Processivity of Mitochondrial DNA Polymerase from Drosophila Embryos EFFECTS OF REACTION CONDITIONS AND ENZYME PURITY*

polymerase from Drosophila embryos has been characterized with regard to its mechanism of DNA synthesis under the influence of a variety of compounds in moderate salt (120 mM KCl), where the enzyme is most highly active and only moderately processive, and in low salt (30 mM KCl), where it is less active yet most highly processive. Disparate activity and processivity optima were obtained in low salt in the presence of varying pH or MgClz and ATP concentrations; in moderate salt, optimal activity and processivity were achieved coincidentally. Whereas no correlation between processivity and activity optima was observed upon addition of polyethylene glycol in either low or moderate salt, the optima were coincident at both salt levels on addition of glycerol. None of the reaction conditions examined allowed DNA polymerase 7 to exhibit maximal activity and processivity con- currently; maximal activity was always achieved in moderate salt and the highest processivity in low salt. However, while limiting the availability of primer termini had no effect on the mechanism of DNA synthesis, we found that the ability of mitochondrial DNA polymerase to singly primed M13 DNA was enhanced then diminished during the course of purification, suggesting loss of an accessory factor.

eukaryotic DNA replication generally correlates with processive synthesis (6-12). Whereas DNA polymerases responsible for continuous synthesis are highly processive, catalyzing the incorporation of thousands of nucleotides (nt) into the growing DNA chain in a single binding event, those involved in discontinuous synthesis may be processive for only tens or hundreds of nucleotides. While mitochondrial DNA polymerase might be expected to catalyze highly processive DNA synthesis, we found previously that when assayed under conditions optimized for DNA synthetic rate, Drosophila pol y is only quasi-processive, incorporating 30-90 nt per binding event (13).
The mitochondrial matrix environment is much different from that of the nucleus; the matrix contains a higher protein concentration (calculated as -50% by weight; Ref. 14) and higher nucleoside triphosphate levels (15). In addition, the matrix volume varies up to -2-fold depending on the metabolic state of the mitochondria ( N ) , resulting in fluctuating concentrations of metabolites that contribute to the compartmental ionic strength. These alterations may be important in the regulation of the metabolic pathways within the matrix (14), including mtDNA replication. Interestingly, the mechanism of DNA synthesis by E. coli DNA polymerases I and I11 and eukaryotic DNA polymerases a and 6 was found to vary depending on reaction conditions (17-20). The sensitivity of these replicative enzymes to their assay environments coupled with the unusual mitochondrial environment suggests that mitochondrial DNA polymerase might also be influenced by the composition of the reaction solvent. We have examined the effect of various reagents on the rate and processivity of DNA synthesis by Drosophila DNA polymerase y on a singlestranded DNA template to address the apparent contradiction of continuous DNA synthesis catalyzed by a moderately processive enzyme. The elucidation of in vitro reaction conditions that enable mitochondrial DNA polymerase to catalyze highly processive and efficient DNA synthesis concurrently would allow correlation of enzyme function with models of mtDNA replication, as well as suggest a mechanism for its regulation. In contrast, an inability to define such conditions may implicate as yet unidentified DNA polymerase accessory factors in mtDNA replication.

