|Home | About | Journals | Submit | Contact Us | Français|
Ino80 is an ATP-dependent nucleosome-remodeling enzyme involved in transcription, replication, and the DNA damage response. Here, we characterize the fission yeast Ino80 and find that it is essential for cell viability. We show that the Ino80 complex from fission yeast mediates ATP-dependent nucleosome remodeling in vitro. The purification of the Ino80-associated complex identified a highly conserved complex and the presence of a novel zinc finger protein with similarities to the mammalian transcriptional regulator Yin Yang 1 (YY1) and other members of the GLI-Krüppel family of proteins. Deletion of this Iec1 protein or the Ino80 complex subunit arp8, ies6, or ies2 causes defects in DNA damage repair, the response to replication stress, and nucleotide metabolism. We show that Iec1 is important for the correct expression of genes involved in nucleotide metabolism, including the ribonucleotide reductase subunit cdc22 and phosphate- and adenine-responsive genes. We find that Ino80 is recruited to a large number of promoter regions on phosphate starvation, including those of phosphate- and adenine-responsive genes that depend on Iec1 for correct expression. Iec1 is required for the binding of Ino80 to target genes and subsequent histone loss at the promoter and throughout the body of these genes on phosphate starvation. This suggests that the Iec1-Ino80 complex promotes transcription through nucleosome eviction.
The structure of chromatin is modulated by chromatin-remodeling factors, such as histone-modifying enzymes and ATP-dependent chromatin-remodeling complexes (7, 35, 45, 54). The latter are usually multisubunit protein complexes containing an ATPase from the SWI2/SNF2 superfamily, which uses the energy derived from ATP hydrolysis to disrupt histone-DNA interactions (6, 13, 26).
Ino80 is a member of this SWI2/SNF superfamily and is the catalytic subunit of the Ino80 chromatin-remodeling complex (65). Initially characterized in budding yeast (Saccharomyces cerevisiae), the Ino80 complex plays a central role in DNA-mediated processes, such as DNA double-strand break repair, homologous recombination, and the regulation of the DNA damage cell cycle checkpoint response, chromosomal-DNA replication, and transcription (5, 10, 18, 23, 24, 38, 39, 52, 56, 57, 65, 69, 74, 76, 80, 82; reviewed in reference 4). Ino80 complexes purified from budding yeast and mammalian cells contain core subunits, which are conserved across species, as well as species-specific proteins (36, 65). In budding yeast, the core Ino80 complex is composed of INO80, ARP5, ARP8, ARP4, RVB1, RVB2, IES2, and IES6 (63, 65). The actin-like proteins ARP5 and ARP8 are unique to the Ino80 complex and are required for an active complex in budding yeast (38, 63). ARP4 plays a role in the response to DNA damage in budding yeast and is also a member of the chromatin-remodeling complex SWR1 and the histone acetyltransferase complex NuA4 (16, 42, 52, 76). The Rvb1 and Rvb2 proteins (also called Pontin/Tip49 and Reptin/Tip48, respectively, in mammalian cells) are AAA+ ATPases related to the Holliday junction-resolving RUVB proteins of bacteria (65). They are integral members of the Ino80- and SWR1-remodeling complexes (36, 42, 43, 50, 65), as well as the histone acetylase Tip60 complex (34) and the transcription activator c-Myc complex (reviewed in references 25 and 81). Ies2 and Ies6 are conserved from yeast to humans and have direct roles in transcriptional regulation in mammalian cells (10, 63).
A YY1-containing Ino80 complex was recently identified in mammalian and Drosophila cells (10, 41, 82). Yin yang-1 (YY1) is a zinc finger-containing Polycomb group (PcG) transcription factor that regulates genes essential for growth and development (1, 9, 15, 68). The mammalian YY1-Ino80 complex has been found to play roles in transcription and homologous-recombination-based repair (10, 82).
The precise mechanisms by which Ino80 regulates cellular processes are unclear. Unlike budding yeast, which has very little higher-order chromatin, the fission yeast (Schizosaccharomyces pombe) offers a more complex chromatin structure that is more similar to that of higher eukaryotes in many respects (for a review, see reference 28). Therefore, the characterization of the Ino80 complex in this model organism may give new insights into the functions and mechanisms of this versatile complex.
