Switching the Cell Cycle. Kip-Related Proteins in Plant Cell Cycle Control

Cell cycle modes during development. During the growth of a multicellular organism, different cell cycle programs are executed. Typically, in young and less differentiated tissues, a proliferative cell cycle mode occurs by which more cells are generated. As cells differentiate, the rate of cell proliferation decreases and eventually, cells exit the cell cycle. However, before complete withdrawal from the cell cycle, and along with differentiation, many plant and animal cells switch from a mitotic cell cycle to an endoreduplication cycle, in which the nuclear DNA becomes replicated without subsequent nuclear and cellular division, leading to polyploid cells. Cell cycle progression is controlled by an evolutionarily conserved molecular mechanism. A central role is played by kinase complexes, which in their minimal configuration consist of a Ser/Thr kinase (the cyclin-dependent kinase [CDK]) and a regulatory cyclin subunit. CDKs phosphorylate a plethora of substrates, thereby triggering the transition from one cell cycle phase into the next one. The sequential and transient activation of different CDK-cyclin complexes dictates the unidirectional progression through the cell cycle. Because of its importance in growth and development, the CDK-cyclin activity must be strictly controlled. In addition, mechanisms must be operational to ensure a correct exit of the cell cycle in response to antimitogenic stimuli. In yeast and mammals, one of the major regulators of CDK activity are CDK inhibitory molecules (CKIs) that bind and inhibit or sequester CDKs. Recently, putative orthologs of CKI proteins have been identified in plants as well. In this Update, we review the current knowledge on the biochemical properties of plant CKIs and discuss their physiological relevance during plant growth and development. Similar to animals, progression through the cell cycle in plants is regulated by CDKs (De Veylder et al., 2003; Inzé, 2005). In Arabidopsis (Arabidopsis thaliana), at least two classes of CDKs are involved in cell cycle regulation: the A-type CDKs that are represented by only one gene in the model species Arabidopsis (designated Arath;CDKA;1) and the B-type CDK family that has four members, grouped into the B1 (Arath;CDKB1;1 and Arath;CDKB1;2) and B2 (Arath;CDKB2;1 and Arath;CDKB2;2) subclasses (Vandepoele et al., 2002). A-type CDKs display kinase activity from late G1 phase until the end of mitosis, suggesting a role for this particular CDK at both the G1-to-S and G2-to-M transition points (Magyar et al., 1997; Porceddu et al., 2001; Sorrell et al., 2001). A central role for CDKA;1 in controlling cell number has been demonstrated using transgenic tobacco (Nicotiana tabacum) plants with reduced A-type CDK activity (Hemerly et al., 1995). The requirement for Arath;CKDA;1 at least for entry into mitosis has been demonstrated as well by cdka;1 null mutants that fail to progress through the second mitosis during male gametophytic development (Nowack et al., 2005). The group of B-type CDKs displays a peak of activity at the G2-to-M phase transition only (Magyar et al., 1997; Porceddu et al., 2001; Sorrell et al., 2001), suggesting that they play a role at the onset of, or progression through, mitosis. Correspondingly, cells of plants with reduced B-type CDK activity arrest in the G2 phase of the cell cycle (Porceddu et al., 2001; Boudolf et al., 2004a). Although titration of CDK activity by the expression of dominant negative versions of both A- and B-type CDKs resulted in cell cycle defects, no extra cell divisions were stimulated by the overexpression of wild-type Arath;CDKA;1, Arath;CDKB1;1, and Arath;CDKB1;2 alleles in plants (Hemerly et al., 1995; Schnittger et al., 2003; Boudolf et al., 2004a). This observation is consistent with the view that a cofactor is required for full CDK activity, i.e. a cyclin. Cyclins are regulated both transcriptionally and posttranslationally, mainly by controlled protein degradation. By regulating the abundance of specific cyclins, the CDK activity is precisely tuned and targeted to substrates in a spatial and temporal manner. The plant cyclin gene family is very complex. For instance, the Arabidopsis genome codes for at least 49 different cyclins (Vandepoele et al., 2002; Wang et al., 2004) that are classified into seven different subclasses (A, B, C, D, H, P, and T). To date, only a few members of the A-type, B-type, and D-type cyclins have been characterized. With a few exceptions, the expression patterns and activity profiles mimic those of their mammalian counterparts. A-type