HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 10, 8235-8242, March 8, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/10/8235    most recent M107286200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Beard, W. A. Articles by Wilson, S. H. Search for Related Content PubMed PubMed Citation Articles by Beard, W. A. Articles by Wilson, S. H.

Loss of DNA Polymerase  Stacking Interactions with Templating Purines, but Not Pyrimidines, Alters Catalytic Efficiency and Fidelity^*

William A. Beard, David D. Shock, Xiao-Ping Yang, Saundra F. DeLauder, and Samuel H. Wilson^§

From the Laboratory of Structural Biology, NIEHS, National Institutes of Health, Research Triangle Park, North Carolina 27709 and  Department of Chemistry, North Carolina Central University, Durham, North Carolina 27707

Received for publication, July 31, 2001, and in revised form, December 20, 2001

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Structures of DNA polymerases bound with DNA reveal that the 5'-trajectory of the template strand is dramatically altered as it exits the polymerase active site. This distortion provides the polymerase access to the nascent base pair to interrogate proper Watson-Crick geometry. Upon binding a correct deoxynucleoside triphosphate, -helix N of DNA polymerase  is observed to form one face of the binding pocket for the new base pair. Asp-276 and Lys-280 stack with the bases of the incoming nucleotide and template, respectively. To determine the role of Lys-280, site-directed mutants were constructed at this position, and the proteins were expressed and purified, and their catalytic efficiency and fidelity were assessed. The catalytic efficiency for single-nucleotide gap filling with the glycine mutant (K280G) was strongly diminished relative to wild type for templating purines (>15-fold) due to a decreased binding affinity for the incoming nucleotide. In contrast, catalytic efficiency was hardly affected by glycine substitution for templating pyrimidines (<4-fold). The fidelity of the glycine mutant was identical to the wild type enzyme for misinsertion opposite a template thymidine, whereas the fidelity of misinsertion opposite a template guanine was modestly altered. The nature of the Lys-280 side-chain substitution for thymidine triphosphate insertion (templating adenine) indicates that Lys-280 "stabilizes" templating purines through van der Waals interactions.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

DNA polymerase (pol)¹  is an attractive model to study polymerase strategies employed to assure efficient and faithful DNA synthesis. Its small size (39-kDa) and lack of accessory proteins facilitates its biochemical, kinetic and structural characterization. DNA polymerase  is also the only eukaryotic polymerase for which there is a high-resolution crystal structure. It is structurally and kinetically suited to function on short DNA gaps during DNA repair or replication (1). Although pol  (pol X family) appears to have evolved separately from other families of polymerases of known structure (2), it shares many general structural and mechanistic features with these polymerases. The polymerase domain is composed of three functionally distinguishable subdomains. The polymerase catalytic subdomain binds two divalent metal cations that assist the nucleotidyl transferase reaction. Two additional subdomains that have primary roles in duplex DNA binding and nucleoside 5'-triphosphate (dNTP) selection surround the catalytic subdomain. These subdomains will be referred to as C (catalytic), D (duplex DNA binding), and N (dNTP selection) subdomains to discern their intrinsic function. These would correspond to the palm, thumb, and fingers subdomains, respectively, according to the nomenclature that utilizes the architectural analogy to a right-hand (3).²

DNA polymerase  also utilizes a kinetic mechanism similar to other polymerases. Steady-state and pre-steady-state kinetics analyses indicate that DNA polymerases follow an ordered binding of substrates, with DNA binding first (4). Comparison of structures of pol  bound to substrate and product DNA (binary complexes), with the structure of pol  bound to substrate DNA and a complementary incoming nucleotide (ternary complex), indicates that numerous structural transitions occur upon binding a correct nucleotide (5). The significance or contribution of these transitions toward fidelity (i.e. substrate specificity) remains to be determined.

The polymerase subdomains are structurally distinct among the different polymerase families. Comparison of DNA polymerase structures bound to DNA with those that include an incoming complementary dNTP reveals that the N-subdomain repositions itself to "close" upon the nascent base pair (5-10). These structures also reveal that the template strand is radically bent as it exits the polymerase active site. This bend in the template strand serves at least two functions. First, it provides the polymerase N-subdomain access to one face of the nascent base pair and the DNA minor groove. This access gives the polymerase the opportunity to check whether geometrical constraints imposed by correct Watson-Crick hydrogen bonding occur. Secondly, it displaces the next templating base away from the polymerase active site, discouraging incorrect template base reading (deletion mutagenesis) (11). For pol , the nascent base pair is sandwiched between -helix N (residues 275-288) and duplex DNA (Fig. 1A). Loss of minor groove hydrogen bonding and/or van der Waals interactions with the templating nucleotide of the nascent base pair through alanine substitution for Arg-283 results in dramatically reduced catalytic efficiency (12, 13), base substitution fidelity (12-14), and frameshift fidelity (11). In contrast, loss of hydrogen bonding and/or van der Waals contact with the incoming nucleotide and Asp-276 (15) or Asn-279 (12, 16) of -helix N results in little or no effect on fidelity or catalytic efficiency. In the "closed" ternary pol  complexes (5, 6), Asp-276 and Lys-280 stack with the bases of the incoming nucleotide and template, respectively (Fig. 1C). Here we kinetically characterize a series of mutant enzymes with substitutions for Lys-280 to probe whether this side chain plays a critical role in template positioning and/or fidelity.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials-- Poly(dA), p(dT)₁₀, p(dT)₂₀, ultrapure deoxynucleoside triphosphates, [-³²P]ATP, and MicroSpin G-25 columns were from Amersham Biosciences, Inc.. [-³²P]dTTP was from PerkinElmer Life Sciences and DE81 filters were from Whatman.

