DNA Polymerase II (ε) of Saccharomyces cerevisiaeDissociates from the DNA Template by Sensing Single-stranded DNA*

Two forms of DNA polymerase II (ε) ofSaccharomyces cerevisiae, Pol II* and Pol II, were purified to near homogeneity from yeast cells. Pol II* is a four-subunit complex containing a 256-kDa catalytic polypeptide, whereas Pol II consists solely of a 145-kDa polypeptide derived from the N-terminal half of the 256-kDa polypeptide of Pol II*. We show that Pol II* and Pol II are indistinguishable with respect to the processivity and rate of DNA-chain elongation. The equilibrium dissociation constants of the complexes of Pol II* and Pol II with the DNA template showed that the stability of these complexes is almost the same. However, when the rates of dissociation of the Pol II* and Pol II from the DNA template were measured using single-stranded DNA as a trap for the dissociated polymerase, Pol II* dissociated 75-fold faster than Pol II. Furthermore, the rate of dissociation of Pol II* from the DNA template became faster as the concentration of the single-stranded DNA was increased. These results indicate that the rapid dissociation of Pol II* from the DNA template is actively promoted by single-stranded DNA. The dissociation of Pol II from the DNA template was also shown to be promoted by single-stranded DNA, although at a much slower rate. These results suggest that the site for sensing single-stranded DNA resides within the 145-kDa N-terminal portion of the catalytic subunit and that the efficiency for sensing single-stranded DNA by this site is positively modulated by either the C-terminal half of the catalytic subunit and/or the other subunits.

In Saccharomyces cerevisiae, DNA polymerase II (⑀), as well as DNA polymerases I (␣) and III (␦), are required for DNA replication (1,2). Deletion of the POL2 gene, which codes for the catalytic subunit of DNA polymerase II, causes cell death with a terminal dumbbell morphology, the hallmark of a defect in DNA replication (3). Direct measurements of in vivo DNA synthesis in temperature-sensitive pol2 mutants revealed that chromosomal DNA replication ceases at the restricted temperature (4,5). The observation that a pol2 mutant defective in 3Ј35Ј exonuclease activity exhibits a mutator phenotype at a variety of genetic markers throughout the genome supports the idea that participation of the polymerase in chromosomal DNA replication is not restricted to certain sites on the genome (6). Recent work by Aparicio et al. (7) revealed that DNA polymerase II is recruited to the origin at the time of initiation of DNA replication and proceeds along the DNA with the replication fork. This finding, together with others, strongly supports the idea that DNA polymerase II is a component of the replication apparatus and that it is responsible for DNA synthesis on either the leading or lagging strand. However, whereas DNA polymerase I is responsible for laying down RNA-DNA primers, specific roles for DNA polymerases II and III at the fork have not been determined (8,9).
DNA repair is another cellular process for which the function of DNA polymerase II might be required. Among the DNA polymerase mutants of S. cerevisiae, pol2 mutants were exclusively deficient in the repair synthesis of base-damaged DNA (10). For nucleotide excision repair, DNA polymerase II and III are potentially responsible, as pol2 and pol3 double mutants showed accumulation of single-strand breaks in their chromosomes after UV irradiation (11).
Besides DNA replication and repair, DNA polymerase II has recently been assigned a new role in the S-phase checkpoint. Certain pol2 mutants are defective in their transcriptional response to DNA damage, specifically in S-phase, and are unable to prevent entry into mitosis when DNA replication is blocked by hydroxyurea (12,13). The Dpb11 protein, which has been suggested to be an additional component of the DNA polymerase II complex on the basis of genetic studies, is also implicated in both DNA replication and checkpoint control (14). A temperature-sensitive dpb11 mutant showed a defect in Sphase progression at the restricted temperature with accompanying loss of viability owing to abnormal nuclear segregation. Furthermore, an involvement of Rfc5 and Rfc2, which are components of replication factor C (RF-C) 1 , in the S-phase checkpoint was also suggested (15,16). 2 Therefore, the view has emerged that the replication apparatus acts as a sensor for aberrant DNA replication as it progresses along the template with the replication fork. DNA polymerase II may play a central role in this process.
A high degree of conservation in the primary structure of the catalytic subunit of DNA polymerase ⑀ between yeast and human, as well as a similarity in their biochemical properties, suggests that the role of DNA polymerase II in yeast may apply to eukaryotic cells in general (17,18). The finding that expression of human DNA polymerase ⑀ depends on cell proliferation is in support of its involvement in chromosomal DNA replication (19). However, experiments by Zlotkin et al. (20) designed to capture DNA polymerases in contact with the nascent DNA by the photolabeling method detected only DNA polymerases ␣ and ␦ on SV40 DNA, consistent with their requirement in in vitro SV40 DNA replication (21)(22)(23). DNA polymerase ⑀, in addition to polymerases ␣ and ␦, was photolabeled by nascent cellular DNA; however, unlike DNA polymerases ␣ and ␦, the signal responded poorly to mitogenic stimulation, which increases the proportion of replicative DNA synthesis relative to repair synthesis. From these results, Zlotkin et al. (20) proposed that the major replicative polymerases, not only in SV40 but also in nuclear DNA synthesis, are DNA polymerases ␣ and ␦. DNA polymerase ⑀ was proposed to be required for cellular DNA replication, such as that observed during post-replicational repair coupled to fork movement, and in checkpoint activities, where it would assure accurate and coordinated DNA synthesis during the cell cycle.
In any case, little is known about the mechanism of action of DNA polymerase II (⑀) in specific replicative reactions except for its ability to catalyze DNA synthesis. This is largely the result of a lack of biochemical characterization of the polymerase, especially of the intact polymerase associated with all of its subunits. In accordance with the complexity of the biological functions of DNA polymerase II in S. cerevisiae, the enzyme possesses a rather complex structure. In its most intact form, DNA polymerase II purified from S. cerevisiae consists of four subunits (24). These comprise a catalytic 256-kDa subunit and auxiliary subunits of 80 and 34 kDa, which are encoded by the POL2, DPB2, and DPB3 genes, respectively (3,25,26). The gene for the smallest subunit of 29 kDa has been identified recently. 3 The large catalytic subunit is composed of two domains; the N-terminal half encodes the catalytic domain, which is conserved among aphidicolin-sensitive DNA polymerases, and the C-terminal half is a region unique to DNA polymerase II (2,3,27). Although the N-terminal domain is sufficient for polymerase and 3Ј35Ј exonuclease activities in vitro, mutational analyses suggest that the C-terminal domain is also involved in DNA replication (5,12). In addition, the C-terminal domain is involved in the functioning of the S-phase checkpoint and the response mechanism to DNA damage, as mutations in this domain selectively cause defects in these processes (12). In as much as the C-terminal domain is known to be required for holding the auxiliary subunits together as a complex (3), the defects caused by mutations may be either the direct effect of impaired functioning of the C-terminal domain or the result of loss of auxiliary subunits from the complex. The functions of the auxiliary subunits are mostly unknown.
