Cell Motility and the Cytoskeleton 3549-58 (1996) Differential Expression of Tubulin lsotypes During the Cell Cycle Charles Dumontet, George E. Duran, Katherine A. Steger, Gloria L. Murphy, Howard H. Sussman, and Branimir 1. Sikic Departments of Medicine and Pathology, Stanford University School of Medicine, Stanford, California Microtubules play an essential role in cell division. Little is known about possible variations of total tubulin and tubulin isotype expression during the cell cycle. We analyzed the total tubulin content, tubulin polymerization status and tubulin isotype content in resting and dividing human K562 leukemic cells and human MES-SA sarcoma cells. Although the total cellular tubulin content increases as the cells progress toward mitosis, the total tubulidtotal protein ratio is stable during the cell cycle. Reverse transcriptase-polymerase chain reaction was applied to analyze the levels of expression of a,p, and y-tubulin isotypes. Whereas a-tubulin isotype and y-tubulin transcripts were found to be expressed at constant levels throughout the cell cycle, some of the P-tubulin isotype transcripts were found to be more highly expressed in dividing then in resting cells. Both of the class IV P-tubulin isotype transcripts (human 5p and p2, Class IVa and IVb, respectively) were expressed in dividing K562 and MES-SA cells at twice the levels found in resting cells. Increased expression of the class IV isotype proteins in dividing cells was confirmed by immunoblotting, both in K562 and in MES-SA cells. A larger fraction of total cell tubulin was found to be polymerized in dividing cells (36-40%) than in resting cells (27-30%). The degree of polymerization of class IV tubulin in dividing and resting cells was similar to that of total tubulin. These results show that total tubulin is expressed as constant levels throughout the cell cycle but that the degree of polymerization is increased as cells are committed to division. The relative overexpression of the two class IV p-tubulin isotypes in dividing cells suggests functional specificity for these isotypes and a regulatory role of these isotypes on the microtubule network during mitosis. 0 1996 Wiley-Liss, Inc. Key words: cell cycle, tubulin isotypes, microtubule, polymerization INTRODUCTION Microtubules are dynamic polymeric organelles involved in numerous cellular processes including mitosis [Borisy and Taylor, 1967; Kirschner, 19781. Tubulin, first described as a “colchicine-binding protein” was shown to be the building block of all microtubular structures [Wilson and Meza, 19731. The demonstration that tubulin dimers were heteropolymers of a and p subunits [Luduena et al., 19771 was later followed by the demonstration that both of these subunits existed under the form of various isotypes [Cleveland et al., 19801. Thanks to the work of Cowan and coworkers in mam0 1996 Wiley-Liss, Inc. mals and Cleveland and coworkers in chickens, a number of isotypes of each subunit have been identified, including seven a-tubulin and seven P-tubulin isotypes [Cowan et al., 1986; Hall et al., 1983; Lewis et al., 1985; Sullivan, 19881. Tubulin isotypes in humans belong to a multigene family of 15 to 20 genes, many of which are pseudogenes [Lewis et al., 1987; Wilde et al., 19821. More recently a less abundant form of tubulin has been identified, termed y-tubulin, which appears to be inReceived December 19, 1995; accepted May 20, 1996. Address reprint requests to Dr. B.I. Sikic, Room M-211, Oncology Division, Stanford Medical Center, Stanford, CA 94305-5306. 50 Durnontet et al. volved in nucleation of microtubules from the centrioles [Zheng et al., 19911. Presently only one isotype of y-tubulin has been reported [Oakley and Oakley, 19891. Tubulin isotypes differ essentially by their terminal carboxy portion [Sullivan, 19881. Tubulin isotypes, although quite different among themselves in a given species, have been remarkably conserved throughout evolution [Little and Seehaus, 19881. It has thus been tempting to attribute functional specificities to some or all of these isotypes. A number of attempts to identify functional characteristics specific for tubulin isotypes have been made. Although the question of the functional specialization of tubulin isotypes remains controversial, there is definite evidence that some isotypes are differentially expressed and required for some specific processes [Luduena, 1993; Raff, 19941. Some of the most compelling evidence for functional specialization of isotypes has been provided by the demonstration that some isotypes are critically involved in meiosis and neurite outgrowth [Gard and Kirschner, 1985; Joshi and Cleveland, 1989; Kemphues et al., 19801. More recently, analysis of the effect of isotype composition on microtubule stability in vitro has shown that different isotypes form microtubular structures with different dynamic properties [Banerjee et al., 1992; Lu and Luduena, 1993; Panda et al., 19941. In the present work we have analyzed the expression of total tubulin and tubulin isotypes during the cell cycle. We have chosen as models the human erythroleukemic cell line K562 which grows in suspension and the adherent