Identification of a Mutant DNA Polymerase in Saccharomyces cerevisiae With an Antimutator Phenotype for Frameshift Mutations

Corresponding author: Linda J. Reha-Krantz, Department of Biological Sciences, CW405 BioSciences Bldg., University of Alberta, Edmonton, Alberta T6G 2E9, Canada., lreha{at}gpu.srv.ualberta.ca (E-mail)

Communicating editor: S. SANDMEYER

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

We propose that a ß-turn-ß structure, which plays a critical role in exonucleolytic proofreading in the bacteriophage T4 DNA polymerase, is also present in the Saccharomyces cerevisiae DNA pol . Site-directed mutagenesis was used to test this proposal by introducing a mutation into the yeast POL3 gene in the region that encodes the putative ß-turn-ß structure. The mutant DNA pol has a serine substitution in place of glycine at position 447. DNA replication fidelity of the G447S-DNA pol was determined in vivo by using reversion and forward assays. An antimutator phenotype for frameshift mutations in short homopolymeric tracts was observed for the G447S-DNA pol in the absence of postreplication mismatch repair, which was produced by inactivation of the MSH2 gene. Because the G447S substitution reduced frameshift but not base substitution mutagenesis, some aspect of DNA polymerase proofreading appears to contribute to production of frameshifts. Possible roles of DNA polymerase proofreading in frameshift mutagenesis are discussed.

DNA polymerases replicate DNA with high accuracy due to their ability to discriminate between "right" and "wrong" nucleotides in the nucleotide incorporation reaction and due to exonucleolytic proofreading, which removes misinserted nucleotides. We used a genetic approach to study proofreading by the bacteriophage T4 DNA polymerase, a model for a large family of DNA polymerases that includes the human DNA polymerase (SPICER et al. 1988 ; TSURIMOTO and STILLMAN 1990 ). Several T4 DNA polymerase mutations were identified as suppressors of the excessive proofreading activity produced by antimutator DNA polymerases (STOCKI et al. 1995 ). The proofreading suppressor mutants are called "active-site-switching" mutants because they affect the ability of the T4 DNA polymerase to switch between polymerase and exonuclease activities. The goal of studies presented here is to identify active-site-switching mutants in other DNA polymerases, specifically the DNA pol of Saccharomyces cerevisiae.

The most frequently identified T4 DNA polymerase active-site-switching mutant has a serine substitution for glycine at position 255, the G255S substitution. The T4 G255S-DNA polymerase and other active-site-switching mutants are not defective in cleaving the phosphodiester bond; instead, these mutants are impaired in one or more earlier steps in the proofreading pathway that prepare the primer terminus for the hydrolysis reaction. Proofreading begins with recognition of a mismatched base pair in the polymerase active center and then the primer end with the incorrect nucleotide is separated from the template strand and transferred to the exonuclease active center where hydrolysis takes place. The G255S-DNA polymerase is slow to form the partially strand-separated editing complex compared to the wild-type T4 DNA polymerase (MARQUEZ and REHA-KRANTZ 1996 ; BAKER and REHA-KRANTZ 1998 ). Structural studies show that residue G255 resides in the loop of a ß-turn-ß structure in the exonuclease domain of the T4 DNA polymerase (WANG et al. 1996 ) and a glycine residue is also found in the same position in the closely related bacteriophage RB69 DNA polymerase (SHAMOO and STEITZ 1999 ). We proposed that the ß-hairpin structure may act as a wedge between the template and primer strands to drive the equilibrium between base pairing and strand separation toward denaturation (MARQUEZ and REHA-KRANTZ 1996 ). This proposal was confirmed recently by structural studies of the RB69 DNA polymerase editing complex (SHAMOO and STEITZ 1999 ). The ß-hairpin projects into the junction between the template and primer strands in position to stabilize the partially strand-separated structure.

The participation of the ß-hairpin structure in forming the strand-separated editing complex is a critical step in the T4 DNA polymerase proofreading pathway. The reduced ability of the T4 G255S-DNA polymerase to form editing complexes means reduced proofreading and increased DNA replication errors. A strong mutator phenotype is observed for the T4 G255S-DNA polymerase in vivo, similar in magnitude to mutant T4 DNA polymerases with amino acid substitutions in the exonuclease active center that prevent the hydrolysis reaction (REHA-KRANTZ 1988 ; STOCKI et al. 1995 ). Thus, inhibition of an early step in the proofreading pathway can be nearly as effective in inhibiting proofreading as prevention of the hydrolysis reaction.

Given the importance of the ß-hairpin structure for proofreading by the T4 DNA polymerase, we have carried out a series of genetic experiments with the DNA pol of S. cerevisiae to determine if this DNA polymerase has a similar structure that functions in proofreading. Eukaryotic DNA pol is the major replicative DNA polymerase in the cell and DNA pol also functions in DNA repair (WAGA and STILLMAN 1998 ). Thus, studies reported here can provide information about the mechanism for production of DNA polymerase-induced errors during chromosome replication and during repair of damaged DNA.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Protein sequence alignments:
DNA polymerase protein sequence comparisons were performed by using the PepTools software (BioTools, Edmonton, Alberta, Canada).

Construction of yeast strains:
All yeast strains used for this study were derived from MS71 (STRAND et al. 1995 ), a Leu⁺ derivative of AMY125 ( ade5-1 leu2-3,112 trp1-289 ura3-52 his7-2), obtained from A. Morrison and A. Sugino, Osaka University, Osaka, Japan. Strains EAS74 (msh2) and EAS38 (msh6) were described previously (SIA et al. 1997 ). EAS56 is a MATa derivative of EAS74 and EAS48 is a MATa derivative of EAS38.

