Phosphorylation of Mcm4 at specific sites by cyclin-dependent kinase leads to loss of Mcm4,6,7 helicase activity.

Mcm proteins that play an essential role in eukaryotic DNA replication are phosphorylated in vivo, and cyclin-dependent protein kinase is at least in part responsible for the phosphorylation of Mcm4. Our group reported that the DNA helicase activity of Mcm4,6,7 complex, which may be involved in initiation of DNA replication, is inhibited following phosphorylation by Cdk2/cyclin A in vitro. Here, we further examined the interplay between mouse Mcm4,6,7 complex and cyclin-dependent kinases and determined the sites required for the phosphorylation of Mcm4. Six Ser and Thr residues, in all, were required for the phosphorylation. Inhibition of Mcm4,6,7 helicase activity by Cdk2/cyclin A was largely relieved by introducing mutations in these residues of Mcm4. Anti-phosphothreonine antibodies raised against one of these sites reacted with Mcm4 prepared from HeLa cells at mitotic phase but did not bind to those at G(1) and G(1)/S, suggesting that this site is mainly phosphorylated in the mitotic phase. Mcm4,6,7 complex purified from HeLa cells at the mitotic phase exhibited a low level of DNA helicase activity, compared with the complexes prepared from cells at other phases. These results suggest that phosphorylation of Mcm4 at specific sites leads to loss of Mcm4,6,7 DNA helicase activity.

are consistent with the notion that the binding of Mcm proteins to chromatin allows it to be replicated once in a cell cycle.
Failure of the regulation of DNA replication can result in overreplication of genomic DNA. Overreplication was observed when Cdc6 protein was overproduced in yeast cells (13)(14)(15) and when cyclin-dependent protein kinase was inactivated in yeast cells (16 -19), in Drosophila (20), and in human cells (21). The kinase inhibits the assembly of preinitiation complex at the replication origin (19,22,23). One of the targets of this negative regulation is Cdc6 protein. In Saccharomyces cerevisiae, the phosphorylation of Cdc6 by cyclin-dependent kinase leads to its degradation (24). In mammalian cells, phosphorylation of Cdc6 by the kinase causes exclusion of the protein from the nucleus (25)(26)(27). It has been also shown that cyclin-dependent kinase can exclude Mcm4 from the nucleus in yeast (28,29). Thus, the cyclin-dependent protein kinase that plays an essential role in the initiation of DNA replication is critical for the regulation of the replication (reviewed in Ref. 30). In addition to cyclin-dependent kinase, geminin (31), which accumulates during the S and G 2 phases, appears to contribute to the negative regulation of DNA replication by inactivating Cdt1 (7,32).
All of the members of the Mcm2-7 protein family contain DNA-dependent ATPase motifs in the central domain (33). Partly consistent with this notion, we (34 -36) and Lee and Hurwitz (37,38) observed that Mcm4,6,7 proteins form a hexameric complex and function as a DNA helicase in vitro. This DNA helicase activity is not processive under standard conditions of DNA helicase assay, but it is processive when a tailed substrate was used for the DNA helicase assay (38). These results, together with in vivo findings (39,40) that suggest that Mcm proteins are involved in elongation of DNA replication, raise the possibility that Mcm4,6,7 helicase or Mcm2-7 complex is responsible for replication fork movement (41)(42)(43). Mcm proteins are phosphorylated in vivo and in vitro (44 -49), and Cdc2/cyclin B is mainly responsible for the phosphorylation of Mcm4 at the G 2 /M phase in mouse FM3A cells (47). We reported that the phosphorylation of Mcm4 with Cdk2/cyclin A is associated with the inactivation of DNA helicase of the Mcm4, 6,7 complex (50). The kinase phosphorylated the aminoterminal region of Mcm4 in the complex. In this study, we further characterized the phosphorylation of Mcm4 by determining sites critical for the phosphorylation. In addition, we demonstrated that the phosphorylation results in inhibition of the DNA helicase activity of the Mcm4,6,7 complex. We propose that this phosphorylation reaction contributes to the regulation of DNA replication.

