License Withheld—Geminin Blocks DNA Replication

Zoi Lygerou and Paul Nurse

Science, December 2000

For cells to survive they must receive a complete copy of their genome every time they divide. Two events enable dividing cells to achieve this goal – S-phase (during which the DNA of the chromosomes is replicated) and M phase of mitosis (during which the replicated chromosomes segregate into the two newly divided cells). To ensure genomic stability, S-phase is tightly regulated so that replication of the chromosomes is initiated only once in each cell cycle. A process called licensing ensures that chromatin becomes competent for a further round of DNA replication only after passage through mitosis. Building on the secure foundations provided by studies of prokaryotes and viruses, work in budding and fission yeasts and with frog egg extracts has identified and characterized many of the components that tightly regulate S-phase onset. Reports by Wohlschlegel et al. and Tada et al. in Nature Cell Biology now connect the activities of two of these components—the positive regulator Cdtl and the negative regulator Geminin—and provide further insight into the licensing of DNA replication in human and frog cells.

The mechanisms leading to the initiation of DNA replication depend on the sequential association of proteins with chromatin. A collection of proteins called the origin recognition complex (ORC), which is thought to bind to origins of replication in the chromatin, is associated with chromatin throughout the cell cycle. This association is necessary for the binding of other replication proteins but does not appear to regulate the timing of S-phase onset. The onset of S phase appears to be controlled by six proteins that form the MCM (min-chromosome maintenance) complex. As the cells exit from mitosis, the MCM interacts with chromatin and licenses the DNA for replication. Although the molecular basis for MCM action is still not clear, the complex is likely to open up the chromatin, providing access for enzymes that replicate DNA. This activity is consistent with the finding that MCM family members are helicase enzymes that unwind DNA.

The loading of MCM proteins onto chromatin is, therefore, a key step in controlling the initiation of DNA replication. Loading requires the initiating factor Cdc6/18, which accumulates in the nucleus as cells exit mitosis and enter Gi; an increase in the amount of Cdc6/18 prompts inappropriate entry into S phase. A second MCM loading factor, Cdtl, has been identified in fission yeast, the fruit fly Drosophila (where it is called DUP), and the frog Xenopus . Cdtl is expressed when cells exit mitosis, becomes associated with chromatin, and can form a complex with Cdc6/18, potentiating its activity. The overall regulation of this sequence of events involves the cyclin-dependent kinanses (CDK.S), which have both positive and negative regulatory effects. The Gi/S CDK promotes progression through G1 into S phase, whereas the G2/M CDK inhibits the licensing of DNA for replication during G2. The G2/M CDK does this by phosphorylating, among other proteins, Cdc6/18 and MCM family members, leading to their inactivation by degradation, nuclear exclusion, or inhibition of chromatin binding.

A candidate molecule for blocking licensing is a protein called Geminin, identified in Xenopus and human cells. Geminin is present in the cell nucleus from S phase until mitosis and is degraded as cells complete mitosis. Addition of Geminin to an in vitro replication assay containing Xenopus egg extracts blocks the association of MCM proteins with G1 chromatin, thereby inhibiting DNA replication. The addition of Geminin blocks initiation of DNA replication at the same stage as does depletion of Cdtl from Xenopus extracts. Wohlschlegel et al. and Tada et al. now extend these findings by showing that Geminin stably interacts with Cdtl in Xenopus and human cells.

Wohlschlegel et al. looked for proteins that would form a complex with Geminin. They found two proteins (with molecular sizes of 130 kD and 65 kD) that could be co-precipitated with Geminin from extracts of cultured human HeLa cells. The 65-kD protein was identified as the human homolog of Cdtl, but the identity of the 130-kD protein remains unknown. Human Cdtl co-precipitated with a single major protein of 35 kD from human cell extracts, which was found to be Geminin. The authors cloned a full-length cDNA and showed that the protein it encodes interacts with Geminin in vitro. The amino-terminali 101 amino acids of human Cdtl are sufficient to ensure binding to Geminin; this region is the least conserved among Cdtl homologs from different species and was not found in a previously identified partial human cDNA.

In a parallel set of experiments, Tada and co-workers set out to identify the targets of Geminin using a Xenopus in vitro DNA replication assay. The investigators showed that Geminin binds to RLF-B, a licensing activity present in Xenopus egg extracts, the components of which have not yet been identified. They then demonstrated that RLF-B is, in fact, Cdtl. Both groups used the Xenopus in vitro DNA replication system to show that addition of excess Cdtl reverses the block on licensing imposed by Geminin, and restores DNA replication. Thus, Cdtl can bind to and block Geminin activity, suggesting that Geminin inhibits the initiation of DNA replication by binding to and inactivating Cdtl. It is rather more difficult to establish unambiguously that Geminin acts through, and only through, Cdtl. The Geminin interaction domain is at the end of Cdti's amino terminus. To establish whether Geminin acts only through Cdtl, one could engineer a truncated Cdtl protein missing its amino-terminal domain and then see whether this protein is able to support DNA replication in vitro, even in the presence of excess Geminin.

What can we say about Geminin's job in vivo? Geminin accumulates when Xenopus eggs are arrested in metaphase, just before the separation of chromosomes to the two daughter celts, and is degraded when cell division proceeds. Tada et al. show that depletion of Geminin from metaphase extracts prompted licensing of chromatin for DNA replication. Geminin could therefore be important for repressing DNA licensing until cell division is complete. Geminin also accumulates in the nucleus of G2 cells, and could act redundantly to ensure inhibition of S phase if licensing factors were to become inappropriately expressed in G2 cells. Geminin is also present in S-phase cells - here it may bind to and inactivate Cdtl to ensure that DNA replication is not reinitiated at origins that have just replicated (a process called origin refiring). Geminin may also be important after DNA damage, halting S-phase to give cells time to repair their DNA. It will be interesting to see whether Geminin expression is induced in cells arrested in G1, due to either DNA damage or withdrawal of growth factors. Genetic approaches will help to clarify the various tasks of Geminin in vivo.

Is Geminin's regulatory role conserved in evolution? Geminin homologs are present in mammals and Xenopus. We have identified a putative Drosophila Geminin homolog in DNA databases (unpublished observations), and Drosophila Cdtl has an amino-terminal domain that could potentially interact with Geminin. This suggests that Geminin's job may be conserved among all metazoans (multi-celled organisms). If a Drosophila homolog of Geminin exists, then flies carrying mutant forms of this protein could be engineered enabling the different functions of Geminin to be elucidated.

Cdtl homologs from fission yeast and the plant Arabidopsis are truncated at their amino termini, in contrast to their metazoan orthologs, and no proteins with significant similarity to Geminin have been identified in these organisms. They may yet turn out to have divergent Geminin homologs that operate through a different Cdtl domain or a different protein, but it is possible that Geminin exists only in Metazoa. Geminin may have evolved to couple S-phase regulation to developmental and growth signals found only in metazoans. Given the fact that Geminin is a crucial negative regulator of the cell cycle, it will be important to establish whether it operates as a tumor-suppressor protein and whether it is mutated in cancer cells.

Answer the questions.

1. What is the role of the MCM complex in the cell?
2. What is the function of Geminin in mitosis?
3. What practical application can further investigation into Geminin lead to?

2. Find in the text synonyms for the following words: to ensure, onset, to bind, assay, depletion, putative, to truncate, to elucidate.