human sarcoma cell line MES-SA. Total tubulin was analyzed by immunoblotting and fluorescence-activated cell sorter (FACS) analysis of viable cells. Tubulin isotype transcript levels were analyzed by semi-quantitative reverse transcriptase-polymerase chain reaction (rtPCR) , using specific primers which we designed and immunoblotting of class IV P-tubulin. Our results demonstrate that in both cell lines, the two class IV P-tubulin isotypes, which share a common carboxy-terminal portion, are expressed at higher levels in dividing cells than in resting cells, suggesting a role of these isotypes in mitosis. in a humidified atmosphere containing 5% CO,. Cell stocks were screened routinely for Mycoplasma by the DNA hybridization method (Gen-Probe, Inc., San Diego, CA) and by rt-PCR. Log phase cells were obtained by seeding the cells at low concentration (5 X lo4 K562 cells/ml; 1 X 106 MES-SA cells in a 80-cm2 flask) and collecting them after 48 hours of culture. Plateau cells were obtained by collecting cells 24-48 hours after they had reached plateau phase of growth (K562 cells) or confluence (MES-SA). Analysis of Tubulin Content and Cell Cycle by FACS To determine the percentages of cells in each phase of the cell cycle, cells were collected, fixed with ice-cold methanol containing 2 mM EGTA, washed, and resuspended in propidium iodide (50 pg/ml), then analyzed by FACS. Percentages of cells in G1, S, and G2/M phase were determined using CellFit@Software (Becton-Dickinson, San Jose, CA). For analysis of tubulin content in the cell cycle, cells were first labeled with murine monoclonal antibody directed against all P-tubulin isotypes and a secondary fluoresceinated antibody, then resuspended in propidium iodide, as previously described [Jaffrezou et al., 19951. Briefly, cells were permeabilized prior to staining with cold (-20°C) methanol, then exposed to pan-P-tubulin antibody (Sigma, 1:1,OOO) for 1 hour at room temperature, washed thrice, and exposed to FITC-labeled goat-antimouse antibody for 1 hour at room temperature. Cells were washed and resuspended in a solution of propidium iodide (50 pg/ml), then analyzed by FACS. Evaluation of Total Tubulin and Class IV p-Tubulin lsotype Protein by lmmunoblotting Cells were harvested in log or plateau phase of growth, and pellets were resuspended in lysis buffer containing Tris-HC1 pH 6.80, 1 mM MgCl,, 2 mM EGTA, and 0.2% Tween 20 and protease inhibitors (phenylmethylsulfonyl fluoride [PMSF] 1 mM, leupeptin 50 pg/ ml, pepstatin 1 pg/ml, trypsin inhibitor 1 mg/ml, and aprotinin 20 kg/ml [Sigma, St. Louis, MO]). Total protein was quantified by the Lowry assay [Lowry et al., MATERIALS AND METHODS 19511, and samples were prepared in sodium dodecyl Cell Culture sulfate (SDS), then boiled before being applied to a 12% The human erythroleukemic cell line K562 was polyacrylamide gel, as previously described. Proteins purchased from the American Type Culture Collection. were then blotted onto a Hybond-ECL nitrocellulose The MES-SA cell line was derived in our laboratory membrane (Amersham, Buckinghamshire, UK) using a from sarcomatous elements of a uterine mixed mullerian Sartorius apparatus (Hayward, CA). The membrane was tumor [Harker and Sikic, 19851. All cell lines were cul- blocked with buffer containing 5% milk and 1% bovine tured in McCoy 5A medium supplemented with 10% albumin, then incubated 2 hours at room temperature newborn calf serum, 2 mM glutamine, 200 unitdm1 pen- with pan+ monoclonal antibody (Sigma Immunochemicillinlml, and 100 pg streptomycin (all from Irvine Sci- icals, 1:3,000 dilution) or anti-class IV P-tubulin isotype entific, Santa Ana, CA). Cells were maintained at 37°C monoclonal antibody (Biogenex, San Ramon, CA, 1: Tubulin Isotypes and the Cell Cycle 1,000 dilution), washed, incubated in goat anti-mouse antibody, washed, incubated in streptavidin-biotin, washed, and processed in ECL reagents. Phosphocellulose-purified tubulin prepared from bovine brain (generously provided by M.A. Jordan, University of Santa Barbara) was used as a control. Preliminary experiments were performed to determine the dilution of antibodies allowing quantification of tubulin content. 0 0 0 0 0 Evaluation of Polymerized and Soluble Tubulin Protein Polymerized and soluble tubulin were distinguished using a modified version of the method reported Minotti et al., . Briefly, cells were harvested, washed, and lysed as described above. Cell lysates were then incubated for 5 min at 37°C in the dark, and the cellular residues were mixed briefly and centrifuged at 14,000 rpm for 10 min at room temperature. The resultant supernatants were transferred to a separate centrifuge tube and kept on ice. The pellet was resuspended in Ling's Lysis buffer (10 mM Tris pH 7.5, 1.5 mM MgCl,, and 10 mM KC1) in a volume equal to the supernatant. A volume equivalent to 100 kg of sample protein was incubated in 4 x Laemmli buffer, vortexed, and boiled for 10 min prior to loading onto a 12% acrylamide-SDS gel. Gels were then processed for immunoblotting as described above. The percent of polymerized tubulin was calculated by comparing the relative ratio of polymerized tubulin (pellet fraction) to the