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

J. Biol. Chem., Vol. 276, Issue 33, 30615-30622, August 17, 2001

This Article Abstract Full Text (PDF) All Versions of this Article: 276/33/30615    most recent M102694200v1 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 Vaisman, A. Articles by Woodgate, R. Search for Related Content PubMed PubMed Citation Articles by Vaisman, A. Articles by Woodgate, R. Social Bookmarking                      What's this?

Human DNA Polymerase  Promiscuous Mismatch Extension^*

Alexandra Vaisman, Agnès Tissier^§, Ekaterina G. Frank, Myron F. Goodman^¶, and Roger Woodgate

From the  Section on DNA Replication, Repair, and Mutagenesis, NICHD, National Institutes of Health, Bethesda, Maryland 20892-2725, and ^¶ Departments of Biological Sciences and Chemistry, Hedco Molecular Biology Laboratories, University of Southern California, University Park, Los Angeles, California 90089-1340

Received for publication, March 26, 2001, and in revised form, May 30, 2001

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Human DNA polymerase  is a low-fidelity template copier that preferentially catalyzes the incorporation of the wobble base G, rather than the Watson-Crick base A, opposite template T (Tissier, A., McDonald, J. P., Frank, E. G., and Woodgate, R. (2000) Genes Dev. 14, 1642-1650; Johnson, R. E., Washington, M. T., Haracska, L., Prakash, S., and Prakash, L. (2000) Nature 406, 1015-1019; Zhang, Y., Yuan, F., Wu, X., and Wang, Z. (2000) Mol. Cell. Biol. 20, 7099-7108). Here, we report on its ability to extend all 12 possible mispairs and 4 correct pairs in different sequence contexts. Extension from both matched and mismatched primer termini is generally most efficient and accurate when A is the next template base. In contrast, extension occurs less efficiently and accurately when T is the target template base. A striking exception occurs during extension of a G:T mispair, where the enzyme switches specificity, "preferring" to make a correct A:T base pair immediately downstream from an originally favored G:T mispair. Polymerase  generates a variety of single and tandem mispairs with high frequency, implying that it may act as a strong mutator when copying undamaged DNA templates in vivo. Even so, its limited ability to catalyze extension from a relatively stable primer/template containing a "buried" mismatch suggests that polymerase -catalyzed errors are confined to short template regions.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

It has been known for some time that DNA polymerases can be classified into discrete families based upon phylogenetic relationships (1, 2). The most recently discovered is the Y family (UmuC/DinB/Rev1/Rad30) of DNA polymerases (3). The Y family of polymerases (4-17), together with others from the A and X families that have recently been identified (18-21), have helped increase the known number of polymerases from 3 to 5 in Escherichia coli and from 6 to at least 15 in humans (22, 23). The cellular function of many members of the Y family of DNA polymerases is currently unknown. The notable exceptions are Escherichia coli pol¹ V, which clearly plays a pivotal role in SOS-induced mutagenesis by copying TT cis-syn dimers, TT (6-4) photoproducts and abasic moieties, lesions that typically block normal DNA replication (5-7, 24-26), and human pol , encoded by POLH (also known as the xeroderma pigmentosum variant (XPV) gene, and/or RAD30A), which plays an important role in the avoidance of sunlight-induced skin cancer by accurately copying cis-syn thymine dimers in vivo (11, 27-29).

The Y family DNA polymerases have several important properties in common including low processivity and extremely inaccurate nucleotide incorporation fidelity accompanied by the absence of intrinsic exonucleolytic proofreading in vitro. In addition to these properties, human pol , a paralog of pol  (30), exhibits a remarkable template-dependent misincorporation spectrum on undamaged DNA in vitro (12-14). Pol  uniquely favors the incorporation of G rather than A opposite a template base T, with roughly equal incorporation of T and A opposite T (12-14). On damaged DNA, pol  catalyzes limited unassisted error-prone bypass of a cis-syn cyclobutane thymine dimer and the misinsertion of two nucleotides opposite a TT (6-4) pyrimidine-pyrimidone photoproduct (31), one base opposite an abasic site (13, 14, 32) and an acetyl aminofluorene-guanine adduct (32, 33).

The in vitro properties of pol  suggest that it might play a role in error-prone translesion replication and possibly in elevated spontaneous mutagenesis, leading to increased carcinogenesis (12, 32). Another potential role for pol  is in the somatic hypermutation of human immunoglobulin genes, enabling the production of an efficient immune response to a wide variety of antigens (12, 34, 35). Indeed, the pattern of nucleotide misincorporation by pol  appears consistent with the mutation spectra observed during somatic hypermutation (12, 23, 35-37). Yet another putative role for this enzyme arises from a recent experiment showing that pol  contains an intrinsic 5'-deoxyribose phosphate lyase activity, such that it might substitute for pol  during base excision repair (38). An interesting hypothesis is that pol -catalyzed misinsertion of G opposite T, generated by deamination of 5-methyl cytosine, might actually protect the genome against C·G to T·A transition mutations (38).

