Division of labor at the eukaryotic replication fork.

DNA polymerase delta (Pol delta) and DNA polymerase epsilon (Pol epsilon) are both required for efficient replication of the nuclear genome, yet the division of labor between these enzymes has remained unclear for many years. Here we investigate the contribution of Pol delta to replication of the leading and lagging strand templates in Saccharomyces cerevisiae using a mutant Pol delta allele (pol3-L612M) whose error rate is higher for one mismatch (e.g., T x dGTP) than for its complement (A x dCTP). We find that strand-specific mutation rates strongly depend on the orientation of a reporter gene relative to an adjacent replication origin, in a manner implying that >90% of Pol delta replication is performed using the lagging strand template. When combined with recent evidence implicating Pol epsilon in leading strand replication, these data support a model of the replication fork wherein the leading and lagging strand templates are primarily copied by Pol epsilon and Pol delta, respectively.

Mismatch repair balances leading and lagging strand DNA replication fidelity.

The two DNA strands of the nuclear genome are replicated asymmetrically using three DNA polymerases, α, δ, and ε. Current evidence suggests that DNA polymerase ε (Pol ε) is the primary leading strand replicase, whereas Pols α and δ primarily perform lagging strand replication. The fact that these polymerases differ in fidelity and error specificity is interesting in light of the fact that the stability of the nuclear genome depends in part on the ability of mismatch repair (MMR) to correct different mismatches generated in different contexts during replication. Here we provide the first comparison, to our knowledge, of the efficiency of MMR of leading and lagging strand replication errors. We first use the strand-biased ribonucleotide incorporation propensity of a Pol ε mutator variant to confirm that Pol ε is the primary leading strand replicase in Saccharomyces cerevisiae. We then use polymerase-specific error signatures to show that MMR efficiency in vivo strongly depends on the polymerase, the mismatch composition, and the location of the mismatch. An extreme case of variation by location is a T-T mismatch that is refractory to MMR. This mismatch is flanked by an AT-rich triplet repeat sequence that, when interrupted, restores MMR to > 95% efficiency. Thus this natural DNA sequence suppresses MMR, placing a nearby base pair at high risk of mutation due to leading strand replication infidelity. We find that, overall, MMR most efficiently corrects the most potentially deleterious errors (indels) and then the most common substitution mismatches. In combination with earlier studies, the results suggest that significant differences exist in the generation and repair of Pol α, δ, and ε replication errors, but in a generally complementary manner that results in high-fidelity replication of both DNA strands of the yeast nuclear genome.

Differential correction of lagging-strand replication errors made by DNA polymerases {alpha} and {delta}.

Mismatch repair (MMR) of replication errors requires DNA ends that can direct repair to the newly synthesized strand containing the error. For all but those organisms that use adenine methylation to generate nicks, the source of these ends in vivo is unknown. One possibility is that MMR may have a "special relation to the replication complex" [Wagner R, Jr., Meselson M (1976) Proc Natl Acad Sci USA 73:4135-4139], perhaps one that allows 5' or 3' DNA ends associated with replication to act as strand discrimination signals. Here we examine this hypothesis, based on the logic that errors made by yeast DNA polymerase α (Pol α), which initiates Okazaki fragments during lagging-strand replication, will always be closer to a 5' end than will be more internal errors generated by DNA polymerase δ (Pol δ), which takes over for Pol α to complete lagging-strand replication. When we compared MMR efficiency for errors made by variant forms of these two polymerases, Msh2-dependent repair efficiencies for mismatches made by Pol α were consistently higher than for those same mismatches when made by Pol δ. Thus, one special relationship between MMR and replication is that MMR is more efficient for the least accurate of the major replicative polymerases, exonuclease-deficient Pol α. This observation is consistent with the close proximity and possible use of 5' ends of Okazaki fragments for strand discrimination, which could increase the probability of Msh2-dependent MMR by 5' excision, by a Msh2-dependent strand displacement mechanism, or both.

Genome-wide model for the normal eukaryotic DNA replication fork.

To investigate DNA replication enzymology across the nuclear genome of budding yeast, deep sequencing was used to establish the pattern of uncorrected replication errors generated by an asymmetric mutator variant of DNA polymerase δ (Pol δ). Sequencing of 16 genomes identified 1,206-bp substitutions generated over 33 generations by L612M Pol δ in a mismatch repair defective strain. Alignment of sequences flanking these substitutions identified "hotspot" motifs for Pol δ replication errors. The substitutions were distributed evenly across all 16 chromosomes. The vast majority were transitions that occurred with a strand bias that varied in a predictable manner relative to known functional origins of replication. This strand bias strongly supports the idea that Pol δ is primarily a lagging strand polymerase during replication across the entire nuclear genome.

