Planta (1998) 206: 215±224
Cell cycle regulation by plant growth regulators: involvement of auxin and cytokinin in the re-entry of Petunia protoplasts into the cell cycle Christophe TreÂhin1, SeÂverine Planchais1, Nathalie Glab1, Claudette Perennes1, James Tregear1, Catherine Bergounioux1 1
Institut de Biotechnologie des Plantes, CNRS ERS 569, Universite de Paris-Sud, BaÃt. 630, Plateau du Moulon, F-91405 Orsay Cedex, France 2 Laboratorium voor Genetica, via Department of Genetics aliated to the Flanders Interuniversity Institute for Biotechnology, Universiteit Gent, K.L. Ledeganckstraat 35, B-9000 Gent, Belgium Received: 2 September 1997 / Accepted: 9 March 1998
Abstract. In order to understand the mode of action of auxins and cytokinins in the induction of cell division, the eects of the two plant growth regulators 2,4dichlorophenoxyacetic acid (2,4-D) and N6-benzyladenine (BA) were investigated using mesophyll protoplasts of Petunia hybrida, cultivated in either complete medium or in medium de®cient in cytokinin, auxin or both. Firstly we studied DNA synthesis, using 5bromodeoxyuridine/bisbenzimide Hoechst/propidium iodide ¯ow cytometry analyses and by the monitoring of histone H4 transcript levels. Roscovitine, a cyclindependent kinase (CDK) inhibitor, was found to block the cell cycle prior to entry into the S and M phases in the cultured P. hybrida protoplasts. This suggests that in Petunia cells there is a requirement for CDK activity in order to complete the G1 and G2 phases. Further experiments using roscovitine showed that neither 2,4-D nor BA were individually able to induce cell cycle progression beyond the roscovitine G1 arrest. We also monitored the phytohormonal induction of S phase by studying variations in transcript levels of the gene for mitogenactivated protein kinase (PMEK1) and transcript levels of the cell division cycle gene cdc2Pet. Only 2,4-D, and not BA, was able to stimulate PMEK1 gene transcription; thus, the more rapid S-phase induction in 2,4-D-treated protoplasts may be attributable to the activation of this transduction pathway. In contrast, both plant growth regulators were required to induce the appearance of cdc2Pet mRNA transcripts prior to S-phase engagement.
Accession number: The cdc2Pet sequence has been placed in the EMBL database under the accession number Y13646 Abbreviations: BA N6-benzyladenine; BrdU 5-bromodeoxyuridine; CDK cyclin-dependent kinase; 2,4-D 2,4-dichlorophenoxyacetic acid; HO bisbenzimide Hoechst; MAPK mitogen-activated protein kinase; PI propidium iodide Correspondence to: C. TreÂhin; E-mail:
[email protected]; Fax: 33-1-69336423
Key words: cdc2Pet ± Cell cycle ± Mitogen-activated protein kinase ± Petunia (cell cycle) ± Plant growth regulator ± Roscovitine
Introduction Current progress in the elucidation of the cascade of intracellular events induced by plant growth regulators can be subdivided into a number of key areas. Using auxin as an example, these areas are: the perception of the signal and the function(s) of auxin-binding proteins, the study of auxin-transport proteins, the electrophysiological consequences of auxin action and, ®nally, molecular approaches to the study of auxin action (Millner 1995). Despite more recent advances in our understanding of these processes, very little information is available on the molecular link between auxin response and cell cycle progression (Takahashi et al. 1995). Studies of the cell cycle response to auxins are complicated because of the heterogeneity of plant cell populations with respect to their physiological state and gene expression activities. Cells often respond dierently at dierent developmental stages, and interactions between the auxin and cytokinin signalling pathways may occur (Shi et al. 1994), thus further complicating the situation. One way to elucidate the mode of action of auxins and cytokinins is to de®ne the steps of the cell cycle which they promote or control. Previously, Bergounioux et al. (1988) showed that mesophyll protoplasts of Petunia hybrida, the majority of which display a 2C DNA content, re-enter the mitotic cell cycle after a few hours when incubated in a medium supplemented with 2,4-dichlorophenoxyacetic acid (2,4-D) and N6-benzyladenine (BA). The presence of the latter two substances is an absolute requirement for re-entry into the cell cycle, implying that they are together involved in one or more events in the cell cycle prior to S phase. This system is thus an ideal experimental model for the study of plant growth regulator control of the cell cycle.
