Cell-cycle control as a target for calcium, hormonal and developmental signals: the role of phosphorylation in the retinoblastoma-centred pathway

Abstract

Background

During the life cycle of plants, both embryogenic and post-embryogenic growth are essentially based on cell division and cell expansion that are under the control of inherited developmental programmes modified by hormonal and environmental stimuli. Considering either stimulation or inhibition of plant growth, the key role of plant hormones in the modification of cell division activities or in the initiation of differentiation is well supported by experimental data. At the same time there is only limited insight into the molecular events that provide linkage between the regulation of cell-cycle progression and hormonal and developmental control. Studies indicate that there are several alternative ways by which hormonal signalling networks can influence cell division parameters and establish functional links between regulatory pathways of cell-cycle progression and genes and protein complexes involved in organ development.

Scope

An overview is given here of key components in plant cell division control as acceptors of hormonal and developmental signals during organ formation and growth. Selected examples are presented to highlight the potential role of Ca²⁺-signalling, the complex actions of auxin and cytokinins, regulation by transcription factors and alteration of retinoblastoma-related proteins by phosphorylation.

Conclusions

Auxins and abscisic acid can directly influence expression of cyclin, cyclin-dependent kinase (CDK) genes and activities of CDK complexes. D-type cyclins are primary targets for cytokinins and over-expression of CyclinD3;1 can enhance auxin responses in roots. A set of auxin-activated genes (AXR1–ARGOS–ANT) controls cell number and organ size through modification of CyclinD3;1 gene expression. The SHORT ROOT (SHR) and SCARECROW (SCR) transcriptional factors determine root patterning by activation of the CYCD6;1 gene. Over-expression of the EBP1 gene (plant homologue of the ErbB-3 epidermal growth factor receptor-binding protein) increased biomass by auxin-dependent activation of both D- and B-type cyclins. The direct involvement of auxin-binding protein (ABP1) in the entry into the cell cycle and the regulation of leaf size and morphology is based on the transcriptional control of D-cyclins and retinoblastoma-related protein (RBR) interacting with inhibitory E2FC transcriptional factor. The central role of RBRs in cell-cycle progression is well documented by a variety of experimental approaches. Their function is phosphorylation-dependent and both RBR and phospho-RBR proteins are present in interphase and mitotic phase cells. Immunolocalization studies showed the presence of phospho-RBR protein in spots of interphase nuclei or granules in mitotic prophase cells. The Ca²⁺-dependent phosphorylation events can be accomplished by the calcium-dependent, calmodulin-independent or calmodulin-like domain protein kinases (CDPKs/CPKs) phosphorylating the CDK inhibitor protein (KRP). Dephosphorylation of the phospho-RBR protein by PP2A phosphatase is regulated by a Ca²⁺-binding subunit.

CELL DIVISION PARAMETERS RESPOND TO HORMONAL AND STRESS STIMULI

As discussed in several comprehensive reviews, the basic regulatory mechanisms in cell-cycle progression rely on a multicomponent system including transcriptional regulation, protein-protein interaction, phosphorylation–dephosphorylation and protein degradation (De Veylder et al., 2007; Dudits et al., 2007; Francis, 2007, Berckmans and De Veylder, 2009; Jurado et al., 2008). A close functional link between plant hormones, primarily auxin, cytokinins and regulatory proteins of cell-cycle events, has been demonstrated by different experimental approaches. The recent review by Perrot-Rechenmann (2010) summarizes auxin actions on cell division. Plant hormones can directly determine cell-cycle entry and progression or can act indirectly, through a variety of regulatory proteins involved in developmental programmes (see Fig. 1). The auxin [2,4-dichlorophenoxy acetic acid (2,4-D)]-stimulated entry into the S-phase is shown in Fig. 2, where histone H4 promoter activity was increased as indicated by the number of β-glucuronidase (GUS)-positive cells in transgenic maize tissues. Auxin, ethylene and wounding can also directly stimulate G₂/M specific events, as reflected by activation of the promoter of the mitotic CDKB2,1 gene in Medicago leaves (Zhipanova et al., 2006). Treatment of Arabidopsis seedlings with auxins and cytokinins activated both CDKA and CyclinD1;1–D2;1 genes, while KRP4 transcription was down-regulated (Cho et al., 2010).

Fig. 1.

