If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Vakgroep Moleculaire Genetica and Departement Plantengenetica, Vlaams Interuniversitair Instituut voor Biotechnologie (VIB), Universiteit Gent, B-9000 Gent, Belgium
Vakgroep Moleculaire Genetica and Departement Plantengenetica, Vlaams Interuniversitair Instituut voor Biotechnologie (VIB), Universiteit Gent, B-9000 Gent, Belgium
Vakgroep Moleculaire Genetica and Departement Plantengenetica, Vlaams Interuniversitair Instituut voor Biotechnologie (VIB), Universiteit Gent, B-9000 Gent, Belgium
Vakgroep Moleculaire Genetica and Departement Plantengenetica, Vlaams Interuniversitair Instituut voor Biotechnologie (VIB), Universiteit Gent, B-9000 Gent, Belgium
Vakgroep Moleculaire Genetica and Departement Plantengenetica, Vlaams Interuniversitair Instituut voor Biotechnologie (VIB), Universiteit Gent, B-9000 Gent, Belgium
Vakgroep Moleculaire Genetica and Departement Plantengenetica, Vlaams Interuniversitair Instituut voor Biotechnologie (VIB), Universiteit Gent, B-9000 Gent, Belgium
Vakgroep Moleculaire Genetica and Departement Plantengenetica, Vlaams Interuniversitair Instituut voor Biotechnologie (VIB), Universiteit Gent, B-9000 Gent, Belgium
Vakgroep Moleculaire Genetica and Departement Plantengenetica, Vlaams Interuniversitair Instituut voor Biotechnologie (VIB), Universiteit Gent, B-9000 Gent, Belgium
To whom correspondence should be addressed: Vakgroep Moleculaire Genetica, Universiteit Gent, K.L. Ledeganckstraat 35, B-9000 Gent, Belgium. Tel.: 32-9-264-5170; Fax: 32-9-264-5349; [email protected]
Affiliations
Vakgroep Moleculaire Genetica and Departement Plantengenetica, Vlaams Interuniversitair Instituut voor Biotechnologie (VIB), Universiteit Gent, B-9000 Gent, Belgium
Vakgroep Moleculaire Genetica and Departement Plantengenetica, Vlaams Interuniversitair Instituut voor Biotechnologie (VIB), Universiteit Gent, B-9000 Gent, Belgium
‡ These authors contributed equally to this work. § Present address: Istituto Miglioramento Genetico Vegetale, Università di Perugia, 74 Borgo XX Giugno, I-06121 Perugia, Italy. Recipient of a predoctoral fellowship of the Ministry of University and Scientific and Technological Research (Italy). ¶ Recipient of a predoctoral fellowship of the Vlaams Instituut voor de Bevordering van het Wetenschappelijk-Technologisch Onderzoek in de Industrie. ‖ Present address: Laboratoire de Physiologie et Biologie Moléculaire des Plantes, Université UMR CNRS 5565, 52, Avenue de Villeneuve, F-66860 Perpignan, France. Recipient of a postdoctoral fellowship of the European Union for Research Training Project Grant ERBFMBICT961274. * This work was supported by the Interuniversity Poles of Attraction Program (Belgian State, Prime Minister's Office, Federal Office for Scientific, Technical and Cultural Affairs; P4/15), the International Human Frontier Science Program (IHFSP RG-434/94M), and the Fund for Scientific Research (Flanders) (G.0121.96).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ** Recipient of a postdoctoral fellowship from the Research Fund of the Ghent University (Belgium). §§ Present address: CropDesign N.V., Technologiepark 3, B-9052 Zwijnaarde, Belgium.
Cyclin-dependent kinases (CDKs) control the key transitions in the eukaryotic cell cycle. All the CDKs known to control G2/M progression in yeast and animals are distinguished by the characteristic PSTAIRE motif in their cyclin-binding domain and are closely related. Higher plants contain in addition a number of more divergent non-PSTAIRE CDKs with still obscure functions. We show that a plant-specific type of non-PSTAIRE CDKs is involved in the control of the G2/M progression. In synchronized tobacco BY-2 cells, the corresponding protein, accumulated in a cell cycle-regulated fashion, peaking at the G2/M transition. The associated histone H1 kinase activity reached a maximum in mitosis and required a yet unidentified subunit to be fully active. Down-regulation of the associated kinase activity in transgenic tobacco plants using a dominant-negative mutation delayed G2/M transition. These results provide the first evidence that non-PSTAIRE CDKs are involved in the control of the G2/M progression in plants.