Materials
Nucleotides and Nucleic Acids-Unlabeled deoxy-and ribonucleoside triphosphates were purchased from P-L Biochemicals; for use at concentrations above 30 p~, ATP, GTP, and ADP solutions were adjusted to pH 7.5 with Tris base (Research Organics 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-insoluhle material in 60 min at 30 "C using DNase Iactivated calf thymus DNA as the substrate. Here, we define standard activity as that exhibited hy pol y in the presence of 120 mM KC1 on singly primed M1:l DNA. Analysis of Products of I'rocpssiuc DNA SynthPsis bv G P~ FJcctrophoresis-Reactions were as above except that reaction mixtures contained 30 p~ each of dATI', dCTP, dGTI', and 10 p~ [o-:"I'] d T T P (2 X lO'cpm/pmal), 20 p~ singly primed M13 DNA, and 0.02 units of Fraction V I enzyme. Incuhation was at 30 "C for 30 min unless otherwise noted. Products to he analyzed by denaturing polyacrylamide gel electrophoresis were made l!c in SDS and 10 mM in EDTA. heated for 4 min at 80 "C. and precipitated with ethanol in the presence of 5 pg of tRNA as carrier. The ethanol precipitates ~ were resuspended in 80% formamide and 90 mM Tris borate. Aliquots were denatured for 2 min at 100 "C and electrophoresed in a 6';, polyacrylamide slah gel (13 X 30 X 0.15 cm) containing 7 M urea in 90 mM Tris-borate (pH 8.3) and 25 mM EDTA. Alternatively, the ethanol precipitates were resuspended in 0.3 M NaOH and 20 mM EDTA, and aliquots electrophoresed in a 1.5% agarose slah gel (13 X 18 X 0.7 cm) containing 30 mM NaCl and 2 mM EDTA in 30 mM NaOH and 2 mM EDTA. Approximately equal amounts of radioactivity (-1000cpm) were loaded in each lane. In addition, equal sample volumes were loaded on each t.ype of gel to allow direct comparison of product size distrihution. Gels were washed in distilled water for 20 min, dried under vacuum, and exposed at -80 "C to Kodak X-Omat AR x-ray film using DuPont Quanta Ill intensifying screens. Quantitation of the data was performed by scanning of the autoradiographs using a Rio-Image Visage 110 digital imager. The area under the peaks was determined hy computer integration analysis and was normalized to the nucleotide level to correct for the uniform labeling of the DNA products. In the determination of processivity values, the length of the primer (15 nt) was subtracted from the DNA product strand lengths.

RESULTS
Drosophila DNA Polymerase y Is Highly Processive a t KC1 Levels Suboptimal for DNA Synthesis-Condensation of the mitochondrial matrix during periods of active respiration likely results in an increase in the concentration of metabolites that contribute to ionic strength (16). Increasing ionic strength decreases both the activity and processivity of E. coli DNA polymerases I and 111, calf thymus DNA polymerase ( y , and T4 phage DNA polymerase (17)(18)(19)23). Recause Drosophila mitochondrial DNA polymerase is stimulated by moderate salt (13, 22), we investigated its processivity on singly primed M13 DNA under conditions of varying monovalent salt concentration. Polymerase y activity varied -6-fold when assayed over the range of 0-210 mM KCI, with an optimum a t -120 mM KC1 (Fig. 1A). At the same time, processivity varied dramatically (-600-fold). with the greatest abundance containing no KC1 (1anv.s I and 2 nverngr prncessive unit (apt11 = 3400 nt). 30 mM KC1 (Inncs : { and 4 ; npu = 2500 n t ) . GO mM KC1 (lanm .5 and fi; apu = 1600 nt ).90 mM K('l (lanes 7 a n d H; apu = 140 nt), 120 mM KC1 (lanps 9 and 1 0 ; npu = 45 n t ) . 160 rnv K('I (lnnm I 1 and 12: apu = 20 nt), 210 mM K('1 ([ones nntl 1.1; npu = G nt). NumhPrs at lrfl indicate the position and sire (in nt) of Hpnll restriction fragments of M1:Khril replientive form D N A (46) and Hind111 restriction fragments of X I)NA that were electrophoresed in adjacent lanes. In the determination of proressivity vnlrre.;. the Irnnh of the primer (15 n t ) wns subtracted from the 1)NA product strnnd lengths measured relative to the moieculnr weight marker.; shown. Products obtained after 41) min of incut~ntion werr similnr in size nnd distrihution to those obtained after 2 0 min of incrrhntinn. indicating that they result from single binding events; suhsec~~rent nnnlyses w w c . performed using 30-min incubations nt :IO "('. of full-length products synthesized in the absence of KC1 (Fig.   1R). Reduced polymerase activitv under conditions where pol y is capable of copying a complete 6407-nt template in a single binding event suggests that either primer recognition or enzyme dissociation is rate-limiting.
pH Has Little Effect on the Processivity of Drosophila Mitochondrial DNA Polymerase-The mitochondrial inner membrane is generally impermeable to charged and highly polar molecules, allowing the formation of a pH gradient across the lipid bilayer. The extent of this gradient varies with matrix pH, as the cytosol seems to be effectively buffered (24). Lowering the pH of the reaction solvent from pH 8.0 to 6.0 increased the processivities of calf thymus DNA polymerases a and 6 (pol a and pol 6, respectively) approximately 30-fold (20). In contrast, polymerase activity was reduced 5-7-fold at pH 6.0 relative to its optimum at pH 7.0. When assayed over the range of pH 6-10, the activity of Drosophila pol y varied -3.5-fold at moderate salt and -10-fold at low salt, with optimal activity achieved between pH 8.0 to 9.6 and at pH 10, respectively ( Fig. 2A). However, processivity varied less than 4-fold regardless of KC1 concentration, and the enzyme was most highly processive at pH 8.5 at both moderate and low salt (Fig. 2B). Thus, optimal polymerase activity and processivity were achieved concurrently at 120 mM KCl. However, like pol a and 6, pol y was most processive at pH suboptimal for activity at 30 mM KCl.
