The base substitution fidelity of eucaryotic DNA polymerases. Mispairing frequencies, site preferences, insertion preferences, and base substitution by dislocation

The base substitution fidelity of DNA polymerase-alpha, -beta, and -gamma (pol-alpha, -beta, and -gamma, respectively) has been determined in vitro, for all 12 possible mispairs at 96 sites in a forward mutational target. Averaging all errors over all known detectable sites, pol-gamma is the most accurate enzyme, producing one error for every 10,000 bases polymerized. Pol-beta is much less accurate, with an error rate of 1/1,500, while pol-alpha has an intermediate accuracy of 1/4,000. The relative differences in fidelity between the DNA polymerases are strongly influenced by the nature of the mispair. For example, G(template):dATP mispairs and G:dGTP mispairs are formed with about equal frequency by all three classes of DNA polymerases, yet pol-gamma produces T:dGTP mispairs at a 100-fold lower frequency than does pol-beta. The DNA polymerases exhibit distinct differences in template site preferences as well as substrate insertion preferences. The increase in accuracy apparent in proceeding from the least selective to the most accurate enzyme results primarily from a decrease in mispair formations at template A and T residues and a decrease in misinsertion of pyrimidine deoxynucleotides. These data clearly demonstrate a major role for eucaryotic DNA polymerases in modulating base mispair frequencies at the level of insertion. In addition to direct mispair formation due to an incorrect incorporation event, an examination of the errors produced by each of the three classes of DNA polymerases at two particular sites in the target sequence suggests that some base substitution errors result from transient misalignment of the primer-template. A model is presented to explain this phenomenon, termed "Dislocation Mutagenesis." The data are discussed in relation to the extensive literature on base substitution errors and to the origin of spontaneous base substitutions in animal cells.     

The mutational specificity of DNA polymerases-alpha and -gamma during in vitro DNA synthesis

The frequency and specificity of mutations produced during in vitro DNA synthesis of the lacZ alpha gene in M13mp2 DNA by eucaryotic DNA polymerase-alpha (pol-alpha) and DNA polymerase-gamma (pol-gamma) have been determined. Pol-alpha, purified from five different sources, produces mutations resulting in loss of alpha-complementation at a frequency of 0.8-1.6%/single round of gap-filling DNA synthesis. DNA sequence analysis of 420 independent mutants produced by pol-alpha demonstrates three classes of errors. The majority of mutations result from single base substitutions, while single base frameshifts are detected at a lower but substantial frequency. Large deletions are also observed, with a frequency and specificity suggesting that they too are produced by pol-alpha in vitro. In contrast, pol-gamma is more accurate, producing mutants at a frequency of 0.3-0.5%. The specificity of pol-gamma errors is also different, since more than 90% of the mutants result from single base substitutions, while frameshift errors are not observed at a frequency significantly above background. The pol-gamma mutant spectrum also contains deletion mutations (10 of 179 mutants) presumably resulting from aberrant in vitro synthesis. When considered together with previous results using pol-beta (Kunkel, T. A. (1985) J. Biol. Chem. 260, 5787-5796) the relative accuracy of the three classes of purified vertebrate DNA polymerases for base substitutions, frameshifts, and deletions is in the order gamma greater than alpha greater than beta. These data demonstrate a correlation between the accuracy and processivity of DNA polymerization. Thus, the most accurate DNA polymerase (pol-gamma) also incorporates the most nucleotides per association with the primer-template, while the least accurate enzyme (pol-beta) is the least processive. This correlation exists both for base substitution mutations and for single base frameshifts, and is most obvious for minus-one-base frameshifts in runs of pyrimidines. In support of this correlation, increasing the processivity of pol-beta from 1 to 4-6 incorporations per association increases the accuracy of in vitro DNA synthesis by severalfold. The data imply that the processivity of DNA synthesis could be an important factor in controlling the levels of spontaneous and perhaps induced mutations.     

