Plant growth and proliferation control is coming into a global focus due to recent ecological and economical developments. Plants represent not only the largest food supply for mankind but also may serve as a global source of renewable energies. However, plant breeding has to accomplish a tremendous boost in yield to match the growing demand of a still rapidly increasing human population. Moreover, breeding has to adjust to changing environmental conditions, in particular increased drought. Regulation of cell-cycle control is a major determinant of plant growth and therefore an obvious target for plant breeding. Furthermore, cell-cycle control is also crucial for the DNA damage response, for instance upon irradiation. Thus, an in-depth understanding of plant cell-cycle regulation is of importance beyond a scientific point of view. The mere presence of many conserved core cell-cycle regulators, e.g. CDKs, cyclins, or CDK inhibitors, has formed the idea that the cell cycle in plants is exactly or at least very similarly controlled as in yeast or human cells. Here together with a recent publication we demonstrate that this dogma is not true and show that the control of entry into mitosis is fundamentally different in plants versus yeast or metazoans. Our findings build an important base for the understanding and ultimately modulation of plant growth not only during unperturbed but also under harsh environmental conditions.
Get your grade
or your money back
using our Essay Writing Service!
Progression through the cell cycle is not only a decisive event for a single cell but also of key importance for organ growth in multicellular organisms such as plants.1,2 Moreover, coupled to and overlapping in space and time with proliferation, cell differentiation takes place and thus, a tight control of the cell cycle is one of the foundations of development.3 Thus, not very surprisingly, an elaborated machinery controlling cell-cycle regulation has evolved and overall, many proteins appear to be conserved between humans and plants.4,5 However, there are also clear differences in the repertoire of cell-cycle regulators in plants and functional studies have often not yet been conducted to elucidate the specific role of many regulators.
In metazoans, a switch-like activation of the central cyclin-dependent kinase, Cdk1 (or its homologous proteins, e.g. Cdc2+ or CDC28p) plays one of the most important mechanisms in cell-cycle control.6 Wee1-type kinases, e.g. Wee1 or Myt1, phosphorylate Cdk1-type kinases at Thr14 and Tyr15 (or the homologous positions) and inhibit their activity (Fig. 1A).7 The function of these kinases is opposed by Cdc25 that acts as dual specificity phosphatase and removes these phosphate groups leading to the rapid activation of Cdk1-type kinases. This inhibition of Cdk1 activity by Wee1 and its release by Cdc25 fulfill a fundamental function during metazoan cell-cycle control ensures the unidirectionality of the cell cycle.8,9 The underlying molecular mechanism for this is a wiring of Cdk1 with Cdc25 or Wee1 by positive and antagonistic (double-negative) feedback loops, i.e. Cdk1 activates its activator Cdc25 and inactivates its inhibitor Wee1 (Fig. 1C). Thus, there are only two stable steady states, inactive or active; this bistability generates a biological switch. The transition from one state to the other is thought to be brought about by rising and falling levels of cyclins as activating subunits of CDKs. Moreover, due to the positive feedback wiring, the two steady states are buffered against small changes in cyclin levels, i.e. it takes a much higher concentration of cyclins to switch from G2-phase into mitosis than to stay in mitosis. This property of feed-back wiring, called hysteresis, greatly reinforces the unidirectionality of the cell cycle (Fig. 1A and C).10,11
Cdc25 and the feedback loops sketched above are also major targets of a checkpoint response and interruption of this feedback loop can effectively arrest the cell cycle. For instance, in animals, DNA damage is sensed by ATM and ATR kinases that in turn activate Chk1 and Chk2 kinases which then will phosphorylate and inactivate Cdc25 allowing the cell to repair its damage.12,13 In parallel, Chk1/2 activate Wee1 by phosphorylation and reinforce the checkpoint.
Previously, candidate genes for Cdc25 and Wee1 homologs have been identified in Arabidopsis as well as other plants.14-18 Along with the finding that plants contain Cdk1-like kinases with a PSTAIRE cyclin binding signature, designated CDKAs, which can rescue yeast cdc2/cdc28 mutants,19-22 this has given rise to the notion that the wiring of the regulatory triangle CDKA-CDC25-WEE1 is conserved in plants.
Always on Time
Marked to Standard
Here and in an accompanying publication by Dissmeyer et al.,23 we have probed this notion by a detailed structure-function analysis. Our data demonstrate that the regulatory connection between these three components is not conserved and that plants must have evolved different mechanisms to stably progress through a mitotic cycle and arrest the cell cycle upon DNA damage.