CDK(s)
cyclin-dependent kinase(s)
PAGE
polyacrylamide gel electrophoresis
Progression through the major transitions of the eukaryotic cell cycle is driven by a family of serine/threonine kinases known as cyclin-dependent kinases (CDKs).1 The catalytic activity of these protein kinases is regulated by the association with their regulatory subunits, cyclins. The activity of the complexes is further controlled by a number of mechanisms including phosphorylation/dephosphorylation, interaction with inhibitory proteins, proteolysis, and intracellular trafficking (
). Of the five mammalian CDKs strongly implicated in cell cycle control, three (CDC2/CDK1, CDK2, and CDK3) are closely related to the prototypical yeast cdc2 and have the same characteristic PSTAIRE motif in the cyclin-binding domain (
). The other two CDKs, CDK4 and CDK6, form a distinct subfamily of CDKs in which PSTAIRE is substituted with either PISTVRE or PLSTIRE, respectively. Both CDK4 and CDK6 are known to function exclusively in the G1 phase (
). The best characterized group (A-type) comprises plant CDKs that are most closely related to the mammalian CDC2 and CDK2 and that contain the same PSTAIRE motif. A-type CDKs can partially complement yeast cdc2/CDC28 mutations and are therefore supposed to be functional homologs of the yeast CDKs (
). Supporting this notion, a dominant-negative mutant of an A-type CDK from Arabidopsis thaliana was found to affect negatively cell cycle progression, most probably both at the G1/S and G2/M transitions (
Currently, functions of plant CDKs other than A-type remain essentially not understood. Here, we studied the function of a plant-specific subfamily of CDKs (B-type; Ref.
). B-type CDKs form a group of closely related kinases from diverse plant species, sharing ∼70–80% identity and comprising a distinct cluster of CDKs. Instead of PSTAIRE, all of them bear unique motifs, either PPTALRE (B1 group) or PPTTLRE (B2 group) (
). Unlike typical CDKs, the expression of B-type CDKs is under strict cell cycle control, and attempts to complement yeast CDK-deficient mutants have been unsuccessful.
We have chosen tobacco as the experimental system because of the high degree of synchronization attainable in the tobacco Bright Yellow-2 (BY-2) cell line (
). The close similarity among B-type CDKs across species has allowed us to use polyclonal antibodies against CDC2bAt from A. thaliana to identify and characterize a B-type CDK in tobacco. In synchronized BY-2 cells, the accumulation of the protein and the related kinase activity is cell cycle regulated and is linked to mitosis. We further demonstrate that the specific activity is present in the form of high molecular weight complexes, whereas the fraction corresponding to the monomeric protein is inactive, which implies the presence of a yet unidentified activating subunit. To assess the function of B-type CDKs in vivo, we have expressed a dominant-negative mutant of the CDC2bAt in transgenic tobacco. These plants were found to have an increased fraction of 4C cells. Accordingly, we propose that B-type plant CDKs play a unique role in the G2/M progression.
EXPERIMENTAL PROCEDURES
Plasmid Construction and Plant Transformation
The cDNA for CDC2bAt-D161N was obtained by mutating the codon GAT (Asp-161) for AAT (Asn-161) by in vitro mutagenesis (
). Both CDC2bAt and CDC2bAt-D161N cDNAs were fused to the Triple-Op promoter, which is a derivative of the cauliflower mosaic virus 35S promoter and has an activity comparable to that of the cauliflower mosaic virus35S promoter (
). The resulting plasmids were transferred intoAgrobacterium tumefaciens C58C1RifR by conjugation. Tobacco plants were transformed by leaf disc transformation (
) and 15 independent T0 primary transformants were analyzed by protein gel blotting. The T0 plants were self-fertilized and the segregating T1 first generation was analyzed further.
Maintenance of the Cell Suspension and Synchronization
The tobacco BY-2 (Nicotiana tabacum L. cv. Bright Yellow-2) suspension was maintained at a weekly dilution (1.8/100) of cells in fresh Murashige and Skoog medium modified according to Nagata et al. (
). A stationary culture was diluted 1/5 in fresh medium supplemented with 4 μg/ml aphidicolin (Sigma). After 24 h of culture, the drug was removed by extensive washes, and the cells were resuspended into fresh medium. DNA synthesis was determined by pulse-labeling with [3H]thymidine as described (
). Mitotic index was determined by ultraviolet light microscopic analysis of 500 cells stained with 0.1 μg/ml 4′,6-diamino-2-phenylindole (Sigma) in the presence of 0.2% Triton X-100.
Purification and Flow Cytometrical Analysis of Nuclei
Tobacco BY-2 protoplasts were obtained from 5 × 104 cells by digestion with 1 ml of enzyme solution (2% cellulase Onozuka R10, 0.1% pectolyase (Kikkoman Co.), 0.66m sorbitol) for 1 h at 37 °C. After treatment, the protoplasts were pelleted by centrifugation (5 min, 1000 ×g), washed once with Murashige and Skoog medium supplemented with 4.5% mannitol. Nuclei were released from the protoplast pellet in Galbraith's extraction buffer (
). After addition of 1% formaldehyde, nuclei were stored at 4 °C until analysis by flow cytometry. Before analysis, nuclei were filtered through a 10-μm nylon filter, treated with RNase A, and stained with propidium iodide (50 μg/ml). Cytometrical analyses were performed on 104nuclei with a fluorescence-activated cell sorter scan flow cytometer (Becton Dickinson). For flow cytometrical analysis of nuclear DNA content in plant tissues, cotyledons or individual calli were chopped with a razor blade in Galbraith's buffer and analyzed as described above.