Schematic overview of potential links between calcium, hormonal and developmental signals and cell cycle regulators; note that the proposed scheme is not complete. Abbreviations: CK, cytokinin; E2F/DP, transcription factors; RBR, retinoblastoma-related protein; P, phospho-protein; CYC, cyclin; CDK, cyclin-dependent kinase; PP2A, phosphatase; SCR, SCARECROW; SHR, SHORT ROOT; SCF, SKP1 + CULLIN + F-box (SKP2); EBP1, plant homologue of epidermal growth factor-binding protein; SKP2, F-box protein; STM, SHOOT MERISTEMLESS; KRP, CDK inhibitor; CaM, calmodulin; CPK, calmodulin-like domain protein kinase; ABAP1, armadillo BTB Arabidopsis protein 1; TCP24, transcription factor; CDT1, DNA replication-licensing factor; ABP1, auxin-binding protein 1; ANT, aintegumenta; ARGOS, auxin-regulated gene in organ size; AXR1, RUB1-activating enzyme; ABA, abscisic acid; GL2, GLABRA (root hair); GEM, GL2 expression regulator; ACS5, 1-aminocyclo-propane-1-carboxil acid synthase. For detail see the text.

Fig. 2.

Auxin (2,4-D) activates while abscisic acid (ABA) depresses wheat histone H4 promoter function serving as a marker of S-phase cells in transgenic maize tissues. The GUS (β-glucuronidase) reporter gene indicates the promoter activity by blue staining.

Plant protoplast cultures offer an ideal experimental system for studies on the hormone-dependence of cell division and the function of cyclin-dependent kinase (CDK) complexes (Pasternak et al., 2007). After the regeneration of the cell wall, single alfalfa cells may enter into the division cycle synchronously, when cultured in a medium containing 2 mg L⁻¹ 2,4-D as the auxin component. Cytokinin (0·5 mg L⁻¹ zeatin) was required for progression through S-phase and completion of cell division. These responses are linked to CDKA1;1 and CDKB1;1 kinase activities. In the absence of cytokinin, CDKA1;1 kinase protein is synthesized but lacks histone H1 phosphorylation activity. In contrast, the other G₂/M kinase (CDKB1;1) was functional in the absence of this hormone (Pasternak et al., 2007). The expression of cyclin genes is activated by auxins and cytokinins, as shown in various cell types (Nieuwland et al., 2007).

Abscisic acid (ABA) regulates plant growth and development through its signalling network responding to stress factors (see review by Cutler et al., 2010). Exogenously applied ABA inhibits S-phase progression, as reflected by histone H4 promoter function in maize callus tissues (Fig. 2). Moreover, the CDK activity in alfalfa leaves was significantly reduced by ABA even in the presence of auxin and cytokinin (Mészáros et al., 2000). Treatment of BY2 tobacco G₁ cells with ABA inhibited the G₁/S transition after application of an aphidicolin block (Swiatek et al., 2002). ABA is generally viewed as a growth inhibitor having significant roles under abiotic stresses such as drought (Skirycz and Inzé, 2010). Drought stress can essentially reduce cell division in both roots and leaves; e.g. water limitation decreased cell division rate in the maize root tip at the basal region of the meristem (Sacks et al., 1997). However, in both wheat and maize leaves, the amount of Cdc2-like protein (CDK) was not altered by drought treatment, as shown by western blotting with PSTAIRE antibody. In the same basal shoot tissues, p13^SUC1-bound kinase activities were 50 % lower in extracts from the stressed compared with the non-stressed leaves (Schuppler et al., 1998; Granier et al., 2000). Primary roots of Arabidopsis responded to salt treatment (0·5 % NaCl) by a reduction in the number of dividing cells and a transient decrease in CDK activities (West et al., 2004). Activation of CDK inhibitor genes (ICK/KRP) under stress conditions or by ABA treatment also provides a basis for the reduction of cell division activities (Wang et al., 2008). Different stressors may differentially induce various members of the ICK/KRP gene family. Drought or cold stresses can activate an additional CDKA1;1 inhibitor encoded by the rice EL2 gene (Peres et al., 2007).

