View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

FEBS Letters 580 (2006) 3375–3380

An essential role of Cdc42-like in mitosis of HeLa cells

Shingo Yasudaa,b, Hiroyuki Taniguchia,b, Fabian Oceguera-Yaneza, Yoshikazu Andoa, Sadanori Watanabea, James Monypennya, Shuh Narumiyaa,* a Horizontal Medical Research Organization, Kyoto University, Faculty of Medicine, Kyoto 606-8501, Japan b Department of Pharmacology, Kyoto University, Faculty of Medicine, Kyoto 606-8501, Japan

Received 9 February 2006; revised 2 May 2006; accepted 2 May 2006

Available online 11 May 2006

Edited by Ulrike Kutay

apparent discrepancy suggests possible redundancy amongst Abstract Here we used RNA interference and examined possi- ble redundancy amongst Rho GTPases in their mitotic role. Rho GTPases for mitotic role. Chromosome misalignment is induced significantly in HeLa cells To date, five Cdc42-like GTPases including Cdc42, TC10, by Cdc42 depletion and not by depletion of either one or all of TCL, and Wrch2 (Chp) have been described [12]. These the other four Cdc42-like GTPases (TC10, TCL, Wrch1 or GTPases are highly related except variations in the C-terminal Wrch2), four Rac-like GTPases or three Rho-like GTPases. sequences. Some or all of these Cdc42-like GTPases might Notably, combined depletion of Cdc42 and all of the other four compensate for the loss of Cdc42 in mitosis, because they all Cdc42-like GTPases significantly enhances chromosomal mis- show biological activities similar to those of Cdc42 when over- alignment. These observations suggest that Cdc42 is the primary expressed, such as the induction of filopodia. Furthermore, GTPase functioning during mitosis but that the other four other Rho GTPases may also compensate for the loss of Cdc42-like GTPases can also assume the mitotic role in its Cdc42, because mDia3 that functions as a Cdc42 effector in absence. 2006 Federation of European Biochemical Societies. Published mitosis can bind to the GTP-bound form of Rac and Rho as by Elsevier B.V. All rights reserved. well [4]. To test this possibility we generated siRNA for each of the five Cdc42-like GTPases (Cdc42, TC10, TCL, Wrch1, Keywords: Rho GTPase; RNA interference; Mitosis; Wrch2) as well as three Rho GTPases (RhoA, RhoB, RhoC) Chromosome segregation; Cytokinesis and four Rac-like GTPases (Rac1, Rac2, Rac3, RhoG), and examined the effect of depletion of these GTPases either alone or combined on mitosis of HeLa cells.

1. Introduction 2. Materials and methods

Rho GTPases including Rho, Rac, and Cdc42 regulate the 2.1. Plasmids and antibodies assembly of actin cytoskeleton and the local dynamics of pGEX-Cdc42, pGEX-Rac1, pGEX-RhoA, pGEX-RhoB, and microtubules, thereby contributing to cell morphogenesis [1]. pGEX-RhoC were described [13]. pEGFP-TC10 was provided by I. Macara. Antibodies used are mouse monoclonal antibody to Cdc42 Rho GTPases also control cell cycle progression. Rho, Rac and Rac1 (BD Transduction Laboratories), monoclonal antibody to and Cdc42 regulate G1 to S phase progression, presumably RhoA (Santa Cruz Biotechnology), rabbit monoclonal antibody to through the regulation of cell–substrate adhesion, while Rho TC10 (Epitomics), polyclonal antibody to GFP (MBL, Japan) and also regulates cytokinesis through induction and maintenance FITC conjugated anti-b- antibody (mouse, mAb; Sigma– Aldrich). of the contractile ring. However, involvement of Rho GTPases in mitosis remained elusive [2]. Using Clostridium difficile toxin B that inhibits all members of Rho GTPases [3] and expressing 2.2. cDNA cloning and plasmid construction cDNAs for TCL, Wrch1, Wrch2 (Chp), Rac2, Rac3 and RhoG were dominant negative mutant of Cdc42 in HeLa S3 cells, we re- isolated by PCR-based cloning using the primers listed below from cently reported that Cdc42 regulates chromosomal segregation cDNA libraries of HeLa S3 cells and ligated to pCR-Blunt vector [4]. This finding was confirmed further by depletion of Cdc42 (Invitrogen). The cDNAs were then recovered from the plasmids, or its activator, Ect2, with RNAi [5]. While these studies sug- and were subcloned into pEGFP vector. cDNAs for Cdc42, Rac1, RhoA, RhoB, and RhoC were amplified by PCR using the primers gest a critical role of Cdc42 in mitosis, there are several obser- listed below from pGEX-Cdc42, pGEX-Rac1, pGEX-RhoA, pGEX- vations opposing such a role. For example, Chen et al. [6] RhoB and pGEX-RhoC, respectively, and subcloned into pEGFP vec- established a Cdc42-null ES cell line and reported that it pro- tor. pEGFP-WT-TC10, pEGFP-T31N-TC10, and pKH3-Q75L-TC10 liferates normally. Recently, Czuchra et al. [7] also generated were provided by Dr. Ian Macara. Cdc42-deficient fibroblastoid cell lines and reported that loss Primers for cDNA amplification: Cdc42(G25K); (M57298) of Cdc42 neither affected filopodium formation nor had signif- forward primer ATGCAGACAATTAAGTGTGTTGTTGTGGG- icant influence on mitosis. There have also been several Cdc42 CGA RNAi studies reporting no mitotic phenotype [8–11]. This reverse primer TCATAGCAGCACACACCTGCGGCTCTTCTT TC10; (M31470) forward primer ATGCCCGGAGCCGGCCGCAG reverse primer TCACGTAATTAAACAACAGTTTATACATCTTG- *Corresponding author. Fax: +81 75 753 4693. ATCCTATT E-mail address: [email protected] (S. Narumiya). TCL; (AJ276567)