The fidelity of DNA synthesis catalyzed by derivatives of Escherichia coli DNA polymerase I

The fidelity of DNA synthesis by an exonuclease-proficient DNA polymerase results from the selectivity of the polymerization reaction and from exonucleolytic proofreading. We have examined the contribution of these two steps to the fidelity of DNA synthesis catalyzed by the large Klenow fragment of Escherichia coli DNA polymerase I, using enzymes engineered by site-directed mutagenesis to inactivate the proofreading exonuclease. Measurements with two mutant Klenow polymerases lacking exonuclease activity but retaining normal polymerase activity and protein structure demonstrate that the base substitution fidelity of polymerization averages one error for each 10,000 to 40,000 bases polymerized, and can vary more than 30-fold depending on the mispair and its position. Steady-state enzyme kinetic measurements of selectivity at the initial insertion step by the exonuclease-deficient polymerase demonstrate differences in both the Km and the Vmax for incorrect versus correct nucleotides. Exonucleolytic proofreading by the wild-type enzyme improves the average base substitution fidelity by 4- to 7-fold, reflecting efficient proofreading of some mispairs and less efficient proofreading of others. The wild-type polymerase is highly accurate for -1 base frameshift errors, with an error rate of less than or equal to 10(-6). The exonuclease-deficient polymerase is less accurate, suggesting that proofreading also enhances frameshift fidelity. Even without a proofreading exonuclease, Klenow polymerase has high frameshift fidelity relative to several other DNA polymerases, including eucaryotic DNA polymerase-alpha, an exonuclease-deficient, 4-subunit complex whose catalytic subunit is almost three times larger. The Klenow polymerase has a large (46 kDa) domain containing the polymerase active site and a smaller (22 kDa) domain containing the active site for the 3'----5' exonuclease. Upon removal of the small domain, the large polymerase domain has altered base substitution error specificity when compared to the two-domain but exonuclease-deficient enzyme. It is also less accurate for -1 base errors at reiterated template nucleotides and for a 276-nucleotide deletion error. Thus, removal of a protein domain of a DNA polymerase can affect its fidelity.     

High fidelity DNA synthesis by the Thermus aquaticus DNA polymerase.

We demonstrate that despite lacking a 3'----5' proofreading exonuclease, the Thermus aquaticus (Taq) DNA polymerase can catalyze highly accurate DNA synthesis in vitro. Under defined reaction conditions, the error rate per nucleotide polymerized at 70 degrees C can be as low as 10(-5) for base substitution errors and 10(-6) for frameshift errors. The frequency of mutations produced during a single round of DNA synthesis of the lac Z alpha gene by Taq polymerase responds to changes in dNTP concentration, pH, and the concentration of MgCl2 relative to the total concentration of deoxynucleotide triphosphates present in the reaction. Both base substitution and frameshift error rates of less than 1/100,000 were observed at pH 5-6 (70 degrees C) or when MgCl2 and deoxynucleotide triphosphates were present at equimolar concentrations. These high fidelity reaction conditions for DNA synthesis by the Taq polymerase may be useful for specialized uses of DNA amplified by the polymerase chain reaction.     

Fidelity of mammalian DNA replication and replicative DNA polymerases.

Current models suggest that two or more DNA polymerases may be required for high-fidelity semiconservative DNA replication in eukaryotic cells. In the present study, we directly compare the fidelity of SV40 origin-dependent DNA replication in human cell extracts to the fidelity of mammalian DNA polymerases alpha, delta, and epsilon using lacZ alpha of M13mp2 as a reporter gene. Their fidelity, in decreasing order, is replication greater than or equal to pol epsilon greater than pol delta greater than pol alpha. DNA sequence analysis of mutants derived from extract reactions suggests that replication is accurate when considering single-base substitutions, single-base frameshifts, and larger deletions. The exonuclease-containing calf thymus DNA polymerase epsilon is also highly accurate. When high concentrations of deoxynucleoside triphosphates and deoxyguanosine monophosphate are included in the pol epsilon reaction, both base substitution and frameshift error rates increase. This response suggests that exonucleolytic proofreading contributes to the high base substitution and frameshift fidelity. Exonuclease-containing calf thymus DNA polymerase delta, which requires proliferating cell nuclear antigen for efficient synthesis, is significantly less accurate than pol epsilon. In contrast to pol epsilon, pol delta generates errors during synthesis at a relatively modest concentration of deoxynucleoside triphosphates (100 microM), and the error rate did not increase upon addition of adenosine monophosphate. Thus, we are as yet unable to demonstrate that exonucleolytic proofreading contributes to accuracy during synthesis by DNA polymerase delta. The four-subunit DNA polymerase alpha-primase complex from both HeLa cells and calf thymus is the least accurate replicative polymerase. Fidelity is similar whether the enzyme is assayed immediately after purification or after being stored frozen.(ABSTRACT TRUNCATED AT 250 WORDS)     

Fidelity of DNA polymerase I and the DNA polymerase I-DNA primase complex from Saccharomyces cerevisiae.