Elucidation of the biochemical functions of the C-terminal domain of the Pol2 catalytic subunit and the auxiliary subunits of DNA polymerase II is critical for our understanding of the specific roles of the polymerase in vivo. As a first step, we have taken advantage of the fact that different forms of DNA polymerase II can be purified from cells of S. cerevisiae. Besides the four-subunit complex described above, a single polypeptide of 145 kDa, a proteolytic product of Pol2, has been purified to near homogeneity (24). Because the polypeptide retains both polymerase and 3Ј35Ј exonuclease activities, it probably contains most, if not all, of the N-terminal domain but lacks the Cterminal domain. Therefore, we compared the biochemical ac-tivities of the two forms of DNA polymerase II. The intact complex consisting of four subunits is called Pol II*, whereas the 145-kDa single polypeptide is called Pol II, following the proposal of the previous study (24). Here, we show that Pol II* is readily displaced from a DNA template by single-stranded DNA, whereas, in comparison, the sensitivity of Pol II to displacement by single-stranded DNA is greatly reduced. This difference in sensitivity to single-stranded DNA defines for the first time a biochemical activity specific for the C-terminal domain of Pol2 and/or the auxiliary subunits.
Nucleic Acids-The homopolymer poly(dA) 300 and oligo(dT) 10 were mixed at a weight ratio of either 20:1 or 5:1 in 20 mM Tris-HCl (pH 8.0) containing 20 mM KCl and 1 mM EDTA, heated at 65°C for 5 min, and then slowly cooled at room temperature. X174 ssDNA was singly primed with an 18-mer (map positions 11-28) synthetic oligonucleotide (30). The primer DNA and the X DNA were mixed at a molar ratio of 3:1 in 10 mM Tris-HCl (pH 8.0) containing 270 mM NaCl and 1 mM EDTA. The mixture was heated at 80°C for 10 min, transferred to 56°C for 15 min, and slowly cooled at room temperature. A linear form of X ssDNA was made as follows. The 20-mer DNA (map positions 4772-4791) (30) was labeled at its 5Ј-end by T4 polynucleotide kinase and hybridized to X ssDNA under the conditions described above for primed X DNA. The hybridized region contains the unique Alw44I site of X DNA. After diluting the reaction mixture 8-fold, the X DNA was digested completely with Alw44I and extracted with phenol-chloroform. The sample was further diluted 8-fold with TE (pH 7.5), heated at 65°C for 15 min, and chilled quickly on ice. The denatured oligonucleotides were removed by centrifugation of the sample in a Centricon-100 (Amicon). Concentrated sample was diluted again with TE (pH 7.5) to the initial volume, and the procedures were repeated again to remove denatured oligonucleotides. The removal of the oligonucleotides was complete as monitored by counting radioactivity in the sample. The 60-mer (d60) and 15-mer (d15) synthetic oligonucleotides were labeled at their 5Ј-ends by T4 polynucleotide kinase. The 15-mer DNA was hybridized to the 60-mer DNA (d60:d15) (14,000 cpm/pmol) by mixing 5Ј-labeled 15-mer and unlabeled 60-mer (or unlabeled 15-mer and 5Јlabeled 60-mer) at a molar ratio of 1:1 in 20 mM Tris-HCl (pH 8.0) containing 20 mM KCl and 1 mM EDTA. The mixture was heated at 90°C for 3 min, incubated at 65°C for 2 h, and cooled slowly at room temperature. Hook-structure DNA was a gift of Dr. H. Maki, Nara Institute of Science and Technology (31).
Purification of DNA Polymerase II (Pol II* and Pol II)-Pol II* and Pol II were purified from CB001 yeast cells (24). Cells (2 kg) were grown, and extracts (fraction I) were subjected to ammonium sulfate precipitation (fraction II) and SP-Sepharose Fast Flow (Pharmacia Biotech) column chromatography (fraction III), as described previously (24). Fraction III was dialyzed against buffer A (50 mM Tris-HCl, pH 7.5, 10% glycerol, 1 mM EDTA, 10 mM NaHSO 3 , 1 mM PMSF, 10 mM 2-mercaptoethanol) until the conductivity was equivalent to buffer A ϩ 0.1 M NaCl. After removal of insoluble material by centrifugation (12,000 ϫ g, 4°C, 15 min), the sample was loaded onto a POROS Q column (5 ϫ 10 cm) (PerSeptive Biosystems) equilibrated with buffer A ϩ 50 mM NaCl. The column was washed with 2 column volumes of the equilibration buffer, and polymerase activity was eluted by 11 column volumes of a linear gradient from 50 to 500 mM NaCl in buffer A at a flow rate of 3 ml/min. Four major peaks of the polymerase activity were obtained. Immunological detection with antiserum against Pol II* revealed that the fourth peak eluting at 300 mM NaCl corresponded to that of Pol II* (four-subunit complex of 256, 80, 34, and 29 kDa polypeptides), whereas the third peak eluting at 250 mM NaCl contained several different subassemblies of DNA polymerase II. The major constituents of the third peak seemed to be complexes of 145-and 34-kDa polypeptides and of 256-and 80-kDa polypeptides. Pol II* and Pol II were purified from the peaks eluting at 300 and 250 mM NaCl, respectively.