0014-5793/$32.00 2006 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.febslet.2006.05.009 3376 S. Yasuda et al. / FEBS Letters 580 (2006) 3375–3380 forward primer ATGAACTGCAAAGAGGGAACTGACAGCAGC 2.3. siRNA preparation and transfection reverse primer TCAGATAATTGAACAGCAGCTGTGACCCT- We amplified the full coding sequence or internal coding sequence CAG of RhoGTPases from DNA of respective pEGFP-RhoGTPases by Wrch1; (AF378087) the primers listed above, and generated siRNA against each forward primer ACCGAGTACATCCCTACTGCCTTCGACAAC RhoGTPase by using the BLOCK-iT RNAi-TOPO Transcription reverse primer TCATACGAAACAGCAGTACTTCTTCCACCA and the BLOCK-iT Dicer RNAi kits (Invitrogen) [14]. The dou- Wrch2; (AY059636) ble-strand RNA obtained was cleaved into 21–23 nucleotide frag- forward primer GGGCATCAAGTGCGTGCTGGT ments with Dicer and purified by using the RNA reverse primer AGGTAGCAGCAGGCTCGGAT purification regents supplied in the kit. As a control, we used siRNA RhoA; (L25080) for the Escherichia coli LacZ gene as provided by the manufacturer. forward primer ATGGCTGCCATCCGGAAGA HeLa S3 cells were seeded onto coverslips in 35-mm dishes at a den- reverse primer TCACAAGACAAGGCAACCAG sity of 2 · 105 per dish and were maintained in DMEM supple- RhoB; (NM_004040) mented with 10% FCS and antibiotics. The following day, cells forward primer ATGGCGGCCATCCGCAAGA were subjected to double thymidine block as described [15]. Just reverse primer TCATAGCACCTTGCAGCAGT after the first thymidine block, HeLa cells were transfected with RhoC; (NM_175744) appropriate siRNA for Rho GTPases using Lipofectoamine 2000 forward primer ATGGCTGCAATCCGAAAGAA (Invitrogen) as described [14]. reverse primer GGGGCTGTCCCATTCTCTGA Rac1; (NM_006908) forward primer ATGCAGGCCATCAAGTGTGTGG 2.4. Immunofluorescence reverse primer TTACAACAGCAGGCATTTTCTC The cells were fixed using 3.7% formaldehyde 12 or 16 h after Rac2; (NM_002872) the second thymidine block, washed three times with PBS(À) forward primer ATGCAGGCCATCAAGTGTGTGG and permeabilized with 0.1% Triton X-100 in PBS(À) for 5 min reverse primer CTAGAGGAGGCTGCAGGCGC at room temperature. After additional three washes with PBS(À), Rac3; (NM_005052) the samples were incubated for 60 min at room temperature in forward primer ATGCAGGCCATCAAGTGCGTGG PBS(À) with a FITC-conjugated antibody to b-tubulin (1:200, reverse primer CTAGAAGACGGTGCACTTCTTCCC mouse, Sigma). DNA was stained with TOPRO-3 (Molecular RhoG; (AF498974) Probe). Cells were examined with a Zeiss LSM510 confocal imag- forward primer ATGCAGAGCATCAAGTGCGTGGTG ing system or Leica AS MDW Multi-Dimensional Workstation for reverse primer TCACAAGAGGATGCAGGACCGC Fig. 4C.