We have determined the fidelity of DNA synthesis by DNA polymerase I (yPol I) from Saccharomyces cerevisiae. To determine whether subunits other than the polymerase catalytic subunit influence fidelity, we measured the accuracy of yPol I purified by conventional procedures, which yields DNA polymerase with a partially proteolyzed catalytic subunit and no associated primase activity, and that of yPol I purified by immunoaffinity chromatography, which yields polymerase having a single high-molecular-weight species of the catalytic subunit, as well as three additional polypeptides and DNA primase activity. In assays that score polymerase errors within the lacZ alpha-complementation gene in M13mp2 DNA, yPol I and the yPol I-primase complex produced single-base substitutions, single-base frameshifts, and larger deletions. For specific errors and template positions, the two forms of polymerase exhibited differences in fidelity that could be as large as 10-fold. Nevertheless, results for the overall error frequency and the spectrum of errors suggest that the yPol I-DNA primase complex is not highly accurate and that, just as for the polymerase alone, its fidelity is not sufficient to account for a low spontaneous mutation rate in vivo. The specificity data also suggest models to explain -1 base frameshifts in nonrepeated sequences and certain complex deletions by a direct repeat mechanism involving aberrant loop-back synthesis.     

Saccharomyces cerevisiae DNA polymerase delta: high fidelity for base substitutions but lower fidelity for single- and multi-base deletions

Eukaryotic DNA polymerase delta (Pol delta) plays an essential role in replicating large nuclear genomes, a process that must be accurate to maintain stability over many generations. Based on kinetic studies of insertion of individual dNTPs opposite a template guanine, Pol delta is believed to have high selectivity for inserting correct nucleotides. This high selectivity, in conjunction with an intrinsic 3'-exonuclease activity, implies that Pol delta should have high base substitution fidelity. Here we demonstrate that the wild type Saccharomyces cerevisiae three-subunit Pol delta does indeed have high base substitution fidelity for the 12 possible base-base mismatches, producing on average less than 1.3 stable misincorporations/100,000 nucleotides polymerized. Measurements with exonuclease-deficient Pol delta confirm the high nucleotide selectivity of the polymerase and further indicate that proofreading enhances the base substitution fidelity of the wild type enzyme by at least 60-fold. However, Pol delta inefficiently proofreads single nucleotide deletion mismatches in homopolymeric runs, such that the error rate is 30 single nucleotide deletions/100,000 nucleotides polymerized. Moreover, wild type Pol delta frequently deletes larger numbers of nucleotides between distantly spaced direct repeats of three or more base pairs. Although wild type Pol delta and Pol epsilon both have high base substitution fidelity, Pol delta is much less accurate than Pol epsilon for deletions involving repetitive sequences. Thus, strand slippage during replication by wild type Pol delta may be a primary source of insertion and deletion mutagenesis in eukaryotic genomes.     

The fidelity of human DNA polymerase gamma with and without exonucleolytic proofreading and the p55 accessory subunit

Mutations in human mitochondrial DNA influence aging, induce severe neuromuscular pathologies, cause maternally inherited metabolic diseases, and suppress apoptosis. Since the genetic stability of mitochondrial DNA depends on the accuracy of DNA polymerase gamma (pol gamma), we investigated the fidelity of DNA synthesis by human pol gamma. Comparison of the wild-type 140-kDa catalytic subunit to its exonuclease-deficient derivative indicates pol gamma has high base substitution fidelity that results from high nucleotide selectivity and exonucleolytic proofreading. pol gamma is also relatively accurate for single-base additions and deletions in non-iterated and short repetitive sequences. However, when copying homopolymeric sequences longer than four nucleotides, pol gamma has low frameshift fidelity and also generates base substitutions inferred to result from a primer dislocation mechanism. The ability of pol gamma both to make and to proofread dislocation intermediates is the first such evidence for a family A polymerase. Including the p55 accessory subunit, which confers processivity to the pol gamma catalytic subunit, decreases frameshift and base substitution fidelity. Kinetic analyses indicate that p55 promotes extension of mismatched termini to lower the fidelity. These data suggest that homopolymeric runs in mitochondrial DNA may be particularly prone to frameshift mutation in vivo due to replication errors by pol gamma.     

Distinctive properties of mammalian DNA polymerases.