CALCIUM AS A CENTRAL SIGNAL MOLECULE MEDIATING HORMONAL AND STRESS EFFECTS

Stimulus-specific and dynamic alterations in the level of free calcium in the cytosol function as a cellular signalling control for a variety of processes, including cell division, metabolism and gene expression (Tuteja and Mahajan, 2007). A whole set of binding proteins acting as Ca²⁺ sensors such as calmodulin, calmodulin-like proteins; calcineurin B-like proteins, phospholipase D, annexins, calreticulin and pistil-expressed binding protein transmit signals eliciting downstream responses. Protein kinases and phosphatases mediate Ca²⁺-dependent phosphorylation events in the control of gene expression or enzyme activities in a specific manner (Kudla et al., 2010). In mammalian cells Ca²⁺ signalling is required for cell-cycle progression and the Ca²⁺ signalling apparatus is remodelled in cancer cells (Roderick and Cook, 2008). In the zygote of the brown alga Fucus serratus both S-phase and zygotic polarization were shown to be dependent on Ca²⁺ elevation in the pre-S-phase (Bothwell et al., 2008). In plants, a transient increase in cytosolic Ca²⁺ can regulate cell-cycle progression in response to abiotic stresses. Sano et al. (2006) have demonstrated the increase of Ca²⁺ in BY2 tobacco cells through the application of oxidative stress (KMnO₄) or hypoosmotic treatment. These oxidative stresses inhibited the entry of cells into mitosis and delayed the cell cycle in a Ca²⁺-dependent manner. Out of several elements of a complex signalling cascade linking cellular Ca²⁺ to cell-cycle regulation, the calcium-dependent, calmodulin-independent or calmodulin-like domain protein kinases (CDPKs/CPKs) have been proposed as active signal mediators (Dudits et al., 1998). In the Arabidopsis genome >30 genes encode CDPKs and members of this kinase family are activated by Ca²⁺ and show autophosphorylation (Bögre et al., 1988, Cheng et al., 2002). The medicago kinase (MsCPK3) gene responded to treatment with a high concentration of auxin (2,4-D) known to be an inducer of asymmetric cell division and somatic embryogenesis in leaf protoplasts (Davletova et al., 2001). In cucumber seedlings during adventitious root formation induced by auxin (β-indole acetic acid) or nitric oxide treatments, the CDPK enzyme was activated in a Ca²⁺-dependent manner (Lanteri et al., 2006).

The link between CPK-mediated Ca²⁺ signals and cell-cycle control was established by experiments showing the elevated activity of an alfalfa CDK inhibitor protein (MtKRP) after phosphorylation by the recombinant MsCPK3 protein (Pettkó-Szandtner et al., 2006). The inhibitory function of the MtKRP protein was also demonstrated by using the histone H1 protein and the alfalfa RBR recombinant fragment as CDK substrate. Both MsCPK3 and MtKRP transcript levels were increased by ABA and salt treatments, which are known as inhibitors of cell division. The Arabidopsis KRP2 inhibitor protein regulating the endoreduplication cycle can serve as a substrate for mitotic CDKB1;1 kinase and this phosphorylation can reduce KRP2 stability (Verkest et al., 2005). In the alfalfa experimental system G₂/M kinase complexes were more sensitive towards the recombinant MtKRP inhibitor than were S-phase complexes (Pettkó-Szandtner et al., 2006). Calmodulin, like other essential Ca²⁺ sensors in plants has a primary role in abiotic and biotic stress responses as shown by transcriptional data (Kim et al., 2009). In Arabidopsis and tobacco cells a kinesin-like calmodulin-binding protein (KCBP) plays a role in the formation of microtubule arrays (Bowser and Reddy, 1997). Calcineurin B-like (CBL) proteins, as members of the Ca²⁺ signalling cascade, can regulate the biosynthesis of ethylene and polyamines (Oh et al., 2008).

In summary, the specificity of cellular responses to the increase in the level of cytosolic Ca²⁺ is dependent on the complexity of a signalling cascade aimed at well-defined targets. In cell-cycle control, CDK activities can represent one of these key targets that play a central role by phosphorylating a set of regulatory proteins including cyclins, histones and retinoblastoma-related proteins (RBRs).

PLANT RETINOBLASTOMA-RELATED PROTEINS AS PHOSPHO-PROTEINS IN THE CONTROL OF THE CELL DIVISION CYCLE

The RBRs, as structural and functional counterparts of the mammalian tumour suppressor pRb proteins in higher plants, have divergent roles in the regulation of the division cycle and development (Gutierrez, 1998; Durfee et al., 2000; Wildwater et al., 2005; Wyrzykowska et al., 2006; Sablowski, 2007; Costa and Gutierrez-Marcos, 2008; Paz Sanchez et al., 2008; Chen et al., 2009; Sabelli and Larkins, 2009). RBR cDNA clones have been identified from both dicot and monocot plant species. While dicot plants have only a single gene, monocot cereal species carry at least two distinct genes with characteristic expression patterns (Lendvai et al., 2007; Miskolczi et al., 2007). Microarray analysis showed an elevated expression of the Arabidopsis RBR1 gene in young roots, stems and light-grown seedlings (de Almeida et al., 2009). In the mammalian cell division cycle retinoblastoma proteins (pRbs) act as a ‘pocket domain’ protein regulating G₁- to S-phase transition through phosphorylation-dependent interaction with the E2F family transcription factors either repressing or activating genes required for cell-cycle check point transition. The Rb–E2F complexes are involved in several basic cellular events such as carcinogenesis, apoptosis and cell differentiation (reviewed by Poznic, 2009).

