INTRODUCTION

Top Abstract Introduction Procedures Results Discussion References

Maintaining the integrity of chromosomal DNA in eukaryotes is of critical importance to the cell and requires a series of complicated enzymatic processes. This is reflected in the complexity and redundancy of the enzyme systems that participate in DNA metabolism, such as replication, repair, and recombination (1, 2). In addition, DNA metabolism is tightly linked to cellular control pathways that regulate the cell division cycle (3-9). One of the enzymes required to achieve DNA replication, repair, or recombination is the DNA helicase, which uses the energy of ATP to translocate in a specific direction along a DNA strand melting the duplex regions it encounters (10-14). The single-stranded DNA (ssDNA)¹ generated by the helicase is utilized by other enzymes that participate in the subsequent steps in DNA metabolic pathways. Recently, the DNA2 gene of Saccharomyces cerevisiae was implicated in chromosomal DNA replication (15, 16). DNA2 was originally identified by screening for cell division cycle mutants of S. cerevisiae and was shown to be essential for cell viability and to encode a 172-kDa protein with characteristic DNA helicase motifs (15). Analyses of a temperature-sensitive mutant of DNA2 demonstrated that the mutant cell arrested in the S phase of the cell cycle and was deficient in DNA synthesis but not RNA synthesis upon shift to the nonpermissive temperature (15). Immunoaffinity purified Dna2 fusion protein displayed a DNA-dependent ATPase activity as well as 3' to 5' DNA helicase activity specific for fork-structured substrates (15). In addition, a mutation in the ATP binding motif of DNA2 led to the inactivation of the ATPase and helicase activities and rendered the mutant cell inviable (16). These findings recommend the Dna2 protein as a candidate for a cellular replicative DNA helicase. Therefore, we decided to study Dna2 in the hopes of gaining understanding of initiation events of DNA replication in eukaryotes.

Despite the results reported previously, it was not demonstrated unambiguously that Dna2, by itself, constituted a DNA helicase. The enzyme preparation used in these studies was obtained by an immunoaffinity purification step and contained other protein(s). One such protein present in the Dna2 enzyme preparation was yeast Fen1 (yFen1, also called Rad27/Rth1), a structure-specific endonuclease. A specific association of yFen1 and Dna2 was demonstrated both genetically and biochemically (17). The Fen1 protein is a multifunctional enzyme found in various organisms; it has been shown to be involved in (i) Okazaki fragment maturation in lagging-strand DNA synthesis in human and simian virus 40 replication in vitro (also called 5'- to 3'-exonuclease, DNase IV, or MF1) (18-21), (ii) an alternative pathway for completion of DNA base-excision repair in human cells (22), and (iii) maintenance of dinucleotide repeat stability in yeasts (23). Given these functions for Fen1, the observation that Dna2 is associated biochemically with yFen1 suggests that Dna2 is not directly involved in advancing replication forks as suggested from previous studies (15, 16). Rather, it is likely to participate in one of the other key aspects of DNA metabolism. In addition to its true physiological function in vivo, two important issues related to the native structure of Dna2 have not been resolved yet: (i) are there any unidentified polypeptide(s) other than yFen1 that are associated with Dna2, and (ii) if there are any, do the associated protein(s) influence the biochemical activities of Dna2?

It is, therefore, necessary to define the biochemical activities of Dna2 protein alone in order to address the issues involved in the native structure of the Dna2 protein. For this purpose, we cloned S. cerevisiae DNA2 using a gap repair strategy and constructed a recombinant baculovirus in order to overexpress Dna2 protein in insect cells. The initial biochemical characterization of recombinant Dna2 purified to homogeneity gave rise to the unexpected finding that it possessed an intrinsic endonuclease activity in addition to the ssDNA-dependent ATPase activity reported previously. Interestingly, Dna2 appeared to melt the duplex DNA only partially, based on the observation that the 3'-end duplex was susceptible to the endonucleolytic activity of Dna2. This finding suggests that the Dna2 protein by itself does not possess marked DNA helicase activity. If it is to participate in DNA replication as a processive DNA helicase, it may require an additional protein(s) such as yFen1 or other unidentified protein(s) that enable Dna2 to attain a vigorous DNA unwinding activity. Alternatively, Dna2 may be involved in other essential aspects of DNA metabolism that require its unique endonuclease activity.

