The multiple biological roles of the 3'-->5' exonuclease of Saccharomyces cerevisiae DNA polymerase delta require switching between the polymerase and exonuclease domains

Until recently, the only biological function attributed to the 3'-->5' exonuclease activity of DNA polymerases was proofreading of replication errors. Based on genetic and biochemical analysis of the 3'-->5' exonuclease of yeast DNA polymerase delta (Pol delta) we have discerned additional biological roles for this exonuclease in Okazaki fragment maturation and mismatch repair. We asked whether Pol delta exonuclease performs all these biological functions in association with the replicative complex or as an exonuclease separate from the replicating holoenzyme. We have identified yeast Pol delta mutants at Leu523 that are defective in processive DNA synthesis when the rate of misincorporation is high because of a deoxynucleoside triphosphate (dNTP) imbalance. Yet the mutants retain robust 3'-->5' exonuclease activity. Based on biochemical studies, the mutant enzymes appear to be impaired in switching of the nascent 3' end between the polymerase and the exonuclease sites, resulting in severely impaired biological functions. Mutation rates and spectra and synergistic interactions of the pol3-L523X mutations with msh2, exo1, and rad27/fen1 defects were indistinguishable from those observed with previously studied exonuclease-defective mutants of the Pol delta. We conclude that the three biological functions of the 3'-->5' exonuclease addressed in this study are performed intramolecularly within the replicating holoenzyme.     

The Correcting Role of Autonomous 3" →="" 5"="" exonucleases="" contained="" in="" mammalian="" multienzyme="" dna="" polymerase="" complexes"="

A study was made of the correcting role of autonomous 3" 5" exonucleases (AE) contained in multienzyme DNA polymerase complexes of rat hepatocytes or calf thymocytes. DNA was synthesized on phage X174 amber3 or M13mp2 primer–templates, and used to transfect Escherichia coli spheroplasts. Frequencies were estimated for direct and reverse mutations resulting from mistakes made in the course of in vitro DNA synthesis. The error rate of the hepatocyte complex was estimated at 3·10^–6 with equimolar dNTP, and increased tenfold when proteins accounting for 70% of the total 3" 5" exonuclease activity of the complex were removed. The fidelity of DNA synthesis was completely restored in the presence of exogenous AE ( subunit of E. coli DNA polymerase III). Nuclear (Pol _n) and cytosolic (Pol _c) forms of DNA polymerase were isolated from calf thymocytes. The former was shown to contain an AE (TREX2) absent from the latter. As compared with Pol _c, Pol _n had a 20-fold higher exo/pol ratio and allowed 4–5 times higher fidelity of DNA synthesis. The error rate of DNA polymerase complexes changed when dNTP were used in nonequimolar amounts.     

Autonomous 3'-->5' exonucleases can proofread for DNA polymerase beta from rat liver

Autonomous 3'-->5'exonucleases are not bound covalently to DNA polymerases but are often involved in replicative complexes. Such exonucleases from rat liver, calf thymus and Escherichia coli (molecular masses of 28+/-2 kDa) are shown to increase more than 10-fold the accuracy of DNA polymerase beta (the most inaccurate mammalian polymerase) from rat liver in the course of reduplication of the primed DNA of bacteriophage phiX174 amber 3 in vitro. The extent of correction increases together with the rise in 3'-->5' exonuclease concentration. Extrapolation of the in vitro DNA replication fidelity to the cellular levels of rat exonuclease and beta-polymerase suggests that exonucleolytic proofreading could augment the accuracy of DNA synthesis by two orders of magnitude. These results are not explained by exonucleolytic degradation of the primers ("no synthesis-no errors"), since similar data are obtained with the use of the primers 15 or 150 nucleotides long in the course of a fidelity assay of DNA polymerases, both alpha and beta, in the presence of various concentrations of 3'-->5' exonuclease.     

PCNA-induced DNA synthesis past cis-syn and trans-syn-I thymine dimers by calf thymus DNA polymerase delta in vitro.

