Modulation of DNA Synthesis in Saccharomyces cerevisiae Nuclear Extract by DNA Polymerases and the Origin Recognition Complex*

We have analyzed the modulation of DNA synthesis on a supercoiled plasmid DNA template by DNA polymerases (pol), minichromosome maintenance protein complex (Mcm), topoisomerases, and the origin recognition complex (ORC) using an in vitro assay system. Antisera specific against the four-subunit pol α, the catalytic subunit of pol δ, and the Mcm467 complex each inhibited DNA synthesis. However, DNA synthesis in this system appeared to be independent of polϵ. Consequently, DNA synthesis in the in vitro system appeared to depend only on two polymerases, α and δ, as well as the Mcm467 DNA helicase. This system requires supercoiled plasmid DNA template and DNA synthesis absolutely required DNA topoisomerase I. In addition, we also report here a novel finding that purified recombinant six subunit ORC significantly stimulated the DNA synthesis on a supercoiled plasmid DNA template containing an autonomously replicating sequence, ARS1.

Studies on the replication of genomes of bacteriophages, plasmids, viruses, and prokaryotic and eukaryotic cells have established general mechanisms of DNA replication (1). The eukaryotic origins of DNA replication were first discovered in the budding yeast, Saccharomyces cerevisiae, and were named autonomously replicating sequences (ARS) 1 (2)(3)(4). Many yeast ARS elements have been characterized in detail, which has allowed the discovery of many cis-acting replication proteins and regulators of DNA replication (5,6). The discovery by Bell and Stillman (7,8) that the in vivo initiation of DNA replication in S. cerevisiae requires a six subunit origin recognition complex (ORC) is a major advancement in understanding the mechanism of eukaryotic DNA replication. In the G 1 phase of the cell cycle, Cdc6p, Cdt1p, minichromosome maintenance (Mcm), and Cdc45p proteins bind sequentially to DNA to form a prereplication complex (pre-RC) (9 -11). At the G 1 /S boundary, S phase-specific cyclin-dependent protein kinases (Cdks) and particularly the Cdc7/Dbf4 kinase transform the pre-RC into an active replication complex (10 -12). Finally, the initiation coincides with the association of the polymerase ␣/primase (pol ␣/primase) complex to the RPA-coated unwound origins (10). S and M phase-specific Cdks block the re-binding of Mcm to the chromatin and prevent a new round of initiations until mitosis (13).
The Mcm protein family that is required for DNA replication in eukaryotes consists of six proteins (Mcm 2-7) with highly conserved amino acids between the six polypeptides. All of these proteins are essential for the cell viability (13,14). The Mcm467 proteins form a relatively stable core complex and other Mcm proteins are loosely associated to it (14 -17). It has been demonstrated that a DNA helicase activity is associated with the Mcm467 complex, suggesting that this complex is involved in the initiation of DNA replication as a DNA unwinding enzyme (18,19). Furthermore, Mcm467 remains associated with the replication fork during the elongation step of DNA replication (20). Among a number of candidate DNA helicases, Mcm467 DNA helicase appears to be the replication fork DNA helicase (14, 18 -22).
Previous studies in various laboratories have identified three DNA polymerases: ␣ (pol ␣), ␦ (pol ␦), and ⑀ (pol ⑀) that are required for the eukaryotic DNA replication (23). Pol ␣ has both primase and polymerase activities making it suitable for priming well as the extension of the primers toward the synthesis of Okazaki fragments (24,25). Earlier, in vitro SV40 DNA replication, reconstituted with purified proteins, has defined the roles of many components of the replisome (26 -29). Once RNA-DNA primers are synthesized by pol ␣/primase complex on RPAcoated, single-stranded DNA template (30), RFC binding triggers the loading of PCNA, and association of pol ␦ and the RNA primers are extended (23). Unlike pol ␦, the role of pol ⑀ remains unclear. Pol ⑀ is a highly processive enzyme that has been implicated in the genomic DNA replication, DNA repair, recombination, as well as the S phase checkpoint control (31)(32)(33)(34)(35)(36).
The in vitro system used in our study for the dissection of DNA synthesis requires nuclear protein extract from budding yeast S. cerevisiae, synchronized to the S phase (37,38). DNA synthesis in this system is bidirectional, aphidicolin-sensitive, semi-conservative, and requires supercoiled plasmid as DNA template (37). Even though the initiation event does not require an ARS sequence, DNA synthesis is observed only in the protein extract prepared from S phase yeast nuclei (37,38). However, it is not known whether yeast ORC may play a role in the DNA synthesis in this system. As the plasmid DNA template must be supercoiled, the question arises as to whether a DNA topoisomerase (Topo) is necessary in this in vitro system. Both Topo I and II can relax positively and negatively supercoiled DNA (39 -41). It has been established earlier that the function of Topo II is essential during mitosis in order to separate intertwined sister chromatids, while Topo I acts as a swivelase during DNA replication and transcription (39 -43). It has been shown to be required in the late stage of DNA syn-thesis in SV40 DNA replication (43).