Apart from the fact that mRNA for both the human (POLI) and the mouse (Pol ) gene is expressed at high levels in testis (30, 35) and at low levels in other tissues, very little is known about the in vivo properties of the enzyme. Most clues to its cellular function come from biochemical characterization of the enzyme in vitro (39). Here, we report our analysis of the ability of pol  to extend mispairs in a variety of sequence contexts. Our results suggest that the efficiency of pol -catalyzed mispair extension is largely dependent upon the template sequence 5' to the mispair, and even though pol  can extend most mispairs relatively efficiently, a "buried" mispair limits pol  synthesis to short regions of DNA.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

DNA Templates-- The templates and primers used for primer extension and steady-state kinetic experiments are listed in Fig. 1. All oligonucleotides were synthesized by Loftstrand Laboratories (Gaithersburg, MD) using standard techniques and were gel-purified before use. The primers were 5'-end-labeled using T4 polynucleotide kinase and [-³²P]ATP. DNA substrates were prepared by annealing templates with ³²P-labeled primers at a 1.5:1 molar ratio. Hybridization was achieved by heating the required mixture of oligonucleotides in an annealing buffer (50 mM Tris-HCl (pH 8), 5 mM MgCl₂, 50 µg/ml bovine serum albumin, and 1.42 mM 2-mercaptoethanol) for 10 min at 100 °C, followed by slow cooling to room temperature over a period of about 2 h. Annealing efficiencies were >95%, as evidenced by the different mobility of the ³²P-labeled primers before and after hybridization to the template on nondenaturing polyacrylamide gels.

Primer Extension Assays-- Glutathione S-transferase-tagged pol  was purified as described previously (12). 100 fmol of primer-template (expressed as primer termini) was incubated with 25 fmol of pol  at 37 °C for 15 min in 10-µl reactions containing 100 µM of either all four dNTPs or each dNTP individually, 40 mM Tris-HCl (pH 8.0), 5 mM MgCl₂, 10 mM dithiothreitol, 250 µg/ml bovine serum albumin, 60 mM KCl, and 2.5% glycerol. Reactions were terminated by the addition of 10 µl of formamide loading dye solution containing 500 mM EDTA, 0.1% xylene cyanol, and 0.1% bromphenol blue in 90% formamide. Before loading onto the gel, the reactions were denatured by heating at 100 °C for 10 min and immediately transferred to ice for 2 min. Products were resolved by denaturing polyacrylamide gel electrophoresis (7 M urea and 20% acrylamide, 4 h at 2000 V) and then visualized using a Molecular Dynamics PhosphorImager.

Kinetic Analysis of Replication Products-- Steady-state measurements for the apparent K_m and V_max for dNTP incorporation were carried out using standing start reactions as described previously (12, 40, 41). Initial time course studies were performed to ensure that the reactions were in the linear range. Less than 20% of the primers were extended under the steady-state conditions, ensuring single hit conditions (42). 100 fmol of DNA substrates was replicated at 37 °C for 2 min in 10-µl reaction mixtures containing 10 fmol of pol  and variable concentrations of dNTP. Reaction products were separated in a 20% polyacrylamide gel containing 7 M urea and then visualized and quantified using a Molecular Dynamics PhosphorImager and ImageQuant software (Molecular Dynamics). The velocity of dNTP incorporation () was determined by dividing the percentage of product generated by the respective time of the reaction. The relationship between  and dNTP concentration conformed to a Michaelis-Menten equation, as indicated by linearity in a Hanes-Woolf plot of [dNTP]/ versus [dNTP]. The apparent V_max and K_m were determined from a Hanes-Woolf plot by linear least squares fit using Sigma Plot software. The specificity of nucleotide incorporation by polymerase (f) was calculated as V_max/K_m measured in units percentage of primer elongation product per minute per µmol nucleotide. The nucleotide misincorporation ratio was determined as f_inc = (V_max/K_m)_incorrect/(V_max/K_m)_correct.

Elongation from a mismatch relative to extension from the correctly paired primer-template termini is calculated as f⁰_ext = (V_max/K_m)_mismatched/(V_max/K_m)_matched (43). The parameter f⁰_ext is the "intrinsic" mismatch extension ratio measuring the relative probability that polymerase bound to a mismatched compared with a correctly matched primer end catalyzes the addition of a next correct nucleotide, in the limit when the concentration of next correct dNTP  0 (43). In general, the relative mismatch extension efficiency depends explicitly on enzyme-DNA equilibrium binding constants to matched and mismatched primes-template DNA and on the absolute concentration of next correct dNTP. The relative efficiency of extending mismatches decreases dramatically with decreasing dNTP concentration, and maximum discrimination defined uniquely by f⁰_ext is achieved as the concentration of next correct dNTP  0. In contrast to the kinetic measurements for f_inc, which were performed using saturating primer/template DNA levels, measurements to determine f⁰_ext were carried out using subsaturating DNA concentrations in the linear region (where the pol apparent equilibrium dissociation constant (K_D) > primer/template DNA concentration). In this region, values of f⁰_ext are independent of any differences in the binding of pol  to matched versus mismatched primer 3'-ends (41, 43).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Pol -catalyzed Extension of Mispaired Primer-template Termini-- Pol -dependent nucleotide incorporation opposite template A is accurate and efficient, whereas pol -dependent nucleotide incorporation opposite template T is inaccurate and inefficient (12-14). As a consequence, we determined the ability of pol  to elongate 12 mismatched and 4 correctly matched primer-template termini when the "target" template base was either A or T (Fig. 1A, N³ = A or T).

