Rsp5, a ubiquitin-protein ligase, is involved in degradation of the single-stranded-DNA binding protein rfa1 in Saccharomyces cerevisiae

Abstract

In Saccharomyces cerevisiae, RAD1 and RAD52 are required for alternate pathways of mitotic recombination. Double-mutant strains exhibit a synergistic interaction that decreases direct repeat recombination rates dramatically. A mutation in RFA1, the largest subunit of a single-stranded DNA-binding protein complex (RP-A), suppresses the recombination deficiency of rad1 rad52 strains (J. Smith and R. Rothstein, Mol. Cell. Biol. 15:1632-1641, 1995). Previously, we hypothesized that this mutation, rfa1-D228Y, causes an increase in recombinogenic lesions as well as the activation of a RAD52-independent recombination pathway. To identify gene(s) acting in this pathway, temperature-sensitive (ts) mutations were screened for those that decrease recombination levels in a rad1 rad52 rfa1-D228Y strain. Three mutants were isolated. Each segregates as a single recessive gene. Two are allelic to RSP5, which encodes an essential ubiquitin-protein ligase. One allele, rsp5-25, contains two mutations within its open reading frame. The first mutation does not alter the amino acid sequence of Rsp5, but it decreases the amount of full-length protein in vivo. The second mutation results in the substitution of a tryptophan with a leucine residue in the ubiquitination domain. In rsp5-25 mutants, the UV sensitivity of rfa1-D228Y is suppressed to the same level as in strains overexpressing Rfa1-D228Y. Measurement of the relative rate of protein turnover demonstrated that the half-life of Rfa1-D228Y in rsp5-25 mutants was extended to 65 min compared to a 35-min half-life in wild-type strains. We propose that Rsp5 is involved in the degradation of Rfa1 linking ubiquitination with the replication-recombination machinery.

FIG. 1

The direct repeat recombination constructs for leu2 and SUP4 are depicted. Both assays utilize a direct repeat of two alleles that are separated by plasmid and selectable marker sequences. Recombination events between the repeats that result in the retention of one allele and deletion of the intervening sequences can be selected. The leu2 and SUP4 direct repeats are 2.4 kb and are not drawn to scale. (A) In the leu2 assay (46), both alleles contain a frameshift mutation created by filling in a restriction enzyme site (EcoRI or BstEII). Leu⁺ recombinants are first selected for on medium lacking leucine and further identified as Ura⁻ colonies after replica plating on medium lacking uracil. (B) The SUP4 assay (24) was modified by disrupting URA3 with HIS3 as described in Materials and Methods. The alleles differ by a single nucleotide change in the anticodon. Can^r recombinants are selected on canavanine-containing medium. The strain contains the ochre suppressible ade2-1 allele, which enables colony color to designate which SUP4 allele is retained in the genome after direct repeat recombination: sup4 colonies are red, while SUP4-o colonies remain white. Deletions are confirmed by their failure to grow on histidine-less medium.

FIG. 2

Identification of mrr mutants and cloning of the mrr1 mutation. (A) Papillation phenotype of the mrr1 mutation in rad1 rad52 rfa1-D228Y after replica plating to the selective media for SUP4 and leu2 direct repeat recombination assays. (B) mrr1 is allelic to RSP5. The five ORFs present in pWJ670 are illustrated, as well as the five subclones constructed from this plasmid by using the indicated restriction enzymes. Each subclone was used to determine which of the five ORFs is MRR1. The Sau3A sites represent the junction of the genomic yeast sequences and the plasmid sequences. The thin lines depict the plasmid sequences. The plasmid, pWJ914, was derived from pWJ671 and carries a deletion of an internal BstEII fragment indicated as the dashed line. Each plasmid was tested for complementation of both the ts and the reduced recombination phenotypes in rad1 rad52 rfa1-D228Y mrr1 strains, and the results are shown adjacent to each plasmid.