Structural considerations of the putative Arabidopsis CDC25 and the WEE1 homolog
Due to its central role in the animal and yeast cell cycle, plant breeders and cell biologists have intensively searched for a plant homolog of the Cdc25 phosphatase and a putative Arabidopsis CDC25 homolog (At5g03455) was presented.16 A detailed sequence alignment of At5g03455 with Cdc25 phosphatases known to play a key role during cell-cycle regulation shows that Cdc25 proteins are highly diverse in different species but all display a general bimodular organization (Fig. 2A). A large N-terminal domain harbors all the regulatory regions important for the central function of Cdc25 as an integrator of developmental and environmental cues, for instance the phosphorylation sites for Chk1/2 as identified in human Cdc25A and Cdc25C and putative phospho-sites, especially serines, are present around a similar stretch in the other Cdc25 proteins (Fig. 2A).13,24 The actual catalytic rhodanese-like domain is located in the C-terminus of the enzyme with a catalytic cysteine residue in the active site present in all tested sequences.
While the C-terminal catalytic domain of At5g03455 aligned well with all Cdc25 enzymes, we found that the putative CDC25 proteins from Arabidopsis and other plant species do not contain a large N-terminal regulatory region (Fig. 2A). Thus, these putative Cdc25 homologs lack all of the important regulatory motifs present in yeast or metazoans, raising the question how At5g03455 could carry out a central role in plant cell-cycle control.
In animals and yeast, Wee1 kinases are the counterplayers of Cdc25. Wee1 has a key function during G2-M control and an important feature of Wee1 regulation is the phosphorylation by Cdk1 (see above; Fig. 1 C). In a detailed alignment of the WEE1 homolog of Arabidopsis with other Wee1-type kinases revealed that overall, Wee1 sequences are much more conserved than those of Cdc25-like enzymes (Fig. 2A and B). In particular, Arabidopsis WEE1 is similar in length to other Wee1 homologs and all typical protein kinase consensus sites such as DFG, HRD, ATP- and water-binding loops, and multiple catalytic residues forming a putative active site can be found in the catalytic segment.25 Arabidopsis WEE1 also harbors some possible regulative phospho-acceptors at homologous sites in the N-terminal part. This implies that these sites - similar to those found in e.g. yeast, human and Drosophila Wee1 (Fig. 2 B) - might play a similar role in regulating kinase activity and protein stability. For instance, the activation domain plays a crucial role in Wee1-degradation via the ubiquitin proteasome pathway in HeLa cells.26 The C-terminal parts of Wee1 kinases containing the catalytic segment are highly conserved and the active sites can be clearly identified in all homologs.
Despite of the higher overall similarity of Arabidopsis WEE1 with other Wee1 kinases in comparison to the Cdc25 proteins, Arabidopsis WEE1 also lacks some motifs present in animal and yeast Wee1, such as a surface loop conferring a high PEST score predictive of low protein stability that is present in mammalian Wee1, a membrane association motif and a Cdk1-interacting interface (both in human Myt1; Fig. 2B). Importantly, also the serine phospho-acceptor sites for activation via Cdk1 cannot be properly aligned. Thus, Arabidopsis WEE1 and especially the putative CDC25 display significant sequence differences in comparison to their animal and yeast relatives suggesting a highly adapted function in plants.
Assessing the function of the putative Arabidopsis CDC25 under DNA damage conditions
In our accompanying work, we analyzed the growth properties of null mutants of the putative CDC25 gene, i.e. At5g03455, as well as transgenic lines overexpressing the putative CDC25 to high levels. In contrast to modulation of CDC25 levels in animals, both the gain- and the loss-of-function in Arabidopsis did neither alter plant growth nor result in an altered cell-cycle profile in comparison to wild-type plants.23 Consistent with this, wee1 mutants in Arabidopsis were not found to display growth defects.27 However, RNAi-mediated down-regulation of WEE1 expression in tomato resulted in reduced plant growth and a smaller fruit size,28 possibly indicating that WEE1 controls growth in a species-specific manner.
This Essay is
a Student's Work
This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.Examples of our work
In contrast to the not yet clearly defined function of WEE1 in plant growth under unperturbed conditions, WEE1 was found to be required under DNA damaging conditions similar to the role of Wee1 in metazoans (Fig. 3B and D).27,29 This suggested that perhaps a regulation of CDKA;1 by P-loop phosphorylation is restricted to checkpoint activity in Arabidopsis.