View larger version (26K): [in this window] [in a new window]   Fig. 1.   Sequence of the DNA templates used in primer extension and kinetic studies. A, DNA substrates with two target template bases (A and T) containing all 12 mispairs and 4 correct pairs at the 3'-primer terminus were used to determine the ability of pol  to extend mismatched primer termini. B and C, DNA substrates with varying bases 5' to template T were used to study the sequence context effect on pausing opposite template T. D-G, DNA substrates with varying bases at a +1 position relative to the T:T mispair were used to study the distal influence of mismatch formation. Mismatched bases are shown in bold.

Reactions extending matched and mismatched primer ends were performed using either one or all four dNTP substrates present at relatively high concentration (100 µM). Pol  exhibits different misincorporation patterns opposite template T (Fig. 2, I) and A (Fig. 3, I) when extending correctly paired primers. Extensive misincorporation of G and T occurs opposite template T (Fig. 2, I). Although T is preferentially incorporated opposite template A, dAMP, dCMP, and dGMP misincorporations also occur directly opposite A and further downstream (Fig. 3, I). These data are in complete agreement with previous studies (12-14). However, our studies also show that similar misincorporation patterns occur using all four correctly paired primer termini, allowing us to conclude that the identity of the paired 3'-primer terminus has little influence on subsequent (mis)insertions at the next template base.

View larger version (69K): [in this window] [in a new window]   Fig. 2.   Extension of matched and mismatched primer termini by pol  with T as the next template base. Primer extension studies were performed for 15 min using 100 fmol of DNA template, 25 fmol of pol , and either a mixture of all four dNTPs or each dNTP individually present at a 100 µM concentration. Lane 4, incubation with dNTP mixture; lane C, incubation with dCTP; lane G, incubation with dGTP; lane T, incubation with dTTP; lane A, incubation with dATP. The template sequence is indicated on the left (pr, primer). The primer-template structure is shown below each gel.

View larger version (71K): [in this window] [in a new window]   Fig. 3.   Extension of matched and mismatched primer termini by pol  with A as the next template base. Primer extension studies were performed for 15 min using 100 fmol of DNA template, 25 fmol of pol , and either mixture of all four dNTPs or each dNTP individually present at a 100 µM concentration. The template sequence is indicated on the left (pr, primer). The primer-template structure is shown below each gel.

The ability of pol  to extend from a mispaired primer terminus, in contrast, clearly depends upon the template base 5' to the mismatch (Figs. 2 and 3). Like other "normal" polymerases (43), pol  generally extends purine:pyrimidine (Pu:Py) mispairs more efficiently than Pu:Pu or Py:Py mispairs. When the next template base is T, extension from the mispaired primer is inefficient and inaccurate. In most cases, incorporation of dAMP, dGMP, and dTMP occurs at low levels (Fig. 2, II, III, and IV). The sole exception is the G:T mismatch, which is extended relatively efficiently by the formation of A:T, T:T, G:T, and, to a lesser extent, C:T base pairs (Fig. 2A, III). In general, extension of mispairs appears considerably more efficient and accurate when A is the target template base (Fig. 3, II, III, and IV). Under these conditions, misincorporation of dGMP, dAMP, and dCMP is virtually undetectable. The only notable exceptions are faint products corresponding to A:A and G:A mispairs formed by extension from a G:T primer-template mismatch (Fig. 3A, III). Therefore, these studies reveal that in both sequence contexts, the G:T mispair, which is formed by pol  in preference to the correct A:T base, is also easily extended by the enzyme.

Analysis of Mismatched and Matched Primer Extension Efficiencies Using Steady-state Kinetics-- By using a steady-state gel kinetic assay (41), we have characterized the catalytic efficiency (V_max/K_m) and fidelity of extension from correctly paired and mispaired bases. In particular, we assayed the efficiency of extension by incorporation of dTMP opposite template A. Under these conditions, the efficiency of dTMP incorporation from a correctly paired G, A, or T primer terminus was similar to that reported previously (12). However, the V_max/K_m for incorporation of dTMP was roughly 2.5-fold lower when extending from a C primer terminus (Table I). Nucleotide misinsertion leads to the creation of a distorted primer-template termini that is geometrically less favored than a correct base pair. As a result, the catalytic efficiencies of mismatch extension are much smaller than those observed when extending matched primer ends and vary over a 100-fold range. The relative mismatch extension efficiencies, f⁰_ext (Table I), vary between 3 × 10⁴ (T:T extension) and 8.4 × 10² (T:G extension) and are considerably higher for Pu:Py mispairs compared with Pu:Pu and Py:Py mispairs, as anticipated from the mismatch extension gel profiles at a single dNTP concentration (Fig. 3).

