal ␥-tubulin numbers and structure in tastin-depleted cells (Fig. 7Fa, b). We next inspected centriole structure more closely using an anti1969

centrin 2 antibody. Mitoses in MA-treated tastin-depleted cells contained two sets of paired centrioles at the center of the spindle pole, which was defined as “normal” (Fig. 7Fc, d). Thus, Eg5 function is also required for tastin suppression to induce centrosomal abnormalities. Tastin depletion did not disrupt the localization of spindle pole components Although a large population of tastin-depleted mitoses had an asymmetrical multipolar spindle architecture, these spindle MTs still appeared to focus normally into spindle poles. To further characterize the structure of the spindle poles, we examined the distribution of NuMA and p150Glued in the spindle pole region in mitotic HeLa cells treated with tastin siRNA. In control cells, MTs organized into astral arrays, with NuMA and p150Glued localized to a crescent-shaped structure at the spindle poles (Fig. 8). After the depletion of tastin, NuMA and p150Glued remained in the spindle poles and their intensity was also not changed compared with controls, although present in multiple foci instead of just two foci (Fig. 8). Moreover, the expression of hTPX2 in spindles of tastin-depleted cells was examined. It was revealed that the expression patterns of hTPX2 in spindle poles were not changed compared with control, although present in multiple polar (Supplemental Fig. S1B). Unlike cells lacking TOGp (28), the diameter of the single aster in tastin-depleted cells was not significantly different than in control cells (data not shown). Given that the dynein/dynactin/NuMA complex is required for the organization of spindle poles (29) and both NuMA and p150 colocalized with MT minus ends in tastin-depleted cells (Fig. 8), we conclude that tastin is not required to focus free MT minus ends at the spindle poles during mitosis.

DISCUSSION The data presented here demonstrate that tastin plays an essential role in bipolar spindle assembly and cen-

Figure 8. Tastin depletion does not disrupt the localization of spindle pole components. Immunofluorescence micrographs showing the localization of the spindle pole components NuMA and p150Glued in tastin-depleted cells. The cells were fixed and stained for NuMA and p150Glued at ⬃65 h after tastin RNAi. DNA was visualized with DAPI. Insets show the GFP fluorescence expressed by the pSUPER vector. Scale bar ⫽ 10 ␮m. 1970

Vol. 22

June 2008

trosome integrity in cultured vertebrate cells. We have shown that tastin localizes to centrosomes, MTs, and the mitotic spindle and is most highly expressed in the G2/M phase. Overexpression of tastin leads to monopolar spindles, whereas tastin depletion causes the fragmentation of pericentrosomal materials and the splitting of centrioles at spindle poles, which lead to multipolar spindle formation. Tastin is a cycling protein essential for cell cycle progression In the present study, we showed that the mRNA and protein levels of tastin were regulated over the cell cycle, with highest level seen in the G2/M and an abrupt decline after cell division, similar to the expression pattern of other mitotic molecules, including RHAMM, TPX2, and Nudel (30, 31). Although the increased protein expression in the G2/M correlated well with transcription levels, the rapid disappearance of the protein in the G1 suggests rapid degradation after cytokinesis. Tastin exhibits potential sumoylation sites (aa 261–264, predicted in the EML database), suggesting that it may be targeted to the ubiquitin/ proteasome pathway (32). Rapid protein degradation is required to prevent excess tastin from disrupting mitotic progression and to guarantee proper exit from mitosis. Indeed, as shown in Fig. 4, the overexpression of exogenous tastin ([␴␫␮]8-fold excess over endogenous tastin, data not shown) induced cell cycle abnormalities. Consistent with this, we also found that overexpression of tastin disrupted normal spindle structure and caused monopolar spindle formation. Thus, failure to remove excess exogenous tastin, which leads to abnormal mitotic progression, may be one of the main causes of the cell cycle abnormalities induced by tastin overexpression. However, depletion of tastin also increased the mitotic index and caused mitotic block, indicating a role for tastin in guaranteeing normal mitotic progression. Moreover, our immunohistochemistry and time-lapse microscopy analysis revealed that many of the mitotic tastin-depleted cells had achieved centrosome duplication and separation and chromosome condensation, but not metaphase chromosome alignment (Figs. 5 and 6); instead, they were in a state comparable to prometaphase (33). Taken together, these observations indicate that tastin is a cycling protein essential for cell cycle progression, and the levels of endogenous tastin must be tightly controlled during mitosis. Our current data also demonstrated that tastin is phosphorylated in mitosis, similar to Protein4.1 and NuMA (15, 34). A motif analysis in the ELM database shows that tastin contains many potential sites for serine/threonine phosphorylation by protein kinases, including CDK1, polo-like kinase, and mitogen-activated protein kinase, which are highly activated during mitosis (35). However, the exact phosphorylation sites of tastin and its significance in mitosis are still unclear and will need to be addressed in future studies.