EXPERIMENTAL PROCEDURES
Purification of Mouse Mcm4,6,7 Complex-Site-directed mutagenesis of the Mcm4 gene was conducted using the QuikChange site-directed mutagenesis kit (Stratagene). The oligonucleotide 5Ј-GTCCCCGGCAT-CCGCCCCGAGCCGCC-3Ј was used as primer to introduce a change from Ser to Ala at amino acid 7, 5Ј-GAGGAAAGCAGGTCGGCACCC-AATCGGAG-3Ј was used for the change at amino acid 32, and 5Ј-GA-AGTGGGGTGAGAGGCGCACCTGTAAGGCAGAGGCC-3Ј was used for the change at amino acid 109 . The oligonucleotide 5Ј-ATGTCGGC-CCCGGCATCCGCCCCGAGC-3Ј was used as the primer to change both  Ser-3 and Thr-7 to Ala, 5Ј-CGACGCGGACGAGTCGCCCCAACCC-AGTCCCTTCGAAGTG-3Ј was used for the change at amino acid 19, and 5Ј-CTACTGCCAATGCCCACCGCACCAGGAGCCGACCTG-3Ј was used for the change at residue 53. The mutated Mcm4 gene was cloned into a pAcUW31 vector (Pharmingen) in which the Mcm6 gene had been cloned (36). All of the mutated sites in Mcm4 were confirmed by nucleotide sequencing. High 5 insect cells were co-infected with recombinant baculovirus expressing proteins Mcm4 and Mcm6 and virus expressing Mcm7 protein (pVL1392 vector). The Mcm4,6,7 complex was purified by nickel column chromatography and then by glycerol gradient centrifugation without histone H3/H4 column chromatography (36).
DNA Helicase Activity-A 17-mer oligonucleotide (5Ј-GTTTTC-CCAGTCACGAC-3Ј) was labeled at the 5Ј-end with polynucleotide kinase in the presence of [␥-32 P]ATP and then annealed to M13 DNA. The annealed oligomer (5 fmol) was incubated at 37°C for 30 min with Mcm4,6,7 complex in 50 mM Tris-HCl, pH 7.9, 20 mM ␤-mercaptoethanol, 10 mM magnesium acetate, 10 mM ATP, and 0.5 mg/ml bovine serum albumin. The reaction was terminated by adding 0.2% SDS, and an aliquot was electrophoresed on a 12% acrylamide gel in Tris borate/ EDTA. The labeled oligomer in the gel was detected by using a Bio-Image analyzer (FLA2000, Fuji).
To examine the effect of phosphorylation on the DNA helicase activity of the Mcm4,6,7 complex, Cdk2/cyclin A was added to the reaction mixture from the start of the DNA helicase assay, and the displaced oligonucleotide was analyzed as described above.
Preparation of Anti-phosphothreonine Antibodies-Antiserum against phosphothreonine at amino acid 19 of mouse Mcm4 was obtained by immunizing rabbits with a synthetic peptide of NH 2 -SRRGRV(phospho-T)PTQSLRSEC-COOH conjugated with keyhole limpet at the carboxyl terminus. Anti-phosphothreonine antibodies were first purified by phosphopeptide column chromatography. The serum was loaded onto a phosphopeptide column that had been prepared by fixing the peptide with CNBr-activated Sepharose, and antibodies were eluted with 0.1 M glycine (pH 2.5) and 0.15 M NaCl. The eluted solution was neutralized and dialyzed against phosphatebuffered saline. Then, the purified antibodies were loaded onto a nonphosphopeptide column. The pass-through fraction was concentrated and used as anti-phosphothreonine antibodies. The specificity of binding of the purified antibodies was confirmed by enzyme-linked immunosorbent assay.
Preparation of Other Proteins-Human Cdk2/cyclin A was purified as reported (50), and the protein concentration of the purified protein was 150 g/ml. Human Cdc2 (His)/cyclin B (Myc) was purified from High 5 cells co-infected with baculoviruses containing these two genes by nickel-nitrilotriacetic acid affinity column chromatography (these viruses were kindly provided by R. A. Laskey). The concentration of the kinase activity of the purified Cdc2/cyclin B was similar to that of the Cdk2/cyclin A. The kinase activity was measured by the ability that phosphorylates H1 histone as reported (50).
