BIOLOGICAL OVERVIEW

Cdc2 kinase activity is required for triggering entry into mitosis in all known eukaryotes. Elaborate mechanisms have evolved for regulating Cdc2 activity so that mitosis occurs in a timely manner, when preparations for its execution are complete (see Drosophila Cdc2). In Schizosaccharomyces pombe, Wee1 and a related Mik1 kinase are Cdc2 inhibitory kinases that are required for preventing premature activation of the mitotic program. To identify Cdc2 inhibitory kinases in Drosophila, cDNA clones were screened for rescue of S. pombe wee1- mik1- double mutants from lethal mitotic catastrophe. One of the genes identified in this screen, Drosophila wee, encodes a Wee1 homolog. Drosophlila Wee kinase is closely related to human and Xenopus Wee1 homologs, and can inhibit Cdc2 activity by phosphorylating a critical tyrosine residue. Drosophlila Wee mRNA is maternally provided to embryos, and is zygotically expressed during the postblastoderm divisions of embryogenesis. Expression remains high in the proliferating cells of the central nervous system well after cells in the rest of the embryo have ceased dividing. The loss of zygotically expressed wee does not lead to mitotic catastrophe during postblastoderm cycles 14 to 16. This result may indicate that maternally provided Wee is sufficient for regulating Cdc2 during embryogenesis, or it may reflect the presence of a redundant Cdc2 inhibitory kinase, as in fission yeast (Campbell, 1995).

Studies of a number of different experimental organisms have shown that activation of the conserved eukaryotic regulator, Cdc2 kinase, is required to trigger mitosis. Cdc2 activity is modulated by post-translational phosphorylations and by association with Cyclins and other interacting proteins, providing opportunities for an array of regulatory inputs. In the fission yeast Schizosaccharomyces pombe, inhibitory phosphorylation of Cdc2 on tyrosine residue 15 (Tyr15), plays at least two roles in cell cycle control. The inhibitory phosphate is added by either one of two partially redundant kinases, Wee1 and Mik1, and is removed by the Cdc25 phosphatase (Drosophila homolog: String). Activity of either Wee1 or Mik1 kinase is required during S phase to prevent premature progression of the cell cycle into mitosis and subsequent 'mitotic catastrophe', where incompletely replicated chromosomes sustain catastrophic damage as they attempt to segregate. The exceptionally small size of yeast cells deficient in Wee1 suggests that Wee1 plays a unique role in regulating the cell size at which mitosis normally occurs, by imposing a G2 arrest until a minimum size requirement is met. Thus, in fission yeast, tyrosine phosphorylation of Cdc2 is part of a mechanism for checking advance of the cell cycle prior to completion of S phase, and plays a regulatory role in coupling cell division with growth (Campbell, 1995 and references therein).

Studies of cycling frog extracts have suggested another role for inhibitory phosphorylation of Cdc2. Progress of the extracts to 'mitosis' requires Cyclin synthesis, and mitotic degradation of Cyclins resets the system, producing oscillations. The accumulation of Cyclin is gradual, but the resulting activation of Cdc2 is abrupt. This discordance can be explained in terms of the kinetics of phosphorylation and dephosphorylation of Cdc2. Phosphorylation of threonine 161 (Thr161) by CAK (Cdc2 Activating Kinase) is required for Cdc2 to become activated, providing one level of control. Between pulses of kinase activation, Xenopus egg extracts accumulate triply phosphorylated, inactive Cdc2 (Thr14, Tyr15, and Thr161), complexed with a Cyclin. The kinase is then activated following the removal of inhibitory phosphates by Cdc25. The abrupt nature of this transition is thought to result from positive feedback: once a small amount of active Cdc2 appears, it can promote further activation by inhibiting a Wee1 kinase and stimulating a Cdc25 phosphatase. Thus, in the frog extract system, tyrosine phosphorylation plays a role in converting a gradual signal (cyclin accumulation) into an on/off switch (Campbell, 1995 and references therein).