216
To further characterize steps leading to S phase following auxin and/or cytokinin addition, two approaches were used in the work described here. Firstly, we monitored DNA synthesis through incorporation of the thymidine analogue 5-bromodeoxyuridine (BrdU). After BrdU incorporation, the discriminatory multiparametric analysis of Hoechst 33258/versus propidium iodide (PI) allows the monitoring of cells that have progressed from one cell cycle to the next (Glab et al. 1994). Secondly, we monitored levels of accumulation of histone H4 transcripts under various plant growth regulator regimes, since the H4 gene has been shown to be transcriptionally regulated during the cell cycle, its peak occurring in S phase (Shaul et al. 1996). To further characterize molecular events underlying the hormonal signaling pathway, we focused on cyclindependent kinases (CDKs). The latter have attracted considerable interest, since they play an essential role in cell division cycle regulation (Hunter 1995; Jacobs 1995). The control mechanisms associated with the G1/S and G2/M transitions, in which CDKs participate, involve modi®cations of the covalent binding of phosphate groups to the serine, threonine and tyrosine residues of substrate proteins. The use of speci®c inhibitors of CDK activity allows monitoring of the cell cycle G1/S and G2/ M transition checkpoints (for a review on chemical inhibitors of CDKs, see Meijer 1995). For instance, olomoucine was found to inhibit proliferation at both the G1/S and the G2/M boundary in Arabidopsis and P. hybrida cells (Glab et al. 1994; Abraham et al. 1995), suggesting that CDKs complexes act in the control of these two transitions in the plant cell cycle, as has previously been shown for animal systems. However, olomoucine displays a quite narrow selectivity amongst the 35 kinases tested, of which it inhibits only CDC2, CDK2, CDK5 and, to a lesser extent, ERK1 (extracellular regulated protein kinase; Vesely et al. 1994). Roscovitine, a novel purine inhibitor of CDKs which is both highly ecient and selective, is tenfold more active than olomoucine (Meijer et al. 1997). This inhibitor belongs to a family of C2,N6,N9-substituted adenines and interacts with the ATP-binding site of the CDKs (Vesely et al. 1994). Roscovitine was recently shown to inhibit in vitro the kinase activity of puri®ed CDK-cyclin complexes and to block tobacco BY-2 cellsuspension cultures at the G1 and G2 cell cycle checkpoints (Planchais et al. 1997). In the work described here, roscovitine was used to determine the ability of each plant growth regulator to allow progression of Petunia protoplasts beyond the G1 checkpoint, characterized by the roscovitine block. Neither CDK transcripts nor p34cdc2 protein are detectable in either non-cycling dierentiated cells of mature leaves (Segers et al. 1996) or in stem pith (Zhang et al. 1996). Therefore, de-novo transcription of the cell cycle gene cdc2 must be required for dierentiated cells to re-enter the cell cycle. Zhang et al. (1996) have shown that cytokinin is stringently required only in late G2 phase to produce CDK activity and that auxin alone promotes DNA synthesis and the inactivation of CDK. However, in contrast to these results, Sauter et al. (1995) have
C. TreÂhin et al.: Control of cell cycle entry by phytohormones
demonstrated gibberellin induction of cdc2Os1 and cdc2Os2 transcripts in rice internodes. Moreover, cytokinin alone can promote cdc2at promoter-driven expression of the b-glucuronidase gene (gus) in tobacco mesophyll protoplasts (Hemerly et al. 1993). Therefore, the extent to which cdc2Pet gene transcript levels are in¯uenced by these hormonal signals becomes an important issue. Given the key role played by mitogen-activated protein kinases (MAPKs) in animal cell cycle regulation, it is likely that similar mechanisms exist in plant cells. A range of dierent growth factors and mitogens stimulate the activities of MAPKs in animal cells. Mitogenactivated protein kinases contribute to cellular signal transduction pathways by phosphorylating various proteins in dierent cellular compartments (reviewed in Hunter 1995). The identi®cation and characterization of MAPK, MAPK kinase (MAPKK) and MAPK kinase kinase (MAPKKK) homologues in plants (Mizoguchi et al. 1997) has provided strong evidence that the latter also use MAPK signalling modules. In tobacco BY-2 cells which had been treated with 2,4-D for 5 min, a rapid and transient activation of MAPK phosphorylating activity was found to occur (Mizoguchi et al. 1994). At the mRNA level, the Arabidopsis MAPKKK homologue, ATMEKK1, and MAPK homologue, ATMPK3, respond to touch or mechanical pressure (Mizoguchi et al. 1996). However, neither auxin nor cytokinin has previously been shown to stimulate MAPK transcription. We have thus investigated the eects of 2,4-D and BA on the accumulation of transcripts of the P. hybrida MAPK gene, PMEK1 (``PMEK'' stands for Petunia MAPK/ERK-related kinase), previously characterized by Decroocq-Ferrant et al. (1995). The aim of the work described here was to obtain new insights into the involvement of the synthetic plant growth regulator 2,4-D and the phytohormone BA in the commitment of cell-derived protoplasts to their ®rst cell cycle. The phenomenon of cell cycle induction was studied by means of an experimental system based on the use of isolated mesophyll protoplasts requiring an exogenous supply of plant growth regulators for cell division. This provided a useful experimental system with which to de®ne the sequential events of plant growth regulator action. We used the CDK inhibitor, roscovitine, to identify G1/S and G2/M transition checkpoints and then to compare the activities of BA and 2,4-D with respect to progression towards S phase. We also analyzed the expression of the P. hybrida cdc2 and PMEK1 genes with respect to variations in auxin and cytokinin supply. This has allowed us to formulate new hypotheses regarding the involvement of auxin and cytokinin in this experimental system, by the identi®cation of cell cycle transitions which have greater or lesser dependence upon a given plant growth regulator. Materials and methods Protoplast culture, bromodeoxyuridine (BrdU) incorporation. Mesophyll protoplasts were isolated from leaves of Petunia hybrida
C. TreÂhin et al.: Control of cell cycle entry by phytohormones (hybrid F1 P ´ PC6, kindly provided by Dr. Cornu, INRA Dijon, France). An aliquot of 105 protoplasts per ml was cultured in the light at 26 °C as previously described (Glab et al. 1994). Roscovitine (provided by L. Meijer) dissolved at 5 mM in dry dimethyl sulfoxide (DMSO), previously stored at )20 °C, was administered for 24 h to P. hybrida protoplast cultures either immediately after protoplast isolation or after the protoplasts had been cultivated for 24 h. Oryzalin (Dow Elanco, Antwerpen, Belgium) dissolved in dry DMSO was added to 3 lM immediately after protoplast isolation (0 h). Roscovitine-treated cells, washed once by centrifugation, were either used for cell cycle analysis or resuspended in 1/2 volume of conditioned medium obtained from protoplast cultures of the same age, and returned to culture. Controls consisted of freshly isolated protoplasts (0 h) and 24-h or 48-h uninterrupted cultures. A 30 mM aqueous BrdU (Sigma, St. Louis, Mo., USA) stock was added to 30 lM ®nal concentration to the P. hybrida protoplast medium. During BrdU incorporation, protoplasts were cultured in the dark at 26 °C. Isolation of nuclei and cytometric analysis. Nuclei were released from the protoplast pellet in Galbraith buer (Glab et al. 1994) supplemented with 1% (w/v) Triton X-100 (pH 7). After 1 min, the remaining membranes were mechanically disrupted by repeated passages of the suspension through a Pasteur pipette. Finally, 1% formaldehyde was added and the nuclei were stored at 4 °C. The samples were ®ltered through nylon (pore size 30 lm). In simple analysis without BrdU treatment, the P. hybrida nuclei in the ®ltrate were stained directly with 2 lg á ml)1 (®nal concentration) bisbenzimide Hoechst (HO) 33342 (Sigma). Cytometric analysis was performed on 2 ´ 104 nuclei with an EPICS V ¯ow cytometer (Coulter, Margency, France) according to conditions described in Perennes et al. (1993). Histograms were processed with Multicycle (Phoenix Flow Systems, San Diego, Calif., USA). In the case of BrdU experiments, the nuclei were stained for 15 min with 1 lg á ml)1 Hoechst 33258 (Sigma). Finally, 3 lg á ml)1 PI (Sigma) was added for a further 15 min. It was observed that the RNase treatment previously used for PI staining was not required. In the biparametric analysis described by Glab et al. (1994), nuclei were excited with UV (351±364 nm) light and a bivariate cytogram of red (PI > 610 nm) vs. blue (408 nm < HO < 500 nm) ¯uorescence was recorded. Care was taken to eliminate both debris and doublets through light scatter and pulse-shape analysis. Northern blot analysis. Leaves or protoplasts were frozen in liquid nitrogen immediately after harvesting and total RNA was extracted by grinding the material in a mortar in the presence of TRIzol Reagent (GIBCO/BRL, Cergy Pontoise, France) according to the manufacturer's instructions. Samples of RNA were treated for 15 min at 65 °C in loading buer and 0.3 ll ethidium bromide (10 mg á ml)1 stock) was added before loading. After electrophoresis in a 1% agarose gel and blotting onto Hybond N+ membrane
217 (Appligene, Illkirch, France), hybridizations with 32P-labelled probes, corresponding to the coding regions of cdc2Pet, PMEK1 and the Arabidopsis histone H4 genes were performed at 65 °C and 63 °C, respectively, according to Church and Guilbert (1984).