Synchronization of HeLa Cells-HeLa cells were cultured in minimum Eagle's medium supplemented with 5% calf serum at 37°C. The cells were synchronized at G 2 /M by adding nocodazole at 50 ng/ml for 19 h. HeLa cells at G 1 /S phase were prepared by incubating them with 1 mM hydroxyurea for 18 h. For G 1 synchronization, the following procedures were employed. HeLa cells were treated with 1 mM hydroxyurea for 11 h. After being washed, the cells were incubated in the presence of nocodazole for 12 h. Mitotic cells were detached from the surface by shaking the bottle, and the collected cells were cultured for 6 h without nocodazole. The cellular proteins were fractionated into chromatin-unbound and chromatin-bound forms (50,51). Briefly, the cells were suspended with CSK buffer (10 mM Pipes, pH 6.8, 100 mM NaCl, 1 mM MgCl 2 , 1 mM EGTA, and 1 mM dithiothreitol) containing 0.1% Triton X-100 and 1 mM ATP at the concentration of 2-4 ϫ 10 7 cells/ml. The cells were placed on ice for 15 min, and soluble proteins were recovered after centrifugation and used as the chromatin-unbound fraction. The Mcm proteins in the chromatin-unbound form were puri-fied by histone-Sepharose column chromatography and then by glycerol gradient centrifugation (34,35). Usually, 1-2 ϫ 10 8 cells were used for the purification. An Mcm4,6,7 complex purified from HeLa cells at the G 2 /M phase was treated with 12 units of lambda protein phosphatase (New England BioLabs) for 30 min at 30°C under the conditions recommended by the manufacturer. these 35 amino acids play an important role in phosphorylation with Cdk2/cyclin A. However, we showed that deletion of up to 112 amino acids from the amino terminus was required for the almost complete loss of phosphorylation in Mcm4 (50). The DNA helicase activity of these mutant Mcm4,6,7 complexes was examined (Fig. 1A). Deletion of 35 amino acids resulted in a 5-fold reduction in the DNA helicase activity of the Mcm4,6,7 complex, and the removal of 112 amino acids resulted in an ϳ10-fold reduction. Thus, the amino-terminal portion of Mcm4 is required for the DNA helicase activity of the mouse Mcm4,6,7 complex. In contrast, single-stranded DNA-dependent ATPase activity was not affected by the deletion of this portion (data not shown).

Requirement of an Amino-terminal Region of
Sites of Mcm4 Required for Phosphorylation by Cdk2- Fig.  2A shows the amino acid sequence in the amino-terminal region of mouse Mcm4. There are six SerPro(SP) and five Thr-Pro(TP) sites (11 in total) that are potentially phosphorylated by the cyclin-dependent kinases. First, we mutated all three consensus sites (amino acids 7, 32, and 109) of the kinase to alanine in this region and prepared a mutant Mcm4,6,7 complex containing the mutagenized Mcm4 (Fig. 2B). After phosphorylating the mutant complex with Cdk2/cyclin A, proteins in the complexes were digested with lysyl endopeptidase to analyze phosphorylation in the amino-terminal region of Mcm4 (Fig. 2C). As the amounts of Cdk2/cyclin A in the reaction increased, the incorporated radioactivity enhanced and the phosphorylated bands shifted upward; these changes were similar to those in the wild-type complex. We introduced a fourth mutation at either amino acid 3, 19,  We also examined the phosphorylation of Mcm4 by Cdc2/ cyclin B using the same set of mutant complexes. As shown in Fig. 3, the wild-type and mutant Mcm4,6,7 complexes were incubated with Cdc2/cyclin B, and the extent of phosphorylation was examined by measuring the gel-mobility shift of Mcm4. Mcm4 bands that shifted upward were detected in the presence of larger amounts of the kinase when the wild-type complex or the triple mutant complex (7/32/109m) was phosphorylated. Such a mobility-shifted band appeared to be decreased in the four kinds of complexes (  2), was examined for DNA helicase activity in the presence of Cdk2/cyclin A (Fig. 4) To further understand the relationship between the phosphorylation of Mcm4 and the level of Mcm4,6,7 helicase activity, we raised anti-phospho antibodies