To test this hypothesis we analyzed the growth response of cdc25 mutants and CDC25 overexpression lines on medium containing hydroxyurea (HU; Fig. 3B and D). HU induces replication stress by inhibiting the RIBONUCLEOTIDE REDUCTASE (RNR) required for dNTP synthesis.30 This reversibly blocks the replication fork in S-phase and causes DNA fragmentation and chromosomal abnormalities.30,31 If the putative Arabidopsis CDC25 gene has the conserved Cdc25 function, we expected that overexpressing plants should show hypersensitivity towards genotoxic stress similar to the situation in human cells.12,32 However, neither the gain- nor the loss-of-function mutants showed a phenotype distinguishable from the wild type (Fig. 3A-E).
In addition, dephosphomutants of CDKA;1 were generated where Thr14 and Tyr15 were replaced by the non-phosphorylatable amino acids Val and Phe. These lines should functionally mimic the loss of all inhibiting WEE1 activity or the ectopic activation of CDC25 phosphatases. However, expression of this construct under the CDKA;1 promoter could completely rescue the cdka;1 mutants and the generated transgenic plants were indistinguishable from wild-type plants.23
Finally, it was recently shown that the putative CDC25 protein has arsenate reductase activity in vivo.23,33,34 To complicate the situation, the putative CDC25/arsenate reductase might play a role in the detoxification of HU in planta since in other studies mutants were found to be slightly more sensitive to HU and overexpression of the putative CDC25/arsenate reductase was reported to result in plants that were a little more resistant than wild-type plants.35 This is the opposite to what would be expected for a cell cycle function of CDC25.
Taken together, we conclude that cell-cycle control in Arabidopsis is efficiently carried out without one of the most central regulators of the metazoan cell cycle.
Possibilities for plant-specific cell-cycle regulation
Although Arabidopsis does not utilize a CDC25-WEE1 relay module to control entry into mitosis, it appears that the mechanistics of CDKs are conserved since a phospho-mimicry of Thr14 and Tyr15 in the P-loop of CDKA;1 resulted in strongly reduced CDKA;1 activity.23 Similarly, phosphorylation of the T-loop, a highly conserved domain in the activation segment of CDKs, was found to be needed in Arabidopsis as in animals to achieve high levels of kinase activity.36,37 Moreover, it seems more than likely that also Arabidopsis CDKA;1 operates as a biological switch and that the system of CDK regulation results in hysteresis.
To generate hysteresis, a double-negative feedback wiring is necessary and therefore obvious candidates for the implementation of such a wiring in plants are CDK inhibitors (CKIs). Indeed, Arabidopsis contains two classes of CKIs each with multiple members, the ICK/KRP class (INHIBITOR/INTERACTOR OF CDK 1/KIP-RELATED-PROTEIN 1) and the SIAMESE/SMR (SIAMESE RELATED) class.38-41 It is tempting to speculate that the prominent classes of CKIs have a key role in controlling the plant cell cycle.
To test the hypothetical involvement of CKI in plant CDK activity hysteresis, we calculated whether fundamental properties of cell-cycle regulation could also be accomplished based on a different wiring of CDKs. We compared two scenarios. In the first case, hysteresis was generated as expected and previously published when CDK is wired by a CDC25-WEE1 feedback loop (Fig. 1A and C).11 In the second scenario, we replaced the CDC25-WEE1 module by a double-negative feedback-loop of CDKs with CKIs taking into account that CKIs have previously been found to be destabilized by CDK action.42 Indeed, this also generated hysteresis and is consistent with previously obtained data for a CKI feedback loop in yeast yielding a robust cell cycle.43
We have recently found evidence that rising and falling levels of ICK/KRPs might drive entry into S-phases.44 This together with the above calculations raises the intriguing hypothesis that hysteresis at the G2-M transition point might be generated by SMRs. This possibly represents a case of how the biochemical constraint of a switch-like Cdk1-activation was solved by a modulation of a recurrent theme in cell-cycle regulation, i.e. inhibition of CDKs through CKIs versus a regulation through inhibitory phosphorylation.
The authors would like to thank John C. Larkin for helpful comments and critical reading of the manuscript. This work was supported by an ATIP grant from the Centre National de la Recherche Scientifique (CNRS) to A.S.