Antibody Preparation, Immunoblotting, and Immunoprecipitations
Polyclonal antibodies against CDC2aNt (courtesy of P. John, Australian National University, Canberra, Australia) and CDC2bAt were raised with the peptides ARNALEHEYFKDIGYVP and SAKTALDHPYFDSLDKSQF derived from the C termini of the respective proteins. The sera were purified with protein A-Sepharose (Amersham Pharmacia Biotech). Because tobacco possesses multiple nearly identical CDKs that share the same C-terminal peptides for both A-type (accession numbers L77082, L77083, D50738, and AF289467) and B1-type (accession numbers AF289465 and AF289466) CDKs, the data obtained with the sera are cumulative.
Protein extracts from BY-2 cells were prepared by grinding cells with sea sand in homogenization buffer as described (
). Tobacco plants were ground in liquid nitrogen. Protein concentrations were determined with the Protein Assay kit (Bio-Rad, Hercules, CA). SDS-PAGE and protein gel blots were performed according to standard procedures with primary anti-CDC2aNt and anti-CDC2bAt antibodies diluted 1/500 and 1/2500, respectively, and a secondary peroxidase-conjugated antibody (Amersham Pharmacia Biotech) diluted 1/5000.
For immunoprecipitation experiments, 100–200 μg of protein extracts were preincubated for 1 h at 4 °C on a rotating platform with 25% (v/v) protein A-Sepharose (Amersham Pharmacia Biotech). After centrifugation, equal amounts of the supernatant were incubated for 4 h with purified antibodies (1/50 diluted anti-CDC2aNt or 1/50 diluted anti-CDC2bAt) and subsequently for 1 h with 25% (v/v) protein A-Sepharose beads. Beads were washed three times with homogenization buffer and a fourth time with kinase buffer (
). The histone H1 kinase assay was carried out by incubating 25 μl of the packed beads with 5 μCi [γ-32P]ATP (3000 Ci/mmol) in the presence of 1 mg/ml histone H1 (Sigma), 50 mm Tris-HCl (pH 7.8), 15 mm MgCl2, 5 mm EGTA, 10 μm ATP, and 1 mm dithiothreitol, for 20 min at 30 °C in a final volume of 35 μl. Samples were analyzed on a 12% SDS-PAGE gel stained with Coomassie Brilliant Blue and autoradiographed.
For competition experiments the C-terminal CDC2bAt peptide was purified to homogeneity by reverse-phase chromatography on a PRLP-S column (8 μm, 300 Å) (Polymer Laboratories) equilibrated in 0.1% trifluoroacetic acid, and eluted with a 0–70% acetonitrile gradient in 0.1% trifluoroacetic acid. The competition was achieved by preincubation of the antiserum with the purified CDC2bAt peptide at 4 °C overnight before the antiserum was used for immunoblotting and immunoprecipitation.
Biochemical Separation and Analysis of Active Kinase Complexes
Proliferating tobacco BY-2 cell suspension culture cells were collected three days after subculturing. Proteins were extracted in homogenization buffer as described previously (
) and bound to a Q-ceramic HyperD column (10 × 25 cm; Biosepra), equilibrated with homogenization buffer. Proteins were eluted by a one-step elution with 1 m NaCl in 0.5× homogenization buffer and further fractionated by size on a gel filtration column (Superdex 200 (Amersham Pharmacia Biotech) 1.7 × 100 cm Omnifit column). The column was equilibrated with Tris buffer (50 mm, pH 7.8) containing 15 mm MgCl2, 5 mm EGTA, 5 mm β-glycerophosphate, 1 mm NaF, 1 mm dithiothreitol, 0.1 mmNaVO4, 100 mm NaCl, and protease inhibitors (Roche Diagnostics, Brussels, Belgium). Kinase activity associated with either A- or B-type CDKs was isolated by immunoprecipitation with anti-CDC2aNt and anti-CDC2bAt antisera, respectively. The activity of 100-μl aliquots of 5-ml fractions was assayed in vitro in the presence of histone H1 (1 mg/ml), cAMP-dependent kinase inhibitor, 15 μm ATP, and 10 μCi [γ-32P]ATP (3000 Ci/mmol) in a final volume of 35 μl at 30 °C. The reactions were terminated after 20 min by heating the samples in sample buffer at 95 °C for 10 min. The proteins were separated on a 15% SDS-PAGE gel and stained with Coomassie Brilliant Blue, and the incorporated radioactivity was measured with a PhosphorImager (Amersham Pharmacia Biotech).
