Interaction of chromatin-associated Plk1 and Mcm7.

Plk1 is a multifunctional protein kinase involved in regulation of mitotic entry, chromosome segregation, centrosome maturation, and mitotic exit. Plk1 is a target of DNA damage checkpoints and aids resumption of the cell cycle during recovery from G2 arrest. The polo-box domain (PBD) of Plk1 interacts with phosphoproteins and localizes Plk1 to some mitotic structures. In a search for proteins that interact with the PBD of Plk1, we identified two of the minichromosome maintenance (MCM) proteins, Mcm2 and Mcm7. Co-immunoprecipitation and immunoblot analysis showed an interaction between full-length Plk1 and all other members of the MCM2-7 protein complex. Endogenous Plk1 co-immunoprecipitates with basal forms of Mcm7 as well as with slower migrating forms of Mcm7, induced in response to DNA damage. The strongest interaction between endogenous Plk1 and Mcm7 was detected in a soluble chromatin fraction. These findings suggest a new function for Plk1 in coordination of DNA replication and mitotic events.

Normal progression of mammalian cells through the cell cycle requires precise integration of positive and negative cell cycle regulators. This function is fulfilled by biochemical networks, called cell cycle checkpoints, that prevent the initiation of the next cell cycle event before correct execution of the previous one. One of the critical circuits for cell cycle control activates Cdc2/cyclin B kinase, called maturation-promoting factor (MPF), 1 at the onset of mitosis. Wee1 and Myt1 kinases keep MPF inactive by inhibitory phosphorylation of Cdc2, whereas phosphatases from Cdc25 family remove these phosphates and activate MPF. Polo-like kinase 1 (Plk1) is one of the protein kinases that activate Cdc25C.
Plk1 is a positive cell cycle regulator that is a member of the Polo-like kinase (PLK) family. Plk1 is expressed in S, G 2 , and M phases of the cell cycle, and becomes activated by phosphorylation at the G 2 /M boundary. Plk1 regulates a cleavageindependent mechanism for dissociation of cohesin from chromosomes, which is important for the separation of sister chromatids (1). Plk1 assists in MPF activation through phosphorylation-dependent import of cyclin B1 and activation of Cdc25C, an activator of Cdc2. After metaphase, Plk1 enhances ubiquitin-dependent degradation of cyclin B1 and inactivation of MPF by activation of anaphase-promoting complex. Results of experiments with a dominant negative mutant of Plk1 suggest a possible role for Plk1 in cytokinesis (2). The centrosome cycle requires Plk1, because microinjection of ␣-Plk1 antibodies prevents centrosome maturation (3).
DNA damage response pathways, including ATM and ATR, inhibit Plk1 after DNA damage (4,5). Interactions of Plk1 with Chk2, Claspin, and Brca1 suggest that there are additional roles of Plk1 in DNA structure checkpoints (6 -8). Plk1 is also involved in restarting the cell cycle after DNA damage-induced checkpoint arrest, through down-regulation of Wee1 (9).
PLKs have one or two conserved elements, called polo boxes, which form carboxyl-terminal phosphopeptide-binding domains (PBDs) (10). The PBD is probably involved in regulation of Plk1, because binding of the PBD to a phosphopeptide elevates Plk1 kinase activity (11). PBDs are important for localization of PLKs to mitotic structures (12) and PLK interactions with other proteins, including Cdc25C, MKlp2, and Chk2 2 (10,13). Identification of additional Plk1 PBD-interacting proteins may lead to discovery of new functions of Plk1.
Six of the minichromosome maintenance (MCM) proteins, Mcm2-Mcm7, form complexes that participate in initiation and elongation steps of DNA replication (14). They share a conserved 200-amino acid nucleotide-binding region and form different subcomplexes (dimers, trimers, and a hexamer) (15 (14). MCM proteins are associated with chromatin in late telophase and at the beginning of the G 1 phase of the cell cycle (16). Interaction of MCMs with chromatin depends on proteins of the origin recognition complex (ORC), Cdc6 and Ctd1 (17). These proteins form a pre-replication complex at origins of DNA replication. Prereplication complexes are activated by cyclinE/Cdk2, cyclinA/ Cdk2, and Dbf4/Cdc7 protein kinases at the G 1 /S phase transition (14). During S phase, Mcm proteins are released from origins of replication after initiation of DNA replication and move with replication forks where they are thought to function as a DNA helicase. Mechanisms that assure the replication of DNA only once per cycle release Mcm proteins from chromatin after firing of the origins of replication and prevent the reloading of Mcm proteins on chromatin until telophase. Mcm7 participates in processes other than DNA replication, interaction with the MYCN transcription factor (18) (20).