Mutagenesis of the Human Pol  Gene-- Oligonucleotide site-directed mutagenesis was performed using a procedure described previously (17). The full-length wild-type pol  gene was sub-cloned into pBluescript II SK (Stratagene). Amino acid changes were generated by PCR with primers containing the desired mutation. The following mutations were introduced into the Lys-280 codon (AAG, 5' to 3'): GCC (K280A), GGA (K280G), ATT (K280I), CTG (K280L), ATG (K280M), CAG (K280Q), CGA (K280R). To ensure that the resulting pol  gene contained the desired change, the entire coding sequence of each mutant was confirmed by DNA sequence analysis. Each mutant was cloned into pWL-11 (18), a bacterial expression plasmid containing the  PL promoter and overexpressed in Escherichia coli TAP56 cells.

Protein Purification-- Wild-type and mutant proteins were purified as described previously (19). Enzyme concentrations were determined by Coomassie dye binding using purified pol  as the standard (20). The concentration of purified pol  was determined by total amino acid analysis.

DNA Preparation-- A 34-mer oligonucleotide DNA substrate containing a single-nucleotide gap was prepared by annealing three gel-purified oligonucleotides (Oligos Etc., Wilsonville, OR) to create a single-nucleotide gap at position 16. Each oligonucleotide was resuspended in 10 mM Tris-HCl, pH 7.4, and 1 mM EDTA, and the concentration was determined from their UV absorbance at 260 nm. The annealing reactions were carried out by incubating a solution of 10 µM primer with 12 µM each of downstream and template oligonucleotides at 90-100 °C for 3 min followed by 30 min at 65 °C and then slow cooling to room temperature. The sequence of the gapped DNA substrate was primer, 5'-CTGCAGCTGATGCGC-3', downstream oligonucleotide, 5'-GTACGGATCCCCGGGTAC-3', and template, 3'-GACGTCGACTACGCGXCATGCCTAGGGGCCCATG-5', where the X represents A, C, G, or T. The primer was 5'-labeled with [-³²P]ATP using T4 polynucleotide kinase (New England BioLabs), and unincorporated radioactive ATP was removed with a MicroSpin G-25 column. The downstream primer was synthesized with a 5'-phosphate.

Poly(dA) and p(dT)₁₀ or p(dT)₂₀ were annealed at a template to primer nucleotide ratio of 5 (i.e. 5-fold more template than primer nucleotides) or 10. These homopolymeric oligonucleotides were annealed by heating to ~100 °C for 1 min and slowly cooling to room temperature. This procedure has been shown to eliminate primer stacking on the template (21). The concentration of annealed template-primers is expressed as the concentration of primer 3'-termini.

Kinetic Assays-- Steady-state kinetic parameters for polymerization on poly(dA)-oligo(dT) were determined as described previously (12). Enzyme activities were determined using a standard reaction mixture (50 µl) containing 50 mM Tris-HCl, pH 7.4 (22 °C), 100 mM KCl, 5 mM MnCl₂, and the appropriate substrate concentrations. Additional reaction conditions and details are described in the figure and table legends. Reactions were initiated by addition enzyme and stopped after an appropriate time interval with 20 µl of 0.5 M EDTA. Quenched reaction mixtures were spotted onto Whatman DE-81 filter disks and dried. Unincorporated [-³²P]dTTP was removed, and filters were counted as described (22).

Steady-state kinetic parameters for single-nucleotide gap filling were determined by initial velocity measurements where the heteropolymeric DNA concentration was held constant at 200 nM and the dNTP concentration varied. In general, the conditions were similar to those described above for pol(dA)-oligo(dT) except that MgCl₂ replaced MnCl₂. In some instances requiring high dNTPs concentrations (e.g. misincorporation assays), the MgCl₂ concentration was increased to 10 mM so that there was at least 5 mM free Mg²⁺. Enzyme concentrations and reaction time intervals were chosen so that substrate depletion or product inhibition did not influence initial velocity measurements. The quenched samples were mixed with an equal volume of formamide dye, and the products were separated on 12% denaturing polyacrylamide gels. The dried gels were analyzed using a PhosphorImager (Molecular Dynamics) to quantify product formation.