Calf thymus proliferating cell nuclear antigen (PCNA) promoted DNA synthesis past cis-syn and trans-syn-I cyclobutane thymine dimers by calf thymus DNA polymerase delta (Pol delta) in vitro. Templates containing site-specific cis-syn and trans-syn-I thymine dimers were prepared via a combination of solid phase synthesis with photoproduct building blocks and DNA ligation. Extension of a 15-mer primer on the UV dimer-containing templates by Pol delta produced termination and bypass products in a dNTP and PCNA dependent manner. In the absence of PCNA and at dNTP concentrations varying between 1 and 100 microM, Pol delta could not bypass the cis-syn dimer and terminated elongation one nucleotide prior to the 3'-T of the dimer. DNA synthesis past the trans-syn-I dimer was even less efficient. In the presence of PCNA, termination occurred primarily one nucleotide prior to the 3'-T of both dimers at 1 microM dNTPs but opposite the 5'-T of the dimers at 100 microM dNTPs. In addition, under the latter conditions, bypass of the dimers was observed, to the extent of about 30% of the products for the cis-syn dimer and about 15% for the trans-syn-I dimer.     

Antimutagenic role of autonomous 3'-->5'-exonucleases

Our own and literary data about antimutagenic role of autonomous 3'-->5'-exonucleases (AE) are analyzed. AE are not bound covalently to DNA polymerases but often involved in replicative complexes. Intracellular overproduction of AE in bacteria is accompanied with the sharp suppression of mutagenesis, whereas the inactivation of AE in bacteria and higher fungi results in the increase of mutation rates by 2-3 orders of magnitude. The addition of AE in biologically meaningful concentrations to DNA polymerases elevates substantially the accuracy of their work in vitro. In these cases, the reverse mutation rates were measured in the DNA from phage (X174 amber 3, whereas the direct mutation rates--in the DNA from phage M13mp2, both being used as primer-templates for DNA synthesis and then transfected into spheroplasts of Escherichia coli. The accuracy of action of nuclease-free DNA polymerases alpha and beta are shown to raise in the presence of AE by 2-3 orders, the accuracy of moderately processive DNA polymerase I--by 2 orders, the accuracy of highly processive DNA polymerase delta--by 5-10 times, though the latter 2 polymerases display and their own 3'-->5'-exonucleolytic activity. AE, involved in the multienzyme DNA polymerase complexes, augment the accuracy of complexes action by 5-10 times. The model of "external" corrective role of AE in DNA biosynthesis is proposed. Study of 30 objects from all 3 kingdoms of live beings (from archae- and eubacteria to mammalia including human) has shown that AE account, as minimum, from 30 to 90% of the total cellular 3'-->5'-exonucleolytic activity. So AE increase essentially the intracellular ratio of values of 3'-->5'-exonuclease to DNA polymerase activities in the very various representatives from a phylogenetic tree that results always in the augmentation of the accuracy of DNA biosynthesis.     

Characterization of two DNA polymerases from the hyperthermophilic euryarchaeon Pyrococcus abyssi.

The complete genome sequence of the hyperthermophilic archaeon Pyrococcus abyssi revealed the presence of a family B DNA polymerase (Pol I) and a family D DNA polymerase (Pol II). To extend our knowledge about euryarchaeal DNA polymerases, we cloned the genes encoding these two enzymes and expressed them in Escherichia coli. The DNA polymerases (Pol I and Pol II) were purified to homogeneity and characterized. Pol I had a molecular mass of approximately 90 kDa, as estimated by SDS/PAGE. The optimum pH and Mg(2+) concentration of Pol I were 8.5-9.0 and 3 mm, respectively. Pol II is composed of two subunits that are encoded by two genes arranged in tandem on the P. abyssi genome. We cloned these genes and purified the Pol II DNA polymerase from an E. coli strain coexpressing the cloned genes. The optimum pH and Mg(2+) concentration of Pol II were 6.5 and 15-20 mm, respectively. Both P. abyssi Pol I and Pol II have associated 3'-->5' exonuclease activity although the exonuclease motifs usually found in DNA polymerases are absent in the archaeal family D DNA polymerase sequences. Sequence analysis has revealed that the small subunit of family D DNA polymerase and the Mre11 nucleases belong to the calcineurin-like phosphoesterase superfamily and that residues involved in catalysis and metal coordination in the Mre11 nuclease three-dimensional structure are strictly conserved in both families. One hypothesis is that the phosphoesterase domain of the small subunit is responsible for the 3'-->5' exonuclease activity of family D DNA polymerase. These results increase our understanding of euryarchaeal DNA polymerases and are of importance to push forward the complete understanding of the DNA replication in P. abyssi.     