Here, we characterize the fission yeast Ino80 complex. The purified complex mediated ATP-dependent nucleosome remodeling in vitro. We found that the complex is highly conserved through evolution and contains a novel factor, Iec1 (67). Iec1 bears sequence similarity to YY1 and other members of the GLI-Krüppel family of zinc finger proteins. Iec1 and the Ino80 complex subunits arp8, ies6, and ies2 are important for efficient DNA damage repair and the response to replication stress. The double deletion of iec1 and ies2 suppresses defects of the single deletions, suggesting that the different Ino80-interacting proteins may program the complex in different, possibly opposing ways. We show that Iec1 is important for the regulation of several genes involved in nucleotide and phosphate metabolism. Iec1 is required for the binding of Ino80 to the promoter and downstream regions of these genes, leading to reduced nucleosome density over the locus. We suggest that the Iec1-Ino80 complex regulates the transcription of genes involved in nucleotide metabolism by mediating nucleosome eviction.
The S. pombe strains used in this study are listed in Table Table1.1. Cells were grown in rich yeast extract with supplements (YES) or Edinburgh minimal medium (EMM) at 30°C unless otherwise indicated (51). Histidine, leucine, uracil, and lysine were added at 250 mg/liter and adenine at 250 mg/liter, or 2 mg/liter for low-adenine media. Low-phosphate EMM was prepared by replacing potassium hydrogen phthalate and Na2HPO4 with NaH2PO4 (10 mg/liter) and NaOAc (1 g/liter). Inorganic phosphate was removed from YES by precipitation as magnesium phosphate under alkaline conditions (62). Hydroxyurea (HU) (Sigma) was added to the medium at 7.5 mM and bleocin (Calbiochem) at 3 μg/ml. For UV irradiation, the plates were placed in a Stratalinker UV cross-linker and exposed to 100 J/m2.
The SA001 strain expressing Ino80 with C-terminal FLAG tags from the endogenous locus was generated using strain FY2317. A plasmid was constructed that drove expression of FLAG- and V5-tagged S. pombe Ino80 from the NMT81 promoter, using the pNMT81-TOPO system (Invitrogen). In this plasmid the ino80 cDNA was followed by a kanamycin resistance cassette, followed by a 174-bp fragment of sequence just 3′ to the ino80 open reading frame (ORF). The replacement cassette was amplified from this construct with Pfu Ultra (Stratagene) with primers 5′-ACTTTCAACAAGGGCAGGTGGTTTGGG-3′ and 5′-GTCGACTGTTTACAACATTTCATCTAA-3′. The FLAG-tagged protein has a modified flag tag at the C terminus (only one D), followed by a 6-amino-acid spacer, then by the V5 tag, then by a 5-amino-acid spacer, and by a double FLAG tag.
The entire iec1 ORF was deleted by homologous recombination (29). The URA4 gene was ligated into the HindIII site of pBluescript, giving rise to pURA. PCR products representing the 5′ upstream region of the iec1 ORF were prepared with the upstream primers (Table (Table2)2) and digested with EcoRI and EagI. The 3′ downstream region of the iec1 ORF was prepared with primers listed in Table Table22 and digested with XhoI and KpnI. Both PCR fragments were inserted on either side of the URA4 gene in pURA. This URA4 deletion cassette was amplified by PCR and transformed into cells by the lithium acetate procedure (19). Colonies were selected for growth in medium lacking uracil. Positives were verified by PCR using primer pairs outside the cassette.
A PCR fragment representing the iec1 ORF and incorporating a C-terminal HA tag was prepared with the primers 5′-GAGACGGCCGCATGTCTATATCCTCGGATACCTC-3′ and 5′-GAGAGAATTCCGCATAATCCGGCACATCATACGGATAACGACTGATGCTGGGATCCACTT-3′, digested with EagI and EcoRI, and, along with the 3′ downstream region of iec1 digested with XhoI and KpnI, inserted on either side of the URA4 gene in pURA as described above. The cassette was then amplified by PCR and transformed into yeast. Positive clones were selected for growth on medium lacking uracil and were verified for correct integration by PCR and by Western blot analysis of whole-cell extracts using anti-HA antibody (Roche).
PCR products representing the iec1 ORF prepared with the primers 5′-GAGAGAGTCGACGGATGTCTATATCCTCGGATACCTCA-3′ and 5′-TCTCTCTTAATTAAACGACTGATGCTGGGATCCACTTG-3′ digested with SalI and PacI and the 3′downstream region of the iec1 ORF prepared with the primers 5′-GAGAGAGAGCTCGATGCATCTTCAAATGCCTTACG-3′ and 5′-TCTCTCATCGATGATACAACCACACTAAGGCAG-3′ digested with SacI and ClaI were inserted on either side of the 13MYC-natMX6 cassette in the pFA6a-13Myc-natMX6 plasmid (78). The resulting cassette was amplified by PCR and transformed into yeast using the lithium acetate procedure (19). Positive clones were selected for growth on medium containing 100 μg/ml nourseothricin (ClonNAT; Werner Bioagents) and were validated by PCR and by Western blot analysis of whole-cell extracts using anti-Myc antibody (Cell Signaling).