The FASEB Journal

YANG ET AL.

A role for tastin in spindle assembly and centrosome integrity during mitosis The role of tastin in spindle assembly was addressed by the overexpression and depletion of tastin in the present study. The most striking phenotype observed after tastin overexpression is the appearance of monopolar spindles with grouped centrosomes, indicating that an appropriate level of the protein is required for bipolar spindle assembly. Defects in the separation or maturation of centrosomes can lead to the formation of mitotic monoasters of MTs. The ability of tastin to localize at centrosomes in the presence of NOC and to interact with other proteins localized in centrosome, such as dynein/dynactin and ␥-tubulin (Figs. 2 and 3), together suggest that tastin overexpression may act as a dominant factor to perturb the normal function of centrosomes in mitosis. Depletion of tastin primarily resulted in the formation of multipolar spindles, further indicating that tastin is necessary for the bipolar organization of the spindle. Induction of multipolar spindles by tastin depletion occurred through the fragmentation of pericentrosomal materials and the aberrant splitting of centrioles within diplosomes in mitosis (Figs. 6 and 7). Thus, tastin contributes to centrosome integrity and in this way could regulate spindle assembly. Tastin depletion influenced neither the timing nor the extent of the formation of parental centrosomes in interphase but selectively perturbed the cohesion and dynamics of centrosomes in mitosis. This is a novel finding and begins to identify aspects of centrosome organization and function that are influenced by tastin in mitosis. How does tastin regulate centrosome integrity? The cohesion and dynamics of centrosomes are regulated by a network of specific factors (35). Tastin depletion may interfere with crucial factors implicated in centrosome organization during mitosis. Such factors might be activated or be capable of establishing crucial interactions at the centrosomal level in a tastin-dependent manner. Interestingly, the loss of several proteins that also localize in spindle poles, including TPX2 (19), TOGp (28), and Nek2B (36), leads to multipolar spindles as a consequence of centrosome fragmentation. In contrast, the overexpression of RanBP1 can lead to the same outcome (25). Multipolar spindles and the fragmentation of centrosomes after the depletion of tastin may also be interpreted as a loss of the cohesive activity of tastin within the spindle poles. Thus, tastin is likely to be a key structural or scaffold protein for maintaining centrosome and spindle integrity as various MT-dependent motors exert force on the centrosome and spindle. In the present study, we showed that tastin interacts with the motor proteins dynein and dynactin and the centrosomal protein ␥-tubulin. It is possible that these molecules exert force on the spindle complex with the aid of tastin and coordinately participate in centrosome and spindle assembly. Thus, loss of tastin could lead to deregulation of their coordination, thereby impairing TASTIN IN SPINDLE ORGANIZATION