MIH1 (pol3-447) was constructed by first subcloning 2.2 kb of the 5' end of the POL3 (DNA pol ) gene from the full-length gene. The full-length gene was obtained from a plasmid provided by Peter Burgers (St. Louis). Greater plasmid stability in bacteria was obtained with this partial POL3 construct, which contains the exonuclease domain. Site-directed mutagenesis to produce the G447S substitution was done by the method of KUNKEL 1985A . Sequencing was done to confirm the presence of the mutation and then the mutant pol3 gene fragment was subcloned into the yeast-integrating plasmid, YIp5. A yeast strain expressing the mutant allele pol3-447 was created by targeted integration of the mutant pol3 gene fragment into the chromosomal POL3 gene of strain MS71 (ROTHSTEIN 1983 ). Southern blotting was used to confirm integration of the plasmid in the transformant and restoration of a single gene copy after recombination. The DNA pol gene was sequenced to verify the presence of the mutation that encodes the G447S substitution and the absence of other mutations (HADJIMARCOU 1999 ).

RJK224 (pol3-447 msh2) was derived by sporulating the diploid formed by crossing haploid strains MIH1 and EAS56. RJK346 (pol3-447 msh6) was derived by sporulating the diploid formed by crossing the haploid strains MIH1 and EAS48. The msh2 and msh6 genotypes were identified by PCR analysis as described previously (SIA et al. 1997 ). Spore colonies were analyzed for the presence of the POL3 genotype by two separate PCR reactions. One reaction, capable of amplifying the POL3 sequence but not the pol3-447 sequence, utilized an upstream primer from open reading frame positions 1318 to 1340 of POL3 (5' GTGTTCTCTTCGAAGGCTTATGG) and a downstream primer from positions 1860 to 1836 (5' GTGCGCCATCATAATACTTGGATAT). The 3' terminal GG nucleotides of the upstream primer (shown in boldface) correspond to the nucleotides that are altered to TC in the pol3-447 allele. Conversely, the second PCR reaction utilized the same downstream primer as the first reaction with an upstream primer specific to the pol3-447 sequence (5' GTGTTCTCTTCGAAGGCTTATTC).

Determination of spontaneous mutation rates and mutational spectra in yeast:
The forward mutation rate at the CAN1 locus was determined by innoculating 20 colonies of each strain into 5 ml of liquid growth medium (YPD) and growing the cultures to saturation. Each culture was washed once with 1 ml of sterile distilled water. The number of Can^r mutants was determined by plating appropriate dilutions onto solid synthetic minimal medium lacking arginine and containing 60 mg/liter canavanine. Each culture was also titered onto complete minimal medium without canavanine to determine the viable cell count.

Cultures were prepared in a similar manner for determining the rate of his7-2 reversion. The number of His⁺ revertants was determined by plating appropriate dilutions onto minimal medium lacking histidine and the viable cell count was determined by plating onto complete minimal medium.

Mutation rates for each experiment were determined by using the method of the median (LEA and COULSON 1949 ). The rates reported for canavanine resistance represent the averages from two experiments using 20 independent cultures each. The rates reported for His⁺ reversion were based on the results of one 20- and one 10-culture experiment. Statistical comparisons were performed as described (WIERDL et al. 1996 ).

The mutations present in individual Can^r isolates were identified by sequencing a PCR-amplified copy of the 1.8-kb CAN1 gene made by using the primers CAN1UP (5' CAGAGTTCTTCAGACTTC) and CAN1DOWN (5' AGGGTGAGAATGCGAAAT). Three separate primers were used to sequence this fragment: CAN1583C, corresponding to nucleotides 602 to 583 of the CAN1 open reading frame (5' AGTGGAACTTTGTACGTCCA); CAN1552, corresponding to nucleotides 533 to 552 (5' CAATCACTTTTGCCCTGGAA); and CAN1DOWN (sequence given above).

The sequences of His⁺ revertants were determined by sequencing a PCR fragment containing this region. This fragment was amplified using primer F1, corresponding to positions (-56) to (-34) relative to the HIS7 open reading frame (5' GAAGTAGCAGTATCCAGTTTAGG), and primer R1, corresponding to positions 886 to 865 (5' ATGTTACTTCATCCGCACCCTG). Sequencing was performed using the R1 primer.

Determination of spontaneous mutation frequencies in bacteriophage T4:
Bacterial and T4 strains and culture conditions have been described (REHA-KRANTZ and NONAY 1993 ; STOCKI et al. 1995 ). DNA replication fidelity was determined by reversion tests using rII mutations that revert by defined mutational pathways. The rUV199oc mutant reverts primarily by an AT GC transition mutation (REHA-KRANTZ 1995 ); the rIIP7oc mutant reverts by AT TA or AT CG transversion mutations (RIPLEY 1975 ); and the rII131 and rII117 mutants revert primarily by +1 frameshift mutations in tracts of five As, (A)₅ (A)₆, to restore the wild-type sequence (PRIBNOW et al. 1981 ; data reported here). The rII131 and rII117 alleles are the classic hotspot sites in the rII genes (BENZER 1960 ). Ten or more independent phage cultures for each of the DNA polymerase-rII combinations were grown for each reversion test. The cultures were titered for the total number of phage and for the number of rII⁺ revertants. The median values are reported. The sequence of the rII131 and rII117 revertants was done by sequencing the transcripts (MCPHEETERS et al. 1986 ) using the primer (5' GAAACAATACGGAATTTCTTGG), which anneals 26 nucleotides from the rII131 site, and the primer (5' CCTTCAATTCGAACATCGCC), which anneals 35 nucleotides from the rII117 site.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Identification of the G255-type ß-hairpin in other DNA polymerases by protein sequence comparisons:
From protein sequence similarities we proposed that the G255 ß-hairpin in the T4 DNA polymerase may be conserved in the yeast and human DNA pol 's (STOCKI et al. 1995 ; MARQUEZ and REHA-KRANTZ 1996 ). This proposal was strengthened by recent structural studies of the thermal stable archaeal DNA polymerase from the Desulfurococcus strain TOK (ZHAO et al. 1999 ). A ß-hairpin is present in the exonuclease domain of the DTOK DNA polymerase in the same relative position as observed for the T4 and RB69 DNA polymerases. Protein sequence comparisons of the T4, RB69, and DTOK DNA polymerases using the PepTools alignment program from BioTools produced the alignment shown in Fig 1. A glycine corresponding to G255 in the T4 DNA polymerase (indicated by a star in Fig 1) is found in the turn region of a ß-hairpin structure for all three DNA polymerases. Protein sequence comparisons were then extended to DNA polymerases for which there is no structural information at this time: the Vent DNA polymerase, Escherichia coli DNA pol II, and the S. cerevisiae and human DNA pol 's (Fig 1). The sequence alignments begin with a highly conserved exonuclease motif (SPICER et al. 1988 ; WANG et al. 1989 ; BRAITHWAITE and ITO 1993 ). The conserved Asp residue, residue D219 in the T4 DNA polymerase, is located in the exonuclease active center. Although there are only a few identical or similar residues in the sequences that follow the conserved exonuclease motif for the DNA polymerases compared, amino acid residues important for forming a ß-turn-ß structure appear to be conserved. A Gly residue (indicated by a star), corresponding to G255 in the T4 DNA polymerase, is present in the turn region of the RB69 and DTOK ß-hairpins and a Gly residue is also present in the proposed ß-hairpins in the Vent DNA polymerase, E. coli DNA pol II, and in the yeast and human DNA pol 's. The ß-hairpin is followed by a short ß-strand and two -helices; conserved Gly, Asp, and Leu residues, which demarcate these structures in the T4, RB69, and DTOK DNA polymerases, appear to be present in the other DNA polymerases. In the RB69 DNA polymerase editing complex, residue Arg260 projects into the junction where the single-stranded primer terminus is separated from the template strand (SHAMOO and STEITZ 1999 ). A basic Arg or Lys residue is found at this position in the DNA polymerases compared, except for the Vent DNA polymerase, which has a Ser residue (Fig 1).