DNA Synthetic Product Length Is Diminished with Increas- Polymerase activity varied -10-fold at 120 mM KC1 and was optimal from 0.5 to 6 mM MgC12 (Fig. 3A). At 30 mM KC1, pol y activity varied -30-fold and the enzyme exhibited two optima at 0.25 and 20-25 mM MgC12. Thus, the enzyme was capable of synthesizing long products (>3000 apu) at a high rate (0.2-0.5 mM MgC12 at 30 mM KCl), yet was also able to make short products (545 apu) quickly (0.25-7 mM MgC12 at 120 mM KCl, and 18-27 mM MgC12 at 30 mM KCl). ATP Specifically Stimulates DNA Synthesis by Drosophila y Polymerase-Seventy percent of the total cellular ATP is localized within the mitochondrial matrix (25), making it at least 4-fold more abundant than the other nucleoside triphosphates present (15). ATP hydrolysis is involved in primer recognition by, and increases the processivities of prokaryotic, viral, and eukaryotic DNA polymerases (26-29). ATP, but not GTP or ADP, stimulated the activity of Drosophila mitochondrial DNA polymerase 4-fold at low salt in the presence of ATP concentrations from 7.5 to 12 mM, such that the activity observed at moderate salt in the absence of ATP was achieved (Fig. 4, A-C). This specific stimulation by ATP was accompanied by a non-proportional decrease in processivity (Fig. 40). Notably, while ATP did not stimulate activity at 120 mM KC1, it decreased processivity above 5 mM ATP (Fig.  4D), as did GTP and ADP at both moderate and low KC1 concentrations (data not shown). Titration experiments indicated that chelation of M e by ATP was unrelated to the specific stimulation observed; the MgClz optimum remained at 4 mM as the ATP concentration varied from 5 to 12 mM (data not shown).
Effects of Macromolecular Crowding Agents-Protein concentration in the mitochondrial matrix has been calculated to be -50% by weight; thus, the behavior of matrix enzymes may be expected to lie between that of enzymes in solution and that of enzymes in ordered complexes (14). The conditions provided by macromolecular crowding agents in uitro may mimic those found in the matrix. Crowding agents (dextran and polyethylene glycol (PEG)) enhance binding of E. coli DNA polymerase I and T4 DNA polymerase to template- primers, resulting in increased activity, yet having no effect on processivity (30, 31). Protein-protein interactions are also enhanced, as indicated by an increased processivity of T 4 DNA polymerase in the presence of its accessory proteins (31). Drosophila y polymerase exhibited a 7-fold variation in activity upon the addition of PEG in the range of 0-24%: activity decreased in the range of 4-12% a t both 120 and 30 mM KC1 and was restored either partially or fully at -16% PEG at moderate and low salt, respectively (Fig. 5A). pol y synthesized longer DNA products in the presence of 8% PEG a t both 120 and 30 mM KC1 (Fig. 5B), where its activity was 2-3-fold reduced. In contrast, polyvinyl alcohol had little effect on either the activity or processivity of the mitochondrial DNA polymerase (data not shown).