	EXPERIMENTAL PROCEDURES

Top Abstract Introduction Procedures Results Discussion References

Oligonucleotides, DNA, Nucleoside Triphosphates (NTPs), and Enzymes-- All oligonucleotides used for the construction of various DNA substrates (20, 24, 26, 27, 52, 73, and 98 nucleotides (nt) in length) and for PCR primers to clone the DNA2 gene were synthesized commercially (BioServe, MD). Oligonucleotides greater than 30-mer were gel-purified prior to use. The 52- and 73-mers were described elsewhere (24). The 24-mer (5'-CGG CAA ATG GAA CGC ACA TGC GCA-3') was complementary to the 5'-end region of the 73-mer, and the 26-mer (5'-GGA AAA CAT TAT TAA TGG CGT CGA GC-3') was complementary to the 3'-end region of the 73-mer (see Fig. 6D for the substrate constructed with these oligonucleotides). The 98-mer (5'-GAA TAC AAG CTT GGG CTG CAG GTC GAC TCT AGA GGA TCC CCG GGC GAG CTC GAA TTC CGG TCT CCC TAT AGT GAG TCG TAT TAA TTT CGA TAA GCC AG-3') contained 5'-end sequences complementary to 20-mer (5'-CTG CAG CCC AAG CTT GTA TT-3') and 3'-end to 27-mer (5'-CTG GCT TAT CGA AAT TAA TAC GAC TCA-3'). The oligonucleotides used as PCR primers were as follows: Dna2A (5'-CCG GAA TTC GGA ACT ACT TCA AAG CTA C-3') and Dna2B (5'-CGC GGA TCC TGA CGA TCT CTT CAA TTG-3') were used to amplify a 5'-flanking region of the DNA2 gene, whereas Dna2C (5'-CGC GGA TCC ATA ACA GGA AAG CTA TCA C-3') and Dna2D (5'-CTA GTC TAG AGT TGC TTG GTG CCA CGA-3') were used for a 3'-flanking region of the gene; Dna2E (5'-CGC GGA TCC ATG CCC GGA ACG CCA CAG AA-3') and Dna2F (5'-CCG GAA TTC AGT CGA TTA GGG ACT ATA G-3') were used to introduce BamHI site (underlined) at the initiation codon (bold type) of the DNA2 gene. X174 sscDNA was purchased from New England Biolabs. Nucleoside triphosphates were obtained from Boehringer Mannheim and [-³²P]dCTP (6000 Ci/mmol) and [-³²P]ATP (>5000 Ci/mmol) were purchased from Amersham Pharmacia Biotech. The following proteins were obtained commercially: restriction endonucleases, the Klenow fragment of Escherichia coli DNA polymerase I, and terminal deoxynucleotidyltransferase were from Promega, and polynucleotide kinase was from Bio-Rad.

Preparation of DNA Helicase and Nuclease Substrates-- DNA substrates used to examine the DNA unwinding and ATP-dependent nuclease activities of the Dna2 protein were prepared by hybridizing the 52-mer or 73-mer to X174 sscDNA as described (24). A partial duplex substrate used to quantify the endonuclease activity of Dna2 was prepared by annealing both 20- and 27-mers (40 pmol each) to the 98-mer (10 pmol) under the conditions as described previously (24). This substrate contained a single-stranded region (49-nt) flanked by duplex regions (20- and 27-base pairs (bp)). The 3'-ends of the two short oligonucleotides annealed to the 98-mer were labeled by incorporating [-³²P]dCTP, chased with excess unlabeled dCTP in the presence of the Klenow fragment. The substrate was gel-purified prior to use, and its specific activity was approximately 3,000 cpm/fmol. Substrates used to analyze the ssDNA-specific endonuclease activity of Dna2 were prepared by annealing either the 24-mer (40 pmol) or 26-mer (40 pmol) to 10 pmol of the 73-mer (named 3'-overhang and 5'-overhang partial duplex, respectively). In order to prepare 3'-overhang partial duplex substrates, the 73-mer (10 pmol) was first labeled at either its 3'-end by incorporating [-³²P]dideoxy-ATP with terminal deoxynucleotidyltransferase or at its 5'-end by incorporating [-³²P]ATP with polynucleotide kinase. The 73-mers thus labeled were then annealed to the 24-mer (40 pmol). The 5'-overhang partial duplex was labeled with Klenow by incorporating [-³²P]dCTP at the 3'-end of the 73-mer annealed to the 26-mer. In order to construct a substrate with partial duplexes at each end, both the 24- and 26-mers (40 pmol each) were annealed to the 73-mer (10 pmol) and labeled at the 3'-end of the 73-mer as described above for the 5'-overhang partial duplex. These substrates were also gel-purified prior to use. The specific activities of both substrates were comparable and ranged from 2,000 to 3,000 cpm/fmol.

ATPase, DNA Unwinding, and Nuclease Assays-- Standard assays for measuring DNA-dependent ATPase activity were carried out in a reaction mixture (20 µl) containing 25 mM Tris-HCl (pH 7.8), 2 mM MgCl₂, 2 mM DTT, 0.25 mg/ml bovine serum albumin (BSA), 250 µM cold ATP, 20 nM [-³²P]ATP (>5000 Ci/mmol), and 50 ng of M13 sscDNA. After incubation at 37 °C for 10 min, an aliquot (2 µl) was spotted onto a polyethyleneimine-cellulose plate (J. T. Baker, Inc.), which was then developed in 0.5 M LiCl, 1.0 M formic acid solution. The products were analyzed using a PhosphorImager (Molecular Dynamics).

Cloning of the DNA2 Gene and Construction of Recombinant Baculoviruses-- The 5'-flanking (397 bp) and 3'-flanking (365 bp) regions of the DNA2 gene were first cloned by PCR with the primers described above and inserted into the yeast plasmid pRS316 (25) to obtain pDNA2FL. This plasmid was introduced into S. cerevisiae strain YPH499 (MATa, ade2-101, ura3-52, lys2-801, trp1-63, his3-200, leu2-1), and the DNA2 gene was cloned with the use of a gap repair strategy as described (26). The retrieved plasmid (pDNA2) contained an intact DNA2 gene, which was confirmed by both restriction and DNA sequencing analyses (27) of the 5'- and 3'-regions. A BamHI site was introduced by using a PCR-amplified fragment that contained a BamHI site prior to the start codon (see above for the sequence of PCR primer Dna2E), and then the DNA2 gene was subcloned into the BamHI site of pBlueBacHis2A (Invitrogen) to generate pHX-DNA2. To prepare DNA2 with N-terminal 405 amino acid deletion, the NdeI-EcoRI fragment of pHX-DNA2 was subcloned into pBlueBacHis2C (Invitrogen) cleaved with BamHI and EcoRI. The ssDNA overhangs generated by NdeI and BamHI were made blunt using Klenow prior to the second cleavage with EcoRI. Recombinant baculoviruses were constructed as recommended by the manufacturer (Invitrogen). The recombinant baculoviruses encoded wild type DNA2 and mutant DNA2 with N-terminal 405-amino acid deletion produce recombinant wild type and mutant Dna2 proteins (HX-Dna2 and HX-Dna2405N, respectively) with additional 37 amino acid residues (MPRGSHHHHHHGMASMTGGQQMGRDLYDDDDKDRWGS). These included six histidines (bold type) and the Xpress epitope (underlined) fused to its N-terminal methionine to facilitate detection and purification of the recombinant Dna2 proteins.