against Thr at amino acid 19 of Mcm4, which plays a role in the phosphorylation by cyclin-dependent kinase in vitro (Figs. 2 and 3). Mcm proteins were purified from HeLa cells synchronized at various stages of the cell cycle, and Mcm4 in the purified fraction was examined for binding with the anti-phosphothreonine antibodies. As shown in Fig. 6A, the anti-phosphothreonine antibodies reacted with Mcm4 from HeLa cells at the mitotic phase but did not react with Mcm4 from cells at G 1 and G 1 /S, indicating that phosphorylation of Thr-19 mainly occurs at the G 2 /M in HeLa cells. The Mcm4,6,7 complex purified from these synchronized HeLa cells was examined for DNA helicase activity (Fig. 6B). Hexameric protein complex formation of the purified Mcm4,6,7 complex was confirmed by native gel electrophoresis. Compared with the Mcm4,6,7 complex from cells in G 1 and G 1 /S, the complex from cells in G 2 /M showed much weaker activity. These results show a correlation between phosphorylation at amino acid 19 of Mcm4 and the lowered level of DNA helicase activity of the purified Mcm4,6,7 complex.

DISCUSSION
In this study, we determined the sites in Mcm4 that are required for phosphorylation by the cyclin-dependent kinases in vitro. One of these sites was indeed phosphorylated in vivo, as was determined by using anti-phosphothreonine antibodies. We demonstrated that the phosphorylation of specific sites in the amino-terminal region of Mcm4 leads to inactivation of Mcm4,6,7 helicase. The correlation between the Mcm4 phosphorylation and the lowered level of DNA helicase activity was also observed when DNA helicase activity was compared among Mcm4,6,7 complexes prepared from HeLa cells at various stages of the cell cycle. These results suggest that one of the roles of the Mcm4 phosphorylation by the cyclin-dependent kinase is to inactivate the DNA helicase activity of the Mcm4,6,7 complex that is probably involved in DNA replication in vivo.
Mcm proteins are phosphorylated in vivo. The mobility in SDS gel of the Mcm4 that was prepared from HeLa cells at the G 2 /M was greatly retarded, suggesting that several sites in Mcm4 are phosphorylated at G 2 /M in vivo. Fujita et al. (47) have shown that Cdc2/cyclin B is mainly responsible for the phosphorylation of Mcm4 at the G 2 /M in mouse FM3A cells. In Xenopus, phosphorylation of Mcm4 by the cyclin-dependent kinase has been reported by several groups (45,46,48,49). Hendrickson et al. (46) have shown that Cdc2/cyclin B is mainly responsible for the phosphorylation of Mcm4 at the mitotic phase, and Pereverzeva et al., (49) have indicated that a kinase(s) in addition to Cdc2/cyclin B are involved in the phosphorylation. The function of the Mcm4 phosphorylation is not clear, but it may inhibit the binding of Mcm complexes to chromatin (46,48). Conversely, it is suggested that the phos-phorylation of Mcm4 at mitotic phase is required for binding of the Mcm complex to chromatin at the G 1 phase (49). Here, we suggested another function of the Mcm4 phosphorylation. We determined the specific sites in the amino-terminal region of mouse Mcm4 that are required for phosphorylation by Cdk2/ cyclin A in vitro. The same sites in Mcm4 appeared to be required for the phosphorylation by Cdc2/cyclin B in vitro, suggesting that both Cdk2/cyclin A and Cdc2/cyclin B have similar specificity of substrate recognition. In addition to three consensus sites for these kinases in the amino-terminal region of Mcm4, we identified two sites of SerPro(SP) and one site of ThrPro(TP) in this region that are required for the phosphorylation with Cdk2/cyclin A and Cdc2/cyclin B. One of these sites (Thr- 19) was phosphorylated mainly at the G 2 /M phase in human HeLa cells, which was determined by using anti-phosphothreonine antibodies. It is tempting to speculate that this site is phosphorylated by Cdc2/cyclin B at G 2 /M in vivo, but further experiments are needed to test this notion. The antiphosphothreonine would be useful to analyze the mechanism of the Mcm4 phosphorylation.