To directly measure the rate of the first insertion (k_pol) and the equilibrium nucleotide dissociation constant (K_d), single-turnover kinetic assays (enzyme/DNA = 5) were performed as outlined previously (15) employing a KinTek Model RQF-3 chemical quench-flow apparatus (KinTek Corp., State College, PA). Briefly, a solution of pol  (1 µM) was preincubated with single-nucleotide gapped DNA (200 nM) with a templating adenine in the gap. This solution was rapidly mixed (2-fold dilution) with various concentrations of dTTP/Mg²⁺. Specific conditions (pH, temperature) and final salt concentrations were as described for the steady-state assay. After various time periods, the reactions were stopped with 0.25 M EDTA, and the quenched samples were mixed with an equal volume of formamide dye. Products were separated and quantified as described above. Under these conditions, the first-order rate constant of the exponential time courses was dependent on the concentration of dTTP. A secondary plot of the concentration dependence of k_obs was hyperbolic and fitted to Equation 1, where k_pol is the intrinsic rate constant for the step limiting the first insertion.

(Eq. 1)

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Vertical-scanning Mutagenesis of Lys-280-- To probe the functional significance of the template base interactions with Lys-280 (Fig. 1), seven alternate side chains were individually introduced at this position. The alternate side chains differ in their size, hydrophobicity, and hydrogen-bonding potential. The mutant proteins were expressed in E. coli and purified. After purification, SDS-polyacrylamide gel analysis indicated that the mutant pol  polypeptides were greater than 99% homogeneous (data not shown). The catalytic efficiency and fidelity of the purified mutant proteins were determined.

View larger version (50K): [in this window] [in a new window]   Fig. 1.   Buried solvent-accessible surface of the nascent base pair in the DNA polymerase  active site. A, structure of the pol  ternary complex bound to one-nucleotide gapped DNA and ddCTP. The template strand (red) is radically bent as it exits the polymerase active site. The molecular surface of one face of the binding pocket for the nascent base pair is contributed by -helix N (magenta). The incoming ddCTP (dark blue) identifies the 3'-end of the primer strand (semi-transparent blue) and forms a Watson-Crick base pair with the templating guanine. The bases and sugars of the other nucleotides are omitted for clarity. The active site metals (orange) coordinate the triphosphate of the incoming ddCTP. An expanded view of the structural elements in the boxed area is shown in panel B. B, Lys-280 (Corey-Pauling-Koltun representation) of -helix N (magenta cylinder) of the N-subdomain makes van der Waals contact with the base of the templating residue (red). The semi-transparent van der Waals surface of the templating guanine is shown, but the other template and primer residues are omitted for clarity. C, view from -helix N, illustrating the buried solvent-accessible surface area (magenta) occurring between pol  and the nascent base pair (cyan). Residues of -helix N that form the binding pocket are illustrated. Asp-276 and Lys-280 stack with the bases of the incoming nucleotide (ddCTP) and templating base, respectively. Asn-279 and Arg-283 form the "minor groove face" of the binding pocket. The major groove edge of the nascent base pair is exposed to solvent. This figure was made with GRASP (42), Molscript (43), and Raster3D (44).

Relative Catalytic Efficiency on Homopolymeric Template-Primers-- To survey the influence of altering the chemical nature of the side chain at residue 280, we analyzed the steady-state kinetics of thymidine triphosphate incorporation on a simple template-primer system, poly(dA)-oligo(dT) (Table I). This template-primer system has proven useful in characterizing other pol  mutant enzymes (12, 23, 24). The catalytic activity (k_cat) and Michaelis constant for the template-primer (K_m,DNA) of all the mutant enzymes were similar to that of wild-type enzyme. In contrast, K_m,dTTP and k_cat/K_m,dTTP for some of the mutant enzymes were modestly elevated. In general, the effect correlated with the size of the mutant side chain. Whereas glycine, alanine, and glutamine substitutions for Lys-280 resulted in a small reduction in catalytic efficiency relative to wild-type enzyme, the other mutant enzymes (i.e. leucine, isoleucine, methionine, and arginine) had little or no effect. The pol  enzymes in Table I are listed in order of increasing size of residue 280 (25).

Influence of the Nature of the Templating Nucleotide on Catalytic Efficiency of K280G-- Because the K280G mutant enzyme exhibited a significant decrease in catalytic efficiency on a homopolymeric template-primer, we extended our kinetic analysis employing a more natural substrate, a heteropolymeric DNA template-primer that has a single-nucleotide gap. This substrate has served as a model substrate for base excision repair assays examining pol -dependent repair synthesis (26). In particular, the repair of the promutagenic G-U base pair is commonly examined in cell extracts or reconstituted systems (26, 27), and crystallographic structures of pol  bound to a single-nucleotide gapped DNA with a templating guanine have been solved (5). In contrast to the small but significant decrease in catalytic efficiency observed on homopolymeric DNA with this mutant, glycine substitution for Lys-280 resulted in a much larger decrease in catalytic efficiency for insertion of dCTP opposite a templating deoxyguanine (Table II and Fig. 2). Surprisingly, the magnitude of the loss of catalytic efficiency for correct nucleotide insertion was strongly dependent on the identity of the templating base. Deoxypyrimidine triphosphate insertion was effected to the greatest extent (templating dA and dG), whereas insertion of a deoxypurine triphosphate (templating dC and dT) was hardly affected (less than 4-fold). The altered catalytic efficiency was due entirely to changes in K_m, since k_cat was not altered by the glycine substitution (Table II). In comparison, k_cat and K_m for formation of all four Watson-Crick base pairs is not affected by the conservative arginine substitution (K280R; Table II).

