Distinct roles of two separable in vitro activities of yeast Mre11 in mitotic and meiotic recombination
1998; Springer Nature; Volume: 17; Issue: 21 Linguagem: Inglês
10.1093/emboj/17.21.6412
ISSN1460-2075
AutoresMunenori Furuse, Yuko Nagase, Hideo Tsubouchi, Kimiko Murakami‐Murofushi, Takehiko Shibata, Kunihiro Ohta,
Tópico(s)Endoplasmic Reticulum Stress and Disease
ResumoArticle2 November 1998free access Distinct roles of two separable in vitro activities of yeast Mre11 in mitotic and meiotic recombination Munenori Furuse Munenori Furuse Cellular and Molecular Biology Laboratory, The Institute of Physical and Chemical Research (RIKEN), Wako-shi, Saitama, 351-0198 Japan The Graduate School of Science and Engineering, Saitama University, Urawa-shi, Saitama, 338-8570 Japan Search for more papers by this author Yuko Nagase Yuko Nagase Cellular and Molecular Biology Laboratory, The Institute of Physical and Chemical Research (RIKEN), Wako-shi, Saitama, 351-0198 Japan Faculty of Science, Ochanomizu University, Bunkyo-ku, Tokyo, 112-8610 Japan Search for more papers by this author Hideo Tsubouchi Hideo Tsubouchi Department of Biology, Osaka University, Toyonaka, Osaka, 560 Japan Search for more papers by this author Kimiko Murakami-Murofushi Kimiko Murakami-Murofushi Faculty of Science, Ochanomizu University, Bunkyo-ku, Tokyo, 112-8610 Japan Search for more papers by this author Takehiko Shibata Takehiko Shibata Cellular and Molecular Biology Laboratory, The Institute of Physical and Chemical Research (RIKEN), Wako-shi, Saitama, 351-0198 Japan The Graduate School of Science and Engineering, Saitama University, Urawa-shi, Saitama, 338-8570 Japan Search for more papers by this author Kunihiro Ohta Corresponding Author Kunihiro Ohta Cellular and Molecular Biology Laboratory, The Institute of Physical and Chemical Research (RIKEN), Wako-shi, Saitama, 351-0198 Japan Search for more papers by this author Munenori Furuse Munenori Furuse Cellular and Molecular Biology Laboratory, The Institute of Physical and Chemical Research (RIKEN), Wako-shi, Saitama, 351-0198 Japan The Graduate School of Science and Engineering, Saitama University, Urawa-shi, Saitama, 338-8570 Japan Search for more papers by this author Yuko Nagase Yuko Nagase Cellular and Molecular Biology Laboratory, The Institute of Physical and Chemical Research (RIKEN), Wako-shi, Saitama, 351-0198 Japan Faculty of Science, Ochanomizu University, Bunkyo-ku, Tokyo, 112-8610 Japan Search for more papers by this author Hideo Tsubouchi Hideo Tsubouchi Department of Biology, Osaka University, Toyonaka, Osaka, 560 Japan Search for more papers by this author Kimiko Murakami-Murofushi Kimiko Murakami-Murofushi Faculty of Science, Ochanomizu University, Bunkyo-ku, Tokyo, 112-8610 Japan Search for more papers by this author Takehiko Shibata Takehiko Shibata Cellular and Molecular Biology Laboratory, The Institute of Physical and Chemical Research (RIKEN), Wako-shi, Saitama, 351-0198 Japan The Graduate School of Science and Engineering, Saitama University, Urawa-shi, Saitama, 338-8570 Japan Search for more papers by this author Kunihiro Ohta Corresponding Author Kunihiro Ohta Cellular and Molecular Biology Laboratory, The Institute of Physical and Chemical Research (RIKEN), Wako-shi, Saitama, 351-0198 Japan Search for more papers by this author Author Information Munenori Furuse1,2, Yuko Nagase1,3, Hideo Tsubouchi4, Kimiko Murakami-Murofushi3, Takehiko Shibata1,2 and Kunihiro Ohta 1 1Cellular and Molecular Biology Laboratory, The Institute of Physical and Chemical Research (RIKEN), Wako-shi, Saitama, 351-0198 Japan 2The Graduate School of Science and Engineering, Saitama University, Urawa-shi, Saitama, 338-8570 Japan 3Faculty of Science, Ochanomizu