cyclins are important from S until M phase, while B-type cyclins primarily control the G2-to-M transition. D-type cyclins, whose expression is mainly correlated with the proliferative status of cells, are presumed to drive cells through the G1-to-S checkpoint in a mitogen-dependent manner (De Veylder et al., 2003; Inzé, 2005). In contrast to animals, some evidence points to an additional function of plant D-type cyclins at the G2-to-M transition (Schnittger et al., 2002; Kono et al., 2003; Koroleva et al., 2004). In addition to binding of cyclins, CDK activity is regulated by docking of small proteins, generally known as CKIs, which have been found to induce cell cycle arrest or to delay cell cycle progression in response to intracellular or extracellular signals. CKIs have been identified in many different organisms, and, although all of them display CKI activity, they control a broad spectrum of often species-specific physiological processes. For example, in budding yeast (Saccharomyces cerevisiae), three CKIs have been described, Pho81, Far1, and Sic1 (Mendenhall, 1998). Pho81 inhibits a CDK-cyclin complex that controls gene expression under low-phosphate conditions; Far1 binds and inactivates G1 CDK complexes to mediate pheromone-dependent cell cycle blockage; and Sic1 plays a role in the timing of S-phase onset. In fission yeast (Schizosaccharomyces pombe), the CKI Rum1 is structurally and functionally related to Sic1, inhibits mitotic CDKs, and plays a central role in the regulation of the G1 phase. In mammals, based on shared structural features and biochemical functions, CKIs have been divided into two major classes, the INK4 and the Kip/Cip class (Sherr and Roberts, 1999). Members of the INK4 family (p15INK4b, p16INK4a, p18INK4c, and p19INK4d) are structurally similar to the Pho81 inhibitor of budding yeast and are characterized by the presence of multiple ankyrin-type repeats for CDK binding. They bind and inhibit a small subset of CDKs (CDK4 and CDK6) that are primarily responsible for passage through G1. INK4 protein binding to monomeric CDKs or CDK-cyclin complexes causes allosteric changes that impair cyclin binding or lead to the dissociation of the CDK-cyclin complex, respectively. In contrast, inhibitors of the Kip/Cip family (p21Cip1, p27Kip1, and p57Kip2) bind and inhibit a broader range of CDKs and function in dimeric as well as heterotrimeric complexes with CDKs and cyclins; all share a conserved inhibitory domain at their N terminus. Kip/Cip binding does not dissociate the CDK-cyclin complex but distorts the catalytic ATP-binding center of the CDK subunit. The first plant CKIs were detected in yeast two-hybrid screens performed to identify CDKA;1-associating proteins (Wang et al., 1997; Lui et al., 2000; De Veylder et al., 2001; Jasinski et al., 2002a). Additional plant CKIs have been discovered in silico through genome data mining (De Veylder et al., 2001; Coelho et al., 2005). Overall, the plant CKIs have only low sequence identity to each other and the nonplant CKIs. Interspecies sequence similarity is restricted to a short amino acid region shared between the plant CKIs and the mammalian Kip/Cip inhibitors. Because of this sequence similarity, which suggests that the plant proteins are homologous to the animal Kips, the name Kip-related proteins (KRPs) was suggested for the seven CKIs found in the Arabidopsis genome (De Veylder et al., 2001), but some family members are also known under the names ICK1 (KRP1) and ICK2 (KRP2; Wang et al., 1997; Lui et al., 2000). No Arabidopsis homologs to the INK4 or yeast inhibitors have been identified so far (Vandepoele et al., 2002). Phenotypes of KRP-misexpressing plants. A, Wild-type Columbia plants. B, Plants misexpressing the Arath;KPR2 gene under the control of the shoot meristem-specific STM promoter, resulting in smaller and more elongated leaves than observed for wild-type plants. Plants are at the same developmental stage and magnification as in A. C and D, Rosette leaves from a wild-type Columbia and a transgenic plant misexpressing the Arath;KRP1 gene under the control of the stomatal lineage-specific TMM promoter, respectively. Expression of Arath;KRP1 in TMM cells results in an altered leaf morphology because of reduced leaf cell numbers. E and F, Close-up of the leaves shown in C and D, respectively. Note the enlarged epidermal cells in F. Bars = 1 mm (C and E) and 100 μm (D and F). (C–F are from Weinl et al. [2005], reprinted with permission.) Besides the CDK subunit, KRP binding is also directed by the cyclin subunit. Yeast two-hybrid assays have revealed interactions of Arabidopsis and tobacco KRPs with