Fidelity and Processivity of Saccharomyces cerevisiae DNA Polymerase η*

The yeast RAD30 gene functions in error-free replication of UV-damaged DNA, and RAD30 encodes a DNA polymerase, pol η, that has the ability to efficiently and correctly replicate past a cis-syn-thymine-thymine dimer in template DNA. To better understand the role of pol η in damage bypass, we examined its fidelity and processivity on nondamaged DNA templates. Steady-state kinetic analyses of deoxynucleotide incorporation indicate that pol η has a low fidelity, misincorporating deoxynucleotides with a frequency of about 10−2 to 10−3. Also pol η has a low processivity, incorporating only a few nucleotides before dissociating. We suggest that pol η's low fidelity reflects a flexibility in its active site rendering it more tolerant of DNA damage, while its low processivity limits its activity to reduce errors.

correct bypass of a T-T dimer may derive from an unusual active site that is more tolerant of DNA distortions. In that case, pol would be expected to have a low fidelity (see also the "Discussion"). Here we determine the fidelity of pol by measuring the steady-state kinetics of correct and incorrect deoxynucleotide incorporation and also examine its processivity. We find that pol is a low fidelity polymerase, and it incorporates only a few nucleotides before dissociating from the primer-template DNA substrate. This low processivity may restrict pol synthesis to short patches, thereby minimizing the error frequency, and thus accounting for its role in error-free bypass of UV-damaged DNA.
Purification of Pol -DNA polymerase was expressed in yeast strain BJ5464 and purified as described previously (3,4).
Analysis of Fidelity-Analysis of the deoxynucleotide incorporation assays was done as described previously (12)(13)(14). Gel band intensities of the substrates and products were quantitated using a PhosphorImager and the ImageQuant software (Molecular Dynamics). For each concentration of dNTP, the observed rate of deoxynucleotide incorporation (V obs ) was determined by dividing the relative amount of the extended product by the incubation time. The observed rate of deoxynucleotide incorporation was plotted as a function of dNTP concentration, and the data were fit by nonlinear regression using SigmaPlot 4.0 to the Michaelis-Menton equation describing a hyperbola as follows (Equation 1).
Apparent K m and V max steady-state parameters for the incorporation of the correct and incorrect deoxynucleotides were obtained from the fit and used to calculate the frequency of deoxynucleotide misincorporation (f inc ) using the following equation (Equation 2).
Processivity Assay-Pol (20 nM) was preincubated with the primertemplate DNA substrate (20 nM) in 25 mM NaPO 4 , pH 7.0, 5 mM dithiothreitol, 100 g/ml bovine serum albumin, and 10% glycerol for 20 min at 25°C. Reactions were initiated by adding all four deoxynucleotides (200 M each), 5 mM MgCl 2 , and excess sonicated herring sperm DNA (1 mg/ml) as a trap. To demonstrate the effectiveness of the trap, pol was preincubuated with the trap DNA and the primer-template * This work was supported by National Institutes of Health Grant GM19261. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
substrate before the addition of dNTPs and MgCl 2 . After various times, reactions were quenched and run on a 10% polyacrylamide gel as described for the deoxynucleotide incorporation assays.
Analysis of Processivity-The processivity, P n , after each deoxynucleotide incorporation was calculated by a method derived from one previously described (15). Briefly, gel band intensities of the deoxynucleotide incorporation products at the 240 s incubation time were quantitated using the PhosphorImager and ImageQuant software. First, for each deoxynucleotide addition n, the percentage of active polymerase molecules incorporating at least n deoxynucleotides is given by the following equation (Equation 3), where I 1 is the intensity of the band at position 1, I n is the intensity of the band at position n, and so on. For each deoxynucleotide incorporation n, processivity, P n , is the probability that the polymerase will incorporate the next deoxynucleotide rather than dissociating (15) and is given by the following equation (Equation 4). P n ϭ % active polymerases at n ϩ 1/% active polymerases at n (Eq. 4)

Applying Equation 3 to Equation 4
gives an expression for P n in terms of gel band intensities, and this equation was used to calculate the processivity as follows (Equation 5).