The kanMX6 cassette was cleaved out of the pFA6a-kanMX6 plasmid by digestion with EcoRI and BamHI and ligated into EcoRI/BamHI-cleaved pBluescript, giving rise to pKAN (3, 32, 78). The 5′ upstream region and 3′ downstream regions of the nhp10, arp8, ies6, and ies2 ORFs were amplified with the primers listed in Table Table22 and digested with EagI and BamHI for the 5′ fragment and EcoRI and HindIII for the 3′ fragment. The 5′ and 3′ fragments were inserted on either side of the kanx cassette in pKAN. This kanx deletion cassette was amplified by PCR, and nhp10, arp8, ies6, and ies2 were deleted by homologous recombination (29). Positive clones were selected for growth in YES medium supplemented with 150 mg/liter G418 (Invitrogen) and verified by PCR using one primer upstream of the cassette and one within the kanx cassette. ino80 was deleted by replacing the entire ORF with the kanx cassette in a diploid strain that was generated by mating JZ1 with JZ5.
The protocol for the purification of the FLAG-tagged Ino80 complex and the MYC- or HA-tagged Iec1 complexes from S. pombe was adapted from the method of Tsukiyama et al. (73). Briefly, cultures of the SA001, FY2317 (as a control), CH006, and CH004 strains were grown to saturation in EMM. The cells were pelleted and washed once with ice-cold water, once with H buffer (25 mM HEPES, pH 7.6, 0.5 mM EGTA, 0.1 mM EDTA, 2 mM MgCl2, 20% glycerol, 0.02% NP-40, 1 mM dithiothreitol [DTT])-300 mM KCl and once with H buffer-300 mM KCl supplemented with Complete EDTA-free protease inhibitor cocktail (Roche). The cells were then frozen in liquid nitrogen and broken in dry ice using a coffee mill. The cells were resuspended in H buffer-500 mM KCl with Complete EDTA-free protease inhibitor (Roche) and spun at 37 K in a SW40 Beckman rotor. The supernatant (~20 ml) was recovered and incubated on a rotating wheel with 100 μl FLAG-M2 agarose beads (Sigma) or 100 μl anti-HA (Roche) monoclonal antibodies coupled to protein A for 3 h at 4°C. The beads were washed 5 times with H buffer-500 mM KCl and twice with H buffer-150 mM KCl. The Ino80 complex was then eluted from the beads by incubating them 4 times with 100 μl 2.5-mg/ml FLAG-peptide (Sigma) in 12.5 mM HEPES, pH 7.6, 7.5 mM Tris, pH 7.4, 75 mM KCl, 75 mM NaCl, 0.5 mM EGTA, 0.1 mM EDTA, 2 mM MgCl2, 20% glycerol, 0.02% NP-40, 1 mM DTT, and Complete EDTA-free protease inhibitor (Roche). For the immunoprecipitation of the HA-tagged complexes, strains CH006 and SA001 (as a control) were used. For mass spectrometric analysis, beads were washed with 25 mM ammonium bicarbonate and proteins were eluted with 6 M guanidine hydrochloride. The amount of protein was determined by SDS-PAGE following Coomassie blue staining by comparing the band intensity with those from a bovine serum albumin standard.
For the immunoprecipitation of the MYC-tagged Iec1 complex, strains CH016 and CH006 were used. The complexes were captured using rabbit anti-MYC antibody coupled to agarose (Sigma) and eluted with SDS loading buffer (Bio-Rad).
Recombinant Xenopus histone octamers were prepared and assembled on a 12× tandem repeat of Lytechinus variegatus 5S rDNA as described previously (47, 79). The accessibility assay was performed as described in reference 79.
Five microliters of eluted fractions from the FLAG-tagged Ino80 and HA-tagged Iec1 purifications, and from the corresponding mock purifications, was reduced, carbamidomethylated, and then purified using C4 ZipTips (Millipore). The purified proteins were dried by vacuum centrifugation, redissolved in 5 μl of 25 mM ammonium bicarbonate in 30% acetonitrile containing 10 ng/μl modified trypsin (Promega), and incubated overnight at 30°C. The resulting peptide mixtures were dried by vacuum centrifugation and redissolved in 5 μl of 0.1% formic acid, and 1 μl was analyzed by liquid chromatography-tandem mass spectrometry (LC-MS/MS). Bands cut from the SDS-PAGE gel were destained by washing them with 25 mM ammonium bicarbonate/50% acetonitrile (5 times) and then reduced, carbamidomethylated, and digested as described above. The tryptic peptides generated from in-solution or in-gel digestions were separated on a reversed-phase column (Vydac C18; 0.05 by 50 mm; 5-μm particle size), with an acetonitrile gradient (0 to 30% over 60 min) containing 0.1% formic acid, at a flow rate of 120 nl/min. The column was coupled to a nanospray ion source (Protana Engineering) fitted to a quadrupole-time of flight (TOF) mass spectrometer (Qstar Pulsar i; Applied Biosystems/MDS Sciex). The instrument was operated in information-dependent acquisition mode, with an acquisition cycle consisting of a 0.5-s TOF scan over the m/z range of 350 to 1,500, followed by 3 2-s MS/MS scans (triggered by 2+ or 3+ ions) recorded over the m/z range of 100 to 1,700 with a 60-s dynamic exclusion of former target ions. Mass spectrometric data were searched against the eukaryote entries in Uniprot 14.2 using Mascot software (Matrix Science).