D-type cyclins, suggesting that KRPs are potential regulators of CDK-cyclinD complexes (Wang et al., 1998; Lui et al., 2000; De Veylder et al., 2001; Jasinski et al., 2002a; Zhou et al., 2002). Furthermore, in vivo binding specificity between plant CKIs and different D-type cyclins has been proven by the observation that the aberrant cell and leaf phenotypes seen upon KRP overexpression can be complemented by co-overexpression of D-type cyclins (Jasinski et al., 2002a; Schnittger et al., 2003; Zhou et al., 2003b). Recently, it was demonstrated that not only D-type but also A-type cyclin-harboring CDK complexes can be inhibited by KRPs in vitro (Coelho et al., 2005); so it seems that the plant CKIs resemble the mammalian Kip/Cip inhibitors, which bind and inhibit a broad range of CDKs, including both A- and D-type cyclin-containing CDK complexes. Like the mammalian Kip/Cip inhibitors, the plant CKIs have low sequence similarity to each other. Detailed analysis identified several sequence elements shared by different KRPs, but only three C-terminally located motifs are conserved in all plant inhibitors (De Veylder et al., 2001). This region of the KRPs shows partial homology with the Kip/Cip protein domain necessary for interaction with the CDK subunit, suggesting that the plant CKI function resides at their C terminus. Indeed, Wang et al. (1998) showed in a yeast two-hybrid interaction assay that the C-terminal domain of KRPs is sufficient for interaction of Arath;KRP1 with Arath;CDKA;1 and Arath;CYCD3;1. Moreover, the functionality of this domain for CDK binding and inhibition was proven in vitro and in vivo (Schnittger et al., 2003; Zhou et al., 2003a). The role of the highly diverse N-terminal plant CKI sequences remains unclear. In Arath;KRP1, the N-terminal region was suggested to negatively regulate CKI function; deletion of this region increased the yeast two-hybrid physical interaction of Arath;KRP1 with CDKs and cyclins, and enhanced the phenotype of Arath;KRP1 overexpression in Arabidopsis (Wang et al., 1998; Schnittger et al., 2003). One possible function of the N terminus could be the regulation of the KRP stability. Arath;KRP2 protein is highly unstable and its degradation depends on the proteasome (Verkest et al., 2005). Indeed, removal of the N-terminal region increased the Arath;KRP1 protein level. However, the mechanism regulating this protein stability remains unknown (Zhou et al., 2003a; Weinl et al. 2005). Yeast and mammalian CKIs are regulated at the transcriptional, translational, and posttranslational levels through mechanisms that affect their abundance rather than their intrinsic activity. Most plant tissues coexpress various KRP genes but at different mRNA levels, suggesting different transcriptionally regulatory mechanisms and possibly distinct roles for the plant CKIs within a single tissue (Wang et al., 1998; De Veylder et al., 2001; Jasinski et al., 2002b; Ormenese et al., 2004). A detailed spatial expression analysis by mRNA in situ hybridizations in the Arabidopsis shoot apex revealed different groups of KRP genes with similar expression patterns. Whereas Arath;KRP4 and Arath;KRP5 expression was confined to mitotically dividing tissues within the shoot apex, other KRP genes could be detected in both dividing and maturing cells (Arath;KRP3, Arath;KRP6, and Arath;KRP7) or exclusively in maturing cells (Arath;KRP1 and Arath;KRP2; Ormenese et al., 2004). These data hint at a function of Arath;KRP1, Arath;KRP2, Arath;KRP3, Arath;KRP6, and Arath;KRP7 during the process of cell cycle exit and onset of differentiation, whereas Arath;KRP4 and Arath;KRP5 might direct specific aspects of the mitotic cell cycle, such as functioning of the checkpoints that control the correct timing of S- and M-phase onset. A role for the KRPs during the regular cell cycle is also suggested by their observed cell cycle phase-dependent temporal regulation (Menges et al., 2005). Transcript levels peak during S-phase for Arath;KRP3 and Arath;KRP5, in G2-phase for Arath;KRP4, during late G2-to-M for Arath;KRP1, and at M-to-G1 for Arath;KRP6. The expression of Arath;KRP2 and Arath;KRP7 is constitutive during the cell cycle. Furthermore, transcript levels of the tobacco NtKIS1a and the Arath;KRP1 accumulated with flower bud and leaf aging, respectively (Wang et al., 1998; Jasinski et al., 2002b). This temporal increase in transcripts during the course of cell cycle arrest and cellular differentiation suggests possible functions for these KRPs in development. KRP mRNA is controlled not only in a spatial and temporal manner, but also through the generation of alternative splicing