Overexpression and Purification of the Recombinant Dna2 Proteins in Insect Cells-- The recombinant baculoviruses containing the DNA2 gene were infected into Hi-5 insect cells for 48 h at a multiplicity of infection of 10. The infected cells (1 × 10⁶ cells/ml, 2 liters) were harvested, resuspended in 160 ml of buffer T (25 mM Tris-HCl (pH 7.5), 1 mM EDTA, 10% glycerol, 1 mM DTT, 0.1 mM PMSF, 0.15 µg/ml leupeptin and antipain) containing 100 mM NaCl, and disrupted by sonication (7 cycles of a 30-s pulse and a 2-min cooling interval). The extracts (3.35 mg/ml) were cleared by centrifuging at 37,000 rpm for 1 h in a Beckman 45 Ti rotor, and the supernatant was applied directly to a heparin-Sepharose (Amersham Pharmacia Biotech) column (2.5 × 8 cm, 39.2 ml) equilibrated with buffer T plus 100 mM NaCl (buffer T100, hereafter, the number indicates the concentration of NaCl added to buffer T). The column was washed with 10 volumes of the same buffer containing neither DTT nor EDTA and eluted with buffer T600 (DTT, EDTA) plus 5 mM imidazole. The peak protein (1.82 mg/ml, 45 ml) was pooled and loaded onto a Ni²⁺-NTA agarose (Quiagen) column (1.5 × 3.5 cm, 6.2 ml) equilibrated with buffer T600 (DTT, EDTA) plus 5 mM imidazole. After extensive washing with buffer T600 (DTT, EDTA) plus 20 mM imidazole, the column was eluted with a 60-ml linear gradient of 20-400 mM imidazole in buffer T600 (DTT, EDTA). Fractions containing the DNA-dependent ATPase activity were pooled and dialyzed for 3 h against buffer T100. The dialysate (0.25 mg/ml, 180 ml) was loaded onto an SP-Sepharose column (Amersham Pharmacia Biotech) (1.5 × 3 cm, 5.3 ml) equilibrated with buffer T100. After washing with 5 column volumes of buffer T100, the column was eluted with a 50-ml linear gradient of 100-600 mM NaCl in buffer T. The DNA-dependent ATPase activity, which peaked at 300 mM NaCl, was pooled (1.0 mg/ml, 20 ml) and concentrated 5-fold. Aliquots (250-µl) were applied to glycerol gradients (5 ml, 15-35% glycerol in buffer T) in the presence of 500 mM NaCl (high salt gradient) and subjected to centrifugation for 24 h at 45,000 rpm in a Beckman SW55 Ti rotor. Fractions (220 µl) were collected from the bottom of the gradients and assayed for the DNA-dependent ATPase activity and nuclease activities. The active fractions, which contained more than 90% of the total DNA-dependent ATPase activity, were pooled and stored at 80 °C. One microgram of the purified enzyme hydrolyzed 8.9 nmol of ATP/min at 37 °C, and this preparation was used to examine the ATPase and nuclease activities, unless otherwise stated. A 5'-³²P label at the substrate duplex end was not removed after prolonged incubation with this preparation, thus the pool did not contain phosphatase activity. HX-Dna2405N was purified employing the same procedure used for HX-Dna2 as above.

Preparation of Polyclonal HX-Dna2 Antibody-- A monoclonal antibody against the Xpress epitope was purchased from Invitrogen. Polyclonal antibody specific for Dna2 was generated as follows: HX-Dna2 (2 mg) from the high salt gradient was subjected to 8% SDS-PAGE, and proteins were visualized by Coomassie staining. Gel slices containing protein were pulverized, then emulsified in (MPL + TDM + CWS) adjuvant (Sigma), and injected into two rabbits (3.5 kg, New Zealand White) according to the manufacturer's instructions (250 µg, each injection). At 21 and 50 days after the primary injections, rabbits were injected a second time (a "boost") with equal amounts of HX-Dna2. Sera were collected 7 days after boosting. The IgG fraction was purified with the use of protein-A Sepharose (Amersham Pharmacia Biotech) as described (28).

	RESULTS

Top Abstract Introduction Procedures Results Discussion References

Purification of the Recombinant Dna2 Protein-- Crude extracts prepared from insect cells infected with the recombinant Dna2 baculovirus contained significant levels (approximately 1to 3% of total proteins) of full-length HX-Dna2 protein (Fig. 1, lane 2), whereas HX-Dna2 was not observed in control extracts prepared from uninfected cells (Fig. 1, lane 1). The expression of recombinant HX-Dna2 was confirmed by Western blot analysis using two antibodies, anti-Dna2 polyclonal antibody (anti-Dna2) (Fig. 1, lane 5), which was raised against full-length HX-Dna2, and anti-Xpress monoclonal antibody (anti-Xpress) (Fig. 1, lane 9), which only detects protein with an intact N terminus. Uninfected cell extracts did not contain any material that cross-reacted with either of the two antibodies (Fig. 1, lanes 4 and 8), whereas purified protein was detected by both antibodies (Fig. 1, lanes 6 and 7, anti-Dna2; lane 10, anti-Xpress). A collection of polypeptides was present in the crude extracts (Fig. 1, lanes 5 and 9), as well as in the purified enzyme preparation obtained from the high salt gradient (Fig. 1, lane 6). These polypeptides apparently arose by proteolysis, as they cross-reacted with anti-DNA2 or anti-Xpress. Our control antibody (polyclonal antibody directed against bovine papillomavirus (E1) did not cross-react with HX-Dna2 at all (data not shown).

