Identification of Stimulators and Inhibitors of Cdc7 Kinase in Vitro*

Cdc7 is a serine-threonine kinase that regulates initiation and progression of DNA replication. The activity of purified Cdc7 kinase is significantly stimulated by polyamines such as spermine or spermidine. Positively charged polymers of lysine or arginine also stimulate its kinase activity, whereas the negatively charged substances such as polyglutamate or nucleic acids significantly inhibit the kinase activity. Spermine affects both the Km and Vmax of Cdc7 kinase for a minichromosome maintenance (MCM) substrate. We also found that histones, lysine- and arginine-rich basic proteins, can stimulate Cdc7 kinase activity, and a MCM complex in association with histone is a more efficient substrate of Cdc7 than the free MCM complex. These results identify potential cellular inhibitors and stimulators of Cdc7 kinase and suggest that Cdc7 may be another target of cellular polyamines and that histones may stimulate Cdc7-mediated phosphorylation of chromatin-bound substrates. Ectopic expression of an antizyme, known to reduce the cellular polyamine levels, resulted in reduction of Cdc7-mediated phosphorylation of MCM4 protein, suggesting physiological roles of polyamines in regulation of Cdc7 kinase activity in the cells.

regulated during cell cycle, and this may be achieved by regulation of the levels of the proteins (16 -18). So far, exogenous stimulators or inhibitors of the Cdc7 kinase are not known.
Here we examined effects of positively or negatively charged substances on Cdc7 kinase activity in vitro. The results indicate that Cdc7 kinase is stimulated by positively charged polymers such as spermine or polylysine or histones and is inhibited by negatively charged polymers including nucleic acids.
Proteins-GST-fused Cdc7 in a complex with the full-length ASK, reported previously (5), or a similar construct generated at Carna Bioscience Inc. was expressed in insect cells and purified as before (5). Cdc7 in a complex with the minimum ASK was generated by coexpression of N-terminal His-tagged Cdc7 and the minimum ASK segment (amino acid residues 171-350) in insect cells. The complex was purified by the nickel column. Mouse MCM2-4-6-7 complex was purified from insect cells as described (37). GST-fused Dfp1/Him1 and hemagglutinin-tagged Hsk1 proteins were coexpressed in insect cells, and the kinase was purified by glutathione-Sepharose 4B. Similarly, GST-tagged Hsk1 and hemagglutinin-tagged Dfp1/Him1 were coexpressed in insect cells and purified. GST-tagged Hsk1 alone was also similarly expressed and purified. Mrc1, tagged with RGS-penta-His at the N terminus, was expressed in insect cells and purified. 3 Transfection, Fluorescence-activated Cell Sorter Analyses, and Cell Extracts-Transfection of 293T cells, fluorescenceactivated cell sorter analyses of DNA content, bromodeoxyuridine incorporation, and preparation of Triton-soluble and -insoluble extracts were conducted as previously described (28).

Cdc7 Kinase Activity Is Stimulated by Polyamines in Vitro-
Previous results indicate that Cdc7 kinase may target serine/ threonine residues with an acidic environment (8,9). We, therefore, examined effects of various positively or negatively charged compounds on Cdc7 kinase activity in vitro. As shown in Fig. 1, polyamines such as spermidine and spermine showed significant stimulation of human Cdc7-ASK kinase activity. Maximum (6-fold) stimulation was observed at 1 mM spermidine or spermine at 8 mM MgOAc. Spermine also stimulated fission yeast Hsk1-Dfp1/Him1 kinase activity by 4 -5 times (Fig.  2), suggesting that polyamine-mediated stimulation of Cdc7 kinase activity may be conserved for different species. Polylysine at 10 M also showed a 4-fold stimulation, whereas polyglutamate at more than 100 M inhibited kinase activity almost completely (Fig. 1). These results indicate that positively charged polymers stimulate Cdc7 kinase activity, whereas negatively charged polymers inhibit it.
Kinetics of Cdc7 Activation by Spermine-We examined the effect of magnesium concentration on spermine-mediated activation of Cdc7 kinase activity. The optimum concentration of magnesium in the presence and absence of 2.5 mM spermine was 2-5 and 5-10 mM, respectively. More than a 10-fold stimulation by spermine was observed at 2 mM magnesium, whereas the stimulation was about 4-fold at 10 mM magnesium (Fig. 3). This result suggests that Cdc7 kinase activation by spermine may be related to the actions of magnesium.