RNA was prepared with the FastRNA Pro Red Kit (MPBio) or extracted using the hot-acid-phenol method (83). Nine μg of RNA was separated on 1.2% formaldehyde agarose (FA) gels, transferred to a Hybond N+ membrane (Amersham), UV cross-linked with a Stratalinker (Stratagene), and hybridized to 32P-labeled pho4, pho1, apt1, cdc22, and actin probes by the dextran sulfate method. The signal generated by the mRNA was quantified using a phosphorimager (Fuji) linked to Advanced Image Data Analysis (AIDA) software (Fujifilm) and normalized either relative to actin or relative to the ethidium bromide-stained 25S rRNA band with the AIDA software.
Hot-phenol-extracted RNA was cleaned using a Qiagen RNeasy minikit (Qiagen). To remove contaminating genomic DNA, 5 μg of isolated RNA was treated with 5 mM DTT, 2 U/μl RNasin (Promega), 1 mM MgCl2, and 1 U/μg RNA DNase I (Promega). Digestion was carried out at 37°C for 30 min, and the reaction was terminated by boiling the mixture at 95°C for 5 min. One μg of DNase I-treated RNA, 100 ng random primers (Promega), and 500 μM deoxynucleotide triphosphate (dNTP) mix (Bioline) in a total volume of 20 μl were incubated at 65°C for 5 min and then cooled on ice. After a brief spin, 1× SS II buffer (Invitrogen), 10 mM DTT, and 2 U/μl RNasin (Promega) were added. The reaction mixture was incubated for 2 min at 25°C, and 200 units of SS II reverse transcriptase (Invitrogen) was added. The reaction mixture was further incubated at 25°C for 10 min and then at 42°C for 1 h. The reaction was terminated by incubation at 70°C for 15 min. The presence of specific transcripts was quantified by quantitative RT-PCR using a Bio-Rad CFX96 with SYBR green PCR Mastermix from Applied Biosystems in duplicate using the 2−ΔΔCT method (46). The primer sequences used are indicated by asterisks in Table Table33.
The chromatin immunoprecipitation (ChIP) protocol was adapted from Kurdistani et al. and Robyr and Grunstein (44, 59). At least two independent experiments were performed per strain. The cells were grown to a density of 2 × 108 cells/ml. The cells were cross-linked by adding formaldehyde to a final concentration of 1% at room temperature. The cross-linking reaction was quenched after 1 h by the addition of glycine to 0.125 M. The cells were harvested, washed in ice-cold PBS, resuspended in cell lysis buffer (0.1% SDS, 50 mM HEPES-KOH, pH 7.5, 1% Triton X, 0.1% sodium deoxycholate, 1 mM EDTA, 150 mM NaCl) with protease inhibitor cocktail (Roche), and lysed with a Fastprep machine (MPBio). The fixed chromatin was fragmented by sonication to an average size of 500 to 1,000 bp. The chromatin was immunoprecipitated with 2 μg of anti-FLAG antibody (Sigma), 5 μl anti-MYC antibody (Cell Signaling), or 5 μg of anti-histone H3 antibody (Abcam). Protein A-Sepharose beads (GE Healthcare) were washed in wash buffer (0.1% SDS, 50 mM HEPES-KOH, pH 7.5, 1% Triton X, 0.1% sodium deoxycholate, 1 mM EDTA, 500 mM NaCl). Cross-links were reversed overnight at 65°C, and protein was removed by digestion with proteinase K and phenol extraction. The purified DNA was resuspended in 40 μl Tris-EDTA, pH 8.0. The products were quantified using SYBR green I incorporation and measured using an ABI Prism 7700 Sequence Detection System (Applied Biosystems) with the primer pairs listed in Table Table3.3. The results were corrected for background (no-antibody control) and expressed as percent chromatin immunoprecipitated from the input.
