PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
 
J Mol Biol. Author manuscript; available in PMC Dec 17, 2011.
Published in final edited form as:
PMCID: PMC2991499
NIHMSID: NIHMS246145
Reversal of a Mutator Activity by a Nearby Fidelity-Neutral Substitution in the RB69 DNA Polymerase Binding Pocket
Anna Trzemecka,1 Agata Jacewicz,1 Geraldine T. Carver,2 John W Drake,2 and Anna Bebenek1*
1 Department of Molecular Biology, Institute of Biochemistry and Biophysics, Polish Academy of Sciences, Pawinskiego 5a, 02-104 Warsaw, Poland
2 Laboratory of Molecular Genetics, National Institute of Environmental Health Sciences, Research Triangle Park, NC 27709, USA
* Corresponding author. Department of Molecular Biology, Institute of Biochemistry and Biophysics, Polish Academy of Sciences, Pawinskiego 5a, 02-104 Warsaw, Poland. aniab/at/ibb.waw.pl
Phage RB69 family-B DNA polymerase is responsible for the overall high fidelity of RB69 DNA synthesis. Fidelity is compromised when conserved Tyr567, one of the residues that form the nascent polymerase base-pair binding pocket, is replaced by alanine. The Y567A mutator mutant has an enlarged binding pocket and can incorporate and extend mispairs efficiently. Ser565 is a nearby conserved residue that also contributes to the binding pocket, but a S565G replacement has only small impacts on DNA replication fidelity. When the Y567A and S565G replacements were combined, mutator activity was strongly decreased compared to that of Y567A alone. Analyses conducted both in vivo and in vitro revealed that, compared to Y567A alone, the double mutant mainly reduced base substitution mutations and to a lesser extent frameshift mutations. The decrease in mutation rates was not due to increased exonuclease activity. Based on measurements of DNA binding affinity and of mismatch insertion and extension, we propose that the recovered fidelity of the double mutant may result in part from increased dissociation of the enzyme from DNA followed by the binding of the same or another polymerase molecule in either the exonuclease or polymerase mode. An additional antimutagenic factor may be a structural alteration in the polymerase binding pocket described in the following report.
Keywords: Exo, the exonuclease domain or site; Pol, the polymerase domain or site
The DNA polymerases (gp43s encoded by gene 43) of the related bacteriophages T4 and RB69 synthesize DNA with high fidelity, inserting one wrong nucleotide per 105 replicated bases1. These polymerases belong to the B family, the same family that is populated by the eukaryotic replicative polymerases α, δ, and ε. For the RB69 DNA polymerase, crystal structures are available for the apo enzyme2, a binary editing complex with DNA bound in the exonuclease domain3, and ternary complexes (replicating complex) with DNA and dNTP bound in the polymerase site4,5. Because of the availability of crystal structures in so many conformations, the RB69 polymerase is a good structural model to study the mechanisms that are responsible for high base selectivity.
RB69 DNA polymerase achieves its fidelity by the coordinated action of two activities, a polymerase (Pol) activity that is responsible for correct nucleotide selection and an exonuclease (Exo) activity that removes mismatched nucleotides from the primer terminus. As with all polymerases that replicate DNA with high fidelity, binding the correct incoming dNTP introduces conformational changes in both the enzyme and the DNA. The fingers subdomain undergoes the largest conformational change, moving toward the palm subdomain so as to enclose the incoming dNTP and the complementary template nucleotide and thus to generate a binding pocket for the nascent base pair 48.
In the RB69 polymerase, the active Pol binding site is formed by several highly conserved residues located in the fingers and palm. Arg482 and Lys560 in the fingers are involved in hydrogen bonding to the phosphate groups of the incoming dNTP. Hydrophobic interactions with the nascent base pair are formed by Leu415 and Tyr416 on the minor groove side and by Tyr567 and Gly568 on the template side. Asn564 and Ser565 form the rear wall of the binding pocket and ensure a coplanar base arrangement of the nascent base pair1,4,5,9.
All of these residues play important roles in RB69 DNA replication fidelity. Leu415 and Tyr567 are critical for substrate discrimination. L415F/G mutants have elevated rates for a variety of base-substitution errors and single-base deletions in repetitive as well as nonrepetitive sequences9. Replacing Tyr567 by Ala, Ser, or Thr confers a strong mutator phenotype, particularly for base-substitution errors1. Tyr416 is responsible for discriminating between ribose and deoxyribose moieties in incoming nucleotides10, while Leu561 discriminates sterically against mismatches in the nascent base-pair binding pocket11. Thus, the size and shape of the polymerase binding pocket plays a crucial role in replication fidelity.
Insertion of a wrong nucleotide compromises the rate of primer extension. The polymerase stalls, altering the balance between extension by the polymerase and excision by the mismatch-editing exonuclease. Newly incorporated mismatches reduce the efficiency of subsequent nucleotide insertions and extensions by 102- to 106-fold7,8,12,13, the magnitude depending on the mismatch and the polymerase.
Polymerases with compromised fidelity due to alterations at the Pol site may display reduced partitioning to the Exo site and thus favor mismatch extension over proofreading, notably with the T4 PolL412M and RB69 PolL415F/G mutants9,14. The same mechanism probably applies to the RB69 PolY567A mutant, as previously proposed1. Conversely, T4 polymerase mutants such as T4A737V that display an antimutator phenotype, mainly reducing A·T → G·C rates, tend to display decreased abilities to translocate after nucleotide incorporation, thus favoring the formation of exonuclease complexes and increasing the efficacy of proofreading15,16. In preliminary screens, we observed that the RB69 S565G replacement had only weak impacts on fidelity but, surprisingly, when the Y567A and S565G replacements were combined within the highly conserved B motif KX3NSXYG, the powerful mutator activity of the PolY567A mutation was strongly reduced. To better understand the role of Ser565 in fidelity, we characterized some of the biochemical properties of RB69 polymerase mutants containing S565G, Y567A, or both replacements. We also investigated the impact of these replacements using fidelity assays both in vivo and in vitro. Our results lead us to propose that introducing the S565G replacement lowers the DNA binding affinity of the polymerase and increases its ability to dissociate from the primer-template, thus increasing the opportunity for proofreading by the same or another polymerase molecule (proofreading in trans) and contributing to the antimutagenic impact of S565G on Y567A.