Also in plants, the RBR functions are controlled by phosphorylation and protein–protein interactions. Similarly to human pRBs, plant RBR proteins are composed of an amino-terminal region, A and B domains in the pocket region and a C-terminal domain. These proteins have several potential CDK phosphorylation sites (Durfee et al., 2000; Boniotti and Gutierrez, 2001; Miskolczi et al., 2007). Interactions between CDKs and D-type cyclins are required for the formation of active kinase complexes that can phosphorylate RBR proteins, as shown by Nakagami et al. (1999). These plant cyclins may have the LxCxE motif that mediates the binding of a variety of proteins to RBR proteins (Huntley et al., 1998; Soni et al., 1995). The NtRBR1 protein of tobacco was phosphorylated by the CYCD3;3/CDKA complex and this kinase activity was detected in G₁ and S-phase cell extracts from synchronized tobacco BY-2 cells (Nakagami et al., 2002).

The cell-cycle phase-dependent phosphorylation of plant RBRs relies on various kinase complexes and determines differential RBR functions. In synchronized alfalfa cells, western blotting of total protein extracts detected the 115-kDa full-size MsRBR protein in all cell-cycle phases with limited variation (Ábrahám et al., 2011). Immunoblotting the same protein samples with anti-phospho-pRb antibodies indicated a low phospho-MsRBR protein level in the control culture with G₁ cells. This was elevated in samples with S-phase cells. Protein amounts cross-reacting with anti-phospho-pRb peptide antibodies were found to be high in extracts isolated from cell populations representing a significant frequency of G₂/M-phase cells. Subsequently the phospho-MsRBR protein was reduced in samples containing G₁ cells. In these studies, both the MedsaCDKA1;1 PSTAIRE motif and the mitotic MedsaCDKB2;1 kinases phosphorylated the His-tagged C-terminal fragment of the MsRBR protein produced in vitro. Kawamura et al. (2006) generated antibodies against the C-terminal region of tobacco NtRBR1 protein and different phospho-serine peptides containing sequences from NtRBR1. The NtRBR1 protein was phosphorylated by both CDKA and CDKB kinases immunoprecipitated from actively growing cells. Antibodies, recognizing specific phospho-serine residues, cross-reacted differentially with the NtRBR1 protein in various phases of the cell cycle. Hirano et al. (2008) demonstrated by pull-down assay that the non-phosphorylated AtRBR1 protein could bind to the E2Fa protein, whereas the hyper-phosphorylated form did not interact with the E2Fa protein. These findings are in agreement with the general model of G₁/S-phase transition control, where the hypophosphorylated Rb proteins act as transcriptional repressor by inhibition of E2F functions in the regulation of the expression of S-phase-specific genes (Poznic, 2009).

The presence of phospho-MsRBR proteins in interphase and mitotic cells was also demonstrated by immunolocalization experiments (Ábrahám et al., 2011). These studies showed intensive staining of the interphase nuclei and the signals were concentrated into spots. As shown by Fig. 3, in mitosis the prophase cells have labelled granules. These structures cannot be recognized in later phases such as prometaphase, metaphase, anaphase and telophase. These observations emphasize the presence of phospho-RBR protein in the G₂ phase and mitosis in plant cells. The functional significance of pRB tumour suppressor and Rbf1 proteins has been demonstrated in mammalian and Drosophila G₂/M cells (Nair et al., 2009; Lavoie, 2008). Analysis of the MsRBR protein in cells from cultures at different growing phases revealed an increase in both forms of this protein during exponential growth. Linking division activity to the presence of the MsRBR proteins was also supported by hormone starvation experiments. The prolonged withdrawal of synthetic auxin and kinetin stopped the growth of A2 alfalfa cultures and the MsRBR protein could not be detected in these cells (Ábrahám et al., 2011). The cell division-dependent presence of plant RBR proteins could also be concluded from additional experiments. Only the non-phosphorylated AtRBR1 protein was detected in stationary-phase MM2d Arabidopsis cells. Transfer into fresh medium resulted in the elevation of detectable amounts of AtRBR1 protein and after 8 h its phosphorylated form was detected by western blotting (Hirano et al., 2008). During the growth of Arabidopsis roots the auxin-binding protein 1 (ABP1) is essential for the G₁/S transition through interaction with the cyclinD/RBR pathway (Tromas et al., 2009).

Fig. 3.

Immunodetection of phospho-MsRBR protein in prophase alfalfa cells as nuclear granules by antibodies produced against the human phospho-pRb peptide (green labelling). These nuclear granules cannot be recognized in later mitotic phases as shown by anaphase cells. The red staining of DNA was carried out by DAPI (diamidino-2-phenylindole).

