BACKGROUND: A complex of MCM proteins is implicated in ensuring that DNA replicates only once in each cell cycle, by 'replication licensing'. The nuclear membrane is also implicated in replication licensing, but the relationship between the MCM proteins and the nuclear membrane is unclear. Here, we investigate the relationship between XMCM3 (a component of the Xenopus MCM complex), nuclear envelope permeability and the initiation of DNA replication once per cell cycle. RESULTS: Our results show that the nuclear envelope does not prevent the entry of XMCM3 into the nucleus, but that it does prevent the binding of XMCM3 to chromatin. We have also identified another component of the Xenopus MCM complex as a homologue of the Schizosaccharomyces pombe protein Cdc21. XMCM3 does not preferentially co-localize with sites of DNA replication. Instead, it is almost uniformly distributed on chromatin and is suddenly lost during replication. XMCM3 crosses intact nuclear membranes of G2-phase HeLa cells but cannot then bind to chromatin. Permeabilization of the nuclear envelope allows the binding of XMCM3 to G2-phase chromatin. We have therefore resolved replication licensing into two stages. The first requires the entry of a cytosolic 'loading factor' that is excluded by the nuclear membrane; subsequently, MCM3 can bind to chromatin in the presence or absence of a nuclear membrane, but only if the loading factor has gained access in the absence of the membrane. CONCLUSIONS: The Xenopus MCM complex contains homologues of yeast MCM2, MCM3, MCM5 and Cdc21 proteins. XMCM3 is displaced from chromatin during replication. The nuclear envelope allows entry of XMCM3 into the nucleus, but regulates its binding to chromatin; binding requires a loading factor which cannot cross the nuclear envelope. Based on these results we present a two-stage model for replication licensing.
BACKGROUND: A complex of MCM proteins is implicated in ensuring that DNA replicates only once in each cell cycle, by 'replication licensing'. The nuclear membrane is also implicated in replication licensing, but the relationship between the MCM proteins and the nuclear membrane is unclear. Here, we investigate the relationship between XMCM3 (a component of the XenopusMCM complex), nuclear envelope permeability and the initiation of DNA replication once per cell cycle. RESULTS: Our results show that the nuclear envelope does not prevent the entry of XMCM3 into the nucleus, but that it does prevent the binding of XMCM3 to chromatin. We have also identified another component of the XenopusMCM complex as a homologue of the Schizosaccharomyces pombe protein Cdc21. XMCM3 does not preferentially co-localize with sites of DNA replication. Instead, it is almost uniformly distributed on chromatin and is suddenly lost during replication. XMCM3 crosses intact nuclear membranes of G2-phase HeLa cells but cannot then bind to chromatin. Permeabilization of the nuclear envelope allows the binding of XMCM3 to G2-phase chromatin. We have therefore resolved replication licensing into two stages. The first requires the entry of a cytosolic 'loading factor' that is excluded by the nuclear membrane; subsequently, MCM3 can bind to chromatin in the presence or absence of a nuclear membrane, but only if the loading factor has gained access in the absence of the membrane. CONCLUSIONS: The XenopusMCM complex contains homologues of yeastMCM2, MCM3, MCM5 and Cdc21 proteins. XMCM3 is displaced from chromatin during replication. The nuclear envelope allows entry of XMCM3 into the nucleus, but regulates its binding to chromatin; binding requires a loading factor which cannot cross the nuclear envelope. Based on these results we present a two-stage model for replication licensing.