Immunoprecipitated chromatin from two independent ChIP experiments was hybridized to GeneChip S. pombe Tiling 1.0FR arrays (Affymetrix). ChIP DNA was amplified to 5 μg as described previously (17, 37). Five mM dUTP was added to the second amplification (round B), and ChIP DNA was recovered with QiaQuick PCR columns. Fragmentation, labeling, and hybridization to the Affymetrix GeneChip S. pombe Tiling 1.0FR was performed by the Affymetrix core facility at Novum, Sweden, according to standard protocols. Duplicate raw data from Affymetrix (.CEL format) were analyzed with Tiling Analysis Software (TAS) v1.1 using quantile normalization plus scaling and run with a bandwidth of 100. The normalized data from each probe were assigned to S. pombe genome coordinates in TAS (S. pombe September 2004 genome; Sanger). The data were visualized with the Affymetrix Integrated Genome Browser. The resulting linear ratio was extracted for each probe position, defined as the center (12th) base coordinate for each 25-nucleotide probe. The 5′ and 3′ intergenic regions (IGR) were a maximum of 450 bp from the start and the stop codon, respectively, of the gene (depending on the intergenic distance between genes) and were divided into a maximum of 10 fragments of 50 bases each. Missing values were replaced by linear interpolation, using values spanning the missing data point. Values for probes of 5′ IGR positions were averaged to give one value for each promoter region. Gene ontology analysis was performed using AMIGO (http://amigo.geneontology.org/cgi-bin/amigo/go.cgi).
Proteins were separated on 4 to 12% NuPage Novex Bis-Tris gels (Invitrogen) and transferred to Hybond ECL nitrocellulose membranes (GE Healthcare). The blots were then incubated in anti-HA antibody (Roche; dilution, 1:1,000), M2 anti-FLAG antibody (dilution, 1:1,000; Sigma), or anti-MYC (dilution, 1:1,000; Cell Signaling). For the blots in Fig. Fig.1,1, anti-FLAG coupled to horseradish peroxidase (HRP) (Sigma) was used. The immobilized antigens were detected using enhanced chemiluminescence (ECL) techniques (Millipore).
Cells were grown in YES or EMM overnight at 30°C. The cells were then diluted in H2O to a concentration of 6 × 107 cells/ml. Five 10-fold serial dilutions were prepared with H2O, and 2 μl from each dilution was spotted onto agar plates. Photographs were taken 3 to 5 days after spotting.
A control strain with empty vector (CH008) and an ino80 null strain complemented with a plasmid driving expression of Ino80 in trans (SA003) were inoculated in 100 ml nonselective rich medium (YES) and grown overnight. One ml of this culture was used to inoculate a fresh 100-ml culture, and the process was repeated two more times. Approximately 1,000 cells were plated onto nonselective (NS) rich medium and selective medium (minus Leu) and incubated at 30°C.
Cells were grown in EMM or low-phosphate or low-adenine EMM for 4 days. The cells were DAPI (4′,6′-diamidino-2-phenylindole) stained and imaged using an Olympus BX40 microscope. Two hundred cells each of the wild-type (WT), Δiec1, and Δiec1 plus pIec1 strains grown in either EMM, low-phosphate EMM (EMM/low Pi), or EMM/low adenine were scored as either ascus or vegetative, and the percentage of asci was calculated.
Yeast cells were harvested in mid-log phase, with about 108 cells used for each extraction. Duplicate extractions were done for each experiment, and the data shown are based on three different experiments. The cells were centrifuged, and the pellet was resuspended in 3 ml of ice-cold 60% aqueous methanol. Each suspension was held at −20°C for 1 h, followed by centrifugation. The supernatants were removed, and each pellet was washed with 1 ml 60% methanol. The washes were added to the original supernatants, and then each solution was held in a boiling water bath for 3 min. This was followed by centrifugation for 15 min at 17,000 × g. Each supernatant was transferred to a fresh tube and taken to dryness under vacuum. Each residue was dissolved in water and stored at −20°C for later analysis. Reaction mixtures (50 μl) contained 100 mM HEPES buffer, pH 7.5, 10 mM MgCl2, 0.1 unit of Escherichia coli DNA polymerase I Klenow fragment (United States Biochemical), 0.25 μM oligonucleotide template, and 0.67 μM [3H]dATP (30 Ci/mmol; 1.0 μCi per assay; Amersham Biosciences) or [3H]dTTP (PerkinElmer Life Sciences). Incubation was carried out for 60 min at 37°C. Deoxyribonucleotide pools were expressed as mole amount per mg yeast protein.
Tiling data are available on MIAMexpress (http://www.ebi.ac.uk/microarray/) under accession number E-MEXP-2284.