View larger version (26K): In this window In a new window Download PPT slide   Figure 1. DNA polymerase protein sequence alignments in the region of the ß-hairpin loop. Structure has been determined for the T4 (WANG et al. 1996 ), RB69 (WANG et al. 1997 ; SHAMOO and STEITZ 1999 ), and DTOK DNA polymerases (ZHAO et al. 1999 ). Helical structures are indicated by the and ß-strands are indicated by the ß. The sequences of the T4, RB69, and DTOK DNA polymerases are aligned with proposed similar sequences in the Vent DNA polymerase, the E. coli DNA pol II, and in the yeast and human DNA pol 's. The star indicates G255 in the T4 DNA polymerase. Apparent conserved amino acids found in all of the DNA polymerases compared are indicated by large boldface letters. Amino acids that are found in most of the DNA polymerases are underlined. The EXO designation indicates the conserved Asp residue that binds a divalent metal ion in the exonuclease active center. The R/K designation indicates the position of Arg260 in the RB69 DNA polymerase, which extends from the ß-hairpin into the junction between the separated primer and template strands in the editing complex.

As a first attempt to determine if the proposed protein sequence similarities mean similar function, the S. cerevisiae G447S-DNA pol was constructed by site-directed mutagenesis. Residue G447 in the yeast DNA pol appears to be analogous to G255 in the T4 DNA polymerase (Fig 1); thus, the yeast G447S-DNA pol was predicted to have reduced proofreading as observed for the T4 G255S-DNA polymerase. Construction of the pol3-447 allele to encode the yeast G447S-DNA pol was achieved by using standard site-directed mutagenesis procedures, which are described in MATERIALS AND METHODS section. Growth rates were the same for the mutant and wild-type yeast strains and the mutant strain was not temperature sensitive (HADJIMARCOU 1999 ). The G255S substitution in the T4 DNA polymerase also does not impair the polymerization reaction or confer a temperature-sensitive phenotype (STOCKI et al. 1995 ). DNA replication fidelity by the T4 G255S-DNA polymerase and the yeast G447S-DNA pol are described below.

Base substitution and frameshift mutagenesis by the T4 G255S-DNA polymerase and the exonuclease-deficient D324G-DNA polymerase:
Although a strong mutator phenotype was observed for the T4 G255S-DNA polymerase by previous reversion and forward mutation tests (REHA-KRANTZ 1988 ; STOCKI et al. 1995 ), the types of mutations produced were not fully characterized. DNA replication fidelity by the T4 G255S-DNA polymerase for specific types of mutations is reported in Table 1 and is compared to the replication fidelity by the exonuclease-deficient D324G-DNA polymerase, which has a glycine substitution for an essential aspartate in the exonuclease active center (REHA-KRANTZ and NONAY 1993 ). The hydrolysis reaction by the D324G-DNA polymerase is reduced 10⁴-fold compared to the wild-type T4 DNA polymerase (ELISSEEVA et al. 1999 ).

Transition and transversion mutation frequencies were increased 100- to 200-fold for the G255S- and D324G-DNA polymerases compared to the wild-type T4 DNA polymerase (Table 1). While very similar mutation frequencies were observed for the G255S- and D324G-DNA polymerases for base substitution errors, a lower frameshift mutation frequency was observed for the G255S-DNA polymerase compared to the D324G-DNA polymerase (Table 1). A 7-fold increase in +1 frameshifts compared to the wild-type DNA polymerase was detected at the rII131 site for the G255S-DNA polymerase and a 20-fold increase was observed at the rII117 site; however, much larger 150- to 200-fold increases were observed for the D324G-DNA polymerase. Some of the rII131⁺ and rII117⁺ revertants were sequenced. For the wild-type T4 DNA polymerase, 67% (14/21) of the rII131⁺ revertants were +1 frameshifts that expanded the run of five As to six, (A)₅ (A)_6, which restores the wild-type sequence. The other revertants were pseudorevertants with mutations at nearby sites: four +1 frameshift mutations, one -2 deletion, and two complex duplications. For the G255S-DNA polymerase, 94% (15/16) of the rII131⁺ revertants restored the wild-type sequence from (A)₅ (A)₆ and a similar number, 92% (11/12), was detected for the D324A-DNA polymerase. Less sequencing was done at the rII117 site, but 75% or more of the revertants were +1 frameshifts, (A)₅ (A)₆, that restored the wild-type sequence.

These results show that the G255S-DNA polymerase, which is defective in forming the partially strand-separated editing complex, is more prone to base substitution mutations than to frameshifts. The exonuclease-deficient D324G-DNA polymerase, on the other hand, which can form the editing complex but is defective in the hydrolysis reaction, is a strong mutator for both base substitution and frameshift mutations. Thus, amino acid substitutions that affect different steps of the proofreading pathway have different effects on base substitution and frameshift fidelity.

