Cyclin-Dependent Protein Kinases and the Control of the Cell Cycle

Previous and current research
We study the structure, function and destruction of CDKs, which control cell cycle transitions: the initiation of chromosome replication, the onset of mitosis and the return to interphase at the end of cell division. We are particularly interested in the role of cyclins in the activation of CDK1 and CDK2, and are trying to understand how particular cyclins control particular processes. For example, cyclin E/CDK2 is required for the initiation of S phase, and cannot initiate mitosis. Cyclin B/CDC2 causes nuclear envelope breakdown and chromosome condensation, but normally cannot initiate DNA replication.

We use a variety of approaches to these problems, using recombinant DNA methods to make active CDKs in bacteria, and frog egg extracts to perform assays for DNA replication or entry into mitosis. We have identified a number of substrates for cyclin A/CDK2 and cyclin B/CDK1 and have compared their specificity. Rather to our surprise, many substrates are equally well phosphorylated by either enzyme, and it looks as though their subcellular localisation may be more important than their intrinsic biochemical specificity. We would like to know how many substrates have to be phosphorylated how much to get cells into mitosis.

We would also like to know why cyclin E/CDK2 cannot by itself get cells into mitosis. We recently discovered that the anaphase-promoting complex (APC/C) directly recognises the 'destruction box' of cyclin B, and that this binding is tighter in mitosis than it is in interphase. How is the recognition regulated?

It is also a great mystery how cyclin A is degraded at the onset of mitosis, while the disappearance of cyclin B is delayed by the spindle assembly checkpoint. We know that cyclin A is recognised slightly differently than cyclin B, but have yet to fully understand the basis of its recognition by the APC/C. The destruction box in cyclin A is necessary, but not sufficient for its recognition.

Future projects
Recent work in the laboratory reveals that the established consensus sequence for CDKs, S/T-P-X-K/R, is not necessary for phosphorylation of some protein, at least, although the proline residue following the serine or threonine is probably more or less mandatory. We need to identify more mitotic substrates, and to find ways to count how many of them there are. Possible approaches include the use of phosphospecific binding proteins and antibodies to conduct a proteomic analysis.

A further aspect of this kind of approach is to ask about the roles of mitotic phosphorylation, to what extent these protein modifications produce new binding sites for recognition by other proteins. For example, it was recently discovered that the polo box in polo kinase recognises a consensus sequence that is phosphorylated by CDKs. There are almost certainly other examples.