Fig. 1. Depletion of Rho GTPases by RNAi. (A) Depletion of EGFP-GTPase. HeLa cells were transfected with siRNA for an indicated GTPase together with the corresponding pEGFP–GTPase fusion construct. The cells were collected at 0, 24 and 48 h after transfection and subjected to immunoblots with an antibody to EGFP. (B,C) Depletion of endogenous Rho GTPases. HeLa cells were transfected with siRNA for each indicated Rho GTPase or groups of Rho GTPases (five Cdc42-like GTPases, Cdc42, TC10, TCL, Wrch1 and Wrch2; four Cdc42-like GTPases, TC10, TCL, Wrch1 and Wrch2; four Rac-like GTPases, Rac1, Rac2, Rac3 and RhoG; three Rho-like GTPases, RhoA, RhoB, and RhoC). After 24 and 48 h, the cells were lysed and subjected to immunoblotting with the respective antibody. S. Yasuda et al. / FEBS Letters 580 (2006) 3375–3380 3377

3. Results NAs for three RhoA-like proteins (RhoA, RhoB and RhoC). RNAi for Rac1, and that for the four Rac-like proteins sup- 3.1. RNAi for each Rho GTPases pressed expression of endogenous Rac1 protein to <10% of We transfected HeLa cells with siRNA for each Rho that of control cells. Similarly, RNAi for RhoA and that for GTPase together with a pEGFP-plasmid harboring the respec- the three RhoA-like proteins effectively suppressed expression tive GTPase and examined effects of RNAi on the expression of endogenous RhoA protein. of the GFP-fused protein by anti-GFP antibody. As a control, In order to confirm depletion of individual GTPase when up we used siRNA for the Escherichia coli LacZ gene. The cells to five different genes were simultaneously subjected to RNAi, were collected 24 and 48 h after transfection and subjected to we examined the levels of endogenous TC10 and Cdc42, the immunoblotting [5]. siRNA treatment suppressed the expres- levels of endogenous Rac1 and Cdc42 and the levels of endog- sion of individual GFP-fused GTPase almost completely, enous RhoA and Cdc42 in RNAi for five Cdc42-like GTPases, including five Cdc42-like GTPases (Cdc42, TC10, TCL, RNAi for Cdc42 combined with four Rac-like GTPases and Wrch1, Wrch2), four Rac-like GTPases (Rac1, Rac2, Rac3, RNAi for Cdc42 combined with three Rho-like GTPases, RhoG), and three Rho-like GTPases (RhoA, RhoB, RhoC) respectively (Fig. 1C). Each GTPase examined exhibited al- (Fig. 1A). most equal level of depletion among them and as compared We next used antibodies against Cdc42, Rac1, or RhoA, and to that found in RNAi for the respective GTPase alone. examined depletion of endogenous protein by siRNA for the respective protein and a set of siRNAs for related GTPases 3.2. Effect of RhoGTPase depletion on chromosome alignment (Fig. 1B). We first transfected HeLa cells with siRNA for LacZ during mitosis (Control), or that for Cdc42, or sets of siRNAs either for five To examine the effect of depletion of Rho GTPases on mito- Cdc42-like proteins (Cdc42, TC10, TCL, Wrch1, and Wrch2), sis, we synchronized HeLa cells at the beginning of S phase or four Cdc42-like proteins (TC10, TCL, Wrch1, and Wrch2). using a double-thymidine block and transfected them with siR- As expected, RNAi for Cdc42 as well as that for five Cdc42- NA immediately after release from the first thymidine block. like proteins effectively suppressed expression of the endoge- dsRed2-Histone H2BK was co-transfected to visualize nuclear nous Cdc42 protein 24 or 48 h after transfection to less than morphology and to identify transfected cells. We fixed cells 10% of that of control cells. On the contrary, depletion was 12 h after the second thymidine block, when the majority of not observed in cells subjected to RNAi for LacZ or to that cells entered mitosis, and determined the number of cells with for four Cdc42-like proteins other than Cdc42. These observa- misaligned chromosomes that localized close to the poles and tions indicate that RNAi for Cdc42 is specific to Cdc42, did not congress at the metaphase plate (Fig. 2A). Quantitative though the nucleotide sequences for Cdc42-like proteins are analysis revealed that the proportion of cells with misaligned highly conserved. We then transfected HeLa cells with siRNA chromosomes significantly increased in the Cdc42-depleted cell for Rac1 or a set of siRNAs for four Rac-like proteins (Rac1, population compared with that in the control population Rac2, Rac3 and RhoG), or siRNA for RhoA or a set of siR- (Fig. 2B). No increase was found either in cells subjected to

