Rtt107/Esc4 binds silent chromatin and DNA repair proteins using different BRCT motifs

Abstract

Background: By screening a plasmid library for proteins that could cause silencing when targeted to the HMR locus in Saccharomyces cerevisiae, we previously reported the identification of Rtt107/Esc4 based on its ability to establish silent chromatin. In this study we aimed to determine the mechanism of Rtt107/Esc4 targeted silencing and also learn more about its biological functions.

Results: Targeted silencing by Rtt107/Esc4 was dependent on the SIR genes, which encode obligatory structural and enzymatic components of yeast silent chromatin. Based on its sequence, Rtt107/Esc4 was predicted to contain six BRCT motifs. This motif, originally identified in the human breast tumor suppressor gene BRCA1, is a protein interaction domain. The targeted silencing activity of Rtt107/Esc4 resided within the C-terminal two BRCT motifs, and this region of the protein bound to Sir3 in two-hybrid tests. Deletion of RTT107/ESC4 caused sensitivity to the DNA damaging agent MMS as well as to hydroxyurea. A two-hybrid screen showed that the N-terminal BRCT motifs of Rtt107/Esc4 bound to Slx4, a protein previously shown to be involved in DNA repair and required for viability in a strain lacking the DNA helicase Sgs1. Like SLX genes, RTT107ESC4 interacted genetically with SGS1; esc4Delta sgs1Delta mutants were viable, but exhibited a slow-growth phenotype and also a synergistic DNA repair defect.

Conclusion: Rtt107/Esc4 binds to the silencing protein Sir3 and the DNA repair protein Slx4 via different BRCT motifs, thus providing a bridge linking silent chromatin to DNA repair enzymes.

Figure 1

Esc4 was identified in an HMR targeted silencing screen. (A) In the presence of both wild-type silencers, HMR-E and HMR-I, the URA3 reporter gene at HMR is completely silenced in a SIR-dependent manner leading to growth in the presence of the drug 5-FOA. (B) When the HMR-E silencer is deleted (and here replaced with a binding site for the Gal4 protein (G)), the HMR locus is no longer silenced and the URA3 reporter gene is expressed. This loss of silencing at hmr::URA3 leads to no growth on 5-FOA media. (C) When the strain in (B) is transformed with a G_BD plasmid (G_BD-X) capable of causing targeted silencing, this leads to restoration of silencing at hmr::URA3 and growth on 5-FOA. In this way, a G_BD hybrid protein library was screened for factors capable of targeted silencing. A G_BD-Esc4(1–1070) hybrid was identified.

Figure 2

Targeted silencing by Esc4 is SIR-dependent. Targeted silencing by G_BD-Esc4 (1–1070) was assessed in hmr::URA3 silencing reporter strains containing different HMR-E elements replaced by a Gal4 DNA binding site (G), similar to the situation shown in Figure 1B. A TRP1-marked plasmid expressing G_BD alone served as a negative control and targeted silencing by the potent targeted silencing factor G_BD-Esc1(1124–1658) served as a positive control [17]. Targeted silencing by G_BD-Esc4 in strains with wild-type SIR genes (SIR⁺) or sir2Δ, sir3Δ, or sir4Δ mutations was examined.

Figure 3

Saccharomyces cerevisiae Esc4 has six BRCT motifs that are conserved in both sequence and domain architecture among yeast homologs. The six BRCT motifs in Esc4 are indicated with brackets and yellow color and were described previously [49]. The motifs in Esc4 are shown as they align with a subset of Esc4 homologs: three from other budding yeasts Ashbya gossipii, Kluyveromyces lactis and Candida glabrata and one from the evolutionarily distant fission yeast Schizosaccharomyces pombe Brc1. The highly conserved aromatic residue, a hallmark of BRCT motifs, is marked above the alignment with a red asterisk. Blue color indicates similarity with the predominant residue at the position in the alignment, with darkness correlated with greater similarity as calculated by ClustalW software and displayed with Jalview alignment editor's BLOSUM62 score viewing option [67, 68].

Figure 4

Targeted silencing activity is achieved by the C-terminus of Esc4. Three different regions of Esc4, containing either the first four BRCT motifs (1–480), linker region (480–836), or last two BRCTs (818–1070) were tested for targeted silencing in a strain with an HMR-E Aeb::G silencer and URA3 reporter gene at HMR. GBD-Esc1 served as a potent positive control and GBD as a negative control. SIR2-dependence of targeted silencing was also determined.

Figure 5

Genetic evidence that ESC4 and SLX4 function together in a parallel pathway to SGS1 for repair of damaged DNA. Strains were grown in YPD medium to saturation and then ten-fold serial dilutions were spotted onto YPD, YPD + HU (100 mM), or YPD + MMS (0.032% in part A and 0.014% in part B) media. Cells were incubated at 30°C for 2 days (SC and SC + MMS) or 3 days (SC + HU) before being photographed. WT, wild type. An asf1Δ mutant was used as a positive control for MMS and HU sensitivity.