We identified the fission yeast ORF SPAC29B12.01 as the Ino80 homologue by sequence homology analysis. The endogenous Ino80p was FLAG tagged, and the associated complex was affinity purified. Mass spectrometric analysis revealed highly conserved core complex a similar to the budding yeast and mammalian Ino80 complexes (Fig. (Fig.1A1A and Table Table4).4). Ino80 copurifies with seven proteins that have homologues in both budding yeast and humans (36, 42, 43, 50, 65): SPAC664.02 (homologue of budding yeast ARP8), SPBC365.10 (ARP5), SPAPB8E5.09 (RVB1), SPBC83.08 (RVB2), Alp5 (ARP4), SPAC222.04 (IES6), and SPAC6B12.05 (IES2).
In addition to this “core complex,” fission yeast Ino80, like its budding yeast counterpart but unlike the human Ino80 complex, contains (i) an HMG box-like protein (SPAC10F6.08) with similarities to the NHP10 protein of budding yeast, which has been found to be involved in DNA damage repair (52); (ii) the TATA binding protein-associated factor Taf14/Tfg3; and (iii) actin (36, 65). We also identified several uncharacterized proteins that are unique to the fission yeast Ino80 complex: SPCC1259.04, SPCC16C4.20, SPAC23G3.04, and SPAC144.02 (Table (Table4).4). A comparative proteomic analysis of yeast chromatin modification and remodeling complexes identified most fission yeast Ino80 complex subunits presented here without further functional analysis (67). However, there are also notable differences between our study and that of Shevchenko et al. We assigned Taf14/Tfg3 and the gene product of SPCC16C4.20 as potential Ino80 subunits. Taf14/Tfg3 (the gene product of SPAC22H12.02) was clearly identified in our Ino80-FLAG pulldown (Fig. (Fig.1A)1A) and is the predicted orthologue of S. cerevisiae TAF14. This protein is also called SWP29, TAF30, TFG3, and ANC1 and is a subunit of the S. cerevisiae Ino80 complex subunit (63). It has also been shown to be a subunit of the general transcription factors TFIIF and TFIID and the SWI/SNF complex (11, 30, 31, 40). Unlike Shevchenko et al., we did not identify Iec5p and Pht1p (the fission yeast orthologue of histone variant H2AZ). These differences may be related to the distinct purification protocols (affinity purification using the FLAG tag versus TAP) and growth conditions of the two studies.
We tested whether the purified Ino80 complex exhibits ATP-dependent nucleosome-remodeling activity in vitro. Nucleosomal arrays were assembled with recombinant purified histones. In the presence of the purified Ino80 complex and ATP, there was increased cleavage of the chromatin by the restriction enzyme DraI compared to controls, indicating that Ino80 remodels nucleosomes to render chromatin more accessible (Fig. (Fig.1B1B).
The fission yeast Ino80 complex contains a novel zinc finger protein, Iec1, encoded by the previously uncharacterized open reading frame SPAC144.02 (Fig. (Fig.1A1A and Table Table4)4) (67). Sequence homology analysis indicated that this 249-amino-acid protein contains two C2H2-type zinc finger motifs and a possible zinc finger at its N terminus (Fig. 1C and D) (22). Our analysis revealed no apparent S. cerevisiae orthologue for the protein, but similarity to the zinc finger motifs of the GLI-Krüppel family of transcription factors was found (Fig. (Fig.1C).1C). The Drosophila PcG PHO and its mammalian orthologue YY1 have been found associated with Ino80 (10, 41, 82). These proteins belong to the GLI-Krüppel family and show similarity to SPAC144.02 at the level of their C2H2 zinc finger motifs (Fig. (Fig.1C).1C). Furthermore, a short stretch of 7 amino acids, HTGEKP(F), a motif present in all YY1 proteins across species and GLI-Krüppel family members, was found in SPAC144.02 (Fig. (Fig.1D)1D) (in GLI3, we found HTGEKPH).
The presence of Iec1 in the Ino80 complex was confirmed by coimmunoprecipitation experiments with whole-cell extracts from endogenously MYC-tagged or HA-tagged Iec1- and FLAG-tagged Ino80-expressing strains, followed by Western blot analysis (Fig. 1E and F). HA-tagged Iec1 was used to coimmunoprecipitate associated proteins from whole-cell extracts, followed by mass spectrometric analysis. Iec1 copurified with Ino80, the AAA+ ATPase Rvb2, and actin in what appeared to be an Iec1-Ino80 core complex (Table (Table4).4). This complex appears similar to the YY1/PHO-Ino80 complexes in mammalian cells and Drosophila, which contain YY1/PHO, Ino80, Rvb1/2, actin (only in Drosophila), and Arp5/Arp8 (Table (Table4)4) (10, 41, 82). We concluded that the fission yeast Ino80 interacts with Iec1, a novel C2H2 zinc finger protein related to GLI-Krüppel family members, such as YY1, and that the protein forms a core complex with Ino80 resembling the YY1-Ino80 core complexes of higher eukaryotes.