RESULTS
Identification of B-type CDK Activity in Tobacco
To study the function of B-type CDKs, we first analyzed the associated kinase activity in the course of the cell cycle in the tobacco BY-2 cell suspension (
). Because B-type CDKs had not been described in tobacco, we decided to use an antibody raised against the C-terminal SAKTALDHPYFDSLDKSQF peptide of CDC2bAt of Arabidopsis. This region is well conserved among the known PPTALRE kinases with 17 of 19 amino acids being identical. Thus, we expected the anti-CDC2bAt antibody to recognize tobacco orthologs. Indeed, the antibody recognized specifically a single protein in a crude protein extract of BY-2 cells fractionated by SDS-PAGE (Fig.1a). The apparent molecular mass of the protein, 34.5 kDa, is close to that of theArabidopsis CDC2b. Furthermore, the interaction was specific, as the addition of the peptide used to raise the antibody obliterated the signals corresponding to the protein (Fig.1b). We also analyzed by protein gel blotting with anti-CDC2bAt antibody immunoprecipitates obtained with either an anti-CDC2bAt or an anti-PSTAIRE antibody and found that the 34.5-kDa protein was easily detectable in the former but not the latter precipitate (data not shown). These results suggested that tobacco possesses a CDC2bAt homolog, referred to as CDC2bNt hereafter. Indeed, the sequences of two nearly identical (99.7%) tobacco B-type CDKs have been deposited into public data bases (accession numbersAF289465 and AF289466) with the identical C-terminal peptide SAKAALDHPYFDSLDKSQF, which differs from Arabidopsis CDC2b only in one position (Fig. 1c).
Figure 1The interaction of antibodies against the C-terminal peptide of CDC2bAt with a 34.5-kDa protein in tobacco cells.a, the total protein extracts fromArabidopsis suspension and tobacco BY-2 cells (lanes 1 and 2, respectively) were used for protein gel blot with anti-CDC2bAt antibodies. b, the total protein extract from BY-2 cells was probed either with anti-CDC2bAt antibodies or with the same antibodies preincubated with the C-terminal peptide (lanes 1 and 2, respectively). c,comparison of the C-terminal peptides of CDC2bAt (accession numbersD10851) and two B-type CDKs from tobacco (accession numbersAF289465 and AF289466).
Characteristically, the protein level of CDC2bNt fluctuated through the cell cycle (CDC2b; Fig. 2b), thus displaying the most conspicuous feature of B-type CDKs. The protein was hardly detectable in early S phase, started to accumulate in late S phase, reached the maximal level at the G2/M transition, and declined afterward. After immunoprecipitation with anti-CDC2bAt serum, the kinase activity associated with CDC2bNt was measured by in vitro phosphorylation of histone H1. In the course of the cell cycle, CDC2bNt kinase activity peaked later than the corresponding protein level with a maximum in the middle of mitosis (Fig. 2b). For comparison, we analyzed also the CDC2aNt protein and associated kinase activity using a specific antibody. The level of CDC2aNt protein was constant during the whole cell cycle (Fig.2b). Its activity rose in early S phase and declined during mitosis, slightly earlier than that of CDC2bNt (Fig.2b).
Figure 2CDC2aNt and CDC2bNt protein levels and histone H1 kinase activity in synchronized BY-2 cell suspensions.BY-2 cells were synchronized with aphidicolin (see “Experimental Procedures”) and samples were taken every hour. a, cell cycle progression was monitored by flow cytometry and by measurement of the mitotic index (MI). b, crude protein extracts were used for protein gel blot with anti-CDC2aNt (CDC2a) and anti-CDC2bAt (CDC2b) antibodies or to prepare immunoprecipitates for histone H1 kinase measurements (H1 CDC2a and H1 CDC2b). The duration of the different cell cycle phases, as estimated on the basis of the flow cytometrical analysis and mitotic index, is depicted at the bottom.
Active CDC2bNt Is Present in High Molecular Mass Complexes
To biochemically characterize CDC2bNt, the proteins present in a total extract of actively proliferating BY-2 suspension cells were separated in two fractions by ion exchange chromatography. The bound fraction contained most of the kinase activity (∼95 and 75% for CDC2aNt and CDC2bNt, respectively) and was further fractionated by size exclusion chromatography. The CDK-associated complexes were immunoprecipitated with either anti-CDC2aNt or anti-CDC2bAt polyclonal antibodies and assayed in an in vitro kinase assay with histone H1 as a substrate. The CDC2bNt-associated kinase activity eluted as two minor peaks of ∼250 kDa and more than 700 kDa and one major broad peak in the range of 65 to 100 kDa, referred to as the 80-kDa fraction hereafter (Fig. 3a). In contrast, CDC2aNt activity eluted essentially as a single peak of ∼200 kDa (Fig. 3a). Immunoblotting showed that most of the CDC2bNt protein was present in high molecular mass complexes with a relatively low kinase activity compared with that of the major CDC2bNt activity peak, whereas the majority of the CDC2aNt protein comigrated with the peak of activity (Fig. 3b). The fractions corresponding to monomeric CDC2aNt and CDC2bNt (fractions 20–21; Fig.3) contained detectable, albeit low, amounts of the corresponding proteins and were essentially inactive.