Results The CDK inhibitor roscovitine blocks the cell cycle at the G1/S and G2/M transitions. In P. hybrida, 84% of the protoplasts isolated from the fully expanded sixth and seventh leaves exhibited a 2C nuclear DNA distribution (Fig. 1). As previously found (Bergounioux et al. 1988), less than 3% of nuclei were in S phase and no mitotic ®gures were observed in the protoplast suspensions. Protoplasts were then cultivated in a medium supplemented with the synthetic plant growth regulators 2,4-D and BA (Glab et al. 1994). After 24 h, the cells derived from protoplasts (referred to hereafter as protoplasts) progressed in the cell cycle (45% 2C, 33% S and 22% 4C) but no mitotic ®gures were observed by microscopy. Oryzalin, which blocks the cells at metaphase and therefore prevents them from re-entering the next G1 phase (G1¢), was added to examine the frequency of cells able to divide during a 48-h period. A comparison between G2 frequency in protoplasts cultivated for 48 h in complete medium in either the presence or absence of 3 lM oryzalin indicated clearly that a higher G2 and M cell frequency (72% 4C nuclei with 30% prophase and metaphase) was attained in the presence of oryzalin. Therefore after 48 h, P. hybrida protoplasts are able to reach the new G1¢ phase of the next generation. Roscovitine, an inhibitor of CDK activity, was added to the complete medium at time 0 of the culture (Fig. 2a±d). At 5 lM roscovitine, the percentage of cells in S phase started to decrease (Fig. 2a), while at 25 lM, progression of the cells to S phase was completely prevented (Fig. 2c). The DNA content distribution was found to remain identical to that obtained immediately after protoplast isolation. The 25 lM treatment was reversible, in that, 24 h after roscovitine treatment, washed cells reached the G2 and M phases (Fig. 2i). This clearly indicates that the roscovitine inhibitor blocked P. hybrida cells in the G1 phase, prior to S-phase entry, in a reversible manner.
Fig. 1. Petunia leaf protoplasts cultured for 48 h reach the G1 phase of the next cell generation. Nuclear DNA content was monitored by ¯ow cytometry after Hoechst (2 lg á ml)1) staining. Freshly isolated protoplasts normally show a main peak of 2C cells. After 24 h, oryzalin (ORY; 3 lM) was added for 24 h. For a better illustration of the number of nuclei in the 2C and 4C peaks, a log scale is used for the DNA histograms. For each histogram, 105 nuclei were analysed
218
C. TreÂhin et al.: Control of cell cycle entry by phytohormones
Fig. 2a±i. The cyclin-dependent kinase inhibitor roscovitine de®nes both the restriction point and the G2/M transition point in protoplast-derived cells. Roscovitine (R) was added, in complete medium (2H), at time 0 of the culture (a±d) or following a 24h initial culture-period (e±h). R5, R10, R25 and R50 represent the roscovitine concentrations used: 5, 10, 25 and 50 lM, respectively. Samples were cultured in the presence of roscovitine for 24 h and processed for ¯ow cytometry (see Materials and methods). i Cells were ®rst cultured for 24 h in complete medium supplemented with roscovitine (25 lM), washed, and then cultured for 24 h in complete medium. Flow diagrams depict the time course of the experimental culture procedure
To test for a G2 block, roscovitine was administered for 24 h following an initial culture period of 24 h. As shown in Fig. 1, the initial culture period of 24 h allows cells to enter S phase and subsequently G2 phase, but prevents them from reaching mitosis. The initial culture was then divided into aliquots and cultured for 24 h in a medium supplemented with roscovitine at concentrations ranging from 5 to 50 lM (Fig. 2e±h). The frequency of 4C nuclei (G2) increased when the roscovitine concentration was increased tenfold (Fig. 2e,h). The highest percentages of S and 4C cells were obtained for 25 lM and 50 lM concentrations which gave 70% and 67%, respectively. In these conditions, no mitotic ®gures were observed by staining the cells, as has been previously described (Planchais et al. 1997). This result demonstrates that roscovitine brings about an ecient G2 block. Neither BA nor 2,4-D will allow cells to pass beyond the cell cycle transition de®ned by the roscovitine-G1 block, when added separately. We next used roscovitine to assess whether 2,4-D or BA separately allow cells to pass beyond the cell cycle transition point de®ned by the roscovitine-G1 block and to progress closer to the G1/S transition. Protoplasts were ®rst cultivated for 24 h with
both 2,4-D plus BA or with 2,4-D or BA alone or without any plant growth regulators, and after 24 h, the distribution of nuclear DNA content was analysed following nuclei release and Hoechst staining (Fig. 3a± d). Next, roscovitine (50 lM ®nal concentration) was added for 24 h to cells ®rst cultured in the complete medium (Fig. 3e), simultaneously with BA to cells ®rst cultured in a medium containing 2,4-D alone (Fig. 3f), simultaneously with 2,4-D to cells ®rst cultured in a medium containing BA alone (Fig. 3g), or simultaneously with BA plus 2,4-D to cells ®rst cultured in a medium containing no plant growth regulators (Fig. 3h). After 24 h, the nuclear DNA distribution was determined. The result showed that neither the 2,4D-containing cultures subsequently supplemented with BA, nor those containing BA subsequently supplemented with 2,4-D, were able to progress to S phase. The frequency of G2 cells increased (to 75%) only when cells were initially cultured (during the ®rst 24 h of the experiment) in the presence of the two plant growth regulators simultaneously. This result implies that the two plant growth regulators perform speci®c roles and that both of them are required for passage beyond the point de®ned by the roscovitine block, before the onset of the S phase.