View larger version (35K): [in this window] [in a new window]   Fig. 1.   Western blot analysis of the recombinant HX-Dna2 protein. Crude extracts prepared from insect cells infected with a recombinant baculovirus overproducing HX-Dna2 and purified HX-Dna2 were subjected to 8% SDS-PAGE (37), and the gel was Coomassie-stained (indicated as Coomassie) and analyzed in Western blot analysis. The antibodies, indicated at the top of the figure, are polyclonal antibody specific for HX-Dna2 (-Dna2) and monoclonal antibody specific for Xpress epitope (-Xpress). Lanes 1, 4 and 8, crude extracts (20 µg) from uninfected insect cells; lanes 2, 5 and 9, crude extracts (20 µg) from insect cells infected with recombinant HX-Dna2 baculovirus; lane 3, purified HX-Dna2 (2 µg); lanes 6 and 10, purified HX-Dna2 (200 ng); lane 7, purified HX-Dna2 (50 ng). The numbers at the left of the figure indicate the molecular sizes (in kDa) of marker proteins (Bio-Rad) (indicated as M), which include myosin (200 kDa), -galactosidase (116 kDa), phosphorylase B (97.4 kDa), BSA (66 kDa), and ovalbumin (45 kDa). The position of HX-Dna2 is also indicated.

Purified Recombinant Dna2 Protein Lacks an Explicit DNA Unwinding Activity but Possesses Nucleolytic Activities-- An SDS-PAGE analysis of Mono Q column fractions yielded results identical to those above in that the 172-kDa polypeptide (Fig. 2A) copurified with the ssDNA-dependent ATPase activity (data not shown). As shown in Fig. 2, we investigated whether HX-Dna2 was able to displace the 73-base oligonucleotide annealed to X174 sscDNA (5'-tailed substrate with a non-complementary 21-nt tail at the 5'-end). When the DNA unwinding activity of HX-Dna2 was examined in the presence (Fig. 2B) and absence of ATP (Fig. 2C), we obtained a somewhat puzzling observation. The DNA species generated by HX-Dna2 were not dependent on the presence of ATP (Fig. 2, B and C) but required MgCl₂ in the reaction mixture (data not shown). In addition, the DNA products did not appear to correspond to the 73-base oligonucleotide, as they were not uniform in their migration; rather, migration retardation of these DNA was inversely proportional to the amount of HX-Dna2 present in each fraction (Fig. 2, A and B). Notable was the observation that the ³²P label at the 3'-end of the annealed oligonucleotide was released in an ATP-dependent fashion when HX-Dna2 concentration was enriched (compare Fig. 2, B and C). This observation suggests that the enzyme may possess an exonucleolytic activity that is able to act on dsDNA in the 3' to 5' direction and in an ATP-dependent manner. This activity, termed ATP-dependent nuclease activity, of HX-Dna2 was analyzed further and is discussed in detail below. Taken together, our observation suggests that the DNA species observed in Fig. 2 most likely result from a nuclease activity of HX-Dna2, indicating that Dna2 possesses at least one type of nuclease activity but not a strong DNA unwinding activity. However, we were able to detect a weak DNA unwinding activity using a 20-mer oligonucleotide-annealed X174 sscDNA as substrate under the condition where the endonuclease activity was inhibited (see below; Fig. 9)

View larger version (50K): [in this window] [in a new window]   Fig. 2.   Nuclease activities are observed and comigrate with the purified recombinant HX-Dna2 protein. The peak fraction of DNA-dependent ATPase activity from the first glycerol gradient (high salt gradient) carried out in the presence of 0.5 M NaCl was loaded onto the Mono Q column as described under "Experimental Procedures." A, SDS-PAGE (8%) of the Mono Q fractions (2 µl) was performed as described above, and the gel was stained with silver. The load (L), flow-through (F), and Mono Q fractions analyzed are indicated at the top of the figure. Protein molecular size markers (indicated as M) are as described in Fig. 1. B and C, autoradiograms of DNA products formed after incubation of each Mono Q fraction (2 µl) with 15 fmol of 5'-tailed X174 sscDNA (73-mer annealed, see "Experimental Procedures") substrate in 20-µl reactions in the presence of 2 mM ATP (B) and in the absence of ATP (C) as described. The reaction products were analyzed on a 10% polyacrylamide gel containing 0.1% SDS in 0.5× TBE (see "Experimental Procedures"). The load (L), flow-through (F), substrate alone (S), and Mono Q fractions analyzed are indicated at the top of the figure. The arrow indicates the migration position of the 73-mer oligonucleotide. The schematic structure of the substrate used is shown at the left of the figure. The asterisk indicates the 3'-32P-labeled end.