University, Bunkyo-ku, Tokyo, 112-8610 Japan 4Department of Biology, Osaka University, Toyonaka, Osaka, 560 Japan *Corresponding author. E-mail: [email protected] The EMBO Journal (1998)17:6412-6425https://doi.org/10.1093/emboj/17.21.6412 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info In Saccharomyces cerevisiae, Mre11 protein is involved in both double-strand DNA break (DSB) repair and meiotic DSB formation. Here, we report the correlation of nuclease and DNA-binding activities of Mre11 with its functions in DNA repair and meiotic DSB formation. Purified Mre11 bound to DNA efficiently and was shown to have Mn2+-dependent nuclease activities. A point mutation in the N-terminal phosphoesterase motif (Mre11D16A) resulted in the abolition of nuclease activities but had no significant effect on DNA binding. The wild-type level of nuclease activity was detected in a C-terminal truncated protein (Mre11ΔC49), although it had reduced DNA-binding activity. Phenotypes of the corresponding mutations were also analyzed. The mre11D16A mutation conferred methyl methanesulfonate-sensitivity to mitotic cells and caused the accumulation of unprocessed meiotic DSBs. The mre11ΔC49 mutant exhibited almost wild-type phenotypes in mitosis. However, in meiosis, no DSB formation could be detected and an aberrant chromatin configuration was observed at DSB sites in the mre11ΔC49 mutant. These results indicate that Mre11 has two separable functional domains: the N-terminal nuclease domain required for DSB repair, and the C-terminal dsDNA-binding domain essential to its meiotic functions such as chromatin modification and DSB formation. Introduction Double-strand DNA lesions caused by factors such as alkylating agents, ionizing radiation, active oxygen radicals and abortive DNA replication, or arising from meiotically induced double-strand DNA breaks (DSBs) at recombination hot spots, are repaired by DSB-repair pathways. In the yeast Saccharomyces cerevisiae, there are at least three major pathways for DSB repair. The first two are homologous recombination between homologous sequences (Shinohara and Ogawa, 1995) and non-homologous end-joining (NHEJ) (Boulton and Jackson, 1996; Tsukamoto and Ikeda, 1998) which is a direct ligation of DSB ends. DNA sequences are usually preserved after repair by homologous recombination (accurate repair). In contrast, the other single-strand annealing pathway (Kramer et al., 1994; Mezard and Nicolas, 1994; Haber, 1995; Boulton and Jackson, 1996) is accompanied by a deletion in DNA sequence (error-prone repair). Mre11, Rad50 and Xrs2 proteins in yeast are assumed to play central roles in the homologous recombination and NHEJ pathways probably as a multiple protein complex (Johzuka and Ogawa, 1995; Ogawa et al., 1995). This is consistent with the results that the phenotypes of disruption in each gene are indistinguishable from each other, and there is no synergistic effect in double null mutations. Null mutations of these three genes confer hypersensitivity to methyl methanesulfonate (MMS) and ionizing radiation (Game and Motimer, 1974; Farnet et al., 1988; Ivanov et al., 1992; Johzuka and Ogawa, 1995). Disruption of these genes results in deficiency in NHEJ or illegitimate recombination in mitosis (Milne et al., 1996; Moore and Haber, 1996; Tsukamoto et al., 1996; Boulton and Jackson, 1998), and exhibits deficiency in meiotic homologous recombination and meiotic DSB formation at recombination hot spots (Cao et al., 1990; Ivanov et al., 1992; Johzuka and Ogawa, 1995). In addition, meiotic DSBs are not repaired properly in a subset of separation-of-function mutations in RAD50 (rad50S) (Cao et al., 1990) and MRE11 (mre11S) (Nairz and Klein, 1997; Tsubouchi and Ogawa, 1998), in