the balance between different motor forces and resulting in disorganized spindle morphology and centrosome fragmentation. In support of this, excess tastin impaired spindle organization, leading to monopolar spindles, and multipolar spindle formation depended on Eg5 (motor protein) activity in the absence of tastin, further indicating that tight and precise regulation of the tastin-associated motor machinery is required for structural support of centrosomal integrity and normal spindle assembly. Taken together, these findings suggest that a balance of various spindle proteins is crucial for maintaining centrosome and spindle integrity. Our finding that NOC prevented the disruptive effect of tastin depletion further strengthens the conclusion that centrosome integrity is sensitive to tastin levels during mitosis and, furthermore, implicates MTs in the loss of centrosome integrity induced by tastin depletion. Tastin, a MAP, localizes to the mitotic spindle, and tastin depletion might alter MT dynamics at the spindle poles, thereby promoting the aberrant separation of centrosomes. MTs themselves or motor proteins might thus play a role in cohesion and dynamics within centrosomes in a manner that is sensitive to tastin levels. Tastin and cancer We demonstrated here that overexpression of tastin disrupted normal spindle structure and caused monopolar spindle formation. These findings predispose mitotic cells to undergo chromosome missegregation. Common consequences of abnormal spindle assembly include chromosome missegregation to daughter cells, which can lead to genomic imbalance and thereby contribute to oncogenesis (21, 37). Indeed, a recent study using a cDNA microarray analysis has indicated that tastin expression is upregulated in prostate cancer tissues (38), suggesting that tastin is a tumor-associated antigen in prostate cancer and perhaps other cancers. Thus, the cell cycle-dependent expression pattern of tastin could also make it a potential cancer marker with high diagnostic and therapeutic value. We thank Dr. Duane A. Compton (Dartmouth Medical School, Hanover, NH, USA) for the NuMA and hTPX2 pAb, Dr. Stephen M. King (University of Connecticut Health Center, Farmington, CT, USA) for the Tctex-1 pAb, Dr. Jun Yang (Harvard Medical School, Boston, MA, USA) for rootletin pAb, and Dr. Michel Bornens (CNRS-Institute Curie, Paris, France) for the centrin 2 pAb. We thank Drs. Yixian Zheng, Xueliang Zhu, Shiyuan Cheng, Yichang Jia, Fubin Wang, and Chengyong Shen for their kind help and inspiring discussions. This work was supported by grants from the Chinese Academy of Sciences, Shanghai Metropolitan Fund for Research and Development (04BZ14005; 06DZ22032), Natural Science Foundation of China (30525041 and 30623003), State Key Program for Basic Research of China (2006CB500704), and U.S. Department of Defense (W81XWH04 –1-0917 to M.N.F.). 1971

REFERENCES 1. 2. 3. 4. 5. 6.

7. 8. 9. 10. 11.

12.

13.

14.

15.

16.

17.

18. 19.

20.

