The mitotic cell cycle consists of alternating rounds of DNA replication (which occurs during the S phase) and chromosome segregation (mitosis or M phase) interrupted by gaps known as G1 (the interval before S phase) and G2 (the interval after S phase). Events that occur in each phase are regulated to ensure that DNA is replicated only once in each cell cycle and that each daughter cell ends up with a complete complement of the genome. Regulation of phase transitions is principally achieved through the action of a host of cyclin-dependent kinases (CDKs) and their corresponding regulatory cyclin proteins. Although the basic cell cycle machinery is highly conserved among all eukaryotes, there are a number of important differences in cell cycle control between higher plants and other eukaryotes. In this issue of The Plant Cell, we highlight three articles related to different aspects of the cell cycle that focus attention on unique characteristics of plant cell cycles and point the way to new discoveries in cell cycle control and organelle replication.
Mitosis usually progresses into cytokinesis, resulting in cell division and the production of two new daughter cells. Our first stop on the cell cycle tour is to view an exception to this rule in plant cells: the development of syncytia in endosperm tissue, which is the focus of work presented by Boisnard-Lorig et al. (pages 495–509). Syncytia, which are typical of plant endosperm tissue, are cells with multiple nuclei that originate from the fusion of multiple cells or from single cells in which DNA replication and nuclear divisions proceed without cytokinesis. Unlike other eukaryotes, flowering plants typically undergo double fertilization: the pollen tube delivers two male gametes to the embryo sac; one fuses with the egg to form the zygote, and the other fuses with the two polar nuclei of the diploid central cell to produce a triploid endosperm nucleus. Division of the endosperm nucleus gives rise to endosperm tissue, which surrounds the developing embryo and plays an important role in the nutrition of the embryo during embryogenesis and seed germination. In numerous species, the endosperm initially develops as a syncytium that contains up to a few hundred nuclei. Ultimately, the syncytium divides into individual cells in a process called cellularization. Boisnard-Lorig observed mitosis in the endosperm tissue of developing Arabidopsis seedlings via constitutive expression of a histone 2B::YFP fusion chromatin marker. The authors observed the development of three spatially distinct domains with syncytial endosperm that exhibit differences in the activity of cell cycle control parameters. Thus, spatial regulation of cell cycle control genes may play a role in the development of the endosperm. For example, it is shown that endoreduplication is likely limited to one of the three endosperm domains, the chalazal endosperm, because this domain accumulates very few nuclei and these nuclei are larger and appear to contain a much greater quantity of chromatin than the nuclei present in the other two domains. Endoreduplication is the most common mechanism of polyploidization in plants, a widespread phenomenon among higher plants and of particular importance in many crop species. Mechanisms of endoreduplication in plants are still poorly understood (Joubès and Chevalier, 2000), and the work of Boisnard-Lorig et al. may point the way to further knowledge of this process.
The second stop on our tour is at a spindle assembly checkpoint in the M phase of embryo cells of the brown alga Fucus. Corellou et al. (pages 585–598) studied aspects of the embryonic cell cycle in Fucus spiralis, which produces large populations of synchronously developing external zygotes that are easy to manipulate and observe experimentally. In animals, there are major differences in cell cycle control between somatic cells and embryo cells. The somatic cell cycle is tightly regulated via a series of “checkpoints,” typically involving CDKs, which monitor cell cycle progression, ensuring that only one round of DNA replication occurs per cycle and preventing mitosis until DNA synthesis and repair are complete. In contrast, checkpoints involving tight regulation of CDK activity appear to be less stringent or absent from early embryonic cells in animals. From experiments using the drugs nocodazole (which inhibits mitotic spindle formation) and olomoucine (a specific inhibitor of CDK activity), Corellou et al. show that, similar to somatic cell cycles, the Fucus zygote cell cycle includes a spindle assembly checkpoint targeted at CDK activity. This checkpoint appears to operate by maintaining high levels of CDK activity, and the authors suggest that inactivation of CDKs may be required for cells to undergo chromatin decondensation and exit mitosis. Previous work by this group (Corellou et al., 2000) showed that Fucus zygotes also contain a DNA replication checkpoint. In the present work, the authors show that the initial embryonic cell cycles of Fucus resemble the somatic rather than the early embryo cell cycles of animals in a number of other important characteristics as well. For example, animal cells in early embryogenesis undergo rapid cell cycles that consist only of S and M phases. Corellou et al. show that, as in somatic cells, the Fucus embryonic cell cycle consists of well-defined G1, S, G2, and M phases. Furthermore, CDKs show complex regulation involving translational control and post-translational control via tyrosine phosphorylation. Complex regulation of CDKs is another characteristic of somatic cells that is lacking in animal embryo cells, in which CDK control over the switch between the S and M phases appears to be regulated principally by periodic synthesis and degradation of cyclin B, a positive regulator of CDK activity.
Although it is tempting to view the Fucus system as a model for photosynthetic organisms, including higher plants, it must be remembered that brown algae (phylum Phaeophyta) belong to the division of Stramenopiles (also called Chromista), a group of protists that includes the diatoms (Bacillariophyta), yellow-green algae (Xanthophyta), and water molds (Oomycota). Phaeophytes and other stramenopiles, although traditionally classified as plants, actually are not closely related to green plants and occupy their own division within the kingdom Protista (or perhaps they may be placed in a separate kingdom called Chromista). Eukaryotic phylogenies show that animals and fungi may be more closely related to plants than are brown algae and other stramenopiles. Thus, work with Fucus may not provide information about higher plant embryogenesis and must be viewed with caution in this context. Rather, this work provides valuable information for comparative studies of eukaryotes, including plants, animals, and fungi. It remains to be seen whether characteristics of cell cycle control in Fucus are shared with plants and/or other eukaryotes.