Figure 3Characterization of protein complexes with CDC2aNt and CDC2bNt-associated kinase activity in BY-2 cell suspensions. The total protein extracts were purified by ion exchange chromatography and fractionated by size exclusion chromatography as described under “Experimental Procedures.”a, CDC2bNt and CDC2aNt-associated kinase activity (in relative units) as determined by histone H1 phosphorylation in immunoprecipitated complexes. Position of the size markers is indicated at the top of the panel. b, the level of CDC2aNt (CDC2a) and CDC2bNt proteins (CDC2b) in individual fractions as assessed by immunoblotting with CDC2aNt- and CDC2bAt-specific antibodies. c, specificity of the immunoprecipitated complexes as assayed by peptide competition. The bound fraction prior to size exclusion chromatography was subjected to immunoprecipitation using the anti-CDC2bAt antiserum preincubated with increasing concentrations of the competing peptide followed by histone H1 kinase activity assay (H1 CDC2b) and immunodetection (CDC2b) of the B-type CDKs.
To confirm that the kinase activity in the immunoprecipitated protein complexes was specific for CDC2bNt, we performed peptide competition assays for the bound protein fraction used in the size fractionation experiments above. As can be seen from Fig. 3c, preincubation of the serum with the peptide strongly reduced immunoprecipitated kinase activity. This corresponds well with our results obtained with a dominant-negative mutant of CDC2bAt and also indicated that at least 80% of the kinase activity immunoprecipitated with the same serum from the whole cell tobacco extracts could be attributed to CDC2bNt (see below). Additionally, it should be stressed that the peptide SAKTALDHPYFDSLDKSQF is highly specific for B-type CDKs. All the significant hits retrieved by a BLAST search with the peptide were confined to B-type CDKs, with at least 17 amino acids identical for PPTALRE kinases and only 12 for PPTTLRE kinases.
Gel filtration separation of the flow-through fraction showed that the majority of the CDC2aNt protein was present in the inactive monomeric fractions and accounted for approximately half of the total CDC2aNt protein (data not shown). In contrast, only a very limited amount of the total CDC2bNt protein was found in the flow-through fraction and it eluted after gel filtration as a single peak of ∼200 kDa, co-migrating with the peak of kinase activity (data not shown). Thus, the pool of monomeric CDC2bNt is very small compared with that of CDC2aNt. Our data demonstrate that large amounts of both CDC2aNt and CDC2bNt are kept in inactive forms, albeit by different means.
Expression of the Mutant CDC2bAt-D161N in Tobacco Down-regulates Endogenous CDC2bNt Activity
The residue Asp-161 ofArabidopsis CDC2b belongs to the triad of catalytic residues that are conserved in all eukaryotic protein kinases (Arg-33, Glu-51, and Asp-145 in the human CDK2) and are strictly required for the kinase activity by providing vital contacts for correct ATP orientation and Mg2+ coordination (
) presumably because of the competition of the mutant proteins for the association with the rate-limiting interacting proteins such as cyclins. As it is well known, orthologous CDKs, even from very divergent species, are functionally interchangeable; hence, dominant-negative mutants are efficacious in heterologous systems (
). To see whether kinase-negative mutants of B-type CDKs are dominant-negative as well, we generated transgenic tobacco plants that express the mutantCDC2bAt–D161N under control of a strong constitutive promoter (Fig. 4). We also produced tobacco plants expressing wild-type CDC2bAt under control of the same promoter. For 15 independent lines from each transformation, the level of the protein was analyzed in 2-week-old plantlets in the primary T0 transformants and in the self-fertilized T1 population. The level of CDC2bAt-D161N was consistently lower than that of CDC2bAt (data not shown), suggesting that a higher level of CDC2bAt-D161N is incompatible with plant regeneration and/or development. Two lines from both transformations that segregated the T-DNA as a single locus were selected for further analysis. We compared histone H1 kinase activity of protein complexes immunoprecipitated with CDC2bAt antiserum in the transgenic and parental lines (Fig. 5). Production of the CDC2bAt-D161N protein correlated with an ∼5-fold inhibition of the B-type CDK activity in both DN-1 and DN-27 lines (Fig. 5). In contrast, ectopic production of the wild-type CDC2bAt (WT-14 and WT-23; Fig. 5) did not affect the corresponding kinase activity (in this case cumulative for CDC2bNt and CDC2bAt), even when produced to a much higher level than the mutant forms (WT-23). Effectively, these results mean also that at least 80% of the kinase activity in the immunoprecipitates associates with CDC2bNt and that cross-reacting contaminants, if any, contribute to less than 20% of the total activity.