C. TreÂhin et al.: Control of cell cycle entry by phytohormones
219
Controls were performed without roscovitine but with oryzalin added in all samples after the ®rst 24 h of culture. Cells were found to progress towards G2 (Fig. 3i±k). However, there were clearly more 4C (G2) cells amongst the population initially cultivated in 2,4-D alone then supplemented with BA than amongst either those cultivated in BA alone for 24 h before adding 2,4D or those cultivated in the absence of plant growth regulators during the ®rst 24 h. This suggests that although 2,4-D did not allow the cells to progress through the G1/S transition, it prepared the cells to enter S phase more rapidly than did BA alone.
Fig. 3a±k. Protoplasts cultured with either 2,4-D only or with BA only do not pass beyond the restriction point. Left hand column, protoplasts were cultivated for 24 h either in complete medium, 2H (a), with 2,4-D only (b), with BA only (c), or without plant growth regulators, )H (d). Middle column, each of the previous cultures were supplemented for 24 h with roscovitine (50 lM) together with the missing plant growth regulator(s). Right hand column, oryzalin (ORY; 3 lM) and the missing plant growth regulator(s) were added for 24 h to each of the samples described in the left-hand column. Samples were processed for ¯ow cytometry and nuclear DNA content measured (see Materials and methods)
2,4-Dichlorophenoxyacetic acid and BA act sequentially. In conventional univariate DNA-Hoechst ¯ow cytometry, the G1¢ products of mitosis are indistinguishable from those of the original G1 phase. The BrdU/HO/PI technique has however proven to be an ecient tool for studying the time course of cell division following mitogenic stimulation (Ormerod and Kubbies 1992). Incorporation of the thymidine analogue BrdU during the S phase results in Hoechst 33258 (HO) ¯uorescence being quenched and therefore no longer proportional to cellular DNA content, enabling one to follow various eects from one cycle to the next. Therefore, singleparameter staining with HO does not yield information about the cell's location within the cell cycle. This additional information can be obtained by staining nuclei simultaneously with a DNA-speci®c dye, such as PI, that is not sensitive to the BrdU-induced quenching. A bivariate analysis based on dual staining with HO and PI can provide complete information about the cell cycle progression within the time interval between the beginning of BrdU incorporation and cell harvest (Glab et al. 1994). Here, Petunia protoplasts were ®rst cultivated for 24 h in medium containing either 2,4-D or BA alone. Then, media were supplemented with BrdU in combination with the missing plant growth regulator and protoplasts were further cultivated for 18 h. Afterwards, nuclei were released, stained with HO and PI and biparametric analysis was performed to follow protoplast division (Fig. 4b,c). In the control, protoplasts were ®rst cultivated for 24 h in complete medium, then for 18 h in the same medium supplemented with BrdU (Fig. 4a). Cells which had incorporated BrdU were easily distinguishable by their HO ¯uorescence decrease (denoted * throughout). During the 18-h period in the presence of BrdU, 2C cells became 4C*cells (G2), which then became the 2C* cells (G1¢) of the next cycle. The 8C* cells, derived from the 4C cell population, were also present in the leaf tissue. When 2,4-D had been added alone during the ®rst 24 h, BrdU incorporation upon BA addition was observed in 4C*, 8C* and few 2C* cells (Fig. 4b). In contrast, when BA was added alone ®rst, only 4C* and 8C* cells were obtained (Fig. 4c). The percentage of cells engaged in cell division which progressed through the next generation (2C* cells) was clearly higher when protoplasts were ®rst cultivated with the two plant growth regulators or with 2,4-D alone as compared to protoplasts ®rst cultivated with BA alone. This result con®rms that the presence of 2,4-D alone
220
C. TreÂhin et al.: Control of cell cycle entry by phytohormones
(during the ®rst step of the culture) allowed the cells to enter S phase more rapidly than did BA alone. In yeast, histone genes are expressed at the onset of the S phase (reviewed in Koch and Nasmyth 1994). In tobacco and Arabidopsis, histone H4 is transcriptionally regulated during the cell cycle, its peak occuring in S phase (Shaul et al. 1996). Therefore, in a second approach, histone H4 transcript levels were followed in order to study the behaviour of S-phase-engaged cells with respect to the dierent conditions (Fig. 5). Immediately after protoplast isolation or when protoplasts were cultivated in a medium lacking plant growth regulators, no histone H4 transcripts were observed (Fig. 5, lanes 1 and 2). In the 2,4-D-containing culture, a slight signal was observed (Fig. 5, lane 3), but not in the case of the BA-treated culture (Fig. 5, lane 4). The signal was maximal when protoplasts were cultivated for 24 h in 2,4-D medium later supplemented for 18 h with BA (Fig. 5, lane 6). This result, obtained using an S-phasespeci®c marker, con®rms that a more rapid entry of cells into S phase was induced by an initial treatment with 2,4-D than by a corresponding treatment with BA. Auxin induces expression of the MAPK gene whereas cytokinin does not. Previous reports have indicated that MAPK gene transcription may be stimulated after exposure of quiescent animal and yeast cells to a variety of stimuli (Errede and Levin 1993). We examined whether the dierences observed in protoplasts treated with 2,4-D or BA with respect to entry into S phase were correlated with levels of transcripts of the MAPK gene PMEK1.