unpolymerized tubulin dimers (supernatant) by densitometry. rt-PCR 0 0 0 0 0 0 0 0 0 0 51 P4 reverse primer: (223, 243) AAA GGC CCC TGAGCGGACACT 5P (Class IVa) forward primer: (-85, -68) TCT CCGCCGCATCTTCCA 5p reverse primer: (167, 186) TCT GGG GAC ATA ATT TCC TC p2 (Class IVb) forward primer: (-42, -22) GTC TACTTCCTCCTCTTCCC P2 reverse primer: (291, 300) GTT GTT CCC AGC ACC ACT CT B a l forward primer: (1003, 1020) ATC AAG ACC AAG CGT ACC B a l reverse primer: (1363, 1380) CAG CAC CTT TGT GAC GTT K a l forward primer: (1000, 1017) ACC ATC AAA ACC AAG CGC K a l reverse primer: (1363, 1380) TGC AGG GCC AAA AGG AAT Ha44 forward primer: (139, 158) CCT TCA CCA CCT TCT TCT GT Ha44 reverse primer: (230, 149) TCG GTA TGG GCC ATT TCG GA H2a forward primer: (1059, 1075) GTG GGC ATT AAC TAC CAG H2a reverse primer: (1383, 1400) CAA CGT GGA AGC AGC CAT y forward primer: (1055, 1072) AGT TGG CCA ACT TCA TCC y reverse primer: (1349, 1367) TGC CCC AGG AGA TGT AGT Primers for a-tubulins were designed using published sequences [Cowan et al., 1983; Dobner et al., Total RNA extraction and rt-PCR were performed 1987; Villasante et al., 19861. Primers for y-tubulin were as previously described [Chen et al., 19941. PCR was designed using the sequence available from GeneBank performed in a Perkin-Elmer Cetus DNA Thermal Cycler (accession number M61764). The P-tubulin isotype clas(Norwalk, CT) using the following profile: 10 s at 94"C, sification used is the one described by Sullivan [Sullivan, 30 s at 55"C, and 30 s at 72°C. The amplimers used in 19881. Primers for P-tubulin isotypes were designed usthis study were synthesized by Operon Technologies I, class IVa, and ing published sequence data for class (Alameda, CA). class IVb isotypes [Lewis et al., 19851 or, in the case of We designed the following primers for analysis of I11 isotype a consensus forward primer and partial class tubulin isotypes (in the case of P-tubulin isotypes, Arasequence information generously provided by Kevin Sulbic numerals refer to the gene, Roman numerals refer to I, livan (Scnpps Research Institute, La Jolla, CA). Class the tubulin protein isotype class): 111, IVa, and IVb P-tubulin primers were designed to 0 M40 (Class I) forward primer: (-42, -22), CCA span introns. In the case of the class I1 P-tubulin isotype, TAC ATA CCT TGA GGC GA sequence was obtained from the EMBL GeneBank (ac0 M40 reverse primer: (226, 246) GCC AAA AGG cession number X7 and 9353), and using the peptide ACCTGAGCGAA sequence previously reported by Cowan et al. [19861. In 0 P9 (Class 11) forward primer: (1 31, 1150) CGC separate experiments the presence of tubulin pseuATC TCC GAG CAG TTC AC dogenes was analyzed by performing PCR directly on the 0 P9 reverse primer: (1301, 1319) TCG CCC TCC RNA, and by digesting RNA with DNAse (Gene Hunter TCC TCC TCG A Kit, Brookline, MA) before cDNA synthesis. 0 P4 (Class 111) forward primer: (1 15) ATG AGG To allow semi-quantification, 28s ribosomal GAA ATC GTG cDNA was used as an endogenous control for PCR since Dumontet et al. 52 TABLE I. Cell Cycle Distribution of Log and Plateau K562 and MES-SA Cells* K562 log phase K562 plateau phase MES-SA log phase MES-SA Dlateau Dhase 24 + 4 88 k 8 18 5 86 k 9 * 70 k 6 9*3 72 2 5 10 -t 3 6 + 2 3 k l 10 k 2 4+ 1 ~ *K562 and MES-SA log phase cells were collected 48 hours after having been seeded at low concentration. Plateau phase cells were collected 24 hours after having been seeded at maximal cell concentrations achieved by spontaneously growing cells. the gene for rRNA is expressed at uniformly high levels and may be used as a normalization factor for total RNA content of cells. The amplimers used for ribosomal RNA were the following: rRNA-A (1846-1826) TTA CCA AAA GTG GCC CAC TA; rRNB-B (1501-1520) GAA AGA TGG TGA ACT ATG CC. Furthermore, samples were run at different cDNA concentrations and different number of PCR cycles to ensure that the reaction was not at the plateau phase. PCR samples were analyzed by 8% polyacrylamide gel electrophoresis, stained with ethidium bromide, and analyzed by densitometric reading of bands on an Alpha Innotech IS-1000 image analyzer (San Leandro, CA). RESULTS Cell Cycle Distribution of Log and Plateau Phase Cells Analysis of cell cycle distribution, as determined by propidium iodide staining and FACS analysis, of log phase cells showed that, under the conditions of growth used, both K562 and MES-SA cells were predominantly in S phase (Table I). Conversely, in the plateau phase, 86 to 88% of the cells were in GO/Gl phase. These results validate the use of log and plateau conditions of culture to compare dividing and non-dividing cells. Total Tubulin Content in Dividing and Non-Dividing MES-SA and K562 Cells Total tubulin content per cell was evaluated by double staining of perrneabilized log phase cells with propidium iodide and pan-p tubulin antibody and a secondary FITC-labeled anti-murine antibody. As shown in Figure 1, total tubulin content per cell increased by approximately 6 0 4 0 % as the cells progressed from G1 through S phase and into G2/M phase. The percent increase was similar in K562 and MES-SA cells. To determine whether the increase in tubulin content