In this report, we have analyzed the roles of yeast DNA polymerases, Mcm467 helicase, ORC complex, and DNA topoisomerases in the in vitro plasmid DNA synthesis. Our results suggest that DNA synthesis in the nuclear extract of S. cerevisiae depended on Topo I, pol ␣, and ␦, ORC, as well as Mcm467 helicase but did not require pol ⑀. We also report here that recombinant yeast ORC significantly stimulated the in vitro DNA synthesis on supercoiled plasmid DNA templates.
Nucleic Acids, Enzymes, Plasmids, and Other Reagents-Ultrapure deoxy-and ribonucleotides were obtained from Amersham Biosciences (Piscataway, NJ) and were used without further purification. Radiolabeled nucleotides ([␣-32 P]dATP, [␣-32 P]CTP, and [␣-32 P]GTP) were obtained from PerkinElmer Life Sciences (Boston, MA). Yeast RPA and recombinant yeast PCNA were purified as described previously (44,45). Zymolase was purchased from Seikagaku America Inc. (Falmouth, MA). Peptide ␣-factor was obtained from Research Genetics Inc. (Mobile, AL). NBT and BCIP, were purchased from Invitrogen. All other chemicals used to prepare buffers and solutions were reagent grade and were purchased from Fisher. Protease inhibitors, leupeptin, pepstatin A, and antipain, were obtained from Bachem Inc. (Los Angeles, CA). Plasmid Bluescript pSK, termed pSK, was obtained from Stratagene (La Jolla, CA). pSKARS1 was designed by cloning 837-bp HindIII/EcoRI fragment of the ARS1-Trp1 gene from plasmid yRP7 into the HindIII/EcoRI site of Bluescript pSK. Nicked and linear DNAs were selectively removed by acid-phenol extraction. The purified plasmid was greater than 95% in supercoiled form. Recombinant S. cerevisiae RPA was expressed in Escherichia coli from an expression plasmid obtained from Dr. Steven Brill of the Department of Molecular Biology and Biochemistry, Rutgers University, Piscataway, NJ (46).
Yeast Synchronization and Yeast Nuclear Extract Preparation-Wild-type or mutant yeast cells were grown in YPD medium in a shaker incubator at 30 or 25°C, respectively. S phase-synchronized yeast cells were obtained by ␣-factor synchronization as described (47). Synchronized cells were resuspended in buffer A and stored at Ϫ80°C. Unsynchronized cells were grown in a shaker incubator at 30°C (wild type) or 25°C (mutant strains) in YPD medium to OD 600 of 0.6, harvested and stored as described above. Spheroplasts were obtained following the procedure described by Verdier et al. (47). Nuclear extracts were prepared as described previously (38). Fractionation of yeast nuclear extract by phosphocellulose chromatography was carried out as described earlier (38).
Expression and Purification of Yeast ORC-ORC was purified following a modification of a procedure described earlier by Lee and Bell (48). Briefly, the yeast cell extract was fractionated using S-Sepharose and then Q-Sepharose chromatography. Following this, the glycerol gradient step of Lee and Bell was replaced with size exclusion liquid chromatography (S.E.-HPLC). All chromatographic procedures were carried out using a BIOCAD HPLC (Applied Biosystems Inc., Worcester, MA). The S.E.-HPLC utilized a Superose 6 column (AP Biotech, Piscataway, NJ), equilibrated and eluted with buffer B containing 150 mM NaCl at the rate of 0.4 ml/min. A total of 75 fractions were collected. Pure ORC complex eluted out in fractions 33-36. The ORC-containing fractions were pooled, dialyzed against buffer C, and concentrated by ultrafiltration, and analyzed by SDS-PAGE (Fig. 6A).
In Vitro DNA Synthesis Assay-The replication reaction was carried out in a 25 l of reaction mixture in buffer E containing 300 ng of supercoiled plasmid DNA, 40 mM creatine phosphate, 0.125 mg/ml creatine phosphokinase, 4 mM ATP, 70 M of each rNTP, 100 M of each dNTP, 0.125 Ci of [␣-32 P]dATP, and proteins, as indicated (33). Standard assays were carried out for 60 min at 30°C (or as indicated). The optimal protein concentration of each extract was determined by titration. Inhibition studies were carried out under identical conditions, except that the extracts were preincubated with specific antisera for 15 min at 30°C in the absence of template DNA and the replication reactions were initiated by the addition of a plasmid DNA template. Replication reactions in the presence of recombinant ORC were done in the same reaction mixture; however, the plasmid template was preincubated with recombinant ORC for 30 min at 30°C in the presence of all other reagents, except dNTPs. The replication reactions were initiated upon addition of dNTP and S phase nuclear extract. The reactions were terminated by the addition of 1 ml of 10% trichloroacetic acid/0.1 M sodium pyrophosphate, followed by incubation for 15 min on ice. The reaction mixtures were filtered onto GFC glass microfiber filters (Whatman Inc.) and counted in a liquid scintillation counter.