The disruption of the budding yeast Ino80 complex impairs DNA repair, replication, and transcription (reviewed in references 4 and 77). We wanted to establish if the fission yeast complex also plays a role in these processes. In a plasmid loss assay, a strain with the endogenous ino80 deleted and with a plasmid expressing Ino80 in trans retained the plasmid for survival. This indicates that Ino80 is essential for cell viability (Fig. (Fig.2A).2A). A spore viability assay supported this finding, because ino80 mutant spores did not germinate after ascus dissection of ino80−/+ diploid cells (data not shown). The arp8, nhp10, ies2, and ies6 subunits of the Ino80 complex were deleted without loss of viability. The responses of Δarp8, Δies2, Δies6, and Δnhp10 mutant cells to DNA replication stress and DNA damage were tested by plating serial dilutions of each strain onto plates containing the drug HU or bleocin or exposing the plates to UV irradiation and assaying viability and growth. HU sensitivity indicates an impaired response to replication stress, as the drug depletes deoxyribonucleotide pools, resulting in stalled replication forks (8, 57). Bleocin causes single- and double-strand DNA breaks (53, 65). UV irradiation leads to the formation of pyrimidine dimers and other DNA damage (70). Δarp8 and Δies2 cells, but not Δnhp10 cells, were sensitive to HU, bleocin, and UV irradiation compared to wild-type cells (Fig. (Fig.2B).2B). As the Δies6 mutant could not grow on minimal medium, it was plated on rich medium (YES), where it was also sensitive to HU, bleocin, and UV irradiation (Fig. (Fig.2C).2C). Thus, the fission yeast Ino80 complex is required for the DNA damage and DNA replication stress responses. We then tested whether Δiec1 shows similar phenotypes in the presence of HU and bleocin or after UV irradiation. Under all conditions, the deletion of iec1 severely decreased cell viability (Fig. (Fig.2D).2D). Expression of Iec1 in trans in the Δiec1 background improved survival (Fig. (Fig.2D).2D). Therefore, there is a functional interaction between Iec1 and the Ino80 complex, as Iec1 and the subunits Arp8, Ies2, and Ies6 of the Ino80 complex are required for the physiological response to DNA damage and replication stress.
To explore the functions of Iec1 in addition to its role in DNA repair/replication stress, we tested the viability of wild-type and Δiec1 strains under various conditions. We found that Δiec1 strains do not grow well at high temperature (37°C); in 1% formamide; in 0.5 mM CdSO4, a heavy metal; in 10 μg/ml benomyl, a microtubule inhibitor; and in 0.004% methyl methanesulfonate (MMS), a DNA-alkylating agent (Fig. (Fig.3A3A).
Δiec1 cells showed altered phenotypes when grown under low-phosphate or low-adenine conditions (Fig. (Fig.3B).3B). When intracellular adenine levels are low, the de novo adenine synthesis pathway is activated. Due to a mutation in ade6 in the experimental strain, the pink pigment 5′-phosphoribosyl-5-aminoimidazole (AIR), a purine precursor, accumulated (Fig. (Fig.3C).3C). Unlike wild-type cells, Δiec1 cells grown in low-phosphate rich medium (YES) accumulated this pink pigment (Fig. (Fig.3B).3B). This indicates a switch from salvage to de novo AMP biosynthesis (58). Since the pink pigment is seen upon phosphate starvation when the adenine supply is not limited, this suggests that the de novo biosynthesis pathway is being erroneously activated in Δiec1 cells. It also indicates cross talk between phosphate and purine metabolic pathways in the cell. In line with this, Δiec1 cells grow well in low-adenine rich medium (YES), but unlike wild-type cells, they do not turn pink (Fig. (Fig.3B).3B). Therefore, the switch from adenine salvage to the de novo pathway is not correctly regulated in Δiec1 cells.
Having established that deletion of iec1 affects phosphate and adenine metabolism, we investigated whether other components of the Ino80 complex were also involved in these pathways. We found that Δarp8, Δies6, and Δies2 cells cannot grow on phosphate-depleted medium, whereas Δnhp10 cells remain viable (Fig. (Fig.3B).3B). Δarp8, Δies6, and Δies2 cells also failed to turn pink in low-adenine medium (Fig. (Fig.3B).3B). As the deletion of iec1, arp8, ies2, or ies6 produced the same phenotype, this suggests that these proteins act together in a common pathway. The mutant with a double deletion of iec1 and ies2 (Δiec1 Δies2) retained the ability to turn pink under low-adenine conditions and was viable in low phosphate, whereas the single- deletion mutants were not (Fig. (Fig.3B).3B). This finding confirmed that Iec1 functions with the Ino80 complex and may suggest that Iec1 and Ies2 control one another (see Discussion).