In this study, we identify members of the MCM protein complex as Plk1-interacting proteins. Interaction of Plk1 with Mcm2 and Mcm7 requires the Plk1 PBD. The interaction between Plk1 and Mcm7 is regulated in response to DNA damage. Chromatin-associated Plk1 and Mcm7 interact more strongly than their cytosolic forms. These results provide evidence for the existence of a link between the DNA replication apparatus and mitotic regulation.
Cell Cultures, Treatments, and Transfection-HEK293 cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum in the presence of antibiotics in a humidified incubator at 37°C. Cells were treated with nocodazole (250 ng/ml, Sigma) for 16 h, hydroxyurea (1 mM, Sigma) for 16 h, or Adriamycin (0.5 or 2 M, Sigma) for 2 h. Cells were transfected with FuGENE 6 (Roche Applied Science), 5 g of plasmid DNA per 10-cm culture dish, and analyzed after 40 h.
For GST pull-down assays, 5 g of soluble GST or GST-Plk1 bound to 20 l of glutathione-Sepharose beads (Amersham Biosciences AB) were incubated with 500 g of lysate from HEK293-T cells for 12 h at 4°C. The assays were analyzed by immunoblotting (IB) with ␣-MCM7 antibody and by Coomassie Brilliant Blue staining.
Isolation and Identification of Plk1 PBD Interacting Proteins-30 mg of protein lysate from mock-transfected or transiently expressing FLAG-Plk1-(330-CT) HEK293-T cells was incubated with 100 l of ␣-FLAG affinity agarose gel for 16 h. FLAG immune complexes were analyzed by SDS-PAGE, and proteins were visualized with Coomassie Blue G-250 (Pierce). The bands of interest were excised from the gel and sent to the Keck Foundation Biotechnology Resource Laboratory at Yale University for trypsin digestion and gel matrix-assisted laser desorption ionization mass spectrometry (MALDI-MS), followed by peptide mass data base searching to identify known proteins. A ProFound data search with peptides yielded by the 90-kDa protein band revealed 24 peptides that match the sequence of human Mcm7 and cover 28% of its sequence. Among the peptides are YIAMCREK, SQLLSYIDR, GSS-GVGLTAAVLR, DVLDVYIEHR, and QIAEEDFYEK. A ProFound data search with peptides obtained from the 125-kDa protein band showed 28 peptides, including FVVGSHVR, MYSDLRK, ITNHIHVR, and AGIVTSLQAR, that match the sequence of Mcm2 and cover 34% of the protein.
Cell Fractionation-Soluble chromatin fractions were prepared according to the procedure described by Fujita et al. (22). In brief, 293-T cells were lysed on ice with CSK buffer (10 mM PIPES, pH 6.8, 100 mM NaCl, 1 mM MgCl 2 , 1 mM EDTA, 1 mM dithiothreitol, and 300 mM sucrose) containing 0.1% Triton X-100, and centrifuged at 1,000 ϫ g for 5 min. The supernatants were clarified by centrifugation at 15,000 ϫ g for 5 min to prepare Triton X-100-extractable fractions. The pellets, containing nuclei, were washed, resuspended in CSK buffer containing 1 mM ATP and digested with 10 units/l DNase I at 25°C for 30 min to prepare DNase I-released fractions.

RESULTS AND DISCUSSION
To identify Plk1-interacting proteins, ␣-FLAG antibody was used to immunoprecipitate lysates from HEK293-T cells transiently expressing a FLAG-tagged portion of Plk1, amino acids 330 to the carboxyl terminus, that constitutes the PBD (10). Immunoprecipitated proteins were separated by SDS-PAGE and stained with Coomassie Blue G-250 (Fig. 1). Two protein bands with molecular masses of 125 and 90 kDa that were specific for immunoprecipitates from cells expressing FLAG-Plk1-(330-CT) were excised from the gel and analyzed by MALDI-MS. They were identified as minichromosome maintenance proteins 2 (Mcm2) and 7 (Mcm7).
Six Mcm proteins, Mcm2-7, can associate in a single complex and are also assorted among various sub-complexes. We determined whether other members of the complex, in addition to Mcm2 and Mcm7, also interact with Plk1. ␣-FLAG immunoprecipitates from cells expressing FLAG-Plk1, FLAG-Plk1-(1-330), and FLAG-Plk1-(330-CT) were probed with an ␣-Mcm3 antibody. Mcm3 was only detected in immunoprecipitates from cells expressing full-length FLAG-Plk1 (Fig. 3A, lane 2). This differs from the binding requirements for Mcm7, which also binds well to the 330-CT fragment containing the PBD (Fig. 2B,  lane 5).