Pol II* Purification-Fractions containing the peak eluting at 300 mM NaCl were pooled, proteins were precipitated with ammonium sulfate (0.4 g/ml sample), and the precipitate was collected by centrifugation (39,000 ϫ g, 4°C, 30 min). The suspension of the pellet in buffer C (50 mM HEPES, pH 7.4, 10% glycerol, 1 mM EDTA, 10 mM NaHSO 3 , 1 mM PMSF, 10 mM 2-mercaptoethanol) was dialyzed against the same buffer until the conductivity reached that of buffer C ϩ 50 mM NaCl (fraction IV) and loaded onto a Mono S HR 10/10 column (Pharmacia Biotech) equilibrated with buffer C ϩ 50 mM NaCl. After washing the column with 3 column volumes of the equilibration buffer, the activity was eluted by a linear gradient of 20 column volumes of 50-500 mM NaCl in buffer C at a flow rate of 0.25 ml/min. The activity was recovered in a peak eluting at 280 mM NaCl with a shoulder at 290 mM NaCl. Four polypeptides of 256, 80, 34, and 29 kDa, which were crossreactive with the Pol II* antiserum, coincided with the activity in the peak at 280 mM NaCl, whereas the activity in the shoulder peak coincided with another set of four bands of 256, 80, 30, and 29 kDa. Because the 34-and 30-kDa polypeptides are known to be encoded by the same gene, DPB3 (26), the complex containing the 30-kDa polypeptide was probably the result of degradation of the 34-kDa polypeptide and was separated from the Pol II* complex at this step. The fractions in the peak at 280 mM NaCl were pooled (fraction V) and applied directly to a HiTrap heparin column (5 ml) (Pharmacia Biotech) equilibrated with buffer D (10 mM sodium phosphate, pH 7.0, 10% glycerol, 1 mM EDTA, 10 mM NaHSO 3 , 1 mM PMSF, 10 mM 2-mercaptoethanol) containing 300 mM NaCl. The column was washed with 3 column volumes of the equilibration buffer, and the activity was eluted by 20 column volumes of a linear gradient of 300 mM to 1 M NaCl in buffer D. The activity was eluted in a single peak at 640 mM NaCl and coincided with the presence of four bands of 256-, 80-, 34-, and 29-kDa polypeptides, as detected by SDS-PAGE. As the activity in this fraction was unstable at 4°C and sensitive to freezing and thawing, presumably as a result of the low protein concentration, eluates were collected into tubes containing BSA and Triton X-100R at final concentrations of 0.5 mg/ml and 0.01%, respectively. This treatment prevented enzyme inactivation. The overall recovery of the activity was 24%, assuming that the Pol II* activity recovered from the POROS Q column was 100%. The specific activity of the final sample of Pol II* was 4,400 units/mg protein (without BSA) (total 6,200 units), and the purity was approximately 70% (without BSA) as estimated by SDS-PAGE. The pool of the peak fractions of the HiTrap heparin column was dialyzed against 50 mM Tris-HCl, pH 7.5, 50% glycerol, 1 mM EDTA, 50 mM NaCl, 10 mM 2-mercaptoethanol, and 5 mM DTT, and aliquots were stored at Ϫ80°C (fraction VI).
Pol II Purification-Fractions that had eluted as a peak at 250 mM NaCl from the POROS Q column were pooled, and proteins were precipitated with ammonium sulfate (fraction IVЈ) and chromatographed on the Mono S HR 10/10 column as described above. The activity was recovered in three major peaks eluting at 200, 240, and 310 mM NaCl. The activity in the peak at 200 mM NaCl was sensitive to high salt (120 mM KCl), and the proteins in the peak showed no cross-reactivity with Pol II* antiserum. However, they were cross-reactive with Pol I antiserum. The activities in the peaks at 240 and 310 mM NaCl were resistant to 120 mM KCl, which is a characteristic of DNA polymerase II. Immunological detection with the Pol II* antiserum revealed that the peak at 240 mM NaCl contained two polypeptides of 145 and 34 kDa, whereas the peak at 310 mM NaCl contained polypeptides of 256 and 80 kDa. The activity in the former peak was further purified. The peak fractions were pooled and dialyzed against buffer D (fraction VЈ) and loaded onto a HiTrap heparin (5 ml) equilibrated with buffer D ϩ 100 mM NaCl. After washing the column with 3 column volumes of the same buffer, the activity was eluted with linear gradient of 10 column volumes of 100 mM to 1 M NaCl in buffer D. The activity recovered in a single peak at 730 mM NaCl coincided with the presence of only the 145-kDa polypeptide, as detected by immunoblot analysis using the Pol II* antiserum. The 34-kDa polypeptide appeared in slightly earlier fractions, suggesting that the association between the two polypeptides, if any, is not maintained during purification. The peak fractions were pooled, dialyzed against buffer B (10 mM potassium phosphate, pH 7.0, 10% glycerol, 10 mM NaHSO 3 , 1 mM PMSF, 10 mM 2-mercaptoethanol) (fraction VIЈ), and loaded onto a Macro-Prep ceramic hydroxyapatite, type I (CHT-I) column (1 ml) (Bio-Rad) equilibrated with buffer B. After washing the column with 3 column volumes of the same buffer, the activity was eluted with a linear gradient of 15 column volumes of 10 -500 mM potassium phosphate at a flow rate of 0.1 ml/min and recovered in one major peak at 200 mM potassium phosphate. The activity coincided with the presence of the 145-kDa polypeptide, namely Pol II. There was also a very small peak at 280 mM NaCl, which co-eluted with the 145-and 34-kDa polypeptides. This peak probably represents the complex of the two polypeptides that contaminated the previous fraction. The specific activity of the final sample of Pol II was 58,000 units/mg (total 1600 units), and the purity of the sample was estimated to be approximately 90% as judged by SDS-PAGE and silver staining of the gel. To prevent enzyme inactivation, BSA, EDTA, DTT, and Triton X-100R were added to the final sample at final concentrations of 0.3 mg/ml, 1 mM, 5 mM, and 0.01%, respectively, and aliquots were stored at Ϫ80°C (fraction VIIЈ).
Amino Acid Sequence Analysis of Pol II-The amino acid sequences of peptides generated by digesting Pol II with lysylendopeptidase were analyzed using a PSQ-10 protein sequencer (Shimadzu). Among the sequences determined, the ones most proximal to the N terminus and the C terminus of the catalytic subunit (total 2222 amino acids) were NH 2 -LSFVNSNQLFEARK (amino acids 191-204) and NH 2 -RNQLT-NEEDPLVLPSEIPSMDEDYV (amino acids 1246 -1270), respectively (3). As the molecular mass of the polypeptide (amino acids 191-1270) was calculated to be 126 kDa, Pol II contains most, if not all, of the N-terminal half of the catalytic subunit but lacks most of the C-terminal region.
Assay for Polymerase Activity during Purification-The assay measures incorporation of [␣-32 P]dTTP into trichloroacetic acid-insoluble material using poly(dA) 300 oligo(dT) 10 (20:1) as the DNA template. The assay conditions were as described previously (24), except the reaction was carried out in a volume of 20 l at 30°C for 30 min. One unit of polymerase activity corresponds to the incorporation of 1 nmol of dTTP per h.
