Functional Interactions of Mitochondrial DNA Polymerase and Single-stranded DNA-binding Protein

Functional interactions between mitochondrial DNA polymerase (pol γ) and mitochondrial single-stranded DNA-binding protein (mtSSB) from Drosophila embryos have been evaluated with regard to the overall activity of pol γ and in partial reactions involving template-primer binding and initiation and idling in DNA strand synthesis. Both the 5′ → 3′ DNA polymerase and 3′ → 5′ exonuclease in pol γ are stimulated 15–20-fold on oligonucleotide-primed single-stranded DNA by native and recombinant forms of mtSSB. That the extent of stimulation is similar for both enzyme activities over a broad range of KCl concentrations suggests their functional coordination and a similar mechanism of stimulation by mtSSB. At the same time, the high mispair specificity of pol γ in exonucleolytic hydrolysis is maintained, indicating that enhancement of pol γ catalytic efficiency is likely not accompanied by increased nucleotide turnover. DNase I footprinting of pol γ·DNA complexes and initial rate measurements show that mtSSB enhances primer recognition and binding and stimulates 30-fold the rate of initiation of DNA strands. Dissociation studies show that productive complexes of the native pol γ heterodimer with template-primer DNA are formed and remain stable in the absence of replication accessory proteins.

Single-stranded DNA-binding proteins (SSBs) 1 serve critical roles in DNA replication, repair and recombination (1). Whereas high affinity DNA binding by SSBs can occur independently of other proteins, both functional and physical interactions between SSBs and a variety of enzymes involved in the above processes have been documented. In particular, interactions between SSBs and replicative DNA polymerases have been demonstrated in bacterial, nuclear, and viral systems (1).
The near-homogeneous mitochondrial DNA polymerase from Drosophila embryos catalyzes relatively efficient DNA synthesis on both predominantly double-and single-stranded DNA templates (2,3), yet its activity and processivity are greatly affected by reaction conditions (4). Mitochondrial SSBs share similar physical and biochemical properties with Escherichia coli SSB (5-10), with which they exhibit a high degree of amino acid sequence conservation (10 -12). Considering the roles served by E. coli SSB in bacterial replication in helix destabilization (13) and in enhancing DNA polymerase processivity (14,15) and fidelity (16,17), we purified Drosophila mtSSB and studied its effects in in vitro DNA synthesis by pol ␥, in an assay that mimics lagging DNA strand synthesis in mitochondrial replication (9). These studies allowed the first demonstration of stimulation by a mtSSB of DNA synthesis by a nearhomogeneous pol ␥. Our biochemical data are consistent with an important role for mtSSB in mitochondrial DNA replication that has been documented genetically by the fact that a null mutation in the gene for the yeast homolog (RIM1) results in complete loss of mitochondrial DNA in vivo (7). Furthermore, we found that Drosophila mtSSB stimulates pol ␥ by a mechanism highly similar to that which we found for E. coli SSB (9,18). Here we demonstrate a dual role for mtSSB in initiation and elongation of DNA strand synthesis catalyzed by pol ␥, and evaluate for the first time the effects of mtSSB on the mispairspecific 3Ј 3 5Ј exonuclease in pol ␥.

Materials
Nucleotides and Nucleic Acids-Unlabeled deoxy-and ribonucleotides were purchased from Amersham Pharmacia Biotech. Enzymes and Proteins-Drosophila DNA polymerase ␥ (Fraction VI, Ͼ90% homogeneous) was prepared as described by Wernette et al. (3). Drosophila mtSSB (Ͼ90% homogeneous) was prepared from embryonic mitochondria essentially as described by Thommes et al. (9). Bovine pancreatic DNase I (Type IV) and T4 polynucleotide kinase were purchased from Sigma and Roche Molecular Biochemicals, respectively. Sequenase, version 2.0, was purchased from United States Biochemical Corp.