The 3'-->5' exonucleases of DNA polymerases delta and epsilon and the 5'-->3' exonuclease Exo1 have major roles in postreplication mutation avoidance in Saccharomyces cerevisiae

Replication fidelity is controlled by DNA polymerase proofreading and postreplication mismatch repair. We have genetically characterized the roles of the 5'-->3' Exo1 and the 3'-->5' DNA polymerase exonucleases in mismatch repair in the yeast Saccharomyces cerevisiae by using various genetic backgrounds and highly sensitive mutation detection systems that are based on long and short homonucleotide runs. Genetic interactions were examined among DNA polymerase epsilon (pol2-4) and delta (pol3-01) mutants defective in 3'-->5' proofreading exonuclease, mutants defective in the 5'-->3' exonuclease Exo1, and mismatch repair mutants (msh2, msh3, or msh6). These three exonucleases play an important role in mutation avoidance. Surprisingly, the mutation rate in an exo1 pol3-01 mutant was comparable to that in an msh2 pol3-01 mutant, suggesting that they participate directly in postreplication mismatch repair as well as in other DNA metabolic processes.     

Analogs of nucleotides, modified by a sugar residue and pyrimidine base, in a DNA synthesis reaction, catalyzed by Thermus aquaticus DNA polymerase]

Substrate properties of dNTP analogues in the DNA synthesis reaction catalyzed by Thermus aquaticus DNA polymerase were studied. It was shown that most of dNTP analogues which were known as terminators of DNA synthesis of E. coli DNA polymerase I were able to terminate DNA synthesis catalyzed by Thermus aquaticus DNA polymerase. An interesting feature of Thermus aquaticus DNA polymerase was the ability to utilize 3'-azido-2',3'-dideoxythymidine triphosphate as terminating substrate. Relative efficiency of tested dNTP analogues incorporation into the DNA growing chain was estimated.     

DNA polymerase mutagenic bypass and proofreading of endogenous DNA lesions

DNA polymerases differentiate between correct and incorrect substrates during synthesis on undamaged DNA templates through the biochemical steps of base incorporation, primer-template extension and proofreading excision. Recent research examining DNA polymerase processing of abasic, alkylation and oxidative lesions is reviewed in light of these discrimination mechanisms. Inhibition of DNA synthesis results from correct polymerase discrimination against utilization of geometrically incorrect template bases or 3' terminal basepairs. The efficiency of translesion synthesis is thus related to the physical structure of the lesion containing DNA. However, variations in enzyme structure and kinetics result in translesion synthesis efficiencies that are also dependent upon the DNA polymerase. With a low probability, polymerase misinsertion events create a 3' lesion terminus which is geometrically favored over the correct lesion basepair, resulting in mutagenic translesion synthesis. For example, both polymerase alpha and polymerase beta appear to require the formation of a stable 3' primer-template structure for efficient abasic site translesion synthesis. However, the enzymes differ as to the precise molecular make-up of the stable DNA structure, resulting in different mutational specificities. Similar mechanisms may be applicable to oxidative damage, where mutational specificities dependent upon the DNA polymerase also have been observed. In vitro reaction conditions also influence DNA polymerase processing of lesions. Using an in vitro herpes simplex virus thymidine kinase (HSV-tk) gene forward mutation assay, we demonstrate that high dNTP substrate concentrations affect the mutagenic specificity of translesion synthesis using alkylated templates. The exonuclease-deficient Klenow polymerase error frequency for G-->A transition mutations using templates modified by N-ethyl-N-nitrosourea (ENU) was four-fold higher at 1000 microM [dNTP], relative to 50 microM [dNTP], consistent with an increased efficiency of extension of the etO6G.T mispair. Moreover, the frequency of other ENU-induced polymerase errors was suppressed when polymerase reactions contained 50 microM dNTP, relative to 1000 microM dNTP. The efficiency of proofreading as a polymerase error discrimination mechanism reflects a balance between the competing processes of 3'-->5' exonuclease removal of mispairs and polymerization of the next correct nucleotide. Polymerases that are devoid of a proofreading exonuclease generally display enhanced abasic site translesion synthesis relative to proofreading-proficient enzymes. In addition, the proofreading exonucleases of Escherichia coli Pol I and T4 DNA polymerases have been found to remove mispairs caused by abasic sites and oxidative lesions, respectively, resulting in lowered polymerase error rates. However, the magnitude of the exonuclease effect is small (less than 10-fold), and highly dependent upon the DNA polymerase-exonuclease. We have studied proofreading exonuclease removal of alkylation damage in the HSV-tk forward assay. We observed no significant reduction in the magnitude of the mutant frequency vs. dose-response curves when N-methyl-N-nitrosourea or ENU-treated templates were used in exonuclease-proficient Klenow polymerase reactions, as compared to the exonuclease-deficient polymerase reactions. Thus, available data suggest that proofreading excision of endogenous lesion mispairs does occur, but the efficiency is dependent upon the lesion and the DNA polymerase-exonuclease studied.     