Bacteriophage RB69 gp43 Y567 and S565 are highly conserved residues within the B family of DNA polymerases and are located in the region that is involved in dNTP binding. We showed previously that Y567 plays a crucial role in base selection: when Y567 was replaced with A, S or T, the mutant polymerase displayed severely decreased fidelities both in vivo and in vitro1. To study the role of S565, we replaced it with G in both Exo+ and Exo backgrounds. We also examined the combinatorial effect of the double replacement Y567A/S565G. Most of the assays were performed using Exo+ derivatives because of the surprising antimutator effect of the double replacement in this background.
Polymerase activity in vivo and specific activity in vitro
In assays in vivo, we often used a hybrid system in which T4 whose own gp43 is inactivated by amber mutations at codons 202 and 386 (T4 43amam) is replicated by wild-type or mutant RB69 gp43s expressed from plasmids1. Polymerase activity in vivo was determined only for Exo+ derivatives. DNA synthesis by T4 gp43 was totally blocked by a 43amam mutation and was performed instead by one of the RB69 DNA polymerases expressed from a plasmid. DNA synthesis was measured in infected E. coli cells at a time when host DNA synthesis had ceased completely1. DNA synthesis was moderate in the presence of S565G either alone (30% of wild-type activity) or in combination with Y567A (40% of wild-type activity), while the Y567A mutant retained 50% activity.
We also measured polymerase specific activity in vitro for the RB69 DNA polymerases using activated DNA as a substrate. The specific activities for the Exo+ versions of PolY567A, PolY567A/S565G, and PolS565G were 80%, 72%, and 47%, respectively, of the Pol+ Exo+ value.
Fidelity in vivo
In rII reversion assays (Table 1), RB69 PolS565G has almost negligible impacts on replication fidelity for both base substitutions and frameshifts; the average change in mutation rates was a two-fold increase compared to wild-type enzyme. Similarly, in rI forward-mutation assays (Table 2), PolS565G in an Exo+ background had a mutation rate elevated by a factor of 5 compared to wild-type enzyme; in an Exo background, no significant difference (rate decreased 0.8-fold) was observed between the parental and mutant enzymes. Surprisingly, the double mutant (PolS565G/Y567A Exo+) had strongly decreased mutator activities compare to the PolY567A Exo+ mutant in rII reversion assays: the second replacement, at S565, suppressed most of the mutator trait of the Y567A replacement. The most striking result was the decrease in base substitution mutation rates by 90-fold at the UV375 A·T site and 60-fold at the UV256 G·C→A·T site, while −1 and +1 frameshift mutations were reduced by 50-fold and 6-fold, respectively, at other sites (Table 1). In rI forward-mutation assays, the PolS565G/Y567A double mutant decreased mutation rates by 40-fold compared to the single PolY567A mutant (Table 2). In the Exo background, the PolS565G/Y567A mutant did not support the production of viable T4 43amam progeny, perhaps because of a lethally high mutation rate from the combination of mutator effects of the Y567A replacement and the Exo deficiency1,17, a combination that the S565G mutation may reduce insufficiently to restore viability.
Table 1
Table 1
Mutation rates for RB69 Exo+ DNA polymerase mutants in rII reversion assays
Table 2
Table 2
Forward-mutation rates of RB69 DNA polymerase mutants in vivo
We sequenced independent rI mutants from experiments in which T4 43amam phage was replicated by the RB69 polymerase mutants PolS565G Exo+ and PolS565G/Y567A Exo+, respectively. The kinds of mutations made by these polymerases are shown in Table 3, together with the kinds obtained previously for the Pol+ and PolY567A enzymes1,17. Mutations produced by the PolS565G Exo+ polymerase were very similar to those produced by the Pol+ Exo+ polymerase; the complex mutations were almost all GCG→CTA at the 146–148 hotspot1719. There were no strong differences between the kinds of mutations produced by the PolS565G/Y567A Exo+ and PolY567A Exo+ polymerases, each making predominantly base substitutions. G→A predominated among transition mutations, with as many as 70% localized at a hotspot at rI base 247. C→T and C→A mutations at another hotspot at rI base 202 were also present in the spectra produced by both the PolY567A and the PolS565G/Y567A Exo+ polymerases (Fig. 1). Only A→G mutations were underrepresented in this spectrum, only 1 appearing while 11 were found in the PolY567A Exo+ spectrum (Table 3). Thus, the decreased mutation rates in the double mutant hardly affected specificity but merely lowered the overall mutation rate.
Table 3
Table 3
Classes of rI mutations generated by Exo+ RB69 DNA polymerase mutants
Fig. 1
Fig. 1
rI mutational spectra for the RB69 Pol S565G/Y567A Exo+ and PolY567A Exo+ polymerases. The sequence is the wild-type rI strand complementary to the coding strand with every tenth base in bold face. Capital letters indicate base substitutions. The addition (more ...)
Fidelity in vitro
To learn more about the role of Ser565 in fidelity, we conducted studies in vitro using the lacZα DNA template in gapped M13mp2 DNA (Table 4). The lacZα mutant frequencies for the Pol+ Exo+ and PolS565G Exo+ mutants were 8 × 10−4 and 4 × 10−4, respectively, values indistinguishable from each other and similar to the (5–7) × 10−4 typically obtained with gapped DNA17,20 and representing only the mutations that accumulated during the growth of the M13 phage stock. The lacZα mutant frequency for the Pol Y567A Exo+ mutant in earlier studies was 55 × 10−4 and we observed a five-fold decrease in mutation frequency when the DNA was copied by the PolS565G/Y567A Exo+ polymerase. This frequency, 10 × 10−4, was close to that of the Pol+ Exo+ polymerase. These results were fully consistent with those obtained in our fidelity studies conducted in vivo.