Fig. 2. Effects of Rho GTPase RNAi on chromosome alignment. HeLa cells were transfected with siRNA for E. coli LacZ or siRNA for each Rho GTPase or indicated combinations of GTPases together with pdsRed2-HistoneH2BK after the first thymidine block. The cells were fixed 12 h after the second thymidine block and subjected to immunofluorescence. (A) Typical images of mitotic cells treated with control RNAi, RNAi for Cdc42 and the combined RNAi for the five Cdc42-like GTPases (Cdc42, TC10, TCL, Wrch1 and Wrch2). (B) Percentages of cells with chromosome misalignment among cells subjected to indicated RNAi and expressing dsRed2-HistoneH2BK. Results are the means ± S.E.M. of 100 cells for each condition (n = 4). *, P < 0.05. 3378 S. Yasuda et al. / FEBS Letters 580 (2006) 3375–3380

RNAi for any of the other Rho GTPases alone or in cells sub- cant compared with RNAi for Cdc42 alone. We then carried jected to RNAi for the four Rac-like GTPases together or the out RNAi for TC10, TCL and Cdc42 or RNAi for Wrch1, three Rho GTPases together. On the other hand, the propor- Wrch2 and Cdc42. These treatments significantly increased tion of cells with misaligned chromosomes dramatically in- the number of cells with misaligned chromosomes when com- creased in cells subjected to combined RNAi for the five pared with the RNAi for Cdc42 alone. These observations Cdc42-like GTPases (Cdc42, TC10, TCL, Wrch1 and Wrch2), indicate that each of four Cdc42-like GTPases can work and the extent of misalignment also appeared stronger com- redundantly to compensate for the loss of Cdc42. Finally we pared to that found in the cells with the Cdc42 RNAi alone examined whether Rac-like or Rho-like GTPases can work (Fig. 2B). In contrast, no significant increase in the population redundantly to compensate for the loss of Cdc42 in this pro- with misaligned chromosomes was found in the cells subjected cess. Neither the combined RNAi for Cdc42 and four Rac- to combined RNAi for four Cdc42-like GTPases (TC10, TCL, like GTPases nor that for Cdc42 and three Rho-like GTPases Wrch1, and Wrch2). These observations suggest that Cdc42 is significantly increased the percentage of cells with misaligned primarily responsible for the regulation of chromosome align- chromosome compared with RNAi for Cdc42 alone. ment in mitosis under normal conditions, but that other Cdc42-like GTPases can compensate for the loss of Cdc42 3.3. Effect of Rho GTPase depletion on chromosome segregation function during this process. We next fixed cells 16 h after thymidine removal, and exam- In order to evaluate the contribution of each of the Cdc42- ined nuclear morphology of cells that had entered interphase. like GTPases to chromosome alignment, we depleted individ- As reported [5] and consistent with the above findings, we ob- ual Cdc42-like GTPases (TC10, TCL, Wrch1 or Wrch2) served multinucleate cells with macro- and micronuclei and nu- together with Cdc42. These treatments tended to enhance clear bridges in the population of cells subjected to RNAi for chromosome misalignment but the difference was not signifi- Cdc42 (Fig. 3A). Quantitative analysis revealed that the