Assay with Singly Primed X ssDNA-The reaction mixture (320 l) contained 35 mM Bis Tris-HCl, pH 6.3, 8 mM MgCl 2 , 10% glycerol, 100 g/ml BSA, 2 mM DTT, 100 M each of dATP, dCTP, and dGTP, 50 M [␣-32 P]dTTP (1000 -2000 cpm/pmol), 2.7 g of singly primed X174 ssDNA, 58 g of RF-A, and the polymerase to be assayed. After preincubation of the components above, except for the [␣-32 P]dTTP, at 30°C for 5 min, the reaction was initiated by the addition of [␣-32 P]dTTP. After incubation at 30°C for the times specified, 40-l samples were withdrawn, quenched by adding 40 l of 50 mM EDTA, and divided into two halves. One was used for counting the radioactive dTTP incorporated into acid-insoluble material as described previously (24). The other was used for an analysis of the DNA products of the reaction. The DNA was precipitated by ethanol precipitation, the resulting pellet was dissolved in 10 l of alkaline loading buffer (30 mM NaOH, 30 mM EDTA, 10% w/v sucrose, 0.04% bromcresol green, 0.2% SDS), and the suspension was separated by electrophoresis through a 1% alkaline agarose gel (15 ϫ 13.5 ϫ 0.3-cm gel, 40 V, 11 h). An EcoT14I digest of DNA was labeled at its 5Ј ends by T4 polynucleotide kinase and used as a size standard. After neutralizing the gel in 7% trichloroacetic acid, the gel was dried and visualized using a Bio-Imaging Analyzer BAS-1500 (Fuji film).
Measurement of the Rate of Dissociation of the Pol:DNA Complex-DNA polymerase (Pol II* or Pol II) and poly(dA) 300 oligo(dT) 10 (5:1) were mixed as indicated in the figure legends in a reaction mixture containing 35 mM Bis Tris-HCl, pH 6.3, 5 mM MgCl 2 , 10% glycerol, 100 g/ml BSA, and 2 mM DTT and pre-incubated at 30°C for 2 min to allow the Pol:DNA complex to reach equilibrium. X ssDNA circle was then added to the reaction at the concentrations specified in the figure legends. At different time points after the chase with the X DNA, samples were withdrawn, and [␣-32 P]dTTP (20 M) was added to initiate DNA synthesis. The reaction was terminated after 2 min at 30°C and the radioactivity incorporated into acid-insoluble material was counted. When the rate of dissociation of Pol:DNA complex was measured by a gel mobility shift assay, 5Ј-labeled d60:d15 was used for complex formation. The Pol:DNA complex was formed and chased with the X DNA as described above. After the times indicated, samples were removed, fixed with glutaraldehyde (0.8%) for 5 min at 30°C, and subjected to non-denaturing gel electrophoresis as described below.
Gel Mobility Shift Assays-Samples (6 l) containing the Pol:DNA complex cross-linked with glutaraldehyde as described above were mixed with 1 l of loading buffer (20% w/v sucrose, 1 mg/ml bromophenol blue) and subjected to non-denaturing gel electrophoresis at 4°C in a 4% polyacrylamide gel (15 ϫ 15 ϫ 0.1 cm) for 3.5 h at 100 V in Tris-glycine buffer (50 mM Tris, pH 8.5, 0.38 M glycine, and 2 mM EDTA). The gel was fixed with 12% (v/v) methanol and 10% (v/v) acetic acid, washed with distilled water, dried at 70°C, and exposed for quantification using a Bio-Imaging Analyzer BAS-1500 (Fuji film).
Other Methods-Protein concentration was determined by Bio-Rad protein assay system based on the method of Bradford using bovine gamma globulin as a standard (32). DNA concentration was determined spectrophotometrically.

RESULTS
Both Pol II* and Pol II Catalyze Highly Processive DNA Synthesis at the Same Rate-First, we compared the polymerase activities of Pol II* and Pol II by measuring their processivity and rate of DNA-chain elongation. In these assays, Pol II* or Pol II was pre-incubated with the singly primed circular X174 ssDNA coated with RF-A in the presence of dATP, dGTP, and dCTP, and then DNA synthesis was initiated by adding 32 P-labeled dTTP. The products of elongation were analyzed by electrophoresis on an alkaline agarose gel. During the first 3 min, a burst in DNA synthesis was observed in both reactions with Pol II* and Pol II (Fig. 1A). The product DNA at these early time points ran as rather distinct bands, and the size increased to reach the full length within the time period of the initial burst in DNA synthesis (Fig. 1B). Because the levels of incorporation during the burst DNA synthesis were less than 3% of the total incorporation obtained when all the input tem-plate primers were utilized and fully elongated, these results indicated that both Pol II* and Pol II started elongation uniformly and that the polymerases have a capacity to elongate DNA all the way around the viral DNA circle without dissociating from it. However, it was also evident that a considerable proportion of the polymerase molecules paused at, or dissociated from, the DNA template before completing the elongation of the viral DNA. The broad distribution of the elongation products on the gel suggested that polymerases paused at (or dissociated from) many positions on the viral DNA, although some specific stop sites were evident. The rates of DNA elongation by Pol II* and Pol II at 30°C, calculated from the maximum size of the elongation products at the first three time points, were 30 and 36 nucleotides/s, respectively. After 3 min in the time course, the overall rate of DNA synthesis decreased owing to the contribution of recycling events by the polymerases. The recycling of Pol II seemed to occur at a slightly higher rate than that of Pol II*. DNA synthesis with low processivity, carried out by DNA polymerase III (␦) (Pol III) of S. cerevisiae in the absence of PCNA and RF-C, showed no biphasic kinetics. Accordingly, an accumulation of short elongation products was observed, even after 30 min when the level of incorporation was equivalent to that of Pol II* or Pol II. From these results, we concluded that the intrinsic capacity of Pol II* for highly processive DNA synthesis is fully retained by the N-terminal half of the catalytic subunit, namely Pol II. Also, the rates of DNA elongation by Pol II* and Pol II were almost indistinguishable.
Pol II* Dissociates from the DNA Template Much Faster than Pol II-In search for a specific function for the auxiliary subunits of Pol II* and/or the C-terminal half of the catalytic subunit, we compared the ability of Pol II* and Pol II to interact with the DNA template and found an unexpectedly large difference in their rates of dissociation from the template. The assay system was based on the one reported by Maga and Hü bscher (29). DNA polymerase ⑀ has been shown to have an affinity for single-stranded DNA, and this can be used to trap the polymerase if it dissociates from the DNA template (29,33). In our assay, the polymerase (Pol II* or Pol II) was incubated first with the DNA template, poly(dA) 300 oligo(dT) 10 (5:1), to allow the Pol:DNA complex to reach equilibrium. The reaction was then chased with circular single-stranded X174 DNA, samples were removed at different time points, and the level of the remaining Pol:DNA replication complexes was quantified by measuring DNA synthesis.
