Functional analyses of the C-terminal half of the Saccharomyces cerevisiae Rad52 protein

Abstract

The Saccharomyces cerevisiae Rad52 protein is essential for efficient homologous recombination (HR). An important role of Rad52 in HR is the loading of Rad51 onto replication protein A-coated single-stranded DNA (ssDNA), which is referred to as the recombination mediator activity. In vitro, Rad52 displays additional activities, including self-association, DNA binding and ssDNA annealing. Although Rad52 has been a subject of extensive genetic, biochemical and structural studies, the mechanisms by which these activities are coordinated in the various roles of Rad52 in HR remain largely unknown. In the present study, we found that an isolated C-terminal half of Rad52 disrupted the Rad51 oligomer and formed a heterodimeric complex with Rad51. The Rad52 fragment inhibited the binding of Rad51 to double-stranded DNA, but not to ssDNA. The phenylalanine-349 and tyrosine-409 residues present in the C-terminal half of Rad52 were critical for the interaction with Rad51, the disruption of Rad51 oligomers, the mediator activity of the full-length protein and for DNA repair in vivo in the presence of methyl methanesulfonate. Our studies suggested that phenylalanine-349 and tyrosine-409 are key residues in the C-terminal half of Rad52 and probably play an important role in the mediator activity.

Figure 1.

The C-terminal region of Rad52 depolymerizes Rad51. (A) Schematic representation of the Rad52 fragment. The functional regions in S. cerevisiae Rad52 are shown at the top. The region spanning amino acid residues 34–169 corresponds to the highly conserved region among Rad52 orthologs. Amino acid 34 is the N-terminus of the protein. (B) Purified Rad52^233–504. The proteins (2 µg each) were fractionated through a 12% polyacrylamide gel. WT, F and Y denote wild-type, F349A and Y409A mutants, respectively. (C) Size exclusion chromatography analyses of Rad52^233–504, Rad51 and a mixture of the two proteins. The elution profiles of the proteins from a Superdex 200 gel filtration column were monitored by the absorbance at 280 nm. The boxed portion (9.5–14 ml) containing the peak was fractionated in 0.5 ml portions, and each fraction was analyzed by SDS-PAGE (lanes 1–9), as shown. Sedimentation equilibrium analysis of the Rad51–Rad52^233–504 complex (D) or Rad52^233–504 (E). The distribution of the complex in the cell reached equilibrium after 20 h of centrifugation at 16 000 rpm (Rad51–Rad52^233–504 complex) or 32 000 rpm (Rad52^233–504). The distribution of the complex was determined by UV scanning of the cell (280 nm) at incremental steps, as plotted in the lower panel. The radius indicates the distance from the center of the rotor. The upper panel shows the residual differences between the experimental data and the fit for each data point. The estimated molecular mass in the table below is the result of the molecular mass analysis, in which the data were fitted to an ideal single-component model. The theoretical molecular mass was calculated from the amino acid sequences of Rad52^233–504 and Rad51.

Figure 2.

Effects of Rad52^233–504 on the formation of Rad51–ssDNA and Rad51–dsDNA complexes. (A) Schematic representation of the assay. ssDNA (1 µM in nucleotides) (B) or dsDNA (1 µM in nucleotides) (C) was added to the reaction mixture containing Rad52^233–504 and Rad51 (1 µM). Products were stabilized by glutaraldehyde fixation and fractionated through an agarose gel. The fixed Rad51–DNA complexes migrated to two major locations in the gel (B and C, lane 2). Increases in the Rad52^233–504 concentration caused greater increases in the unbound dsDNA, rather than the unbound ssDNA (B and C, compare lanes 6–8). The Rad52^233–504 concentrations were 0.5 µM (lane 4), 1 µM (lane 5), 2 µM (lane 6), 4 µM (lane 7) and 8 µM (lanes 3, 8 and 9). Lanes 3 and 9 are identical, except for the addition of proteinase K after glutaraldehyde fixation.

Figure 3.