DNA polymerase-alpha and -beta can be distinguished from one another by the differential effects of N-ethylmaleimide, KCl, ara-CTP and temperature, as well as on the basis of sedimentation. The sensitivity of DNA polymerase-beta to elevated temperatures as compared to DNA polymerase-alpha provides a new means of distinguishing between these two enzymes even in crude extracts and a possible probe for determining their function. DNA polymerase-alpha and -beta share several properties in common, including the ability to readily incorporate dUTP in place of dTTP. The Km for dUTP varies from 10 to 30 micron with different preparations of DNA polymerase-alpha and -beta. Thus, in mammalian cells, dUMP could be incorporated into DNA, and if excised by an endonuclease, would lead to discontinuities. Initial analyses of fidelity in direct comparative studies indicate that beta-class DNA polymerases are highly accurate in base selection when copying poly[d(A-T)]. Less than one molecule of dGMP is incorporated for every 12 000-45 000 molecules of dAMP and dTMP polymerized. DNA polymerase-alpha is somewhat less accurate, making one mistake for every 4000-10 000 correct nucleotides incorporated. Since both polymerases lack an exonucleolytic activity, this accuracy must be the result of selectivity for the complementary nucleotide by the polymerase.     

Low fidelity DNA synthesis by a y family DNA polymerase due to misalignment in the active site

Sulfolobus solfataricus DNA polymerase IV (Dpo4) is a member of the Y family of DNA polymerases whose crystal structure has recently been solved. As a model for other evolutionarily conserved Y family members that perform translesion DNA synthesis and have low fidelity, we describe here the base substitution and frameshift fidelity of DNA synthesis by Dpo4. Dpo4 generates all 12 base-base mismatches at high rates, 11 of which are similar to those of its human homolog, DNA polymerase kappa. This result is consistent with the Dpo4 structure, implying lower geometric selection for correct base pairs. Surprisingly, Dpo4 generates C.dCMP mismatches at an unusually high average rate and preferentially at cytosine flanked by 5'-template guanine. Dpo4 also has very low frameshift fidelity and frequently generates deletions of even noniterated nucleotides, especially cytosine flanked by a 5'-template guanine. Both unusual features of error specificity suggest that Dpo4 can incorporate dNTP precursors when two template nucleotides are present in the active site binding pocket. These results have implications for mutagenesis resulting from DNA synthesis by Y family polymerases.     

The mutational specificity of DNA polymerase-beta during in vitro DNA synthesis. Production of frameshift, base substitution, and deletion mutations

The frequency and specificity of mutations produced in vitro by eucaryotic DNA polymerase-beta have been determined in a forward mutation assay using a 250-base target sequence in M13mp2 DNA. Homogeneous DNA polymerase-beta, isolated from four different sources, produces mutations at a frequency of 4-6%/single round of gap-filling DNA synthesis. DNA sequence analyses of 460 independent mutants resulting from this error-prone DNA synthesis demonstrate a wide variety of mutational events. Frameshift and base substitutions are made at approximately equal frequency and together comprise about 90% of all mutations. Two mutational "hot spots" for frameshift and base substitution mutations were observed. The characteristics of the mutations at these sites suggest that certain base substitution errors result from dislocation of template bases rather than from direct mispair formation by DNA polymerase-beta. When considering the entire target sequence, single-base frameshift mutations occur primarily in runs of identical bases, usually pyrimidines. The loss of a single base occurs 20-80 times more frequently than single-base additions and much more frequently than the loss of two or more bases. Base substitutions occur at many sites throughout the target, representing a wide spectrum of mispair formations. Averaged over a large number of phenotypically detectable sites, the base substitution error frequency is greater than one mistake for every 5000 bases polymerized. Large deletion mutations are also observed, at a frequency more than 10-fold over background, indicating that purified DNA polymerases alone are capable of producing such deletions. These data are discussed in relation to the physical and kinetic properties of the purified enzymes and with respect to the proposed role for this DNA polymerase in vivo.     