Methods
Bacterial Subcloning, Overexpression, and Purification of Recombinant mtSSB-The 372-base pair coding sequence of Drosophila mtSSB was engineered by polymerase chain reaction amplification of a fulllength cDNA clone (12) to contain NdeI restriction endonuclease sites at its ends; on the amino-terminal end, an NdeI site was created to contain an ATG at amino acid position 16, corresponding to the residue prior to the first amino acid in the mature Drosophila mtSSB (9), and on the carboxyl-terminal end, the NdeI site was positioned at a site 11 base pairs distal to the termination codon. The resulting DNA fragment was purified by gel electrophoresis, cleaved with NdeI, and cloned into the bacteriophage T7 promoter-based expression vector pET-11a (Novagen) at its unique NdeI site. The E. coli strain BL21 (DE3) (Novagen) was used for transformation, and ampicillin-resistant plasmid-containing cells were screened for insert size and orientation of recombinant DNA by restriction analyses.
For overexpression, pET-11a recombinant plasmid-containing BL21 (DE3) cells (400 ml) were grown at 37°C with aeration, in Luria broth containing 100 g/ml ampicillin. When the bacterial cells reached an optical density of 0.6 at 595 nm, isopropyl-1-thio-␤-D-galactopyranoside was added to 0.3 mM, and the culture was incubated further for 2 h. Cells were harvested by centrifugation, washed in 50 mM Tris-HCl, pH 7.5, 10% sucrose, recentrifuged, frozen in liquid nitrogen, and stored at Ϫ80°C.
For preparation of cell extracts and purification of recombinant mtSSB, frozen cells were thawed on ice, and all further steps were performed at 0 -4°C. Cells were suspended in 1 ⁄25 volume of original cell culture in 50 mM Tris-HCl, pH 7.5, 10% sucrose, 2 mM EDTA, 5 mM dithiothreitol (DTT), 1 mM phenylmethylsulfonyl fluoride (PMSF), 10 mM sodium metabisulfite, 2 g/ml leupeptin. Cells were lysed by incubation for 30 min in the presence of 0.3 mg/ml final concentration of hen egg white lysozyme (Roche Molecular Biochemicals) and 20 mM spermidine, 0.25 M NaCl, followed by freezing in liquid nitrogen and thawing on ice. The suspension was then centrifuged at 17,500 ϫ g for 30 min. The supernatant fluid was recovered for use as the soluble protein fraction for Cibacron blue agarose chromatography.
DNA Polymerase ␥ Assay-Reaction mixtures (0.05 ml) contained 50 mM Tris-HCl, pH 8.5, 4 mM MgCl 2 , 10 mM DTT, 0 -180 mM KCl as indicated, 400 g/ml bovine serum albumin, 20 M each of dGTP, dATP, dCTP, and [ 3 H]dTTP (1000 cpm/pmol), 10 M (as nt) singly primed recombinant M13 DNA, and 0.1 unit of Fraction VI enzyme (6-fold excess of primer ends over pol ␥ molecules). mtSSB (0.8 g) was added as indicated in the figure legends. Incubation was at 30°C for 30 min. Specific modifications are described in the figure legends. One unit of activity is that amount that catalyzes the incorporation of 1 nmol of deoxyribonucleoside triphosphate into acid-insoluble material in 60 min at 30°C using DNase I-activated calf thymus DNA as the substrate.
3Ј 3 5Ј Exonuclease Assay-Reaction mixtures (0.05 ml) contained 50 mM Tris-HCl, pH 8.5, 4 mM MgCl 2 , 10 mM DTT, 0 -180 mM KCl as indicated, 400 g/ml bovine serum albumin, 4 M 5Ј-end-labeled singly primed recombinant M13 DNA containing a 3Ј-terminal mispair, and 0.1 unit of Fraction VI enzyme. mtSSB (0.4 g) was added as indicated in the figure legends. Incubation was for 30 min at 30°C. Samples were then made 1% in SDS and 10 mM in EDTA, heated for 10 min at 65°C, and precipitated with ethanol in the presence of 1 g of sonicated salmon sperm DNA as carrier. The ethanol precipitates were resuspended in 80% formamide and 90 mM Tris borate. Aliquots were denatured for 2 min at 100°C, chilled on ice, and electrophoresed in an 18% polyacrylamide slab gel (13 ϫ 23 ϫ 0.075 cm) containing 7 M urea in 90 mM Tris borate, pH 8.3, and 25 mM EDTA. After electrophoresis, the gel was washed in 15% glycerol for 20 min and exposed to a PhosphorImager screen (Molecular Dynamics). The data were analyzed using the ImageQuant version 4.2a software.