View this table: [in this window] [in a new window]   Table I Steady-state kinetic analysis of dTMP insertion opposite A using DNA substrates with matched and mismatched primer-template termini Standing start kinetic assays were performed for 2 min using 10 fmol of pol  and 100 fmol of the matched and mismatched primer-templates shown in Fig. 1A. dTTP concentrations ranged from 0.004 to 10 µM for the matched primer-templates and templates with G:A, T:A, and C:A mismatched primer-templates. For the remaining mismatched primer-templates, dTTP concentrations ranged from 5 to 320 µM. The nucleotide insertion specificity (Vmax/Km) for dTMP incorporation by pol  was determined using Hanes-Woolf plots. The efficiency of mismatch extension relative to the extension of correctly paired primers was determined as f0ext = (Vmax/Km)mismatch/(Vmax/Km)match. Data for Vmax/Km are means (± standard error) from three to five experiments.

Pol -dependent Extension of a G:T Mispair, a Special Case-- Based on qualitative data (Fig. 3), extension of most mispairs was generally accurate when the template base was A. The exception was the G:T mispair, where significant extension occurred by misincorporation of dAMP and, to a lesser extent, dGMP opposite A (Fig. 3, III). By comparison, extension of 11 of 12 mispairs was inefficient and inaccurate when the 5' template base was T (Fig. 2). Again, the exception was extension from the G:T mispair, where extension was robust. In addition, based on the amount of primer utilization, dAMP appeared to be incorporated more often opposite template T than did the other three nucleotides (Fig. 2, III).

In the kinetic experiments, extension from the Watson-Crick A:T base pair favored the misincorporation of G opposite an adjacent template T (V_max/K_m = 5.4% µM¹ min¹) by 11.3-fold over the correct incorporation of A (V_max/K_m = 0.48% µM¹ min¹) (Table II). This result, which has been reported previously (12-14), serves to distinguish pol  from all other polymerases. We have also verified that formation of a T:T mispair is reduced only 2-fold relative to formation of an A:T base pair and that extension from an A:T base pair is significantly more accurate when the target template base is A (Table II).

View this table: [in this window] [in a new window]   Table II Kinetic analysis of nucleotide incorporation on DNA templates with matched and mismatched primer termini Standing start kinetic assays were performed for 2 min using 10 fmol of pol  and 100 fmol of the matched and mismatched primer-templates shown in Fig. 1A. dATP concentrations ranged from 2.5 to 320 µM; dTTP concentrations ranged from 10 to 320 µM; dGTP concentrations ranged from 0.25 to 4 µM for the matched primer-templates and from 20 to 320 µM for the mismatched primer-templates. The insertion efficiency (Vmax/Km) was determined using Hanes-Woolf plots. The misincorporation specificity (finc) was determined as the ratio of incorrect to correct nucleotide incorporation specificities. The efficiency of mismatch extension relative to the extension of correctly paired primers was determined as f0ext = (Vmax/Km)mismatch/(Vmax/Km)match. Data for Vmax/Km are means (± standard error) from three to five experiments.

The "special case" arises when extension takes place from a G:T mismatch by the incorporation of nucleotides opposite an adjacent template T. An original 11.3-fold preference favoring the incorporation of G rather than A opposite T from a matched primer terminus has now switched to a 3.3-fold preference favoring the incorporation of A rather than G opposite T from the mismatched primer (Table II). In other words, formation of a wobble G:T mispair is favored from a correctly paired primer, but formation of a Watson-Crick A:T base pair takes place from the G:T mispair. A similar change in pol  insertion specificity opposite template T, depending on whether the primer-template terminus was matched or mismatched, was also recently observed with a gapped DNA substrate (38). When the template base 5' to the G:T mispair is A, incorporation of dTMP is favored over the incorrect base, but "only" by a factor of ~100-fold as compared with 2000-fold when the primer is correctly paired (Table II).

The efficiencies of mismatch extension relative to the extension of correctly paired primers (f⁰_ext; Table II) using the next correct as well as incorrect nucleotide are remarkably high for both TA-5' and TT-5' template sequences. When the target template base is T, the frequency of mismatch extension using the wrong base, G, is 36-fold lower than that seen by using the correct base, A. By comparison, the frequency of mismatch extension by the incorporation of T opposite T is even slightly greater than the incorporation of A opposite T. Interestingly, the frequency of G:T mismatch extension on templates with a TA-5' sequence is 23-fold greater for the incorporation of the wrong base, A, as compared with incorporation of the correct base, T (Table II).