In Drosophila embryos, inhibitory phosphorylation of Cdc2 plays yet another role. During embryogenesis, the levels of different cell cycle regulators change, and as they change, the rate limiting step in Cdc2 activation shifts. The early nuclear division cycles are run entirely by maternally contributed gene products and consist of rapid oscillations between mitosis and S phase. Constant levels of active, cytoplasmic Cdc2 kinase are detectable throughout cycles 1-8, although it is not clear if this is true in the mitotic apparatus itself. After cycle 9, oscillations in Cyclin levels become detectable and presumably accompany the local inactivation of Cdc2, leading to a gradual slowing of the cycles. At the beginning of cycle 14, newly formed cells arrest in G2 prior to the onset of gastrulation. Subsequent embryonic mitoses occur in an intricate pattern that is regulated both spatially and temporally. Major cytoskeletal rearrangements are required to execute extensive cell movements during gastrulation. These morphogenetic movements are incompatible with mitosis, and it appears that the division program has evolved so as to coordinate their timing with periods of cell division. This coordination is achieved by regulating the expression of String. Events at the beginning of cycle 14, such as destruction of maternally supplied String and the second Cdc25, Twine, allow the accumulation of inhibitory phosphates on Cdc2: the cells arrest in G2 once DNA replication is completed. During subsequent embryonic cell cycles, transcriptionally regulated episodes of String expression lead to transient loss of inhibitory phosphorylation on Cdc2. The resulting pulses of Cdc2 kinase activity drive mitoses in the detailed patterns associated with morphogenesis. Thus, changes in inhibitory tyrosine phosphorylation of Cdc2 play a central role in coordinating cell proliferation with morphogenesis during development in Drosophila (Campbell, 1995 and references therein).

To isolate Drosophila cDNA clones encoding Wee1-like activity, a complementation screen using a Schizosaccharomyces pombe strain that has a temperature-sensitive allele of wee1, and a null allele of the redundant kinase mik1 (wee1-50 mik1::ura+leu-) was set up. This strain undergoes lethal mitotic catastrophe at the restrictive temperature, 37^o C. This strain (selecting for leucine prototrophy) was transformed with a library of Drosophila embryonic cDNAs expressed from the thiamine-repressible NMT promoter and screened for clones that confer viability at the restrictive temperature with the promoter on. Approximately 106 leu+ transformants from the Drosophila cDNA library yielded 53 transformant colonies after 5 days growth at the restrictive temperature. Plasmids were rescued from 48 of these transformants, and restriction mapping suggests that a large number of different genes had been isolated. Rather than re-testing all of these for complementation in the original wee1-50 mik1::ura+leu- background, it was reasoned that a second genetic test for Wee1 function might retrieve a more restricted subset of these clones and still include any genuine Wee1 homologs. Accordingly, each of the cDNA clones was transformed into a second strain: wee1-50 cdc2-3w leu-, which also undergoes temperature-sensitive mitotic catastrophe at a restrictive temperature. The cdc2-3w allele renders entry into mitosis independent of cdc25 expression and sensitizes the cells to reductions in Wee1 function. Ten of the original rescuing plasmids also complemented this strain. Restriction mapping of these clones indicated that this group represented five different genes. One of these clones encodes a protein similar to Wee1 kinase homologs and has been named Drosophila wee. The remaining four genes identified in this screen have been named WMC genes (Wee-Mik-Complementing) (Campbell, 1995).

In Drosophila, the maternally expressed mei-41 and grapes genes are required for successful execution of the nuclear division cycles of early embryogenesis. In fission yeast, genes encoding similar kinases (rad3 and chk1, respectively) are components of a cell cycle checkpoint that delays mitosis by inhibitory phosphorylation of Cdk1. Mutations have been identified in Drosophila wee. Like mei-41 and grp, wee is zygotically dispensable but is required maternally for completing the embryonic nuclear cycles. The arrest phenotype of wee mutants, as well as genetic interactions between wee, grp, and mei-41 mutations, suggest that wee is functioning in the same regulatory pathway as these genes. These findings imply that inhibitory phosphorylation of Drosophila Cdk1 (alternatively termed Cdc2) by Wee is required for proper regulation of the early syncytial cycles of embryogenesis (Price, 2000).