Table 4
Table 4
Fidelities of RB69 polymerases in lacZα forward-mutation tests in vitro
The lacZα mutant frequency for the PolS565G Exo polymerase was 14 × 10−4, modestly lower that the 29 × 10−4 obtained previously with the Pol+ Exo enzyme. The PolS565G/Y567A Exo polymerase was more proficient in proofreading some errors than was the PolY567A Exo polymerase (Table 4: 89 × 10−4 versus 164 × 10−4). However, the PolS565G/Y567A Exo polymerase retained a high mutation frequency, which probably resulted in lethally mutated progeny during growth in vivo.
The contribution of S565G to the fidelity of PolS565G/Y567A
As previously reported17, the RB69 PolY567A Exo+ polymerase is a strong base-substitution mutator in vitro, mostly for transition mutations (and especially for T→C produced by T·G mispairs), but also for frameshift mutations in homopolymeric runs while frameshift mutations in runs were almost absent in vivo. The PolS565G/Y567A Exo+ double mutant also produced base substitutions, again mostly transition mutations. This polymerase also produced frameshift mutations and large deletions not seen in the PolY567A Exo+ spectrum (Table 5). Among 12 large deletions, 11 (of 27–371 nucleotides) occurred between direct repeats. Most of the frameshift mutations occurred at runs of two or more repeated bases.
Table 5
Table 5
RB69 DNA polymerase mutation specificities in lacZα forward-mutation tests in vitro
The roster of mutational classes showed a significant decrease in T→C transitions with the double replacement compared to PolY567A alone. There were at least five sequence contexts where T→C mutations were either absent or appeared only once in the double-mutant spectrum, but were strongly represented in the single-mutant spectrum (Fig. 2a). Overall, base-substitution rates decreased by 14-fold in the double mutant, the decrease in T·dGMP rates being as much as 35-fold compare to the PolY567A polymerase (Table 5). The other transition mutation rates decreased 7-fold for C·dATP, 8-fold for G·dTMP, and 5-fold for A·dCMP. Transversion mutations were less frequent in the spectra of both polymerases. A·dAMP rates decreased 14-fold, G·dGMP rates 7-fold, and C·dTMP rates 6-fold and, for the remaining mismatches, the mutation rates decreased 3- to 1.5-fold (Table 5). Frameshift mutation rates decreased about 8-fold in the double mutant compared to the single mutant.
Fig. 2
Fig. 2
Fig. 2
Fig. 2
lacZ α mutational spectra. The 5′→3′ sequence of the viral template strand of the lacZα sequence in M13mp2 is shown from position −84 through +197 where +1 is the first transcribed base and every tenth base (more ...)
We did not sequence lacZα mutants from the PolS565G Exo+ polymerase background because the mutation frequency (4 × 10−4) for this polymerase was at the level of the historical background frequency (5.7–6.2 × 10−4) for unfilled template DNA. This low mutation frequency probably recorded the intrinsic mutation frequency resulting from phage M13 replication and made further analysis of the PolS565G Exo+ polymerase unreliable.
The contribution of proofreading to the fidelity of the PolS565G/Y567A Exo+ polymerase
The lacZα mutation frequency for the PolS565G/Y567A Exo polymerase (89 × 10−4) was 6-fold higher that for the PolS565G Exo polymerase (14 × 10−4) and was about 2-fold lower than for the PolY567A Exo mutant (164 × 10−4). The PolS565G/Y567A Exo polymerase had reduced nucleotide selectivity compared with the parental Pol+ Exo polymerase but was less error-prone than the PolY567A Exo polymerase. Proofreading improved fidelity by 8-fold in the PolS565G/Y567A Exo+ polymerase but by only 3-fold in the PolY567A Exo+ polymerase (Table 4). Proofreading in the PolS565G/Y567A double mutant removed both base mismatches and frameshift mutations. The error rates for all 12 possible base mismatches over all detectable sites in the lacZα template and for Exo+ and Exo single and double Pol mutants are presented in Table 5. In the Exo background, the spectra for the single and double mutants showed no striking differences, and the distribution of mutations along the lacZα sequence revealed no significant differences between the two enzymes (Fig. 2b,c).
Exonuclease activities of single and double Pol mutants
One possible explanation for the observed decrease in mutation rates in the PolS565G/Y567A mutant would be increased exonuclease activity. We therefore conducted exonuclease assays on substrates with either a correct base pair or a T·dGMP mismatch at the 3′ terminus to compare the various polymerase variants. However, both the PolS565G and PolS565G/Y567A polymerases were less efficient than the wild-type and PolY567A polymerases in degrading dsDNA with either a correct or incorrect primer-terminus base pair (Fig. 3a,b). The exonuclease activity on ssDNA was indistinguishable among the Pol+, PolY567A and PolS565G/Y567A polymerases but was slightly reduced in the PolS565G polymerase (Fig. 3c).
Fig. 3
Fig. 3
Fig. 3
Fig. 3
Exonuclease activities of RB69 DNA polymerases on ssDNA using a 32P-labeled 20-mer, on normal dsDNA using a 32P-labeled 20-mer/35-mer, and on dsDNA containing a terminal G·T mismatch using a 32P-labeled 20-mer/35-mer. Reactions were incubated (more ...)