Figure 4Construction of transgenic tobacco expressing either CDC2bAt or the CDC2bAt-D161Nmutant.a, aligned amino acid sequence of regions of the Arabidopsis (At) CDC2b and CDC2a, human (Hs) CDK2, and S. pombe (Sp) cdc2. The mutated residue in CDC2bAt is indicated. Boxes are drawn around residues conserved in the four kinases. b, schematic diagram of the expression cassettes used for transformation of tobacco SR-1 (see “Experimental Procedures”). CDC2bAt-D161Nmutant cDNA was created by polymerase chain reaction mutagenesis. Both the mutant and wild-type cDNAs were cloned under the control of the Triple-Op (TOP) promoter and introduced into tobacco plants (see “Experimental Procedures”).
Figure 5Transgene expression and histone H1 activity in transgenic tobacco. Seedlings of transgenic plants expressingCDC2bAt (lines WT14 and WT23) orCDC2bAt-D161N (lines DN-1 and DN-27) grown for 2 weeks on selective medium are compared with the SR-1 control of the same age. a, protein gel blot analysis of total proteins with anti-CDC2bAt antibodies (CDC2b) and the kinase activity in protein complexes immunoprecipitated either with anti-CDC2bAt antibodies (CDC2b activity) or with anti-CDC2aNt antibodies (CDC2a activity) (see “Experimental Procedures”). b, relative CDC2b kinase activity estimated by quantification of the radioactively phosphorylated H1 protein.
To see whether CDK activities other than B-type are affected in the transgenics, we analyzed histone H1 kinase activity in protein complexes either immunoprecipitated with CDC2aNt-specific antibodies or purified with p13SUC1 affinity matrix (p13SUC1binds CDC2a but not CDC2b in tobacco).
In neither case, any reliable difference among the lines could be detected (Fig.5a; data not shown).
Production of CDC2bAt-D161N Increases the Number of Cells in G2 Phase
To analyze whether the down-regulation of B-type CDK activity affects cell cycle progression, we compared the nuclear DNA content in segregating T1 plants grown on nonselective medium for two weeks. For each of the lines, 20 plants were chosen randomly. For each plant, half of one cotyledon was transferred to selective medium to identify the siblings that lost the transgene, while the other half was used to regenerate calli on nonselective medium. The regenerated calli and the other cotyledon were subjected to the flow cytometric analysis of nuclear DNA content. The cotyledons and calli were chosen as representatives of terminally differentiated and undifferentiated tissues, respectively. The fractions of 2C (G1 phase) and 4C (G2 phase) cells were determined for individual plants, and the siblings with and without the T-DNA were compared. As shown in Fig.6b, the expression ofCDC2bAt-D161N correlated with an ∼2- and 1.5-fold increase in the 4C/2C ratio in cotyledons and calli, respectively. The ectopic expression of CDC2bAt did not modify the G2/G1 ratio. Thus, these data showed that ectopic expression of CDC2bAt-D161N, but notCDC2bAt, affects cell cycle in transgenic tobacco.
Figure 6Segregation analyses of transgenic plants expressing CDC2bAt orCDC2bAt-D161N. Segregating populations of T1 plants grown on medium without selection were used for the analysis.a, the phenotypes of seedlings (representativelines WT-23 and DN-1) 2 weeks after in vitrogermination. Arrows point to the siblings with no T-DNA. No morphological differences can be observed. b, flow cytometric analysis of transgenic plants. Cotyledons of 2-week-old seedlings were subjected to flow cytometrical analysis (see “Experimental Procedures”). Calli were regenerated from 2-week-old cotyledons and flow cytometrical experiments were performed after 2 weeks. For each line, 20 plants were analyzed individually, and the results are the mean value for the particular lines, calculated separately for the siblings bearing T-DNA, both homo- and hemizygous (filled bars, transgenic lines only) and without T-DNA (open bars, SR-1 included).
Ectopic Production of Either CDC2bAt or CDC2bAt-D161N Does Not Affect Plant Morphology
We further studied whether ectopic production of the CDC2bAt protein and its D161N mutant affects the morphology of plants. T0 seeds were germinated on nonselective medium for two weeks, and the phenotypes of the segregating populations of plants were compared. The plants ectopically expressingCDC2bAt showed no phenotypic alterations (WT-23; Fig.6a). Neither were the plants expressingCDC2bAt-D161N affected, despite a 5-fold reduction of the B-type CDK activity (Fig. 6a, DN-1). Microscopic analyses revealed no cytological modifications in any of the tissues analyzed including epidermal cells in leaves, cotyledons, and all the root cell files or particular modifications in the organization of the shoot and root meristems (the primary sources of all plant cells). Because the ploidy level and morphogenesis of seedlings is known to be under light control (
), we compared also seedlings grown in the dark and could not find any changes either. Furthermore, we did not observe any significant differences in the rate of callus regeneration from the cotyledons (data not shown).