Analysis of the Products of DNA Synthesis on Agarose Gel-In vitro replication assay was carried as described above, over a time range of 5-60 min. The reactions were terminated by the addition of 1 volume of buffer F containing 20 g/ml RNase A and then 100 g/ml proteinase K. After proteinase K digestion for 30 min at 37°C, the plasmid DNA was phenol-extracted, twice ethanol-precipitated, and digested with the restriction endonucleases, EcoRI and HindIII, in order to excise the ARS fragment. The DNA fragments were analyzed on a 1% agarose gel and stained with ethidium bromide. The gels were dried and exposed to Fuji RX film at Ϫ80°C for 2-24 h.
Polyacrylamide Gel Analysis of the Products of DNA Synthesis-Reactions were carried out as described above, except with 5 Ci of [␣-32 P]dATP or 2.5 Ci of each [␣-32 P]CTP and [␣-32 P]GTP were added. They were terminated by the addition of 2 l of 500 mM EDTA/NaOH (pH 8). The DNA products were ethanol-precipitated and resuspended in 1ϫ formamide loading buffer G. The samples were heated at 95°C for 2 min and loaded on a 7 M urea/6% polyacrylamide gel (urea-PAGE) in 1ϫ Tris borate/EDTA (TBE). Electrophoresis was carried out for 4 h in 1ϫ TBE buffer. Gels were dried and exposed to Fuji RX film at Ϫ80°C for 12 h.
Immunological Analysis-Polyclonal antisera against the four subunits of pol ␣, recombinant p125 of pol ␦, recombinant Mcm467 complex, and recombinant 90 kDa N terminus fragment of the catalytic subunit of pol ⑀ were raised in mice using standard immunological protocols. Antibodies (IgG fraction) against Mcm467 complex were further purified by ammonium sulfate (33% saturation) precipitation and dialysis against PBS. Inhibition of the DNA helicase activity was analyzed using a helicase assay, as described by Lee and Hurwitz (19). Western transfer was performed using a Bio-Rad Trans-blot apparatus according to the manufacturer's instructions. Antigen-antibody reaction was detected by utilizing alkaline phosphatase-conjugated second antibody.
Other Methods-Protein concentrations were determined by the method of Bradford (49) using bovine serum albumin as a standard. SDS-PAGE was carried out according to Laemmli (50).

Mcm467 DNA Helicase Is Necessary for in Vitro DNA Synthesis-The
Mcm complex is believed to function as a replicative helicase of Eukaryotes and Archaea (18). An important question is whether the replication fork is driven by the Mcm467 helicase. The presence of Mcm467 in both G 1 phase (wild-type NE G1 ) and S phase (wild-type NE S ) nuclear extracts of wild-type S. cerevisiae was tested by Western blotting using mouse polyclonal antibodies specific against these proteins (Fig. 1A). It was found that Mcm467 proteins were present in wild-type NE G1 derived from cells that were synchronized in the G 1 phase (Fig. 1A, lane 1), as well as in wild-type NE S derived from cells that were synchronized in the S phase of the cell cycle (Fig. 1A, lane 2). In the case of wild-type NE S , the bands corresponding to Mcm4 and Mcm6 were accompanied by additional cross-reacting bands of slightly slower mobility, which are likely caused by phosphorylation of these proteins in the S phase of the cell cycle (Fig. 1A, lane 2). Other posttranslational modifications could not be ruled out.
In order to determine whether Mcm467 complex acts as a helicase in the system, we studied the inhibition effect of anti-Mcm467 antibodies in the in vitro supercoiled DNA synthesis assay (Fig. 1B). Anti-MCM467 antibodies strongly inhibited (Ն75%) the helicase activity of purified Mcm467 complex (data not shown) (27). Wild-type NE S was preincubated with increasing concentrations of anti-Mcm467 antibodies or nonimmune IgG, and the plasmid replication was initiated by the addition of 300 ng of pSKARS and carried out as described under "Experimental Procedures." In vitro plasmid replication of pSKARS was inhibited (ϳ60%) with polyclonal antisera specific against Mcm467 complex (Fig. 1B).
Additional evidence for the requirement of Mcm467 complex for in vitro DNA synthesis in this system was obtained after analysis of the Okazaki fragments synthesized in the presence of Mcm467 IgG (Fig. 1C). Wild-type NE S was preincubated with increasing concentrations (2, 5, and 10 g) of anti-Mcm467 IgG (Fig. 1C, lanes 2-4), or increasing concentrations (2, 5, and 10 g) of nonimmune IgG (Fig. 1C, lanes 5-7) or in the absence of IgG (Fig. 1C, lane 1). DNA synthesis was initiated after the addition of supercoiled plasmid DNA template, and the replication intermediates labeled with [␣-32 P]dATP were analyzed on a urea-PAGE gel (Fig. 1C). Products ranging in size from 50 to Ն1500 bp were observed in the absence of an antibody (Fig. 1C, lane 1) as well as in the presence of nonimmune IgG (Fig. 1C, lanes 5-7). In contrast, preincubation of wild-type NE S with anti-Mcm467 IgG led to a significant inhibition of the Okazaki fragment synthesis (Fig. 1C, lanes 2-4). The ladder of replication intermediates decreased with increasing concentrations of anti-Mcm467 IgG (Fig. 1C, lanes 2-4).