Frameshift fidelity by the S. cerevisiae wild-type and G447S-DNA pol 's:
DNA replication fidelity in yeast was determined by two methods: (1) reversion of the his7-2 allele, which reverts primarily by a +1 frameshift mutation to expand the tract of seven As to eight, (A)₇ (A)₈, which is the wild-type sequence (HADJIMARCOU 1999 ; SHCHERBAKOVA and KUNKEL 1999 ), and (2) by a forward mutation test for production of mutations that confer resistance to canavanine (Can^r). The Can^r mutation assay detects both base substitution and frameshift mutations since any mutation that inactivates the arginine permease will prevent the uptake of canavanine into the cell. Mutation rates were determined by the method of the median (LEA and COULSON 1949 ) from 30 or more independent cultures.

No significant differences in mutation rates for production of His⁺ revertants or Can^r mutants were observed between the wild-type and pol3-447 strains (Table 2). These mutation rates, however, do not provide a true measure of replication errors made by DNA pol since many errors will be corrected by postreplication mismatch repair, which is present in yeast and in many other organisms, but not in phage T4. Replication fidelity by the G447S-DNA pol was measured in an msh2 background since the Msh2 protein is required for repair of both single-base mispairs and insertion/deletion mispairs (MARSISCHKY et al. 1996 ). An 30-fold increase in the Can^r mutation rate was observed for the msh2 strain compared to the wild-type strain and a 170-fold rate increase was observed for reversion of the his7-2 allele (Table 2). Fewer Can^r mutations and his7-2 revertants were observed for the msh2 pol3-447 strain compared to the msh2 strain (Table 2); the Can^r mutation rate was 3-fold lower and the his7-2 rate was 5-fold lower. The lower mutation rates observed for the msh2 pol3-447 strain compared to the msh2 strain are significant by the Fisher exact test (P < 0.0001) for both the his7-2 reversion assay and for production of Can^r mutations.

View this table: In this window In a new window   Table 2. Mutation rates for the wild-type and G447S-DNA pol in the presence and absence of postreplication mismatch repair

Ten independent His⁺ revertants arising in the msh2 strain and 10 His⁺ revertants arising in the msh2 pol3-447 strain were sequenced. All 20 His⁺ revertants were +1 insertions, A₇ A₈, which restored the wild-type sequence. Twenty independent Can^r mutants for the wild-type, pol3-447, msh2, and msh2 pol3-447 strains (80 total) were sequenced (Table 3). In the presence of active mismatch repair, most of the Can^r mutations were single-base substitutions, but most of the mutations for the msh2 strains were single-base insertion/deletion mutations, primarily -1 deletions in short mononucleotide repeat sequences, as observed by MARSISCHKY et al. 1996 . Thus, the reduction in Can^r mutation rate for the msh2 pol3-447 strain is due primarily to a decrease in frameshifts in short homopolymeric tracts.

View this table: In this window In a new window   Table 3. Mutation spectra of CAN1 mutants produced by the wild-type and G447S-DNA pol in the presence and absence of postreplication mismatch repair

Base substitution fidelity by the S. cerevisiae wild-type and G447S-DNA pol 's:
A strong mutator phenotype for base substitution mutations was expected for the G447S-DNA pol as observed for the T4 G255S-DNA polymerase (Table 1). The Can^r mutation rates for the wild-type and pol3-447 strains, however, are indistinguishable (Table 2 and Table 3). Although the base substitution mutator phenotype for the G447S-DNA pol was expected to be strong enough to be detected in the presence of postreplication mismatch repair, mismatch repair may be sufficient to correct base substitution errors made by a weak mutator DNA polymerase. The Msh6 protein is required for repair of base substitution errors (MARSISCHKY et al. 1996 ); thus, inactivation of the MSH6 gene is expected to reveal base substitution errors made by the wild-type and G447S-DNA pol 's. An approximately sevenfold rate increase in Can^r mutants was observed for the msh6 strain compared to the wild-type strain (Table 2) and 95% (19/20) of the Can^r mutants were base substitutions (Table 3); however, no significant differences were seen for the msh6 pol3-447 strain. Thus, the yeast G447S-DNA pol does not appear to be a mutator for base substitution errors, but we cannot rule out the possibility that replication errors made by the G447S-DNA pol may be repaired by another proofreading exonuclease activity, such as by DNA pol (MORRISON and SUGINO 1994 ).

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Genetic (REHA-KRANTZ 1988 ; STOCKI et al. 1995 ), biochemical (MARQUEZ and REHA-KRANTZ 1996 ; BAKER and REHA-KRANTZ 1998 ), and structural (SHAMOO and STEITZ 1999 ) studies of the phage T4 and the closely related phage RB69 DNA polymerase demonstrate a critical role for the G255 ß-hairpin structure in the proofreading pathway. Substitution of serine for residue G255 in the turn region of the ß-hairpin structure in the T4 DNA polymerase (Fig 1) results in a strong mutator phenotype for base substitution errors and a weaker mutator phenotype for frameshifts (Table 1). Although the T4 G255S-DNA polymerase can hydrolyze single-stranded DNA and duplex DNAs that have preformed strand separations at the primer terminus (MARQUEZ and REHA-KRANTZ 1996 ), the reduced ability of the G255S-DNA polymerase to convert duplex DNA to the strand-separated editing complex imposes a higher kinetic barrier to the proofreading pathway (BAKER and REHA-KRANTZ 1998 ). As a consequence, the G255S-DNA polymerase proofreads mismatched primer termini less frequently and there are more DNA replication errors.

The discovery of a similar ß-turn-ß structure in the exonuclease domain of an archaeal DNA polymerase, the DTOK DNA polymerase (ZHAO et al. 1999 ), is consistent with the proposal that proofreading is assisted by a ß-hairpin structure in other DNA polymerases. The T4, RB69, and DTOK DNA polymerases are members of a large family of protein-sequence-related DNA polymerases, which includes the eukaryotic DNA pol , an essential DNA polymerase for chromosome replication (SITNEY et al. 1989 ). Protein sequence alignments were carried out for the T4, RB69, and DTOK DNA polymerases for which structural data are available and were compared to the protein sequences of the Vent DNA polymerase, E. coli DNA pol II, and to the yeast and human DNA pol 's (Fig 1). Although there is only limited sequence similarity in the region of the ß-turn-ß structure, amino acid residues that demarcate ß-strands and -helices in the T4, RB69, and DTOK DNA polymerases appear to be conserved in the other DNA polymerases compared. A star in Fig 1 indicates the position of residue G255 in the turn of the ß-hairpin of the T4 DNA polymerase and large boldface letters indicate this residue and other apparent conserved residues.