Primer Extension Assay-Reaction mixtures (0.05 ml) contained 50 mM Tris-HCl, pH 8.5, 4 mM MgCl 2 , 10 mM DTT, 30 mM KCl, 400 g/ml bovine serum albumin, 30 M each of dGTP, dATP, dTTP, 20 M (as nt) 5Ј-end-labeled singly primed recombinant M13 DNA, and 0.35 unit of Fraction VI enzyme (3.5-fold excess of primer ends over pol ␥ molecules). mtSSB (2.0 g) was added as indicated in the figure legends. Incubation was at 30°C for the indicated times. Samples were then made 1% in SDS and 10 mM in EDTA, heated for 10 min at 65°C, and precipitated with ethanol in the presence of 1 g of sonicated salmon sperm DNA. Samples were electrophoresed, the gel was processed, and the data were quantitated as described above.
DNase I Footprinting-Pol ␥ interactions at the template-primer terminus were examined by DNase I footprinting on singly primed M13 DNA in an experimental scheme modified from Reems and McHenry (19).
A 49-nt primer was annealed to M13mp19 DNA (at a position corresponding to 993-1041 in the latter) and radiolabeled at its 3Ј-end essentially as per Reems and McHenry (19), except that the 3Ј-end was extended by 2 nt with Sequenase, version 2.0, in the presence of [␣-32 P]dATP (3000 Ci/mmol, 0.13 M) and ddTTP (0.8 M). The radiolabeled template-primer DNA was separated from excess unannealed primer and unincorporated nucleotide by gel filtration on a Sephadex G-50 column equilibrated with 10 mM Tris-HCl, pH 8.0, 1 mM EDTA.
The DNase I digestion reactions were performed in a reaction mixture (40 l) containing 50 mM Tris-HCl, pH 8.5, 4 mM MgCl 2 , 5 mM DTT, 30 mM KCl, 5% glycerol, 17 fmol of 3Ј-end-labeled template-primer DNA, and varying levels of pol ␥ Fraction VI and recombinant mtSSB (0.2 g) as indicated. The reaction mixtures were incubated for 5 min at 30°C. DNase I was then added, and the mixtures were incubated further for 2 min at 37°C. The digestions were terminated by addition of 150 l of stop solution (0.5% SDS, 0.2 M sodium acetate, 30 mM EDTA, 25 g/ml tRNA) and then extracted with an equal volume of phenol/ chloroform (1:1) and precipitated with ethanol. The ethanol precipitates were washed with 70% ethanol and resuspended in 95% formamide containing 10 mM NaOH, 1 mM EDTA, and 0.3 mg/ml each of bromphenol blue and xylene cyanol. Aliquots were denatured for 2 min at 100°C, chilled on ice, and electrophoresed in a 10% polyacrylamide slab gel (31 ϫ 38.5 ϫ 0.04 cm) containing 7 M urea in 90 mM Tris borate, pH 8.3, and 25 mM EDTA. After electrophoresis, the gel was processed, and the data were quantitated as described above.