DNA binding activity
Decreased exonuclease activity may result from a change in the dynamic equilibrium between the formation of complexes at the Pol and Exo active sites14. A mismatch at the 3′ terminus usually increases Exo activity by weakening Pol binding to the substrate7. Using gel retardation assays, the strength of dsDNA binding was measured for Pol+, PolY567A, PolS565G and PolS565G/Y567A polymerases in Exo+ backgrounds with either a normal terminal base pair or with a terminal T·dGMP mismatch using the same substrates as for the exonuclease assays (Fig. 4a,b). Kd(DNA) values were then determined from reciprocal plots of enzyme concentration (Table 6). The PolS565G polymerase had the highest dissociation constants when binding correctly paired DNA (Kd(DNA) = 38.9 nM) or the mismatched substrate (Kd(DNA) = 132 nM). These dissociation constants were about 6-fold higher than for the Pol+ polymerase and 3-fold higher than for the PolY567A polymerase on normal dsDNA, and 3- and 5-fold higher, respectively, than for the Pol+ and PolY567A polymerases on dsDNA with the mismatch. The PolS565G/Y567A polymerase also bound less strongly to both substrates compared to the Pol+ and PolY567A polymerases (Kd(DNA) = 29.9 nM for normal and 79.9 nM for mismatched dsDNA compared with 6.9 nM and 13.2 nM for Pol+ and PolY567A polymerases on correctly paired dsDNA and 38.2 nM and 25.7 nM for Pol+ and PolY567A polymerases on mismatched dsDNA, respectively).
Fig. 4
Fig. 4
Fig. 4
RB69 DNA polymerase binding to dsDNA with normal base pairing or a terminal G·T mismatch. The 5′-labeled double-stranded substrates (5 nM) were incubated with 5, 10, 25, 50, 100, and 200 nM of the indicated RB69 polymerases. After gel (more ...)
Table 6
Table 6
DNA binding affinities of Exo+ RB69 DNA polymerase mutants
Weaker DNA binding in the Pol site does not correlate with decreased dsDNA exonuclease activity for the PolS565G or PolS565G/Y567A polymerases. Thus, both enzymes probably dissociate from either normal or mismatched primer-template substrates more often that do either the Pol+ or the PolY567A polymerases.
Primer-extension activities
The error rate of the PolS565G/Y567A Exo+ polymerase is decreased for almost all mispairs compared to the PolY567A Exo+ polymerase. These differences are also observed in an Exo background and thus may not simply reflect partitioning ratios between the Pol and Exo sites. Analysis of mutation spectra for the PolS565G/Y567A Exo+ and PolY567A Exo+ polymerases showed that, among transition mutations, the biggest difference was in extending T·dGMP mismatches (Table 5). Therefore, we used standing-start and running-start assays to measure the abilities of the Pol+, PolY567A, and PolS565G/Y567A polymerases (all in Exo+ backgrounds) to insert dGTP opposite T and to extend the T·dGMP mismatch.
The Pol+ and PolS565G polymerases misinsert dGTP inefficiently opposite a template T (Fig. 5) and also extend a T·dGMP mismatch inefficiently (Fig. 6). The PolY567A polymerase both generates and extends T·dGMP mismatches more efficiently (Figs. 5 and and6).6). The PolS565G/Y567A polymerase is much less efficient in both insertion and extension (Figs. 5 and and6).6). The same pattern was observed in running-start assays: the Pol+ and PolS565G polymerases were unable to detectably insert dGTP opposite template T. The PolY567A polymerase both inserted and extended a mismatch, while the PolS565G/Y567A polymerase was again much less efficient in both steps (Fig. 7). The Exo variants of Pol+, PolS565G and PolS565G/Y567A polymerases were similarly less efficient in both activities, in contrast to the PolY567A polymerase, which inserted and extended the T·dGMP mismatch efficiently (data not shown).
Fig. 5
Fig. 5
Fig. 5
T·dGTP mismatch formation by RB69 DNA polymerases. (a) A 32P-labeled primer-template was present at 50 nM in all reactions. The sequence of a portion of the substrate is shown on the left with an incoming dGTP (“dG”). For each (more ...)
Fig. 6
Fig. 6
Fig. 6
T-dGMP mismatch extension by RB69 DNA polymerases. (a) A 32P-labeled primer-template with a terminal T-dGMP mismatch was present at 50 nM in all reactions. The sequence of a portion of the substrate is shown on the left. For each polymerase, 0.05 nM of (more ...)
Fig. 7
Fig. 7
Running-start reactions for T·dG mismatch formation and extension for the indicated RB69 DNA Exo+ polymerases. A 32P-labeled primer·template was present at 50 nM in all reactions. A partial sequence of the substrate is shown on the left. (more ...)
The PolS565G/Y567A Exo polymerase supports the replication of T4 phage in vivo
It was reported previously that the PolY567A Exo polymerase was unable to support the growth in vivo of either T4 43amam or T4 43+1. This dominant-lethal phenotype was ascribed to the low fidelity of DNA synthesis conducted by this enzyme, resulting in lethally mutated progeny phage. In the complementation assay, the PolS565G/Y567A Exo polymerase is also unable to support the growth of T4 43amam but can support the growth of T4 43+ (Fig. 8), suggesting that Pol+ can compete sufficiently with PolS565G/Y567A to produce some viable progeny whose plaques are, however, small compared to T4 plaques growing on BB cells harboring a plasmid expressing Pol+ Exo or PolS565G Exo.
Fig. 8
Fig. 8
Complementation assay. Phage growth images are spot tests in which 20,000 (left) or 100 (right) viable particles of either T4 43am or 43+ were spotted onto lawns of cells carrying a plasmid producing the indicated Exo RB69 gp43.