Fig. 4a±c. The BrdU/HO/PI analysis of cycling activity in protoplasts subjected to delayed BA or 2,4-D addition. Protoplasts were isolated from the fully expanded fourth post-cotyledonary leaf. The BrdUlabelled cells (4C* and 8C*) correspond to cells which replicated their DNA, the next generation (G1¢) being represented by the 2C* population. a BrdU was added for 18 h after protoplasts had been cultured for 24 h in medium containing both plant growth regulators. b BrdU and BA were added for 18 h after protoplasts had been cultured for 24 h in medium containing 2,4-D only. c BrdU and 2,4-D were added for 18 h after protoplasts had been cultured for 24 h in medium containing BA only
Fig. 5. Histone H4 transcripts accumulate only when both 2,4-D and BA are present. Total RNA was prepared from protoplasts immediately after isolation (lane 1). Total RNA was also extracted from protoplasts cultured for 24 h in the following conditions: lane 2, without plant growth regulators; lane 3, with 2,4-D only; lane 4, with BA only; lane 5, in the presence of both plant growth regulators; lane 6, with 2,4-D subsequently supplemented for 18 h with BA; lane 7, with BA subsequently supplemented for 18 h with 2,4-D. For each sample, 20 lg of total RNA was loaded onto an agarose gel which was Northern-blotted (see Materials and methods) prior to hybridization to a histone H4 cDNA probe. The RNAs were stained with ethidium bromide (EtBr) to check that equal amounts of RNA had been loaded
C. TreÂhin et al.: Control of cell cycle entry by phytohormones
Fig. 6. Accumulation of PMEK1 transcripts in protoplast cultures following delayed BA or 2,4-D addition. Total RNAs were extracted from protoplasts cultured for 24 h as follows: lane 1, without plant growth regulator; lane 2, with 2,4-D only; lanes 3 and 4, with 2,4-D subsequently supplemented for 13 h and 18 h respectively with BA; lane 5, with BA only; lanes 6±8, with BA subsequently supplemented for 13 h, 18 h and 24 h, respectively, with 2,4-D. Samples containing 60 lg of total RNA were electrophoresed, blotted and hybridized (see Materials and methods) using a PMEK MAPK cDNA probe. Identical samples were also stained with ethidium bromide (EtBr) to check that equal amounts of RNA had been loaded
Three P. hybrida protoplast cultures containing either no plant growth regulators, 2,4-D alone or BA alone were initiated. After 24 h, BA was added to the cultures containing 2,4-D alone and 2,4-D was added to the samples containing BA alone, for periods of 13 h, 18 h and 24 h. The RNA was then extracted for Northern blot hybridization using the MAPK PMEK1 cDNA as a probe (Fig. 6). No MAPK transcripts were detected in a cell culture incubated in the absence of plant growth regulators for 24 h (Fig. 6, lane 1). On the other hand, incubation for 24 h with 2,4-D resulted in a high level of transcripts (Fig. 6, lane 2) which remained when BA was added and cells were passing through S to G2 (Fig. 6, lanes 3 and 4). In protoplasts cultivated for 24 h in BA alone, the MAPK transcript level was substantially lower than in protoplasts cultivated with 2,4-D alone (Fig. 6, lane 5), whilst a clear increase was observed after the 24-h period following the addition of 2,4-D (Fig. 6, lane 8). We sought to determine whether auxin maintained a high MAPK transcript level originally present in the leaf, or whether it acted via a transcriptional activation (Fig. 7). The PMEK1 transcript level in fully expanded leaves was found to be very low (Fig. 7a, lane 1). However, a much higher expression was observed in protoplasts derived from the same leaves immediately after their isolation (Fig. 7a, lane 2). After protoplasts had been cultivated for 24 h either without plant growth regulators or with only BA, no transcripts were observed (Fig. 7a, lane 3 and 7b, lane 2). Cultures incubated with 2,4-D only for 24 h, or to which 2,4-D was supplemented for 2 h after the 24-h period without plant growth
221
Fig. 7a,b. Expression of PMEK1 is induced both by enzymatic wounding and by auxin. a total RNAs were isolated from fully developed leaves (lane 1), from freshly isolated protoplasts derived from the same leaves (lane 2) and from similarly obtained protoplasts cultivated for 24 h without plant growth regulators (lane 3). b Total RNAs were isolated from protoplasts cultivated for 24 h with 2,4-D only (lane 1), with BA only (lane 2) and without plant growth regulators prior to the addition of 2,4-D for 1 h or 2 h (lanes 3 and 4, respectively). Total-RNA samples of 20 lg were Northern-blotted and hybridized as described in the legend to Fig. 6
regulators, showed a high level of MAPK gene expression (Fig. 7b, lanes 1, 3 and 4). Auxin without cytokinin is not able to stimulate cdc2Pet transcript accumulation. The cdc2Pet cDNA obtained by reverse transcriptase-polymerase chain reaction (RTPCR) was used as a probe to monitor the transcript levels of this gene in protoplasts treated for 24 h either with 2,4-D alone or with 2,4-D plus BA (Fig. 8, lanes 3 and 5). Transcript levels did not increase over the 24-h culture period when only auxin was added (Fig. 8, lane 3); on the other hand, a clear reproducible increase in
Fig. 8. Induction of cdc2Pet transcript accumulation requires both 2,4-D and BA. Total RNA was extracted from protoplasts cultured for 24 h in the following conditions: lane 1, without plant growth regulators; lane 2, with 2,4-D only in presence of 25 lM roscovitine; lane 3, with 2,4-D only; lane 4, in the presence of both plant growth regulators supplemented with 25 lM roscovitine; lane 5, in the presence of both plant growth regulators. For each sample, 50 lg of total RNA was loaded onto an agarose gel which was Northernblotted (see Materials and methods) prior to hybridization to a cdc2Pet cDNA probe. The RNAs were stained with ethidium bromide (EtBr) to check that equal amounts of RNA had been loaded
222
cdc2Pet transcript accumulation was observed for a culture to which both plant growth regulators were added (Fig. 8, lane 5). Addition of BA alone had no eect on cdc2Pet transcript levels (data not shown). To investigate cdc2Pet trancript levels in cells progressing to S phase and subsequently G2, roscovitine (25 lM) was applied for 24 h to protoplasts in the presence of the two plant growth regulators (Fig. 8, lane 4). The cdc2Pet transcript level in protoplasts cultivated with both plant growth regulators but without roscovitine is clearly higher than the corresponding cdc2Pet transcript level observed when the CDK inhibitor is present. Discussion In both multicellular and unicellular eukaryotes, the ability to enter and exit the cell cycle at the appropriate moment is vital to survival. In yeast, the G1 decision to initiate a new cell cycle is called START (Hartwell et al. 1974) whilst the corresponding restriction point in animal cells is known as R (Pardee 1974). Our task is twofold, namely to characterize the mechanisms which control the succession of events that allow isolated dierentiated cells to re-enter the mitotic plant cell cycle, ®rstly with respect to plant homologues of yeast and animal cell cycle regulators and secondly in terms of plant-speci®c mechanisms, such as those which determine plant cell totipotency. Most plant cell cycle regulators characterized to date resemble those active in