Selective inactivation of the 3'' to 5'' exonuclease activity of Escherichia coli DNA polymerase I by heat

Heat selectively inactivates the 3' to 5' exonuclease activity of E. coli DNA polymerase I, resulting in reduced dNTP turnover and lower fidelity of replication of homopolymer and natural DNA templates.     

Kinetic analysis of the unique error signature of human DNA polymerase ν

The fidelity of DNA synthesis by A-family DNA polymerases ranges from very accurate for bacterial, bacteriophage, and mitochondrial family members to very low for certain eukaryotic homologues. The latter include DNA polymerase ν (Pol ν) which, among all A-family polymerases, is uniquely prone to misincorporating dTTP opposite template G in a highly sequence-dependent manner. Here we present a kinetic analysis of this unusual error specificity, in four different sequence contexts and in comparison to Pol ν's more accurate A-family homologue, the Klenow fragment of Escherichia coli DNA polymerase I. The kinetic data strongly correlate with rates of stable misincorporation during gap-filling DNA synthesis. The lower fidelity of Pol ν compared to that of Klenow fragment can be attributed primarily to a much lower catalytic efficiency for correct dNTP incorporation, whereas both enzymes have similar kinetic parameters for G-dTTP misinsertion. The major contributor to sequence-dependent differences in Pol ν error rates is the reaction rate, k(pol). In the sequence context where fidelity is highest, k(pol) for correct G-dCTP incorporation by Pol ν is ~15-fold faster than k(pol) for G-dTTP misinsertion. However, in sequence contexts where the error rate is higher, k(pol) is the same for both correct and mismatched dNTPs, implying that the transition state does not provide additional discrimination against misinsertion. The results suggest that Pol ν may be fine-tuned to function when high enzyme activity is not a priority and may even be disadvantageous and that the relaxed active-site specificity toward the G-dTTP mispair may be associated with its cellular function(s).     

PCNA-dependent DNA polymerase delta from rabbit bone marrow.

Proliferating cell nuclear antigen (PCNA) and PCNA-dependent DNA polymerase delta were partially purified and characterized from rabbit bone marrow. Rabbit DNA polymerase delta sediments at 8.2 S upon glycerol density gradient centrifugation. Similar to calf thymus PCNA-dependent DNA polymerase delta, a 125-123-kDa doublet and 48-kDa polypeptides correlate with DNA polymerase activity. Western blotting of rabbit DNA polymerase delta with polyclonal antibody to calf thymus PCNA-dependent DNA polymerase delta gives the same results as calf thymus delta; the 125-123-kDa doublet is recognized. PCNA-dependent DNA polymerase delta is resistant to inhibition by dideoxynucleotides and is relatively insensitive to inhibition by N2-[p-(n-butyl)phenyl]dGTP. A 3'-->5' exonuclease copurifies with the DNA polymerase. The processivity of DNA polymerase delta alone is very low but greatly increases with the addition of PCNA from rabbit bone marrow or calf thymus. Comparative studies of the original DNA polymerase delta from rabbit bone marrow demonstrate a lack of recognition by antibodies to calf thymus delta and a high degree of processivity in the absence of PCNA. Additionally, the originally described DNA polymerase delta is a single polypeptide of 122 kDa. These features would recategorize the original delta to the epsilon category by recently proposed convention. PCNA-dependent DNA polymerase delta is a relatively minor component of rabbit bone marrow compared to DNA polymerase alpha and PCNA-independent DNA polymerase delta (epsilon), the relative proportions being alpha, 60%; delta, 7%; and epsilon, 30%.     

Localization of the exonuclease and polymerase domains of Bacillus subtilis DNA polymerase III.