View larger version (17K): [in this window] [in a new window]   Fig. 3.   Glycerol gradient sedimentation in the absence of NaCl confirmed that DNA-dependent ATPase and nuclease activities comigrated with the HX-Dna2 protein. The peak protein fraction of the Mono Q column was analyzed in a 5-ml glycerol gradient (15-35%) in the absence of NaCl (low salt gradient) as described under "Experimental Procedures." A, proteins were analyzed across the glycerol gradient fractions (2 µl) as above. The same protein size markers (indicated as M) were used as in Fig. 1. The load (L) and glycerol gradient fractions analyzed are indicated at the top of the figure. B and C, autoradiograms of DNA unwinding assays carried out as in Fig. 2, B and C. Assays were carried out using the same fractions (2 µl) as for the SDS-PAGE proteingels in the absence (C) and presence of 2 mM ATP (B). The load (L), substrate alone (S), and gradient fractions analyzed are indicated at the top of the figure. The arrow indicates the position where the 73-mer oligonucleotide migrated. The schematic structure of the substrate used is shown at the left. The asterisk indicates the 3'-32P-labeled end. D, quantitation of the ssDNA-dependent ATPase activity. ATP hydrolysis was measured with 2 µl of each glycerol gradient fraction in a 20-µl reaction in the presence of 15 nmol ATP as described under "Experimental Procedures." E, quantitation of the endonuclease and ATP-dependent nuclease activities. ATP-dependent nuclease activities shown in the presence of ATP (B, ) and in the absence of ATP (C, ) were quantified. The endonuclease activity () was measured in the reaction mixtures (20 µl) that contained 0.2 µl of each glycerol gradient fraction with 15 fmol of a substrate (49-nt ssDNA flanked by duplex DNAs at both ends, see "Experimental Procedures").

The Nuclease Activities Are Intrinsically Associated with the Purified Recombinant Dna2-- To exclude the possibility that the ssDNA-specific endonuclease and ATP-dependent nuclease activities derived from insect cell proteins by fortuitous interactions with HX-Dna2 and to confirm that these activities were intrinsic to the HX-Dna2 protein, we performed immunodepletion experiments using polyclonal antibodies specific for HX-Dna2. HX-Dna2 protein obtained from the low salt gradient was first incubated with anti-Dna2, and then with protein A-agarose beads (Amersham Pharmacia Biotech), and the mixture was centrifuged to precipitate the immunocomplex. When we measured the endonuclease, ATP-dependent nuclease, and DNA-dependent ATPase activities remaining in the supernatant, we found that all three activities were depleted in proportion to amount of added antibody (Fig. 4), as would be expected if all the activities reside in one polypeptide. In contrast, the control antibody failed to deplete any of these activities (Fig. 4). Moreover, the endonuclease and ssDNA-dependent ATPase activities were inhibited in the presence of anti-Dna2 antibodies (data not shown). Together with results shown in Figs. 2 and 3, the immunodepletion experiments demonstrate unequivocally that the HX-Dna2 protein possesses all three enzymatic activities.

View larger version (18K): [in this window] [in a new window]   Fig. 4.   Antibodies specific to HX-Dna2 depleted all three activities simultaneously. DNA-dependent ATPase (, ), ATP-dependent nuclease (, ), and endonuclease (, ) activities were measured after the HX-Dna2 preparation was depleted with anti-Dna2 (closed symbols) or with anti-E1 (control, open symbols) polyclonal antibodies. The anti-E1 antibody was raised against the bovine papillomavirus E1 protein (38). The HX-Dna2 enzyme solution (100 µl, 4 µg/ml) was incubated with the indicated amounts of anti-Dna2 or anti-E1 on ice for 30 min with occasional rocking, and then protein A-coupled Sepharose beads (10 µl) were added. The mixtures were then incubated for an additional 15 min. The immunocomplexes were precipitated by centrifuging for 1 min in an Eppendorf centrifuge. Ten µl of each supernatant were used to detect each enzyme activity in the reaction mixtures (20-µl) as described under "Experimental Procedures."

The Reaction Products Formed by HX-Dna2 from Partial Duplex DNA Substrates Were dsDNA, but Not Displaced Oligonucleotides-- If HX-Dna2 is a ssDNA-specific endonuclease, the reaction products should consist of dsDNA derived from the partial duplex region where the oligonucleotide was annealed. This prediction prompted us to carry out an experiment to analyze systematically the products formed by HX-Dna2 using the partial duplex X174 sscDNA as the substrate. To do so, we used the restriction endonuclease HpaII, which is highly specific for dsDNA, because the partial duplex region of the X174 sscDNA substrates contains a single HpaII cleavage site (24). We first analyzed the reaction products generated from the blunt X174 partial duplex substrate (Fig. 5, lanes 1-6). The boiled control oligonucleotide was not susceptible to vigorous digestion with HpaII (Fig. 5, lanes 2 and 3), whereas the product formed by the addition of excess HX-Dna2 (Fig. 5, lane 4) was cleaved completely by the same enzyme under the identical condition (Fig. 5, lane 6), demonstrating that the DNA formed by Dna2 was indeed dsDNA. When the product DNA was heat-denatured, the labeled 52-mer oligonucleotide reappeared (Fig. 5 lane 5), suggesting the 52-mer oligonucleotide was contained in the product and was not affected by HX-Dna2. We obtained identical results from analyses of the DNA products obtained from the 5'-tailed X174 substrate (Fig. 5, lanes 7-12). When we analyzed DNA products obtained from the two partial duplex substrates above in the presence of ATP (4 mM), they were also sensitive to HpaII (data not shown). It was noteworthy that the oligonucleotide released by the boiling of 5'-tailed X174 substrate treated with HX-Dna2 was shortened and migrated in a manner similar to that of the 52-mer (Fig. 5, compare lanes 2 and 11), indicating the unannealed 5' tail was removed. Taken together, these results suggest that the DNA products generated by HX-Dna2 from partial duplex DNA were not ssDNA displaced by an unwinding activity of HX-Dna2 but were dsDNA products that resulted from degradation of the bulky ssDNA of template X174.