of dividing cells was specific to cells undergoing mitosis or was due to growth of cell size and increased protein content during cell division, we performed im- munoblots comparing the amount of total p-tubulin relative to total cellular protein, in log and plateau phase cells. Various protein loadings (25, 50, and 100 pg) were compared, and protein loading was confirmed by Coomassie staining of duplicate gels. Purified bovine brain tubulin (5 pg) was used as a control. As shown in Figure 2 (100 pg loading), the ratio of total p-tubulin to total protein was similar in dividing and non-dividing cells. This was found to be the case both in K562 cells and in MES-SA cells. Determination of total protein content per cell in log and plateau phase cells demonstrated higher total protein content in the log phase cells (443 pg/cell in K562 cells and 164 pgkell in MES-SA cells) than in plateau phase cells (160 pg/cell in K562 cells and 112 pgkell in MES-SA cells). Given the fact that the majority of the cells in the log and the plateau populations are in S phase and in GO/Gl phase, respectively, these differences should be interpreted as reflecting the total protein content of resting cells and cells preparing for mitosis. The difference in total protein content between resting cells and cells in M phase is probably even greater. Expression of Tubulin lsotypes During the Cell Cycle a, p, and y-Tubulin isotype transcripts were analyzed by rt-PCR. In these experiments, isotype content was normalized to 28s ribosomal RNA content. These experiments were performed on three to six different batches of cells. Given the caveats of semi-quantitative PCR, samples were run at different number of cycles and at different concentrations of cDNA in order to ensure that the reaction had not reached saturation. Furthermore to ensure comparability of samples, cDNA concentrations were chosen in order to obtain ribosomal PCR products which did not differ in amount by more than 10%. Among the sets of primers analyzed, only H2a and class IVa p-tubulin generated PCR products compatible with the presence of pseudogenes. In both cases, however, the amount of PCR products produced were low (less than 5% of the corresponding products obtained from cDNA), and pretreatment with DNAse did not significantly modify the amount of PCR product obtained (data not shown). Tubulin isotype content of K562 and MES-SA cells are shown in Table I1 and Figure 3 . The isotype profile of the two cell lines is clearly different, in particular as concerns the a-tubulin content. Both K562 and MES-SA cells were found to express two of the four a-tubulin isotypes. Both cell lines expressed the more abundant K l a isotype. However, K562 cells were found to express Ha44 and MES-SA cells were found to express B l a . None of the cell lines expressed H2a, which had initially been reported to be a testis-specific isotype. Tubulin Isotypes and the Cell Cycle 53 Phase of the Cell Cycle Fig. 1. Analysis of total P-tubulin content during the cell cycle by FACS analysis. K562 and MES-SA cells were collected in log phase of growth, permeabilized with ethanol, and stained with a pan+ monoclonal antibody and secondary FITC-labeled antibody. Cells were incubated in 50 Kglml propidium iodide and analyzed by FACS on a Becton-Dickinson cytofluorometer. Pan R tubulin scripts, as evidenced by the number of cycles required for amplification. a-Tubulin and y-tubulin transcripts were expressed at comparable levels in log and plateau cells. Among the P-tubulin isotypes, class I, class 11, and class I11 transcripts were expressed at constant levels throughout the cell cycle. Conversely, both of the class IV isotype tranFig. 2. Analysis of total P-tubulin content during the cell cycle by scripts were expressed approximately twofold more in immunoblotting. Cells in log or plateau phase of growth were lysed, and 100 Kg of total protein was electrophoresed on a 12% SDS- dividing cells than in non-dividing cells, both in K562 polyacrylamide gel. Total bovine brain tubulin (5 pg) was used as a and in MES-SA cells. In both cell lines, the class IVb control. Samples were transferred onto nitrocellulose and immuno- isotype product was amplified at a lower number of cyblotting was performed with pan-P-tubulin monoclonal antibody. cles than the class IVa isotype product. In separate experiments (data not shown), limited numbers of cells stained with propidium iodide were Both cell lines expressed y-tubulin. Both cell lines ex- sorted into GO/G1, S, and G2/M populations, and impressed all classes of P-tubulin isotypes, but class I11 mediately processed for rt-PCR analysis of their tubulin (considered to be a minor neuronal isotype) was detected isotype transcript levels. Results showed no difference in in very small amounts in K562 cells. Class I transcripts class I P-tubulin isotype levels but a twofold increase in appeared to be the most abundant among P-tubulin tran- class IV P-tubulin isotype levels between GO/G1 and S K562 MES-SA PC-TUB 54 Dumontet et al. RIB0 K1 a Bla Ha44 H2a class I class II class 111 class IVa class IVb Y W I- Fi m (0 c P 0 vl Q, 0 I x 7c I- (Q m m v) m N N D Fig. 3. Tubulin