These peculiar mismatch extension specificities result in a large number of tandem misincorporation events. The formation of the tandem misincorporations with respect to the V_max/K_m values for each nucleotide incorporation and extension step is illustrated in Fig. 4. The values for overall efficiency of the two-step process are underlined. We estimated previously (12) that pol  mainly catalyzes the addition of 1 base/primer-template binding event. Thus, we feel that it is a reasonable approximation to calculate the efficiency of inserting two consecutive bases as the product of the two individual relative apparent V_max/K_m values for each step because they occur essentially independently of each other. The relative probability with which pol  synthesizes dinucleotide base pairs containing single or tandem mispairs compared with a perfectly match dinucleotide base pair is shown in Fig. 4 in parentheses. The first thing to note is that the frequency of dGMP misinsertion opposite template T is ~1.7-fold greater for the TA-5' sequence context compared with the TT-5' context (cf. f_inc = 9.7 (Fig. 4A) versus f_inc = 5.7 (Fig. 4B)). Extension of the resulting G:T mispair proceeds most efficiently by the incorporation of the correct base T opposite template A (V_max/K_m = 2.4% µM¹ min¹; Fig. 4A) and the correct base A opposite template T (V_max/K_m = 0.14% µM¹ min¹; Fig. 4B). Although the efficiency of the G:T mispair extension is ~17-fold higher when the template target base is A compared with T, and the overall efficiency of incorporating GT opposite TA is 13-fold higher than the efficiency of incorporating GA opposite TT (2.3 versus 0.18), the actual probability that pol  will incorporate and extend a G:T mismatch compared with the incorporation of two complementary Watson and Crick base pairs is, remarkably, ~5.3-fold higher for the DNA template with the TT sequence compared with DNA template with the TA sequence (cf. 1.6 versus 0.3; Fig. 4). The explanation for this "nonintuitive" result is based upon the following rationale: the catalytic efficiency of incorporating an A followed by a T opposite TA-5' is actually ~3-fold higher than the misincorporation of G and its subsequent correct extension (cf. 8.1 versus 2.4; Fig. 4A). These differences are largely achieved by the fact that although misincorporation of G is favored over that of A by ~10-fold, the correctly paired A:T primer is extended ~40-fold better than the G:T mispair. Thus, although the G:T mispair is extended most efficiently in the TA-5' sequence, the likelihood that the final product of (mis)incorporation and extension will contain a mutation within this sequence context is only 0.3 relative to the incorporation and extension of two complementary bases. Conversely, incorporation of two As opposite TT-5' only occurs with an overall efficiency of 0.11 as compared with an efficiency of 0.18 for the misincorporation of G opposite T and its correct extension. (Fig. 4B). As a consequence, misincorporation of G opposite T will be more likely "fixed" as a mutation at a template TT-5' sequence than at a template TA-5' sequence. Incorporation of tandem GT and GG mispairs at template TT-5' also occurs with greater probability than misincorporation of G followed by T within the TA-5' context (Fig. 4). Overall, therefore, the relative probability that pol  will fix misincorporations as mutations occurs in the following order: the most prevalent event will be the incorporation of G opposite T and its correct extension at template T (by a factor of ~1.6 relative to the incorporation and extension of two complementary base pairs). Next, we would expect a tandem misincorporation of GT opposite TT-5' (which will occur with approximately the same relative probability as the incorporation of two As). Finally, the GG mispair will occur with slightly greater probability at template TT-5' than the single misincorporation of G opposite T and its correct extension at template TA-5' (0.5 versus 0.3 relative probability) (Fig. 4).

View larger version (16K): [in this window] [in a new window]   Fig. 4.   Comparison of overall efficiency of nucleotide insertion opposite the template T and extension of matched or mismatched primers in different sequence contexts. The efficiency of nucleotide incorporation (Vmax/Km expressed as the percentage of primer elongation product per minute per µM nucleotide) for each step is indicated above each arrow. The target template base in each reaction is shown in bold. The efficiency of nucleotide incorporation opposite the 3'-template T was determined in 2-min kinetic experiments using dATP at concentrations ranging from 2-64 µM and dGTP at concentrations ranging from 0.5-16 µM. The efficiency of nucleotide incorporation opposite the 5'-template T (A) and 5'-template A (B) is taken from Table II. The value for the overall catalytic efficiency of primer elongation by two bases (calculated as the product of the individual values obtained for Vmax/Km incorporation opposite the 3' and 5' bases) are underlined. The relative probability of forming each dinucleotide pair relative to the perfectly matched complementary bases is shown in parentheses. From these numbers, one can see that although extension of the G:T mispair within the TA-5' sequence is the most catalytically favorable, the greatest probability that the G:T mispair will be generated and fixed as a mutation by pol  actually occurs within the TT-5' context.