View larger version (23K): [in this window] [in a new window]   Fig. 5.   Characterization of DNA products formed by HX-Dna2 with partial duplex substrates. Schematic structures of two partial duplex substrates (blunt, 53-mer annealed X174 sscDNA; 5'-tailed, 73-mer annealed X174 sscDNA) used to characterize the DNA products generated by the HX-Dna2 protein are indicated at the top of the figure. The asterisk indicates a radioisotopic label at the 3'-ends of the annealed oligonucleotides. The substrates (15 fmol each) were first incubated at 37 °C for 10 min in the reaction mixtures (20 µl) that contained 80 ng of purified HX-Dna2 (glycerol gradient fraction 11, see Fig. 3) under DNA unwinding assay conditions, but without ATP, as described under "Experimental Procedures." The DNA products were then digested for 2 h with the HpaII and subsequently analyzed by electrophoresis through 10% polyacrylamide gel that contained 0.1% SDS (w/v). Where indicated (+), the products were boiled for 3 min (Heat) prior to electrophoresis, or they were incubated with 5 units of HpaII restriction enzyme (HpaII) for 2 h after incubation with HX-Dna2. The minus () indicates that the treatment or additions were omitted. The positions of the oligonucleotides displaced from the two substrates (blunt and 5'-tailed) by heat treatments are indicated as 52-mer and 73-mer, respectively.

Analysis of DNA Products Confirms That Dna2 Is a ssDNA-specific Endonuclease-- HX-Dna2 appeared to possess an exonuclease-like activity for two reasons: (i) any ssDNA present in the form of unannealed tail was efficiently removed (Fig. 5), and (ii) the 3'-end label at the duplex region was released by HX-Dna2, although removal required ATP (Figs. 2 and 3). As mentioned above, the bulky ssDNA of X174 could have been degraded by either an endonuclease activity alone or by combined endonuclease and exonuclease activities. In order to distinguish between these two possibilities, we attempted to define the HX-Dna2 nuclease activities by analyzing the reaction products in high resolution denaturing polyacrylamide gels using better defined substrates. We first analyzed reaction products generated by incubating a 3'-overhang partial duplex substrate with HX-Dna2. The single-stranded region was efficiently degraded (Fig. 6A), and the products were not exclusively mononucleotide but were oligonucleotides of varying sizes ranging from monomers to decamers. This suggests the Dna2 protein is an endonuclease. Interestingly, the oligonucleotide degradation was limited to the first 10 nucleotides from the 3'-end with the 3'-overhang substrate (labeled at 3'-end) (Fig. 6A). In order to investigate how extensively the ssDNA region is degraded, we tested in the nuclease reaction a 5'-labeled 3'-overhang partial duplex substrate. As shown in Fig. 6B, HX-Dna2 was capable of degrading most of the 3'-overhang ssDNA when high concentrations of enzyme were used (>20 ng) (Fig. 6B). However, the enzyme was not able to remove ssDNA region completely, leaving 3'-ssDNA tail of 4 to 5 nucleotides (Fig. 6B). We also tested a 5'-overhang partial duplex substrate. The HX-Dna2 protein was capable of degrading the ssDNA tail up to the duplex junction, as the 26-nt products appeared with more than 10 ng of enzyme (Fig. 6C). This is in contrast to the finding that some of the 3' ssDNA immediately adjacent to duplex DNA was resistant to HX-Dna2 under the same reaction condition. Finally, we confirmed that the enzyme was able to cleave the ssDNA flanked by duplex regions at both ends (Fig. 6D). This reaction was rather inefficient because less than 10% of the substrate was cleaved at the highest amount (160 ng) of enzyme (Fig. 6D), whereas most of the ssDNA tails in the substrate were cleaved at the same enzyme concentration (Fig. 6, A-C). This result suggests that the enzyme may require stretches of ssDNA longer than 24 nucleotides for efficient cleavage, as a substrate similar in structure, but with 49-nt stretches of internal ssDNA was cleaved efficiently (Fig. 3E). This is also consistent with the previous observation that the ssDNA that is distant from duplex DNA was efficiently cleaved even in the presence of very low enzyme levels (<1 ng) (Fig. 6A), whereas much more enzyme (>20 ng) was required to attack ssDNA that was annealed close to the duplex junction (Fig. 6B). All of the results thus far presented support the hypothesis that HX-Dna2 possesses only an endonucleolytic and not an exonucleolytic activity.

View larger version (24K): [in this window] [in a new window]   Fig. 6.   HX-Dna2 possesses an endonucleolytic activity but not exonucleolytic activity. The schematic structures of each substrate are shown at the top of each figure. The sizes of the duplex region are indicated as bp, and the sizes of ssDNA region as nt at the top of the schematics. The asterisks denote 32P-labeled ends of each substrate. Increasing amounts of HX-Dna2 were added to reaction mixtures (20 µl) as described under "Experimental Procedures" and incubated at 37 °C for 10 min in the presence of 15 fmol of each substrate (A and B, 3'-overhang partial duplexes, 3'- and 5'-end-labeled, respectively; C, 5'-overhang partial duplex; D, partial duplexes at both ends). The amount of HX-Dna2 used in lanes of each panel was as follows: lane 1, no enzyme; lane 2, 1 ng; lane 3, 2 ng; lane 4, 5 ng; lane 5, 10 ng; lane 6, 20 ng; lane 7, 40 ng; lane 8, 80 ng; and lane 9, 160 ng. After incubation, the products were boiled and subjected to electrophoresis for 1.5 h at 35 watts through a 20% denaturing polyacrylamide gel (7 M urea) as described under "Experimental Procedures." The size markers (M) used were prepared by labeling synthetic oligo(dT) mixtures (2-, 3-, 4-, 6-, 8-, 10-, and 12-mers) and commercial size markers ((dGATC)n of 8-32-mer) (Amersham Pharmacia Biotech) at their 5'-ends with the polynucleotide kinase. The 52-mer was labeled separately and mixed with the commercial markers.