which unprocessed DSBs are accumulated during meiosis. In such mutants, a type II topoisomerase-like factor Spo11 (Bergerat et al., 1997; Keeney et al., 1997) is left attached covalently to the 5′ termini of the DSBs (Keeney et al., 1997), thereby no resection occurs at ends of the DSBs. Mre11 and Rad50 are highly conserved proteins also found in Schizosaccharomyces pombe (Tavassoli et al., 1995), mouse and human cells (Petrini et al., 1995). Recently, a functional homologue of XRS2 in human cells turned out to be a gene responsible for Nijmegen breakage syndrome which is a variant of ataxia telangiectasia characterized by ionizing radiation sensitivity, immunodeficiency and an increased incidence of cancer (Carney et al., 1998; Matsuura et al., 1998; Varon et al., 1998). In human cells, Mre11, Rad50 and Xrs2 proteins were shown to form a complex as observed in yeast cells (Dolganov et al., 1996; Carney et al., 1998). These data suggest that the DSB-repair system of Mre11–Rad50–Xrs2 is a key element of eukaryotic DSB repair which is highly conserved in all eukaryotes. Mre11–Rad50–Xrs2 proteins play other important roles in maintenance and stability of chromosomes. Yeast Mre11 and Rad50 proteins are required for telomere maintenance (Kironmai and Muniyappa, 1997; Boulton and Jackson, 1998; Nugent et al., 1998), homologous pairing (Loidl et al., 1994; Weiner and Kleckner, 1994), and suppression of mitotic interchromosomal recombination (Farnet et al., 1988; Alani et al., 1990; Ajimura et al., 1993; Johzuka and Ogawa, 1995). In addition, mutations in these genes confer an aberrant level of hypersensitivity to micrococcal nuclease (MNase) at DSB sites in premeiosis and meiosis (Ohta et al., 1998). DSB sites in S.cerevisiae exhibit hypersensitivity to DNase I (Wu and Lichten, 1994; Fan and Petes, 1996; Keeney and Kleckner, 1996) and MNase (Ohta et al., 1994). MNase hypersensitivity at DSB sites in ARG4 and CYS3 loci increases during meiosis prior to DSB formation (Ohta et al., 1994). Analysis of the chromatin configuration at DSB sites in mre11, rad50 and xrs2 mutants revealed that MNase hypersensitivity at DSB sites in mre11 mutant stays at the premeiotic level, while the other two mutants exhibit increases in MNase hypersensitivity comparable to that in wild-type cells (Ohta et al., 1998). In addition, mre11 mutation is epistatic to rad50 with respect to the effects in chromatin configuration at DSB sites, suggesting that Mre11 protein plays a more fundamental role at DSB sites in chromatin. Mre11 protein has a homology to Escherichia coli SbcD nuclease which forms a complex with SbcC, an E.coli counterpart of Rad50 proteins (Leach et al., 1992; Sharples and Leach, 1995; Connelly and Leach, 1996). SbcD and Mre11 share a consensus motif for various phosphoesterases in the N-terminal domain (Sharples and Leach, 1995). On the other hand, SbcD lacks the clusters of charged amino acids in the C-terminal domain of Mre11. Therefore, Mre11 seems to consist of at least two domains: an N-terminal nuclease domain homologous to SbcD, and a C-terminal eukaryote-specific domain with charged amino acids clusters. To examine the role of the Mre11 protein, we have studied the relationship between in vitro biochemical properties and in vivo functions, using yeast Mre11 proteins along with two mutant proteins having a mutation on either N-terminal or C-terminal domains. The present results indicate that Mre11 consists of two separable functional domains: the N-terminal domain which is essential to the nuclease activity of Mre11 in vitro and recombination repair in vivo, and the C-terminal domain which is important for dsDNA-binding activity of Mre11 in vitro and essential for in vivo meiotic functions