To determine the concentration of X ssDNA required to trap the dissociated polymerases, we first examined the capacity of X DNA to inhibit the DNA synthesis carried out by the two polymerases (Fig. 2, A and B). Addition of the X DNA inhibited the actions of both Pol II* and Pol II, although a higher concentration of DNA was required to inhibit Pol II than Pol II*. For complete inhibition of Pol II* and Pol II, 23 and 45 nM (as DNA circles) of X DNA were required, respectively. Therefore, the presence of 45 nM X DNA in the assay should prevent both Pol II* and Pol II from re-associating with poly-(dA)oligo(dT) after decay of the Pol:DNA complexes, a conclusion that was validated in the next experiment.
The polymerases were preincubated with poly(dA)oligo(dT) to form a complex, followed by the concomitant addition of X DNA (45 nM) and [ 32 P]dTTP to start DNA synthesis. The time course of the reaction with Pol II* showed 6 pmol of burst DNA synthesis with no increase in the incorporation of dTTP thereafter (Fig. 2C). Therefore, this represents DNA synthesis by the polymerase in the preformed complexes with the DNA template, as recycling of the polymerase was prevented by X DNA. When the amount of the polymerase was doubled, the kinetics of the reaction remained the same except that a 2-fold increase in the burst size was observed (data not shown). Thus, the burst size in DNA synthesis reflects the initial amount of Pol II*:DNA complex. Similar kinetics were observed with Pol II (Fig. 2D), and the relative amounts of Pol II:DNA complex formed during the pre-incubation could be measured.
Based on the results shown in Fig. 2, we were able to measure the amounts of Pol:DNA complex remaining after a chase with X DNA, and in this way the rates of dissociation of the complexes could be monitored. We added X DNA at 45 nM to the pre-incubated mixture of the polymerase (Pol II* or Pol II) and poly(dA)oligo(dT) to start the chase. Samples were then taken at different time points and assayed for burst DNA synthesis. As shown in Fig. 3, very little dissociation of the complex between Pol II and poly(dA)oligo(dT) was observed during the 2-min chase at 30°C. In contrast, dissociation of Pol II* from its corresponding complex was quite rapid. We confirmed that no inactivation of Pol II* had taken place during the 2-min incubation at 30°C by carrying out an experiment in the absence of X DNA (data not shown). The half-life of the Pol II*:DNA complex was about 7 s, whereas that of the Pol II:DNA complex was estimated to be 8.8 min by increasing the time course of the dissociation assay (Fig. 7). Therefore, Pol II* dissociates from the DNA template 75-fold faster than Pol II under these conditions. A semi-logarithmic plot of the data obtained using Pol II* did not produce a straight line. It seems that approximately 10% of the complexes are substantially more stable than the rest. There may exist several different kinds of complexes owing to the different forms of Pol II*.
The results were confirmed in experiments using the gel mobility shift assay, which allowed us to detect directly the complexes between the two polymerases and the DNA templates. We developed an assay by following basically the method described by Maga and Hü bscher (29). A synthetic deoxyoligonucleotide of 60-mer hybridized with a 15-mer primer (d60:d15) was used as a DNA probe. Either Pol II* or Pol II was incubated with the 5Ј-32 P-labeled d60:d15 DNA to form Pol:DNA complex, and, at different times after the chase with X ssDNA, aliquots from the reaction were fixed and subjected to electrophoresis. The radioactivity in a shifted band at each time point was taken to be a measure of the amount of remaining complex.
We re-examined the optimal concentration of X DNA for dissociation in the gel mobility shift assay. Fig. 4A shows the inhibition of complex formation between d60:d15 and polymerases by X DNA. In the absence of X DNA, retardation of , and the X ssDNA circle at the final concentrations (as DNA circles) specified in the figure were prewarmed at 30°C for 2 min and mixed at time 0 to start the reaction (total 100 l). After incubation at 30°C for the times indicated, 20-l samples were withdrawn, quenched, and processed for acidprecipitable radioactivity. C and D, a 90-l reaction mixture containing 1.5 units of Pol II* (C) or Pol II (D) and poly(dA) 300 oligo(dT) 10 (5:1) (final concentration of 100 nM) was pre-incubated at 30°C for 2 min and added to a prewarmed 10-l reaction mixture containing [␣-32 P]dTTP (final concentration of 20 M) and X ssDNA circle (final concentration of 45 nM) to start the reaction. After incubation at 30°C for the times indicated, 20-l samples were withdrawn and processed as in A. 32 P-labeled d60:d15 was observed upon addition of either Pol II* or Pol II to the mobility shift assay. That the major shifted band in each experiment was the result of the complex between d60:d15 and either Pol II* or Pol II was verified from its co-appearance with the activity of the polymerases, when the peak fractions of Pol II* or Pol II in their final purification step were analyzed by both gel mobility shift and polymerase assays (data not shown). When X DNA was titrated into the reactions, the formation of the Pol II*:DNA complex was completely inhibited at a concentration of 0.9 nM (as DNA circles) X DNA, whereas a 50-fold higher concentration of X DNA (45 nM) was required to inhibit the formation of Pol II:DNA complex by 92%. As observed previously in Fig. 2, a higher concentration of single-stranded DNA was required for the inhibition of complex formation between Pol II and the DNA template than between Pol II* and the DNA template.
Using 45 nM X DNA as a trap, the rates of dissociation of polymerases from d60:d15 were measured (Fig. 4B). Whereas the half-life of the Pol II:DNA complex was calculated to be 1.1 min, the half-life of the Pol II*:DNA complex was so short that it could not be measured because the shifted band was already undetectable at the first time point in the dissociation assay. Although the rates of complex dissociation observed in the gel mobility shift assay were significantly faster than those observed in the previous assay (Fig. 3), the Pol II*:DNA complex still dissociated faster than the Pol II:DNA complex. The gel mobility assays enabled us to discriminate complex dissociation and polymerase inactivation, which could have occurred if the Pol II*:DNA complex was somehow inactivated by the addition of the X DNA in the burst DNA synthesis assay.
Dissociation of Pol II* from the DNA Template Is Actively Promoted by Single-stranded X174 DNA-DNA polymerases with high processivity tend to have slow dissociation rates from the DNA template. Because both Pol II* and Pol II are highly processive enzymes and are indistinguishable from each other in this respect (Fig. 1), the rapid dissociation of the Pol II* from the DNA template suggested that the observed dissociation rate might not be a true reflection of the intrinsic stability of the Pol II*:DNA complex. The equilibrium dissociation constant (K D ) of a given protein-DNA interaction is a measure of the affinity of the protein for the DNA, and it represents the same quantity k off /k on . If we assume that the rates of association (k on ) do not differ so much from one complex to another because the "on" rate is usually diffusion limited for DNA binding proteins, the K D value of a given complex gives us a good estimate of its rate of dissociation that is intrinsic to the complex. Therefore, we measured the equilibrium dissociation constants (K D ) of Pol II*:DNA and Pol II:DNA complexes.