DNA Sequence Context Influence on Pol -catalyzed Primer Extension-- Pol  extends all 12 mispairs, but the extension efficiencies depend on two key factors, the identity of the mispair at the 3'-primer end (Table I) and the DNA template sequence context. This can be seen clearly when one compares replication products in Figs. 2, I and 3, I, where the pause at template T varies considerably in the presence of all four dNTPs. These observations suggest that elongation of a mismatched primer-template terminus may occur more efficiently in a TA-5' template sequence (Fig. 3) than in a TC-5' sequence (Fig. 2). Alternatively, it is possible that pol  may be more prone to making G:T mismatches in a "standing start" mode of replication compared with "running start" synthesis. To distinguish between these two possibilities, two DNA primer-templates were utilized (Fig. 1, B and C). The primer-template DNA used for running start synthesis has a template T situated 3 bases from the end of the primer with either A or C as the next template base (Fig. 1B, n = A or C). The results show clearly that pausing opposite the +3T site is much weaker when T is followed by A rather than C (Fig. 5A). The primer-template DNA used for standing start synthesis has a template T adjacent to the primer terminus and contains either C, G, T, or A as the next template base (Fig. 1C). Yet again, we find that elongation after the incorporation of a dNMP moiety opposite template T clearly depends on the base 5' to the T and proceeds with a relative efficiency of A > G > T  C (Fig. 5B). The integrated band intensities of the replication products past template T are 58.9 ± 5.0% for 5'A, 20.9 ± 2.4% for 5'G, 9.8 ± 1.4% for 5'T, and 7.3 ± 1.6% for 5'C (data are the means ± S.E. from four experiments similar to the one shown in Fig. 5B). Thus, the pol 's pausing at template T clearly depends on the template sequence context and cannot be attributed to differences in the efficiency of misincorporating dGMP opposite T in running and standing start synthesis modes of replication.

View larger version (45K): [in this window] [in a new window]   Fig. 5.   The effect of sequence context on pol -dependent primer extension. A and B, the nearest 5' neighbor influence on the pause site opposite template T was studied using the DNA templates shown in Fig. 1, B and C. C and D, the efficiency of T:T mispair elongation in different sequence contexts was studied using templates shown in Fig. 1, D and E. E and F, templates shown in Fig. 1, F and G, were used to study the distal effect of T:T mispairs on primer elongation. Primer extension studies were performed for 15 min using 100 fmol of DNA template, 25 fmol of pol , and dNTP present at a 100 µM concentration. The template sequence is indicated on the left, and the template structure and varying base are shown below each panel. These studies indicate that the extent of synthesis by pol  is exquisitely sequence context-dependent and that a mispair "buried" 2-3 bases from the primer terminus may be sufficient to terminate further elongation by pol .

Primer Destabilization by a One-base Buried T:T Mismatch-- Formation of T:T mispairs by pol  occurs almost as well as A:T base pairs (~ 70% efficiency; Ref. 12), yet mismatch extension is poor when the 5'-template base is T (Fig. 2A, II, lane 4) compared with extension of either A:T or G:T pairs (Fig. 2A, I and III, lane 4). The DNA templates shown in Fig. 1, DG, were used to investigate the extent to which a T:T mismatch inhibits continued elongation when present either at a primer-3'-terminus or "buried" 1 base from a correctly matched primer terminus. Extension of a T:T mispair by the incorporation of A opposite T is largely unaffected by the presence of either C or A immediately downstream from the template T site (Fig. 5C). However, the T:T mispair is elongated much more efficiently when the template base adjacent to the mispair is A (Fig. 5D) rather than T (Fig. 5C), and the pattern of mispair extension in this case depends on the sequence downstream of A. In the TAA-5' context, the favored product contains the addition of two Ts downstream from the mismatch, whereas in the TAC-5' context, the elongated primer product contains mostly single additions of T opposite A with much fewer further additions of G opposite C.

When pol  is presented with a preformed A:T (Fig. 5E) or T:A (Fig. 5F) base pair adjacent to the T:T mispair, both matched primers are further extended by the addition of a single nucleotide, G opposite C or T opposite A, with greater primer elongation to some extent when the next template base is A. However, further extension is inhibited in all four cases (Fig. 5, E and F). The key conclusion to be drawn from this experiment is that even when moderately efficient mispair extension occurs via incorporation of Watson-Crick base pairs, a "buried" mispair located 2-3 bases from the nascent primer chain is sufficient to forestall extensive pol -dependent chain elongation and hence promiscuous template mutagenesis, even in the presence of high dNTP concentrations.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Human pol  exhibits a number of peculiar biochemical hallmarks. Its fidelity can be extremely poor when copying undamaged DNA templates, depending on the template base copied. It strongly favors forming G:T rather than A:T base pairs by factors ranging between 3- and 11-fold (Table II; Fig. 4) (12-14, 38), depending on sequence context. No other known polymerase favors forming base mispairs over canonical Watson-Crick base pairs. As an added feature, pol  forms T:T almost as well as A:T base pairs (12). However, the specificity of pol  changes when bound to a mismatched G:T primer terminus, preferring to make A:T pairs ~3-fold more frequently than G:T mispairs (Table II) (38). Pol  also behaves in an unusual manner when catalyzing Watson-Crick base pairs by forming T:A base pair by at least an order of magnitude more efficiently than either A:T, C:G, or G:C base pairs (12-14). Furthermore, replication of template A is generally accurate, with misincorporations occurring with a frequency of 1-2 × 10⁴, a frequency similar to or perhaps only slightly lower than that of other DNA polymerases lacking exonucleolytic proofreading (44, 45). Overall, pol  fidelity varies ~10,000-fold, depending on the template base copied (12). Reminiscent of two other members of the Y family (UmuC/DinB/Rev1/Rad30) of polymerases, E. coli pol V and human pol , pol  is also able to incorporate nucleotides opposite replication-blocking DNA lesions (13, 14, 31-33). However, in contrast to pol V (5-7, 25, 26) and pol  (9, 10), for the damaged DNA tested to date, pol  has significant difficulty synthesizing past the lesion sites and exhibits only a limited ability to bypass a cis-syn TT dimer (31).