DISCUSSION
We have attempted to elucidate the function of the CDK group distinguished by a PPT(A/T)LRE motif, which is specific for plants (referred to as B-type; Ref.
) belongs to the A-group of plant CDKs (PSTAIRE-type). Here, we show that tobacco possesses a putative homolog of the CDC2bAt kinase, recognized by a CDC2bAt-specific antibody and displaying the expression pattern characteristic of this type of CDKs. As shown before for the alfalfa PPTALRE kinase CDC2MsD (
with partially synchronized Arabidopsis suspension cells, albeit not with the same degree of resolution, confirm that the expression pattern of the tobacco CDK described here is also shared by theArabidopsis homolog.
Along with PPTALRE kinases, plant B-type CDKs include PPTTLRE kinases, of which only alfalfa CDC2MsF has been analyzed so far in synchronized cells through the complete cell cycle. The protein level of this kinase reaches a maximum in mitosis distinctively later than PPTALRE CDKs (
). This difference between PPTALRE and PPTTLRE kinases may reflect a general dichotomy of B-type CDKs.
The oscillation of the CDC2b protein level during cell cycle suggests that it may be a rate-limiting factor for the activity, which is not typical for CDKs. To address this question, we have pursued three different approaches. First, we compared the temporal profiles of the CDC2b protein and the associated CDK activity. The regulation of PPTALRE CDK-associated activity in the course of cell cycle had not been analyzed before, although it was shown that the protein complexes containing a PPTALRE CDK from rice possess histone H1 kinase activityin vitro (
). Here, we demonstrated that the histone H1 kinase activity associated with CDC2bNt peaks in mitosis considerably later than the corresponding protein. This result indicates that posttranslational mechanisms may control the B-type CDK activity. The timing of mitosis is almost universally controlled by the phosphorylation of Tyr-15 in the PSTAIRE CDKs (
). Taking into account that CDC2bNt activation occurs at the end of G2, it would be interesting to see whether this regulation is relevant for B-type CDKs. The conservation of Tyr-15 in these CDKs supports this suggestion.
Second, we partially purified protein complexes that contain B-type CDKs from tobacco BY-2 suspension cells to see whether interaction with other proteins may regulate CDC2b kinase activity. Our results demonstrated that monomeric B-type CDKs (fractions 20–21; Fig. 3) have negligible kinase activity, whereas the highest activity was found in the 70–80-kDa fractions (fractions 17–18; Fig. 3) containing a comparable amount of the CDC2bNt protein. This implies that B-type CDKs require an activating subunit with a molecular mass of ∼40–50 kDa. Even though the composition of the protein complexes is unknown, our results indicate that plant B-type CDKs may depend on additional components (most probably cyclins) for their full kinase activity.
It is worth noting that a high amount of the CDC2b protein in tobacco was sequestered in high molecular mass complexes (particularly 160–200 kDa and ∼530 kDa) almost devoid of histone H1 kinase activity (Fig.3). Recently, Yoshizumi et al. (
) presented evidence thatArabidopsis CDC2b plays a role in regulating seedling growth in the dark. It is tempting to speculate that the sequestering of CDC2bNt in the inactive high molecular mass complexes is due to complete inhibition of the corresponding developmental pathways in the dedifferentiated BY-2 cells. At the same time it is important to stress that histone H1 may not be a physiological substrate of CDC2bNt. The high molecular mass complexes we have detected, while being inactive toward histone H1, may nevertheless possess high activity toward the natural substrates still to be identified.
Finally, we overexpressed CDC2bAt in transgenic tobacco under control of a strongly constitutive promoter to see whether the cell cycle-modulated production of B-type CDKs is fundamental for the regulation of the associated activity. In none of the transgenic lines could a significant increase in B-type CDK activity be found despite the high protein level. This result clearly demonstrates that, like for most other CDKs, the protein level of B-type CDKs is not a limiting factor for their activity. Correspondingly, plants overproducing CDC2bAt show no discernible phenotype. Although the results of CDC2bAt overproduction might be simply accounted for by assuming that CDC2bAt does not interact with tobacco cyclins, the dominant-negative effect of the CDC2bAt-D161N mutant argues strongly against this interpretation (see below).