Fig. 3. Induction of multinucleated cells by RNAi. HeLa cells were transfected with siRNA and pdsRed2-HistoneH2BK as described. The cells were fixed 16 h after the second thymidine block and subjected to immunofluorescence. (A) Typical images of cells treated with control RNAi, RNAi for Cdc42, the five Cdc42-like GTPases, or the three Rho-like GTPases (RhoA, RhoB, and RhoC). Arrows show micronuclei. (B) Percentages of multinucleate cells with aberrant nuclei among cells subjected to indicated RNAi and expressing dsRed2-HistoneH2BK. (C) Induction of binucleate cells by combined RNAi for the three Rho-like GTPases. Results in B and C are the means ± S.E.M. of 100 cells for each condition (n = 4). *, P < 0.05. S. Yasuda et al. / FEBS Letters 580 (2006) 3375–3380 3379 proportion of cells that showed this mitotic abnormality was as its dominant active and dominant negative mutants. Previ- significantly higher in the Cdc42-depleted cell population than ously we reported that overexpression of both dominant active the control cell population (Fig. 3B). Furthermore, combined and dominant negative mutants of Cdc42 interfered with mito- RNAi for the five Cdc42-like GTPases dramatically increased sis [4]. We found a significant rise in the population of mitotic the proportion of cells with the abnormal nuclear phenotype, cells with misaligned chromosomes and in that of G1-phase and the phenotype appeared more striking than that produced aberrant multinucleate cells when active Q75L or negative by RNAi for Cdc42 alone (Fig. 3A). RNAi for TC10, TCL and T31 N mutant of TC10 was expressed (Fig. 4). Notably, a little Cdc42 or RNAi for Wrch1, Wrch2 and Cdc42, again, signifi- but significant rise in these cell populations was also found cantly increased the number of cells with the phenotype com- when wild type TC10 was overexpressed. pared with RNAi for Cdc42 alone. In contrast, RNAi for TC10, TCL, Wrch1 and Wrch2 or RNAi for the four Rac-like GTPses or the three Rho GTPases either alone or in combina- 4. Discussion tion did not induce the nuclear phenotype significantly (Fig. 3B). Notably, however, the combined RNAi for the three The Rho GTPase family has now expanded to contain more Rho GTPases and not RNAi for individual member produced than 20 members, and have been grouped into the subfamilies bi-nucleated cells containing two nuclei of equal size, (Fig. 3C). of Rho, Rac, Cdc42 and others. While the members within The above findings indicate that Cdc42-like GTPases other each subfamily show highly conserved amino acid sequences, than Cdc42 can substitute for Cdc42 in its mitotic role. To con- redundant functions amongst these GTPases have not been firm this hypothesis, we overexpressed wild type TC10 as well tested rigorously. Here we have used RNAi, and shown defi-