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Template-dependent effect of glycine substitution for Lys-280 on relative catalytic efficiency. Wild-type pol  and the mutant K280G enzyme were assayed on single-nucleotide gapped DNA substrates as outlined under "Experimental Procedures." The identity of the templating base was altered, and the catalytic efficiency (kcat/Km,dNTP) of the complementary nucleotide was determined. The relative (wild-type/K280G) effect of the glycine substitution was strongly dependent on the identity of the templating base. Glycine substitution diminished catalytic efficiency to a much greater extent with templating purines (i.e. insertion of dCTP and dTTP) than pyrimidines (i.e. insertion of dATP and dGTP). These represent the mean and S.E. of at least two independent determinations. The catalytic efficiencies for insertion of dATP, dCTP, dGTP, and dTTP by wild-type pol  were 0.76 ± 0.17, 0.76 ± 0.10, 1.17 ± 0.23, and 0.29 ± 0.02 s1-µM1, respectively.

The elevated K_m for correct insertion opposite a templating deoxypurine with the glycine mutant of Lys-280 could be due to a lower binding affinity for the incoming nucleotide (i.e. elevated K_d), a diminished rate of nucleotide insertion during the first enzymatic turnover (i.e. decreased k_pol), and/or a loss of DNA binding affinity (i.e. increased DNA dissociation rate constant, k_off). As we have outlined previously (28), the K_m for the incoming nucleotide in rapid equilibrium with the subsequent step is equivalent to K_d[k_off/(k_pol + k_off)].³ To dissect the contribution of k_pol and K_d in the elevated K_m,dTTP observed with the K280G mutant, single-turnover time courses were determined. This approach (polymerase/DNA = 5) eliminates interference from enzyme cycling since under this condition nearly all of the substrate DNA is bound to enzyme so that, upon the addition of dTTP/Mg²⁺, dNTP binding and insertion limit catalysis. Under these single-turnover conditions, the observed rate constant (k_obs) of the exponential time courses was dependent on the concentration of dTTP (Fig. 3). The data fit well to Equation 1, with k_pol of 3.2 s¹ and K_d of 16.6 µM for wild-type enzyme, yielding a specificity constant (k_pol/K_d) of 0.19 s¹-µM¹. For the K280G mutant, k_pol and K_d were determined to be 2.7 s¹ and 289 µM, respectively, so that k_pol/K_d = 0.009 s¹-µM¹. These specificity constants are similar to those determined from a steady-state analysis (i.e. k_cat/K_m; see legend to Fig. 2 and Table II).

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Effect of glycine substitution for Lys-280 on single-turnover analysis of dTTP insertion. Wild-type enzyme or the K280G mutant was preincubated with DNA and mixed rapidly with various concentrations of dTTP/Mg2+ as outlined under "Experimental Procedures." Under this condition (polymerase > DNA), the first-order time courses were fitted to a rising exponential equation. The observed rate constants (kobs) for these time courses were plotted as a function of dTTP concentration and fitted to a hyperbola (Equation 1). The fits yielded kpol of 3.2 ± 0.1 or 2.7 ± 0.1 s1 and Kd of 16.6 ± 2.0 or 289 ± 26 µM for wild-type or K280G enzymes, respectively. This yields specificity constants (kpol/Kd) for insertion of dTTP opposite deoxyadenine of 0.19 ± 0.02 s1-µM1 for wild-type enzyme and 0.0093 ± 0.0009 s1-µM1 for the K280G mutant. These specificity constants are similar to those determined from a steady-state analysis (i.e. kcat/Km; see the legend of Fig. 2 and Table II).

Relative Catalytic Efficiency for Correct Insertion into a Single-nucleotide DNA Gap with a Templating Deoxyadenosine-- The greatly diminished catalytic efficiency for thymidine insertion observed with the K280G mutant provided us with a wide range of catalytic efficiencies to determine the influence of the chemical nature of the residue 280 side chain. Although the magnitude of the effect is larger, the results are very similar to those observed with the homopolymeric template-primer system (Table I). Fig. 4 illustrates the relative (wild-type/mutant) catalytic efficiency for thymidine insertion into a one-nucleotide gap. The data are presented in order of decreasing size of the residue 280 side chain (large to small, left to right). As with the homopolymeric template-primer system, the altered catalytic efficiencies are primarily due to changes in K_m,dTTP (data not shown).

View larger version (15K): [in this window] [in a new window]   Fig. 4.   Influence of the side chain at residue 280 of pol  on catalytic efficiency for dTTP insertion. Wild-type pol  (K*) and the mutant variants at residue 280 were assayed on single-nucleotide gapped DNA substrates with a templating dA, and the catalytic efficiency (kcat/Km,dTTP) of the complementary dTTP was determined as outlined under "Experimental Procedures." The relative (wild-type/mutant) effect was strongly dependent on the volume of the alternate side-chain substitutions. The alternate residue 280 side chains are listed in order (left to right) of decreasing van der Waals volume (25). These represent the mean and S.E. of at least two independent determinations.