Glycerol Stimulates DNA Polymerase Activity and Processiuity Concurrently-Glycerol acts to stabilize enzyme activity and conformation (32) and may enhance the association of proteins in solution. Drosophila pol y activity was stimulated approximately 2-fold when assayed in the presence of -20% glycerol at both moderate and low salt (Fig. 6A). In addition, its processivity increased 1.5-3.5-fold with the inclusion of -15% glycerol (Fig. 6 B ) , demonstrating that pol y is capable of exhibiting optimal activity and processivity concurrently has little effect on the processivity of DNA synthesis by Drosophila pol y. DNA synthesis reactions were carried out in the presence of a DNA "trap" to ensure processive DNA synthesis. Reaction mixtures were as described under "Methods," except that they contained 2,8, or 30 p~ 5'-end-labeled singly primed M13 DNA and 0.4, 0.2, or 0.1 units of pol y, respectively. After incubation a t 30 "C for 5 min, prewarmed (30 "C) DNase I-activated calf thymus DNA (360 p~) and 30 p~ each of dATP, dCTP, dGTP, and dTTP were added simultaneously. Reactions were incubated a t 30 "C for another 30 min, and the product DNA strands isolated and analyzed on denaturing 1.5% agarose (upper panel) and 6% polyacrylamide (lower panel) gels. Reactions were performed a t 30 mM KC1 (lanes I -3) or 120 mM KC1 (lanes 4-6) a t pol y to template-primer ratios of 2:l (lanes I and 4 ) , 1:4 (lanes 2 and 5). 1:30 (lanes 3 and 6).
at each KC1 level examined in the presence of glycerol. Limited Template-Primer Availability Does Not Alter the Processivity of Drosophila y Polymerase-Protein-protein interactions between DNA polymerase molecules and/or accessory proteins have been demonstrated in both prokaryotic and eukaryotic systems (33). To promote such interactions between pol y molecules and/or putative sub-stoichiometric accessory proteins, we analyzed enzyme activity under conditions of limiting primer termini. After preincubation of the near-homogeneous enzyme with an M13 DNA template containing 5'-end-labeled primers, DNA synthesis was carried out in the presence of a 45-fold excess of unlabeled DNase Iactivated calf thymus DNA, which serves to trap unbound enzyme molecules, thereby ensuring a single DNA synthetic cycle. The processivity of Drosophila y polymerase was unaffected when the template-primer to enzyme molecule ratio was varied over a 60-fold range, from 0.5 to 30 primer termini per pol y molecule (Fig. 7).
The Ability of Mitochondrial DNA Polymerase to Copy Efficiently Single-stranded DNA Is Enhunced, Then Diminished during Purification-The extent of DNA synthesis by nearhomogeneous Drosophila pol y varies with the template-  lanes 1-3) or 120 mM KC1 (lanes 4-6) and containing pol y Fraction IV (-8% of homogeneous; lanes 1 and 4; apu = 1800 and 70 nt, respectively), Fraction V (-19% of homogeneous; lanes 2 and 5; apu = 3000 and 65 nt), Fraction VI (near-homogeneous; lanes 3 and 6; apu = 3400 and 50 nt). Less pure fractions could not be used due to the nucleases present.
primer utilized (13). In the two-subunit Fraction VI enzyme, DNA polymerase activity was -10-fold lower on singly primed M13 DNA than on DNase I-activated calf thymus DNA. To investigate the loss of putative polymerase accessory proteins that are required for efficient DNA synthesis on singlestranded DNA substrates, we compared the pol y activity ratio on the above DNAs during the course of purification. Whereas the activity ratio increased in the first few steps of purification, likely reflecting the removal of inhibitors of DNA synthesis on M13 DNA such as nucleases, a 4-fold decrease in the ability of pol y to copy M13 DNA was observed between Fractions V and VI (Fig. 8A). Because there was no change in processivity (Fig. 8B), the data suggest the loss of an accessory factor that stimulates DNA synthesis without affecting the mechanism of nucleotide incorporation.

DISCUSSION
Replication of the Drosophila mitochondrial genome proceeds asynchronously by a mechanism in which up to 98% of the leading DNA strand is copied prior to initiation of lagging DNA strand synthesis (34,35). Although continuous DNA strand synthesis would be consistent with this mechanism, biochemical characterization of Drosophila mitochondrial DNA polymerase does not fully support it; DNA polymerases implicated in continuous DNA strand synthesis are generally highly processive, but we have shown that pol y is only moderately processive under reaction conditions optimal for DNA synthetic rate (13). By altering various parameters of in vitro DNA synthesis, we hoped to increase the processivity of y polymerase while maintaining a high rate of nucleotide polymerization. We show that mitochondrial DNA polymerase, like other prokaryotic and eukaryotic DNA polymerases, is sensitive to its assay environment, but that changes in reaction conditions yield a highly processive enzyme only under conditions that are suboptimal for DNA synthetic rate. As incorporation of a nucleotide is generally more rapid than dissociation and reassociation of a DNA polymerase at the primer terminus (33), we might expect pol y to be most active under conditions that limit enzyme cycling, that is, under conditions favoring high processivity. Instead, our results suggest that initiation and/or termination of a processive product are even more rate-limiting.