DNA Synthesis Assay for the Stability of Pol ␥⅐DNA Complexes-The stability of pol ␥⅐DNA complexes in template-primer binding and enzyme idling was determined in an experimental scheme modified from Hacker and Alberts (20). Reaction mixtures contained 50 mM Tris-HCl, pH 8.5, 4 mM MgCl 2 , 10 mM DTT, 400 g/ml bovine serum albumin, 30 or 120 mM KCl, and 4 M singly primed M13mp7 DNA (as nt). For experiments to examine the effect of mtSSB on the stability of primer binding, mtSSB (0.4 g/200 pmol of DNA as nt) was added to the reaction mixture and incubated for 1 min at 30°C. Pol ␥ Fraction VI (0.2-0.7 unit) was then added, and the incubation was continued for 1 min. After incubaton with pol ␥, a 50-l aliquot was removed and terminated as described below. DNase I-activated calf thymus DNA (180 -600 M as nt) was added to the remainder of the reaction mixture to serve as a DNA trap. After DNA trap addition, 50-l aliquots were removed at varying times to tubes containing dGTP, dATP, and dTTP at a final concentration of 0.3 mM each. Reactions were incubated for an additional 2 min at 30°C to allow primer extension to the GGG trinucleotide position. Reactions were terminated and samples were processed and analyzed as described for the 3Ј 3 5Ј exonuclease assay.
Experiments to examine pol ␥ idling were conducted essentially as above. Pol ␥, in the presence or absence of mtSSB was preincubated with the M13 DNA substrate for 1 min at 30°C. dGTP, dATP, and dTTP were added to a final concentration of 0.5 M, and samples were incubated for 25 s at 30°C, to allow extension to the GGG trinucleotide position. The DNase I-activated calf thymus DNA trap was then added as above, and aliquots were removed at varying times to tubes containing dGTP, dATP, dTTP, and dCTP at a final concentration of 0.3 mM. The reactions were incubated for an additional 2 min at 30°C to allow primer extension to the hairpin by pol ␥ molecules still associated with the template-primer DNA. To ensure that no further primer extension was initiated on unreacted M13 DNA substrate molecules after addition of the DNA trap, control reactions were prepared and incubated as described above, except that they were terminated immediately after the variable time incubation period. Reactions were terminated and analyzed as above. Products resulting from DNA synthesis past the hairpin were analyzed in a 6% polyacrylamide, 7 M urea slab gel (13 ϫ 30 ϫ 0.15 cm).

Bacterial Overexpression and Purification of Recombinant
Drosophila mtSSB-Drosophila mtSSB was purified previ-ously to homogeneity from both whole embryos and embryonic mitochondria with a high yield of 0.5-1.0 g/g of embryos (9). To pursue further biochemical and physical studies of the interaction of mtSSB with Drosophila mitochondrial DNA polymerase, we sought a more amenable source and a simpler purification. To that end, we employed bacterial overexpression and developed a two-step purification scheme. Having determined the amino-terminal sequence of the protein purified from Drosophila embryos (9), we engineered a plasmid construct in the T7 promoter-based vector pET-11a from a full-length cDNA clone, to produce the mature protein lacking the 16 amino acid presequence. Overexpression upon isopropyl-1-thio-␤-D-galactopyranoside induction of plasmid-containing BL21 (DE3) cells yielded ϳ16 g of mtSSB/ml of cell culture, ϳ60% of which remained in the soluble fraction upon cell lysis. Protein analysis by SDS-polyacrylamide gel electrophoresis, followed either by silver staining or by immunoblot analysis with rabbit antiserum against native mtSSB from Drosophila embryos, identifies the overexpressed polypeptide of ϳ18 kDa as recombinant mtSSB (Fig. 1). We then purified the recombinant mtSSB from soluble cell extracts by a single chromatographic step in which the protein bound to Cibacron blue agarose is washed under stringent conditions (in the presence of 0.8 M NaCl), and eluted with increasing concentrations of sodium thiocyanate. The 1.5 M sodium thiocyanate eluate containing the recombinant mtSSB represents a yield of ϳ60% of homogeneous protein; the native tetrameric form was recovered upon glycerol gradient sedimentation of a dialyzed and concentrated blue agarose fraction. Overall, the yield of D. melanogaster mtSSB is 12-fold greater from 400 ml of induced bacterial culture than from 200 g of Drosophila embryos.