Mutations in DNA polymerase gamma cause error prone DNA synthesis in human mitochondrial disorders

This paper summarizes recent advances in understanding the links between the cell's ability to maintain integrity of its mitochondrial genome and mitochondrial genetic diseases. Human mitochondrial DNA is replicated by the two-subunit DNA polymerase gamma (polgamma). We investigated the fidelity of DNA replication by polgamma with and without exonucleolytic proofreading and its p55 accessory subunit. Polgamma has high base substitution fidelity due to efficient base selection and exonucleolytic proofreading, but low frameshift fidelity when copying homopolymeric sequences longer than four nucleotides. Progressive external ophthalmoplegia (PEO) is a rare disease characterized by the accumulation of large deletions in mitochondrial DNA. Recently, several mutations in the polymerase and exonuclease domains of the human polgamma have been shown to be associated with PEO. We are analyzing the effect of these mutations on the human polgamma enzyme. In particular, three autosomal dominant mutations alter amino acids located within polymerase motif B of polgamma. These residues are highly conserved among family A DNA polymerases, which include T7 DNA polymerase and E.coli pol I. These PEO mutations have been generated in polgamma to analyze their effects on overall polymerase function as well as the effects on the fidelity of DNA synthesis. One mutation in particular, Y955C, was found in several families throughout Europe, including one Belgian family and five unrelated Italian families. The Y955C mutant polgamma retains a wild-type catalytic rate but suffers a 45-fold decrease in apparent binding affinity for the incoming dNTP. The Y955C derivative is also much less accurate than is wild-type polgamma, with error rates for certain mismatches elevated by 10- to 100-fold. The error prone DNA synthesis observed for the Y955C polgamma is consistent with the accumulation of mtDNA mutations in patients with PEO. The effects of other polgamma mutations associated with PEO are discussed.     

On the fidelity of DNA replication. Studies with human placenta DNA polymerases.

The fidelity of DNA synthesis with purified DNA polymerase alpha and beta from human placenta has been studied. With poly[d(A-T)] as the template-primer and Mg2+ as the metal activator, DNA polymerase alpha incorporates 1 mol of dGMP for every 6,000 to 12,000 mol of complementary nucleotides polymerized. Under the same conditions, DNA polymerase beta is more accurate, the error rate being 1/20,000 to 1/60,000. This greater accuracy of DNA polymerase beta is observed with a variety of homopolymer templates. With both enzymes, substitution of Mg2+ with activating concentrations of Mn2+ or Co2+ enhances the frequency of misincorporation. At greater than activating concentrations of Mn2+ and Co2+, there is an inhibition of complementary nucleotide incorporation, further increasing the frequency of misincorporation. Nearest neighbor analysis of the products synthesized with both enzymes indicates that the noncomplementary nucleotides are incorporated predominantly as single base substitutions. The greater accuracy of DNA polymerase beta over DNA polymerase alpha should be considered in relationship to their possible roles in DNA replication and repair.     

Unique error signature of the four-subunit yeast DNA polymerase epsilon

We have purified wild type and exonuclease-deficient four-subunit DNA polymerase epsilon (Pol epsilon) complex from Saccharomyces cerevisiae and analyzed the fidelity of DNA synthesis by the two enzymes. Wild type Pol epsilon synthesizes DNA accurately, generating single-base substitutions and deletions at average error rates of 5' exonuclease activity is less accurate to a degree suggesting that wild type Pol epsilon proofreads at least 92% of base substitution errors and at least 99% of frameshift errors made by the polymerase. Surprisingly the base substitution fidelity of exonuclease-deficient Pol epsilon is severalfold lower than that of proofreading-deficient forms of other replicative polymerases. Moreover the spectrum of errors shows a feature not seen with other A, B, C, or X family polymerases: a high proportion of transversions resulting from T.dTTP, T.dCTP, and C.dTTP mispairs. This unique error specificity and amino acid sequence alignments suggest that the structure of the polymerase active site of Pol epsilon differs from those of other B family members. We observed both similarities and differences between the spectrum of substitutions generated by proofreading-deficient Pol epsilon in vitro and substitutions occurring in vivo in a yeast strain defective in Pol epsilon proofreading and DNA mismatch repair. We discuss the implications of these findings for the role of Pol epsilon polymerase activity in DNA replication.     