Thus, anti-Mcm467 IgG inhibited the DNA synthesis significantly in this in vitro system. Therefore, the Mcm467 complex appeared likely to be the functional DNA helicase in this system.
Pol ␣ and ␦ Are Absolutely Required in in Vitro DNA Synthesis-Wild-type NE S was fractionated by phosphocellulose chromatography as described earlier (38). The fraction containing the major polymerase activity (wild-type Pol S fraction), bound to phosphocellulose, and the unbound fraction (wild-type Aux S fraction) were analyzed for the presence of DNA polymerase ␣/primase complex (pol ␣) and DNA polymerase ␦ (pol ␦) by Western blotting using mouse antisera specific against the whole pol ␣/primase complex, and catalytic subunit of pol ␦ (125 kDa) ( Fig. 2A). Wild-type Pol S fraction contained the pol ␣/primase complex (Fig. 2A, lane 3), as well as pol ␦ (Fig. 2A, lane 5), whereas wild-type Aux S fraction was completely devoid of these DNA polymerases ( Fig. 2A, lanes 4 and 6, respectively).
Wild-type Pol S fraction was preincubated with antiserum specific against pol ␣/primase complex, pol ␦, or nonimmune serum. The system was reconstituted upon addition of purified yeast RPA, recombinant PCNA, and supercoiled plasmid DNA template (38). Inhibition of DNA synthesis was determined by the incorporation of [␣-32 P]dATP for 60 min as described under "Experimental Procedures." Titrations of wild-type Pol S frac- tion with different amounts of anti-pol ␣, anti-pol ␦, and nonimmune serum are shown in Fig. 2B. The maximum inhibition observed with anti-pol ␣ and anti-pol ␦ sera were 79 and 73%, respectively, suggesting that both of these DNA polymerases are required for in vitro replication of plasmid DNA. In contrast, nonimmune serum did not interfere significantly (less then 10% inhibition) with the extent of the DNA synthesis (Fig. 2B).
The 32 P-labeled products of DNA synthesis inhibited by antipol ␣, anti-pol ␦, or nonimmune serum were analyzed on urea-PAGE (Fig. 3). The products in the absence of an antiserum or in the presence of nonimmune serum could be observed as a ladder of replication intermediates (Fig. 3B, lanes 1-4, respectively). Similar distribution of replication intermediates was detected as a result of [␣-32 P]CTP/[␣-32 P]GTP incorporation into the RNA primers and their extension with unlabeled dNTP to Okazaki fragments (Fig. 3A, lanes 1-4, respectively). When wild-type Pol S fraction was preincubated with anti-pol ␣ or anti-pol ␦ sera, in vitro incorporation of both [␣-32 P]dATP (Fig. 3B, lanes 5-10, respectively) and [␣-32 P]CTP/[␣-32 P]GTP (Fig. 3A, lanes 5-10, respectively) was remarkably inhibited and a ladder of replication intermediates was not detected. The accumulation of smaller products (Ͻ72 bp) with [␣-32 P]CTP/[␣-32 P]GTP incorporation suggested that the elongation had been inhibited in an early stage of Okazaki fragment synthesis by both anti-pol ␣ and anti-pol ␦ sera (Fig. 3A, lanes 5-10, respectively). Some inhibition of DNA synthesis was observed at higher amounts of antiserum, which was likely caused by the presence of ribonuclease in the antiserum.
DNA Polymerase ⑀ in the in Vitro DNA Synthesis-We have analyzed the in vitro plasmid replication for a possible function of pol ⑀ using an S phase synchronized nuclear extract of ⌬pol⑀ mutant of S. cerevisiae containing a pol ⑀ N-terminal active site truncation. The maximum extent of DNA synthesis observed with NE S derived from ⌬pol⑀ (19 pmol) was similar to that observed with wild-type NE S (19 pmol) as shown in Fig. 4A. In the absence of functional pol ⑀, in vitro replication of supercoiled plasmid template remained unchanged compared with the wildtype extract (Fig. 4A). This observation was consistent with the further analysis of the 32 P-labeled replication intermediates on urea-PAGE gel (Fig. 4B). Plasmid replication was carried out with increasing concentrations (17-68 g) of either wild-type NE S (Fig. 4B, lanes 1-5, respectively) or ⌬pol⑀ NE S (Fig. 4B, lanes 6 -10, respectively). With both extracts, a ladder of replication intermediates (size range 50 -1500 nt) could be observed after incorporation of [␣-32 P]dATP into the newly synthesized DNA chains (Fig. 4B, lanes 1-5 and 6 -10, respectively).