such as meiotic DSB formation, meiotic recombination and meiotic changes in chromatin configuration at DSB sites. Results DNA-binding activity of the Mre11 protein: a role for the C-terminal domain To study biochemical properties of Mre11 protein, we generated three constructs for expression in E.coli including wild-type and two mutants: an N-terminal phosphoesterase mutation and a C-terminal truncated protein (Figure 1). In the N-terminal mutation, we introduced an amino acid substitution at Asp16 to Ala (Mre11D16A) (Figure 1). This residue is well conserved in all known Mre11 proteins and various phosphoesterases (Figure 1). From the analysis of X-ray diffraction studies, this residue was shown to be in the active center and important for the binding of the Fe3+ ion in a phosphoesterase calcineurin A (Griffith et al., 1995; Kissinger et al., 1995). Site-directed mutagenesis at the corresponding Asp in bacteriophage λ phosphoprotein phosphatase (Figure 1) resulted in a 106-fold reduction in Kcat (Zhuo et al., 1994). The C-terminal truncation mutant (Mre11ΔC49) lacks 49 amino acids in the C-terminus of Mre11. This region consists of a cluster of basic amino acids which is absent in E.coli SbcD protein (Figure 1). Figure 1.Sites of mre11 mutations. (A–C) Amino acid sequences around mre11D16A, mre11S, mre11-58 and mre11-58S mutations in MRE11 and other related phosphoesterases. Residues indicated with dark or light shadings represent identical or similar amino acids, respectively. Arrowheads show the positions of the mutations indicated. Numbers represent positions of amino acids from the first methionine. ScMre11, S.cerevisiae Mre11 (Johzuka and Ogawa, 1995); SpRad32, S.pombe Rad32 (Tavassoli et al., 1995); CeMre11, Caenorhabditis elegans Mre11 (DDBJ/EMBL/GenBank accession No. Z73978); MmMre11, Mus musculus (DDBJ/EMBL/GenBank accession No. U58987); HsMre11, Homo sapiens (Petrini et al., 1995; Paull and Gellert, 1998); SbcD (Leach et al., 1992); calcineurin A (Guerini and Klee, 1989); λPPase (Cohen et al., 1988). (D) Positions of mre11D16A and mre11ΔC49 mutations in a schematic drawing of Mre11 protein. Domain structures of other Mre11 and SbcD proteins are also indicated. HsMre11A is from Petrini et al. (1995) and HsMre11B from Paull and Gellert (1998). Black boxes indicate motifs for metal-phosphoesterases. Hatched and shaded boxes represent basic and acidic amino acid clusters. Download figure Download PowerPoint All types of proteins were expressed in E.coli, and the cell extracts were fractionated through a Co2+ chelating affinity column, since they have a (His)6-tag in the N-termini. We confirmed that tagging in the N-terminus does not prevent in vivo functions of Mre11 (data not shown). TALON–purified fractions were further purified through a heparin–Sepharose or a dsDNA–cellulose column. All proteins bound to the heparin–Sepharose column efficiently. The wild-type Mre11 and Mre11D16A proteins could bind to a dsDNA–cellulose column (Figure 2A and B). They were eluted from dsDNA–cellulose at a moderately high concentration of salt (400–500 mM NaCl). Mre11ΔC49 did not bind to dsDNA–cellulose efficiently, but could be fractionated through heparin–Sepharose and Q–Sepharose columns. The fraction purified by heparin–Sepharose or dsDNA–cellulose was further fractionated through phenyl–Sepharose chromatography. The purity of the Mre11 proteins after the heparin or DNA–cellulose chromatography was nearly homogeneous (Figure 2B). Figure 2.Fractionation of wild-type and mutant Mre11 proteins. (A) The protein elution profile of a DNA affinity chromatography of the wild-type (His)6-Mre11 protein. (His)6-Mre11 proteins were expressed in E.coli, and the cell extracts were applied to a TALON