Kinetic analysis of base substitution mutagenesis by transient misalignment of DNA and by miscoding

We measured the insertion fidelity of DNA polymerases alpha and beta and yeast DNA polymerase I at a template site that was previously observed to yield a high frequency of T----G transversions when copied by DNA polymerase beta but not by the other two polymerases. The results provide direct biochemical evidence that base substitution errors by DNA polymerase beta can result from a dislocation mechanism governed by DNA template-primer misalignment. In contrast to DNA polymerase beta, neither Drosophila DNA polymerase alpha nor yeast DNA polymerase I appear to misinsert nucleotides by a dislocation mechanism in either the genetic or kinetic fidelity assays. Dislocation errors by DNA polymerase beta are characterized primarily by a substantial reduction in the apparent Km for inserting a "correct," but ultimately errant, nucleotide compared to the apparent Km governing direct misinsertion. For synthesis by DNA polymerase beta, dislocation results in a 35-fold increase in dCMP incorporation opposite template T (T----G transversion) and a 20-35-fold increase in dTMP incorporation opposite T (T----A transversion); these results are consistent with parallel genetic fidelity measurements. DNA polymerase beta also produces base substitution errors by direct misinsertion. Here nucleotide insertion fidelity results from substantial differences in both Km and Vmax for correct versus incorrect substrates and is influenced strongly by local base sequence.     

The high fidelity and unique error signature of human DNA polymerase epsilon

Bulk replicative DNA synthesis in eukaryotes is highly accurate and efficient, primarily because of two DNA polymerases (Pols): Pols δ and ε. The high fidelity of these enzymes is due to their intrinsic base selectivity and proofreading exonuclease activity which, when coupled with post-replication mismatch repair, helps to maintain human mutation rates at less than one mutation per genome duplication. Conditions that reduce polymerase fidelity result in increased mutagenesis and can lead to cancer in mice. Whereas yeast Pol ε has been well characterized, human Pol ε remains poorly understood. Here, we present the first report on the fidelity of human Pol ε. We find that human Pol ε carries out DNA synthesis with high fidelity, even in the absence of its 3'→5' exonucleolytic proofreading and is significantly more accurate than yeast Pol ε. Though its spectrum of errors is similar to that of yeast Pol ε, there are several notable exceptions. These include a preference of the human enzyme for T→A over A→T transversions. As compared with other replicative DNA polymerases, human Pol ε is particularly accurate when copying homonucleotide runs of 4-5 bases. The base pair substitution specificity and high fidelity for frameshift errors observed for human Pol ε are distinct from the errors made by human Pol δ.     

Strand asymmetry of +1 frameshift mutagenesis at a homopolymeric run by DNA polymerase III holoenzyme of Escherichia coli

We have recently shown that single-base frameshifts were predominant among mutations induced within the rpsL target sequence upon oriC plasmid DNA replication in vitro. We found that the occurrence of +1 frameshifts at a run of 6 residues of dA/dT could be increased proportionally by increasing the concentration of dATP present in the in vitro replication. Using single-stranded circular DNA containing either the coding sequence of the rpsL gene or its complementary sequence, the +1 frameshift mutagenesis by DNA polymerase III holoenzyme of Escherichia coli was extensively examined. A(6) --> A(7) frameshifts occurred 30 to 90 times more frequently during DNA synthesis with the noncoding sequence (dT tract) template than with the coding sequence (dA tract). Excess dATP enhanced the occurrence of +1 frameshifts during DNA synthesis with the dT tract template, but no other dNTPs showed such an effect. In the presence of 0.1 mM dATP, the A(6) --> A(7) mutagenesis with the dT tract template was not inhibited by 1.5 mM dCTP, which is complementary to the residue immediately upstream of the dT tract. These results strongly suggested that the A(6) --> A(7) frameshift mutagenesis possesses an asymmetric strand nature and that slippage errors leading to the +1 frameshift are made during chain elongation within the tract rather than by misincorporation of nucleotides opposite residues next to the tract.     

Fidelity of DNA synthesis by the Thermus aquaticus DNA polymerase

We have determined the fidelity of in vitro DNA synthesis catalyzed at high temperature by the DNA polymerase from the thermophilic bacterium Thermus aquaticus. Using a DNA substrate that contains a 3'-OH terminal mismatch, we demonstrate that the purified polymerase lacks detectable exonucleolytic proofreading activity. The fidelity of the Taq polymerase was measured by two assays which score errors produced during in vitro DNA synthesis of the lacZ alpha complementation gene in M13mp2 DNA. In both assays, the Taq polymerase produces single-base substitution errors at a rate of 1 for each 9000 nucleotides polymerized. Frameshift errors are also produced, at a frequency of 1/41,000. These results are discussed in relation to the effects of high temperature on fidelity and the use of the Taq DNA polymerase as a reagent for the in vitro amplification of DNA by the polymerase chain reaction.