Abundant ribonucleotide incorporation into DNA by yeast replicative polymerases.

Measurements of nucleoside triphosphate levels in Saccharomyces cerevisiae reveal that the four rNTPs are in 36- to 190-fold molar excess over their corresponding dNTPs. During DNA synthesis in vitro using the physiological nucleoside triphosphate concentrations, yeast DNA polymerase epsilon, which is implicated in leading strand replication, incorporates one rNMP for every 1,250 dNMPs. Pol delta and Pol alpha, which conduct lagging strand replication, incorporate one rNMP for every 5,000 or 625 dNMPs, respectively. Discrimination against rNMP incorporation varies widely, in some cases by more than 100-fold, depending on the identity of the base and the template sequence context in which it is located. Given estimates of the amount of replication catalyzed by Pols alpha, delta, and epsilon, the results are consistent with the possibility that more than 10,000 rNMPs may be incorporated into the nuclear genome during each round of replication in yeast. Thus, rNMPs may be the most common noncanonical nucleotides introduced into the eukaryotic genome. Potential beneficial and negative consequences of abundant ribonucleotide incorporation into DNA are discussed, including the possibility that unrepaired rNMPs in DNA could be problematic because yeast DNA polymerase epsilon has difficulty bypassing a single rNMP present within a DNA template.

Genome instability due to ribonucleotide incorporation into DNA.

Maintaining the chemical identity of DNA depends on ribonucleotide exclusion by DNA polymerases. However, ribonucleotide exclusion during DNA synthesis in vitro is imperfect. To determine whether ribonucleotides are incorporated during DNA replication in vivo, we substituted leucine or glycine for an active-site methionine in yeast DNA polymerase ϵ (Pol ϵ). Ribonucleotide incorporation in vitro was three-fold lower for M644L and 11-fold higher for M644G Pol ϵ compared to wild-type Pol ϵ. This hierarchy was recapitulated in vivo in yeast strains lacking RNase H2. Moreover, the pol2-M644G rnh201Δ strain progressed more slowly through S phase, had elevated dNTP pools and generated 2-5-base-pair deletions in repetitive sequences at a high rate and in a gene orientation-dependent manner. The data indicate that ribonucleotides are incorporated during replication in vivo, that they are removed by RNase H2-dependent repair and that defective repair results in replicative stress and genome instability via DNA strand misalignment.

A specific loop in human DNA polymerase mu allows switching between creative and DNA-instructed synthesis.

Human DNA polymerase mu (Polmu) is a family X member that has terminal transferase activity but, in spite of a non-orthodox selection of the template information, displays its maximal catalytic efficiency in DNA-templated reactions. As terminal deoxynucleotidyl transferase (TdT), Polmu has a specific loop (loop1) that could provide this enzyme with its terminal transferase activity. When loop1 was deleted, human Polmu lacked TdT activity but improved DNA-binding and DNA template-dependent polymerization. Interestingly, when loop1 from TdT was inserted in Polmu (substituting its cognate loop1), the resulting chimaera displayed TdT activity, preferentially inserting dGTP residues, but had a strongly reduced template-dependent polymerization activity. Therefore, a specialized loop in Polmu, that could adopt alternative conformations, appears to provide this enzyme with a dual capacity: (i) template independency to create new DNA information, in which loop1 would have an active role by acting as a 'pseudotemplate'; (ii) template-dependent polymerization, in which loop1 must allow binding of the template strand. Recent in vivo and in vitro data suggest that such a dual capacity could be advantageous to resolve microhomology-mediated end-joining reactions.

The fidelity of DNA synthesis by yeast DNA polymerase zeta alone and with accessory proteins.

DNA polymerase zeta (pol zeta) participates in several DNA transactions in eukaryotic cells that increase spontaneous and damage-induced mutagenesis. To better understand this central role in mutagenesis in vivo, here we report the fidelity of DNA synthesis in vitro by yeast pol zeta alone and with RFC, PCNA and RPA. Overall, the accessory proteins have little effect on the fidelity of pol zeta. Pol zeta is relatively accurate for single base insertion/deletion errors. However, the average base substitution fidelity of pol zeta is substantially lower than that of homologous B family pols alpha, delta and epsilon. Pol zeta is particularly error prone for substitutions in specific sequence contexts and generates multiple single base errors clustered in short patches at a rate that is unprecedented in comparison with other polymerases. The unique error specificity of pol zeta in vitro is consistent with Pol zeta-dependent mutagenic specificity reported in vivo. This fact, combined with the high rate of single base substitution errors and complex mutations observed here, indicates that pol zeta contributes to mutagenesis in vivo not only by extending mismatches made by other polymerases, but also by directly generating its own mismatches and then extending them.