The direct involvement of plant RBRs in the control of cell division, endoreduplication and differentiation was shown by reduction of the expression of NtRBR1 gene through virus-induced gene silencing in tobacco plants (Park et al., 2005). The transcription of E2F and S-phase genes such as ribonucleotide reductase (RNR), proliferating cell nuclear antigen (PCNA), mini chromosome maintenance (MCM), histone H1 and replication origin activation protein (CDC6) was significantly up-regulated in infected tissues of these plants. Leaf cells were observed with enlarged nuclei representing higher ploidy levels. Endoreduplication was also enhanced in transgenic plants over-expressing the E2Fa-DP genes (De Veylder et al., 2002; Kosugi and Ohashi, 2003). In cultured Arabidopsis cells (MM2d cell line) the RNAi-induced down-regulation of AtRBR1 increased G₂-phase cells (Hirano et al., 2008). Dependence of the activation of S-phase entry and cell proliferation on RBR mRNA levels can be also demonstrated in monocot cells. Fig. 4 presents an increase in the frequency of S-phase cells incorporating 5-ethynyl-2′-deoxyuridine (EdU) and biomass production in rice callus culture after antisense transformation of the OsRBR1 gene. A recent publication (Kotogány et al., 2010) describes the use of EdU labelling of S-phase cells in plant tissues.

Fig. 4.

The transcript level from the rice retinoblastoma-related gene (RBR) influences the number of S-phase cells and biomass in transgenic rice cell cultures. Over-expression of the rice RBR1 gene reduces while down-regulation of this gene increases the frequency of DNA-synthesizing S-phase cells labelled with 5-ethyl-2′-deoxyuridine (EdU) as shown in yellow and green. The over-expressing culture (line 80) could not be cultured for a prolonged time.

PHOSPHATASE PP2A REGULATORY SUBUNIT, A PHOSPHOPROTEIN WITH CALCIUM-BINDING MOTIVES INTERACTS WITH PLANT RBR PROTEINS

Post-translational modifications of regulatory proteins by both kinases and phosphatases are essential in controlling molecular and physiological responses. Protein phosphatases regulate practically every step of the mitotic cycle (reviewed by Bollen et al., 2009). PP1, PP2A, PP4 and the dual specificity phosphatases Cdc25 and Cdc14 are involved in regulatory processes. During mitosis, PP1 and PP2A control mitotic kinases (Cdk1, Nek2A, Plk1, Aurora A), and dephosphorylate specific pools of mitotic kinase substrates. Upon mitotic exit, they function as guides to promote the controlled removal of mitotic phosphorylations. In animal cells, PP1c interacts with pRB and directly dephosphorylates the protein (Vietri et al., 2006; Kiss et al., 2008). In addition, PP2A was reported to be implicated in the dephosphorylation of RBRs, particularly upon oxidative stress (Cicchillitti et al., 2003). A PP2A regulatory subunit (PR70) was shown to associate with pRb and mediate its dephosphorylation (Magenta et al., 2008). In chondrocytes, FGF promoted the dephosphorylation of p107 by inducing an association between PP2A and p107 (Kolupaeva et al., 2008). Although phosphorylation of RBRs by CDKs in plants is well documented (Nakagami et al., 1999), up till now not a single report has been published on plant phosphatases that are responsible for the dephosphorylation of phospho-RBRs.

In yeast two-hybrid screens, using full-length MsRBR1 protein of alfalfa as bait, a Medicago truncatula interactor was identified, which showed a significant level of similarity to PP2A protein phosphatase B′ regulatory subunits (39 % identity and 57 % homology with murine PR59 PP2A regulatory subunit). Further pairwise assays revealed that this protein discriminated between the rice RBR proteins in interaction strength. Like the alfalfa RBR protein, OsRBR1 showed strong association with the PP2A regulatory subunit, whereas OsRBR2 failed to interact with this prey (Lendvai et al., 2007). Similar selectivity was reported for mammalian pocket domain proteins. Murine RBs were shown to interact with the PR59 PP2A B′ regulatory subunit in a differential manner; the p107 protein was the only one to show a strong association but pRB did not (Voorhoeve et al., 1999). The PR59-associated PP2A complex dephosphorylated p107 in vivo, and over-expression of the phosphatase regulatory subunit in U2OS cells resulted in inhibition of cell-cycle progression and accumulation of the cells in G₁ phase (Voorhoeve et al., 1999). This analogy has also supported the suggestion that plant RBRs belonging to the RBR1 or RBR2 subfamilies have different roles in the regulation of plant cell division and differentiation. Previous results have also suggested that PP2A has an important cell division regulatory role, since its activity contributes to the control of mitotic kinases and microtubule organization in alfalfa (Ayaydin et al., 2000).