Mitochondrial SSB Stimulates Both the DNA Polymerase and 3Ј 3 5Ј Exonuclease Activities of Drosophila DNA Polymerase ␥-We showed previously that SSB stimulates the DNA polymerase activity of Drosophila DNA polymerase ␥ on singly- primed M13 DNA, in an assay that mimics lagging DNA strand synthesis in mitochondrial replication (9,18). To extend these results and provide a comparative analysis for the native and recombinant proteins, we assayed both forms in the DNA synthesis reaction, and also evaluated their effects in mispair hydrolysis by pol ␥ on the same DNA substrate containing 3Ј-terminal mispaired primers. Native and recombinant Drosophila mtSSB stimulate similarly the DNA polymerase activity of pol ␥ over a broad range of KCl concentrations (Fig. 2). DNA polymerase activity is stimulated 18-and 21-fold, respectively, at 15 mM KCl, thereby lowering the KCl concentration required to achieve optimal DNA synthetic rate 8-fold relative to the 120 mM KCl concentration that is optimal in the absence of SSB.
Likewise, the native and recombinant Drosophila mtSSBs stimulate the 3Ј 3 5Ј exonuclease activity of pol ␥. Both the maximal stimulation, 16-and 13-fold, respectively, and the KCl titration curves are highly similar to those for the DNA synthetic reaction. This suggests the likelihood of effective coordination of the two activities in native pol ␥ under the fluctuating ionic conditions present in the mitochondrial matrix (21,22). Notably, the mispair specificity of the 3Ј 3 5Ј exonuclease is unchanged in the presence of mtSSB: over the entire KCl range, less than 10% of the paired termini generated by 3Јterminal mispair hydrolysis were hydrolyzed (data not shown). Thus, although we have shown previously that mtSSB enhances the processivity of pol ␥ in nucleotide polymerization, contributing severalfold to the overall stimulation of DNA synthetic rate (9), the mechanism of stimulation of the 3Ј 3 5Ј exonuclease is clearly unrelated to enhanced processivity. At the same time, it may be likely that mtSSB would enhance the processivity of pol ␥ in hydrolysis of multiple 3Ј-terminal mispairs.
Our working hypothesis for the mechanism of pol ␥ stimulation is that mtSSB increases the rate of initiation on single-stranded substrates for both DNA synthesis and exonucleolytic hydrolysis. We examined the former experimentally using a time course analysis of DNA synthesis that involves limited primer extension on M13 DNA in the absence of dCTP, such that DNA strand termination occurs after polymerization of 8 or 11 deoxynucleotides. We find that mtSSB stimulates the production of short nascent strands 10 -30-fold from 5-120 s of incubation (Fig. 3). This stimulation results in 40 versus 1.3% of the substrate being utilized at 2 min of incubation in the presence versus the absence of mtSSB, respectively.