The previous findings suggest that the active site of pol  can accommodate aberrant base pairing geometries when inserting nucleotides at specific template sites such as T but might be considerably less accommodating when attempting to extend distorted primer ends. Alternatively, because pol  exhibits low fidelity when copying T but not A, it might also exhibit selective mismatch extension, extending some distorted ends more efficiently than others. To examine these issues, we measured the relative extension efficiency and fidelity for the 12 possible mismatched and 4 correctly matched base pairs in different sequence contexts.

Pol -catalyzed Misincorporation and Mismatch Extension Behavior-- A compilation of pol -catalyzed misincorporation and mismatch extension efficiencies is shown in Fig. 6, where points above the dashed line correspond to a higher relative efficiency of mispair extension compared with the frequency of nucleotide misinsertion (f⁰_ext > f_inc), whereas points below the dashed line represent the reverse situation. The mispair formed with highest efficiency, the wobble G:T (f_inc = 11), is readily extended (~ 3% as efficiently as A:T). The T:G mismatch is formed with high frequency (f_inc = 0.13) and is extended even more easily than the G:T mispair (f⁰_ext = 8.4%). In general, a rough correlation exists between the relative efficiency of mismatch formation and extension; however, there are several notable exceptions. The T:T mispair is formed with the second highest efficiency (f_inc = 0.7) but is poorly extended (f⁰_ext = 3 × 10⁴). The converse is true for the inefficiently formed C:A mispair (f_inc = 10⁴), which is easily extended (f⁰_ext = 4 × 10²). Thus, the general behavior of pol  is clearly distinct from that of pol  (13, 46) and from that of avian myeloblastosis virus and HIV-1 reverse transcriptases, which favor mispair extension over nucleotide misincorporation (f⁰_ext > f_inc) for almost all mispairs (43). Our study also suggests that in general, pol  extends most mispairs less efficiently than that reported for Saccharomyces cerevisiae pol  but to some extent more efficiently than that reported for human pol  (47).

View larger version (15K): [in this window] [in a new window]   Fig. 6.   Comparison of nucleotide misincorporation and mispair extension efficiencies by pol . Efficiencies of mispair extension (f0ext) taken from Table I were plotted versus efficiencies of misincorporation (finc) taken from Ref. 12. The dashed line corresponds to f0ext = finc. In general, pol  appears able to extend mispairs more efficiently than the related human pol  but to a lesser extent than S. cerevisiae pol .

Both, f_inc and f⁰_ext, determine the probability that nucleotide misincorporation will be "fixed" as a mutation. As we have shown above, the efficiency of pol -catalyzed mismatch extension depends not only on the identity of primer-template mispairs but also on the DNA template sequence context. One of the manifestations of this dependence is the increased probability of fixing the favored G:T mispair as a mutation when T is the 5'-template base. Within this sequence context, misincorporation of a G:T mispair followed by incorporation of the correct A:T base pair occurs with a probability ~1.6-fold greater than the incorporation of two complementary A:T base pairs (Fig. 4B). There is also an excess of pol -catalyzed tandem misincorporations occurring as a consequence of the unusual misincorporation and mismatch extension specificities (Fig. 4). GT/TT tandem mismatches occur slightly more often than the AA/TT correct matches, and GG/TT mismatches, although less frequent, nevertheless occur half as often as the correct AA/TT base pairs (Fig. 4B). In contrast, tandem mismatches occur much less frequently at TA-5' template sites because of the strong preference of pol  for inserting T opposite A (Fig. 4A).

Sequence Context Effects on Mismatch Extension-- Zhang et al. (14) have recently reported that synthesis by pol  frequently terminates opposite template T. This process, termed "T-stop," is hypothesized to occur by the inefficient extension of the favored G:T mismatch (14). However, our data demonstrate that although incorporation of G opposite T causes a reduction in catalytic efficiency, it is not an "impenetrable" block to further replication. On the contrary, pol  extends all 12 mispairs, but the efficiency of extension depends on the DNA template sequence context (Figs. 2, 3, and 5). One observes, for example, a distinctive replication pause site occurring opposite a template T when the 5' base is C (Fig. 5B), but the pause band is reduced substantially when the 5' base is A (Fig. 5B).