Orthologues of all PP2A subunits have been described in plants. The PP2A catalytic subunits are localized to various cell compartments and play diverse functions ranging from metabolism to cell-cycle control. In Arabidopsis, the catalytic subunit isoforms of PP2A are encoded by five genes, each of which appears to be expressed in all tissues, albeit at different levels (Arino et al., 1993; Casamayor et al., 1994; Pérez-Callejón et al., 1998). Database searches in the rice genome revealed the presence of five PP2A catalytic subunits (Os02g0217600, Os03g0167700, Os3g0805300, Os06g0574500 and Os10g0410600) and a single PP2A A regulatory subunit (Os09g0249700). Since the active holoenzyme in general consists of one catalytic, one A- and one B-type regulatory subunit, the correct characterization of this RBR-dephosphorylating plant PP2A phosphatase holoenzyme requires the identification of all PP2A subunits.

Sequence analysis of the OsRBR1-interacting OsPP2A B′ regulatory protein (encoded by the Os10g0476600 gene) revealed that it contains EF-hand domains potentially regulating its function by Ca²⁺-binding. Several examples in the literature have already demonstrated such regulation; experiments showed that the PR70 member of the PPP2R3 family of human regulatory subunits targets protein phosphatase 2A to Cdc6 (Davis et al., 2008). Two functional EF-hand calcium-binding motifs mediate the calcium-enhanced interaction of PR70 with PP2A. Another report suggested that Ca²⁺ binding to EF-hand 1 of the B′/PR70 subunit was able to further increase PP2A affinity for certain substrates like Thr-75-DARPP32 protein, or to alter the orientation of phospho-serine or -threonine in the active site of the catalytic subunit (Ahn et al., 2007). In the yeast two-hybrid experiments it was possible to demonstrate that the interaction between the OsRBR1 and OsPP2A B′ proteins needs an intact pocket domain of the RBR and the presence of the EF-hand domains on the regulatory subunit (P. Yu et al., BRC Szeged, Hungary, unpubl. res.). Such a finding may support the hypothesis that the dephosphorylation of plant RBR proteins might be influenced by the increase in intracellular Ca²⁺; thus it could respond to extracellular stimuli, e.g. environmental stress factors.

AN OVERVIEW OF POTENTIAL LINKS BETWEEN HORMONAL AND DEVELOPMENTAL SIGNALLING AND CELL-CYCLE REGULATORS

Over the past 20 years, starting with the report on the cloning of the Cdc2 gene from pea (Feiler and Jacobs, 1990), biochemical and genetic approaches involving recombinant DNA have led to a substantial increase in understanding the functions of the key cell-cycle regulators and have provided good evidence that their role is integrated into the developmental programme of plants in a complex way (de Jager et al., 2005; Maughan et al., 2006; De Veylder et al., 2007; Busov et al., 2008). However, the picture is far from complete. Figure 1 summarizes the previously discussed data, demonstrating selected cases where plant hormones interact with regulatory components of cell-cycle control. These growth regulators can directly influence events mediated by cyclin, CDK, KRP, E2F/DP and RBR. It is possible to see the pivotal role of auxins that can also be realized through pathways regulating organ formation or plant structure. This summary outlines the links with Ca²⁺ signalling. The proposed scheme emphasizes the presence of the RBR protein and its phosphorylated form in the G₂/M cell-cycle phase and cites examples for interplay between regulators of plant structure and cell proliferation or differentiation.

Studies in root hair development offer an excellent example for a joint component in the regulation of DNA replication and epidermal cell fate (Caro et al., 2007). Root hair initiation from epidermal cells occurs in trichoblast progenitors that do not express the homeobox transcription factor GL2 (GLABRA2). Activity of this gene is controlled by the GEM (GL2expression modulator) protein interacting with a DNA replication licensing factor, CDT1. Over-expression of GEM represses GL2 and results in ectopic root hair formation, while GEM acts as a repressor of cell division. The armadillo BTB Arabidopsis protein 1 (ABAP1) has been described as a regulator of cell-cycle progression in leaves by integrating plant developmental signals with DNA replication and transcription controls (Masuda et al., 2008). The 2- and 5-fold reduction in ABAP1 levels in Arabidopsis plants caused an increase in the expression of AtCDT1 genes, resulting in stimulation of rosette and leaf growth. ABAP1 is a nuclear G₁/early S-specific protein that interacts with the transcription factor AtTCP24 and this complex binds to promoters of AtCDT1 genes. The authors suggest that ABAP1 acts as a member of a negative feedback loop to control DNA replication during leaf development.