To determine if the apparent conserved ß-turn-ß structure in other DNA polymerases is important for proofreading, as observed for the T4 DNA polymerase, site-directed mutagenesis was used to construct the S. cerevisiae G447S-DNA pol . The expectation was that the yeast G447S-DNA pol would confer a strong mutator phenotype for base substitutions and a weaker mutator phenotype for frameshifts, as detected for the T4 G255S-DNA polymerase (Table 1). DNA replication fidelity in yeast was determined in vivo by reversion and forward mutation tests (Table 2 and Table 3). A mutator phenotype for base substitution errors was not detected for the yeast G447S-DNA pol , but an antimutator phenotype for +1 and -1 frameshifts in short homopolymeric tracts was observed (Table 2 and Table 3). The frameshift antimutator phenotype was observed only in the absence of postreplication mismatch repair, which was achieved by inactivation of the MSH2 gene. These results will be discussed first with respect to the role of the ß-hairpin structure in the yeast DNA polymerase proofreading pathway and second with respect to the role of DNA polymerase proofreading in frameshift mutagenesis.

Because a strong mutator phenotype for base substitution errors was not observed for the yeast G447S-DNA pol , a ß-hairpin structure may not exist or, if the structure is present, the structure does not function in proofreading in the yeast DNA pol as it does for the T4 DNA polymerase. Alternatively, if yeast DNA pol has a ß-hairpin structure, as supported by the protein sequence alignments (Fig 1), the serine substitution for G447 may not be sufficient to alter the yeast ß-hairpin structure as much as the G255S substitution alters the T4 ß-hairpin structure. Additional mutational analysis of the yeast DNA pol is required to determine which of the proposals is correct, but we favor the proposal that the yeast DNA pol has a ß-hairpin that functions in proofreading since the G447S substitution does affect DNA replication fidelity (Table 2 and Table 3). The antimutator phenotype of the G447S-DNA pol most likely indicates a direct role of DNA pol in frameshift mutagenesis; an indirect role would involve the G447S substitution allowing proofreading by another, more accurate DNA polymerase.

The antimutator phenotype observed for the yeast G447S-DNA pol for frameshift mutations, but not for base substitutions, emphasizes the differences in the mechanisms responsible for producing the two types of mutations. The antimutator phenotype is even more interesting because of the location of the G447S substitution in the proofreading 3' 5' exonuclease domain since antimutator alleles of the T4 DNA polymerase (NOSSAL 1998 ) and the E. coli DNA pol III holoenzyme (SCHAAPER 1993 ) have amino acid substitutions in protein domains that function in nucleotide incorporation and translocation. The G255S substitution in the T4 DNA polymerase also affects base substitution and frameshift fidelity differently; while a strong mutator phenotype for transition and transversion mutations was detected, a weaker mutator phenotype was observed for +1 frameshifts in tracts of five As (Table 1). In contrast, mutant T4 and yeast DNA polymerases defective in the hydrolysis reaction, the T4 D324G-DNA polymerase (Table 1) and the yeast D321A/E323A-DNA pol , the pol3-01 allele (MORRISON et al. 1993 ), display strong mutator phenotypes for both base substitution and frameshift mutations.

Base substitutions are believed to arise by misinserted nucleotides that are not corrected by exonucleolytic proofreading or by postreplication mismatch repair. Frameshifts, however, are believed to arise by transient misalignment of the primer and template strands, most likely in sequences with simple repeats (STREISINGER et al. 1966 ). Transient strand misalignments may also result in base substitutions (KUNKEL and SONI 1988 ). In both cases, the misaligned DNA strands form Watson-Crick base pairs at the primer terminus and, thus, would be expected to be at least partially refractory to exonucleolytic proofreading.

Two models for production of misaligned DNA strands have been proposed: one model requires DNA polymerase dissociation and the second occurs during processive DNA replication. Since frameshift mutagenesis is stimulated in in vitro assays by conditions that favor DNA polymerase dissociation, strand misalignments are proposed to either occur spontaneously when the DNA strands are unbound or to be formed during reassociation of the DNA polymerase to the primer template (KUNKEL 1985B ; SCHLOTTERER and TAUTZ 1992 ; KUNKEL et al. 1994 ; BEBENEK et al. 1995 ). Replicative DNA polymerases, however, are "clamped" to the DNA template and do not readily dissociate (HACKER and ALBERTS 1994 ), which raises the possibility that strand misalignments may occur during chromosome replication while the DNA polymerase remains bound. Misalignment errors have been detected in in vitro reactions during the highly processive replication by the T4 DNA polymerase holoenzyme (KROUTIL et al. 1998 ). DNA polymerases have a potential opportunity to produce strand misalignments during exonucleolytic proofreading when the 3' end of the primer strand is separated from the template and transferred to the exonuclease active center (Fig 2). Normally, the primer end is returned to the polymerase active center after exonuclease trimming so that correct base pairing with the template strand is restored. We (HADJIMARCOU 1999 ) and others (FUJII et al. 1999 ) have proposed that the primer terminus may sometimes be returned to the polymerase active center misaligned and that the misalignment could be stabilized by complementary but out-of-register base pairing within repeat sequences (Fig 2). Extension of the misaligned primer template would then result in the expansion or contraction of the number of repeat units depending on whether the misalignment was in the primer or template strands.