To evaluate the presence of the other Mcm proteins in FLAG-Plk1 immunoprecipitates, we used an ␣-MCM2-7 antibody (generously provided by S. Bell) that recognizes a conserved epitope on all MCM2-7 subunits (23). The antibody detected Mcm2-7 in the lysates as well as in ␣-FLAG-Plk1 immunoprecipitates from cells expressing FLAG-Plk1 (Fig. 3B, lane 4).
To find a biological context for the interaction between Plk1 and Mcm7, we determined if DNA damage affects the interaction. To introduce DNA damage, we used Adriamycin, a topoisomerase inhibitor that causes DNA breaks. This activates DNA damage checkpoints in multiple phases of the cell cycle (24). Treatment of HEK 293 cells with Adriamycin increased the amount of slower migrating forms of Mcm7 (Fig. 4A, lanes 2 and  3). Interestingly, slower migrating forms of Mcm7 were proportionately increased relative to basal Mcm7 forms in Plk1 immunoprecipitates (Fig. 4B, lanes 3 and 4). A protein band with the same mobility was detected by the ␣-MCM2-7 antibody that recognizes all MCM2-7 subunits (23) in Plk1 immunoprecipitates from Adriamycin-treated cells (Fig. 4C, lane 5), but not from control untreated cells (lanes 4 and 6), and in antibody only immunoprecipitates (lane 3). Treatment of the immunoprecipitates with -phosphatase decreased the slower migrating forms detected with ␣-Mcm7 (Fig. 4B, lane 5) and ␣-Mcm2-7 2 antibodies, suggesting that these are phosphoproteins.
Mass spectrometry and IB data showed that Plk1 interacts with Mcm2 through Mcm7. However, it is not clear if Plk1  (Fig. 3A).
Mcm proteins are localized to the cytosol and nucleus and are associated with chromatin in a cell cycle-dependent manner. Cytosolic extracts were used for the immunoprecipitation experiments in Figs. 1-4. Endogenous Plk1 and MCMs might be expected to interact at chromatin, because Mcm proteins are associated with chromatin in parts of late mitosis, G 1 , and S phases, and Plx1 is present at stalled replication forks in association with Claspin (8). To examine the association of Mcm7, as a representative Mcm protein, and Plk1 with chromatin, we fractionated cellular lysates into cytosolic (Triton X-100-extractable) and "soluble chromatin" fractions prepared by DNase treatment of Triton X-100-extracted nuclei (22). Plk1 and Mcm7 proteins were present in both sets of fractions (Fig.  5, lanes 2-7). Orc2, a chromatin-associated protein, was used as a chromatin marker, and Grb2, a cytosolic protein was used as a cytosolic marker (Fig. 5, lower two panels). Mcm7 was coimmunoprecipitated with endogenous Plk1 from both cytosolic and DNase-released chromatin fractions (Fig. 5, lanes 5 and 2). Interestingly, the ␣-Plk1 antibody co-immunoprecipitated more Mcm7 from the DNase-released soluble chromatin fraction than from the cytoplasmic fraction, even though both Plk1 and Mcm7 were more abundant in the cytosolic fraction (Fig. 5,  lanes 2 and 5). Treatment of cells with nocodazole for 16 h synchronized them in G 2 /M. Under these conditions, the amount of total Mcm7 and Mcm7 in ␣-Plk1 immunoprecipitates from the soluble chromatin fraction was reduced. Likewise, S-phase arrest imposed by hydroxyurea treatment for 24 h did not alter the relative association between Plk1 and Mcm7 (Fig. 5, lanes 4 and 7). Hydroxyurea treatment inhibits DNA synthesis, actuating the intra-S phase replication checkpoint, and activating DNA damage checkpoint pathways through production of single-stranded DNA and aberrant DNA structures resulting from replication fork collapse.
A subset of the MCM proteins are phosphorylated in S phase by Cdk2 and Dbf4/Cdc7 protein kinases, which are required for S phase entry and for activation of replication forks, and by the checkpoint phosphoinositide 3-kinase-related kinase ATR, which negatively regulates cell cycle progression after DNA damage, and which might protect stalled replication forks (20,25). The requirement of the Plk1 PBD, a phosphopeptide-binding motif, for Plk1 binding to MCM7 and MCM3 proteins, suggests that a phosphorylation-dependent mechanism mediates the interaction (Figs. 2B and 3A). Perhaps phosphorylation of MCMs by Dbf4/Cdc7 and S phase cyclin-dependent kinases, or by ATR or ATM kinases creates binding sites for Plk1 PBD. In a Xenopus DNA replication checkpoint response to stalled replication forks, ATR phosphorylates chromatinassociated Claspin, a checkpoint mediator protein, creating a docking site through which the Polo-like kinase Plx1 interacts with Claspin (8). Plx1 bound to Claspin, phosphorylates it, and causes Claspin dissociation from chromatin (8). Such a model, together with the fact that proteins from the MCM complex are also located at the replication fork, would be consistent with the stronger co-immunoprecipitation of Mcm7 with Plk1 from chromatin, relative to cytosolic fractions (Fig. 4).