Effect of Divalent Metal on Relative Catalytic Efficiency-- As noted above, the influence of the glycine substitution had a much more pronounced effect on thymidine insertion employing a template-primer with a one-nucleotide gap (Fig. 4) than a homopolymeric DNA substrate (Table I). Although the rate-limiting step(s) may be different in these two kinetic assays, catalytic efficiency (k_cat/K_m) is expected to be the same (28). An important distinction between these two assays is the divalent ion utilized during catalysis. The poly(dA)-oligo(dT) template-primer system strictly requires Mn²⁺. In contrast, the heteropolymeric template-primer system can utilize either Mn²⁺ or Mg²⁺. Because the above assays examining single-nucleotide gap filling utilized Mg²⁺, the effect of glycine substitution on thymidine insertion into a single-nucleotide gap was repeated with a reaction mixture where MnCl₂ replaced MgCl₂. Consistent with the assays described above, the identity of the divalent metal cofactor influenced the magnitude of the effect of the glycine substitution. In the presence of 5 mM MnCl₂, the wild-type enzyme exhibited a catalytic efficiency of 1.8 ± 0.2 s¹-µM¹ for insertion of dTTP opposite a templating dA in a one-nucleotide gap. In contrast to the nearly 50-fold reduction observed utilizing Mg²⁺ (Fig. 2), the catalytic efficiency for the K280G mutant was diminished less than 25%, i.e. 1.4 ± 0.2 s¹-µM¹.

Relative Fidelity of K280G-- The base excision repair of a deaminated cytosine (i.e. uracil) involves an intermediate where dG serves as the templating base for pol  single-nucleotide insertion. Because the catalytic efficiency for insertion of deoxypyrimidine triphosphates (i.e. templating purines) is significantly reduced with the glycine mutant, we examined the impact of the glycine substitution on the fidelity of single-nucleotide gap-filling with guanine or thymidine serving as the templating base (Table III and Fig. 5). As expected, the catalytic efficiency is dramatically reduced for the misinsertion of dCTP and dGTP opposite a templating dT for wild-type enzyme (Tables II and III), and glycine substitution for Lys-280 resulted in small reductions in comparison to wild-type enzyme in catalytic efficiencies for these mispairs. Because the impact of the glycine substitution on dATP (i.e. correct) insertion was about the same, the resulting fidelity was not altered relative to wild-type enzyme (Fig. 5). In contrast, there was a small but significant change in relative fidelity with the K280G mutant for dATP and dTTP misinsertion opposite guanine. Interestingly, K280G is a mutator polymerase for misinsertion of dATP and an anti-mutator for misinsertion of dTTP opposite the templating guanine relative to wild-type enzyme (Fig. 5).

View this table: [in this window] [in a new window]   Table III Steady-state kinetic summary for incorrect insertion on single-nucleotide gapped DNA substrates Assays were performed as described under "Experimental Procedures." The DNA substrates (200 nM) were identical except for the identity of the templating base (dG or dT) in the single-nucleotide gap. The concentration of a non-complementary dNTP was varied from at least 0.3 to 3 Km. Initial velocities were fitted to the Michaelis equation by nonlinear least squares methods. The results represent the mean (S.E.) of at least two independent determinations. The catalytic efficiencies (kcat/Km) for correct insertion opposite a templating dG and dT were 0.76 ± 0.10 and 0.76 ± 0.17 for wild-type enzyme and 0.042 ± 0.005 and 0.31 ± 0.07 s1  µM1 for K280G, respectively.

View larger version (14K): [in this window] [in a new window]   Fig. 5.   Template-dependent effect of glycine substitution for Lys-280 on relative fidelity. Wild-type pol  and the mutant K280G enzyme were assayed on single-nucleotide gapped DNA substrates as outlined under "Experimental Procedures." The identity of the templating base was either dG or dT, and the catalytic efficiency (kcat/ Km,dNTP) for formation of a non-Watson-Crick base pair was determined. Fidelity was calculated from [(kcat/Km,dNTP)correct + (kcat/ Km,dNTP)incorrect]/(kcat/Km,dNTP)incorrect (Table II). Whereas glycine substitution for Lys-280 had no effect on fidelity for misinsertion of dCTP and dGTP opposite dT, there was a small but significant change in relative fidelity with the K280G mutant for dATP and dTTP misinsertion opposite guanine. Interestingly, K280G is a mutator polymerase for misinsertion of dATP and an anti-mutator for misinsertion of dTTP opposite the templating guanine relative to wild-type enzyme. These represent the mean and S.E. of at least two independent determinations. The fidelity for misinsertion of dCTP and dGTP opposite dT and for dATP and dTTP opposite dG by wild-type pol  was 51,000, 8,700, 245,000, and 63,000, respectively.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

DNA polymerases must efficiently select (bind and incorporate) the correct dNTP from a pool of structurally similar molecules to ensure accurate DNA synthesis during DNA replication and repair. Structural and kinetic characterization of a variety of diverse DNA polymerases has hastened our understanding of the molecular strategies employed by polymerases to achieve efficient nucleotide insertion. It is generally accepted that polymerases require that the nascent base pair conform to Watson-Crick geometry for efficient DNA synthesis. However, the functional significance of Watson-Crick hydrogen bonding and the interactions (hydrogen bonding and van der Waals) occurring between the polymerase and nascent base pair appear to be dependent on the DNA polymerase. For example, exonuclease-deficient forms of Klenow fragment and T7 DNA polymerase are able to efficiently insert the nonpolar isostere of thymidine (i.e. difluorotoluene), which lacks hydrogen-bonding capacity, opposite a templating adenine base. In contrast, pol  and Moloney murine leukemia virus reverse transcriptase are not able to insert this thymidine isostere (29).