RESULTS
Fidelity of Pol -Fidelity is a measure of the frequency of incorporating a correctly base-paired versus an incorrectly base-paired deoxynucleotide (7,12,13,16). To determine the frequency of misincorporation by pol , we measured the V max and K m steady-state parameters for the incorporation of correct and incorrect deoxynucleotides opposite each template residue, using a standing-start assay, wherein the target template residue immediately follows the end of the primer (13).
Pol (2 nM) was incubated with the primer-template DNA substrate (50 nM) and various concentrations of one of the four deoxynucleotides, after which reaction products were resolved by polyacrylamide gel electrophoresis and band intensities quantified. Fig. 1A shows the deoxynucleotide incorporation pattern opposite a template T residue. The concentrations of dGTP, dTTP, and dCTP were varied from 0 to 200 M, whereas the concentrations of dATP was varied from 0 to 10 M. Under the reaction conditions, the kinetics of deoxynucleotide incorporation were linear with time.
The kinetics of single deoxynucleotide incorporation opposite the template T residue are shown in Fig. 1B. These data were fit to the Michaelis-Menton equation (Equation 1) and used to determine the apparent K m and V max values for each deoxynucleotide (Table I). The frequency of misincorporation, f inc , of G, T, and C opposite the template residue T was then calculated using Equation 2 (12)(13)(14)17): f inc ϭ (V max /K m ) incorrect / (V max /K m ) correct. As shown in Table I, for the incorporation of the incorrect deoxynucleotide G opposite the template T, the V max was 0.89 nM/min and K m was 99 M, whereas for the incorporation of the correct deoxynucleotide A opposite the template T, the V max was 2.9 nM/min and the K m was 1.7 M. Thus f inc for G opposite T is 5.3 ϫ 10 Ϫ3 (Table I). Similarly, f inc of T and C opposite the template T residue were 8.8 ϫ 10 Ϫ3 and 6.5 ϫ 10 Ϫ3 , respectively (Table I).
The V max and K m steady-state parameters for the misincorporation of deoxynucleotides opposite the other three template bases and the calculated f inc values are also shown in Table I. The values for the f inc ranged from 3.1 ϫ 10 Ϫ4 for the misincorporation of A opposite a template G to 2.3 ϫ 10 Ϫ2 for the misincorporation of C opposite template C, and the average value for f inc was 6.0 ϫ 10 Ϫ3 .
Processivity of Pol -Processivity is a measure of how many deoxynucleotides a DNA polymerase incorporates in a single DNA binding event (15,18). To ensure that we were observing deoxynucleotide incorporation resulting from a single DNA binding event, we monitored DNA synthesis in the presence of an excess of nonradiolabeled, sonicated herring sperm DNA as a trap (Fig. 2A). The reactions in lanes 1-6 ( Fig. 2A) were performed by first preincubating pol with the DNA substrate for 20 min. Excess herring sperm DNA, MgCl 2 , and all four deoxynucleotides were then added to initiate the reaction. The excess herring sperm DNA is included to trap all pol molecules that dissociated from the substrate ensuring that all DNA synthesis resulted from a single DNA binding event. The reactions in lanes 7-12 ( Fig. 2A) were performed by first preincubating pol with the DNA substrate and the excess herring sperm DNA for 20 min followed by the addition of MgCl 2 and deoxynucleotides. The lack of DNA synthesis in these lanes shows that the excess herring sperm DNA is sufficient to trap all pol molecules. The processivity of a DNA polymerase is quantitatively expressed as the probability, P n , for each deoxynucleotide incorporation event n that the polymerase will move ahead to incorporate the next nucleotide n ϩ 1 rather than dissociate from  Table I, and these parameters were used to calculate the frequency of deoxynucleotide misincorporation, f inc .

Fidelity of Yeast Pol 36836
the DNA template (15). First, the percentage of active polymerase molecules incorporating at least n deoxynucleotides was calculated using Equation 3 (see "Materials and Methods").
The percentage of active polymerases adding at least one deoxynucleotide was set as 100%, and the percentage of active polymerases decreased after each subsequent addition because of the dissociation of some fraction of polymerase molecules. For example, 93% added at least two deoxynucleotides, 88% added at least three deoxynucleotides, 80% added at least four deoxynucleotides, and so on (Fig. 2B). On this DNA substrate, ϳ50% of the pol molecules incorporate at least six deoxynucleotides before dissociating from the DNA.
Next, we calculated the processivity P n after each deoxynucleotide incorporation using the following equation (Equation 5) (15): P n ϭ (I nϩ1 ϩ I nϩ2 ϩ . . . )/(I n ϩ I nϩ1 ϩ I nϩ2 ϩ . . . ), where I n is the intensity of band n, I nϩ1 is the intensity of band n ϩ 1, and so on. For example, of the polymerase molecules that incorporated one nucleotide, 93% incorporated at least one additional nucleotide. Thus, P 1 ϭ 0.93. The values of P n ranged from 0.94 in the case of n ϭ 2 deoxynucleotide additions to 0.40 in the case of n ϭ 7 deoxynucleotide additions with an average value of 0.76 Ϯ 0.20. Thus after each nucleotide incorporation, on average 76% of the bound pol molecules incorporate at least one additional deoxynucleotide, while 24% dissociate from the DNA substrate. Thus pol synthesizes DNA with low processivity.