1972

Mitchison, T. J. (1989) Mitosis: basic concepts. Curr. Opin. Cell Biol. 1, 67–74 Rieder, C. L. (1991) Mitosis: towards a molecular understanding of chromosome behavior. Curr. Opin. Cell Biol. 3, 59 – 66 Gadde, S., and Heald, R. (2004) Mechanisms and molecules of the mitotic spindle. Curr. Biol. 14, R797– 805 Pihan, G. A., and Doxsey, S. J. (1999) The mitotic machinery as a source of genetic instability in Cancer Semin. Cancer Biol. 9, 289 –302 Hyman, A. A., and Karsenti, E. (1996) Morphogenetic properties of microtubules and mitotic spindle assembly. Cell 84, 401– 410 Maxwell, C. A., Keats, J. J., Crainie, M., Sun, X., Yen, T., Shibuya, E., Hendzel, M., Chan, G., and Pilarski, L. M. (2003) RHAMM is a centrosomal protein that interacts with dynein and maintains spindle pole stability. Mol. Biol. Cell 14, 2262–2276 Doxsey, S. J., Stein, P., Evans, L., Calarco, P. D., and Kirschner, M. (1994) Pericentrin, a highly conserved centrosome protein involved in microtubule organization. Cell 76, 639 – 650 Compton, D. A., and Cleveland, D. W. (1993) NuMA is required for the proper completion of mitosis. J. Cell Biol. 120, 947–957 Zimmerman, W., and Doxsey, S. J. (2000) Construction of centrosomes and spindle poles by molecular motor-driven assembly of protein particles. Traffic 1, 927–934 Merdes, A., and Cleveland, D. W. (1997) Pathways of spindle pole formation: different mechanisms; conserved components. J. Cell Biol. 138, 953–956 Suzuki, N., Zara, J., Sato, T., Ong, E., Bakhiet, N., Oshima, R. G., Watson, K. L., and Fukuda, M. N. (1998) A cytoplasmic protein, bystin, interacts with trophinin, tastin, and cytokeratin and may be involved in trophinin-mediated cell adhesion between trophoblast and endometrial epithelial cells. Proc. Natl. Acad. Sci. U. S. A. 95, 5027–5032 Fukuda, M. N., and Nozawa, S. (1999) Trophinin, tastin, and bystin: a complex mediating unique attachment between trophoblastic and endometrial epithelial cells at their respective apical cell membranes. Semin. Reprod. Endocrinol. 17, 229 –234 Suzuki, N., Nakayama, J., Shih, I. M., Aoki, D., Nozawa, S., and Fukuda, M. N. (1999) Expression of trophinin, tastin, and bystin by trophoblast and endometrial cells in human placenta. Biol. Reprod. 60, 621– 627 Nadano, D., Nakayama, J., Matsuzawa, S., Sato, T. A., Matsuda, T., and Fukuda, M. N. (2002) Human tastin, a proline-rich cytoplasmic protein, associates with the microtubular cytoskeleton. Biochem. J. 364, 669 – 677 Huang, S. C., Liu, E. S., Chan, S. H., Munagala, I. D., Cho, H. T., Jagadeeswaran, R., and Benz, E. J., Jr. (2005) Mitotic regulation of protein 4.1R involves phosphorylation by cdc2 kinase. Mol. Biol. Cell 16, 117–127 King, S. M., Dillman, J. F., 3rd, Benashski, S. E., Lye, R. J., Patel-King, R. S., and Pfister, K. K. (1996) The mouse t-complexencoded protein Tctex-1 is a light chain of brain cytoplasmic dynein. J. Biol. Chem. 271, 32281–32287 Tai, A. W., Chuang, J. Z., and Sung, C. H. (1998) Localization of Tctex-1, a cytoplasmic dynein light chain, to the Golgi apparatus and evidence for dynein complex heterogeneity. J. Biol. Chem. 273, 19639 –19649 Mills, J. C., Stone, N. L., and Pittman, R. N. (1999) Extranuclear apoptosis. The role of the cytoplasm in the execution phase. J. Cell Biol. 146, 703–708 Garrett, S., Auer, K., Compton, D. A., and Kapoor, T. M. (2002) hTPX2 is required for normal spindle morphology and centrosome integrity during vertebrate cell division. Curr. Biol. 12, 2055–2059 Kapoor, T. M., Mayer, T. U., Coughlin, M. L., and Mitchison, T. J. (2000) Probing spindle assembly mechanisms with monas-

Vol. 22

June 2008

21. 22.

23. 24. 25.

26. 27.

28.

29.

30.

31.

32. 33.