variants, as illustrated by the NtKIS1 locus that generates two splice variants, NtKIS1a or NtKIS1b (Jasinski et al., 2002b). The splice variant NtKIS1b lacks the most C-terminal motif found in NtKIS1a and other plant CKIs. Consistently, NtKIS1b does not interact with A-type CDKs and D-type cyclins and is unable to inhibit CDK activity in vitro and in vivo. Currently, little is known about the regulation of plant CKIs at the protein level. In mammals, regulation of CKI activity is complex and is accomplished through several mechanisms. Kip/Cip inhibitors can be inactivated through out-titration by CDK-cyclinD complexes. Other mechanisms control the subcellular localization. Kip/Cip proteins have distinct nuclear and cytoplasmic functions, and their cytoplasmatic compartmentalization releases and activates nuclear CDK-cyclin complexes (Coqueret, 2003). However, the best studied posttranslational regulatory mechanism of the mammalian CKIs affects their abundance through ubiquitin-dependent proteolysis. Two alternative proteolytic pathways control p27Kip1 stability (Hengst, 2004). One pathway acts in the nucleus and requires p27Kip1 phosphorylation at Thr-187 by CDK2-cyclinE complexes and subsequent recognition and degradation at the S-phase by the SCFSkp2 ubiquitin-ligase complex. The other one acts at the G1-phase, is independent of Skp2 and Thr-187 phosphorylation, and involves cytoplasmic sequestration of p27Kip1 and its degradation through the recently identified Kip ubiquitination-promoting complex (Hengst, 2004). There is evidence that at least some plant KRPs are regulated through proteolysis. As described above, functional analysis of the Arath;KRP1 the presence of a regulatory motif for protein in its N-terminal domain (Zhou et al., 2003a; Weinl et al., 2005). both and Arath;KRP2 are regulated at the posttranslational during and Arabidopsis leaf development, demonstrated by their in protein levels while their transcript levels (Coelho et al., et al., 2005). In the of Arath;KRP2, protein stability is regulated by the Moreover, in vitro analysis illustrated that Arath;KRP2 is a CDK-cyclin and that its phosphorylation is at least in responsible for Arath;KRP2 proteolysis. Although both Arath;CDKA;1 and complexes phosphorylate Arath;KRP2, the cell cycle phase this occurs the specific ubiquitin-ligase that is responsible for Arath;KRP2 degradation is mechanism of posttranslational regulation has been identified through analysis of the NtKIS1a and NtKIS1b splice (Jasinski et al., 2002b). the NtKIS1b does not interact with and D-type cyclins, NtKIS1b the of NtKIS1a to inhibit CDK activity in The two splice have a different expression Whereas NtKIS1a is during the cell cycle, NtKIS1b transcript levels peak at These with their subcellular that NtKIS1b NtKIS1a inhibition of CDK activity at the G2-to-M transition. However, the mechanism by which this occurs remains to be In animals, CKI function depends on its intracellular localization. CKI p27Kip1 its inhibitory function in the nucleus and entry into the nucleus to be as a control mechanism. In addition, p27Kip1 degradation is precisely regulated and is also at least to some with its intracellular and subcellular of A, expression of the as revealed by protein B, of Arath;KRP1 with from the from the into the cells. C and D, Close-up of the protein is found exclusively in the the to the nucleus and the Bars In animals, regulation of nuclear and of p27Kip1 is complex and involves phosphorylation at and of these phosphorylation is conserved in plant However, the Arath;KRP1 in addition to a also a one is that Arath;KRP1 becomes in the and that a motif in the N terminus of the protein is required for this degradation. be to to the intracellular of CKIs is an important regulatory mechanism in plants as well. plant CKIs function in a manner. proteins between and Arath;KRP1 were found at least two to three cells from their of expression et al., and Other plant cell cycle regulators have also been observed to between cells and A. Currently, it is it is the mRNA or the protein that or this is based on a targeted a mechanism. the functional relevance of this is the one the could be a of of these the other the of Arath;KRP1 the to on a cellular with the and growth in tissues and For instance, during leaf development, epidermal cells have been observed to exit the cell cycle before cells et al., 1999). A mechanism can be in which the of KRPs in the the cell cycle exit of the with that of the during development. In the molecular of the of KRPs remains to be in The observation that several plant CKIs are transcriptionally regulated during development that