We titrated various amounts of 32 P-labeled d60:d15 DNA into reactions containing a fixed amount of DNA polymerase (Pol II* or Pol II), and after cross-linking the Pol:DNA complexes formed at each equilibrium state with glutaraldehyde, the complexes were separated from free oligonucleotide by nondenaturing PAGE. The amount of Pol:DNA complex was measured by counting the radioactivity in the shifted band. The formation of the complex between polymerase (Pol II* or Pol II) and the singly primed oligonucleotide showed a hyperbolic dependence on the concentration of the DNA template. The half maximal point in the curve represents the K D of each complex The mixtures were incubated at 30°C for 5 min and then added with 1 l of 5% glutaraldehyde. After incubation of the mixtures for another 5 min at 30°C, 1 l of loading buffer was added to the samples, and they were subjected to nondenaturing polyacrylamide gel electrophoresis. B, a 20-l reaction mixture containing 0.5 unit of Pol II* (or 2 units of Pol II) and 32 P-labeled d60:d15 (final concentration of 96 nM) was incubated at 30°C for 5 min, and then 5-l of reaction mixture containing X ssDNA was added (final concentration of 45 nM as DNA circles). At indicated times after the chase with X DNA, 5-l samples were withdrawn, added with glutaraldehyde, and subjected to electrophoresis as in A. (Fig. 5). The K D values of the Pol II*:DNA and Pol II:DNA complexes were estimated to be 16 and 10 nM, respectively, indicating that complexes of Pol II* and Pol II with the DNA template are almost of equal stability.
To account for the 75-fold difference in the rate of dissociation of Pol II*:DNA and Pol II:DNA complexes, two possibilities were considered. First, the association rate of the Pol II*:DNA complex is rapid enough to compensate for the its rapid dissociation, so that the K D value of this complex does not differ so much from that of the Pol II:DNA complex. Alternatively, the association rate for the Pol II:DNA complex may be slow enough to compensate for its slow dissociation. Second, the dissociation rate of Pol II*:DNA complex obtained in the previous section may not be the rate intrinsic to the complex. The single-stranded DNA used as a trap may carry out a rapid, second-order displacement of Pol II* from its complex with the DNA template.
To test the second possibility, we examined whether the dissociation of the complex shown in Figs. 3 and 4 was a second-order process. Good evidence for a first-order process would be that the rate of dissociation is not affected by varying the concentration of the X DNA as long as the DNA is present at high enough concentration to prevent recycling of the polymerase. If it were the second-order process, the rate would be expected to vary depending on the concentration of the X DNA. To perform a titration of the X DNA in the dissociation reaction of the Pol II*:DNA complex, we took advantage of the fact that the concentration of the X DNA can be lowered to 0.9 nM while still serving as a trap in the gel mobility shift assay (Fig. 4A). The experiment was carried out essentially as shown in Fig. 4B, except the concentration of X DNA was varied from 0.9 to 4.5 nM. As shown in Fig. 6, the decrease in the amounts of the shifted band during the time course became more prominent as higher concentration of the X DNA was added, indicating that the rate of dissociation of Pol II*:DNA complex increases in a manner dependent on the concentration of the X DNA. The half-lives of the complex were estimated to be 26, 10, and 6 s at 0.9, 2.3, and 4.5 nM X DNA, respectively. At 45 nM X DNA, the rate was faster than could be measured (Fig.  4B). The cross-linking efficiency of the complex with glutaraldehyde was not affected by increasing the concentration of X  Fig. 4. A, a 5-l reaction mixture contained 0.1 unit of Pol II* and 32 P-labeled d60:d15 at the concentrations (as DNA molecules) specified in the figure. The reaction mixtures were incubated at 30°C for 5 min, added with 1 l of 5% glutaraldehyde, and subjected to nondenaturing polyacrylamide gel electrophoresis. Quantification of the radioactivity in the bands corresponding to the complex and unbound DNA was performed as described under "Experimental Procedures." B, the experiment was done as in A, except the reaction mixtures contained 0.4 unit of Pol II instead of Pol II*. DNA (data not shown). Therefore, we conclude that the observed dissociation of the Pol II*:DNA complex is not a firstorder process. The process is rather a second-order displacement of Pol II* from the DNA template promoted by X DNA. Thus, Pol II* has the capacity to dissociate from the DNA template by sensing some element in the structure of X DNA, most probably a structure specific to single-stranded DNA.
Dissociation of Pol II Is Also Promoted by Single-stranded X174 DNA, Although at a Much Slower Rate-The rate of dissociation of the Pol II:DNA complex observed in Fig. 3 was very slow compared with that of Pol II*. This implies that the Pol II:DNA complex is not sensitive to the displacement promoted by X DNA. However, the fact that two different values for the half-life of the complex were obtained from two different experiments (Figs. 3 and 4) suggested the possibility that this complex is also subject to second-order displacement.
Dissociation of the Pol II:DNA complex was re-examined as shown in Fig. 7. The experiment was done essentially as the one shown in Fig. 3, except the time course of the dissociation assay was prolonged to 20 min, and two different concentrations of X DNA were used. No inactivation of the polymerase took place during the 20-min incubation at 30°C, as judged from the results of an experiment carried out in the absence of the X DNA. A semi-logarithmic plot of the data produced straight lines, suggesting that the complex formed between Pol II and the DNA template is a single entity. The half-life of the Pol II:DNA complex was estimated to be 8.8 min in the presence of 45 nM X DNA. It was shortened to 3.2 min when the X concentration was doubled to 90 nM. Therefore, dissociation of the Pol II:DNA complex is also promoted by X DNA. However, the Pol II displacement from this complex by X DNA was inefficient compared with that of Pol II*.
Deoxyoligonucleotides Also Promote the Dissociation of Both Pol II* and Pol II from the DNA Template-We then tested a 60-mer (d60) synthetic deoxyoligonucleotide for its ability to displace the polymerases from complexes with the DNA template. The assay system was the same as described in Fig. 3, except d60 was used as a trap for the dissociated polymerase. The concentration of d60 sufficient to trap dissociated poly-merases was determined to be 6.4 M as DNA molecules (380 M as nucleotides). Under these conditions, the half-lives of the Pol II*:DNA and Pol II:DNA complexes were 10 and 40 s, respectively (Fig. 8), indicating that oligomer DNA can also promote dissociation of both Pol II*:DNA and Pol II:DNA complexes. These results, taken together with those obtained using X DNA, strongly suggest that it is single-stranded DNA that is sensed by the DNA polymerase when it is displaced from the DNA template. Consistent with the previous data (Figs. 3 and  7), the semi-logarithmic presentation of the dissociation curve for the Pol II:DNA complex produced a straight line, whereas that of the Pol II*:DNA complex did not.