isotype PCR products in dividing and resting K562 and MES-SA cells. Log and plateau phase K562 and MES-SA cells were analyzed for their tubulin isotype contents by rt-PCR. Samples were run at various number of cycles using various concentrations of cDNA. 28s ribosomal RNA was used as an endogenous control. Products were run on a 8% polyacrylamide gel, stained with ethidium bromide, and quantified by densitometry. phase, with no difference between S phase and G2/M phase cells. Overexpression of class IV P-tubulin was confirmed at the protein level by immunoblotting with a monoclonal antibody which recognizes the C-terminal portion common to both class IVa and Class IVb isotypes (Fig. 4). Consistent results were found at different protein loadings (25 to 100 pg/lane), on at least three different batches of cells. Immunoblotting of cell lysates with a polyclonal anti-y-tubulin antibody (generously provided by Tim Steams, Stanford University) confirmed that there was no difference in y-tubulin protein content in dividing and non-dividing cells (data not shown). TABLE 11. Expression Ratios of a,p, and y-Tubulin Isotypes in K562 and MES-SA Cells* Polymerization Status of Total Tubulin and Class IV P-Tubulin Comparison of polymerized (cytoskeletal) and soluble tubulin was performed by precipitation of polymer- K562 MES-SA Kla Bla Ha44 H2a Class I p Class I1 p Class I11 p Class IVa p Class IVb f3 Y *The ratios are expressed as log divided by plateau values for each cell type. Tubulin isotype expression was analyzed by semi-quantitative PCR, at different concentrations of cDNA and at different number of PCR cycles. Values were obtained by densitometric reading of ethidium bromide gels and normalized to 28s ribosomal transcript levels. The results are expressed as the ratios of log/plateau cells. Numbers in parentheses represent the number of cycles used for amplification. N.E., isotype not expressed at 44 cycles of PCR. Tubulin Isotypes and the Cell Cycle Class IV R tubulin K562 MES-SA PC-TUB Fig. 4. Immunoblotting of K562 and MES-SA total cell lysates with anti-class IV P tubulin monoclonal antibody. Log and phase cells were collected, and total lysates were run on a 12% SDS-polyacrylamide gel. Purified bovine brain microtubule protein was used as a control. Samples were transferred onto nitrocellulose, and immunoblotting was performed with anti-class IV P-tubulin monoclonal antibody. ized tubulin and immunoblotting of both fractions with pan+ monoclonal antibody and class IV-specific p-tubulin monoclonal antibodies. As shown in Table 111, a larger fraction of total P-tubulin was found to be polymerized in dividing cells (36-40%) than in resting cells (27-30%), both in K562 and in MES-SA cells. The polymerized fraction were quite similar in K562 cells, which grow in suspension, and in MES-SA cells, which are adherent. The fraction of class IV P-tubulin under polymerized form in dividing and resting cells was comparable to that of total P-tubulin (Fig. 5). DISCUSSION Progression through the cell cycle is dependent upon the assembly and dynamic behavior of the mitotic spindle. Constitution of the mitotic spindle allowing chromosome separation at anaphase is an essential step of mitosis. A number of mitotic regulatory molecules, such as cyclin B/p34cdc2, have been shown to bind both to cytoplasmic and to spindle microtubules [Ookata et al., 19931. Certain antimitotic drugs, including depolymerizing agents (such as vinca alkaloids) and stabilizing agents (such as taxanes), are targeted toward the mitotic spindle and inhibit cell division by modifying microtubular dynamics [Jordan et al., 1993; Wilson and Jordan, 19951. Considerable data are available regarding the roles of microtubules in mitosis [McIntosh, 1979; Wordeman and Mitchison, 19941. However most of these data are qualitative, and little is known concerning the variations in total tubulin, specific isotype contents, and polymerization status during the cell cycle. In our experiments, the total cellular tubulin content was found to increase as cells progressed through mitosis, a phenomenon attrib- 55 utable to the increase in total protein content as the cells increase in size before undergoing cytokinesis. However, a larger percentage of total cell tubulin was in the polymerized form in dividing cells. Microtubules form an essential component of the cytoskeleton, and these data suggest that cells preparing for or undergoing mitosis are submitted to specific structural constraints. Our results indicate that dividing cells have higher transcript and protein levels of class IV P-tubulin than non-dividing cells. This was found to be true both in the leukemic cell line K562, which grows in suspension, and in the adherent sarcoma cell line MES-SA. These results suggest that the regulation of class IV P-tubulin during the cell cycle is not organ specific. A number of investigators have analyzed the distribution of isotypes among the different microtubular structures and shown that in most cases all cellular microtubules are copolymers of all available tubulin isotypes [Lewis et al., 1987; Lopata and Cleveland, 19871. Immunofluorescence studies by ourselves (unpublished data) and others have not demonstrated preferential distribution of the class IV isotype with the mitotic spindle [Lopata and Cleveland, 1987; Sawada and Cabral, 19891. However, the level of resolution in these experiments has not allowed the quantitative analysis of the different isotypes in individual subsets of spindle microtubules. Sisodia et al.  have isolated spindle and cytoplasmic microtubules from CHO cells and reported that the ratio of isotypes is comparable in both fractions. Altered isotype content may affect the spindle’s dynamic behavior through a number of ways. To allow chromosomal positioning and separation, the microtubules composing the spindle must be dynamic. It has been shown that microtubule turnover is approximately 20-fold faster in mitotic than in interphase cells [Saxton et al., 19841. Recent data suggest that different P-tubulin isotypes, in particular class 111, differ in their dynamic behavior [Panda et al., 19941. Thus a modification in the ratios of P-tubulin isotypes may play an important role in the regulation of the spindle microtubule dynamics during mitosis. Another possible mechanism of dynamic regulation may rely on the interactions between given isotypes and specific microtubule-associated proteins (MAPs). P-Tubulin isotypes differ primarily between their C-terminal region. This region has been shown to be the binding region for a number of MAPs [Littauer et al., 1986; Paschal et al., 1989; Serrano et al., 19851. Certain MAPs such as tau and MAP2 have been shown to modulate microtubule dynamics as well as flexibility [Dye et al., 1993; Lee and Rook, 19921. Furthermore, coordinated regulation of MAP and tubulin genes has been reported [Oblinger and Kost, 19941. It is possible that microtubules with different isotype contents may display various affinities for the binding of MAPs, and Dumontet et al. Total R Tubulin Soluble Polymerized ao I 70 60 50 40 30 20 10 0 K562 log MES-SA log K562 plateau MES-SA plateau Class IV t3 Tubulin f 52 70 w2 60 a . 3 50 W 40 I 80 I I 0 30 $ 8 a 20 10 0 K562 log K562 plateau Fig. 5. Analysis of polymerized (cytoskeletal) and soluble tubulin in MES-SA and K562 cells in log or plateau phase of growth by precipitation of both fractions with pan+ and Class IV-specific P-tubulin monoclonal antibodies. Fractions were electrophoresed on a 12% consequently be more or less susceptible to their modulating effects. One notable aspect of our findings is that the isotypes which are differentially regulated during the cell cycle are present in relatively small amounts, as suggested by the number of cycles of PCR required to amplify products. Considering the functional multiplicity of microtubules in cells, as well as their mechanistic contraints during mitosis, it should not be considered surprising that an isotype-based regulation of spindle behavior would depend on modifications of a small but functionally important class of regulatory isotypes, rather than on the bulk of structural tubulin isotypes. Quantification of isotypes in different non-neuronal cell - MES-SA log MES-SA plateau SDS-polyacrylamide gel, and transferred onto nitrocellulose. Immunoblotting was performed with the appropriate monoclonal antibody, and quantified by densitometry. TABLE 111. Percentages of Polymerized Tubulin in K562 and MES-SA Cells* K562 Total P-tubulin Class IV 13-tubulin MES-SA Dividing Resting Dividing Resting 40kl 35 k 4 30k3 30 f 6 36kl 39 ? 6 27k4 32 k 3 *Total and polymerized P-tubulin and class IV P-tubulin were evaluated by immunoblotting in dividing and resting K562 and MES-SA cells. Values shown are the average (2 SD) of the percentages of polymerizedkotal tubulin. Tubulin Isotypes and the Cell Cycle types has shown the class I isotype to be most abundant [Lopata and Cleveland, 1987; Sisodia et al., 19901, with the class IV isotype representing 10-30% of all P-tubulin. In our cell lines, PCR data also suggest that class I is the most abundant, with class IVb representing a smaller contingent and class IVa being a minor fraction. The fact that class IV accounts for only a fraction of total tubulin explains why the total tubulidtotal protein ratio is not different in dividing and resting cells, in spite of the overexpression of class IV tubulin protein. Of note is the fact that the “isotypic profile” of 01 and P-tubulins differs among cells, with some of the isotypes not being expressed at all in a given cell type. y-Tubulin, reported to be associated with the centrosome, was found to be expressed at constant levels throughout the cell cycle, a finding consistent with the fact that centrosome function is independent of the cell cycle [Tournier and Bornens, 19941. It is remarkable that the two isotypes which are modulated during the cell cycle share the same C terminus and are both considered as belonging to class IV in the Sullivan classification [Sullivan, 19881. These two isotypes differ only by ten amino acid residues, spaced out along the amino acid sequence. This functional redundancy, as well as the strong conservation of these two isotypes during evolution, suggests that this class plays an important role in cell physiology. However, although