The polymerase dNTP binding site is dynamic and unique for each incorporation; i.e. the identity of the primer terminus and templating base is altered with each insertion. Consequently, substrate-polymerase van der Waals and hydrogen-bonding interactions are unique during each catalytic step. For example, crystal structures of pol  bound to a single-nucleotide gap (binary substrate complex), single-nucleotide gap and incoming dNTP (ternary substrate complex), and nicked DNA (binary product complex) indicate that Tyr-271 donates a hydrogen bond to a unique DNA minor groove acceptor in each case (1, 5). In addition, there are numerous protein conformational changes observed when the correct dNTP binds to a binary polymerase complex. Most notable is the movement of the N-subdomain to close upon the nascent base pair (30).

DNA polymerases decode the template strand in an attempt to preserve Watson-Crick base-pairing rules. In doing so, they must be able to decipher the identity of the templating base. This requires that the template base be positioned so that the incoming dNTP can examine geometric constraints imposed by the polymerase active site. The templating base is a critical component of the polymerase active site, and optimum positioning is therefore essential for efficient DNA synthesis (4). Interestingly, structures of binary complexes of DNA polymerases from the pol A family bound to template-primer DNA indicate that the templating base is positioned outside of the DNA helix (1, 8, 31). Closure of the N-subdomain of pol  also repositions the templating residue (11), but this repositioning is more subtle than what must occur for the pol A family of DNA polymerases. Consistent with a stabilization of the templating residue with the additional polymerase contacts in the closed conformation is the 20-Ų decrease in the average B-factor for the templating base in the closed complex relative to the open binary DNA complex (5). Arg-283 of pol  has an essential role in this positioning, and alanine substitution for this residue results in a dramatic loss of catalytic efficiency and fidelity (11-14). The R283A pol  mutant represents the greatest loss of fidelity engineered by a single-point mutation for any DNA polymerase. Because of the critical role that template positioning has on catalytic efficiency and fidelity, we have employed a steady-state kinetic analysis to explore the role that Lys-280 plays in template base positioning and/or stabilization. As predicted by current kinetic models for polymerization, we have demonstrated that substrate specificity, as determined by k_cat/K_m, is equivalent to k_pol/K_d determined by pre-steady-state analysis (15). In this later analysis, k_pol is the first-order rate constant for the insertion step and is limited by a conformational change and/or chemistry, and K_d is the equilibrium dissociation constant for dNTP binding. Lys-280 is observed to be stacked with the templating guanine of the nascent base pair in the pol  DNA complex with an incoming ddCTP (Fig. 1) (5, 6).

The observation that glycine substitution for Lys-280 results in a decrease in catalytic efficiency that is strongly dependent on the identity of the templating base indicates that interactions with the nascent base pair may be energetically unique for formation of each Watson-Crick base pair (Fig. 2). Thus, residue 280 interactions with templating purines are more important than they are for templating pyrimidines. This suggests that template positioning and stabilization is unique for each base pair. This asymmetry is not unexpected. Li and Waksman (32) have determined the crystal structure of four closed ternary complexes of Klentaq, each with a different Watson-Crick nascent base pair. Polymerase interactions that were specific for the dC-ddGTP base pair were noted in the DNA major and minor grooves. Furthermore, in the absence of a templating base (i.e. an abasic site), deoxypurines are typically inserted with a much higher efficiency than deoxypyrimidine triphosphates. Extending that observation to correct nucleotide insertion, the purine may play a primary role during formation of the Watson-Crick base pair independent of whether it is in the templating position or is being selected for insertion. Thus, an incoming deoxypurine triphosphate may have a critical role in template positioning and stabilization.

The loss in catalytic efficiency with the K280G mutant with templating purines was primarily due to an elevated K_m (Table II). Because K_m for an incoming dNTP is a reflection of the dNTP binding affinity and the identity of the rate-limiting step (DNA dissociation and/or k_pol; for a discussion, see Ref. 28), a single-turnover approach was employed to determine the contribution of k_pol and K_d to the specificity constant. The increase in K_m was primarily due to an increase in the equilibrium binding affinity (K_d) of the incoming dTTP (Fig. 3). This suggests that the role of the Lys-280 side chain is to stabilize the templating purine in a conformation that optimizes nucleotide binding. Because k_pol was hardly affected by the glycine substitution, the modified interaction with the templating base doesn't perturb the step that limits insertion (i.e. chemistry or conformational change).