FIG. 3

Localization of the mutations in rsp5-25. (A) The rsp5-25 mutation was mapped by using a plasmid that contains the wild-type RSP5 allele, digested with indicated restriction enzymes for use in plasmid gap repair experiments. The gapped regions are indicated as thick lines. After transformation into an rsp5-25 strain, plasmid gap repair results in the copying of the corresponding genomic region onto the plasmid. The results of complementation experiments in rad1 rad52 rfa1-D228Y rsp5-25 are depicted. The black boxes on the RSP5 ORF depict the WW domains, and the hatched box represents the ubiquitination domain of the protein. (B) Localization of the two rsp5-25 mutations in the ubiquitination domain. The mutated tryptophan residue is indicated as a boldface letter, and the ochre codon is indicated by an asterisk. The protein sequences of several Rsp5 homologs are also shown to indicate the conservation near the mutated residues.

FIG. 4

Protein blot analysis of different RSP5 alleles. Extracts from two rsp5-25, rsp5-Y647och, rsp5-W650L, and RSP5 strains were prepared, and the total amount of Rsp5 protein was detected by protein blot analysis by using Rsp5 antibody. The full-length Rsp5 proteins (92 kDa) is indicated as Rsp5. In rsp5-25 and rsp5-Y647och strains, an additional truncated form of Rsp5 (72 kDa) was observed and is indicated as Rsp5*. This is due to the ochre codon present at the position of residue 647. The SUP4-o suppressor also present in these strains recognizes the ochre codon to produce full-length Rsp5. However, since SUP4-o suppression is only partial, both full-length and truncated forms of Rsp5 are observed. The other bands are background obtained with the Rsp5 antibody.

FIG. 5

UV survival curves. Wild-type (▴), rfa1-D228Y (■), and rfa1-D228Y rsp5-25 (⧫) strains were exposed to increasing amounts of UV irradiation, and their survival was plotted. Both rsp5-25 strains and rfa1-D228Y strains overexpressing rfa1-D228Y exhibit identical survival to wild-type strains (reference and data not shown).

FIG. 6

Protein stability and mRNA levels of rfa1-D228Y. (A) A protein blot analysis of the stability of Rfa1-D228Y in the RSP5 and rsp5-25 mutant strains was performed. Rfa1-D228Y protein is tagged with HA and is indicated on the blots as Rfa1^•. Other bands are due to background obtained with the HA antibody. The half-lives were calculated as described in Materials and Methods. (B) Protein levels of Rfa1-D228Y in rfa1-D228Y strains overexpressing rfa1-D228Y (lane 1) and rfa1-D228Y rsp5-25 strains (lane 3) were compared by a protein blot analysis. rfa1-D228Y strains (lane 2) were included as a control. (C and D) Measurement of mRNA and protein levels in rfa1-D228Y strains during the cell cycle. To measure mRNA and protein levels of rfa1-D228Y during the cell cycle, cells were released from alpha-factor arrest, samples were taken every 15 min, total mRNA was extracted, and total yeast protein extracts were prepared as described in Materials and Methods. (C) RNA blots were performed by using an internal fragment of RFA1 as a probe and URA3 as a loading control. The mRNA levels for URA3 were unchanged (data not shown). (D) Protein blot analysis was performed by using Rfa1 antibody. As a control, the same blot was reprobed with Rsp5 antibody. No variation in Rsp5 protein levels was detected (data not shown). After release from arrest, cells with small buds started to appear at 30 min and peaked at 45 min for the first cell cycle. The peak for cells with small buds for the second cell cycle occurs at 105 min.

FIG. 7

Degradation of Rfa1 during S phase and its in vivo ubiquitination. (A and B) Protein blot analyses of the stability of Rfa1 and Rfa1-D228Y during S phase. RFA1 and rfa1-D228Y cells were arrested with HU at S phase, and samples were taken every 15 min. Total yeast protein extracts were prepared and subjected to protein blot analysis by using the Rfa1 antibody. The half-lives of each were identical to that determined for logarithmically growing cells (see Fig. 6A). (C) In vivo ubiquitination of Rfa1. Wild-type, cim3, and cim5 mutant strains that overexpress rfa1-D228Y under the GAL1 promoter were pulse labeled with [³⁵S]methionine. After incubation at 37°C, the nonpermissive temperature for the cim mutants, total yeast protein extract was prepared, and immunoprecipitations were performed with Rfa1 antibody. By immunoprecipitation, putative ubiquitinated forms of Rfa1 were visualized in cim3, cim5, and wild-type strains respectively. The bands corresponding to the nonubiquitinated forms of the Rfa1 are indicated. The higher-molecular-weight bands, indicated by brackets, represent the putative ubiquitinated forms.