Reversal of a powerful mutator activity by a nearby fidelity-neutral substitution
The phage RB69 DNA polymerase mutant PolY567A displays a powerful mutator activity that can exceed 103-fold at some sites1,17. In the crystal structure of the RB69 DNA polymerase ternary complex, Y567 displays an unfavorable rotamer conformation that is maintained by hydrogen-bonding interactions with two nearby residues21. These interactions are essential for forming a nucleotide binding pocket. The aromatic phenyl ring of Y567 plays an important role in base discrimination, forming a hydrogen bond with the minor-groove edge of the DNA duplex at the primer terminus and thus helping to check Watson-Crick base-pair geometry. Removing the γ-hydroxyl from Y567 (Y567F) disrupts the hydrogen-bonding network, allows Y567F to adopt a more favorable rotamer conformation that distorts the nucleotide binding pocket, reduces the affinity of the incoming dNTP, and blocks DNA replication21. On the other hand, replacing the phenolic side chain with a methyl group (Y567A) produces a vigorous polymerase that is a powerful base-substitution mutator both in vivo and in vitro and a moderate mutator for frameshift mutations in vitro. The PolY567A mutant polymerase is better able to accommodate noncanonical base pairings and the Y→A replacement does not interfere with the geometry of correctly paired bases1,17. Thus, Y567 plays a major role in base discrimination by providing a steric gate to check pairing (Fig. 9).
Fig. 9
Fig. 9
A view of the major-groove edge of the nascent dTTP·dA base-pair binding pocket illustrating the positions of Ser565 and Tyr567 in α-helix P (wide gray ribbon). “n” indicates the template dA and “n − 1” (more ...)
The role of Ser565 in maintaining fidelity has only recently come under scrutiny. Because Ser565 contributes to the rear wall of the binding pocket, we expected that the S565G replacement would enlarge the size of the pocket, thus more readily accepting base mispairs and displaying mutator activity. However, the PolS565G polymerase displays at most very weak mutator and antimutator activities both in vivo and in vitro (Tables 15). We also anticipated that the double-mutant PolS565G/Y567A polymerase would contain a further-expanded nucleotide binding pocket compared with the PolY567A polymerase and would therefore display further-increased mutation rates. Surprisingly, however, the PolS565G/Y567A polymerase displayed sharply decreased mutation rates both in vivo and in vitro compared to the PolY567A polymerase (Tables 15). On average, PolS565G/Y567A was 50-fold more accurate than PolY567A in rII reversion assays and 40-fold more accurate in the rI forward-mutation assay.
L561 is another residue contributing to the geometry of the Pol pocket, and when reduced in bulk by the L561A replacement, displays modest mutator activities in mutation tests in vivo and in kinetic assays in vitro11. In contrast to the antimutagenic action of S565G in the PolS565G/Y567A context, we observed no antimutagenic effect of L561A in the PolL561A/S565G context (data not shown). Combining the L561A and Y567A replacements produced a polymerase with exceptionally high mutation rates, and the strong mutator phenotype was again suppressed when S565G was added to the combination (data not shown).
Associated biochemical parameters
We searched for altered behaviors that might explain the sharply contrasting fidelities of the PolS565G/Y567A and the PolY567A polymerases. Polymerase activities were modestly reduced both in vitro and in vivo, but not in informative ways. Proofreading estimated by comparing mutation rates in Exo+ and Exo versions of the mutant polymerases (Table 4), and exonuclease activities on either dsDNA or ssDNA (Fig. 3), were uninformative as to the fidelity paradox. In mispair formation and extension assays in vitro under both standing-start and running-start conditions, the Pol+, PolS565G and PolS565G/Y567A polymerases were inefficient while the PolY567A polymerase was substantially more efficient, results consistent with but not explaining the mutation patterns. Pre-steady-state kinetic data showed that the Pol+, PolS565G, and PolS565G/Y567A polymerases display high Kd,app values for both G·dTMP and T·dGMP mispairs, while the PolY567A polymerase has reduced Kd,app values for both mispairs22. Our present kinetic data for the Pol+ and PolY567A polymerases are in good agreement with the published data.
The accompanying paper5 explores the kinetic parameters of mispair formation, extension, and proofreading using templates that mimic the sequences of either a mutational hotspot or a region of apparently low mutation rate in the rI mutation reporter used here. Although not addressing the question of the antimutagenic properties of the S565G replacement, the assays do largely reproduce in vitro the relative mutation rates observed by us in vivo as a function of local sequence, the first such demonstration known to us. Because spontaneous mutation rates vary greatly from site to site (for instance, by about 103-fold in phage T423), rII reversion rates are expected to vary considerably among sites and mutation spectra typically show a wide range of site-specific rates of forward mutation. Thus, it is often desirable to use a diversity of substrates for kinetics studies of polymerase accuracy.
Measuring DNA binding using gel-retardation assays with correct or mispaired termini, the PolS565G and PolS565G/Y567A polymerases displayed weaker binding compared to the Pol+ or PolY567A polymerases (Fig. 4 and Table 6). Weaker DNA binding for both PolS565G and PolS565G/Y567A polymerases compared to Pol+ and PolY567A polymerases may reflect less-specific polymerase-DNA interactions whose in vivo consequences could be manifested in less efficient DNA synthesis, but direct extrapolation to replication in vivo is difficult and requires caution. Using a different assay and different interaction conditions, the accompanying paper reports no such difference5.