Mcm proteins are present in cells as a chromatin-bound form as well as an unbound form, and at G 2 /M phase, almost all of the Mcm proteins are present as the unbound form. They appear to be present mainly as a heterohexameric Mcm2-7 complex either in the chromatin-bound form (51) or the unbound form (52). We purified the Mcm4,6,7 complex from the chromatin-unbound fraction at various stages of the HeLa cell cycle. We used a histone-Sepharose column chromatography by which Mcm2, Mcm3, and Mcm5 proteins, which can be inhibitory to Mcm4,6,7 helicase activity, are removed (35,37,53). Then, we compared the DNA helicase activity of these Mcm4,6,7 complexes. The activity of the complex prepared from cells at the G 2 /M phase was much lower than that prepared from cells in other phases (Figs. 5 and 6). In vitro, phosphorylation of Mcm4 by Cdk2/cyclin A resulted in inhibition of the DNA helicase activity of the mouse Mcm4,6,7 complex (Fig. 4). This inhibition by Cdk2/cyclin A was largely rescued by changing the specific sites in the amino-terminal region of Mcm4 that were required for phosphorylation by Cdk2/cyclin A. These results indicate that phosphorylation at specific sites of Mcm4 by cyclin-dependent kinases can lead to inhibition of the Mcm4,6,7 helicase activity. Related to this point, the Mcm4 bound to chromatin at S phase is more phosphorylated than the protein in unbound form in Xenopus (45) and in HeLa cells (47). The functional significance of the phosphorylation of the chromatin-bound Mcm4 at S phase remains to be determined.
The ATP binding sites in Mcm4,6,7 proteins appear to play a distinct role in the DNA helicase activity of the Mcm4,6,7 hexamer (36). Namely, it was suggested that the ATP binding of Mcm6 is essential for DNA helicase activity but not for singlestranded DNA-dependent ATPase activity and that the ATP binding site of Mcm4 contributes to the binding of Mcm4,6,7 complex to single-stranded DNA. We reported that phosphorylation of Mcm4 by Cdk2/cyclin A moderately affects the singlestranded DNA binding activity (50). The amino-terminal region of Mcm4 was required for the DNA helicase activity of the Mcm4,6,7 complex (Fig. 1A). In this region of Mcm4, there are multiple SerPro(SP) sites that could also act as a binding motif of double-stranded DNA (54). Based on these findings, we consider that this region plays a role in the interaction with DNA during DNA helicase action, and the phosphorylation of Mcm4 by the cyclin-dependent kinase perturbs the interaction to inactivate the DNA helicase activity of the Mcm4,6,7 complex.
Several proteins are involved in the regulation of DNA replication, which occurs once in a cell cycle. The cyclin-dependent kinase plays a central role in the regulation. One important target of the kinase is Cdc6 protein, and the phosphorylation causes degradation or nuclear export of Cdc6 (24 -27). However, a significant fraction of Cdc6 protein is present on chromatin throughout the cell cycle of human cells (12,55). Another important substrate of the cyclin-dependent kinase may be Mcm proteins. The kinase may play a role in the regulation of DNA replication by preventing the loading of Mcm proteins onto chromatin during the S and G 2 /M phases (19,22,46,56). It is possible that the phosphorylation of Mcm4 at the sites determined here is involved in the detachment of Mcm proteins from chromatin. In this paper, we have suggested another role for the phosphorylation of Mcm4 by cyclin-dependent kinase in the regulation of DNA replication. Namely, Mcm4,6,7 helicase activity may be a target for the negative regulation by the cyclin-dependent kinase.