Western blotting analysis, using mouse anti-sera prepared against a 90 kDa N-terminal fragment of the catalytic subunit of pol ⑀, confirmed its presence in the wild-type Pol s fraction ( Fig. 2A, lane 7) but not in the wild-type Aux S fraction (Fig. 2A,  lane 8). The anti-pol ⑀ serum possessed significant inhibition of the polymerase activity of purified yeast pol ⑀ in activated calf thymus DNA assay (not shown). In order to assess the role of pol ⑀ in the in vitro plasmid replication, wild-type Pol S fraction was pre-incubated with either antiserum specific against pol ⑀ or nonimmune serum (Fig. 2B). In contrast to anti-pol ␣ and anti-pol ␦ sera, anti-pol ⑀ sera did not show any significant inhibition of the replication reaction and the extent of inhibition of DNA synthesis was comparable to that obtained with nonimmune control serum (Fig. 2B). In addition, the products of DNA synthesis in the presence or absence of anti-pol ⑀ or nonimmune serum were analyzed on urea-PAGE after labeling with [␣-32 P]dATP (data not shown). Our results suggest that pol ␣/primase complex and pol ␦ are primarily responsible for carrying out the DNA synthesis in vitro.
Supercoiled Plasmid DNA Replication Requires Topo I-DNA synthesis in this in vitro system requires supercoiled plasmid DNA template, which led to the hypothesis that a topoisomerase is required. In order to test this hypothesis, we examined yeast topoisomerases in this system by using two topoisomerase mutant strains of S. cerevisiae: ⌬Top1 in which the Top1 gene was completely deleted, and top1-1top2-1 ts with several point mutations in both the Top1 and Top2 genes. In the second strain, the Top2 gene possesses normal function at permissive temperature (25°C), the activity of Topo II is eliminated by shifting to restrictive temperature (37°C), whereas the top1 mutations eliminate Topo I activity and is not temperature-sensitive. In order to keep Topo II functional, the top1-1top2-1 ts cells were grown and synchronized at 25°C. We have compared DNA synthesis using supercoiled pSKARS1 plasmid DNA template in wild-type, ⌬Top1, and top1-1top2-1 ts yeast nuclear extracts, as well as Pol fractions derived from synchronized and unsynchronized cells, respectively. The Topo I and Topo II activities were separated in the phosphocellulose chromatography and fractionated in the Pol and Aux fractions respectively (38). Topo I activity in the top1-1top2-1 ts strain was measured in the Pol s fraction, which is devoid of Topo II activity. Topo I activity was not detectable in the Pol s fraction of top1-1top2-1 ts strain, consistent with the findings of Brill et al. (39,40).
High levels of plasmid DNA replication were observed only in wild-type NE S (Fig. 5A). In contrast, ⌬Top1 and top1-1top2-1 ts NE S demonstrated significantly reduced levels of DNA synthesis (Fig. 5A). The extent of the DNA synthesis was ϳ18 pmol for wild-type NE S , ϳ4 pmol for ⌬Top1 NE S , and ϳ2 pmol for top1-1top2-1 ts NE S , respectively. The results with the Pol S fractions that were obtained from wild-type, ⌬Top1, and top1-1top2-1 ts NE S extracts suggest that in vitro DNA synthesis required primarily Topo I (Fig. 5, A and B). The extent of DNA synthesis was ϳ19.8 pmol for wild-type Pol S fraction, ϳ2.5 pmol for ⌬Top1 Pol S fraction, and ϳ2.8 pmol for top1-1 top2-1 ts Pol S fraction. These results suggest a significant role of Topo I in the in vitro DNA synthesis. The role of Topo II remained unclear. A small degree of dependence on Topo II was observed in the absence of Topo I in the ⌬Top1 NE s (Fig. 5A). In the absence of a functional Topo I, the DNA synthesis was inhibited in both ⌬Top1 Pol S and in top1-1top2-1 ts Pol S fractions (Fig. 5B). It should be noted that Topo II remained in the Aux fraction of phosphocellulose chromatography and the Pol S fraction contained Topo I and was devoid of Topo II (data not shown). Consequently, a small difference in DNA synthesis observed in the nuclear extracts but not in the Pol fractions could be due to Topo II in the nuclear extract (Fig. 5, A and B).
Exogenous ORC Stimulates DNA Synthesis in Vitro-The initiation of replication in the yeast, S. cerevisiae, in vivo requires a six subunit origin recognition complex (ORC) (8). The function of ORC is to initiate a competent pre-RC in early G 1 phase, by recruiting Cdc6p and Cdt1p, which in turn load the hexameric Mcm complex (8).
Supercoiled plasmids were preincubated with increasing The reactions were carried out in a standard DNA synthesis assay as described in the "Experimental Procedures." The products were ethanol-precipitated and analyzed on a 7 M urea-PAGE followed by autoradiography.