Candidate mechanisms for the antimutagenic activity of S565G
One possible explanation for the increased fidelity of the PolS565G/Y567A polymerase compared to PolY567A is increased partitioning of the mismatched primer-terminus to the Exo site. Diverse changes in the Pol site can perturb the tight coordination between the Pol and Exo cycles7,24, and changing Y567 to alanine not only enlarges the nucleotide binding pocket but probably affects Pol/Exo partitioning in favor of the Pol site1,17, increasing mismatch extension and thus decreasing proofreading. Conversely, proofreading may become more proficient with any inhibition in polymerization, as might be caused by mutations in the Pol site that reduce the ability to form polymerizing complexes and thus shift the balance in favor of Exo complexes15,16. (A side effect is that increased proofreading lowers discrimination between mismatched and matched primer-termini and increases the degradation of newly synthesized DNA25,26). Because both the PolY567A and PolS565G/Y567A enzymes bind less tightly to dsDNA substrates (Fig. 4 and Table 6), they are expected to form exonuclease complexes more readily and, as a consequence, to display increased exonuclease activities. However, we observed no such increase in exonuclease activity. One explanation for this failure would be increased polymerase dissociation from the primer-template DNA, whether or not mismatched, followed by rebinding to DNA to form either an Exo complex in the presence of a mismatch, or a Pol complex and continued replication. T4 DNA polymerase molecules do exchange during DNA replication27, and RB69 would be expected to do the same. Frequent dissociation will increase the opportunity to excise a mismatch by either the original or another polymerase molecule and thus may be responsible for the increased fidelity of replication by PolS565G/Y567A compared to PolY567A. This explanation for the increased fidelity of PolS565G/Y567A compared to PolY567A is supported by our observation of decreased DNA binding by the PolY567A and PolS565G/Y567A enzymes but is challenged by the failure to observe such a difference in the accompanying paper5. However, the conditions used in the two measurements were different, so that decreased binding remains a possibility in vivo.
Note also that more frequent dissociation from the primer-template would also tend to result in the production of large deletions, as observed in vitro in the absence of the gp45 processivity clamp (Table 5). The inhibition of large-deletion mutagenesis between direct repeats by accessory proteins in vitro was observed previously for RB69 DNA polymerase and yeast DNA polymerase δ17,28.
Another explanation for the increased fidelity of the PolS565G/Y567A polymerase compared to PolY567A is provided by the wealth of crystallographic data in the accompanying paper5. One key observation is that the replacement of the bulky Y567 side chain by alanine reduces the rigidity of the Pol pocket in the Y567A single mutant by disrupting the hydrogen-bonding network involving the OH groups of Y567, Y391 and T587, so that the templating base can be displaced downward to accommodate mispairs. When the S565G replacement was introduced into the Y567A mutant, base selectivity was increased compared to PolY567A even though the volume of the Pol pocket was somewhat greater in the PolY567A/S565 double mutant than in PolY567A. This appeared to be due to an increase in hydrophobic van der Waal interactions of G565 with the templating base, making it more rigid.
The above two candidate explanations for the antimutagenic impact of S565G on Y567A are not mutually exclusive.
Construction, expression, and purification of RB69 DNA polymerase variants
Plasmids pCW19R and pCW.50R were generous gifts from Jim Karam (Tulane University). Plasmid pCW19R carries a wild-type RB69 gene 43 (encoding the gp43 polymerase) under the control of the T7 Φ10 promoter of cloning vector pSP72 (Promega). Plasmid pCW.50R encodes a Pol+ Exo polymerase carrying the exonuclease-inactivating replacements D222A and D327A. Site-directed mutagenesis to create the PolS565G, PolY567A, and PolS565G/Y567A variants was carried out using the Stratagene QuikChange Site-Directed Mutagenesis protocol and was confirmed by sequencing. Expression and purification of the PolS565G, PolY567A and PolS565G/Y567A gp43s were as previously described18.
Mutation tests in vivo
Reversion tests using T4 rII mutations were performed as described1. T4 43amamrII mutants were used to infect E. coli BB cells carrying a plasmid expressing the desired allele of RB69 gene 43. rII131 reverts by +1 (A5→A6), rIIUV232 by −1 (A3→A2), rIIUV256 by G·C→A·T transitions, and rIIUV375 probably by both transitions and transversions at three adjacent A·T sites. Forward-mutation rI assays were also performed as described1. To determine the kinds of rI mutations introduced in vivo by various RB69 polymerase mutants, mutants of independent origin were collected, mutant plaques were resuspended in 40 μl of water, and the rI gene was amplified by PCR and sequenced as described1. When sequencing was not performed, a historical correction factor of 0.64 was used for estimate frequencies of rI mutants among all r mutants.
Mutation tests in vitro
Mutation test in vitro were performed with phage M13mp2 lacZα gapped substrates prepared as described20. The incubation mixture (25 μl) contained 25 mM Tris-acetate (pH 7.5), 10 mM Mg acetate, 150 mM K acetate, 2 mM dithiothreitol, 150 ng of M13mp2 lacZα gapped substrate, 1 mM of each dNTP, and 7–10 pmol of wild-type or mutant polymerase. Reactions were incubated at 37 °C for 30 min, stopped by addition of EDTA, and analyzed by agarose gel electrophoresis. Products of reactions in which gap-filling was complete were introduced into MC1061 cells and plated on CSH50 cells to score M13 plaques as either wild-type (dark blue) or mutant (white to less-dark blue). A representative set of mutants from most collections was sequenced to determine types of errors and to adjust the mutation frequency for rare phenotypic mutants without a mutation in the lacZα reporter and for mutants with more than one detectable mutation; a historical correction factor of 0.95 was applied to mutant frequencies when sequencing was not done. Mutation rates (μ) were calculated as described20 by multiplying the net mutant frequency (adjusted for the historical background, 6.2 × 10−4, of uncopied lacZα DNA) by the proportion of mutants in each class and dividing by 0.6 (a correction factor for detecting errors in E. coli) and by the number of detectable sites (opportunities) for each class of mutation.
DNA synthesis in vivo
T4 43amam was used to infect E. coli BB cells carrying a plasmid expressing the desired allele of RB69 gene 43. At 22 min after infection, [3H] thymidine was added (20 μCi/ml at a specific activity of 20 μCi/μg dT). After 15 min, [3H] labeling was terminated in an ice bath and trichloroacetic acid-precipitable 3H counts were determined. Measurements of DNA synthesis in vivo were performed as previously described1.
Polymerase specific activity
Polymerase specific activity was measured by incorporation of α-32P-dATP (Hartmann-Analytic) into activated calf-thymus DNA (Sigma) as previously described9.
