exo1-Dependent mutator mutations: model system for studying functional interactions in mismatch repair

Abstract

EXO1 interacts with MSH2 and MLH1 and has been proposed to be a redundant exonuclease that functions in mismatch repair (MMR). To better understand the role of EXO1 in mismatch repair, a genetic screen was performed to identify mutations that increase the mutation rates caused by weak mutator mutations such as exo1Delta and pms1-A130V mutations. In a screen starting with an exo1 mutation, exo1-dependent mutator mutations were obtained in MLH1, PMS1, MSH2, MSH3, POL30 (PCNA), POL32, and RNR1, whereas starting with the weak pms1 allele pms1-A130V, pms1-dependent mutator mutations were identified in MLH1, MSH2, MSH3, MSH6, and EXO1. These mutations only cause weak MMR defects as single mutants but cause strong MMR defects when combined with each other. Most of the mutations obtained caused amino acid substitutions in MLH1 or PMS1, and these clustered in either the ATP-binding region or the MLH1-PMS1 interaction regions of these proteins. The mutations showed two other types of interactions: specific pairs of mutations showed unlinked noncomplementation in diploid strains, and the defect caused by pairs of mutations could be suppressed by high-copy-number expression of a third gene, an effect that showed allele and overexpressed gene specificity. These results support a model in which EXO1 plays a structural role in MMR and stabilizes multiprotein complexes containing a number of MMR proteins. A similar role is proposed for PCNA based on the data presented.

FIG. 1

Characterization of the mutator phenotype of the exo1-dependent mutator mutants using patch tests. The indicated strains were transformed with either the control vector or the vector containing the EXO1 gene. Then, three colonies each were patched onto a master plate and replica plated onto an SD-Lys plate to evaluate the lys2::InsE-A10 reversion properties of each strain as described in Materials and Methods.

FIG. 2

Schematic representation of the position of the exo1-dependent and pms1-A130V-dependent mutator mutations. The relative positions of the exo1-dependent (A) and pms1-A130V-dependent (B) mutator mutations on the indicated genes are indicated by the arrows. The black vertical bars in PMS1 and MLH1 indicate the position of motifs important for ATP binding. The stippled boxes in PMS1 and MLH1 indicate the position of motifs important for PMS1-MLH1 interactions. The two open vertical boxes in MSH2, MSH3, and MSH6 indicate motifs important for ATP hydrolysis. The single vertical black bars in MSH3, MSH6, and POL32 indicate the motif that is important for interaction with PCNA. The two stippled boxes in EXO1 indicate motifs thought to be important for exonuclease activity.

FIG. 3

Maps of the positions of the amino acid residues affected by the exo1-dependent and pms1-A130V-dependent mutator mutations onto the crystal structure of the ADP-bound form of the N-terminal fragment of MutL. The indicated structures were generated using the RasMol program from Roger Sayle (University of California at San Diego) with coordinates of the 40-kDa ATPase fragment of E. coli MutL complexed with ADP from the Protein Data Bank (ID1B62). The ribbon diagram of MutL is in blue. The ADP is indicated in yellow. The residues affected by the exo1-dependent mlh1 mutations (A), the exo1-dependent pms1 mutations (B), and the pms1-A130V-dependent mlh1 mutations are indicated in red. The relevant E. coli MutL amino acid residue numbers are indicated followed by the relevant yeast amino acid residue numbers in parentheses.

FIG. 4

Suppression of representation exo1-dependent mutator mutations by increased expression of PCNA. The indicated strains were transformed with either the control vector or the vector containing the POL30 gene encoding PCNA. Then, three colonies each were patched onto a master plate and replica plated onto an SD-Lys plate to evaluate the lys2::InsE-A10 reversion properties of each strain as described in Materials and Methods.