the G2/M transition of yeast and animal cells, although biochemical demonstrations of their roles in plants are at present lacking (reviewed in Jacobs 1995). Previous phosphorylation experiments with the human p9CKShs1 suc1 homolog (p13suc1) covalently linked to a resin allowed the detection of a peak of H1 kinase activity which was higher in G2 and M cells than in G1-arrested cells (Planchais et al. 1997). Similarly, using p13suc1, H1 kinase activity was detected in alfalfa cells in S, G2 and M phases (Magyar et al. 1993). Moreover, the chemical CDK inhibitor olomoucine was shown to be ecient not only at the G2/M transition but also at the G1/S transition both in P. hybrida and in Arabidopsis (Glab et al. 1994). In this study, we have analysed the inhibitory eect of roscovitine which acts as a competitive inhibitor for ATP binding (Meijer et al. 1997). As in mammalian and tobacco cells, roscovitine eciently blocks cells at the G1/S and G2/M transitions. However, roscovitine has been shown to be 10 times more ecient than olomoucine with respect to G1 and G2 blocking at a concentration of 25 lM. Indeed, our initial results using P. hybrida protoplasts with olomoucine required a concentration of 200 lM in order to produce the desired eect (Abraham et al. 1995). The fact that an ecient block occurs suggests that regulation of the G1/ S and G2/M transitions in P. hybrida requires the activity of a CDK-cyclin complex. The roscovitine-G1 block could de®ne a checkpoint in the Petunia cell cycle which might correspond to the restriction point or
C. TreÂhin et al.: Control of cell cycle entry by phytohormones
START, thus con®rming results previously obtained by Glab et al. (1994). To date, experiments in which the addition of plant growth regulators is delayed have shown that mitotic induction requires the presence of auxin, cytokinin being only necessary at a later stage before mitosis (Meyer and Cooke 1979). In contrast, in P. hybrida protoplasts, 2,4D and BA are necessary for entry into the S phase (Bergounioux et al. 1988). However, it was not determined whether both of the plant growth regulators are necessary before or after the restriction point. If either 2,4-D or BA alone induces cells to progress towards the S boundary and beyond the restriction point de®ned by the roscovitine block, roscovitine addition to the medium will not prevent their progression towards the G2 phase. On the other hand, if cultures have not reached the restriction point, they will be blocked when the missing plant growth regulator is added in combination with roscovitine. Our results clearly show that protoplasts cultivated in 2,4-D or BA alone do not pass beyond the restriction point. Therefore, this result suggests that the two plant growth regulators are together necessary to obtain an activation of cyclindependent kinase in the G1 phase of P. hybrida. However, their eects can be distinguished. Indeed, protoplasts cultivated ®rst in 2,4-D alone for 24 h, then for 18 h in a medium containing both phytohormones, display a higher percentage of G2 cells than do those initially cultured in BA alone prior to supplemention with 2,4-D. Auxin therefore seems to act either before cytokinin or in a separate signalling pathway. This observation was con®rmed in the assay of Hoechst quenching following BrdU incorporation. The latter has the advantage of associating DNA content and BrdU incorporation (Glab et al. 1994), such that we can follow the DNA synthesis of a cell population resulting from the eect of a delayed supply of one of the necessary plant growth regulators to the medium. Our results clearly show that a much higher proportion of BrdU labelled nuclei and higher histone H4 transcript levels are obtained when cultures are initiated in 2,4-D. This demonstrates that DNA synthesis is more rapidly engaged in the presence of 2,4-D and therefore that 2,4-D-engaged cells progress towards S in a speci®c manner which is not obtained by BA treatment alone. Therefore, despite the fact that both auxin and cytokinin are necessary for entry into S phase, an auxin pretreatment allows cells to enter S phase more rapidly. The binding of ligands to membrane or soluble cytoplasmic receptors results in the activation of downstream pathways, and in some cases transcriptional induction of rate-limiting mitotic control elements such as MAPK (Errede and Levin 1993). There are few reports describing the regulatory response of MAPK in plants mediated by transcriptional control. However, transcripts for PKABA1 and WPK4 from wheat have been shown to accumulate following treatment with the phytohormones abscisic acid and cytokinins, respectively (Anderberg and Walker-Simmons 1992; Sano and Yousse®an 1994). We did not witness a PMEK1 (P. hybrida MAPK) transcript signal in fully developed
C. TreÂhin et al.: Control of cell cycle entry by phytohormones
223
Fig. 9. Schematic hypothetical model for the induction of mitosis in Petunia leaf protoplasts. The initiation of the mitotic cell cycle involves the two plant growth regulators auxin (2,4-D) and cytokinin (BA) which collectively provide the signals necessary for progression to a restriction point de®ned by the cyclin-dependent inhibitor roscovitine. Both of them induce (Å) the accumulation of cdc2Pet transcripts whereas 2,4-D alone is able to induce the accumulation of PMEK1 transcripts
P. hybrida leaves, an observation similar to that of Seo et al. (1995). In contrast Wilson et al. (1993), Mizoguchi et al. (1994), and Decroocq-Ferrant et al. (1995) each found transcripts of MAPK genes to be present in leaves. Our results show, however, that PMEK1 transcripts accumulate after protoplast isolation from these fully expanded leaves. The enzymatic procedure used for protoplast isolation has already been shown to induce proteins characteristic of stress (Fleck et al. 1982). Moreover, tobacco MAPK has been shown to be a possible mediator in wound signal transduction pathways (Seo et al. 1995). Thus, PMEK1 expression could, in this case, be a response to enzymatic wounding in a situation similar to that described by Mizoguchi et al. (1996). When protoplasts are cultivated for 24 h in medium depleted of both plant growth regulators or in the presence of cytokinin alone, PMEK1 is no longer expressed. On the other hand, its expression remains high when protoplasts are cultivated in the presence of 2,4-D. Moreover, the incubation of protoplasts in the presence of auxin, following plant growth regulator deprivation (during 24 h), results in PMEK1 transcript accumulation over a 2-h period. Thus, PMEK1 gene expression is clearly induced by auxin treatment. By analogy with quiescent mammalian cells in culture, in which reactivation of the cell cycle is mediated by the activation of MAPK in response to growth factors (Owaki et al. 1992), PMEK1 could be involved in the triggering of cells from G0 to G1/S transition. Furthermore, MAPK activity appears to increase rapidly after auxin treatment in auxin-starved BY-2 cells (Mizoguchi et al. 1994). This suggests that 2,4-D is involved in both the transcriptional regulation of MAPK genes and in the reactivation of MAPK proteins. Thus, an important question to be asked is: is 2,4-D-transcriptional regulation of PMEK1 mediated by the presence of an auxin response element in its promoter?