We next titrated Cdc7 kinase in the presence and absence of spermine. At 140 nM Cdc7-ASK, almost 20-fold activation was observed, whereas only 4-fold activation was observed at 560 nM (supplemental Fig. 1).
We then measured K m and V max for the substrate in the presence and absence of spermine. The substrate in our assays is the 3 M. Yokoyama, manuscript in preparation.  MCM2-4-6-7 complex, in which MCM2 and MCM4 are mainly phosphorylated by Cdc7 (5,7,9). MCM2-4-6-7 was titrated in reaction mixtures containing 2 mM magnesium (Fig. 4). K m and V max for the substrate in the absence of spermine was 390 nM and 0.28 pmol/min, whereas those in the presence of 2.5 mM spermine were 125 nM and 0.85 pmol/min, respectively. Thus, both K m and V max for the MCM2-4-6-7 substrate were improved by 3-fold by the presence of spermine.
Cdc7 in Complex with Minimum ASK Cannot Be Efficiently Activated by Spermine-Polyamines may stimulate the Cdc7 kinase activity by directly interacting with the kinase complex, as was shown with casein kinase, the ␤ subunit of which interacts with spermine (19 -21). To answer this, we decided to examine the effect of spermine on the catalytic subunit alone. The fission yeast Hsk1 kinase, unlike budding yeast or mammalian Cdc7 kinase, has basal phosphorylation activity by itself (22). We have purified GST-Hsk1 alone or GST-Hsk1 in complex with Dfp1/Him1 and compared the effect of spermine (Fig. 5). Whereas the addition of spermine stimulated the Hsk1-Dfp1/Him1 complex-mediated phosphorylation of Mrc1 protein by 16-fold, it stimulated the Hsk1-mediated phosphorylation only by 2-fold. This indicates that the spermine exerts its stimulatory effect mainly through the Dfp1/Him1 activation subunit.
To define the domain on the activation subunit responsible for stimulation by spermine, we next compared full-length ASK and the "minimum" ASK for the response to increased concentration of spermine. The minimum ASK carries amino acids 171-350, containing Motif-M and Motif-C sufficient for activation of Cdc7 kinase. Indeed, the purified Cdc7-minimum ASK complex can phosphorylate MCM to the extent similar to that of Cdc7-full length ASK complex. However, the addition of spermine resulted in only less than a 2-fold stimulation of phosphorylation, whereas the Cdc7-full length ASK was stimulated by nearly 20-fold at an optimum spermine concentration (Fig.  6). This result indicates that, although the minimum ASK is sufficient for activation of Cdc7 kinase, the complex is not significantly activated by spermine, suggesting that the segment of ASK which responds to spermine may lie outside the minimum ASK. Because the long C-terminal tail of human ASK is diverged between species and is not present in the yeast counterparts, it is likely that Motif-N missing in the minimum ASK may be involved in stimulation by spermine. We speculate that Motif-N, a BRCA1 C-terminal domain-like motif nonessential for activation of Cdc7 (23,32), may play a role in recognition of the substrate. Because spermine differentially affects the level of phosphorylation of different substrates (data not shown), it may affect the kinase-substrate interactions through Motif-N.
Effect of Other Compounds on Cdc7 Kinase Activity in Vitro; Stimulation by Histones-We then examined other compounds (Fig. 7). All the positively charged compounds including protamine and melittin stimulated Cdc7 kinase activity. The 13-mer oligonucleotide containing five basic residues (lysine or arginine) slightly stimulated Cdc7 kinase activity in vitro, whereas lysine or arginine did not show any stimulation. On the other hand, negatively charged molecules, including polyU and heparin, inhibited Cdc7 kinase activity in vitro ( Fig. 7 and data not shown). The addition of DNA generally inhibited Cdc7 kinase (data not shown).