As with correct base pair formation, the identity of the templating base also influences the fidelity of the K280G mutant. With a templating dT, the glycine mutant exhibited a fidelity for insertion of dCTP or dGTP similar to wild-type enzyme (Fig. 5). In contrast, the fidelity of dATP and dTTP insertion by K280G was altered with the templating purine, dG. However in these cases, K280G is a mutator for insertion of dATP and an anti-mutator for insertion of dTTP. This suggests that the wild-type lysine side chain stabilizes a dTTP-dG wobble mispair and discourages the dATP-dG mismatch intermediate.

The structure of a mispair intermediate in a polymerase active site has not been determined. However, it is expected that mispair intermediates will resemble those observed in duplex DNA. Structures of dA-dG mismatches in duplex DNA display conformational variability (anti-syn glycosidic preferences) that appear to depend on intra-strand stacking interactions (33). In our one-nucleotide gapped DNA substrate, the templating dG is adjacent to another guanine that forms a Watson-Crick base pair with the primer terminus (dC). In this sequence context, the templating guanine may be expected to prefer a syn conformation that could form a Hoogsteen base pair with an incoming dATP (anti). The data suggest that loss of the interactions with Lys-280 through glycine substitution increases the probability of the mismatched syn conformation. The K280R mutant displayed a 2.8-fold increase in fidelity relative to wild-type enzyme (data not shown), indicating that interactions with a basic residue 280 side chain and the templating guanine may discourage the mismatched intermediate.

The dT-dG mispair forms a wobble base pair, with dT projecting into the major groove and dG into the minor groove (34). Because polymerases generally interact with the nascent base pair through the minor groove, the predominant structural alteration is expected to be a shift of thymidine into the major groove to satisfy hydrogen bonding. If the templating guanine base is not stabilized through stacking interactions with Lys-280, the templating guanine could "drift" into the major groove, making it very difficult for an incoming dTTP to form a Wobble base pair. Comparison of the templating guanine position in the open complex, which lacks interactions with Lys-280, with that in the closed ternary complex reveals that the templating guanine is displaced into the major groove in the open form (see Fig. 2A in Ref. 11). For the reciprocal mispair (dGTP-dT), loss of stacking interactions with a templating dT would not be expected to be influenced by removing the residue 280 side chain, since interactions would be diminished with wild-type enzyme for a templating thymidine displaced into the major groove.

The efficiencies for formation of the four Watson-Crick base pairs varies less than 5-fold (Fig. 2 and Table II). In contrast, the catalytic efficiency for incorrect insertion is diminished at least 10⁴-fold for wild-type enzyme and varies over a wider range (30-fold) for the mispairs examined (Table III). The structural accommodation of different mispairs in B-DNA suggests that the polymerase interacts uniquely with each mispair. Thus, it would be surprising that a single alteration in a DNA polymerase would give rise to a general mutator polymerase that makes all mispairs with a frequency greater than wild-type enzyme. As observed with the K280G mutant (Fig. 5), a spectrum of mutation frequencies (i.e. reciprocal of fidelity) relative to the wild-type polymerase is expected. The R283K mutant of pol  has also been reported to exhibit a mutator activity for certain mispairs but produces other mispairs at a frequency similar to wild-type enzyme (14). As noted above, since certain mispairs are structurally sensitive to their sequence context, it is not surprising that a human immunodeficiency virus-1 reverse transcriptase mutant (R72A) has been described that is an anti-mutator in one sequence context and a mutator in another for the same mispair (35). In this case, Arg-72 of reverse transcriptase stacks with the base of the incoming dNTP (9).

Site-directed mutagenesis of Lys-280 indicated that the glycine derivative resulted in the greatest decrease in catalytic efficiency on a homopolymeric template-primer system (Table I), but the effect was significantly lower than observed with heteropolymeric DNA, which utilizes the same templating nucleotide (i.e. dA; Fig. 4). The previous assay requires Mn²⁺ as the divalent metal cofactor, whereas nucleotide insertion on the one-nucleotide gapped heteropolymeric DNA substrate can utilize either Mg²⁺ or Mn²⁺. Substituting Mn²⁺ for Mg²⁺ in this assay resulted in little or no effect on relative catalytic efficiency for thymidine insertion with the glycine mutant. Thus, the metal cofactor can override the contribution that polymerase stacking with the template base confers to overall catalytic efficiency. Crystallographic studies indicate that Mn²⁺ can activate non-template-directed nucleotide insertion, where Mg²⁺ cannot (36). It was suggested that this might be related to the conformation of one of the metal ligands, Asp-192.

The flexible Lys-280 side chain consists of a hydrophobic arm that is positively charged at its end. These characteristics provide the potential for it to stack with the templating base but specifically interact with the N7 of purines by providing a hydrogen bond donor. The influence of the nature of the residue 280 side chain on thymidine insertion into a one-nucleotide gapped DNA substrate indicates that the ability to donate a hydrogen bond is not essential. Isoleucine and methionine substitutions at residue 280 result in mutant enzymes that are very similar to wild-type pol . These hydrophobic residues would not be expected to be able to form a hydrogen bond with N7 of adenine. In general, the magnitude of the decrease in catalytic efficiency for thymidine insertion correlated with the van der Waals volume of the alternate 280 side chains (Table I and Fig. 4) and not with the non-polar accessible surface area. For example, the non-polar accessible surface area of the glutamine and alanine side chains is 53 and 67 Å², respectively, whereas the van der Waals volume of glutamine is nearly 2-fold greater than alanine (25). Relative (wild-type/mutant) k_cat/K_m,dTTP was significantly greater for the alanine substitution than for glutamine (Fig. 4), suggesting that specific interactions with surrounding side chains (i.e. packing) may also contribute to the observed effects. In other pol X family DNA polymerases such as terminal transferase, DNA polymerase , and DNA polymerase µ, sequence alignments predict that the equivalent residue in these enzymes would be an arginine. In contrast, sequence alignment suggests that the African swine fever virus encodes an X family DNA polymerase that has an isoleucine at this position (37). As noted above, these substitutions for Lys-280 of human pol  did not affect catalytic efficiency for the correct insertion of thymidine.