DNA substrates
All nucleotides were purchased from the Laboratory of DNA Sequencing and Oligonucleotide Synthesis (Institute of Biochemistry and Biophysics, Warsaw) and were purified on 20% polyacrylamide, 7-M urea gels. All primers were 5′ end-labeled with [γ-32P]dATP (5000 Ci/mmol; Hartmann Analytic) using T4 DNA polynucleotide kinase (Takara). Annealing of primer to template was in 10 mM Tris-HCl (pH 7.5) with a primer/template molar ratio of 1:1.3. The primer/templates were annealed at 75 °C for 5 min and allowed to cool slowly to 25 °C.
DNA gel retardation assay
The interactions of RB69 wild-type or mutant gp43s with dsDNA were assayed using 5′-end-labeled 20-mer (5′-TTTGATGTATTATCAATTGT-3′) hybridized to a 1.3 molar excess of complementary 35-mer (5′-TGCCTTCGTAATCTTACAATTGATAATACATCAAA-3′). The interactions of RB69 polymerases with a mispaired primer terminus were assayed using a 5′ end-labeled 20-mer (5′-ATGTGCTGCAAGGCGATTAG-3′) hybridized to a 30-mer (5′-GTTACCCAACTTAATCGCCTTGCAGCACAT-3′). The 10-μl reaction mixture contained 10 mM HEPES (pH 8.0), 0.5 mM dithiothreitol, 1 μg of bovine serum albumin, 10 mM EDTA, 5 nM DNA substrate, and increasing concentrations of gp43. After incubating for 5 min at 4° C, products were separated on a precooled 6% polyacrylamide nondenaturing gel and quantitated on Fuji Phosphor- Imager. To determine Kd(DNA), the reciprocal of the fraction of DNA shifted was plotted as a function of enzyme concentration.
Exonuclease activity
Assays of dsDNA and ssDNA 3′→5′ exonuclease activities were done as previously described18. Exo activity on mismatched DNA was assayed under the same conditions using 5 nM of gp43 and 50 nM of 32P 5′ end-labeled 20-mer (5′-ATGTGCTGCAAGGCGATTAG-3′) annealed to 30-mer (5′-GTTACCCAACTTAATCGCCTTGCAGCACAT-3′).
Extension of a primer-terminus mismatch
The assay for the extension of a terminal T·dGMP mismatch contained 50 nM of 32P 5′-end-labeled 20-mer (5′-ATGTGCTGCAAGGCGATTAG-3′) annealed to 30-mer (5′-GTTACCCAACTTAATCGCCTTGCAGCACAT-3′). The 10-μl reaction mixture contained 25 mM Tris acetate (pH 7.5), 10 mM Mg acetate, 2 mM DTT, 0.05 nM of gp43, and 1 mM of dGTP. Reactions were incubated at 37 °C for 1, 3 and 5 min and then quenched by addition of stop-dye solution. Product bands were resolved by electrophoresis on 16% polyacrylamide gel with 7 M urea, analyzed by autoradiography, and quantified by densitometry.
Misinsertion assay
The substrate for misinsertion of dGMP opposite template T in standing-start conditions was 32P 5′ end-labeled 20-mer (5′-ATGTGCTGCAAGGCGATTAC-3′) annealed to a complementary 27-mer (5′-GTAAGATGTAATCGCCTTGCAGCACAT-3′). The reaction mixture contained 25 mM Tris acetate (pH 7.5), 10 mM Mg acetate, 2 mM DTT, 50 nM of DNA substrate, 0.05 nM of gp43, and 1 mM of dGTP. Reactions were incubated at 37 °C for 1, 3 and 5 min and then quenched by addition of stop-dye solution.
Running-start assays were conduced in the same reaction buffer. The 10-μl reaction mixture contained 50 nM of 32P 5′ end-labeled 20-mer (5′-ATGTGCTGCAAGGCGATTAC-3′) annealed to a complementary 27-mer (5′-GAACTCCGTAATCGCCTTGCAGCACAT-3′), 0.05 nM of gp43, and increasing concentrations of dGTP (250, 500 and 1000 μM). Reactions were incubated at 37 °C for 5 min and quenched by addition of 5 μl of stop-dye solution. Product bands were resolved by electrophoresis on 16% polyacrylamide gels with 7 M urea, analyzed by autoradiography, and quantified by densitometry.
Acknowledgments
We thank Bill Beard for providing Fig. 9, and both him and Mike Murray for critical readings of the manuscript, and Bill Konigsberg for advice on the implications of his structural models. This research was supported in part by funds allocated to project number Z01ES061054 of the Intramural Research Program of the National Institutes of Health, National Institute of Environmental Health Sciences, USA, and by grant N301014433 from the Polish Ministry of Science and Higher Education to A.B.