Both 2,4-D and cytokinin were also involved in the induction of cdc2Pet transcript accumulation prior to G1/S transition in Petunia protoplasts. In this connection, it is interesting that cdc2 transcript accumulation was found to be induced in Nicotiana stem pith cultivated only with a-naphthaleneacetic acid (Zhang et al. 1996) and that cytokinin is involved in events preceding S phase in gibberellin-treated rice internodes (Sauter et al. 1995). When both plant growth regulators are present, the increased levels of cdc2Pet transcripts observed in non-roscovitine blocked cells as compared with cells treated with the CDK inhibitor suggest that cdc2Pet transcription continues to increase in cells which progress through S to G2. Therefore, in Petunia cells, plant growth regulators appear to stimulate growth in G1 phase in a similar manner to that seen in animal cells. When Petunia cells are transferred to a medium lacking plant growth regulators, they continue to progress towards G2 (data not shown). It remains to be determined whether hormonal stimulation is still required beyond the restriction point for completion of the cell cycle or whether the internal level of plant growth regulators is sucient to maintain cell cycle progression. In conclusion, the various elements described above which operate in our experimental system may be proposed to interact according to the scheme shown in Fig. 9. It is hoped that further studies of the key cell cycle signalling elements described here will shed further light on the complex mechanisms which regulate cell division in plants. We are grateful to Drs Laurent Meijer (Station Biologique, Rosco, France) for supplying the roscovitine inhibitor, to Martin Kreis and VeÂronique Decroocq-Ferrant for providing the P. hybrida PMEK1 cDNA and to Nicole Chaubet (IBMP, Strasbourg, France) for supplying the Arabidopsis histone H4 cDNA. We appreciate helpful discussions with Drs Dirk Inze (University of Ghent, VIB, Ghent, Belgium) and Spencer Brown (ISV, Gifs/
224 Yvette, France). We thank DanieÁle De Nay (ISV) and Jean-Marc Bureau (ISV) for excellent cytometry. This work was partly supported by the Franco-Flamish cooperation programme Tournesol and the Association pour la recherche sur le cancer. S.P. was supported by an EC Grant, number ERBFMBICT961724.
References Abraham RT, Acquarone M, Andersen A, Asensi A, Belle R, Berger F, Bergounioux C, Brunn G, Buquet-Fagot C, Fagot D, Glab N, Goudeau H, Goudeau M, Guerrier P, Houghton P, Hendriks H, Kloareg B, Lippai M, Marie D, Maro B, Meijer L, Mester J, Mulner-Lorillon O, Poulet SA, Schierenberg E, Schutte B, Vaulot D, Verlhac MH (1995) Cellular eects of olomoucine, an inhibitor of cyclin-dependent kinases. Biol Cell 83: 105±120 Anderberg RJ, Walker-Simmons MK (1992) Isolation of a wheat cDNA clone for an abscisic acid-inductible transcript with homology to protein kinases. Proc Natl Acad Sci USA 89: 10183±10187 Bergounioux C, Perennes C, Brown SC, Gadal P (1988) Cytometric analysis of growth-regulator-dependent transcription and cell cycle progression in P. hybrida protoplast cultures. Planta 175: 500±505 Church GM, Gilbert W (1984) Genomic sequencing. Proc Natl Acad Sci USA 81: 1991±1995 Decroocq-Ferrant V, Decroocq S, Van Went J, Schmidt E, Kreis M (1995) A homologue of the MAP/ERK family of protein kinase genes is expressed in vegetative and in female reproductive organs of Petunia hybrida. Plant Mol Biol 27: 339±350 Errede B, Levin DE (1993) A conserved kinase cascade for MAP kinase activation in yeast. Curr Opin Cell Biol 5: 254±260 Fleck J, Durr A, Fritsch C, Vernet T, Hirth L (1982) Osmoticshock `stress proteins' in protoplasts of Nicotiana sylvestris. Plant Sci Lett 26: 159±165 Glab N, Labidi B, Qin LX, TreÂhin C, Bergounioux C, Meijer L (1994) Olomoucin, an inhibitor of the cdc2/cdK kinases activity, blocks plant cells at the G1 to S and G2 to M cell cycle transitions. FEBS Lett 533: 207±211 Hartwell L, Culotti J, Pringle J, Reid BJ (1974) Genetic control of the cell division cycle in yeast. Science 183: 46±51 Hemerly A, Ferreira P, De Almeira Engler J, Van Montagu M, Engler G, Inze D (1993) cdc2a expression in Arabidopsis is linked with competence for cell division. Plant Cell 5: 1711±1723 Hunter T (1995) Protein kinase and phosphatases: The Yin and Yang of protein phosphorylation and signaling. Cell 80: 225± 236 Jacobs W (1995) Cell cycle control. Annu Rev Plant Physiol Plant Mol Biol 46: 317±319 Koch C, Nasmyth K (1994) Cell cycle regulated transcription in yeast. Curr Opin Cell Biol 6: 451±459 