Structural gene mutants were cloned and exploited to identify the major catalytic domains of Bacillus subtilis DNA polymerase III (BsPolIII), a 162.4-kDa [1437 amino acids (aa)] polymerase: 3'-5' exonuclease (Exo) required for replicative DNA synthesis. Analysis of the sequence, mutagenicity, and catalytic behavior of natural and site-directed point mutants of BsPolIII unequivocally located the domain involved in exonuclease catalysis within a 155-aa residue segment displaying homology with the Exo domain of Escherichia coli DNA polymerase I. Sequence analysis of four structural gene mutations which specifically alter then enzyme's reactivity to the inhibitory dGTP analog, 6-(p-hydroxyphenylhydrazino)uracil, and the inhibitory arabinonucleotide, araCTP, defined a domain (Pol) involved in dNTP binding. The Pol domain was in the C-terminal fourth of the enzyme within a 98-aa segment spanning aa 1175-1273. The primary structure of the domain was unique, displaying no obvious conservation in any other DNA polymerase, including the distantly related PolIIIs of the Gram- organisms, E. coli and Salmonella typhimurium.     

Efficiency and fidelity of human DNA polymerases λ and β during gap-filling DNA synthesis

The base excision repair (BER) pathway coordinates the replacement of 1-10 nucleotides at sites of single-base lesions. This process generates DNA substrates with various gap sizes which can alter the catalytic efficiency and fidelity of a DNA polymerase during gap-filling DNA synthesis. Here, we quantitatively determined the substrate specificity and base substitution fidelity of human DNA polymerase λ (Pol λ), an enzyme proposed to support the known BER DNA polymerase β (Pol β), as it filled 1-10-nucleotide gaps at 1-nucleotide intervals. Pol λ incorporated a correct nucleotide with relatively high efficiency until the gap size exceeded 9 nucleotides. Unlike Pol λ, Pol β did not have an absolute threshold on gap size as the catalytic efficiency for a correct dNTP gradually decreased as the gap size increased from 2 to 10 nucleotides and then recovered for non-gapped DNA. Surprisingly, an increase in gap size resulted in lower polymerase fidelity for Pol λ, and this downregulation of fidelity was controlled by its non-enzymatic N-terminal domains. Overall, Pol λ was up to 160-fold more error-prone than Pol β, thereby suggesting Pol λ would be more mutagenic during long gap-filling DNA synthesis. In addition, dCTP was the preferred misincorporation for Pol λ and its N-terminal domain truncation mutants. This nucleotide preference was shown to be dependent upon the identity of the adjacent 5'-template base. Our results suggested that both Pol λ and Pol β would catalyze nucleotide incorporation with the highest combination of efficiency and accuracy when the DNA substrate contains a single-nucleotide gap. Thus, Pol λ, like Pol β, is better suited to catalyze gap-filling DNA synthesis during short-patch BER in vivo, although, Pol λ may play a role in long-patch BER.     

The frameshift infidelity of human DNA polymerase lambda. Implications for function

DNA polymerase lambda (Pol lambda) is a member of the Pol X family having properties in common with several other mammalian DNA polymerases. To obtain clues to possible functions in vivo, we have determined the fidelity of DNA synthesis by human Pol lambda. The results indicate that the average single-base deletion error rate of Pol lambda is higher than those of other mammalian polymerases. In fact, unlike other DNA polymerases, Pol lambda generates single-base deletions at average rates that substantially exceed base substitution rates. Moreover, the sequence specificity for single-base deletions made by Pol lambda is different from that of other DNA polymerases and reveals that Pol lambda readily uses template-primers with limited base pair homology at the primer terminus. This ability, together with an ability to fill short gaps in DNA at low dNTP concentrations, is consistent with a role for mammalian Pol lambda in non-homologous end-joining. This may include non-homologous end-joining of strand breaks resulting from DNA damage, because Pol lambda has intrinsic 5',2'-deoxyribose-5-phosphate lyase activity.     

The 3'-5' exonuclease of human DNA polymerase delta (pol delta) is regulated by pol delta accessory factors and deoxyribonucleoside triphosphates.

Human DNA polymerase delta (pol delta) is required for the synthesis of leading strand of simian virus 40 (SV40) DNA replication in vitro. Pol delta requires the accessory factors, proliferating cell nuclear antigen (PCNA), activator 1 (A1; also known as replication factor C [RF-C]), human single-stranded DNA binding protein (HSSB; also known as replication protein A [RP-A]) for the elongation of primed template DNA. Since pol delta has an associated 3'-5' exonuclease activity, the effect of pol delta accessory factors on the exonuclease activity was examined. The 3'-5' exonuclease activity was stimulated 8-10 fold by the addition of HSSB, and this stimulatory effect was preferential to HSSB since other SSBs from E. coli, T4 or adenovirus, had a little or no effect. The stimulatory effect of HSSB was markedly inhibited by the combined action of A1 and PCNA. Furthermore, the addition of deoxyribonucleoside triphosphates (dNTPs) completely abolished the effect of HSSB on the 3'-5' exonuclease activity even in the absence of pol delta accessory factors. These results suggest that accessory factors and dNTPs regulate both the polymerase and the 3'-5' exonuclease activities.     