After fertilization of an egg cell, a series of cell divisions is accompanied by early specialization of cells and, in late globular-stage embryos, the primary shoot and root meristems are laid down as the main centres of cell division and starting points for the formation of differentiated cells. The transition from the meristematic zone to the elongation and the differentiation zone either in the root tip or the basal region of leaf depends on the co-ordination between proliferation and differentiation processes. This cell fate determination is under the influence of plant hormones such as auxin, brassinosteroids, cytokinins and gibberellin. In addition ABA, ethylene and jasmonic acid are considered as stress-responsive regulators. Wolters and Jürgens (2009) reviewed the essential proteins and corresponding genes involved in hormone actions and outlined hormonal regulation in meristem functions and in starting differentiation.

The interplay between the basic cell-cycle machinery and auxin could be clearly demonstrated in the regulation of lateral root initiation. Auxin accumulation in the pericycle cells primes lateral root formation through the induction of cell division (see review by Péret et al., 2009). D-type cyclins represent rate-limiting factors in G₁–S phase transition during cell-cycle progression. The loss-of-function mutation in Arabidopsis CYCD4;1 caused a reduction in the number of enlarged pericycle cells (Nieuwland et al., 2009). In the absence of this kind of cyclin D, lateral root density was lowered and this defect could be restored by auxin treatment. Over-expression of another D-type cyclin (CYCD3;1) resulted in enhanced auxin response and increased the lateral root density in the presence of 0·1 or 1 µm naphthalene acetic acid (De Smet et al., 2010). The activation of cell division is an essential but not sufficient prerequisite for lateral root initiation. This was clearly demonstrated by over-expression of the CYCD3;1 gene in the solarity root1 (srl1) mutant (Vanneste et al., 2005). In this mutant background, cell division was induced without lateral root initiation. Based on microarray transcript profiling data, the authors list a number of genes involved in cell cycle, auxin signalling, transport, conjugation and biosynthesis genes among the lateral root initiation genes. Primary auxin-responsive genes out of cell-cycle genes such as CYCA2;4 and CDKB2;1 genes have promoters with auxin-responsive regulatory elements. The heterodimeric transcription factor E2FA/DPA is one of the key activators of S-phase entry. Lateral root density was found to be reduced in transgenic Arabidopsis plants over-expressing these transcriptional factors (De Smet et al., 2010). In contrast, auxin-induced lateral root formation was increased in this line. The authors concluded that enhancement of the division capacity in pericycle cells results in the formation of new lateral roots in an auxin-dependent manner. The SHORT ROOT (SHR) and SCARECROW (SCR) transcription factor network controls root patterning through identification of asymmetric cell division. The recent microarray analysis of cell-type-specific transcriptional effects has identified CDKB2;1, CDKB2;2 and CYCD6;1 as key downstream targets for SHR/SCR as regulators of the cell-cycle machinery (Sozzani et al., 2010). These studies provide experimental support for a direct molecular link between these key developmental regulators and a cell-cycle gene.

Beside auxins, ethylene can directly modulate root growth by modifying cell division. As an example, this can be concluded from studies on the CULLIN3 knockdown mutant of Arabidopsis (Thomann et al., 2009). The CULLIN3-based ubiquitin protein ligase determines the stability of ACS5, a member of the 1-aminocyclo-propane-1-carboxylic acid synthases (ACS) that catalyses a rate-limiting step in ethylene biosynthesis. The lack of the CUL3A/B function can contribute to the stabilization of the ACS5 protein and the induction of ethylene. This change resulted in a reduction in the root meristem size and cell number. The authors suggest premature exit of cells from the meristem and transition to cell expansion.

The functionality of shoot apical meristem (SAM) is derived from co-ordinated actions of several plant hormones with a pivotal role conscribed to auxins, cytokinins and gibberellins (see review by Vernoux et al., 2010). Products of the homeobox genes WUSCHEL (WUS) and SHOOT MERISTEMLESS (STM) control SAM organization by specifying stem cell niche or allowing the proliferation of meristem cells (Lenhard et al., 2002). In transgenic Arabidopsis plants with ectopic expression of the STM gene (AINTEGUMENTA promoter/STM), the CycB1;1::CDBGUS reporter gene as a mitotic marker indicated the promotion of cell division in leaf primordial cells. Wu et al. (2005) characterized the STIMPY (STIP) homeobox gene, which is also required for the growth of SAM, as manifesting the WUS function. The expression of histone H4, as an S-phase-specific marker was not detected in the apical region of mutant stip seedlings. The authors reported that sucrose treatment could restore the stip mutant phenotype.