Magyar Z, Bako L, Bgre L, Dedeoglu D, Kapros T, Dudits D (1993) Active cdc2 genes and cell cycle phase-speci®c cdc2related kinase complexes in hormone-stimulated alfalfa cells. Plant J 4: 151±161 Meijer L (1995) Chemical inhibitors of cyclin-dependent kinases. In: Meijer L, Guidet S, Tung HYL (eds) Progress in cell cycle research, vol 1. Plenum Press, New York, London, pp 351±363 Meijer L, Borgne A, Mulner O, Chong JPJ, Blow JJ, Delcros J-G, Moulinoux J-P (1997) Biochemical and cellular eects of roscovitine, a potent and selective inhibitor of the cyclindependent kinases cdc2, cdk2 and cdk5. Eur J Biochem 243: 527±536 Meyer Y, Cooke R (1979) Time course of hormonal control of the ®rst mitosis in tobacco mesophyll protoplasts cultivated in vitro. Planta 147: 181±185 Millner PA (1995) The auxin signal. Curr Opin Cell Biol 7: 224±231
C. TreÂhin et al.: Control of cell cycle entry by phytohormones Mizoguchi T, Gotho Y, Nishida E, Yamaguchi-Shinozaki K, Hayashida N, Kamada H, Shinozaki K (1994) Characterization of two cDNAs that encode MAP kinase homologues in Arabidopsis thaliana and analysis of the possible role of auxin in activating such kinase activities in cultured cells. Plant J 5: 111±122 Mizoguchi T, Irie K, Hirayama T, Hayashida N, YamaguchiShinozaki K, Matsumoto K, Shinozaki K (1996) A gene encoding a mitogen-activated protein kinase is induced simultaneously with genes for a mitogen-activated protein kinase and an S6 ribosomal protein kinase by touch, cold and water stress in Arabidopsis thaliana. Proc Natl Acad Sci USA 93: 765±769 Mizoguchi T, Ichimura K, Shinozaki K (1997) Environmental stress reponse in plants: the role of mitogen-activated protein kinases. Trends Biotech 15: 15±19 Ormerod MG, Kubbies M (1992) Cell cycle analysis of asynchronous cell populations by ¯ow cytometry using bromodeoxyuridine label and Hoechst-propidium iodide stain. Cytometry 13: 678±685 Owaki H, Makar R, Boulton TG, Cobb MH, Geppert TD (1992) Extracellular signal-regulated kinases in T cells: characterization of human ERK1 and ERK2 cDNAs. Biochem Biophys Res Commun 182: 1416±1422 Pardee AB (1974) A restriction point for control of normal animal cell proliferation. Proc Natl Acad Sci USA 71: 1286±1290 Perennes C, Qin LX, Glab N, Bergounioux C (1993) Petunia p34cdc2 protein kinase activity in G2/M cells obtained with a reversible cell cycle inhibitor, mimosine. FEBS Lett 333: 141±145 Planchais S, Glab N, TreÂhin C, Perennes C, Bureau JM, Meijer L, Bergounioux C (1997) Roscovitine, a novel cyclin-dependent kinase inhibitor, characterizes restriction point and G2/M transition in tobacco BY-2 cell suspension. Plant J 12: 191±202 Sano H, Yousse®an S (1994) Light and nutritional regulation of transcripts encoding a wheat protein kinase homolog is mediated by cytokinins. Proc Natl Acad Sci USA 91: 2582±2586 Sauter M, Mekhedov SL, Kende H (1995) Gibberellin promotes histone H1 kinase activity and the expression of cdc2 and cyclin genes during the induction of rapid growth in deepwater tice internodes. Plant J 7: 623±632 Segers G, Gadisseur I, Bergounioux C, De Almeida Engler J, Jacquemard A, Van Montagu M, Inze D (1996) The Arabidopsis cyclin-dependent kinase gene cdc2bAt is preferentially expressed during S and G2 phases of the cell cycle. Plant J 10: 601±612 Seo S, Okamoto M, Seto H, Ishizuka K, Sano H, Ohashi Y (1995) Tobacco MAP kinase: A possible mediator in wound signal transduction pathways. Proc Natl Acad Sci USA 270: 1988± 1995 Shaul O, Mironov V, Burssens S, Van Montagu MV, Inze D (1996) Two Arabidopsis cyclin promoters mediate distinctive transcriptional oscillation in synchronized tobacco BY-2 cells. Proc Natl Acad Sci USA 93: 4868±4872 Shi LY, Strabala TJ, Hagen G, Guifoyle TJ (1994) Transgenic tobacco plants that overproduce cytokinins show increased tolerance to exogenous auxin and auxin transport inhibitors. Plant Sci 100: 9±14 Takahashi Y, Ishida S, Nagata T (1995) Auxin-regulated genes. Plant Cell Physiol 36: 383±390 Vesely J, Havlicek L, Strnad M, Blow JJ, Donella-Deana A, Pinna L, Letham DS, Kato J, Detivaud L, Leclerc S, Meijer L (1994) Inhibition of cyclin dependent kinases by purine analogues. Eur J Biochem 224: 771±786 Wilson C, Eller N, Gartner A, Vicente O, Heberle-Bors E (1993) Isolation and characterization of a tobacco encoding a putative MAP kinase. Plant Mol Biol 23: 543±551 Zhang K, Letham DS, John PCL (1996) Cytokinin controls the cell cycle at mitosis by stimulating the tyrosine dephosphorylation and activation of p34cdc2-like H1 histone kinase. Planta 200: 2± 12