F349 and Y409 of Rad52 are essential for the interaction with Rad51. Size exclusion chromatography analyses of Rad52^233–504 F349A-Rad51 (A) and Rad52^233–504 Y409A-Rad51 (B) mixtures. The elution profiles of the proteins from a Superdex 200 gel filtration column were monitored by the absorbance at 280 nm. The boxed portion (9.5–14 ml) containing the peak was fractionated in 0.5 ml portions, and each fraction was analyzed by SDS-PAGE (lanes 1–9), as shown below.

Figure 4.

Mutations in F349 and Y409 of Rad52 impair its ability to repair MMS-induced DNA damage. Wild-type and mutant Rad52 proteins were expressed from a single-copy plasmid in rad52Δ haploid transformants, and their abilities to complement the MMS sensitivity exhibited by the haploid strain were examined by a spot assay. (A) Tenfold serial dilutions of the transformants were spotted onto plates containing 0, 0.25, 0.59 or 1.18 mM MMS and incubated at 30°C for 5 days. (B) Quantitative representation of the MMS sensitivities of the haploid strains. The double mutant displayed higher sensitivity toward MMS, as compared with the single mutants, and was similar to a Rad52 deletion mutant lacking the entire C-terminal half (Rad52^34–237). Black circle, RAD52; green triangle, rad52 Y409A; black diamond, rad52 F349A; orange triangle, rad52 F349A/Y409A; white square, rad52^34–237; and white circle, vector.

Figure 5.

Recombination mediator activity of the Rad52 F349A/Y409A mutant assessed by a DNA strand exchange assay. (A) Schematic representation of the DNA strand exchange reaction. Circular ssDNA (ss) and linear dsDNA (ds) base pair to form a joint molecule (jm), in a process called homologous pairing. The joint molecule is converted into a nicked circular DNA (nc) by DNA strand exchange. (B) A standard reaction was performed by pre-incubating Rad51 (9 µM before the addition of dsDNA) with ssDNA (30 µM before the addition of dsDNA), followed by the addition of RPA (2 µM before the addition of dsDNA) and then by the addition of 1 µl dsDNA (30 µM final concentration) to start the reaction (lane 2). Co-incubation of Rad51 and RPA resulted in the severe inhibition of the DNA strand exchange reaction (lane 3). The inclusion of the indicated amounts of Rad52 (lanes 9–12) or the Rad52 F349A/Y409A mutant (lanes 4–8) led to various degrees of reaction restoration. The indicated concentrations of Rad52 are those before the addition of dsDNA.

Figure 6.

Recombination mediator activity of the Rad52 F349A/Y409A mutant assessed by an ATPase assay. (A) Schematic representation of the ATPase assay. The mediator activity of Rad52 was observed by monitoring the ATPase activity of Rad51 in the presence of RPA-bound ssDNA. The Rad52-mediated replacement of RPA with Rad51 on ssDNA would result in increased levels of ATP hydrolysis by Rad51. In the reaction, RPA was preincubated with circular ssDNA, followed by the addition of Rad52 and Rad51, which were preincubated together before addition. (B) Graphical representation of the ATPase assay. The amount of ADP produced as a function of Rad52 concentration was plotted. Black circle, Rad52; black square, Rad52 F349A/Y409A; white triangle, Rad52^34–237; white diamond, Rad51 alone (no RPA); black diamond, RPA only; cross mark, no protein; white circle, Rad52 alone; white square, Rad52 F349A/Y409A alone; and black triangle, Rad52^34–237 alone.

Figure 7.

Conservation of F349 and Y409 in yeast Rad52 orthologs. (A) The C-terminal region of S. cerevisiae Rad52, spanning amino acid residues 349–412, was aligned with the homologous regions of the yeast Rad52 orthologs from Kluyveromyces lactis, Ashbya gossypii, Candida glabrata and Schizosaccharomyces pombe. The boxed regions indicate the FVTA sequence and the previously identified Rad51-binding region (20). The polymerization motifs of Rad51 from the corresponding yeast species are shown below. (B) Crystal structure of the fourth BRC repeat (BRC4) from the human BRCA2 protein complexed with Rad51 (PDB ID, 1N0W). BRC4 (blue) forms two important interactions with Rad51 (yellow). Both interactions involve phenylalanine residues (F1524 and F1546) that are ∼20 amino acid residues apart. F1524 is part of the FHTA motif that is highly similar to the FVTA sequences used by yeast Rad51 and Rad52.