The budding yeast Polo-like kinase Cdc5 interacts with the Dbf4 subunit of Dbf4/Cdc7 kinase that activates the pre-replication complex (26). Dbf4/Cdc7 itself phosphorylates some of the Mcm proteins (27). Also, Rad53, the Chk2 ortholog in budding yeast, regulates Dbf4/Cdc7 kinase after DNA damage (28). These results, taken together with the fact that Plk1 and Chk2 themselves interact (6), imply the existence of a regulatory network involving Mcm proteins, Dbf4/Cdc7, Plk1, and Chk2 kinases. Recently, it was reported that MCM2 and MCM3 are substrates of ATM and ATR kinases that are activated in response to genotoxic stress (20). In a normal cell cycle, ATR phosphorylates MCM2 during S phase, which may be a component of a mechanism that monitors the normal progression of DNA replication (20). It is interesting that the ATR-interacting protein, which forms a stoichiometric complex with ATR, interacts with MCM7 and that reduction of MCM7 protein causes defects in the intra-S-phase checkpoint (20). Also, Mcm7 binds to Rad17, a protein involved in activation of cell-cycle checkpoints, and participates in the transmission of DNA damage  1-4) and Triton X-100-extractable cytosolic fractions (lanes 5-7) from nocodazole- (lanes 3 and 6), hydroxyurea-(lanes 4 and  7), or mock-treated (lanes 1, 2, and 5) HEK 293T cells, were immunoprecipitated with ␣-Plk1 antibody. Fractions and IPs were analyzed by IB with ␣-Plk1, ␣-Mcm7, Orc2, and Grb2 antibodies. signals (29). In the context of these results, it is even more intriguing that slower migrating forms of Mcm7, induced by DNA damage after treatment with Adriamycin, co-immunoprecipitate efficiently with Plk1 (Fig. 4B, lane 4; Fig. 4C, lane 5).
Strict timely regulation of Plk1 expression and activity is important for the passage of cells through mitosis. Overexpression of a kinase-defective mutant of Plk1 (Plk1-KD), the carboxyl-terminal portion of Plk1 (Plk1-CT), and, to a lesser extent, the wild type Plk1, alters the normal cell cycle of some cell culture lines, resulting in the accumulation of cells with G 2 /M DNA content and disorganization of condensed chromosomes (30,31). Furthermore, spindle checkpoint mechanisms are activated after overexpression of Plk1-KD and Plk1-CT (2). These cells later undergo mitotic catastrophe and/or apoptosis. The fact that kinase inactive mutants as well as wild type Plk1 induce this cellular phenotype, rules out the involvement of Plk1 kinase activity. One possible explanation for this phenotype is the titration of a protein or proteins crucial for the transition through mitosis. Overexpressed PBD of Plk1 binds to and inhibits endogenous Plk1, which might also cause this abnormal phenotype (32). Introduction of a polo-box peptide fused to an Antennapedia peptide in cells results in G 2 /M cell cycle arrest, misaligned chromosomes, multiple spindle poles, and apoptosis (33).
The interaction of Plk1-(330-CT) with Mcm proteins suggests new mechanisms for the phenotype observed after expression of Plk1-CT. Plk1-(330-CT) binding to different Mcm protein complexes might affect their functions in DNA replication, such as firing of DNA replication origins. This could produce altered chromosomes with incompletely replicated DNA and/or chromosomes with cohesion problems. Interestingly, mutations in DNA replication-related genes, such as ORC2, MCM2, and the gene encoding DNA polymerase ␦ in budding yeast activate the spindle checkpoint (34). Furthermore, small interference RNA down-regulation of Orc6 produced a phenotype with multipolar spindles, aberrant mitosis, and multinucleated cells, which has similarities to the "Plk1-CT" phenotype (35). Activation of an ATR-dependent checkpoint in response to Plk1 depletion by small interference RNA in U2OS cells provides another clue to a possible link between Plk1 and DNA maintenance (9).
Plk1 is a protein kinase that mainly participates in mitotic regulation, whereas Mcm proteins are part of a complex involved in regulation of DNA replication. The association of Plk1 with Mcm proteins indicates possible new functions for Plk1, as well as for Mcm proteins and the possibility of a new link for coordination between DNA replication and mitosis, in the unperturbed cell cycle and in response to genotoxic stress.