Polymerase mu is a DNA-directed DNA/RNA polymerase.

DNA polymerases are defined as such because they use deoxynucleotides instead of ribonucleotides with high specificity. We show here that polymerase mu (pol mu), implicated in the nonhomologous end-joining pathway for repair of DNA double-strand breaks, incorporates both ribonucleotides and deoxynucleotides in a template-directed manner. pol mu has an approximately 1,000-fold-reduced ability to discriminate against ribonucleotides compared to that of the related pol beta, although pol mu's substrate specificity is similar to that of pol beta in most other respects. Moreover, pol mu more frequently incorporates ribonucleotides when presented with nucleotide concentrations that approximate cellular pools. We therefore addressed the impact of ribonucleotide incorporation on the activities of factors required for double-strand break repair by nonhomologous end joining. We determined that the ligase required for this pathway readily joined strand breaks with terminal ribonucleotides. Most significantly, pol mu frequently introduced ribonucleotides into the repair junctions of an in vitro nonhomologous end-joining reaction, an activity that would be expected to have important consequences in the context of cellular double-strand break repair.

Association of DNA polymerase mu (pol mu) with Ku and ligase IV: role for pol mu in end-joining double-strand break repair.

Mammalian DNA polymerase mu (pol mu) is related to terminal deoxynucleotidyl transferase, but its biological role is not yet clear. We show here that after exposure of cells to ionizing radiation (IR), levels of pol mu protein increase. pol mu also forms discrete nuclear foci after IR, and these foci are largely coincident with IR-induced foci of gammaH2AX, a previously characterized marker of sites of DNA double-strand breaks. pol mu is thus part of the cellular response to DNA double-strand breaks. pol mu also associates in cell extracts with the nonhomologous end-joining repair factor Ku and requires both Ku and another end-joining factor, XRCC4-ligase IV, to form a stable complex on DNA in vitro. pol mu in turn facilitates both stable recruitment of XRCC4-ligase IV to Ku-bound DNA and ligase IV-dependent end joining. In contrast, the related mammalian DNA polymerase beta does not form a complex with Ku and XRCC4-ligase IV and is less effective than pol mu in facilitating joining mediated by these factors. Our data thus support an important role for pol mu in the end-joining pathway for repair of double-strand breaks.

Ribonucleotide incorporation by yeast DNA polymerase ζ.

During replication in yeast, the three B family DNA replicases frequently incorporate ribonucleotides (rNMPs) into DNA, and their presence in the nuclear genome can affect genome stability. This prompted us to examine ribonucleotide incorporation by the fourth B family member, Pol ζ, the enzyme responsible for the majority of damage-induced mutagenesis in eukaryotes. We first show that Pol ζ inserts rNMPs into DNA and can extend primer termini containing 3'-ribonucleotides. We then measure rNMP incorporation by Pol ζ in the presence of its cofactors, RPA, RFC and PCNA and at normal cellular dNTP and rNTP concentrations that exist under unstressed conditions. Under these conditions, Pol ζ stably incorporates one rNMP for every 200-300 dNMPs incorporated, a frequency that is slightly higher than for the high fidelity replicative DNA polymerases. Under damage-induced conditions wherein cellular dNTP concentrations are elevated 5-fold, Pol ζ only incorporates one rNMP per 1300 dNMPs. Functional interaction of Pol ζ with the mutasome assembly factor Rev1 gives comparable rNMP incorporation frequencies. These results suggest that ribonucleotide incorporation into DNA during Pol ζ-mediated mutagenesis in vivo may be rare.

Proofreading of ribonucleotides inserted into DNA by yeast DNA polymerase ɛ.

We have investigated the ability of the 3' exonuclease activity of Saccharomyces cerevisiae DNA polymerase ɛ (Pol ɛ) to proofread newly inserted ribonucleotides (rNMPs). During DNA synthesis in vitro, Pol ɛ proofreads ribonucleotides with apparent efficiencies that vary from none at some locations to more than 90% at others, with rA and rU being more efficiently proofread than rC and rG. Previous studies show that failure to repair ribonucleotides in the genome of rnh201Δ strains that lack RNase H2 activity elevates the rate of short deletions in tandem repeat sequences. Here we show that this rate is increased by 2-4-fold in pol2-4 rnh201Δ strains that are also defective in Pol ɛ proofreading. In comparison, defective proofreading in these same strains increases the rate of base substitutions by more than 100-fold. Collectively, the results indicate that although proofreading of an 'incorrect' sugar is less efficient than is proofreading of an incorrect base, Pol ɛ does proofread newly inserted rNMPs to enhance genome stability.

Inefficient proofreading and biased error rates during inaccurate DNA synthesis by a mutant derivative of Saccharomyces cerevisiae DNA polymerase delta.