View larger version (10K): In this window In a new window Download PPT slide   Figure 2. Strand misalignment during DNA polymerase proofreading. (A) At the initiation of exonucleolytic proofreading, the 3' end of the primer strand is transferred from the polymerase active center, which is illustrated by a semicircle, to the exonuclease active center, which is illustrated by a <. The editing complex is stabilized by the ß-hairpin structure. The terminal phosphodiester bond is cleaved at the position of the arrow to release the 3' nucleotide, which is 5A in the illustration. (B) The "trimmed" primer terminus is then returned to the polymerase active center. Although correct reannealing of the primer and template strand normally occurs, we propose that strand misalignments are possible. Strand misalignments may be stabilized if Watson-Crick base pairing is possible, as is the case for strand misalignments in sequences with simple repeats. Primer extension of the misaligned primer template in this example results in expansion of the tract of five As to six. A repositioning of the template strand within the polymerase active center is indicated, which would increase the likelihood of producing a misalignment, but there is no experimental evidence at this time to support such movement.

The G255S substitution in the T4 DNA polymerase produces a mutant enzyme with decreased ability to form the partially strand-separated editing complex (MARQUEZ and REHA-KRANTZ 1996 ; BAKER and REHA-KRANTZ 1998 ). Reduced formation of the editing complex would mean less strand separation and could account for the weak mutator phenotype for +1 frameshifts observed for the T4 G255S-DNA polymerase instead of the strong mutator phenotype observed for the exonuclease-deficient D324G-DNA polymerase, which can readily form the editing complex (Table 1). If this proposal is true, then the weak mutator phenotype for +1 frameshifts for the G255S-DNA polymerase may indicate two pathways for formation of frameshifts. One pathway would be dependent on forming the editing complex, which would be reduced by the G255S substitution, and a second pathway that would be due to extension of preformed misaligned DNA strands, which would be increased for the G255S-DNA polymerase. There is evidence that active proofreading can stimulate formation of mutations. Exonuclease activity is required for dinucleotide expansion catalyzed by the T4 DNA polymerase in an in vitro assay (FIDALGO DA SILVA and REHA-KRANTZ 2000 ) and for "misalignment misincorporation" catalyzed by the E. coli DNA pol III holoenzyme (BLOOM et al. 1997 ).

Antimutator DNA polymerases provide an opportunity to learn how DNA replication errors are made. The frameshift antimutator G447S-DNA pol differs from the T4 antimutator DNA polymerases, which sharply reduce AT GC transitions, but have little effect or may even increase other types of mutations (DRAKE and GREENING 1970 ; RIPLEY 1975 ; RIPLEY and SHOEMAKER 1983 ; REHA-KRANTZ 1995 ). The yeast G447S-DNA pol also differs from the E. coli dnaE antimutators, which reduce transitions, but not transversions or frameshifts (SCHAAPER 1993 ). These results indicate that mechanisms for producing transitions, transversions, and frameshifts may be different, as proposed by DRAKE 1993 . Another implication of the frameshift antimutator G447S-DNA pol is the link between frameshift mutagenesis and postreplication mismatch repair. Since the frameshift antimutator phenotype for the G447S-DNA pol was detected only in the absence of postreplication mismatch repair, postreplication mismatch repair normally repairs these insertion/deletion mismatches. Thus, postreplication mismatch repair may have evolved to remedy not only nucleotide misincorporation errors and strand misalignments that escape proofreading, but also errors that result from strand misalignments that may be formed during proofreading.

	ACKNOWLEDGMENTS

We thank D. Wishart and BioTools Inc. (Edmonton, Alberta, Canada) for assistance with protein sequence alignments, L. Stefanovic for assistance in determining yeast mutation rates, and S. Stocki and D. Zhao for assistance in determining the T4 mutation rates. We also thank M. Goodman, T. Kunkel, and lab members for helpful comments on the manuscript. This research was supported by a grant from the Medical Research Council of Canada (MT-13651 to L.J.R.-K.) and by grants from the National Institutes of Health (GM-52319 to T.D.P. and GM-17879 to R.J.K). M.I.H. was supported by the International Council for Canadian Studies and the Canadian Commonwealth Scholarship and Fellowship Program. L.R.-K. is a Scientist of the Alberta Heritage Foundation for Medical Research.

Manuscript received November 3, 2000; Accepted for publication February 13, 2001.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

BAKER, R. P. and L. J. REHA-KRANTZ, 1998  Identification of a transient excision intermediate at the crossroads between DNA polymerase extension and proofreading pathways. Proc. Natl. Acad. Sci. USA 95:3507-3512[Abstract/Free Full Text].

BEBENEK, K., W. A. BEARD, J. R. CASAS-FINET, H.-R. KIM, and T. A. DARDEN et al., 1995  Reduced frameshift fidelity and processivity of HIV-1 reverse transcriptase mutants containing alanine substitutions in helix H of the thumb subdomain. J. Biol. Chem. 270:19516-19523[Abstract/Free Full Text].

BENZER, S., 1960  On the topography of the genetic fine structure. Genetics 47:403-415.

BLOOM, L. B., X. CHEN, D. K. FYGENSON, J. TURNER, and M. O'DONNELL et al., 1997  Fidelity of Escherichia coli DNA polymerase III holoenzyme. The effects of ß, complex processivity proteins and proofreading exonuclease on nucleotide misincorporation efficiencies. J. Biol. Chem. 272:27919-27930[Abstract/Free Full Text].

BRAITHWAITE, D. K. and J. ITO, 1993  Compilation, alignment, and phylogenetic relationships of DNA polymerases. Nucleic Acids Res. 21:787-802[Free Full Text].

DRAKE, J. W., 1993  General antimutators are improbable. J. Mol. Biol. 229:8-13[Medline].

DRAKE, J. W. and E. O. GREENING, 1970  Suppression of chemical mutagenesis in bacteriophage T4 by genetically modified DNA polymerases. Proc. Natl. Acad. Sci. USA 66:823-829[Abstract/Free Full Text].

ELISSEEVA, E., S. S. MANDAL, and L. J. REHA-KRANTZ, 1999  Mutational and pH studies of the 3'5' exonuclease activity of bacteriophage T4 DNA polymerase. J. Biol. Chem. 274:25151-25158[Abstract/Free Full Text].

FIDALGO DA SILVA, E. and L. J. REHA-KRANTZ, 2000  Dinucleotide repeat expansion catalyzed by bacteriophage T4 DNA polymerase in vitro.. J. Biol. Chem. 275:31528-31535[Abstract/Free Full Text].