wee has an essential maternal function during the nuclear division cycles of embryogenesis and also implicates zygotic wee function in a cell cycle checkpoint that responds to inhibition of DNA replication. The demonstration that wee has a role during the early syncytial nuclear cycles calls into question a previous assumption that inhibitory phosphorylation does not control these cycles. Analyses of the state of phosphorylation during the early cycles had failed to detect inhibitory phosphorylation of Cdk1 prior to cycle 13. Furthermore, because reduction in the gene dose of cyclin A and cyclin B slows the late nuclear cycles, it has been suggested that progress of these cycles is regulated by accumulation of cyclins to a threshold level. The finding that wee is required for completing the nuclear division cycles suggests that inhibitory phosphorylation plays a role in their regulation after all. The failure to detect inhibitory phosphorylation during these cycles can be explained if only a small pool of Cdk1 is subject to this modification. Wee1-type kinases are predominantly nuclear in Drosophila and other organisms and nuclear Wee1 activity is sufficient to block entry into mitosis even in the presence of high cytoplasmic Cdk1 activity. Hence, it is suggested that inhibitory phosphorylation of a small nuclear pool of Cdk1 contributes importantly to the control of the syncytial cycles. The proposal that inhibitory phosphorylation regulates syncytial cycles is an implicit component of a recently proposed model for the mechanism by which mei-41 and grp regulate the progressive lengthening of these cycles. In response to incompletely replicated DNA, the recognized activities of these conserved checkpoint kinases arrest the cell cycle by preventing the removal of inhibitory phosphates from Cdk1. While this model appears to be at odds with the lack of detectable inhibitory phosphorylation of Cdk1 during the syncytial cycles, the findings that Drosophila wee is required for the early nuclear division cycles supports this proposal. Indeed, the apparent parallels in the phenotypes of mei-41, grp, and wee maternal mutants suggest that these genes operate by a similar mechanism. Because the results implicate this pathway without defining precisely how it is induced, it remains possible that the same pathway could be used in a unique regulatory circuit. In either case, the lesson seems to be that the remarkable conservation of the eukaryotic cell cycle regulatory machinery is coupled with an equally remarkable flexibility in how that machinery can be deployed, depending on the particular developmental constraints of each organism. In early Drosophila embryos, a regulatory pathway that usually serves a surveillance function plays an essential cell cycle role (Price, 2000).

It was unexpected that zygotic wee function would be dispensable under normal growth conditions, since Cdk1 inhibitory phosphorylation appears to play an important role in cell cycle regulation at many stages of development in Drosophila. Following the last syncytial division during interphase of cycle 14, Cdk1 becomes quantitatively inhibited by phosphorylation. This dramatic regulatory transition could result from delocalization of Wee1, activation of a cytoplasmically localized Cdk1 inhibitory kinase, inhibition of cytoplasmic Cdc25, or more active exchange of Cdk1 between the nucleus and cytoplasm during cycle 14. These possibilities are currently being investigated. It has been demonstrated that entry into mitosis 14 depends on zygotic expression of String phosphatase and removal of inhibitory phosphate from Cdk1. Furthermore, String activity is also required during the following postblastoderm mitoses of embryogenesis and during imaginal disc development. Twine activity is also required during meiosis. These requirements for String imply that inhibitory phosphorylation is normally significant at all of these stages of development. In fission yeast, loss of Wee1 kinase can suppress requirements for the Cdc25 phosphatase. In Drosophila, however, loss of zygotic wee function does not bypass the requirement for String activity. The continued requirement for String activity might be due to maternal perdurance of wee function. Alternatively, there might be other Wee1 kinases that can function either redundantly with wee or independently. The gene encoding a Drosophila homolog of Myt1 (see Drosophila Myt1), a Wee1-related kinase, has been cloned; this gene may contribute to some of these activities (Price, 2000).

GENE STRUCTURE

cDNA clone length - 2362

Bases in 5' UTR - 174

Exons - 5

Bases in 3' UTR - 258

PROTEIN STRUCTURE

Amino Acids - 618

Structural Domains

Searches of the GenBank data base indicate that the kinase domain of Drosophila Wee protein is similar to Wee1 homologs from S. pombe, S. cerevisiae, humans and Xenopus. The Wee coding sequence includes the generic, conserved residues shared by all kinases, as well as the signatory 'EGD' motif noted for the Wee1 kinase family. In pair-wise comparisons, Drosophila Wee is most closely related to human Wee1 and Xenopus Wee1 while the three yeast kinases are more closely related to each other than to the three metazoan genes. All of these Wee1-like kinases share a similar structure in that they encode a C-terminal kinase domain and a less conserved, N-terminal domain that in S. pombe functions as a regulatory domain. Drosophila Wee and human Wee1 kinases also share two conserved regions in the N-terminal domain. This region includes a sequence motif (NI/VNPFTPD/QS) that resembles a consensus phosphorylation site for MAP kinases, and is also found in Xenopus Wee1. In addition, eight T/SP motifs are found in the N-terminal region of Dwee1 as well as two in the C-terminal region. These T/SP are potential phosphorylation sites for several mitotic kinases and multiple T/SP sites occur in the N-terminal regions of all three metazoan Wee1 homologs (Campbell, 1995).

date revised: 2 April 2002

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.