Mutations affecting a putative MutLalpha endonuclease motif impact multiple mismatch repair functions

Abstract

Mutations in DNA mismatch repair (MMR) lead to increased mutation rates and higher recombination between similar, but not identical sequences, as well as resistance to certain DNA methylating agents. Recently, a component of human MMR machinery, MutLalpha, has been shown to display a latent endonuclease activity. The endonuclease active site appears to include a conserved motif, DQHA(X)(2)E(X)(4)E, within the COOH-terminus of human PMS2. Substitution of the glutamic acid residue (E705) abolished the endonuclease activity and mismatch-dependent excision in vitro. Previously, we showed that the PMS2-E705K mutation and the corresponding mutation in Saccharomyces cerevisiae were both recessive loss of function alleles for mutation avoidance in vivo. Here, we show that mutations impacting this endonuclease motif also significantly affect MMR-dependent suppression of homeologous recombination in yeast and responses to S(n)1-type methylating agents in both yeast and mammalian cells. Thus, our in vivo results suggest that the endonuclease activity of MutLalpha is important not only in MMR-dependent mutation avoidance but also for recombination and damage response functions.

Fig. 1

(a) Pms1/2 domains and location of E707/E705. The ATPase domain, the Mlh1-interacting domain and a motif critical for MutLα endonuclease activity are indicated by hatched, black and grey boxes, respectively. The “mutated” E707/E705 residue is indicated by an arrow. Numbers correspond to the amino acid position in the protein in accordance with the current data bases. (b) Papillation phenotypes of the wild type, pms1Δ, pms1-E707K strains. The relative mutator effects in wild type, pms1Δ, pms1-E707K strains were detected by monitoring forward mutation at CAN1 (multiple types of mutations) vs. reversion at his1-7 (intragenic missense suppressor mutations), his7-2 (+1 bp frameshifts in a stretch of 7A/T bp) or hom3-10 (−1 bp frameshifts in a stretch of 7A/T bp) by replica plating onto appropriate selective media. It should be noted that pms1Δ strains contain ade2-1 that results in solid red colored papilliae, whereas pms1-E707K strains contain a ade2-9G (+1 bp frameshift in a stretch of 9G/C bp) reporter that gives rise to red/white papilliae due to DNA polymerase slippage in the 9 G/C run.

Fig. 2

MNNG-induced killing of rad52Δ mgt1Δ, rad52Δ mgt1Δ pms1Δ and rad52Δ mgt1Δ pms1-E707K yeast strains: (a) comparison of MNNG responses for wild type, mgt1Δ, rad52Δ and pms1 mutants in combinations as measured by a serial dilution spot test. Late-log phase cells were spotted on CSM and CSM + 0.35 μM MNNG plates at 5-fold serial dilutions as described in the Section 2. (b) MNNG survival curves. mgt1Δ (■), rad52Δ mgt1Δ pms1Δ (▲), rad52Δ mgt1Δ pms1-E707K (●) and rad52Δ mgt1Δ (♦) strains were exposed for 45 min to increasing amounts of MNNG and survival was determined. Error bars represent standard error of the mean. LD₅₀ values for mgt1Δ, rad52Δ mgt1Δ pms1Δ, rad52 mgt1Δ pms1-E707K and rad52Δ mgt1Δ strains were 9.7, 1.8, 1.5 and 0.25 μM MNNG, respectively.

Fig. 3

Cell cycle arrest phenotype of rad52Δ mgt1Δ mutant strains after exposure to MNNG. After 45 min exposure to 0.5 μM MNNG, cells from rad52Δ mgt1Δ, rad52Δ mgt1Δ pms1Δ and rad52Δ mgt1Δ pms1-E707K cultures were collected every hour, fixed and stained with DAPI: (a) after 2 h, rad52Δ mgt1Δ accumulated as large-budded cells with a single nucleus and (b) quantification of cell cycle arrest after MNNG exposure. Distribution of cells from rad52Δ mgt1Δ, rad52Δ mgt1Δ pms1Δ and rad52Δ mgt1Δ pms1-E707K cultures after exposure to MNNG are depicted as percentages. Cells were divided into categories as cells with no buds, small buds, large buds and divided cells. For each time point, at least 100 cells were counted in two separate experiments.

Fig. 4

Survival of MEFs after treatment with 6-tg: (a) 8 days after a 24 h exposure to 6-tg, CFU were determined for Pms2-deficient (♦), hPMS2-E705K expressing (▲) and low and high hPMS2-expressing (■, ●) MEF cells. Error bars represent standard error of the mean. LD₅₀ values were 1.33, 1.31, and 0.28 μM 6-tg for Pms2-deficient, hPMS2-E705K expressing and low hPMS2-expressing MEF cells, respectively. (b) Expression of hPMS2 and hPMS2-E705K in stably-transfected MEFs. To determine the level of hPMS2-E705K vs. wild type hPMS2 expression levels in MEFs, whole cell lysates were prepared and compared by Western analysis. Endogenous Msh6 expression was used to normalize hPMS2 expression.