Functional interdependence of DNA polymerizing and 3'-->5' exonucleolytic activities in Pyrococcus furiosus DNA polymerase I

Pyrococcus furiosus DNA polymerase I (Pol BI) belongs to the family B (alpha-like) DNA polymerases and has a strong 3'-->5' exonucleolytic activity, in addition to its DNA polymerizing activity. To understand the relationship between the structure and function of this DNA polymerase, three deletion mutants, Delta1 (DeltaLeu746-Ser775), Delta2 (DeltaLeu717-Ser775) and Delta3 (DeltaHis672-Ser775), and two substituted mutants of Asp405, D405A and D405E, were constructed. These substitutions affected both the DNA polymerizing and the 3'-->5' exonucleolytic activities. The Delta1 mutant protein had DNA polymerizing activity with higher specific activity than that of the wild-type Pol BI, but retained only 10% of the exonucleolytic activity of the wild-type. The other two deletion mutants lost most of both activities. These results suggest that the DNA polymerizing and exonucleolytic activities are closely related to each other in the folded structure of this DNA polymerase, as proposed in the family B DNA polymerases.     

A 70-kDa chloroplast DNA polymerase from pea (<Emphasis Type="Italic">Pisum sativum</Emphasis>) that shows high processivity and displays moderate fidelity

A 70-kDa chloroplast (ct) DNA polymerase from pea has been purified to apparent homogeneity. The ct DNA polymerase was insensitive to dideoxynucleotides (d(2) NTP) but showed high sensitivity to phosphonoacetic acid. The enzyme lacked any detectable 5'-->3' exonuclease activity but showed 3'-->5' exonuclease activity. The polymerase displayed high processivity (3 kb) and moderate fidelity, which may be sufficient for the faithful replication of the 140-kb pea ct genome. A 43-kDa accessory protein increased the polymerization rate but did not affect the rate of mis-incorporation in vitro, thus indicating that the domains for polymerisation and proof reading may be spatially separate.     

Exonucleolytic proofreading enhances the fidelity of DNA synthesis by chick embryo DNA polymerase-gamma

The high fidelity of chick embryo DNA polymerase-gamma (pol-gamma) observed during in vitro DNA synthesis (Kunkel, T. A. (1985) J. Biol. Chem. 260, 12866-12874) has led us to examine this DNA polymerase for the presence of an exonuclease activity capable of proofreading errors. Highly purified chick embryo pol-gamma preparations do contain exonuclease activity capable of digesting radiolabeled DNA in a 3'----5' direction, releasing deoxynucleoside 5'-monophosphates. The polymerase and exonuclease activities cosediment during centrifugation in a glycerol gradient containing 0.5 M KCl. In the absence of dNTP substrates, this exonuclease excises both matched and mismatched primer termini, with a preference for mismatched bases. Excision is inhibited by the addition of nucleoside 5'-monophosphates to the digestion reaction. In the presence of dNTP substrates to permit competition between excision and polymerization from the mismatched primer, the exonuclease excises mismatched bases from preformed terminal mispairs with greater than 98% efficiency. The preference for excision over polymerization can be diminished by addition of either high concentrations of dNTP substrates or nucleoside 5'-monophosphates to the exonuclease/polymerase reaction. To determine if this exonuclease is capable of proofreading misinsertions produced during a normal polymerization reaction, a sensitive base substitution fidelity assay was developed based on reversion of an M13mp2 lacZ alpha nonsense codon. In this assay using reaction conditions that permit highly active exonucleolytic proofreading, pol-gamma exhibits a fidelity of less than one error for every 260,000 bases polymerized. As for terminal mismatch excision, fidelity is reduced by the addition to the synthesis reaction of high concentrations of dNTP substrates or nucleoside 5'-monophosphates, both hallmarks of exonucleolytic proofreading by prokaryotic enzymes. Taken together, these observations suggest that the 3'----5' exonuclease present in highly purified chick embryo pol-gamma preparations proofreads base substitution errors during DNA synthesis. It remains to be determined if the polymerase and exonuclease activities reside in the same or different polypeptides.