cobalt2+-affinity column (Clontech). The TALON-purified fraction was then fractionated through dsDNA–cellulose column. (B) Fractions indicated by arrowheads in (A) were analyzed by a 7.5% SDS–PAGE. TALON-purified fraction includes proteolytic fragments of Mre11 proteins, but the purity of the fractions after dsDNA–cellulose is nearly homogenous. Molecular weight markers are maltose-binding protein-β-galactosidase, 175 000; maltose-binding protein-paramyosin, 83 000; glutamic dehydrogenase, 62 000; aldolase, 47 500; and triosephosphate isomerase, 32 500. The lanes Ext, FT, and Bd in the TALON chromatography represent E.coli extracts, flow-through and bound fractions, respectively. The lane FT in dsDNA–cellulose chromatography indicate the flow-through fraction. The numbers for each lane indicates the fraction number. (C) Gel purification of Mre11 proteins. Mre11 proteins were fractionated through Superose 12HR as described in Materials and methods. All Mre11 proteins (wild-type, D16A, ΔC49) eluted between catalase and aldolase. Molecular weight standards are catalase (230 000), aldolase (158 000), BSA (67 000), ovalbumin (43 000), chymotrypsin (25 000), RNase A (13 700) and tryptophan (186). The arrowhead indicates the positions of Mre11. Download figure Download PowerPoint We studied Mre11 proteins further using gel-filtration column chromatography to examine their apparent molecular weights. All Mre11 proteins eluted between catalase (mol. wt: 230 000) and aldolase (mol. wt: 158 000) (Figure 2C). Estimated molecular weights of Mre11 proteins are ∼200 000, which is slightly higher than twice the molecular weight for monomeric (His)6-tagged Mre11 proteins that has been deduced from amino acid sequences [wild-type (His)6-Mre11: 82 112, (His)6-Mre11D16A: 82 068, (His)6-Mre11ΔC49: 76 819]. These results suggest that Mre11 protein can exist in a multimeric form as human Mre11 protein (Paull and Gellert, 1998). We could not detect any significant difference in the apparent molecular weights of wild-type Mre11, Mre11D16A and Mre11ΔC49 proteins, indicating that none of the mutations affected overall protein configuration. Using purified Mre11 proteins, we first examined DNA-binding activities using a band-shift assay. A linearized pUC118 DNA substrate was mixed with the purified wild-type Mre11 in the presence of Mg2+. Protein–DNA complexes were analyzed by non-denaturing agarose gel electrophoresis followed by staining with ethidium bromide (Figure 3A). In the presence of Mre11, we detected a mobility shift of a dsDNA band. Binding of Mre11 to dsDNA does not need free DNA ends, since we detected similar mobility shifts using a supercoiled dsDNA substrate (data not shown). We could not find strict sequence specificity in DNA binding of Mre11. Figure 3.DNA-binding activity of the wild-type Mre11 protein. (A) Binding of Mre11 to linearized pUC118 dsDNA. Heparin-purified wild-type Mre11 protein (lanes 1–4; 0, 1, 2, 4 μg, respectively) was added to 50 ng of PstI-digested pUC118 plasmid DNA. Protein–DNA complexes were then analyzed by a non-denaturing agarose gel (0.8%) electrophoresis followed by staining with ethidium bromide. Lane M represents molecular weight markers (HindIII digestion of phage λ DNA, 23.1, 9.4, 6.6, 4.3, 2.3, 2.0, 0.56 kb). (B) Band-shift analysis using 5′ [32P]end-labeled oligonucleotides. Heparin-purified wild-type Mre11 protein (0 μg in lanes 1, 4, 8, 11; 2 μg in lanes 2, 5, 9, 12; 4 μg in lanes 3, 6, 7, 10, 13) was added to 5 nmol phosphate of non-labeled oligonucleotide substrates [lanes 1–3, poly(dT)14 T14; lanes 4–7 and 11–13, ds-oligonucleotide composed of Aarg4 and BantiA; lanes 8–10, ss-oligonucleotide Aarg4, respectively] including 20 fmol of 5′ 32P-labeled molecules. In