Assuming a binding site size of 68 nt per tetramer (9), mtSSB was used in this analysis at a level 2.5-fold in excess of that required to saturate the ssDNA and was preincubated with the DNA substrate in the presence of reaction mix containing dNTPs for 5 min prior to pol ␥ addition. Thus, the lag before maximal stimulation of DNA strand synthesis likely relates to the time required for formation of productive pol ␥ complexes. mtSSB Enhances Primer Binding by Pol ␥-mtSSB may stimulate the rate of initiation of DNA synthesis by enhancing primer recognition and binding, by enhancing formation of stable and/or productive pol ␥⅐template-primer DNA complexes, and/or by stimulating nucleotide polymerization per se. To evaluate these possibilities, we began with a DNase I footprinting analysis of template-primer DNA binding. Because pol ␥ has a high affinity for ssDNA, we examined primer-terminus interactions directly using an M13 DNA substrate to which was annealed an oligonucleotide primer (49 nt) labeled at its 3Јterminus. We find that in the absence of accessory proteins, the pol ␥ heterodimer forms a stable complex with template-primer DNA that results in DNase I protection of 20 nt of the primer DNA strand (Fig. 4A). Whereas addition of mtSSB sufficient to saturate the ssDNA did not alone result in any observable footprint, mtSSB reduced 3-4-fold the amount of pol ␥ required to detect a stable footprint on the primer terminus, to a level corresponding to a single pol ␥ molecule per template-primer  1-8, respectively). B, reactions were performed in the presence of saturating recombinant mtSSB and 0, 7, 14, 27, 55, 83, or 0 fmol Drosophila pol ␥ (lanes 2-8, respectively). Lane 1 represents a control reaction lacking pol ␥ and mtSSB. The sizes of the DNA fragments were determined relative to a DNA sequencing ladder electrophoresed in a parallel lane (not shown). 7 fmol of Drosophila pol ␥ corresponds to 1.1 ng and 0.09 units. DNA (Fig. 4B). Thus, mtSSB enhances primer recognition and binding, yet the 3-4-fold increase can account only partially for the 30-fold stimulation of the rate of initiation observed under similar low salt reaction conditions (Fig. 3). In this regard, the DNase I footprinting analysis does not relate directly substrate DNA binding to catalysis and so does not measure the potential contribution of mtSSB to productive complex formation, which might exhibit a shorter half-life than that required to observe a stable footprint. The footprinting analysis of primer binding may therefore represent an underestimate of the effect of increased binding on catalysis.
Stability of Pol ␥⅐DNA Complexes in Template-Primer Binding and Enzyme Idling-mtSSB may stimulate primer recognition and template-primer binding indirectly, by eliminating nonproductive binding of pol ␥ to ssDNA regions. Alternatively or in addition, mtSSB may stabilize enzyme-DNA interactions, increasing the lifetime rather than the formation of productive complexes. We examined the latter by measuring dissociation rates using a strategy developed by Hacker and Alberts (20) to study the bacteriophage T4 holoenzyme. The experimental scheme is shown in Fig. 5. M13 DNA primed with a 5Ј-endlabeled oligonucleotide is used as the substrate for DNA binding and catalysis. By staging the addition of a DNA trap either before or after dNTP addition, we measured by DNA product strand analysis the rate of dissociation of pol ␥, in the presence or absence of mtSSB, either from the primer terminus or after the incorporation of eight nucleotides, where the enzyme pauses in the absence of dCTP at the position of the three consecutive template dGMP residues.
We find that mtSSB has a modest effect on the dissociation rate of pol ␥ bound at the primer terminus and measured in terms of productive complexes, that is, enzyme bound with the capacity for nucleotide polymerization (Fig. 6). At 30 mM KCl, where stimulation by mtSSB is 30-fold at 2 min in the initiation assay, the half-life of pol ␥ primer binding is 13.2 Ϯ 2.3 and 4.6 Ϯ 0.82 min in the absence and presence of mtSSB, respectively. Because the half-life for dissociation in either case is longer than the time required to observe the effect on initiation, the negative contribution of mtSSB on complex stability is minimized. The relatively long half-life of the pol ␥⅐DNA complexes suggests that formation rather than the stability of productive complexes is rate-limiting in the overall replication scheme. The apparently negative effect of mtSSB on complex stability at low salt is not present at 120 mM KCl, where stimulation is less than 3-fold, and pol ␥ activity alone is highest. Notably, the dissociation rate at 120 mM KCl is 15-fold higher than at 30 mM KCl, with a half-life of 0.87 Ϯ 0.15 min. This likely reflects a salt-stimulated rapid recycling of pol ␥ upon nonproductive DNA binding to ssDNA regions.
