Mutants defective in Rad1-Rad10-Slx4 exhibit a unique pattern of viability during mating-type switching in Saccharomyces cerevisiae

Abstract

Efficient repair of DNA double-strand breaks (DSBs) requires the coordination of checkpoint signaling and enzymatic repair functions. To study these processes during gene conversion at a single chromosomal break, we monitored mating-type switching in Saccharomyces cerevisiae strains defective in the Rad1-Rad10-Slx4 complex. Rad1-Rad10 is a structure-specific endonuclease that removes 3' nonhomologous single-stranded ends that are generated during many recombination events. Slx4 is a known target of the DNA damage response that forms a complex with Rad1-Rad10 and is critical for 3'-end processing during repair of DSBs by single-strand annealing. We found that mutants lacking an intact Rad1-Rad10-Slx4 complex displayed RAD9- and MAD2-dependent cell cycle delays and decreased viability during mating-type switching. In particular, these mutants exhibited a unique pattern of dead and switched daughter cells arising from the same DSB-containing cell. Furthermore, we observed that mutations in post-replicative lesion bypass factors (mms2Delta, mph1Delta) resulted in decreased viability during mating-type switching and conferred shorter cell cycle delays in rad1Delta mutants. We conclude that Rad1-Rad10-Slx4 promotes efficient repair during gene conversion events involving a single 3' nonhomologous tail and propose that the rad1Delta and slx4Delta mutant phenotypes result from inefficient repair of a lesion at the MAT locus that is bypassed by replication-mediated repair.

Figure 1.—

Synthesis-dependent strand annealing model for mating-type switching in S. cerevisiae (adapted from Pâques and Haber 1999). (A) Only the MATa and HMLα loci are shown. Mating-type switching is initiated by a DSB formed by HO endonuclease at the MATa locus near the Y/Z1 junction. This is followed by 5′–3′ resection to create 3′ single-stranded ends, and the 3′-end with homology to the HMLα donor sequence initiates strand invasion and primes DNA synthesis off of the donor template. Strand displacement from the donor sequence followed by annealing onto the broken chromosome results in the formation of a 3′ single-stranded nonhomologous tail that must be excised prior to the subsequent DNA synthesis and ligation steps. Rad1-Rad10-Slx4 is hypothesized to act in 3′ nonhomologous tail removal at this step. (B) Mating-type switching involving two nonhomologous ends due to insertion of KANMX sequence on the distal side of the break. A 3′ nonhomologous tail removal step is required to allow priming of DNA synthesis off of the invading strand. Repair then proceeds as above.

Figure 2.—

Southern blot analysis of mating-type switching in wild-type and rad1Δ strains: (A) Diagram of the MAT locus showing the restriction sites used for Southern blot analysis, expected fragment lengths, and location of the probes used for detection of mating-type switching. (B) Analysis of digested DNA for wild-type and rad1Δ mutants induced for mating-type switching. Experiments were performed at least three times, with representative time courses shown. (C) Quantification of repair efficiency as described in materials and methods.

Figure 3.—

Msh2 localization to the DSB in wild-type, rad1Δ, and donorless mutants. (A) Location of primers used for semiquantitative PCR following Msh2 chromatin immunoprecipitation. (B) Representative chromatin immunoprecipitation and PCR detection of Msh2 localization to MAT during mating-type switching in wild-type, rad1Δ, and donorless mutants. Since the Ya sequence is removed during mating-type switching, the input signal is also shown using primers to an unrelated locus (CRY1). (C) For each time point, the Msh2 ChIP signal was set relative to the t = 0 signal, with the maximum signal for each time course set as 1.0 to compare the relative the timing of Msh2 localization. Each data point represents the mean of three to four experiments ± SEM.

Figure 4.—

FACS analysis of cells undergoing mating-type switching. (A) Bar graphs of the percentage of cells in G₁, S, or G₂/M phases of the cell cycle at 0, 2, 4, and 6 hr following induction of mating-type switching in wild-type, rad1Δ, and donorless strains (average of at least three experiments ± SEM). See materials and methods for details. The increase in the percentage of G₂/M cells in rad1Δ mutants relative to wild type is statistically significant (P < 0.01 at t = 2 and t = 4 hr, Student's t-test). (B) Representative FACS profile for wild-type cells at t = 0, with vertical gates separating the 1n (G₁) and 2n (G₂/M) DNA content.

Figure 5.—

Model for mating-type switching facilitated by DNA replication. We propose that mating-type switching can be mediated by ongoing DNA replication. DSB formation, 5′–3′ resection, strand invasion, synthesis, and repair of the invading strand occur as shown in Figure 1 and as predicted by SDSA models. The partially repaired recombination intermediate shown at the top containing a single-stranded break can then be acted on by either the DNA replication machinery or the Rad1-Rad10-Slx4 complex. In the presence of Rad1-Rad10-Slx4, the 3′ Ya nonhomologous tail is removed efficiently prior to, during, or following DNA replication, and once DNA replication has been completed, the cell can divide to produce two viable cells of the switched mating type. In the absence of Rad1-Rad10-Slx4, mating-type switching products are produced largely by replication of the partially repaired recombination intermediate to yield either one switched and one dead daughter cell or, after the action of inefficient nucleases, two viable switched daughters.