class IVb has been reported to be present at low levels in a number of tissues, the class IVa isotype has previously been reported to be present in neuronal tissues only [Lee et al., 19841. The fact that class IVa isotype transcripts have not yet been described in other tissues may be due to the fact that these transcripts are present in very small amounts, as evidenced by the number of cycles of PCR required to display specific products, and may not have been detected by traditional RNA analysis methods such as Northern blotting. The coordinate regulation of these two redundant isotypes sharing the same C-terminal portion suggests a common transcriptional regulatory mechanism. The simultaneous variations in both isotypes may be interpreted either as a redundant way for the cells to ensure the presence of sufficient amounts of class IV P-tubulin protein, or may be due to discrete functional differences between these two isotypes which we have yet to identify. In conclusion, we report that cells undergoing mitosis contain larger amounts of total tubulin, a larger fraction of polymerized tubulin, and a selectively increased content of class IV P-tubulin isotypes. Our results suggest that these cell cycle-specific phenomena are not tissue-specific. These data support the hypothesis of functional specificity of P-tubulin isotypes, as has been suggested by the strong conservation of these isotypes during evolution. 57 ACKNOWLEDGMENTS This work was supported by American Cancer Society grant DHP-76, Department of the Army grant DAMD 17-94-5-4352, NIH grant R01 CA 68217 (B.I.S.), 1’Association pour la Recherche sur le Cancer, the Ligue Contre le Cancer de la Drdme, and the Philippe Foundation, Inc. (C.D.). We thank S.A. Lewis and N. Cowan for providing vectors and helpful advice and information concerning tubulin isotypes. REFERENCES Banerjee, A., Roach, M.C., Trcka, P. and Luduena, R.F. (1992): Preparation of a monoclonal antibody specific for the class IV isotype of @-tubulin.Purification and assembly of a@ 11, a@ 111, and a@ IV tubulin dimers from bovine brain. J. Biol. Chem. 267:5625-5630. Borisy, G.G. and Taylor, E.W. (1967): The mechanism of action of colchicine. Colchicine binding to sea urchin eggs and the rnitotic apparatus. J. Cell Biol. 34:535-548. Chen, G . , Zaffrezou, J.P., Fleming, W.H., Duran, G.E. and Sikic, B.I. (1994): Prevalence of mukidrug resistance related to activation of the mdrl gene in human sarcoma mutants derived by single-step doxorubicin selection. Cancer Res. 54:4980-4987. Cleveland, D.W., Lopata, M.A., MacDonald, R.J., Cowan, N.J., Rutter, W.J. and Kirschner, M.W. (1980): Number and evolutionary conservation of a-and P-tubulin and cytoplasmic @and y genes using specific cloned cDNA probes. Cell 20:95105. Cowan, N.J., Dobner, P.R., Fuchs, E.V., and Cleveland, D.W. (1983): Expression of human a-tubulin genes: interspecies conservation of 3’ untranslated regions. Mol. Cell. Biol. 3:17381745. Cowan, N.J., Lewis, S.A., Sarkar, S . , and Gu, W. (1986): Functional versatility of mammalian P-tubulin isotypes. In Maccioni, R.B., and Arechaga, J. (eds.): “The Cytoskeleton in Cell Differentiation and Development. ” IRL Press, Oxford, England, pp. 157-166. Dobner, P.R., Kislauskis, E., Wentworth, B.M., and Villa- Komaroff, L. (1987): Alternative 5’ exons either provide or deny an initiator methionine codon to the same a-tubulin coding region. Nucleic Acids Res. 15:199-218. Dye, R.B., Fink, S.P. and Williams, R.C., Jr. (1993): Taxol-induced flexibility of microtubules and its reversal by MAP-2 and Tau. J. Biol. Chem. 268:6847-6850. Gard, D.L., and Kirschner, M.W. (1985): A polymer-dependent increase in phosphorylation of @-tubulinaccompanies differentiation of a mouse neuroblastoma cell line. J Cell Biol 100:764774. Hall, J.L., Dudley, L., Dobner, P.R., Lewis, S.A., and Cowan, N.J. (1983): Identification of two human P-tubulin isotypes. Mol. Cell. Biol. 3354-862. Harker, W.G. and Sikic, B.I. (1985): Multidrug (pleiotropic) resistance in doxorubicin-selected variants of the human sarcoma cell line MES-SA. Cancer Res. 45:4091-4096. Jaffrezou, J.P., Dumontet, C., Derry , W.B., Duran, G.E., Chen, G . , Tsuchiya, E., Wilson, L., Jordan, M.A., and Sikic, B.I. (1995): Novel mechanism of resistance to paclitaxel (Taxol@) in human leukemia K562 cells by combined selection with PSC 833. Oncol. Res. 7:517-527. Jordan, M. A , , Toso, R.J., Thrower, D. and Wilson, L. (1 993): Mechanism of mitotic block and inhibition of cell proliferation by 58 Dumontet et al. taxol at low concentrations. Proc. Natl. Acad. Sci. U.S.A. 9019552-9556, Joshi, H.C., and Cleveland, D.W. (1989): Differential utilization of (3-tubulin isotypes in differentiating neurites. J. Cell Biol. 109: 663-673. Kemphues, K.J., Raff, E.C., Raff, R.A., and Kaufman, T.C. (1980): Mutation in a testis-specific P-tubulin in Drosophila: analysis of its effects on meiosis and map location of the gene. Cell 2 11445-45 1. Kirschner, M.W. (1978): Microtubule assembly and nucleation. Int. Rev. Cytol. 