In one of the most detailed analyses of the embryonic cell cycle of higher plants to date, Sauter et al. (1998) presented reverse transcriptase–mediated polymerase chain reaction analysis of the expression of three cyclin genes and the CDK cdc2 in sperm, egg cells, and other cells of the embryo sac isolated from maize plants and in developing zygotes produced via in vitro fertilization. Whereas cdc2 transcripts were expressed in all gametic cells and in the zygote throughout development to the two-cell stage at 48 hr after fertilization, the cyclin genes exhibited cell-specific expression in the embryo sac and differential expression during zygote development. All cyclins were transcribed de novo after fertilization, with a high degree of expressional regulation during the first embryonic cell cycle. This is in marked contrast to findings in most animal cells, in which the egg provides a store of mRNAs, including cyclins, which control regulation of the initial embryonic cell cycles. Sauter et al. (1998) hypothesized that this zygotic regulation of the first cell cycle may contribute to a greater adaptive ability of plants during early embryogenesis, unlike animal embryos, in which patterns of cell division are more rigidly fixed before fertilization. This analysis also might lead to the prediction that cell cycle control in higher plant zygotes will show some degree of similarity to that in the Fucus model, because Corellou et al. have shown that Fucus also exhibits a high degree of zygotic regulation of the cell cycle during early embryogenesis.
The final stop on this brief tour of the cell cycle is to investigate another feature unique to plant (and algal) cells: plastids and plastid division. Plastids are believed to have originated from an endosymbiosis event in which an early photosynthetic prokaryote invaded or was engulfed by a primitive eukaryotic host (Margulis, 1970; Gray, 1992). Plants contain a number of different types of plastids in addition to chloroplasts: amyloplasts, which accumulate in seeds and tubes; leucoplasts, which are found principally in petals; and chromoplasts, which are found in fruits and flowers. All plastids differentiate from small, colorless proplastids, which are found in dividing cells in meristematic tissue. However, division of mature plastids also occurs in developing cells, and dumbbell-shaped dividing chloroplasts have been observed in young leaves of a number of species (Pyke, 1999).
Plastid division appears to be driven by the formation of a contractile ring called the plastid-dividing (PD) ring (reviewed by Pyke, 1999). A similar dividing ring controls bacterial cell division, a major component of which is FtsZ, a filamentous protein with structural similarity to tubulin. A family of FtsZ homologs was found in Arabidopsis, and transgenic plants constitutively expressing antisense AtFtsZ constructs have mesophyll cells with fewer enlarged chloroplasts compared with wild-type plants, suggesting that FtsZ proteins form part of the PD ring in plant cells (Osteryoung et al., 1998). The PD ring is a double or triple ring. Osteryoung et al. (1998) identified chloroplast-targeted and nontargeted forms of FtsZ in Arabidopsis and hypothesized that these forms make up part of the inner and outer PD rings, respectively.
Miyagishima et al. (pages 707–721) present high resolution ultrastructural analysis of the outer PD ring in dividing chloroplasts of the red alga Cyanidioschyzon merolae. Negative staining of isolated dividing chloroplasts showed that the outer ring consisted of 5-nm-diameter filaments, which formed a rigid structure that did not disassemble in 2 M urea. Immunoblotting of extracted proteins with CmFtsZ-specific antibody suggested that FtsZ was not present in the outer ring, and an unidentified 56-kD protein was isolated as a candidate ring component. The authors previously identified two types of FtsZ protein in C. merolae, one (CmFtsZ2) with similarity to cyanobacterial FtsZ that localized to chloroplast protein fractions and one (CmFtsZ1) that resembles an α-proteobacterial counterpart and is likely to function in mitochondria (Takahara et al., 2000). Antibody specific to CmFtsZ2 was used in the present work. The authors point out the possibility that another FtsZ protein, which does not cross-react with the antibody used, forms part of the outer ring, but their results suggest that the insoluble 56-kD protein is the principal ring component, and insoluble properties have not been reported previously for FtsZ proteins.
However, as in the case of Fucus, it cannot be assumed that green plants necessarily will share features of red algae, which belong to the protist phylum Rhodophyta. In transgenic Arabidopsis plants, expression of either AtFtsZ1 or AtFtsZ2, both of which encode proteins related to cyanobacterial FtsZ sequences, reduces the number of chloroplasts in mature leaf cells from 100 to 1, indicating that both genes are required for chloroplast division in higher plant cells (Osteryoung et al., 1998). It is reasonable to hypothesize that FtsZ2 constitutes part of the outer PD ring, because, unlike FtsZ1, it lacks a chloroplast transit peptide and is not transported into isolated chloroplasts.
The cell cycle of higher plants contains unique features, aspects of which may be ripe for discovery in Arabidopsis. For example, molecular identification of the 56-kD protein component of the outer PD ring in C. merolae will lead to the immediate identification of homologs in the completely sequenced genome of Arabidopsis. It also should be possible to identify all of the CDKs in Arabidopsis and begin a process of systematically investigating their possible functions in cell cycle control and other plant growth processes. The three articles highlighted here should stimulate renewed interest in comparative studies of cell cycle control among higher plants, algae, and other eukaryotes.