The data also reveal that pol -dependent elongation after the mismatch is often inhibited after the addition of one or two next correct nucleotides. This effect is largely caused by the mismatch itself, as well as its location at positions distant from the primer terminus, rather than by further misincorporations. For example, a T:G mismatch in a GT-5' sequence (Fig. 2D, III) is elongated by several base pairs, the first of which is likely to be erroneous, whereas accurate extension of an A:C mismatch in a CA-5' sequence context (Fig. 3C, III) results in elongation by just a single base pair. Thus, pol -catalyzed mutations are likely to be confined to very short template regions even in the presence of high dNTP levels.

Speculations on Biological Roles for Pol -- The biochemical properties of pol  and its distributive synthesis (12), coupled with selectively compromised nucleotide insertion and mismatch extension fidelity, suggest that its action is likely to be confined to short regions of DNA, and one possible function might be the generation of highly localized mutations during somatic hypermutation. Recently, pol  was shown to contain a lyase activity (38), suggesting that the enzyme may be involved in a specialized form of base excision repair. Indeed, one can imagine a scenario in which it is necessary to incorporate a G across from an altered template site that originally contained Cdeamination of 5-methyl cytosine is an obvious example (38). Of course, given its relationship with other members of the Y family of polymerases (3), which are best characterized for their roles in translesion replication, it is also possible that pol  is geared to a specific role in translesion replication of lesions other than those tested to date.

Indeed, a similar degree of "ignorance" also applies to most of the other members of the Y family of DNA polymerases. An entirely distinct level of biochemical complexity concerns mechanisms of enzyme trafficking and targeting: namely, how are one or another of the polymerases chosen for a specific biological task, and once one is chosen, how is the template target site located? Although the specific biochemical characteristics of pol  and the other errant polymerases are obviously important in understanding their cellular functions, in the absence of direct biological data, possible roles for many of these polymerases remain entirely speculative. One thing that is crystal clear, however, is that unless strictly regulated, pol  and its cohort polymerases will cause excessive levels of spontaneous mutations in vivo.

	ACKNOWLEDGEMENTS

We thank members of the Section on DNA Replication, Repair, and Mutagenesis for constructive comments during the course of this work.

	FOOTNOTES

^* This work was supported by the National Institutes of Health Intramural research program (A. V., A. T., E. G. F., and R. W.) and by National Institutes of Health Grants GM21422 and GM42554 (to M. F. G.).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.

^§ Present address: UPR 9003, CNRS, Cancérogenèse et Mutagenèse Moléculaire et Structurale, ESBS, Blvd. S. Brant, 67400 Strasbourg, France.

To whom correspondence should be addressed: Section on DNA Replication Repair and Mutagenesis, Bldg. 6, Rm. 1A13, National Institute of Child Health and Human Development, National Institutes of Health, 9000 Rockville Pike, Bethesda, MD 20892-2725. Tel.: 301-496-6175; Fax: 301-594-1135; E-mail: woodgate@helix.nih.gov.

Published, JBC Papers in Press, June 11, 2001, DOI 10.1074/jbc.M102694200

	ABBREVIATIONS

The abbreviations used are: pol, polymerase; dNTP, deoxynucleotide triphosphate.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				J.-C. Walser, L. Ponger, and A. V. Furano CpG dinucleotides and the mutation rate of non-CpG DNA Genome Res., September 1, 2008; 18(9): 1403 - 1414. [Abstract] [Full Text] [PDF]

				E. G. Frank and R. Woodgate Increased Catalytic Activity and Altered Fidelity of Human DNA Polymerase {iota} in the Presence of Manganese J. Biol. Chem., August 24, 2007; 282(34): 24689 - 24696. [Abstract] [Full Text] [PDF]

				A. J. Picher, M. Garcia-Diaz, K. Bebenek, L. C. Pedersen, T. A. Kunkel, and L. Blanco Promiscuous mismatch extension by human DNA polymerase lambda Nucleic Acids Res., June 28, 2006; 34(11): 3259 - 3266. [Abstract] [Full Text] [PDF]

				W. T. Wolfle, R. E. Johnson, I. G. Minko, R. S. Lloyd, S. Prakash, and L. Prakash Replication past a trans-4-Hydroxynonenal Minor-Groove Adduct by the Sequential Action of Human DNA Polymerases {iota} and {kappa} Mol. Cell. Biol., January 1, 2006; 26(1): 381 - 386. [Abstract] [Full Text] [PDF]

A. E. Vidal, P. Kannouche, V. N. Podust, W. Yang, A. R. Lehmann, and R. Woodgate Proliferating Cell Nuclear Antigen-dependent Coordination of the Biological Functions of Human DNA Polymerase {iota} J. Biol. Chem., November 12, 2004; 279(46): 48360 - 48368.