Fig. 4. Effect of TC10 mutants on mitosis. HeLa cells were transfected with pEGFP (control), pEGFP-WT-TC10, pKH3-Q75L-TC10, pEGFP- T31N-TC10 or pEGFP-T17N-Cdc42 during the second thymidine block. The cells were fixed 10 h (A,B) or 14 h (C,D) after the second thymidine block and subjected to immunofluorescence. Typical images of mitotic cells (A) and G1 cells (C) transfected with control EGFP and EGFP-T31N- TC10 are shown. Percentages of cells with chromosome misalignment and aberrant multinucleate cells among cells transfected with indicated constructs were analyzed at mitosis (B) or G1 phase (D). Results are the means ± S.E.M. of 100 cells for each condition (n = 3). Ã, P < 0.05. 3380 S. Yasuda et al. / FEBS Letters 580 (2006) 3375–3380 nitely the redundant roles amongst Cdc42-like GTPases in Cdc42 and mDia3 regulate microtubule attachment to kineto- mitosis and those of Rho-like GTPases in cytokinesis. These chores. Nature 428, 767–771. findings provide an explanation why several genome-wide [5] Oceguera-Yanez, F., Kimura, K., Yasuda, S., Higashida, C., Kitamura, T., Hiraoka, Y., Haraguchi, T. and Narumiya, S. RNAi screening experiments have not necessarily revealed (2005) Ect2 and MgcRacGAP regulate the activation and implications of Cdc42 in mitosis [8] and why Cdc42-null ES function of Cdc42 in mitosis. J. Cell Biol. 168, 221–232. cell lines proliferate normally [6,7]. Probably in most cells these [6] Chen, F., Ma, L., Parrini, M.C., Mao, X., Lopez, M., Wu, C., redundant mechanisms operate so that they easily compensate Marks, P.W., Davidson, L., Kwiatkowski, D.J., Kirchhausen, T., Orkin, S.H., Rosen, F.S., Mayer, B.J., Kirschner, M.W. and Alt, for the loss of one mechanism and HeLa cells that show mito- F.W. (2000) Cdc42 is required for PIP(2)-induced actin polymer- tic defect to the loss of Cdc42 alone may be an exception. The ization and early development but not for cell viability. Curr. present study also provides an answer to a question whether all Biol. 10, 758–765. of the Rho-like GTPases or its specific member regulate cyto- [7] Czuchra, A., Wu, X., Meyer, H., van Hengel, J., Schroeder, T., kinesis. It can be assumed that TC10, TCL, Wrch1 and Wrch2, Geffers, R., Rottner, K. and Brakebusch, C. (2005) Cdc42 is not essential for filopodium formation, directed migration, cell bind to the same effectors that Cdc42 binds to compensate for polarization, and mitosis in fibroblastoid cells. Mol. Biol. Cell. the mitotic function of Cdc42, because they are highly con- 16, 4473–4484. served in the effector region, and it is known, for example, that, [8] Somers, W.G. and Saint, R. (2003) A RhoGEF and Rho family similar to Cdc42, all bind to Pak1 [12]. On the other hand, GTPase-activating protein complex links the contractile ring to cortical microtubules at the onset of cytokinesis. Dev. Cell 4, 29– these proteins exhibit variations in the C-terminal amino acid 39. sequence, and are suggested to exert specific functions at differ- [9] Jantsch-Plunger, V., Gonczy, P., Romano, A., Schnabel, H., ent locations within the cell. How these GTPases are recruited Hamill, D., Schnabel, R., Hyman, A.A. and Glotzer, M. (2000) and function redundantly during mitosis in the absence of CYK-4: a Rho family GTPase activating protein (GAP) required Cdc42 remains to be clarified by future study. for central spindle formation and cytokinesis. J. Cell Biol. 149, 1391–1404. [10] Kay, A.J. and Hunter, C.P. (2001) CDC-42 regulates PAR Acknowledgment: This work was supported in part by a Grant-in-Aid protein localization and function to control cellular and embry- for Specially Promoted Research from the Ministry of Education, onic polarity in C. elegans. Curr. Biol. 11, 474–481. Culture, Sports, Science, and Technology of Japan. [11] Kamijo, K., Ohara, N., Abe, M., Uchimura, T., Hosoya, H., Lee, J.S. and Miki, T. (2006) Dissecting the role of Rho- mediated signaling in contractile ring formation. Mol. Biol. Cell 17, 43–55. References [12] Wennerberg, K. and Der, C.J. (2004) Rho-family GTPases: it’s not only Rac and Rho (and I like it). J. Cell Sci. 117, 1301– [1] Etienne-Manneville, S. and Hall, A. (2002) Rho GTPases in cell 1312. biology. Nature 420, 629–635. [13] Watanabe, N., Kato, T., Fujita, A., Ishizaki, T. and Narumiya, S. [2] Narumiya, S., Oceguera-Yanez, F. and Yasuda, S. (2004) A new (1999) Cooperation between mDia1 and ROCK in Rho-induced look at Rho GTPases in cell cycle: role in kinetochore-microtu- actin reorganization. Nat. Cell Biol. 1, 136–143. bule attachment. Cell Cycle 3, 855–857. [14] Yasuda, S. and Narumiya, S. (2006) Analysis of a mitotic role of [3] Just, I., Selzer, J., Wilm, M., von Eichel-Streiber, C., Mann, M. Cdc42. Methods Enzymol. 406, 656–665. and Aktories, K. (1995) Glucosylation of Rho proteins by [15] Bostock, C.J., Prescott, D.M. and Kirkpatrick, J.B. (1971) An Clostridium difficile toxin B. Nature 375, 500–503. evaluation of the double thymidine block for synchronizing [4] Yasuda, S., Oceguera-Yanez, F., Kato, T., Okamoto, M., mammalian cells at the G1-S border. Exp. Cell Res. 68, 163– Yonemura, S., Terada, Y., Ishizaki, T. and Narumiya, S. (2004) 168.