The roles of fission yeast exonuclease 5 in nuclear and mitochondrial genome stability

Abstract

The Exo5 family consists of bi-directional, single-stranded DNA-specific exonucleases that contain an iron-sulfur cluster as a structural motif and have multiple roles in DNA metabolism. S. cerevisiae Exo5 is essential for mitochondrial genome maintenance, while the human ortholog is important for nuclear genome stability and DNA repair. Here, we identify the Exo5 ortholog in Schizosaccharomyes pombe (spExo5). The activity of spExo5 is highly similar to that of the human enzyme. When the single-stranded DNA is coated with single-stranded DNA binding protein RPA, spExo5 become a 5'-specific exonuclease. Exo5Δ mutants are sensitive to various DNA damaging agents, particularly interstrand crosslinking agents. An epistasis analysis places exo5⁺ in the Fanconi pathway for interstrand crosslink repair. Exo5⁺ is in a redundant pathway with rad2⁺, which encodes the flap endonuclease FEN1, for mitochondrial genome maintenance. Deletion of both genes lead to severe depletion of the mitochondrial genome, and defects in respiration, indicating that either spExo5 or spFEN1 is necessary for mitochondrial DNA metabolism.

Keywords: Exo5; Exonuclease; Fission yeast; Interstrand crosslink repair; Mitochondrial replication.

Figure 1.. Biochemical activities of spExo5.

(A) Conserved sequence motifs in the Exo5 family. Brown, cysteines that ligand the Fe-S cluster; red, catalytic residues. (B) 12% SDS-PAGE of purified spExo5. Staining was with Coomassie. (C) UV spectrum of purified spExo5; Absorption at 410 nm is indicative of the presence of a Fe-S cluster. (D) Comparison of activities of S. cerevisiae, S. pombe, and human Exo5 on a 34-nucleotide oligonucleotide. Standard assays contained 10 nM 5’-³²P-labeled ssDNA substrate at the indicated concentrations of scExo5, spExo5, or hsExo5 for 4 min at 30 °C. Samples were analyzed on a 7 M urea-17% polyacrylamide gel. (E) Dependence of hExo5 and spExo5 activity on the concentration of 34-mer ssDNA. Data were fit to a Michaelis-Menten model. For spExo5, Km = 26 ± 2.2 nM and Vmax = 1 ± 0.2 sec⁻¹; for hExo5, Km = 3,200 ± 800 nM and Vmax = 0.012 ± 0.002 sec⁻¹. (F) Standard assays on 10 nM of the indicated fully dsDNA (lanes 2-5), ssDNA (lanes 6-8) and partial dsDNA (lanes 9-14) substrates, with indicated concentrations of spExo5. M, ladder from partial digestion of ssDNA with snake venom phosphodiesterase, a 3’-exonuclease. (G) 5’-directionality of spExo5 enforced by human or S. cerevisiae RPA. Standard assay mixtures used 10 nM 5’-³²P-labeled (dT)₆₅ substrate and 0.15 nM spExo5 with either no RPA, or 10, 12.5, 15, 20, 25, 30, or 40 nM of the indicated RPA for 5 min at 30 °C. Lanes 1 and 2 were control assays without Exo5 and with 40 nM of the indicated RPA.

Figure 2.. Mitochondrial function of spExo5.

(A) Exo5Δ strains containing the pBL289 (exo5⁺-FLAG₃) series of plasmids, placed under control of the nmt promoter, or empty vector were grown under non-inducing conditions (5 μM thiamine), and extracts subjected to Western analysis with 3xFLAG antibodies. *, non-specific band. (B) Isogenic strains with the indicated Rad2 or Exo5 deletions, and the Rad2Δ Exo5Δ double mutant containing the pBL288 series of plasmids (last 4 entries), were grown on selective media (-Leu), and ten-fold serial dilutions plated on rich media with either glucose or glycerol. Growth was for 3 days (YES) or for 6 days (YEG) at 30 °C. (C) Isogenic strains with the indicated Rad2 or Exo5 deletions were grown on rich media and chromosomal DNA isolated and digested with restriction enzymes AclI, AgeI, pstI, and PvuII, and separated on a 1% agarose gel. Staining was with GelRed and the fluorescence was recorded with a Typhoon phosphoimager in the fluorescence mode. The enzyme mixture cuts chromosomal DNA into small fragments but cuts the ribosomal DNA array only once per repeat (10.8 kb) and the mtDNA twice (15.7+3.8). The ratio of the 15.7/10.8 bands was quantified, with that of wild-type set to (1). Bottom, 10% of undigested input DNA. (D) Plasmid containing strains from (B) were grown on selective media with 5 μM thiamine. Chromosomal DNA was isolated and restriction enzyme digested, and separated on a 0.7 % agarose gel. Digestion and analysis was exactly as in (C).

Figure 3.. SpExo5 interacts with spRPA and spPCNA.

(A) Co-immunopurification with spExo5-FLAG₃ from S. pombe cells grown under low-expressive conditions (5 μM thiamine). Elutions from the 3xFLAG beads were analyzed on a 10%-SDS PAGE gel and stained with coomassie brilliant blue. (B) Immunoblot analysis of co-immunopurifications of extracts from exo5Δ cells containing either empty vector, or plasmids containing 3xFLAG-Exo5 or the nuclease-defective 3xFLAG-spExo5-D207A mutant. Antibodies used against 3xFLAG (spExo5-FLAG₃), spSsb1 (RPA70 subunit), and spPcn1 (PCNA). (C) Immunoblot analysis of extracts as in (B) either mock-treated or treated with DNaseI. The panels in (B) and (C) were assembled from the full Western blots shown in Supplementary Fig. S3C,D.

Figure 4.. Overexpression of SpExo5 causes checkpoint activation and cell death.

(A) Overexpression of spExo5 leads to cell death in S. pombe cells regardless of checkpoint proficiency. Cells either WT or rad3Δ containing pREP3x-spExo5 or pREP3x-spExo5-D207A were grown overnight in EMM-leu supplemented with 5μM thiamine. Cells were washed serially diluted on plates either containing 15 μM or 0 μM thiamine and incubated 2-3 day at 30 °C. (B) Light microscopy of cells from (A), either grown on plates containing 15 μM or 0 μM thiamine for 2-3 days.

Figure 5.. Exo5 genetic interactions in DNA repair pathways.

(A) Epistasis experiment with exo1Δ Ten-fold serial dilutions were plated on YES plates and exposed to UV-irradiation, or plated on YES supplemented with the indicated concentration of MMS. Right panel was rearranged for presentation; the original is shown in Supplementary Fig. S3E. (B) Sensitivity to acute psoralen treatment as described in the methods. Cells were washed twice with PBS buffer and diluted for spotting on YES plates. (C) Epistasis experiment with pso2Δ for sensitivity to chronic cis-platin treatment as described in the methods. Cells were spotted in ten-fold dilutions. (D) Epistasis experiments with rad2Δ for sensitivity to chronic cis-platin and MMS. (E) Epistasis experiments with fml1Δ for sensitivity to chronic cis-platin. Selected sections are shown; the full figure is in Supplementary Fig. S4B. (F) Epistasis experiments withpli1Δ for sensitivity to chronic cis-platin.