As with template-primer DNA binding, we find that mtSSB has only a modest effect on the dissociation rate of pol ␥ upon enzyme idling after polymerization of 8 nt (Fig. 7). The dissociation rate of pol ␥ upon idling is 16 -40-fold slower at 30 mM than at 120 mM KCl, with half-life values of 17.9 Ϯ 3.2 and 1.08 Ϯ 0.18 min, respectively, in the absence of mtSSB, and 33.8 Ϯ 6.0 (30 mM KCl) and 0.84 Ϯ 0.15 min (120 mM KCl) in its presence. The stabilizing effect of mtSSB on the idling complex likely reflects the structural difference between an elongation complex versus an initiation complex. The functional or physical interactions between the enzyme and mtSSB may also change once pol ␥ engages in processive DNA synthesis. Notably, whereas DNA synthesis is stalled at the position of the 22-base pair hairpin helix in the DNA substrate (see Fig. 5) in the absence of mtSSB, pol ␥ polymerizes 3 nt into the helix at 120 mM KCl (Fig. 7D) and completely through it at 30 mM KCl (Fig. 7B) in the presence of mtSSB. This and the longer half-life of the stalled complex at 30 mM KCl are likely a reflection of both the substantially higher processivity of pol ␥ at low salt (18), and the further enhancement of enzyme processivity by the helix destabilizing function of mtSSB (9).
Taken together, the dissociation experiments indicate that whether bound at the primer terminus in the absence of nucleotides, or stalled in the polymerization mode by a single nucleotide omission, pol ␥⅐DNA complexes are remarkably stable. Thus, the major contribution of mtSSB in stimulating the rate of initiation of DNA synthesis is most likely in complex formation. DISCUSSION We have evaluated mechanistically the effects of mtSSB on the catalytic activities of pol ␥ under in vitro conditions that mimic lagging DNA strand synthesis in mitochondrial replication. Our finding that mtSSB stimulates similarly both the DNA polymerase and 3Ј 3 5Ј exonuclease activities of pol ␥ over a broad range of KCl concentrations suggests functional coordination of the two activities at the replication fork. In evaluating 3Ј 3 5Ј exonuclease activity, we measured specific mispair hydrolysis on oligonucleotide-primed ssDNA. At the replication fork, the growing 3Ј-terminus is paired to the template strand that is coated with mtSSB (23). Thus, coordination of DNA polymerase and 3Ј 3 5Ј exonuclease in pol ␥ is anticipated in the catalysis of proofreading DNA synthesis that is required to ensure replication fidelity, and our results support this. In fact, that mispair hydrolysis is specifically stimulated by mtSSB over base pair hydrolysis may reflect a lower cost of editing in mitochondrial DNA replication as compared with bacterial and bacteriophage systems with replicative DNA polymerases that exhibit high nucleotide turnover (1). FIG. 5. Scheme for assay of template-primer DNA binding and enzyme idling. A, template-primer DNA structure. The 5Ј 32 P-labeled oligonucleotide primer (38 nt) is annealed to M13mp7 DNA at a site corresponding to map positions 6291-6329. Downstream from the primer are eight C, A, or T nucleotides followed by three G nucleotides (GGG) in M13mp7 DNA; 32 nt farther downstream is a 22-base pair DNA hairpin helix. B, experimental scheme. In template-primer DNA binding experiments, dGTP, dATP, and TTP are added after incubation periods with radiolabeled DNA substrate followed by the DNA trap (see under "Experimental Procedures"). Product DNA strands are measured at the position of the GGG trinucleotide. In enzyme idling experiments, dGTP, dATP, and TTP are added in the first incubation with the radiolabeled DNA substrate. dCTP is added after the DNA trap to measure DNA product strand extension from the GGG trinucleotide to the DNA hairpin helix. The experimental scheme was modified from that described by Hacker and Alberts (20).
We have shown previously that SSB increases severalfold the processivity of nucleotide polymerization by pol ␥ (18). Nevertheless, neither the major contribution of mtSSB to increasing the rate of nucleotide polymerization nor its stimulation of 3Ј 3 5Ј exonuclease are apparently a consequence of it.