Interestingly, the Pol II:DNA complex dissociated 13 times faster than it did in the presence of 45 nM X DNA (240 M as nucleotides). In contrast, the half-life of the Pol II*:DNA complex was almost unchanged, regardless of the nature of the single-stranded DNA used in the reactions. Although the dissociation of Pol II:DNA complex is still slower than that of the Pol II*:DNA complex as shown in Fig. 8, the marked difference in the rates of dissociation between the two polymerase complexes observed in the presence of X DNA (Fig. 3) became only a 4-fold difference when d60 was used as single-stranded DNA.
These results suggest an involvement of DNA ends in the process of polymerase displacement. To test the possibility, we made a linear form of the single-stranded X DNA and used it instead of circular single-stranded X DNA in the dissociation experiment. However, the results were exactly the same as those observed with circular X DNA (Fig. 3, data not shown). Furthermore, we carried out a dissociation experiment using a synthetic oligonucleotide with a hook structure, whose 3Ј-end was designed to anneal to its corresponding sequence within the same oligomer. Again, the moderately rapid dissociation of Pol II from its DNA complex, equivalent to that observed with d60, was observed at an equivalent concentration of hook DNA (data not shown). These results seem to exclude the possibility that ends, especially free 3Ј-ends, are necessary for the dissociation of the Pol II:DNA complex. DISCUSSION Among the DNA polymerases in S. cerevisiae, DNA polymerase II (⑀) is unique in that it inherently exhibits highly processive DNA synthesis. In an assay using poly(dA)oligo(dT) as a template, it has been shown that not only Pol II* but also Pol II possess the capacity to polymerize at least 100 nucleotides per binding event to the primer template (24). In the present study, we confirmed the results in an assay using a singly primed X174 DNA (5.4 kilobases) coated with RF-A. Under these conditions, both Pol II* and Pol II were capable of elongating DNA all the way around the X DNA circle without dissociating from it. However, the polymerases also showed a tendency to pause or dissociate at many sites on the template, as pointed out previously (34). The observation that DNA polymerase II becomes responsive in the presence of salt to the processivity factors, PCNA and RF-C, seems to have resulted in a neglect of the importance of the high intrinsic processivity of the polymerase (18,35). Nevertheless, the structural basis for this highly processive DNA synthesis, which must reside within the N-terminal portion of the catalytic subunit, as well as its biological importance, certainly warrants further investigation. Our estimation for the rate of DNA elongation was 30 to 40 nucleotides/s at 30°C by either Pol II* or Pol II. This value is comparable with that of Burgers (34), which was estimated in the presence of PCNA and RF-C (at least 50 nucleotides/s at 30°C). The rate may actually be somewhat stimulated by PCNA and RF-C, as in the case of mammalian Pol ⑀ and PCNA (29). It was impossible to distinguish Pol II* and Pol II on the basis of their polymerase activities both in terms of processivity and rates of DNA replication.
We obtained a clue to the functions of the C-terminal half of the catalytic subunit and/or the auxiliary subunits by characterizing the DNA complexes of Pol II* or Pol II. Although the intrinsic stability of the two complexes, as indicated by their equilibrium dissociation constants, was almost the same, the rates of complex dissociation measured in the presence of single-stranded DNA, which served as a trap for free DNA polymerase, were 75-fold faster for the Pol II*:DNA complex than for the Pol II:DNA complex. The rapid dissociation of Pol II*:DNA complex was shown to be a second-order process whose rate depended on the concentration of single-stranded DNA. Thus, Pol II* bound to a DNA template is displaced through its ability to sense single-stranded DNA. We further showed that the rate of dissociation of Pol II:DNA complex increased at higher concentrations of single-stranded DNA, although the rates were still much slower than those of the Pol II*:DNA complex. Therefore, Pol II can also sense single-stranded DNA but, unlike Pol II*, is displaced from the DNA template with low efficiency. These results suggest that there is at least one site in the N-terminal half of the catalytic subunit, besides the one responsible for interaction with the DNA template, that recognizes single-stranded DNA and promotes dissociation of the Pol:DNA complex. An additional site(s) that stimulates the single-stranded DNA-directed displacement of the polymerase seems to be present in the C-terminal half of the catalytic subunit and/or in the auxiliary subunits of Pol II*. One of the candidates for this site could be the active center for 3Ј35Ј exonuclease activity in the N-terminal half of the catalytic subunit. The polymerase displacement seen in the present study may occur as a result of conformational changes in the primary DNA binding site of the polymerase provoked by an interaction between single-stranded DNA and the exonuclease domain. However, considering the fact that the active site for the exonuclease has a greater preference for the free 3Ј-end of the DNA, this possibility seems unlikely as the displacement of Pol II was promoted with equal efficiency by circular and linear X DNA or by oligomer DNA and hook DNA whose 3Ј-ends are sequestered.
The 75-fold difference in the rates of dissociation between Pol II*:DNA and Pol II:DNA complexes was reduced to 4-fold when 60-mer single-stranded DNA was used instead of X DNA at equivalent amounts in terms of nucleotide concentration. The 60-mer DNA selectively increased the dissociation rate of the Pol II:DNA complex to a level that was comparable with that of the Pol II*:DNA complex. The observation suggests that the small size and high molar concentration of the 60-mer DNA made it accessible to the putative sensor site for singlestranded DNA on Pol II. The site(s) on the C-terminal half of the catalytic subunit and/or on the auxiliary subunits may modulate the affinity for single-stranded DNA of the site that resides in the N-terminal half of the catalytic subunit. Identification of the sites responsible for polymerase displacement and clarification of the mechanism awaits further investigation. In addition, it will be interesting to ascertain if Pol:DNA complexes show active displacement when they are in the elongation or the idling states of DNA synthesis. In as much as the rapid displacement of Pol II* promoted by the single-stranded DNA requires the C-terminal domain of the catalytic subunit and/or other subunits that are structurally unique to this polymerase, it is suggested that the displacement process is a reaction specific to DNA polymerase II (⑀). However, our trial to set up the dissociation assays with other essential DNA polymerases of S. cerevisiae, Pol I (␣) and Pol III (␦), failed because of a difficulty in detecting the DNA complexes of these polymerases owing to their inherent instability. It would be necessary and of interest to compare Pol II* and Pol III in this displacement process in the presence of PCNA and RF-C.