Leaf initiation on the flanks of the SAM and early development rely on the division of cells in regions with increased auxin concentration (Scarpella et al., 2010). The direct involvement of an auxin-binding protein (ABP1) in the alteration of cell-cycle gene expression and leaf size/morphology was demonstrated by transient reduction of this protein (Braun et al., 2008). Conditional repression of ABP1 activity in Arabidopsis leaves caused a reduction in cell size in leaves and these cellular changes are linked to lowered transcript levels of cyclin D genes (CYCD3;1, CYCD6;1). The E2FC and RBR genes were activated in the same tissues. The authors concluded that the ABP1 protein is required for entry to the division cycle that was shown in other experimental systems such as tobacco BY2 cultured cells (David et al., 2007). Proteolysis of cell-cycle control proteins such as mitotic cyclins is dependent on ubiquitination that can be carried out by E3 ligases as components of the anaphase-promoting complexes (APCs). Serralbo et al. (2006) characterized the Arabidopsis HOBBIT (HBT) gene encoding a homologue of the CDC27 protein as a subunit of APC. Analysis of leaves of Arabidopsis plants lacking the HBT functions revealed that both cell division and later elongation were impaired. Importantly, the authors observed the rescue of division in epidermal cells by the underlying T mesophyll cells. The role of HOBBIT function in the regulation of cell division and elongation can differ in roots and leaves. HBT reduction leads to a decrease in endoreduplication in roots, whereas leaf clones showed decreased mitotic activity.

Horváth et al. (2006) described significant alterations in organ size depending on down- or up-regulation of the potato EBP1 gene both in potato and Arabidopsis plants. This plant gene shows functional and structural homology to human EBP1, the ErbB-3 epidermal growth factor receptor-binding protein. The EBP1 function depends on auxin and, in over-expressing potato plants, CYCD3;1 expression in the meristem increased in comparison with the wild type. The expression level of the G₂/M-specific CDKB1;1 gene was found 2–3 times higher in young leaves. In transgenic Arabidopsis plants the endogenous RBR1 protein level was negatively regulated by EBP1 protein. Figure 1 presents a cascade of auxin-responsive proteins (AXR1, ARGOS, AINTEGUMENTA) that can modify plant organ size by elevating the expression of the CYCD3;1 gene (Hu et al., 2003). AXR1 is a component of the RUB1-activating enzyme, which is involved in an early step of auxin signalling (Leyser, 2006). Over-expression of ARGOS (auxin-regulated gene involved in organ size) genes increased leaf size by production of more cells. In these 35S-ARGOS Arabidopsis plants the expression of ANT and CYCD3;1 genes was enhanced in tissues. The ANT gene encodes a transcription factor of the AP2 domain family and its ectopic expression resulted in an increase in cell numbers and enlarged organs (Mizukami and Fischer, 2000). The authors proposed a role for ANT in maintenance of the meristematic competence of cells through stimulation of cyclin D3 gene expression.

CONCLUDING REMARKS

Despite impressive progress in the discovery of plant genes and encoded proteins functioning as key elements of the cell-cycle machinery, there is a need for a deeper knowledge of how cell division contributes to organ growth and the completion of the developmental programme under normal or suboptimal environmental conditions. The plant-specific nature of regulation during cell-cycle progression can be clearly recognized in the mode of action of plant hormones regulating the expression of several division control genes, including cyclins, CDK and CDK inhibitor protein genes (KRP). Hormonal modification of cell division activity can be partially linked to Ca²⁺-sensitive molecular pathways based on post-translational modifications such as protein phosphorylation or dephosphorylation. In the future, extensive studies on the role of Ca²⁺ signalling in cell-cycle events will be expected to substantially improve our understanding of how cell division is dependent on developmental and environmental factors. The central significance of RBRs not only in the regulation of cell division, but also in plant organ development is now widely demonstrated in various experimental systems. RBR functions are regulated by phosphorylation, therefore studies on the roles of CDKs and phosphatases will highlight novel pathways in cell-cycle control. The interactions between E2F transcription factors and RBRs are considered basic molecular regulatory processes in G₁–S phase transition. Detection of the alfalfa RBR protein and its phosphorylated form in G₂ and mitotic cells extends the present models and supports future efforts to clarify the functional role of plant RBRs during mitosis. The present overview has cited several examples for the influence of developmental genes such as SHORT ROOT (SHR), SCARECROW (SCR), SHOOT MERISTEMLESS (STM), AINTEGUMENTA (ANT) and auxin-regulated gene in organ size (ARGOS) that modulate cell division activity through cell-cycle control elements. As presented in Fig. 1, the elements of the cell-cycle control machinery primarily serve as acceptors of signals reflecting the hormonal or metabolic status of cells and participate in the determination of actual cell division activity according to the developmental programme.