54:l-71. Lee, G . , and Rook, S.L. (1992): Expression of tau protein in nonneuronal cells: microtubule binding and stabilization. J. Cell Sci. 102:227-237. Lee, M.G., Loomis, C., and Cowan, N.J. (1984): Sequence of an expressed human P-tubulin gene containing ten Alu family members. Nucleic Acids Res. 12:5823-5836. Lewis, S.A., Gilmartin, M.E., Hall, J.L. and Cowan, N.J. (1985): Three expressed sequences within the human P-tubulin multigene family each define a distinct isotype. J. Mol. Biol. 1 8 2 11-20. Lewis, S.A., Lee, M.G., and Cowan, N.J. (1985): Five mouse tubulin isotypes and their regulated expression during development. J. Cell Biol. 101:852-861. Lewis, S.A., Gu, W., and Cowan, N.J. (1987): Free intermingling of mammalian beta-tubulin isotypes among functionally distinct microtubules. Cell 49:539-548. Littauer, U.Z., Giveon, D., Thierauf, M., Ginzburg, I., and Ponstingl, H. (1986): Common and distinct tubulin binding sites for microtubule-associated proteins. Proc. Natl. Acad. Sci. U.S.A. 8317 162-7166. Little, M., and Seehaus, T. (1988): Comparative analysis of tubulin sequences. Comp. Biochem. Physiol. A 90:655-670. Lopata, M.A., and Cleveland, D.W. (1987): In vivo microtubules are copolymers of available P-tubulin isotypes: localization of each of six vertebrate P-tubulin isotypes using polyclonal antibodies elicited by synthetic peptide antigens. J. Cell Biol. 105:17071720. Lowly, O.H., Rosebrough, N.J., Farr, A.L., and Randall, R.J. (1951): Protein measurement with Folin phenol reagent. J. Biol. Chem. 193:265-275. Lu, Q . , and Luduena, R.F. (1993): Removal of P 111 isotype enhances taxol induced microtubule assembly. Cell Struct. Funct. 18: 173-1 82. Luduena, R.F. (1993): Are tubulin isotypes functionally significant? Mol. Cell. Biol. 4:445-457. Luduena, R.F., Shooter, E.M., and Wilson, L. (1977): Structure of the tubulin dimer. J. Biol. Chem. 252:7006-7014. McIntosh, J.R. (1979) Cell division. In Hyams, J.S. (ed.): “Microtubules.” London: Academic Press, pp. 381-441. Minotti, A.M., Barlow, S.B., and Cabral, F. (1991): Resistance to antimitotic drugs in Chinese hamster ovary cells co with changes in the level of polymerized tubulin. J. Biol. Chem. 266:3987-3994. Oakley, C.E., and Oakley, B.R. (1989): Identification of gamma- tubulin, a new member of the tubulin superfamily encoded by mipA gene of Aspergillus nidulans. Nature 338:662-664. Oblinger, M.M., and Kost, S.A. (1994): Coordinate regulation of tubulin and microtubule associated protein genes during development of hamster brain. Brain Res. Dev. Brain Res. 77:4554. Ookata, K., Hisanaga, S., Okumura, E., and Kishimoto, T. (1993): Association of p34cdc2lcyclin B complex with microtubules in starfish oocytes. J. Cell Sci. 105:873-881. Panda, D., Miller, H.P., Banerjee, A., Luduena, R.F., and Wilson, L. (1994): Microtubule dynamics in vitro are regulated by the tubulin isotype composition. Proc. Natl. Acad. Sci. U.S.A. 91:11358-11362. Paschal, B.M., Obar, R.A., and Vallee, R.B. (1989): Interaction of brain cytoplasmic dynein and MAP2 with a common sequence at the C terminus of tubulin. Nature 342:569-572. Raff, E.C. (1994) The role of multiple tubulin isoforms in cellular microtubule function. In Hyams, J.S., and Lloyd, C.W. (eds.): “Microtubules.” New York: Wiley-Liss, Inc., pp. 85-109. Sawada, T., and Cabral, F. (1989): Expression and function of P-tubulin isotypes in Chinese hamster ovary cells. J. Biol. Chem. 264:3013-3020. Saxton, W.M., Stemple, D.L., Leslie, R.J., Salmon, E.D., Zavortink, M., and McIntosh, J.R. (1984): Tubulin dynamics in cultured mammalian cells. J. Cell Biol. 99:2175-2186. Serrano, L., Montejo de Garcini, E., Hernandez, M.A., and Avila, J. (1985): Localization of the tubulin binding site for tau protein. Eur. J. Biochem. 153595-600. Sisodia, S . S . , Gay, D.A., and Cleveland, D. W. (1990): In vivo discrimination among P-tubulin isotypes: selective degradation of a type IV P-tubulin isotype following overexpression in cultured animal cells. New Biol. 2:66-76. Sullivan, K.F. (1988): Structure and utilization of tubulin isotypes. Annu. Rev. Cell Biol. 4:687-716. Tournier, F., and Bornens, M. (1994): Cell cycle regulation of centrosome function. In Hyams, J.S., and Lloyd, C.W. (eds.): “Microtubules.” New York Wiley-Liss, pp. 303-324. Villasante, A., Wang, D., Dobner, P., Dolph, P., Lewis, S.A., and Cowan, N.J. (1986): Six mouse a-tubulin mRNAs encode five distinct isotypes: testis-specific expression of two sister genes. Mol. Cell. Biol. 6:2409-2419. Wilde, C.D., Crowther, C.E., Cripe, T.P., Gwo-Shu Lee, M., and Cowan, N.J. (1982): Evidence that a human P-tubulin pseudogene is derived from its corresponding mRNA. Nature 297183-84. Wilson, L., and Meza, I. (1973): The mechanism of action of colchicine. Colchicine binding properties of sea urchin sperm tail outer doublet tubulin. J. Cell Biol. 58:709-719. Wilson, L., and Jordan, M.A. (1995): Microtubule dynamics: taking aim at a moving target. Chem. Biol. Interact. 2:569-573. Wordeman, L., and Mitchison, T.J. (1994): Dynamics of microtubule assembly in vivo. In Hyams, J.S., and Lloyd, C.W. (eds.): “Microtubules.” New York: Wiley-Liss, pp. 287-301. Zheng, Y . , Jung, M.K., and Oakley, B.R. (1991): Gamma-tubulin is present in Drosophila melanogaster and Homo sapiens and is associated with the centrosome. Cell 65:817-823.