Role of the Srs2-Rad51 Interaction Domain in Crossover Control in Saccharomyces cerevisiae

Abstract

Saccharomyces cerevisiae Srs2, in addition to its well-documented antirecombination activity, has been proposed to play a role in promoting synthesis-dependent strand annealing (SDSA). Here we report the identification and characterization of an SRS2 mutant with a single amino acid substitution (srs2-F891A) that specifically affects the Srs2 pro-SDSA function. This residue is located within the Srs2-Rad51 interaction domain and embedded within a protein sequence resembling a BRC repeat motif. The srs2-F891A mutation leads to a complete loss of interaction with Rad51 as measured through yeast two-hybrid analysis and a partial loss of interaction as determined through protein pull-down assays with purified Srs2, Srs2-F891A, and Rad51 proteins. Even though previous work has shown that internal deletions of the Srs2-Rad51 interaction domain block Srs2 antirecombination activity in vitro, the Srs2-F891A mutant protein, despite its weakened interaction with Rad51, exhibits no measurable defect in antirecombination activity in vitro or in vivo Surprisingly, srs2-F891A shows a robust shift from noncrossover to crossover repair products in a plasmid-based gap repair assay, but not in an ectopic physical recombination assay. Our findings suggest that the Srs2 C-terminal Rad51 interaction domain is more complex than previously thought, containing multiple interaction sites with unique effects on Srs2 activity.

Keywords: DNA repair; crossover control; genome stability; helicase; protein interaction; recombination.

Figure 1

Mutations targeting the putative Srs2 BRC repeat-like motif weaken the Srs2–Rad51 physical interaction. (A) Schematic of Srs2 domains. (B) Sequence alignment of the putative Srs2 BRC repeat motif with other BRC repeat motifs in the indicated species. Glycine (G): orange; phenylalanine (F), isoleucine (I), leucine (L), alanine (A): blue; serine (S), threonine (T): green; proline (P): yellow; aspartic acid (D), glutamic acid (E): violet. (C) Quantitative β-galactosidase assay analyzing the physical interaction between full-length Srs2 and Rad51. (D) Ni-NTA pull-down with 1.3 μM Rad51 and 0.4 μM His9-tagged Srs2 (WT or srs2-F891A) at increasing concentrations of KCl (0–400 mM). Srs2 was visualized using a Biorad stain-free imaging system. Rad51 bands were detected by immunoblot analysis. The amount of Rad51 pulled down was normalized against the amount of Srs2 pull-down in each lane. Shown are means ±1 SE, n = 2–3.

Figure 2

srs2-F891A mutation has no effect on UV sensitivity, rad55∆ IR sensitivity, or rad18∆ MMS sensitivity. (A) Quantitative UV survival assay. The indicated haploid W303 strains were grown to stationary phase in liquid YPD medium, plated onto YPD, UV irradiated, and then grown at 30° for 2 days. The number of surviving colonies was normalized to the number of viable colonies in the unirradiated control samples. (B) Quantitative IR survival assay. The indicated haploid W303 strains were grown to midlog phase, irradiated with IR (0–200 Gy), and then plated onto YPD and grown at 25° for 2 days. srs2-F891A, unlike srs2∆, does not suppress rad55∆ IR sensitivity. (C) Qualitative MMS survival assay. Strains were grown to stationary phase in liquid YPD. Serial fivefold dilutions of the strains were then spotted onto YPD or 0.0006% MMS and grown at 30° for 1 day. The srs2∆ rad18∆ double mutant appears white because it is ADE2+ unlike the other strains depicted. srs2∆ suppresses rad18∆ MMS sensitivity as expected. (D) Immunoblot analysis of whole cell protein TCA extraction of WT and srs2-F891A strains (W303 background). srs2∆ was used as a negative control. Shown are means ±1 SE, n = 3.

Figure 3

Srs2-F891A and Srs2 disrupt Rad51 filaments at similar rates. In vitro disruption of Rad51 filaments. (A) A schematic of the assay where Rad51 filaments were assembled on ssDNA immobilized to magnetic beads and disrupted by Srs2 or Srs2-F891A. (B) Supernatant Rad51 bound to scavenger DNA with 25 nM of Srs2 or Srs2-F891A was analyzed by immunoblot. A representative blot is shown. (C) Quantification of supernatant Rad51 was normalized to fold increase over no Srs2 added. Shown are means ±1 SD, n = 3. The difference is not significant by a Student’s t-test.

Figure 4

srs2-F891A (W303 strain background) increases the relative CO frequency without affecting overall repair efficiency. (A) Schematic of the gap repair system. Plasmid linearized within the HIS3 reading frame was transformed into yeast strain where his3∆3′ serves as the repair template for the linearized plasmid. The stable presence of the URA3 marker was used as an indicator of plasmid integration, representing CO events, while the unstable presence of the URA3 marker was used an indicator of plasmid repair without integration, representing NCO events. (B) Gap repair assay shows that srs2-F891A (W303 strain background) has no defect in repair efficiency but exhibits a clear shift from NCO repair products to CO repair products similar to srs2∆. All strains were MMR-defective (mlh1∆). Asterisks indicate a significant difference when compared to WT using a Student’s t-test (P $\le$ 0.05). (C) Gap repair analysis of srs2-F891A in SJR strain background recapitulates findings in the W303 strain background.

Figure 5

Ectopic physical recombination assay. (A) Schematic of the ectopic physical recombination assay. (B) Representative Southern blot analysis of the CO and NCO repair products for the strains shown in the figure. (C) Quantitative analysis of the CO to NCO ratio. The lower CO band intensities were divided by the sum of the lower CO and NCO bands. In contrast to the gap repair system, the ectopic physical recombination assay does not exhibit a shift from NCO to CO repair products even in an mph1∆ background. Shown are means ±1 SE, n = 3.

Figure 6

Spontaneous inverted repeat assay. (A) Schematic of the intron-based inverted repeat recombination assay. Inverted repeats of cβ2 intron sequence (blue) are fused to intron splice sites (gray) positioned adjacent to the 5′ and 3′ halves of the HIS3 gene (yellow). Spontaneous CO events at cβ2 intron sequence reorient and bring together the his3 halves. The resulting transcript produces a functional HIS3 mRNA after splicing out the cβ2 intervening sequence. (B) Recombination rates for wild type and srs2∆ recapitulate findings previously published (Spell and Jinks-Robertson 2003, 2004b). srs2-F891A mutant exhibits rates of recombination comparable to that of the wild type.

Figure 7

Both Srs2-F891A and Srs2 dissolve the D-loop in vitro. (A) Schematic of the reconstituted D-loop disruption assay. (B) Representative gel of Srs2 and Srs2-F891A titration in D-loop disruption assay. (C) Quantification of D-loops in the presence of increasing concentrations of Srs2 and Srs2-F891A proteins. Shown are means ±1 SD, n = 3.