A major restriction in characterizing the displacement process is that the single-stranded DNA in our assay plays two roles; one is that it serves as a trap for the dissociated polymerase, and the other is that it promotes displacement of the polymerase from the DNA template. A search for singlestranded DNA species that can act only as a trap but not as a promoter of polymerase displacement has so far been unsuccessful. Double-stranded DNA (blunt-ended linear doublestranded DNA of pUC119) does not act inhibitory to DNA synthesis by Pol II* and Pol II, so it cannot be used as a trap (data not shown). Because we cannot separate the two phenomena, it is difficult to test different DNA species, such as doublestranded DNA and RF-A-coated single-stranded DNA, for their ability to promote polymerase displacement. However, we observed no difference in amounts of Pol II*:DNA complex formed at given concentrations of the polymerase and the template DNA in the presence or absence of double-stranded DNA, showing that the double-stranded DNA does not affect the equilibrium state of Pol II*:DNA complex formation (data not shown). The observation suggests that the double-stranded DNA does not have an ability to promote polymerase displacement.
Maga and Hü bscher (29) have made kinetic analyses of the interaction of DNA polymerase ⑀ (Pol ⑀) from fetal calf thymus with DNA. In their study, they determined the equilibrium dissociation constant of the complex between the Pol ⑀ and a synthetic oligodeoxynucleotide (61-mer) hybridized to a 15-mer primer (d61:d15) to be 6 nM. They also observed that the Pol ⑀:DNA complex dissociated with a half-life of 7 min at 22°C using single-stranded M13 DNA as a trap. The slow dissociation may be due to a lack of the stimulatory function that we described for the second-order displacement of the polymerase by single-stranded DNA because their preparation of Pol ⑀ consisted only of the catalytic subunit of 145 kDa and an auxiliary subunit of 45 kDa. However, a direct comparison of their results with ours is made difficult because of a crucial discrepancy. They observed a higher affinity of Pol ⑀ for d61: d15 than that for d61. On the contrary, we saw no difference in the affinity of Pol II* or Pol II for d60:d15 and d60 DNA (data not shown). Therefore, whereas Pol ⑀ in the complex described by Maga and Hü bscher (29) is exclusively at the site of the primer, the interaction of Pol II* (or Pol II) with the DNA template is the result of nonspecific binding of the polymerase to the single-stranded DNA. This discrepancy may be attributed to a difference in the properties of Pol ⑀ from fetal calf thymus and Pol II* (Pol II) from S. cerevisiae. Titration of Pol II* or Pol II in the gel mobility shift assay showed that one molecule of d60:d15 can be bound by at least two molecules of either polymerase. When we analyzed the Pol:DNA complexes, the conditions were chosen as such that the polymerase and the DNA template formed a complex with 1:1 stoichiometry.
Because Pol II* and Pol II show no preference for binding to the primer terminus, the competitive inhibition (Fig. 4A) of complex formation between the polymerase and DNA templates by single-stranded DNA is as if diluting the specific activity of the radioactive DNA template. In the case of Pol II, complex formation, as detected by the radioactivity in the shifted band, was almost completely inhibited when X DNA was present at a 60-fold higher concentration in terms of nucleotides over that of 32 P-labeled d60:d15, which is consistent with this notion. In contrast, an equimolar concentration of X DNA was sufficient to completely inhibit complex formation between Pol II* and the labeled d60:d15. The binding of Pol II* to the X DNA predominates over that to d60:d15, probably because active displacement of the polymerase is taking place in the reaction. The Pol II* initially bound on d60:d15 is displaced by the X DNA. Also, the Pol II* bound on the X DNA stays on the DNA perhaps because intramolecular displacement is favored. The preference for the intramolecular displacement of Pol II* is also suggested from the time course of DNA synthesis by Pol II* in the presence of X DNA, as shown in Fig. 2A. The dTTP incorporation on the template poly(dA)oligo(dT) reached a plateau after 2 min, irrespective of the concentration of the X DNA added to the reaction. These results suggest that Pol II* is unable to re-associate with poly-(dA)oligo(dT) once the polymerase is trapped by the X DNA. This peculiar nature of X DNA on the inhibition of DNA synthesis by Pol II* may also be attributed to the intramolecular displacement of Pol II* on long single-stranded DNA.
There are several cellular processes in which the sensitivity of DNA polymerase II to single-stranded DNA might play an important role. First, it may be involved in the sensing mechanism for detecting a replication fork block in the initial stage of the S-phase checkpoint pathway (12). A block in the progression of the replication fork by DNA damage or nucleotide deprivation may generate single-stranded DNA at or near the fork (36). DNA polymerase II, which has been shown to be present at the replication fork (7), may sense single-stranded DNA as a signal of aberrant DNA replication. The resulting displacement of the polymerase could be a part of the biochemical process involved in signal transmission to the downstream components in the S-phase checkpoint mechanism (37). In the SOS response, the checkpoint mechanism of Escherichia coli, the single-stranded DNA created by blocking DNA replication acts as a signal for the sensor protein, RecA, which results in activation of the protein and the induction of downstream reactions (38). The pol2 mutants carrying non-sense mutations in the C-proximal region of the catalytic subunit are defective in their response to DNA damage or replication defects, such as induction of damage-inducible genes and prevention of entry into mitosis (12). This is consistent with the fact that Pol II, lacking the C-terminal half of the catalytic subunit, does not have an ability to efficiently sense single-stranded DNA. Second, the activity may efficiently translocate the polymerase to the single-stranded DNA region where DNA synthesis is required. Single-stranded gaps generated as intermediates in the nucle-otide excision repair may attract polymerases. Another possible site is the intermediate structure for DNA recombination. It has recently been reported that Holliday junctions accumulate when DNA elongation is blocked and that DNA polymerase II might be required for the formation of the recombination intermediates, probably through its capacity to stabilize such structures (39). A role for recombination in the restoration of collapsed or blocked replication forks during DNA replication has been proposed (40). Pol II may take a part in this process by efficiently translocating onto the cross-stranded structure and extending the length of the invading strand. One other possible site of action is the single-stranded regions generated on the lagging strand. If the polymerase is involved in lagging strand DNA synthesis, this activity may provide a means for efficient retargeting of the polymerase to the next primer after the completion of the synthesis of an Okazaki fragment. The activity of Pol II* described in this report should further our understanding of the biological role of DNA polymerase II in S. cerevisiae.