FUJII, S., M. AKIYAMA, K. AOKI, Y. SUGAYA, and K. HIGUCHI et al., 1999  DNA replication errors produced by the replicative apparatus of Escherichia coli. J. Mol. Biol. 289:835-850[Medline].

HACKER, K. J. and B. A. ALBERTS, 1994  The slow dissociation of the T4 DNA polymerase holoenzyme when stalled by nucleotide omission. J. Biol. Chem. 269:24209-24220[Abstract/Free Full Text].

HADJIMARCOU, M. I., 1999 Proofreading by yeast DNA polymerase delta. M.Sc. Thesis, University of Alberta, Edmonton, Alberta, Canada.

KROUTIL, L. C., M. W. FREY, B. F. KABOORD, T. A. KUNKEL, and S. J. BENKOVIC, 1998  Effect of accessory proteins on T4 DNA polymerase replication fidelity. J. Mol. Biol. 278:135-146[Medline].

KUNKEL, T. A., 1985a  Rapid and efficient site-specific mutagenesis without phenotypic selection. Proc. Natl. Acad. Sci. USA 82:488-492[Abstract/Free Full Text].

KUNKEL, T. A., 1985b  The mutational specificity of DNA polymerases- and - during in vitro DNA synthesis. J. Biol. Chem. 260:12866-12874[Abstract/Free Full Text].

KUNKEL, T. A. and A. SONI, 1988  Mutagenesis by transient misalignment. J. Biol. Chem. 263:14784-14789[Abstract/Free Full Text].

KUNKEL, T. A., S. S. PATEL, and K. A. JOHNSON, 1994  Error-prone replication of repeated DNA sequences by T7 DNA polymerase in the absence of its processivity subunit. Proc. Natl. Acad. Sci. USA 91:6830-6834[Abstract/Free Full Text].

LEA, D. E. and C. A. COULSON, 1949  The distribution of the number of mutants in bacterial populations. J. Genet. 49:264-285.

MARQUEZ, L. A. and L. J. REHA-KRANTZ, 1996  Using 2-aminopurine fluorescence and mutational analysis to demonstrate an active role of bacteriophage T4 DNA polymerase in strand separation required for 3'5' exonuclease activity. J. Biol. Chem. 271:28903-28911[Abstract/Free Full Text].

MARSISCHKY, G. T., N. FILOSI, M. F. KANE, and R. KOLODNER, 1996  Redundancy of Saccharomyces cerevisiae MSH3 and MSH6 in MSH2-dependent mismatch repair. Genes Dev. 10:407-420[Abstract/Free Full Text].

MCPHEETERS, D. S., A. CHRISTENSEN, E. T. YOUNG, G. STORMO, and L. GOLD, 1986  Translational regulation of expression of the bacteriophage T4 lysozyme gene. Nucleic Acids Res. 14:5813-5826[Abstract/Free Full Text].

MORRISON, A. and A. SUGINO, 1994  The 3'5' exonucleases of both DNA polymerase and participate in correcting errors of DNA replication in Saccharomyces cerevisiae.. Mol. Gen. Genet. 242:289-296[Medline].

MORRISON, A., A. L. JOHNSON, L. H. JOHNSTON, and A. SUGINO, 1993  Pathway correcting DNA replication errors in Saccharomyces cerevisiae.. EMBO J. 12:1467-1473[Medline].

NOSSAL, N. G., 1998  A new look at old mutants of T4 DNA polymerase. Genetics 148:1535-1538[Abstract/Free Full Text].

PRIBNOW, D., D. C. SIGURDSON, L. GOLD, B. S. SINGER, and C. NAPOLI et al., 1981  rII cistrons of bacteriophage T4. J. Mol. Biol. 149:337-376[Medline].

REHA-KRANTZ, L. J., 1988  Amino acid changes coded by bacteriophage T4 DNA polymerase mutator mutants. J. Mol. Biol. 202:711-724[Medline].

REHA-KRANTZ, L. J., 1995  Learning about DNA polymerase function by studying antimutator DNA polymerases. Trends Biochem. Sci. 232:133-140.

REHA-KRANTZ, L. J. and R. L. NONAY, 1993  Genetic and biochemical studies of bacteriophage T4 DNA polymerase 3'5' exonuclease activity. J. Biol. Chem. 268:27100-27108[Abstract/Free Full Text].

RIPLEY, L. S., 1975  Transversion mutagenesis in bacteriophage T4. Mol. Gen. Genet. 141:23-40[Medline].

RIPLEY, L. S. and N. B. SHOEMAKER, 1983  A major role for bacteriophage T4 DNA polymerase in frameshift mutagenesis. Genetics 103:353-366[Abstract/Free Full Text].

ROTHSTEIN, R., 1983  One-step gene disruption in yeast. Methods Enzymol. 101:202-211[Medline].

SCHAAPER, R. M., 1993  The mutational specificity of two Escherichia coli dnaE antimutator alleles as determined from lacI mutation spectra. Genetics 134:1031-1038[Abstract].

SCHLÖTTERER, C. and D. TAUTZ, 1992  Slippage synthesis of simple sequence DNA. Nucleic Acids Res. 20:211-215[Abstract/Free Full Text].

SHAMOO, J. and T. A. STEITZ, 1999  Building a replisome from interacting pieces: sliding clamp complexed to a peptide from DNA polymerase and a polymerase editing complex. Cell 99:155-166[Medline].

SHCHERBAKOVA, P. V. and T. A. KUNKEL, 1999  Mutator phenotypes conferred by MLH1 overexpression and by heterozygosity for mlh1 mutations. Mol. Cell. Biol. 19:3177-3183[Abstract/Free Full Text].

SIA, E., R. KOKOSKA, M. DOMINSKA, P. GREENWELL, and T. D. PETES, 1997  Microsatellite instability in yeast: dependence on repeat unit size and DNA mismatch repair genes. Mol. Cell. Biol. 17:2851-2858[Abstract].

SITNEY, K. C., M. E. BUDD, and J. L. CAMPBELL, 1989  DNA polymerase III, a second essential DNA polymerase, is encoded by the S. cerevisiae CDC2 gene. Cell 56:599-605[Medline].