lane 7, ∼1 μg of affinity-purified anti-Mre11ΔC49 protein antibody was added to the reaction mixture for lane 6. In this case, a supershift of DNA–protein complex was observed (small arrow), indicating that the band-shift activity is due to Mre11 protein. Protein–DNA complex was analyzed in 4% non-denaturing PAGE. A large arrow shows the position of the band shift in the presence of Mre11 protein. (C) Band-shift analysis using various types of non-labeled ds-oligonucleotides. Six micrograms of purified Mre11 protein was incubated with 5 nmol phosphate of non-labeled oligonucleotide substrates as shown in (F), and analyzed as described in (B). Lane M indicates 100 bp DNA ladders. An arrow indicates the position of the band shift. (D–E) Quantitative analysis of the band-shift assay [(D) for the panel (B), and (E) for the panel (C)]. In D, lanes 2 (2 μg of Mre11 with ss-oligo), 3 (4 μg of Mre11 with ss-oligo), 5 (2 μg of Mre11 with ds-oligo) and 6 (4 μg of Mre11 with ds-oligo) were analyzed. Band intensities of the shifted bands were measured and % ratio relative to the intensities at the unbound origin were indicated. A dashed line in (E) shows the level of the blunt end ds-oligonucleotide substrate. (F) Schematic drawing of oligonucleotides used for the band-shift assay. P105-dsDNA has cohesive complementary 4 nt overhangs at both ends. Download figure Download PowerPoint To study further DNA-binding specificity of Mre11, various types of ss- or ds- 32P-labeled or non-labeled oligonucleotide substrates (Figure 3F) were mixed with Mre11 in the presence of Mg2+, and the resulting complex was analyzed by non-denaturing polyacrylamide gel electrophoresis followed by analysis with a radioactivity image analyzer or with a fluorescence image analyzer after staining with ethidium bromide. It was shown that binding to ds-oligonucleotides (30mer) is more efficient than that to ss-oligonucleotides (30mer) (Figure 3B and D). However, Mre11 bound to oligonucleotides less efficiently compared with longer DNA substrates such as plasmid DNA. In fact, the efficiency of the binding was dependent upon the length of the duplex in the oligonucleotides rather than the structure of DNA substrates (Figure 3C and E). Binding of Mre11 to DNA substrates with 3′-overhangs seemed to be slightly stronger than that to substrates with the corresponding length of 5′-overhangs (∼2.2-fold in the case of 7 nt overhangs). We also detected a 1.8- to 2.5-fold enhancement in the binding of Mre11 to ds-oligonucleotide substrates (28mer) with cohesive 4 nt overhangs, compared with a substrate with non-cohesive overhangs of a similar length (Figure 3C and E), suggesting that annealing and catemer formation at the complementary ssDNA overhangs facilitates the binding of Mre11 to DNA. We found that Mre11 protein has a transitional metal-dependent dsDNA-binding activity. In a condition where limited amount of Mre11 exists, Mre11 weakly bound to dsDNA regardless of the presence or absence of Mg2+ (Figure 4B). On the other hand, in such a condition, transition metal divalent cations (e.g. Mn2+, Zn2+, Cu2+, Ni2+) enhanced the binding of Mre11 to dsDNA substrates (Figure 4B). This may not be simply due to the (His)6-tag in the N-terminus of the Mre11 protein, since other DNA-binding proteins [HisMts1, a CREB-type transcription factor in S.pombe (Kon et al., 1997)] with an N-terminal (His)6 did not show such a transitional metal-dependent enhancement (Figure 4C). Figure 4.DNA-binding activity of mutant Mre11 proteins. (A) Band-shift activity of mutant Mre11 proteins using a linearized plasmid dsDNA as in Figure 3A. The amount of Mre11 protein used is 0 (lanes 1), 0.5 (lanes 2), 1 (lanes 3), 2 (lanes 4) and 3 μg (lanes 5). Reaction