We have shown here that in stimulating the rate of initiation of DNA strand synthesis by pol ␥, mtSSB increases the fraction of substrate molecules utilized 30-fold and thus likely recruits pol ␥ to the primer terminus. Whether this occurs by an active or passive mechanism remains to be elucidated, but our DNase I footprinting analysis supports the recruiting model. Remarkably, the footprinting data show that unlike most replicative DNA polymerases, the native pol ␥ heterodimer forms stable complexes with template-primer DNA, and template-primer binding is enhanced in the presence of mtSSB. mtSSB alone does not protect the primer from digestion, nor does it alter the footprinting pattern of pol ␥, which protects two helical turns of the DNA template strand at the primer terminus. Mikhailov and Bogenhagen (24) found mtSSB to inhibit binding of Xenopus pol ␥ to template-primer DNA, in a gel mobility shift analysis on short oligonucleotides. On such substrates, we would anticipate efficient binding of pol ␥ alone because the ssDNA present is less than 50 nt per substrate molecule. Indeed, we find that oligonucleotide template-primer binding by Drosophila pol ␥ is not stimulated by mtSSB nor is the initiation of DNA synthesis on such substrates. 2 We suggest that initiation of lagging DNA strand synthesis in mitochondrial replication should require mtSSB-facilitated primer recognition and binding by pol ␥ where, as in our model M13 assay in vitro, the displaced lagging DNA strand template is thousands of nucleotides in length (25). Our data show clearly that the rate of initiation of DNA strand synthesis to produce an 8 -11-mer on a long ssDNA is stimulated up to 30-fold by mtSSB.
What factors mediate the functional interactions of pol ␥ with mtSSB? Functional interactions may occur upon DNA binding and/or by specific physical interactions. E. coli SSB is known to bind DNA in several modes depending on ionic conditions (26). Studies of mtSSBs also show salt-dependent effects on DNA binding and, in particular, on binding site size and in cooperativity of DNA binding (5,8,9). Ionic conditions likely affect both DNA conformation and SSB structure, and these may vary on different template primers. Likewise, pol ␥ binding to the template DNA at the primer terminus may differ from that in long ssDNA regions. Our data show that mtSSB stimulates DNA synthesis by pol ␥ on M13 DNA over a broad range of salt concentrations, under which the effects on mtSSB on template-primer binding and dissociation by pol ␥ vary substantially. This suggests flexibility in presumptive proteinprotein interactions and perhaps in the mode of mitochondrial DNA replication under the fluctuating ionic conditions that occur in vivo (21,22).
Specific physical interactions have been demonstrated between replicative DNA polymerases and SSB proteins in several systems. Both bacteriophage T4 and T7 DNA polymerases interact physically with their cognate SSBs, the gene 32 and 2.5 proteins, respectively (27)(28)(29)(30). That T7 DNA polymerase is stimulated in DNA synthesis by both the T7 and E. coli SSBs (29,(31)(32)(33) suggests that whereas specific physical interactions may facilitate polymerase activity, they are not essential for functional interactions, at least as measured by in vitro DNA synthesis. In that regard, E. coli DNA pol III holoenzyme has recently been shown to interact physically with E. coli SSB (34,35). This interaction involves the subunit of the former and the acidic carboxyl-terminal tail of the latter and has a dual effect on in vitro DNA synthesis under elevated salt conditions in DNA strand initiation and in chain elongation. Our functional data resemble those studies: mtSSB stimulates both primer recognition by pol ␥ and enhances the processivity of nucleotide polymerization. Interestingly, although mtSSB is a homolog of bacterial SSB, sharing significant sequence (7, 10 -12) and structural similarity (36), it lacks the acidic carboxyl terminus required for physical interaction in both E. coli SSB and in T7 gene 2.5 protein (30). In fact, the carboxyl terminus appears as a disorganized loop in the crystal structure of the human mtSSB (36). Thus, future studies to link functional with physical interactions between pol ␥ and mtSSB have the potential to reveal novel interaction domains.