A previous study concluded that an alanine substitution for Lys-280 had no effect on catalytic efficiency or fidelity for rat pol  (38). Because this study did not employ a DNA substrate that utilized a templating adenine, the changes in catalytic efficiency and fidelity expected with the alanine substitution (K280A) rather than glycine (K280G) may have been too small to detect. These differences underscore how subtle changes in protein and/or substrate structure can result in diverse effects. Thus, glycine substitution for Lys-280 results in a DNA polymerase that exhibits wild-type selectivity for certain mispairs (e.g. dCTP or dGTP insertion opposite templating thymidine) but mutator (dG-dATP) or anti-mutator (dG-dTTP) selectivity for others (Fig. 5).

The strategy employed by pol  to stabilize the templating base is multifaceted. It depends on specific interactions from its surroundings, that is, polymerase (Arg-283 and Lys-280) and the identity of the incoming nucleotide/metal cofactor. The contribution of these interactions (forces) to catalytic efficiency and, thus, fidelity will be characteristic for each base pair (correct and incorrect). Additionally, since the identity of the template base that pairs with the primer terminus can also affect catalytic efficiency, template base-stacking interactions with the 3'-templating base can also contribute significant interactions (39). It is essential that protein substrate (product) interactions suggested from structural studies be confirmed by kinetic and thermodynamic analyses. As the molecular mechanisms of polymerase substrate specificity emerge, it will be necessary to ascertain how each polymerase utilizes these to perform its particular function(s).

	ACKNOWLEDGEMENTS

We are grateful to Matthew Corregan for assistance with the purification of the mutant enzymes and to Drs. K. Bebenek and R. E. London for critical reading of the manuscript.

	FOOTNOTES

^* The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^§ To whom correspondence should be addressed. Tel.: 919-541-3267; Fax: 919-541-3592; E-mail: wilson5@niehs.nih.gov.

Published, JBC Papers in Press, December 26, 2001, DOI 10.1074/jbc.M107286200

² The subdomain nomenclature as originally proposed for pol  (40, 41) utilized the right-hand analogy; however, it is functionally opposite to that employed for other DNA polymerases. To eliminate potential confusion, a functionally based nomenclature is employed as outlined in the text.

³ The full expression for k_cat is (k_pol k_off)/(k_pol + k_off). Thus, k_cat/K_m should theoretically be equivalent to k_pol/K_d and is independent of the relative magnitudes of k_pol and k_off (28).

	ABBREVIATIONS

The abbreviations used are: pol, polymerase; dNTP, 2'-deoxynucleoside 5'-triphosphate; dd-, dideoxy.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

This article has been cited by other articles:

				P. Lin, V. K. Batra, L. C. Pedersen, W. A. Beard, S. H. Wilson, and L. G. Pedersen Incorrect nucleotide insertion at the active site of a G:A mismatch catalyzed by DNA polymerase {beta} PNAS, April 15, 2008; 105(15): 5670 - 5674. [Abstract] [Full Text] [PDF]

				G. C. Lin, J. Jaeger, and J. B. Sweasy Loop II of DNA polymerase beta is important for polymerization activity and fidelity Nucleic Acids Res., May 14, 2007; 35(9): 2924 - 2935. [Abstract] [Full Text] [PDF]

				W. W. Duym, K. A. Fiala, N. Bhatt, and Z. Suo Kinetic Effect of a Downstream Strand and Its 5'-Terminal Moieties on Single Nucleotide Gap-filling Synthesis Catalyzed by Human DNA Polymerase {lambda} J. Biol. Chem., November 24, 2006; 281(47): 35649 - 35655. [Abstract] [Full Text] [PDF]

				M. C. Foley, K. Arora, and T. Schlick Sequential Side-Chain Residue Motions Transform the Binary into the Ternary State of DNA Polymerase {lambda} Biophys. J., November 1, 2006; 91(9): 3182 - 3195. [Abstract] [Full Text] [PDF]

				M. A. Kalam, K. Haraguchi, S. Chandani, E. L. Loechler, M. Moriya, M. M. Greenberg, and A. K. Basu Genetic effects of oxidative DNA damages: comparative mutagenesis of the imidazole ring-opened formamidopyrimidines (Fapy lesions) and 8-oxo-purines in simian kidney cells. Nucleic Acids Res., January 1, 2006; 34(8): 2305 - 2315. [Abstract] [Full Text] [PDF]