Emergence of DNA polymerase ε antimutators that escape error-induced extinction in yeast

Abstract

DNA polymerases (Pols) ε and δ perform the bulk of yeast leading- and lagging-strand DNA synthesis. Both Pols possess intrinsic proofreading exonucleases that edit errors during polymerization. Rare errors that elude proofreading are extended into duplex DNA and excised by the mismatch repair (MMR) system. Strains that lack Pol proofreading or MMR exhibit a 10- to 100-fold increase in spontaneous mutation rate (mutator phenotype), and inactivation of both Pol δ proofreading (pol3-01) and MMR is lethal due to replication error-induced extinction (EEX). It is unclear whether a similar synthetic lethal relationship exists between defects in Pol ε proofreading (pol2-4) and MMR. Using a plasmid-shuffling strategy in haploid Saccharomyces cerevisiae, we observed synthetic lethality of pol2-4 with alleles that completely abrogate MMR (msh2Δ, mlh1Δ, msh3Δ msh6Δ, or pms1Δ mlh3Δ) but not with partial MMR loss (msh3Δ, msh6Δ, pms1Δ, or mlh3Δ), indicating that high levels of unrepaired Pol ε errors drive extinction. However, variants that escape this error-induced extinction (eex mutants) frequently emerged. Five percent of pol2-4 msh2Δ eex mutants encoded second-site changes in Pol ε that reduced the pol2-4 mutator phenotype between 3- and 23-fold. The remaining eex alleles were extragenic to pol2-4. The locations of antimutator amino-acid changes in Pol ε and their effects on mutation spectra suggest multiple mechanisms of mutator suppression. Our data indicate that unrepaired leading- and lagging-strand polymerase errors drive extinction within a few cell divisions and suggest that there are polymerase-specific pathways of mutator suppression. The prevalence of suppressors extragenic to the Pol ε gene suggests that factors in addition to proofreading and MMR influence leading-strand DNA replication fidelity.

Figure 1

Genetic interactions between Pol ε and Pol δ proofreading and MMR. Using a plasmid-shuffling strategy, proofreading-deficient variants of Pol ε (encoded by pol2-4) or Pol δ (encoded by pol3-01) or wild-type controls (POL2 or POL3) were introduced via LEU2 plasmids into BY4733 pol2Δ or pol3Δ strains harboring deletion mutations that partially (A) or completely (B) abrogate MMR. Serial dilutions were plated on FOA-containing media to select for loss of complementing POL2– or POL3–URA3 plasmids and reveal the synthetic phenotype. POL and MMR alleles are indicated at left and above the corresponding panels, respectively.

Figure 2

Effects of Pols ζ and η on pol2-4 msh2Δ synthetic lethality. pol2Δ msh2Δ POL2–URA3 cells defective for Pol ζ (rev3Δ) or Pol η (rad30Δ) were transformed with POL2– or pol2-4–LEU2 plasmids, and 10-fold serial dilutions of isolated transformants were plated onto FOA-containing media as in Figure 1. For comparison, the POL2– and pol2-4–LEU2 plasmids were similarly shuffled into strains that were wild type (WT) for Pols ζ and η (REV3 RAD30) and either msh2Δ or MSH2 (WT MMR). The LEU2 plasmid with no POL2 or pol2-4 gene served as the vector-only control. Colony formation was assessed after incubation at 30° for 3 days. Neither rev3Δ nor rad30Δ rescued pol2-4 msh2Δ synthetic lethality.

Figure 3

Escape from error-induced extinction. (A and B) Plasmid shuffling was used to screen for eex mutants that suppress pol2-4 msh2Δ synthetic lethality. pol2Δ msh2Δ POL2–URA3 strains derived from (A) BY4733 and (B) Y7092 were transformed with pol2-4, POL2, or vector-only LEU2 plasmids. Approximately 10⁴–10⁵ cells from 48 independent pol2-4 transformants of each strain were spotted separately in a 6 × 8 grid on FOA-containing media to select for loss of the POL2–URA3 plasmid and to isolate suppressor mutants. Bona fide suppressors containing pol2-4 as the sole source of Pol ε arose at a rate of 1.9 × 10⁻⁴ eex mutants/cell division in the BY4733 strain [95% confidence interval (C.I.) = 2.6 × 10⁻⁴ – 1.3 × 10⁻⁴] and 4.0 × 10⁻³ in the Y7092 background (95% CI = 4.4 × 10⁻³ – 3.6 × 10⁻³). (C) A similar plasmid-shuffling strategy was used to estimate the rate of escape from pol3-01 msh2Δ synthetic lethality in the Y7092 background (8.8 × 10⁻⁶ eex mutants/cell division; 95% CI = 2.0 × 10⁻⁵ – 1.6 × 10⁻⁶). In C, 10-fold fewer viable cells were plated in each grid position compared to A and B. The POL2 control patches in A are two independent transformants. The POL2 and POL3 control patches in B and C are also from replicate transformants and include 10-fold dilutions of each. Rates (eex mutants/cell division) were calculated as described in Materials and Methods. Red boxes indicate grid positions magnified below each plate.