Increasing concentrations of KC1 induced a 6-fold increase in the activity, but a 600-fold diminution in the processivity of Drosophila pol y. Ionic strength affects DNA and protein structure as well as molecular interactions, and thus it is likely to affect the binding of y polymerase to DNA. Because high processivity correlates with strong enzyme-primer terminus interactions, our data suggest that increasing ionic strength destabilizes them. This destabilization may allow rapid enzyme cycling, resulting in elevated activity if cycling is the rate-limiting step. In addition, the stabilization of DNA secondary structure at high ionic strength (36) probably contributes to the multitude of pause sites observed under these conditions. In contrast, at low ionic strength, the DNA polymerase is tightly bound to the primer terminus and meets fewer impediments on the template DNA, facilitating fully processive DNA synthesis. Here, strong enzyme-DNA associations may inhibit enzyme cycling, resulting in limited overall activity.
Alterations in pH resulted in differential activity optima for Drosophila pol y at low uers'sus moderate salt. However, the enzyme was most processive at pH 8.5 regardless of the KC1 level. The different activity optima likely reflect ionic strength effects on shielding of amino acid side groups. At moderate salt, it appears that side groups which are titratable in the range of pH 8.0 to pH 9.6 are sufficiently shielded such that the changes induced lack significant effects on the structure, and therefore the function, of pol y. On the other hand, at low salt, titration of side groups at more basic pH results in increased DNA polymerase activity. Here, the intrinsic charges of incompatible amino acid side chains may be altered sufficiently to allow their association, stabilizing enzyme structure and stimulating activity.
MgC12 decreased the processivity of the Drosophila mitochondrial DNA polymerase, yet the optimal Mg2+ levels for polymerase activity varied at low uers'sus moderate salt. Mg2' is bound by both DNA and DNA polymerases, and is a required cofactor for enzyme activity, as Mg-dNTP complexes are the substrates for nucleotide polymerization. MgClz also contributes to the ionic strength of the reaction solvent, thus partially explaining the increase in activity at high MgC12 levels at low salt, yet the loss of polymerase activity with the addition of MgC12 at moderate salt. However, this "salt" effect is probably not responsible for the initial peak in DNA polymerase activity observed between 0.25 and 0.5 mM MgC12 at low salt. Furthermore, although differential binding of M C by the DNA, dNTPs and pol y may result in a rapid initial increase in the rate of DNA synthesis, the presence of an approximately 20-fold excess of Mg2+ molecules over potential binding sites renders this potential explanation unlikely as well. Interestingly, E. coli DNA polymerase I possesses at least three types of divalent cation binding sites exhibiting varying affinities for manganese; the high and intermediate (3)(4)(5)(6)(7)(8)(9)(10) affinity sites appear to be stimulatory, while the low affinity sites (800-900 PM) seem to be inhibitory (37). Such a binding scenario involving y polymerase could explain the observed initial peak of polymerase activity. At the same time, diminished processivity may result from the increase in ionic strength due to addition of MgC12. Alternatively, the decrease may be related to the finding of Griep and McHenry (38) that the inclusion of Mg2+ alters the conformation of the processivity factor of E. coli DNA polymerase 111, the 37-kDa p subunit; the presence of 10 mM MgC12 specifically shifts the p subunit from a dimer into a predominantly monomeric form. While the dimeric form is directly implicated in processive DNA synthesis by pol I11 holoenzyme (26,39,40), the role of the monomer has not been addressed. Drosophila mitochondrial DNA polymerase comprises a 125-kDa polymerase catalytic subunit and a 35-kDa subunit of unknown function (22). A pol I11 p subunit-like activity that is affected similarly by Mg2+ may be present in the two subunit pol y.
High levels of ATP, but not GTP or ADP, were able to restore polymerase activity when Drosophila pol y was assayed at low salt. The specificity of this stimulation eliminates the possibility that the ionic strength contribution of added nucleotides alone is responsible. Nucleotide hydrolysis may be involved, as there is some stimulation of pol y activity by GTP but no effect of ADP. However, we were unable to detect any DNA-dependent ATPase activity in Drosophila pol y.' Lack of ATP hydrolysis may be indicative of the lack of a protein factor or factors which work(s) to chaperone the putative processivity factor onto the DNA (28, 41,42). Alternatively, ATP bound to pol y may induce a conformational change that influences polymerase catalytic efficiency. It is clear, although perhaps surprising, that stimulation by ATP is not a result of lowering the effective concentration of Mg2+ via chelation; when the ATP concentration is varied from 5 to 12 mM, the MgClz optimum remains at 4 mM.3 In addition, GTP would be expected to chelate the Mg2+ as well as ATP, yet its effect on enzyme activity is minimal.