FIG. 5. Requirements of topoisomerase I in the in vitro DNA
synthesis. DNA synthesis was analyzed using nuclear extracts derived from ⌬Top1 and top1-1top2-1 ts mutant strains of S. cerevisiae. A, wildtype NE S (f), ⌬Top1 NE S (OE), and top1-1top2-1 ts NE S (E). B, wild-type Pol S fraction (f), ⌬Top1 Pol S fraction (OE) and top1-1top2-1 ts Pol S fraction(E). Replication reactions were carried out as described in "Experimental Procedures" upon addition of 300 ng of plasmid pSKARS1.
concentrations of recombinant ORC (0.6 -5 g) in the presence of ATP as described under "Experimental Procedures." Replication reactions were further initiated upon addition of wildtype NE S (40 g) and dNTP (Fig. 6B). In the absence of recombinant ORC, the DNA synthesis of wild-type NE S was ϳ18 pmol (Fig. 6B). The extent of DNA synthesis remained almost unaffected in the presence of 0.6 -2 g of exogenous ORC and significantly increased above 2 g of exogenous ORC in a 25-l assay (Fig. 6B). The maximum DNA synthesis observed with 40 g of wild-type NE S and 5 g of recombinant ORC (ϳ35 pmol) was significantly higher than that observed in the absence of ORC (Fig. 6B). Our results indicate that the addition of 5 g of exogenous recombinant ORC stimulated in vitro plasmid replication roughly 2-fold in NE S of S. cerevisiae.
We have also analyzed whether ORC stimulation of in vitro plasmid DNA synthesis was dependent on the presence on ARS in the plasmid. In order to ascertain the requirement of ARS, we have compared DNA synthesis with plasmids pSKARS, containing ARS1 and pSK (Fig. 7). Replication reactions with pSKARS and pSK templates, shown in Fig. 7A, were carried out with increasing concentrations of NE S , following preincubation with 4 g of ORC or without ORC. In the absence of exogenous ORC, DNA syntheses with both plasmid templates in the NE S of S. cerevisiae were similar. The maximum extent of DNA synthesis was ϳ14 pmol for pSKARS and ϳ12 pmol for pSK, respectively (Fig. 7A). However, the addition of recombinant exogenous ORC into this in vitro system significantly stimulated replication of pSKARS (ϳ2-fold), whereas, replication of pSK remained unchanged (Fig. 7A).
In the absence of exogenous ORC, both plasmids were equally replicated in the NE S and the maximum extent of DNA synthesis was ϳ18 pmol for pSKARS and ϳ19 pmol for pSK, respectively (Fig. 7B). Under these conditions, control NE G1 demonstrated significantly lower DNA synthesis for pSKARS and pSK (Fig. 7B). Our previous results suggest that the DNA synthesis of NE G1 does not require either RPA or PCNA and is not a bona fide plasmid replication (23). Therefore, DNA synthesis NE G1 is a reasonable assessment of the level of nonspecific TCA-precipitable 32 P label in this system. In the presence of exogenous ORC, the maximum DNA synthesis observed with NE G1 remained unchanged for both plasmids (Fig. 7B). ORC did not alter the levels of DNA synthesis with plasmid pSK (Fig. 7B). In contrast, exogenous ORC caused significant increase of the DNA synthesis with plasmid pSKARS as a template (Fig. 7B). Consequently, exogenous ORC stimulated the in vitro plasmid DNA replication in the NE S of S. cerevisiae.
We further analyzed the products of pSKARS replication after 15, 30, and 60 min of incubation in the NE S of S. cerevisiae, in the presence or in the absence of exogenous ORC (Fig.  7C). The replication products were digested with EcoRI and HindIII (in order to excise the ARS fragment), and analyzed in an agarose gel electrophoresis followed by autoradiography (Fig. 7C). The addition of exogenous ORC into NE S significantly increased the appearance of the labeled DNA synthesis products with increasing incubation times (Fig. 7C). DISCUSSION Much progress has been made in the identification of eukaryotic proteins that are likely involved in the various stages of chromosomal DNA replication (6,9,(51)(52)(53)(54). Chromosomal replication origins or ARSs of S. cerevisiae are the most well characterized in eukaryotes (4,5,8,52,55). ORC binding to replication origins has been shown to be required for the activation of the replication origins in Xenopus and Drosophila eggs (56,57). In addition, a number of other G 1 phase proteins including Cdc6, Mcm 2-7 proteins, and pol ␣ are required for the initiation of DNA replication (9). Mcm protein complex undergoes rearrangement to Mcm467 complex with DNA helicase activity, although its in vivo function as helicase remains unclear (18,58,59). Mcm and ORC proteins are also phosphorylated in the G 1 to S transition (59,60). Recent studies also established that successful DNA synthesis and fork progression in the S phase require participation of the DNA polymerases, ␣, ␦, and perhaps ⑀ (23,31,32,36,37,53,61). Therefore, in vitro analyses to determine putative roles of these proteins in DNA replication are essential.