Figure 4

Amino acid changes in Pol ε eex mutants. Aligned amino-acid sequences of the catalytic subunits of Pols ε and δ (S.c. pol e, S. cerevisiae Pol ε; H.s. pol e, Homo sapiens Pol ε; S.c. pol d, S. cerevisiae Pol δ; H.s. pol d, H. sapiens Pol δ). Secondary structural elements of yeast Pol δ (Swan et al. 2009) are indicated below the alignment and color-coded to depict their domain location (as in Figure 7): rectangles, α-helices; arrows, β-strands; solid lines, loops; dotted line, structure unknown. Amino-acid substitutions encoded by Pol ε eex mutations are shown in black type above the sequence, with the corresponding yeast Pol δ residues in parentheses. Conserved polymerase and exonuclease motifs are framed in green and blue, respectively (Bernad et al. 1989; Wang et al. 1989). Three additional regions of homology (C-1, C-2, and C-3) are framed in black (Huang et al. 1999). Amino-acid conservation is indicated using the following color scheme: red, residues conserved in all four sequences; yellow, residues conserved in three sequences; gray, similar amino acids in at least three sequences.

Figure 5

Growth and antimutator phenotypes conferred by pol2-4 intragenic eex. (A) eex mutations reverse synthetic growth defects associated with pol2-4. The pol2-4,eex mutations were re-engineered into fresh pol2-4–LEU2 plasmids and introduced into wild-type (WT) MMR, msh6Δ, or msh2Δ strains for plasmid shuffling. Strains harboring POL2– or pol2-4–LEU2 plasmids served as controls. Transformants were serially diluted and spotted onto FOA-containing media to assess colony-forming capacity after incubation at 30° for 3 days (WT MMR and msh6Δ) or 4 days (msh2Δ). (B) eex mutations confer antimutator phenotypes. Rates of spontaneous mutation, expressed as Can^r mutants/cell division, were determined from multiple independent fluctuation analyses of each strain. Confidence intervals (95%) for each mutation rate are shown as error bars. The downward red arrow in the gray box indicates the antimutator effect of eex alleles on the pol2-4 mutator phenotype. Symbol patterns indicate POL2 and MSH6 allele status: black left half, POL2; black right half, MSH6; solid black, POL2 MSH6; unfilled left half, pol2-4; unfilled right half, msh6Δ; completely unfilled, pol2-4 msh6Δ. Symbol shapes indicate eex allele status (see key insert): star and hexagon, no eex; triangle, G435C; inverted triangle, V522A; circle, T850M; square, K966Q; diamond, A1153D.

Figure 6

Spontaneous CAN1 mutations from pol2-4 and pol2-4,eex msh6Δ cells. The can1 genes from ∼50 independent Can^r mutants of each strain were PCR-amplified and sequenced. Mutations identified in different strains are color-coded according to the key at the bottom. Each base letter above the wild-type CAN1 sequence indicates an independent base substitution or frameshift (+ or −) mutation. Horizontal lines indicate CAN1 sequences involved in complex mutations, duplications, and deletions. The wild-type POL2 and POL2 msh6Δ spectra are from Herr et al. (2011a).

Figure 7

Locations of Pol ε eex amino-acid substitutions mapped onto the Pol δ structure. (A) Overall distribution of Pol ε eex amino-acid substitutions. The catalytic subunit of yeast Pol δ is depicted as a ribbon diagram with the following color-coded elements: exonuclease domain, red; thumb domain, green; fingers domain, blue; palm domain, purple; amino domain, gray; DNA template strand, brown sticks; primer strand, tan sticks; catalytic carboxylate residues in the polymerase and exonuclease active sites, gray CPK sticks; metal ions, small black spheres; incoming dCTP, green CPK sticks; template G nucleotide, orange CPK sticks. Locations of eex-encoded changes are shown as yellow spheres and labeled by the Pol ε amino-acid substitution with the corresponding Pol δ residue in parentheses. The A1153D substitution is not pictured because it falls in a region where the Pol δ structure is unknown. (B) Locations of Pol ε eex substitutions (yellow spheres) relative to Pol δ eex substitutions (aqua spheres; see (Herr et al. 2011a). The purple sphere at the Exo-Thumb interface is Pol ε G435C, which aligns with Pol δ R475I/G. (C, D, E, and F) Close-up depictions of Pol ε eex substitutions. Important residues are highlighted as space-filling spheres, with yellow indicating positions corresponding to Pol ε eex substitutions. Structure of S. cerevisiae Pol δ is from Swan et al. (2009) (Protein Data Bank accession code 3IAY).