and detection of bands were as described in Figure 3A. Lane M represents molecular weight markers (EcoT14I digestion of phage λ DNA, 19.3, 7.7, 6.2, 4.3, 3.5, 2.7, 1.9, 1.5 kb). Note that Mre11ΔC49 protein has less efficient band-shift activity. (B) Dependence of the band-shift activity of the wild-type Mre11 protein on metal divalent cations. In this experiment, we used 0.6 μg of Mre11 protein. Lanes 1–7, none, 5 mM of MgCl2, MnCl2, ZnCl2, CuSO4, NiSO4, SrCl2, and CaCl2, respectively. (C) Band-shift activity of mutant Mre11 proteins using a linearized plasmid dsDNA in the absence (lanes −) or presence of divalent cations: Mg2+ (lanes Mg) or Mn2+ (lanes Mn). Lane (−prot.) represents a control without addition of Mre11 proteins. HisMts1 is a (His)6-tagged Atf1/Gad7 transcription factor (Kon et al., 1997). In all cases, we used 0.6 μg of proteins. Arrowheads indicate the positions of a supershift. Arrows show the original position of the linearized pUC118 dsDNA. Download figure Download PowerPoint DNA-binding activities of mutant Mre11 proteins were examined by band-shift analysis using plasmid dsDNA (Figure 4C). Mre11D16A exhibited almost the same level of DNA band-shift activity in the wild-type Mre11. On the other hand, Mre11ΔC49 protein showed only a weak band shift even in the presence of transition divalent cations. This is consistent with the results of DNA–cellulose affinity chromatography (see above). These results can be rationalized by a loss of cooperativity in Mre11ΔC49 protein. However, the reduction in the DNA-binding activity of Mre11ΔC49 protein was also observed when we used oligonucleotide substrates (30mer) in which less cooperative binding could be expected (data not shown). From these results, we concluded that C-terminal domain is important for the DNA-binding activity of Mre11 protein. N-terminal phosphoesterase domain is involved in nuclease activity of Mre11 Extensive homology in the N-terminal domain of Mre11 to E.coli SbcD nuclease suggests that Mre11 has a nuclease activity. In fact, human Mre11 protein exhibits exo- and endo-nuclease activities (Paull and Gellert, 1998; Trujillo et al., 1998). To examine nuclease activity of yeast Mre11, E.coli-expressed (His)6-Mre11 was incubated with a circular M13 viral ssDNA substrate in the presence of Mn2+ (Figure 5A). Mre11 cleaved circular M13 ssDNA in the presence of Mn2+ (0.5–10 mM), but we could not detect significant nuclease activity in the absence of Mn2+ or in the presence of other divalent cations (Mg2+, Ni2+, Ca2+, Sr2+, Co2+ and Zn2+). The nuclease activity did not require ATP. A linearized plasmid DNA can be a nuclease substrate of Mre11 in the presence of Mn2+ (Figure 5B). Supercoiled or relaxed closed circular dsDNA was not digested by Mre11 (data not shown), indicating that digestion of dsDNA requires the ends of DNA. We detected an exonuclease activity on ds-oligonucleotides with blunt and 5′-overhangs at both ends, but not on ds-oligonucleotides with 3′-overhangs at both ends (Figure 5C). When we incubated Mre11 protein with 5′ 32P-labeled linearized plasmid DNA with 5′-overhangs, smear bands were detected after the incubation (Figure 5D). These suggests that Mre11 has 3′ to 5′ double-strand-specific exonuclease activity as observed in human Mre11 protein (Paull and Gellert, 1998; Trujillo et al., 1998). Figure 5.Nuclease activities of Mre11 proteins. (A) Nuclease activity on M13 circular ssDNA. DNA–cellulose-purified wild-type Mre11 protein (lanes 1–4: 0, 1, 2, 4 μg, respectively) was incubated with 200 ng of M13mp18 circular ssDNA at 37°C for 1 h as described in Materials and methods. Digested DNA was analyzed in a 0.8% agarose gel. Lane M indicates 1 kb DNA ladder (Gibco-BRL). Lanes WT, D16A, and