Abbreviations used
Exothe exonuclease domain or site
Polthe polymerase domain or site

Footnotes
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1. Bebenek A, Dressman HK, Carver GT, Ng S, Petrov V, Yang G, Konigsberg WH, Karam JD, Drake JW. Interacting fidelity defects in the replicative DNA polymerase of bacteriophage RB69. J Biol Chem. 2001;276:10387–97. [PubMed]
2. Wang J, Sattar AK, Wang CC, Karam JD, Konigsberg WH, Steitz TA. Crystal structure of a pol alpha family replication DNA polymerase from bacteriophage RB69. Cell. 1997;89:1087–99. [PubMed]
3. Shamoo Y, Steitz TA. Building a replisome from interacting pieces: sliding clamp complexed to a peptide from DNA polymerase and a polymerase editing complex. Cell. 1999;99:155–66. [PubMed]
4. Franklin MC, Wang J, Steitz TA. Structure of the replicating complex of a pol alpha family DNA polymerase. Cell. 2001;105:657–67. [PubMed]
5. Wang M, Lee HR, Sinha A, Wang J, Blaha G, Christian T, Xia S, Konigsberg W. The variation in mutation rates caused by RB69pol fidelity mutants can be rationalized on the basis of their kinetic behavior and crystal structures. J Mol Biol (simultaneously submitted) [PMC free article] [PubMed]
6. Li Y, Korolev S, Waksman G. Crystal structures of open and closed forms of binary and ternary complexes of the large fragment of Thermus aquaticus DNA polymerase I: structural basis for nucleotide incorporation. Embo J. 1998;17:7514–25. [PubMed]
7. Kunkel TA, Bebenek K. DNA replication fidelity. Annu Rev Biochem. 2000;69:497–529. [PubMed]
8. Kunkel TA. DNA replication fidelity. J Biol Chem. 2004;279:16895–8. [PubMed]
9. Zhong X, Pedersen LC, Kunkel TA. Characterization of a replicative DNA polymerase mutant with reduced fidelity and increased translesion synthesis capacity. Nucleic Acids Res. 2008;36:3892–904. [PMC free article] [PubMed]
10. Yang G, Franklin M, Li J, Lin TC, Konigsberg W. A conserved Tyr residue is required for sugar selectivity in a Pol alpha DNA polymerase. Biochemistry. 2002;41:10256–61. [PubMed]
11. Zhang H, Rhee C, Bebenek A, Drake JW, Wang J, Konigsberg W. The L561A substitution in the nascent base-pair binding pocket of RB69 DNA polymerase reduces base discrimination. Biochemistry. 2006;45:2211–20. [PMC free article] [PubMed]
12. Echols H, Goodman MF. Fidelity mechanisms in DNA replication. Annu Rev Biochem. 1991;60:477–511. [PubMed]
13. Goodman MF, Creighton S, Bloom LB, Petruska J. Biochemical basis of DNA replication fidelity. Crit Rev Biochem Mol Biol. 1993;28:83–126. [PubMed]
14. Fidalgo da Silva E, Mandal SS, Reha-Krantz LJ. Using 2-aminopurine fluorescence to measure incorporation of incorrect nucleotides by wild type and mutant bacteriophage T4 DNA polymerases. J Biol Chem. 2002;277:40640–9. [PubMed]
15. Spacciapoli P, Nossal NG. A single mutation in bacteriophage T4 DNA polymerase (A737V, tsL141) decreases its processivity as a polymerase and increases its processivity as a 3′-->5′ exonuclease. J Biol Chem. 1994;269:438–46. [PubMed]
16. Reha-Krantz LJ. DNA polymerase proofreading: Multiple roles maintain genome stability. Biochim Biophys Acta 2010 [PubMed]
17. Bebenek A, Carver GT, Dressman HK, Kadyrov FA, Haseman JK, Petrov V, Konigsberg WH, Karam JD, Drake JW. Dissecting the fidelity of bacteriophage RB69 DNA polymerase: site-specific modulation of fidelity by polymerase accessory proteins. Genetics. 2002;162:1003–18. [PubMed]
18. Jacewicz A, Makiela K, Kierzek A, Drake JW, Bebenek A. The roles of Tyr391 and Tyr619 in RB69 DNA polymerase replication fidelity. J Mol Biol. 2007;368:18–29. [PMC free article] [PubMed]
19. Schultz GE, Jr, Carver GT, Drake JW. A role for replication repair in the genesis of templated mutations. J Mol Biol. 2006;358:963–73. [PubMed]
20. Bebenek K, Kunkel TA. Analyzing fidelity of DNA polymerases. Methods Enzymol. 1995;262:217–32. [PubMed]
21. Yang G, Wang J, Konigsberg W. Base selectivity is impaired by mutants that perturb hydrogen bonding networks in the RB69 DNA polymerase active site. Biochemistry. 2005;44:3338–46. [PubMed]
22. Zhang H, Beckman J, Wang J, Konigsberg W. RB69 DNA polymerase mutants with expanded nascent base-pair-binding pockets are highly efficient but have reduced base selectivity. Biochemistry. 2009;48:6940–50. [PMC free article] [PubMed]
23. Ronen A, Rahat A. Mutagen specificity and position effects on mutation in T4rII nonsense sites. Mutat Res. 1976;34:21–34. [PubMed]
24. Muzyczka N, Poland RL, Bessman MJ. Studies on the biochemical basis of spontaneous mutation. I. A comparison of the deoxyribonucleic acid polymerases of mutator, antimutator, and wild type strains of bacteriophage T4. J Biol Chem. 1972;247:7116–22. [PubMed]
25. Clayton LK, Goodman MF, Branscomb EW, Galas DJ. Error induction and correction by mutant and wild type T4 DNA polymerases. Kinetic error discrimination mechanisms. J Biol Chem. 1979;254:1902–12. [PubMed]
26. Gauss P, Doherty DH, Gold L. Bacterial and phage mutations that reveal helix-unwinding activities required for bacteriophage T4 DNA replication. Proc Natl Acad Sci U S A. 1983;80:1669–73. [PubMed]
27. Yang J, Zhuang Z, Roccasecca RM, Trakselis MA, Benkovic SJ. The dynamic processivity of the T4 DNA polymerase during replication. Proc Natl Acad Sci U S A. 2004;101:8289–94. [PubMed]
28. Fortune JM, Stith CM, Kissling GE, Burgers PM, Kunkel TA. RPA and PCNA suppress formation of large deletion errors by yeast DNA polymerase delta. Nucleic Acids Res. 2006;34:4335–41. [PMC free article] [PubMed]
29. Pettersen EF, Goddard TD, Huang CC, Couch GS, Greenblatt DM, Meng EC, Ferrin TE. UCSF Chimera--a visualization system for exploratory research and analysis. J Comput Chem. 2004;25:1605–12. [PubMed]