ATP Hydrolysis Enables the Dna2 Protein to Degrade dsDNA from the 3'-End-- Because HX-Dna2 possesses the ability to hydrolyze ATP in the presence of ssDNA, we investigated the effects of ATP on the HX-Dna2 endonuclease activity. The presence of low ATP concentrations (<2 mM) appeared not to affect the ssDNA-specific endonuclease activity of Dna2 as shown in Figs. 2 and 3. However, when a 3'-³²P-labeled partial duplex X174 was used as a substrate, we observed that the 3'-end label was released in the presence of ATP (Figs. 2B and 3B) but not in its absence (Figs. 2C and 3C). This indicates that HX-Dna2 had a dsDNA-specific nuclease activity that is dependent on ATP. In order to characterize this dependence, we measured the effect of ATP concentration on the dsDNA-specific nuclease activity. The maximal ATP effect was observed between 2 and 4 mM ATP in the reaction mixture (data not shown). When we examined the effects of other nucleoside triphosphates on this activity and analyzed the reaction products (Fig. 7, A and B), only ATP and dATP (approximately 70% of the effect of ATP) stimulated nuclease activity, and their products consisted of oligonucleotides of varying sizes (monomers to octamers) (Fig. 7A). Other nucleoside triphosphates and nonhydrolyzable ATP analogs did not support this activity (<5% ATP). This is in accordance with the ability of HX-Dna2 to hydrolyze only ATP and dATP (Fig. 7B), as previously reported (15). This observation suggests the following mechanism by which the 3'-end labels at duplex junction are removed by HX-Dna2. The Dna2 protein unwinds only partially the 3'-end of duplex DNA using the energy derived from ATP hydrolysis. This process generates a stretch of ssDNA of limited length that is susceptible to Dna2 in the solution or to Dna2 protein engaged in the melting process. Because the maximum size of the released products was 8 nt, the size of the segment of DNA melted by HX-Dna2 is believed to be 12 to 13 bases, taking into consideration that the enzyme leaves 4-5 nt of ssDNA uncleaved even in the presence of excess of HX-Dna2 (Fig. 6B). In order to examine whether the ATP-dependent nuclease activity of HX-Dna2 was specific to the 3'-end of DNA, its effect on a partial duplex substrate labeled at the 5'-end was also examined (Fig. 8). The sizes of major products from this substrate ranged from 30 to 42 nt, indicating that the enzyme preferentially attacks 3'-ends of duplex DNA. In conclusion, Dna2 is able to act upon dsDNA from the 3'-end of duplex DNA, and ATP hydrolysis is necessary for this reaction to proceed.

View larger version (16K): [in this window] [in a new window]   Fig. 7.   ATP hydrolysis is required for stimulation of the dsDNA-specific nuclease. A, an autoradiogram of the products of the reaction (20-µl) measuring the ATP-dependent dsDNA nucleolytic activity measured in the presence of different NTPs and dNTPs (2 mM each), as indicated at the top of figure. The reaction mixtures contained 20 ng of HX-Dna2 protein with 15 fmol of 3'-end-labeled substrate (5'-tailed substrate, a 73-mer annealed X174 sscDNA) and were incubated for 10 min as described under "Experimental Procedures." The size markers (M) and analysis of products were as described above in Fig. 6. The schematic structure of the substrate used is shown at the left. B, quantitation of results shown in A (), and results of ssDNA-dependent (d)NTPase activity (). The ssDNA-dependent (d)NTPase activities of HX-Dna2 were measured in reaction mixtures (20-µl) that contained 20 ng of HX-Dna2 in the presence of each of the (d)NTP (200 µM each) as indicated on the x axis. The reaction mixtures were incubated at 37 °C for 10 min as described under "Experimental Procedures."

View larger version (10K): [in this window] [in a new window]   Fig. 8.   The ATP-dependent nuclease preferentially releases the 3'-end labels of duplex DNA. Schematic structures of the 52-mer annealed X174 sscDNA substrates are shown at the top of the figure. The asterisks indicate 32P-labeled ends. The indicated amount of HX-Dna2 in the 20-µl reaction (see "Experimental Procedures") was incubated with either 3'-labeled (left panel) or 5'-labeled (right panel) substrate (15 fmol each) at 37 °C for 10 min in the presence (+) of 2 mM ATP and absence () of ATP. The size markers (M) and analysis of products were as described for Fig. 6.

Dna2 by Itself Is a Weak DNA Helicase-- As shown above, endonucleolytic cleavage of the 3'-ends of duplex DNA in the presence of hydrolyzable nucleoside triphosphates suggests that Dna2 is intrinsically capable of unwinding duplex DNA. In order to detect a DNA unwinding activity of HX-Dna2, we first investigated the reaction conditions optimal for the ssDNA-specific endonuclease and ATPase activities. We discovered that optimal concentrations of Mg²⁺ for the two reactions were significantly different; the endonuclease activity required the presence of 2.5-10 mM MgCl₂ for optimal activity, whereas the ATPase activity required significantly low concentrations of MgCl₂ (0.15-0.3 mM) (data not shown). In addition, the endonuclease activity was substantially inhibited at high levels (2-4 mM) of ATP, especially in the presence of reduced levels of MgCl₂ (data not shown). Therefore, we examined DNA unwinding activity of Dna2 under varying concentrations (0.5-2 mM) of MgCl₂ in the presence of a fixed concentration of ATP (2 mM) using X174 sscDNA that had 20-mer oligonucleotides annealed (Fig. 9). In this experiment, we tested three different Dna2 proteins as follows: (i) HX-Dna2, (ii) HX-Dna2K1080E which has a substitution at the conserved lysine residue in the ATP binding motif (16), and (iii) HX-Dna2405N with deletion of N-terminal 405 amino acids. HX-Dna2K1080E was prepared as reported (16) and was shown previously that it lacked DNA unwinding and ATPase activities.