The macromolecular crowding agent, polyethylene glycol, stimulated DNA polymerase activity when assayed at low salt, yet inhibited and then partially restored activity when assayed at moderate salt. Enhanced processivity was observed at both KC1 levels but was not coincident with optimal activity. These results may be complicated by the ability of PEG to precipitate protein, DNA, and protein-DNA complexes. However, PEG was shown to enhance the association of E. coli DNA polymerase I, its Klenow fragment, and phage T4 DNA polymerase with DNA (30), resulting in increased DNA polymerase activity. Stimulation was generally most pronounced at high ionic strength, such that the optimal salt concentration for enzyme activity increased with increasing PEG. Predictably, there was no apparent effect on DNA polymerase processivity since * C. M. Wernette protein-DNA association, but not dissociation, should be affected by crowding (43). In contrast, PEG enhanced the assembly of a holoenzyme form of phage T4 DNA polymerase and consequently, its processivity (31). Our data with pol y are consistent with either possibility; PEG may enhance binding of y polymerase to the DNA as a consequence of macromolecular crowding, and/or may promote protein-protein interactions (either polymerase-polymerase or polymerase-accessory factor). Glycerol may also promote these interactions, because its inclusion increased both DNA polymerase activity and processivity, regardless of salt concentration.
The processivity of Drosophila pol y was unchanged when the enzyme molecule to primer terminus ratio was varied over a 60-fold range, perhaps reflecting an inability to enhance protein-protein interactions by non-chemical means in the near-homogeneous enzyme. Alternatively, some associations that occur may not be detectable because they do not alter the processivity of mitochondrial DNA polymerase. Because a dissociable accessory factor(s) may be lacking in the nearhomogeneous pol y, the mechanism and efficiency of DNA synthesis by more crude DNA polymerase fractions were examined. We found that while the mechanism of DNA synthesis by pol y did not change after purification to the Fraction IV stage (8% of homogeneous), enzyme activity did; upon glycerol gradient sedimentation, the activity of pol y on singly primed DNA was decreased 4-fold, suggesting its separation from a dissociable factor present in the less pure fractions. Efforts are currently under way to identify this putative factor.
We have shown here that Drosophila y polymerase is capable of fully processive DNA synthesis on a natural DNA template in vitro only under conditions that are suboptimal for DNA polymerase activity. Thus, it is difficult to make a direct correlation of pol y function with current models of mitochondrial DNA replication or to suggest a mechanism for its regulation. The ATP/ADP ratio and the pH of the mitochondrial matrix have been implicated in the control of oxidative phosphorylation (44). Furthermore, the ATP/ADP ratio has been shown to be the primary regulator of the tricarboxylate cycle in insect flight muscle (45). The mechanism of control of DNA replication in the mitochondrion is unknown, but the ability of mitochondrial DNA polymerase to catalyze DNA synthesis under a wide range of experimental conditions suggests that replication can occur in actively respiring, as well as resting organelles. In contrast, only a narrow set of conditions facilitate processive DNA synthesis, possibly restricting continuous DNA synthesis to specific periods of mitochondrial activity. Although the aqueous i n vitro assay environment may be too unlike that of the mitochondrial matrix to draw accurate conclusions regarding the mechanism of pol y function, our data suggests that other protein factors involved in mitochondrial DNA replication may function to promote high processivity under conditions optimal for DNA synthetic rate. These protein factors may associate with pol y or, as suggested by Sabatino et al. (20), they may exert their effects by altering the microenvironment to which the DNA polymerase is exposed. Furthermore, using prokaryotic, viral, and eukaryotic nuclear systems as precedent, we would predict that mitochondrial DNA polymerase associates with several accessory proteins involved in primer recognition, processivity enhancement, and single-stranded DNA coating. We hope to gain insight into the function of such accessory proteins in mitochondrial DNA replication by reinvestigating the effects of the various reaction parameters examined here, on pol y in association with such factors, as they are identified.