ΔC49 show the data using 2 μg of DNA–cellulose-purified wild-type and heparin-purified D16A Mre11 proteins and Q–Sepharose-purified ΔC49 protein, respectively. Note that there is no detectable nuclease activity in Mre11D16A protein. (B) Nuclease activity on a linearized plasmid dsDNA. Lanes WT, D16A, and ΔC49 show the data using 1 μg of heparin-purified wild-type and D16A Mre11 proteins and Q–Sepharose-purified ΔC49 protein, respectively. Proteins were incubated with 50 ng of EcoRI-digested pUC118 DNA as described in (A). Digested DNA was analyzed on a 0.8% agarose gel. Note that there was no detectable nuclease activity in Mre11D16A protein. Lane M shows molecular size markers of EcoT14I-digested phage λ DNA (see Figure 4A). C) Nuclease activity on ds-oligonucleotides. Oligonucleotides Aarg4 and BantiA, K5over7 and BantiA, Aarg4 and L3over7 (see Materials and methods) were annealed to form ds-oligonucleotides with blunt ends (lanes blunt), 7 nt 5′ overhangs (lanes 5′-7nt), and 7 nt 3′ overhangs (lanes 3′-7nt) at both ends, respectively. As described in (A), oligonucleotide substrates including ss-oligonucleotide Aarg4 (2.35 nmol phosphate, 160 pmol DNA ends) were then incubated with 8 μg of wild-type Mre11 protein at 30°C for 1 h. Digested DNA was separated in a 20% polyacrylamide gel and stained with ethidium bromide. Cleavage was only detected in ds-oligonucleotides with blunt ends or 5′ overhangs at both ends. Lane M indicates 10 bp ladders. Numbers indicate base pair numbers of the bands in 10 bp ladders. (D) Nuclease activity on EcoRI-linearized pUC119 DNA 32P-labeled at the 5′ ends. As described in (A), 5′ 32P-labeled pUC119 plasmid DNA (5 ng) was incubated with 3 μg of wild-type Mre11 protein at 37°C for 0 (lane 0′), 10 (10′), 30 (30′), and 60 min (60′) as described in (A). Bands were detected by autoradiography. Note that smear bands were detected after the incubation. Download figure Download PowerPoint We next examined nuclease activity of mutant proteins (Figure 5B). Mre11ΔC49 protein has a nuclease activity comparable to the wild-type Mre11, while it shows reduced DNA binding. On the other hand, we could detect no significant nuclease activity in Mre11D16A protein fraction. These observations show that the C-terminal DNA-binding domain is not responsible for the nuclease activity and the N-terminal phosphoesterase domain is required for the nuclease activity. This indicates that the nuclease activity is an intrinsic property of Mre11 protein and is not due to trace amounts of contaminating proteins. Phenotypes of mre11D16A and mre11ΔC49 mutations during mitosis To study the relationship between the in vitro properties of Mre11 and in vivo functions, we introduced the corresponding mutations (mre11D16A and mre11ΔC49, see Figure 6A) into the yeast genome. We first examined mitotic effects. The mutant mre11D16A exhibited strong MMS sensitivity, but it was ∼1/10 less sensitive than the case of a null mutant (Figure 6B and C). On the other hand, we could not detect any significant difference in MMS sensitivity of mre11ΔC49 and the wild-type. In both mre11D16A and mre11ΔC49 strains, we did not detect hyper-recombination (Table I) and slow growth phenotypes that are observed in mre11 null mutants. In mre11ΔC49 strain, the interchromosomal recombination rate was reproducibly smaller than the wild-type level (Table I). Figure 6.MMS sensitivity of mre11D16A and mre11ΔC49 mutants. (A) Constructions for gene replacement by homologous recombination. All constructs are flanked with URA3 gene (arrows indicate the direction of transcription) for positive selection of recombinants. Black boxes indicate motifs for metal-phosphoesterases. Hatched and shaded boxes represent basic
Referência(s)