SPICER, E. K., J. RUSH, C. FUNG, L. J. REHA-KRANTZ, and J. D. KARAM et al., 1988  Primary structure of T4 DNA polymerase. Evolutionary relatedness to eucaryotic and other procaryotic DNA polymerases. J. Biol. Chem. 263:7478-7486[Abstract/Free Full Text].

STOCKI, S. A., R. L. NONAY, and L. J. REHA-KRANTZ, 1995  Dynamics of bacteriophage T4 DNA polymerase function: identification of amino acid residues that affect switching between polymerase and 3'5' exonuclease activities. J. Mol. Biol. 254:15-28[Medline].

STRAND, M., M. C. EARLEY, G. F. CROUSE, and T. D. PETES, 1995  Mutations in the MSH3 gene preferentially lead to deletions within tracts of simple repetitive DNA in Saccharomyces cerevisiae.. Proc. Natl. Acad. Sci. USA 92:10418-10421[Abstract/Free Full Text].

STREISINGER, G., Y. OKADA, J. EMRICH, J. NEWTON, and A. TSUGITA et al., 1966  Frameshift mutations and the genetic code. Cold Spring Harbor Symp. Quant. Biol. 31:77-84[Abstract/Free Full Text].

TSURIMOTO, T. and B. STILLMAN, 1990  Functions of replication factor C and proliferating-cell nuclear antigen: functional similarity of DNA polymerase accessory proteins from human cells and bacteriophage T4. Proc. Natl. Acad. Sci. USA 87:1023-1027[Abstract/Free Full Text].

WAGA, S. and B. STILLMAN, 1998  The DNA replication fork in eukaryotic cells. Annu. Rev. Biochem. 67:721-751[Medline].

WANG, J., P. YU, T. C. LIN, W. H. KONIGSBERG, and T. A. STEITZ, 1996  Crystal structures of an NH₂-terminal fragment of T4 DNA polymerase and its complexes with single-stranded DNA and with divalent metal ions. Biochemistry 35:8110-8119[Medline].

WANG, J., A. K. M. A. SATTAR, C. C. WANG, J. D. KARAM, and W. H. KONIGSBERG et al., 1997  Crystal structure of a pol family replication DNA polymerase from bacteriophage RB69. Cell 89:1087-1099[Medline].

WANG, T. S.-F., S. W. WONG, and D. KORN, 1989  Human DNA polymerase : predicted functional domains and relationships with viral DNA polymerases. FASEB J. 3:14-21[Abstract].

WIERDL, M., C. N. GREENE, A. DATTA, S. JINKS-ROBERTSON, and T. D. PETES, 1996  Destabilization of simple repetitive DNA sequences by transcription in yeast. Genetics 143:713-721[Abstract].

ZHAO, Y., D. JERUZALMI, I. MOAREFI, L. LEIGHTON, and R. LASKEN et al., 1999  Crystal structure of an archaebacterial DNA polymerase. Structure Fold Des. 7:1189-1199[Medline].

This article has been cited by other articles:

				M. M. Desai, D. Weissman, and M. W. Feldman Evolution Can Favor Antagonistic Epistasis Genetics, October 1, 2007; 177(2): 1001 - 1010. [Abstract] [Full Text] [PDF]

				E. F. d. Silva and L. J. Reha-Krantz DNA polymerase proofreading: active site switching catalyzed by the bacteriophage T4 DNA polymerase Nucleic Acids Res., August 15, 2007; (2007) gkm591v1. [Abstract] [Full Text] [PDF]

				Z. F. Pursell, I. Isoz, E.-B. Lundstrom, E. Johansson, and T. A. Kunkel Regulation of B family DNA polymerase fidelity by a conserved active site residue: characterization of M644W, M644L and M644F mutants of yeast DNA polymerase {varepsilon} Nucleic Acids Res., May 14, 2007; 35(9): 3076 - 3086. [Abstract] [Full Text] [PDF]

				C. H. Sterling and J. B. Sweasy DNA Polymerase 4 of Saccharomyces cerevisiae Is Important for Accurate Repair of Methyl-Methanesulfonate-Induced DNA Damage Genetics, January 1, 2006; 172(1): 89 - 98. [Abstract] [Full Text] [PDF]

				L. Li, K. M. Murphy, U. Kanevets, and L. J. Reha-Krantz Sensitivity to Phosphonoacetic Acid: A New Phenotype to Probe DNA Polymerase {delta} in Saccharomyces cerevisiae Genetics, June 1, 2005; 170(2): 569 - 580. [Abstract] [Full Text] [PDF]

				Y. H. Jin, P. Garg, C. M. W. Stith, H. Al-Refai, J. F. Sterling, L. J. W. Murray, T. A. Kunkel, M. A. Resnick, P. M. Burgers, and D. A. Gordenin The Multiple Biological Roles of the 3'->5' Exonuclease of Saccharomyces cerevisiae DNA Polymerase {delta} Require Switching between the Polymerase and Exonuclease Domains Mol. Cell. Biol., January 1, 2005; 25(1): 461 - 471. [Abstract] [Full Text] [PDF]

				E. S. Miller, E. Kutter, G. Mosig, F. Arisaka, T. Kunisawa, and W. Ruger Bacteriophage T4 Genome Microbiol. Mol. Biol. Rev., March 1, 2003; 67(1): 86 - 156. [Abstract] [Full Text] [PDF]

				A. J. Rattray, B. K. Shafer, C. B. McGill, and J. N. Strathern The Roles of REV3 and RAD57 in Double-Strand-Break-Repair-Induced Mutagenesis of Saccharomyces cerevisiae Genetics, November 1, 2002; 162(3): 1063 - 1077. [Abstract] [Full Text] [PDF]

				N. P. Degtyareva, P. Greenwell, E. R. Hofmann, M. O. Hengartner, L. Zhang, J. G. Culotti, and T. D. Petes Caenorhabditis elegans DNA mismatch repair gene msh-2 is required for microsatellite stability and maintenance of genome integrity PNAS, February 19, 2002; 99(4): 2158 - 2163. [Abstract] [Full Text] [PDF]