Biochemical studies with the various Mcm protein complexes in yeast and mammals have shown that a dimeric complex of Mcm467 heterotrimer contains 3Ј 3 5Ј helicase, ssDNA binding, and DNA-dependent ATPase activities, whereas its interaction with either Mcm2 or Mcm3-5 inhibit the helicase activity (19,22,62). The Mcm complex is believed to function as a replicative helicase of Eukaryote and Archaea (18,19). Although Mcm4, -6, and -7 proteins are present in both wild-type NE G1 and wild-type NE S , Mcm4 and Mcm6 proteins appeared to be in part phosphorylated or modified in wild-type NE S (Fig. 1). Polyclonal antibodies specific against Mcm467 complex significantly inhibited replication of pSKARS and virtually eliminated the synthesis of Okazaki fragments (Fig. 1C). Therefore, Mcm467 is the likely DNA helicase in this in vitro system. Earlier studies by Braguglia et al. (37) indicated that DNA2 helicase/exonuclease is also required in the in vitro system. DNA2 helicase/exonuclease has been shown to have a possible function in the Okazaki fragment maturation although the helicase activity is dispensable for growth (63,64). Therefore, Mcm467 DNA helicase is likely required for progression of the replication fork, whereas, DNA2 helicase/exonuclease is required in the downstream processing of the Okazaki fragments. Therefore, it is reasonable that both Mcm467 and DNA2 are required for in vitro replication in the yeast nuclear extract. However, further studies are needed for complete delineation of the temporal functions of these two enzymes during S phase of the cell cycle particularly in terms of their DNA helicase activities.
Pol ␣, ␦, and ⑀ have been identified as the principal DNA polymerases that are involved in the eukaryotic DNA replication (23,31,32,36,37,53,61). Among them, pol ␣ has the unique function of catalyzing the initiation of DNA synthesis in conjunction with a primase activity. The DNA pol ␣/primase complex synthesizes RNA-DNA primers of about 35 nucleotides for DNA synthesis on the leading and on the lagging strands, culminating in the synthesis of Okazaki fragments (30,65). It has been postulated that after completion of the primer syn- thesis, DNA pol ␣/primase complex is displaced from the replication fork by a polymerase switch mechanism that is coordinated by RFC and PCNA (29,66,67). Subsequently, the RNA-DNA primers are elongated by other DNA polymerases (23,29).
We have carried out a partial fractionation of the S phase nuclear extract of S. cerevisiae by phosphocellulose chromatography which may be a first step toward complete reconstitution of the in vitro plasmid replication (38). Pol S fraction, devoid of some accessory factors but enriched in DNA pol ␣/primase complex, pol ␦ and pol ⑀ ( Fig. 2A), is capable of supporting the replication reaction upon addition of purified RPA and recombinant PCNA (38). Titration of wild-type Pol S fraction with antisera specific against pol ␣/primase complex and catalytic subunits of pol ␦ demonstrated a dose-dependent inhibition of DNA synthesis (Fig. 2B) comparable to anti-RPA and anti-PCNA sera (38). Additionally, analysis of the replication products after inhibition of in vitro plasmid DNA replication with anti-pol ␣ and anti-pol ␦ sera showed a remarkable reduction of the ladder of replication intermediates as visualized by denaturing polyacrylamide gel electrophoresis (Fig. 3). It has been previously shown that both pol ␣ and ␦ are required for plasmid DNA replication, and our results are in agreement with the findings of Braguglia et al. (37). In addition to a physical interaction, there might be functional coupling between these two DNA polymerases (68,69). S. cerevisiae protein YJR043C (Pol32) interacts with the catalytic subunits of pol ␣ and it is required for the cell cycle progression in G 2 /M phase (68). In contrast, anti-pol ⑀ sera did not demonstrate significant inhibition of the DNA synthesis (Fig. 2B) and did not lead to a reduction of the replication intermediates (Fig. 4C). Furthermore, in vitro DNA synthesis was successfully reconstituted in NE S derived from cells of ⌬pol⑀ mutant (Fig. 4, A and B) and demonstrated the same pattern of replication intermediates as the wild-type NE S (Fig. 4B). At the least, in vitro replication of supercoiled plasmid DNA in vitro depends on pol ␣/primase complex and pol ␦. Similarly, it has been shown that SV40 DNA replication is also pol ⑀-independent (23), but this polymerase might be required for the chromosomal DNA replication in vivo (33). It has been shown that pol ⑀ is one of the replicative polymerases in HeLa cells (35). Studies from several laboratories suggest that pol ⑀ is involved in the cell cycle checkpoint control and may function to ensure accurate DNA synthesis (31,34,70,71). However, the role of pol ⑀ remains unclear (31,36). Recently, it has been shown that pol ␦ depletion causes a more severe defect of elongation in Xenopus egg extracts compared with pol ⑀ depletion and suggests the possibilities that pol ␦ is essential for lagging strand DNA synthesis, and its function cannot be substituted for pol ⑀ (32). Since both leading and lagging strand are somehow synthesized in the absence of pol ⑀, albeit at a slow rate, Fukui et al. (32) speculated that pol ⑀ might be involved in elongation by affecting the fork progression. In the absence of pol ⑀, synthesis of fragments longer than 2 kb could be observed (32). Therefore, it is possible that small DNA templates, such as pSKARS (3.8 kb), could be easily replicated by pol ␣/pol ␦ in the nuclear extracts of S. cerevisiae without involving pol ⑀. This may also be true for SV40 DNA replication in vitro in the mammalian cell extract as well as in the reconstituted systems (28,61,72). Further studies with very large plasmid DNA template are required for evaluating this possibility.