View larger version (21K): [in this window] [in a new window]   Fig. 9.   DNA unwinding activity of HX-Dna2 protein is observed in a condition where nuclease activity is suppressed. Schematic structure of the partial duplex substrate (20-mer (5'-GGC GAT TGC GTA CCC GAC GA-3') annealed to X174 sscDNA) is shown at left. The asterisk indicates a radioisotopic label at the 3'-end. HX-Dna2K1080E has a substitution of glutamate for the conserved lysine residue in the ATP binding motif as reported (16). HX-Dna2405N has a deletion of N-terminal 405 amino acids as described under "Experimental Procedures." Enzymes (40 ng of HX-Dna2 and HX-Dna2K1080E; 20 ng of HX-Dna2 405N) were added to reaction mixtures (20 µl) containing 15 fmol of the 20-mer annealed X174 substrate in the presence of indicated amounts of Mg2+ and ATP and incubated at 37 °C for 10 min. After incubation, the products were subjected to electrophoresis for 1.5 h at 150 V through 12% polyacrylamide gel containing 0.1% SDS in 0.5× TBE. S and B denote substrate alone and boiled substrate controls, respectively. The arrow indicates the position where the 20-mer oligonucleotide migrated. The amounts of substrate unwound and 3'-end labels released were measured with the use of a PhosphorImager, and the results are presented at the bottom of the figure.

	DISCUSSION

Top Abstract Introduction Procedures Results Discussion References

Both ssDNA-specific Endonuclease and ATP-dependent Nuclease Activities Are Intrinsic to Dna2-- In this report, we examine the biochemical activities associated with the recombinant HX-Dna2 protein and present evidence that Dna2 possesses a ssDNA-specific endonuclease activity capable of degrading dsDNA in an ATP-dependent manner, in addition to the ssDNA-dependent ATPase and weak DNA unwinding activities as reported previously (15). To demonstrate that these activities are all intrinsic to the Dna2 protein, we carried out several types of experiments. (i) By using protein purification methods, we found that the endonuclease activity was resistant to separation from ssDNA-dependent ATPase (Figs. 2 and 3). Further extensive purification using the active FPLC Mono Q fractions (Fig. 2) was attempted by the use of phenyl-Sepharose, gel filtration, and FPLC Mono S column chromatographies, but we failed to separate any of these activities (data not shown). In addition, throughout each additional chromatographic step, the ratio of the endonuclease and ATPase activities remained constant (data not shown). (ii) Immunodepletion experiments revealed that antibodies specific for HX-Dna2 depleted concurrently the endonuclease, ATP-dependent nuclease, and ssDNA-dependent ATPase activities. The anti-Dna2 antibody also inhibited both endonuclease and ATPase activities, whereas a control antibody failed to do so (data not shown).

Dna2 by Itself Is a Weak DNA Helicase and May Require an Additional Protein to Gain a Strong Unwinding Activity-- We did not detect any marked DNA unwinding activity for HX-Dna2 with a partial duplex substrate that had 52-base oligonucleotide annealed under any of the conditions that we tested. However, we were able to demonstrate a low unwinding activity of both HX-Dna2 and HX-Dna2405N, although their unwinding activity was observed with a X174 substrate with a 20-bp partial duplex under a condition where the endonuclease activity was minimized by reducing the ratio of MgCl₂/ATP (Fig. 9). This suggests that the Dna2 protein possess an inherently inconspicuous DNA unwinding activity that can be activated by a yet unknown mechanism (see below). The duplex-specific ATP-dependent nuclease activity is likely to be a disclosure of such a concealed DNA helicase activity of Dna2. If this is the case, the preferential release of 3'-end labels from duplex DNA suggests that the Dna2 protein should translocate and unwind DNA in the 5'  3' direction, which is opposite to that reported in previous studies (15).

Potential Roles of DNA2 in Vivo-- We present herein the first example of ATP-dependent nuclease in eukaryotes. However, enzymes with both nuclease and DNA-dependent ATPase/DNA helicase activities are not uncommon in prokaryotes. The Bacillus subtilis AddAB enzyme, which is a multifunctional and multisubunit complex, possesses multiple nuclease activities and ATP-dependent exonuclease activities (29, 30). One of the AddAB subunits is a DNA-dependent ATPase and DNA helicase. This enzyme complex was shown to be involved in genetic recombination, DNA repair, and maintenance of cell viability. Another well known example is the E. coli RecBCD enzyme, which is a heterotrimeric protein that participates in the major homologous recombination pathway in E. coli (31, 32). Most recently, RexAB, a biological homolog of E. coli RecBCD, was identified from Lactococcus lactis (33). The RecBCD acts as a nonspecific nuclease in an ATP-dependent fashion until it comes across a recombination hop spot, the sequence. After this event, the enzyme loses its nuclease activity and acquires DNA unwinding activity, which in turn promotes thereafter ssDNA production. Although Dna2 is not a multisubunit complex, it displays biochemical activities that are analogous to those of the bacterial enzymes described above. If Dna2 is truly homologous to the bacterial enzyme complexes, the protein is likely to be involved in the processes similar to those of the prokaryotic enzymes (that is, DNA repair and recombination).