they share with the mammalian the potential to developmental into the cell cycle Indeed, Arath;KRP1 transcripts are by which with a in CDK activity. Furthermore, Arath;KRP1 expression is by the acid (Wang et al., suggesting that this particular KRP might be in responsible for the growth inhibitory upon acid By contrast, the Arath;KRP2 both in cell and in et al., 2001; et al., 2002). of Arath;KRP2 the of cells into the cell cycle. Arath;KRP2 transcripts have been detected in young at the but not the of the i.e. the at which can In tissues, Arath;KRP2 expression has been observed at both the and with the observation that mainly at the young of the upon the of a Arath;KRP2 expression is in cells the a mechanism by which Arath;KRP2 the of two A role for KRPs in controlling has been by overexpression as illustrated by the observation that Arath;KRP2 overexpression in Arabidopsis results in a in the number of et al., 2002). Kip/Cip inhibitor gene expression has been found to with the onset of endoreduplication et al., 1998; et al., 2000). The is an alternative cell cycle during which DNA is not by mitosis and and often the onset of cell the most mechanisms to increase the cellular DNA in and, although the physiological relevance of the endoreduplication process is are several to that an increase in the DNA cell growth and activity (Schnittger et al., 2003; and Roberts, 2003). In yeast and the onset of endoreduplication with a in CDK activity and 2001; et al., 2001). A similar mechanism is operational in plants because the of endoreduplication in Arabidopsis and the of is by a in CDK activity and 1995; et al., et al., 2005). Recently, KRPs have been demonstrated to in control of this in CDK activity. In CDK activity is inhibited in both mitotically dividing and leaf By contrast, in only the mitotic CDK-cyclin complexes are entry into mitosis but onset and progression through resulting into an increase in DNA The two seen upon or KRP overexpression can be by that KRPs have a binding the CDK-cyclin complexes that control the checkpoint or that levels of CDK-cyclin activity are required for entry into mitosis than for entry into The of KRPs to the onset of the in dividing tissues was by the specific overexpression of Arath;KRP2 in tissues, an inhibition of mitotic CDK activity and a onset of endoreduplication (Verkest et al., 2005). low levels of Arath;KRP1 in cells or its specific expression in the mitotically dividing stomatal cells increased levels et al., 2005). of KRPs controlling the switch between the different cell cycle A, In cells, B-type CDKs phosphorylate KRPs, triggering their In addition, phosphorylation might the of KRPs, with their binding to A-type CDKs. B, In cells to B-type CDK activity resulting in a of the KRPs, which bind and inhibit A-type CDK-cyclin complexes with a role in mitosis. The KRP is not to inhibit as well the CDK-cyclin complexes S-phase cells to the C, During cell cycle KRP expression is in addition to entry into mitosis, CDK-cyclin complexes controlling the entry into S-phase resulting in a complete cell cycle In animals, CKIs might also have functions the cell cycle, such as in differentiation, and cell is no evidence that plant CKIs also function the cell cycle. and constitutive overexpression of Arath;KRP2 not cellular differentiation as illustrated by the timing of stomatal differentiation patterns with to leaf development (De Veylder et al., 2001). onset of endoreduplication by Arath;KRP1 expression in cells does not with the of cell et al., 2005). However, of Arath;KRP1 in Arabidopsis cell (Schnittger et al., 2003). This phenotype seems to be to the developmental of because for other cell no cell phenotypes have been observed upon KRP the it is not the observed cell phenotype is with a endoreduplication with cell as a of a between DNA and cell additional are required to KRPs control cell have about a increase in of CKIs in plants. regulatory pathways are with KRPs functioning as cell cycle KRPs might be important for CDK activity within dividing cells, as well as in the transition between different cell cycle such as endoreduplication or cell cycle exit KRPs could possibly play a central role in cell cycle progression with developmental as well as to be For instance, KRPs are very conserved these proteins share a the CDK and a developmental or physiological for the many different KRP genes within an in is the KRP abundance and have recently been that to these at a and level. In addition, KRPs from different plant species have been the for a and leading to the of in KRP function in plants. and the members of their and for and on the and De for in