Figure 1. Computational analysis of the switch-like activation of Cdk1-like kinases. (A) and (B) show steady-state activity of CDKs as a function of cyclin levels. (A) CDK/cyclin activity regulated via inhibitory Tyr15-phosphorylation of the CDK catalytic subunit of the complex. (B) CDK/cyclin activity control is achieved by stoichiometrically acting CDK inhibitors (CKIs). Both switches allow building up inactivated kinase and once a cyclin level has reached a threshold, high levels of kinase activity are rapidly available that can forcefully promote the entry into the next cell-cycle phase. Importantly, a small drop in cyclin levels is not sufficient to change the activity state, thus the system is buffered and once the decision is taken to enter the next cell-cycle phase, this cannot easily be reverted. (C) Double-negative and positive feedback loops targeting the status of inhibitory CDK phosphorylation. CDK activity is governed via inhibitory phosphorylation by WEE1/MYT1 kinases and activatory dephosphorylation by CDC25 phosphatases. CDK can phosphorylate WEE1/MYT1 to inactivate its own inactivator and CDK activates its own activator CDC25 by phosphorylation.11 (D) Double-negative feedback loop of the CDK-CKI module.43 CKIs inhibit CDKs and, in turn, CDKs promote CKI degradation.42
Figure 2. Alignments of Cdc25 and Wee1 amino acid sequences. (A) Alignment of Cdc25 candidate sequences of budding yeast (S. cerevisiae; Sc_Mih1p; Mitotic inducer homolog 1), fission yeast (S. pombe; Sp_Cdc25+), the three human homologs (Hs_CDC25A to C), Drosophila (Dm_Stg; String), and Arabidopsis (At_At5g03455). Nearly all regulatory sites are located in the N-terminal regulatory domain whereas and the C-terminal comprises a rhodanese-like catalytic domain. In the sequence of the putative Arabidopsis CDC25, this N-terminus is lacking but the C-terminus is conserved. Below the alignment, regulating kinases are listed with their phosphoacceptor-sites ("P") on top of the CDC25 sequences. Phospho-sites in red were shown to be required for Wee1 activation. Specific molecular features, domains and conserved catalytically-required residues are highlighted in color. In the alignment, sequence gaps are shown as white dashed lines, amino acids in light grey, related amino acids in dark grey, and conserved sites in black. The length of the domains and overall proteins is shown to scale. For some additional plant species, the sequences for putative CDC25 phosphatases were available and all are lacking the N-terminal regulatory domain (not included in the alignment): Brachypodium (Bradi1g78020, Bradi3g32550), Lotus (chr1.CM0012.260), rice (Os03g0108000), Sorghum (Sb01g030120), and tomato (SL1.00sc05858). (B) Alignment of Wee1 candidate sequences of budding yeast (Sc_Swe1p; S. cerevisiae Wee1), fission yeast (Sp_Wee1+), human (Hs_Wee1), Drosophila (Dm_Wee1), and Arabidopsis (At_WEE1). The homologs of the Wee1-related kinases Myt1/Mik1 are included for fission yeast (Sp_Mik+; Mitotic inhibitor kinase 1), human (Hs_Myt1; Membrane associated tyrosine/threonine kinase 1), and Drosophila (Dm_Myt1). Most of the regulatory sites are located within the N-terminal part (activation domain). Conserved catalytic sites are found in all homologs (e.g. HxD, DFG, etc.). Mammalian Wee1 sequences harbor a surface loop with a high PEST core, domains serving as membrane-anchors and for substrate recognition in the C-terminal. All these features are missing in the Arabidopsis WEE1. Color code as above. 14-3-3: protein-protein interaction-site; APC: destabilization via Anaphase-promoting complex/cyclosome; ATP: ATP-binding site; NE/LS: nuclear export/localization signal; P: phosphoaccptor site; SCF: destabilization via Skp-Cullin-F-box containing complex; substrate: substrate binding-site; water: water pocket; Zn: Zink binding-site. Source of the sequence information: UniProtKB/Swiss-Prot at uniprot.org and PlantsDB, the MIPS plant database for monocots and dicots at mips.helmholtz-muenchen.de/plant/.
Figure 3. Analysis of wee1 and cdc25 mutants. (A) and (B) Root growth of wild-type (Col-0) and wee1 mutants germinated and grown for 10 d on Murashige and Skoog (MS) agar control medium (A) and (B) on MS containing 1 mM hydroxyurea (HU). wee1 mutants were used as positive and hypersensitive controls for genotoxic stress on HU plates and display a strong reduction in root growth (red arrowhead). Bar in (A) = 1 cm. (C) to (G) kinematic analysis of the root growth. Seedlings were grown under the same conditions on MS plates for 10 d after germination (d.a.g.). (C) Ratio of the mean growth rates of (1 mM HU/control) of the intervals 4 to 7 which correspond to linear root growth for Col-0, two independent At5g03455 overexpression (OE) lines ("CDC25-OE"), the At5g03455 mutant ("cdc25"), and wee1. On top of the columns, the values are given in per cent (rounded up). (D) Absolute root growth on control MS and (F) on 1 mM HU. (E) Root growth rates on control MS and (G) on HU (abscissa, 1 = growth from day 1 to day 2, etc.). Error bars represent SD.