The above results negate the major argument that has been recurrently invoked to substantiate a cell cycle function of plant B-type CDKs (i.e. the cycling expression) and leaves us with the question whether they are involved in cell cycle control at all. To approach it, we resorted to kinase null mutants that had proved to be a useful tool to dissect the function of CDKs (
). Kinase-inactive mutants of CDKs have a dominant-negative effect, presumably because of the competition of the mutant proteins for the association with the rate-limiting interacting proteins, such as cyclins. We have generated a number of transgenic lines expressing the kinase null mutant CDC2bAt-D161N to analyze its impact on the phenotypes and cell cycle progression. We have found that in the lines with highest expression levels, the CDC2bNt-associated kinase activity is reduced approximately by a factor of 5. Importantly, the effect was specific for CDC2bNt-associated activity and exerted only by the mutant, but not the wild-type, protein. This result indicates that the mutant protein can titrate out a rate-limiting protein required for the CDC2bNt activity, whereas the complexes formed by wild-type CDC2bAt are fully active in tobacco cells. These results strongly augment the above biochemical evidence for an activating subunit of B-type CDKs.
The relatively high residual B-type CDK activity in theCDC2bAt-D161N transgenic plants suggests that lines with higher expression levels have been counterselected in the process of transformation, indicating that B-type CDKs may have an essential function. To get insight into that function, we analyzed these transgenic plants at the level of morphology, cytology, and cell cycle progression. While morphologically and cytologically no difference could be observed, the transgenic lines with a decreased level of CDC2b activity showed a considerable increase in the size of the G2 cell population.
Depending on the type of cells analyzed, our results have different implications. In the rapidly dividing callus cells, the increase in the G2 cell population means that the progression through and/or the exit from G2 was compromised. In case of cotyledons, composed mainly of nondividing cells, the higher proportion of G2 cells indicates that more cells exited the cell cycle in the G2 phase in the course of embryogenesis. The apparently weaker impact of the mutation in callus than in cotyledons (∼50% versus 100% G2 increase) is probably a consequence of the much higher representation of G2 cells in callus than in cotyledons (∼4- to 5-fold). Given that a similar effect has been observed in tissues as different as cotyledons and callus, we expect this observation to be rather general for any tissue.
We believe that the observed changes in the cell cycle progression are attributable to the down-regulation of CDC2bNt activity and that this activity is involved in G2/M progression. The molecular mechanism of CDC2b action in the cell cycle is still to be elucidated. It is difficult to completely exclude at present the possibility that the mutant protein perturbs the cell cycle indirectly, for example by interfering with the function of an important regulator of the G2/M progression unrelated to CDC2b. However, this is a very improbable scenario, because no changes were observed in the case of overproduction of the wild-type protein; this means that even if theArabidopsis protein (no matter wild-type or mutant) is able to titrate out an essential regulatory molecule in tobacco cells, the complexes formed by CDC2bAt with that molecule fully substitute for the endogenous counterparts and thus fulfill the same function.
While this work was in progress, Yoshizumi et al. (
) described transgenic Arabidopsis plants with the expression of the CDC2bAt gene inhibited by means of antisense technology. The expression of the antisense construct correlated with short hypocotyl and open cotyledon phenotypes when transgenic seedlings were grown in the dark. We did not observe this type of change in our transgenic tobacco lines. On the other hand, no changes in the DNA content were detected in the CDC2b antisense Arabidopsisplants. One of the probable explanations for the discrepancy is the presence of residual B-type kinase activity in our transgenic lines, which is presumably sufficient to enable normal developmental programs in the dark. However, the comparison between the results of Yoshizumiet al. (
) and ours is compounded by the major differences in experimental set-up. First of all, different technologies were used in the two studies, and it is unknown to what extent CDC2bAt activity was influenced by the production of antisense RNA. Plants are notorious for having multiple paralogous genes, and this seems to be the case for CDC2b both in Arabidopsis and tobacco. The implication is that the antisense technology, being rather DNA sequence-specific, most probably inactivated only one of the paralogs, without altering the activity of the others. In contrast, the use of dominant-negative mutants is expected to affect the activity of all the paralogous proteins. Second, different organs were used to analyze the DNA contents. Finally, the specifics of the plant species used may have also affected the outcome, as it has been observed before by others. The results of the two works may be complementary rather than contradictory. Obviously, many more experiments are required to define more precisely the functions of this plant-specific type of kinases.
) CDC2bAt, CDC2aNt, CDKB1–1 Nt, and CDKB1–2 Nt are designated Arath;CDKB1;1, Nicta;CDKA;1, Nicta;CDKB1;1, and Nicta;CDKB1;2, respectively.
Acknowledgments
We thank Pete John for providing antibodies, Véronique Boudolf for help with the experiments on the dark-grown seedlings, Marcelle Holsters for critical reading of the manuscript, Martine De Cock for help preparing it, and Rebecca Verbanck for help with the figures.
86708-8&id=gr1.jpg){kind=link}
86708-8&id=gr2.jpg){kind=link}
86708-8&id=gr3.jpg){kind=link}
86708-8&id=gr4.jpg){kind=link}
86708-8&id=gr5.jpg){kind=link}
86708-8&id=gr6.jpg){kind=link}