MCM associates with chromatin in late telophase to early G 1 , and Cdc7-mediated phosphorylation of MCM occurs on chromatin. Thus, histones in the chromatin may facilitate the Cdc7mediated phosphorylation of MCM proteins or other chromatin-associated substrates. It was reported before that histone H3 associates with MCM2 protein (24). Thus, we have examined whether MCM associated with histones may be phosphorylated more efficiently than free MCM. As shown in Fig. 9, histones were coimmunoprecipitated with MCM, and MCM in the complex was more efficiently phosphorylated by Cdc7 than the immunoprecipitated MCM alone. This result supports the  idea that histones may play a role in stimulating Cdc7-mediated phosphorylation of MCM protein and other substrates on the chromatin.
Overexpression of Antizyme Results in Decrease of Cdc7-mediated Phosphorylation in the Cell-Ornithine decarboxylase (ODC) is a key enzyme in polyamine biosynthesis, and the cellular polyamine levels can be altered by manipulating the cellular levels of ODC. Antizymes are involved in degradation of ornithine decarboxylase as well as in negative regulation of polyamine transport (38,39). Thus, overexpression of an antizyme is known to reduce intracellular polyamine contents. We have examined the effect of ectopic expression of an antizyme on DNA replication and cellular phosphorylation of MCM4, a known target of Cdc7 kinase (9). Transient transfection of AZ2 expression vector into 293T cells resulted in a decrease of cells incorporating bromodeoxyuridine (from 48.9% in vector-transfected cells to 18.9% in AZ2-transfected cells; Fig. 10A). Cdc7 is known to phosphorylate MCM2 and MCM4 on the chromatin during S phase. Thus, we have examined the phosphorylation state of MCM by using an antibody directed against a specific  1 and 8, 0.05%; lanes 2 and 9, 0.1%; lanes 3 and 10, 0.5%;  lanes 4 and 11, 1%; lanes 5 and 12, 2%; lanes 6 and 13, 5%; lanes 7 and 14, no spermine added. A, autoradiogram. B, silver staining. C, quantification of the results. The Mrc1 protein bands (upper bands) were cut out from the dried gel, their radioactivity (cpm) was measured, and the incorporation without spermine is taken as 100. The broad radioactive bands seen below the GST-Hsk1 protein (A, lower bands; lanes 1-7) represent the autophosphorylated Dfp1/Him1 protein, which appears as smeared bands under this experimental condition.  1-7) or Cdc7 in a complex with the minimum ASK (amino acid residues 171-350). A, the products were run on 7.5% SDS-APGE. Upper, autoradiogram; lower, silver staining. B, quantification of the results. The MCM2 bands were cut out, radioactivity (cpm) was measured, and incorporation without spermine is taken as 100.
serine residue (Ser-6) phosphorylated by Cdc7 (S6T7 antibody (9)). The S6T7 signal in the chromatin-enriched fraction was significantly lower in AZ2-transfected cells than that in control cells (Fig. 10B). Administration of ␣-difluoromethylornithine, an irreversible inhibitor of ODS, also led to decreased S6T7 signal (data not shown). These results are consistent with the idea that decreased cellular polyamine level affects Cdc7-mediated phosphorylation and, thus, the rate of DNA replication, suggesting that polyamines may be involved in regulation of Cdc7 kinase activity in the cells.

Stimulators and Inhibitors of Cdc7
Kinase in Vitro-It has not been known whether Cdc7-mediated phosphorylation events are regulated by exogenous proteins or factors, except that Lee et al. (30) reported that Cdc23 (MCM10) stimulates Hsk1 kinase activity in vitro. In this communication we examined whether positively or negatively charged polymers affect the Cdc7 kinase activity in vitro. We have shown that positively charged polymers such as polyamines and polylysine significantly stimulate Cdc7 kinase-mediated phosphorylation events in vitro, whereas negatively charged polymers such as polyglutamate or heparin inhibited the phosphorylation. Among various polyamines, spermine was most effective, and its stimulatory effect was more significant at lower concentration of magnesium, as was observed for casein kinase (31,33). Titration of substrates show that both affinity (K m ) and the maximum velocity (V max ) for MCM2-4-6-7 substrate is increased by 3-fold in the presence of 2.5 mM spermine.