The requirement of Topo I in the in vitro DNA synthesis was obtained by using two isogenic strains of S. cerevisiae: ⌬Top1, a Topo I deletion mutant, and top1-1top2-1 ts , in which Topo I is not functional while Topo II is temperature sensitive (39,40). Reconstituted systems with both NE S and Pol S fraction derived from synchronized S phase ⌬Top1 and top1-1top2-1 ts mutants did not show appreciable specific DNA synthesis in comparison to wild-type NE S and wild-type Pol S fraction (Fig. 6). In the absence of Topo I, Topo II could be capable of carrying out the unwinding of negatively supercoiled DNA, but it is not capable of supporting in vitro plasmid DNA replication in the NE S of S. cerevisiae. These results are in a good agreement with the earlier observations that inactivation of Topo II does not alter nascent DNA chain synthesis in vivo significantly, whereas the absence of Topo I leads to a significant delay in chain elongation (39,40,43,73). Thus, in wild-type cells, DNA Topo I is likely the major replication swivel in the early stages of S phase.
The initiation of DNA replication in S. cerevisiae is similar to that observed in bacteria in that it occurs at well defined origins of DNA replication that are recognized by specific initiator proteins (8,9,74). In the early S phase, it is likely that nuclear extracts of S. cerevisiae contain low levels of ORC, which could be due to its tight binding to the chromatin (75). ORC appeared to play a role in this in vitro system. Addition of recombinant exogenous ORC led to a significant stimulation of  (Ⅺ). B, DNA synthesis with pSK and pSKARS plasmid DNA templates with or without ORC (4 g) were carried out as in A, with either NE S or NE G1 nuclear extracts. C, analysis of ORC-stimulated DNA synthesis on pSKARS plasmid DNA templates. DNA synthesis was carried out with 35 g of wild-type NE S for 15, 30, and 60 min, respectively, with or without exogenous ORC. The DNA synthesis products were digested with EcoRI and HindIII and analyzed on a 1% agarose gel, followed by autoradiography. Position of the linearized pSK plasmid and the 0.84-kb ARS1 fragments are as indicated.
the DNA synthesis (Fig. 6), as well as a remarkable increase in the products of DNA replication (Fig. 7C). Moreover, the stimulation of in vitro plasmid replication in the S phase nuclear extract of S. cerevisiae by ORC was strictly dependent on the presence of an ARS sequence (Fig. 7). Since Mcm467 helicase appears also essential in this in vitro system (Fig. 1), a possibility exists that recombinant ORC identifies AT-reach non-ARS sequences and directs the replication factors to form a pseudo pre-RC. It is also likely that the ARS1 sequence in pSKARS helps attract more ORC molecules to the plasmid as well as its binding to the plasmid. This may lead to the observed enhancement of DNA synthesis both at ARS and non-ARS sites in the plasmid, as the ORC stimulation is not observed with pSK plasmid lacking ARS1. In addition, another possible explanation could be that the plasmid pSKARS is only 3.8 kb in length, as a result, ORC could activate a number of AT-rich regions near the ARS sequence and this activation is not confined to the ARS region only as shown by Vashee et al. (76). It is also apparent from the results in Fig. 7C that the stimulatory effect of exogenous ORC is far more significant than that observed in the assay measuring only the trichloroacetic acid-precipitable radioactivity. This is primarily due to the fact that the background radioactivity that is trichloroacetic acid-precipitable does not interfere in the agarose gel electrophoretic analysis. It is also important to note that the labeling of the DNA fragments in the absence of exogenous ORC is similar to that observed with exogenous ORC, even though the extent of synthesis with only endogenous ORC is significantly lower.
In conclusion, the major findings in the present studies are as follows: (i) Replication of supercoiled plasmid in S phase nuclear extract of S. cerevisiae depends on pol ␣/primase complex and pol ␦, as well as Mcm467 helicase, but does not appear to require pol ⑀. Perhaps long DNA templates are required to dissect the processive role of pol ⑀ in the in vitro system, which is consistent with the findings of Fukui et al. (53). (ii) Yeast Topo I appeared to be essential for supporting in vitro plasmid replication. (iii) Recombinant ORC significantly stimulates DNA synthesis in the in vitro system, and this stimulation required a supercoiled plasmid with a yeast ARS.