DISCUSSION
Replicative DNA polymerases incorporate wrong deoxynucleotides with very low frequencies (7,12,13,16). This high fidelity has been suggested to arise in part because of the intolerance of the active site to geometric distortions in DNA (7). Structures of several DNA polymerases have indicated that a conformational change in the enzyme is critical for the formation of the phosphodiester bond between the primer and the incoming deoxynucleotide, and it plays an important role in fidelity (for a review, see Ref. 19). Furthermore, these structures are consistent with the proposal that the incoming deoxynucleotide initially binds in the active site of the polymerase in a manner independent of the template base, but precise Watson-Crick base pairing geometry between the incoming deoxynucleotide, and the template base is required for this catalytically essential conformational change to occur (19). Thus, if the geometry is not correct, the phosphodiester bond will not be formed efficiently.
Even though the thymine bases of a cyclobutane pyrimidine dimer can properly base pair with adenines (8,9), the distorted geometry of the dimer (10,11) presumably blocks polymerases, because they cannot tolerate the distortion (7). The ability of pol to efficiently and correctly bypass dimers would suggest that relative to other DNA polymerases, the active site of pol has an increased tolerance of DNA distortions.
The in vitro misinsertion frequencies, f inc , vary for different DNA polymerases. Using a steady-state kinetics assay, the error rates for T4 DNA polymerase and Escherichia coli DNA polymerase III holoenzyme, both replicative DNA polymerases, were found to vary from 10 Ϫ4 to 10 Ϫ7 (14,20). Using an approach involving the synthesis of DNA in a gap across from the ␣-complementation lacZ gene of M13mp2, the eukaryotic replicative DNA polymerase pol ␦ was found to have an error rate of about 10 Ϫ5 (21), whereas pol ␣, required for lagging strand DNA synthesis, and pol ␤, involved in short patch base excision repair, are less accurate and have an error rate of about 10 Ϫ3 to 10 Ϫ4 (21,22). Here, using the steady-state kinetics assay, we find pol to have an error rate of ϳ10 Ϫ2 to 10 Ϫ3 . Thus, relative to these other DNA polymerases, pol has a low fidelity, which may arise from an active site more tolerant of distortions in DNA. It would be of much interest to compare a high resolution structure of pol with the structures of other DNA polymerases to determine the structural basis of the flexibility of FIG. 2. Processivity of pol . A, processive DNA synthesis by pol resulting from a single DNA binding event. In lanes 1-6, pol (20 nM) was preincubated with the primer-template DNA (20 nM) for 20 min at 25°C, and the reactions were initiated by the addition of dNTPs (200 M each), MgCl 2 (5 mM), and the sonicated herring sperm DNA trap (1 mg/ml). Reactions were quenched after the indicated incubation times indicated in seconds (s), and products were resolved by denaturing polyacrylamide gel electrophoresis. The positions of the unextended primer (n ϭ 0) and the extended primers (n ϭ 1 to 10) are indicated. As a control (lanes 7-12), pol was preincubated with the primer-template DNA and the herring sperm DNA trap, and reactions were initiated by the addition of dNTPs and MgCl 2 . B, percentage of pol molecules incorporating at least n deoxynucleotides was calculated using Equation 3 (see "Materials and Methods").
pol 's active site that gives it a higher tolerance for distortions and a lower fidelity.
Despite its low fidelity, pol functions in error-free bypass of UV lesions. Both in yeast and humans, inactivation of pol enhances the frequency of UV-induced mutations (1,3,23,24), and as a consequence, XP-V patients suffer from a high incidence of skin cancers. Additonally, pol has little effect on spontaneous mutations, as the rate of spontaneous CAN1 s to can1 r forward mutations is not affected in the rad30⌬ strain. 2 Thus, there is no evidence that pol contributes significantly to the generation of mutations in vivo. This likely occurs because the DNA synthesis activity of pol is limited. Pol may synthesize just enough DNA to bypass lesions, thus affording little opportunity to make errors. Additionally, the activity of pol may be regulated by the Rad6-Rad18 complex (25), which may have a role in targeting pol to DNA damage sites and in ensuring that the action of pol is limited to damage bypass. Very likely, pol ␦ takes over soon after the damage has been bypassed by pol .