DNA polymerase delta (pol delta) is a high fidelity eukaryotic enzyme that participates in DNA repair and is essential for DNA replication. Toward the goal of dissecting its multiple biological functions, here we describe the biochemical properties of Saccharomyces cerevisiae pol delta with a methionine replacing conserved leucine 612 at the polymerase active site. Compared with wild type pol delta, L612M pol delta has normal processivity and slightly higher polymerase specific activity. L612M pol delta also has normal 3' exonuclease activity, yet it is impaired in partitioning mismatches to the exonuclease active site, thereby reducing DNA synthesis fidelity. Error rates in vitro for L612M pol delta are elevated for both base substitutions and single base deletions but in a highly biased manner. For each of the six possible pairs of reciprocal mismatches that could arise during replication of complementary DNA strands to account for any particular base substitution in vivo (e.g. T-dGMP or A-dCMP for T to C transitions), L612M pol delta error rates are substantially higher for one mismatch than the other. These results provide a biochemical explanation for our observation, which confirms earlier genetic studies, that a haploid pol3-L612M S. cerevisiae strain has an elevated spontaneous mutation rate that is likely due to reduced replication fidelity in vivo.

Evidence for extrinsic exonucleolytic proofreading.

Exonucleolytic proofreading of DNA synthesis errors is one of the major determinants of genome stability. However, many DNA transactions that contribute to genome stability require synthesis by polymerases that naturally lack intrinsic 3' exonuclease activity and some of which are highly inaccurate. Here we discuss evidence that errors made by these polymerases may be edited by a separate 3' exonuclease, and we consider how such extrinsic proofreading may differ from proofreading by exonucleases that are intrinsic to replicative DNA polymerases.

A gradient of template dependence defines distinct biological roles for family X polymerases in nonhomologous end joining.

Three Pol X family members have been linked to nonhomologous end joining (NHEJ) in mammals. Template-independent TdT promotes diversity during NHEJ-dependent repair of V(D)J recombination intermediates, but the roles of the template-dependent polymerases mu and lambda in NHEJ remain unclear. We show here that pol mu and pol lambda are similarly recruited by NHEJ factors to fill gaps when ends have partially complementary overhangs, suggesting equivalent roles promoting accuracy in NHEJ. However, only pol mu promotes accuracy during immunoglobulin kappa recombination. This distinctive in vivo role correlates with the TdT-like ability of pol mu, but not pol lambda, to act when primer termini lack complementary bases in the template strand. However, unlike TdT, synthesis by pol mu in this context is primarily instructed by a template from another DNA molecule. This apparent gradient of template dependence is largely attributable to a small structural element that is present but different in all three polymerases.

Sibling rivalry: competition between Pol X family members in V(D)J recombination and general double strand break repair.

The nonhomologous end-joining pathway is a major means for repairing double-strand breaks (DSBs) in all mitotic cell types. This repair pathway is also the only efficient means for resolving DSB intermediates in V(D)J recombination, a lymphocyte-specific genome rearrangement required for assembly of antigen receptors. A role for polymerases in end-joining has been well established. They are a major factor in determining the character of repair junctions but, in contrast to 'core' end-joining factors, typically appear to have a subtle impact on the efficiency of end-joining. Recent work implicates several members of the Pol X family in end-joining and suggests surprising complexity in the control of how these different polymerases are employed in this pathway.

Translesion synthesis past platinum DNA adducts by human DNA polymerase mu.

DNA polymerase mu (pol mu) is a member of the pol X family of DNA polymerases, and it shares a number of characteristics of both DNA polymerase beta (pol beta) and terminal deoxynucleotidyl transferase (TdT). Because pol beta has been shown to perform translesion DNA synthesis past cisplatin (CP)- and oxaliplatin (OX)-GG adducts, we determined the ability of pol mu to bypass these lesions. Pol mu bypassed CP and OX adducts with an efficiency of 14-35% compared to chain elongation on undamaged DNA, which is second only to pol eta in terms of bypass efficiency. The relative ability of pol mu to bypass CP and OX adducts was dependent on both template structure and sequence context. Since pol mu has been shown to be more efficient on gapped DNA templates than on primed single-stranded DNA templates, we determined the ability of pol mu to bypass Pt-DNA adducts on both primed single-stranded and gapped templates. The bypass of Pt-DNA adducts by pol mu was highly error-prone on all templates, resulting in 2, 3, and 4 nt deletions. We postulate that bypass of Pt-DNA adducts by pol mu may involve looping out the Pt-GG adduct to allow chain elongation downstream of the adduct. This reaction appears to be facilitated by the presence of a downstream "acceptor" and a gap large enough to provide undamaged template DNA for elongation past the adduct, although gapped DNA is clearly not required for bypass.