34. 35. 36. 37. 38.

trol, a small molecule inhibitor of the mitotic kinesin, Eg5. J. Cell Biol. 150, 975–988 Brinkley, B. R. (2001) Managing the centrosome numbers game: from chaos to stability in cancer cell division. Trends Cell Biol. 11, 18 –21 Pihan, G. A., Purohit, A., Wallace, J., Malhotra, R., Liotta, L., and Doxsey, S. J. (2001) Centrosome defects can account for cellular and genetic changes that characterize prostate cancer progression. Cancer Res. 61, 2212–2219 Mayor, T., Stierhof, Y. D., Tanaka, K., Fry, A. M., and Nigg, E. A. (2000) The centrosomal protein C-Nap1 is required for cell cycle-regulated centrosome cohesion. J. Cell Biol. 151, 837– 846 Lingle, W. L., and Salisbury, J. L. (2000) The role of the centrosome in the development of malignant tumors. Curr. Top. Dev. Biol. 49, 313–329 Di Fiore, B., Ciciarello, M. R., Mangiacasale, R., Palena, A., Tassin, A. M., Cundari, E., and Lavia, P. (2003) Mammalian RanBP1 regulates centrosome cohesion during mitosis. J. Cell Sci. 116, 3399 –3411 Yang, J., Adamian, M., and Li, T. (2006) Rootletin interacts with C-Nap1 and may function as a physical linker between the pair of centrioles/basal bodies in cells. Mol. Biol. Cell 17, 1033–1040 Walczak, C. E., Vernos, I., Mitchison, T. J., Karsenti, E., and Heald, R. (1998) A model for the proposed roles of different microtubule-based motor proteins in establishing spindle bipolarity. Curr. Biol. 8, 903–913 Cassimeri, L., and Morabito, J. (2004) TOGp, the human homolog of XMAP215/Dis1, is required for centrosome integrity, spindle pole organization, and bipolar spindle assembly. Mol. Biol. Cell 15, 1580 –1590 Merdes, A., Heald, R., Samejima, K., Earnshaw, W. C., and Cleveland, D. W. (2000) Formation of spindle poles by dynein/ dynactin-dependent transport of NuMA. J. Cell Biol. 149, 851– 862 Yan, X., Li, F., Liang, Y., Shen, Y., Zhao, X., Huang, Q., and Zhu, X. (2003) Human Nudel and NudE as regulators of cytoplasmic dynein in poleward protein transport along the mitotic spindle. Mol. Cell Biol. 23, 1239 –1250 Maxwell, C. A., Keats, J. J., Belch, A. R., Pilarski, L. M., and Reiman, T. (2005) Receptor for hyaluronan-mediated motility correlates with centrosome abnormalities in multiple myeloma and maintains mitotic integrity. Cancer Res. 65, 850 – 860 Dohmen, R. J. (2004) SUMO protein modification. Biochim. Biophys. Acta 1695, 113–131 Echeverri, C. J., Paschal, B. M., Vaughan, K. T., and Vallee, R. B. (1996) Molecular characterization of the 50-kD subunit of dynactin reveals function for the complex in chromosome alignment and spindle organization during mitosis. J. Cell Biol. 132, 617– 633 Compton, D. A., and Luo, C. (1995) Mutation of the predicted p34cdc2 phosphorylation sites in NuMA impair the assembly of the mitotic spindle and block mitosis. J. Cell Sci. 108, 621– 633 Mayor, T., Meraldi, P., Stierhof, Y. D., Nigg, E. A., and Fry, A. M. (1999) Protein kinases in control of the centrosome cycle. FEBS Lett. 452, 92–95 Uto, K., and Sagata, N. (2000) Nek2B, a novel maternal form of Nek2 kinase, is essential for the assembly or maintenance of centrosomes in early Xenopus embryos. EMBO J. 19, 1816 –1826 Doxsey, S. J. (2001) Centrosomes as command centres for cellular control. Nat. Cell Biol. 3, E105–108 Dhanasekaran, S. M., Barrette, T. R., Ghosh, D., Shah, R., Varambally, S., Kurachi, K., Pienta, K. J., Rubin, M. A., and Chinnaiyan, A. M. (2001) Delineation of prognostic biomarkers in prostate Cancer Nature 412, 822– 826

The FASEB Journal

Received for publication July 5, 2007. Accepted for publication December 13, 2007.

YANG ET AL.