Cyclin E Uses Cdc6 as a Chromatin-Associated Receptor Required for DNA Replication

Using an in vitro chromatin assembly assay in Xenopus egg extract, we show that cyclin E binds specifically and saturably to chromatin in three phases. In the first phase, the origin recognition complex and Cdc6 prereplication proteins, but not the minichromosome maintenance complex, are necessary and biochemically sufficient for ATP-dependent binding of cyclin E–Cdk2 to DNA. We find that cyclin E binds the NH(2)-terminal region of Cdc6 containing Cy–Arg-X-Leu (RXL) motifs. Cyclin E proteins with mutated substrate selection (Met-Arg-Ala-Ile-Leu; MRAIL) motifs fail to bind Cdc6, fail to compete with endogenous cyclin E–Cdk2 for chromatin binding, and fail to rescue replication in cyclin E–depleted extracts. Cdc6 proteins with mutations in the three consensus RXL motifs are quantitatively deficient for cyclin E binding and for rescuing replication in Cdc6-depleted extracts. Thus, the cyclin E–Cdc6 interaction that localizes the Cdk2 complex to chromatin is important for DNA replication. During the second phase, cyclin E–Cdk2 accumulates on chromatin, dependent on polymerase activity. In the third phase, cyclin E is phosphorylated, and the cyclin E–Cdk2 complex is displaced from chromatin in mitosis. In vitro, mitogen-activated protein kinase and especially cyclin B–Cdc2, but not the polo-like kinase 1, remove cyclin E–Cdk2 from chromatin. Rebinding of hyperphosphorylated cyclin E–Cdk2 to interphase chromatin requires dephosphorylation, and the Cdk kinase–directed Cdc14 phosphatase is sufficient for this dephosphorylation in vitro. These three phases of cyclin E association with chromatin may facilitate the diverse activities of cyclin E–Cdk2 in initiating replication, blocking rereplication, and allowing resetting of origins after mitosis.