Mechanisms of Stimulation and Inhibition-It is known that polyamines stimulate phosphorylation by casein kinase in vitro by binding to its regulatory subunit ␤ (21, 31). Cdc7 is a rather unique kinase in the phylogenetic tree of kinase families, but it is most related to casein kinase. The acidophilic nature of substrate recognition (5) is also shared by casein kinase and Cdc7 kinase. Negatively charged polymers may compete with the substrate target site embedded in the acidic environment for the substrate recognition surface of Cdc7 kinase complex, thus leading to strong inhibition of the kinase activity. We do not know the mechanism of spermine-mediated stimulation of Cdc7 kinase. The stimulatory effect of positively charged polymers appears to be conserved through species, since strong stimulation was observed also with the fission yeast Hsk1-Dfp1/Him1 kinase complex. Different substrates, MCM and Mrc1, are equally stimulated. Thus, it appears that polyamines may directly activate Cdc7 kinase complex. However, we have observed that different sets of polypeptides exhibit varied responses to spermine in Cdc7-mediated phosphorylation assays; phosphorylation of some polypeptides are highly stimulated by the presence of spermine, whereas that of others are not affected at all. 4 Furthermore, autophosphorylation of Hsk1 was not stimu-  lated by spermine, whereas it stimulated Mrc1 phosphorylation (see the phosphorylation of Hsk1 protein in Fig. 5). This suggests that polyamines may also interact with substrates, facilitating the Cdc7-mediated phosphorylation on specific substrates. Alternatively, interaction of polyamines with Cdc7 kinase complex may somehow facilitate its recognition of specific substrates.
The ␤ subunit of casein kinase possesses a segment rich in acidic residues that was shown to interact with polyamines (19 -21). The purified human Cdc7 polypeptide alone, which is completely inert as a kinase, was not activated by the pres-ence of spermine (data not shown), indicating that spermine is not able to replace the function of the activation subunit. Spermine activated Hsk1-Dfp1/Him1 kinase complex much more efficiently than it did Hsk1 protein alone (Fig. 5), consistent with the notion that the activation subunit is the major target of spermine. Spermine, however, could not very efficiently activate Cdc7 in a complex with minimum ASK (amino acids 171-350), which contains only the conserved motif M and C essential and sufficient for activation of Cdc7 kinase ( Fig. 6 (23, 32)). Thus, spermine interaction domain, if any, may be present outside the minimum ASK segment. We indeed detected binding of spermine to the Cdc7-ASK preparation we have used in the kinase assays (data not shown). At present, however, we are not able to conclude whether this is a genuine, specific interaction, or it simply reflects nonspecific ionic interaction between the positively charged compounds and negatively-charged surface of the kinase complex.
Potential Roles of Histone in Regulation of Cdc7 Kinase Activity-Histones significantly stimulated Cdc7 kinase activity in vitro, presumably due to their acidic nature. MCM2 protein has been shown to bind to histone H3 (24). We have shown that MCM-histone complex is phosphorylated by Cdc7 more efficiently than is the free MCM. Because Cdc7 regulates the processes of DNA replication at the origins as well as at replication forks, its major targets are on the chromatin. Thus, it is an intriguing possibility that histones facilitate the phosphorylation events mediated by Cdc7 on the chromatin.
Cdc7 as a Novel Target of Polyamines in Regulation of Cell Proliferation-Polyamines have been shown to play diverse roles in cell proliferation. They are known to regulate various signal transduction pathways and cell cycle progression through regulating the activity and abundance of transcription/post-transcription factors as well as through modulating protein-DNA interactions by their nucleic acid binding ability (34). Polyamines generally stimulate cell proliferation and carcinogenesis and have been regarded as promising molecular targets for cancer therapy (35,36). The fact that ectopic expression of an antizyme reduced the phosphoryl- Effect of various histone components (from Roche Applied Science) on Cdc7 kinase activity was examined in standard in vitro phosphorylation assays for Cdc7 kinase. The MCM2 bands were cut out, and radioactivity (cpm) was measured. The incorporation in the absence of compounds is taken as 100. All the histones were added at the concentrations of 20 ng/ml, 200 ng/ml, 2 g/ml, 20 g/ml, and 200 g/ml. Amounts of histones added (per 25 l reaction) are indicated in the graph. Histones stimulate Cdc7 kinase activity in vitro.  ation level of MCM4 (Fig. 10) is consistent with the physiological role of polyamine regulation of Cdc7 kinase. The findings in this report show the possibility that Cdc7 is another molecular target of polyamines and shed a novel light into the mechanism of polyamine-induced growth stimulation of normal and cancer cells.