Meiosis and subsequent sporulation are normal responses of fission yeast upon nutrient deprivation. To determine if this pathway was affected, cells were grown on minimal medium (EMM), EMM/low Pi, or low-adenine EMM and then stained with DAPI and visualized using fluorescence microscopy to detect asci (Fig. (Fig.3D).3D). In wild-type cells, asci were detected under all conditions. However, upon iec1 deletion, the sporulation frequency dropped severely in low-phosphate or low-adenine EMM. This could be rescued by expressing Iec1 in trans, indicating that the presence of Iec1 is required for efficient sporulation when cells are deprived of either phosphate or adenine.
We wanted to establish if the sensitivity to low levels of phosphate and adenine seen in the iec1 mutants was due to a direct effect on phosphate-responsive genes and those involved in adenine metabolism. Deletion of Ino80 in budding yeast causes repression of phosphate-responsive genes, such as the PHO5, PHO12, and PHO89 genes, as well as the adenine metabolism genes AAH1 and ADK1 (64, 71, 76). We tested the effect of Iec1 deletion upon the expression of pho1 and pho4, the only two acid phosphatase genes in fission yeast. These genes correspond to the budding yeast PHO5 and PHO3 genes, respectively (49, 60). Pho4 is a constitutive acid phosphatase repressed specifically by thiamine (49, 60). Pho1 is weakly repressed by phosphate, adenine, and thiamine (49, 60, 61). We also tested the expression of apt1, encoding adenine phosphoribosyltransferase 1, an enzyme in the salvage pathway that catalyzes the conversion of adenine to AMP (Fig. (Fig.3C).3C). Northern blot analysis showed that the mRNA levels of all three genes were not affected in Δiec1 cells when the cells were grown at normal phosphate levels (Fig. (Fig.4A4A and data not shown) but were significantly downregulated in low phosphate in Δiec1 cells compared to control cells (Fig. 4A and B). This indicates that, upon phosphate starvation, Iec1 is necessary for the correct expression of genes involved in phosphate response and adenine metabolism. We tested the expression of Iec1 and found that is was markedly upregulated upon phosphate starvation, consistent with a role in regulating the response to this nutritional stress (Fig. (Fig.4C4C).
Ribonucleotide reductase (RNR) is a key enzyme in deoxyribonucleotide metabolism that converts ribonucleotides into deoxyribonucleotides (55). Transcriptional upregulation of the RNR genes is known to occur upon DNA damage (12, 20). In fission yeast, active RNR is a heterotetramer of two large (Cdc22) and two small (Suc22) subunits (21). cdc22 mRNA levels decreased in the iec1 mutant in EMM and, to a lesser degree, in EMM/low Pi (Fig. (Fig.4D).4D). We tested cdc22 expression in cells treated with HU, a condition that in budding yeast leads to upregulation of RNR expression, which is not affected by deletion of INO80 (52, 84). Deletion of iec1 did not affect the upregulation of cdc22 in cells treated with HU (data not shown). We measured deoxyribonucleoside triphosphate pools in wild-type and iec1 mutant cells in EMM and EMM/low Pi and found that, while dCTP and dGTP levels did not change markedly, the dTTP pool increased at the expense of the dATP pool in iec1 mutant cells compared to wild-type cells (Fig. (Fig.4E).4E). These results are consistent with partial inhibition of RNR, as this decrease in dATP and increase in dTTP has been reported in mouse 3T6 cells treated with the ribonucleotide reductase inhibitor HU (8). These observations suggest that the Iec1-Ino80 complex is required for optimal expression of cdc22 under normal and phosphate-depleted growth conditions but that in the presence of HU other activities may override the requirement for Iec1-Ino80.
To test the direct involvement of Ino80 in the regulation of the phosphate-responsive and nucleotide metabolism genes, we performed chromatin immunoprecipitation, followed by hybridization analysis to high-density tiling arrays with 20-bp resolution. Immunoprecipitated Ino80-bound chromatin fragments were isolated from cells grown in EMM or EMM/low Pi. The genome-wide analysis revealed significant differences in Ino80 binding between low and normal phosphate levels. There was a substantial increase in the number of promoter regions (5′ intergenic regions) that showed Ino80 enrichment during phosphate starvation (Fig. (Fig.5A),5A), suggesting a role for Ino80 in low-phosphate stress response. Gene ontology analysis of Ino80 promoter targets revealed that a significant number of the genes targeted in low phosphate are involved in the stress response (Fig. (Fig.5B).5B). The ChIP-on-chip analysis showed that there was an accumulation of Ino80 at the promoters of apt1, ade1, aah1, pho1, and pho4 upon phosphate starvation (Fig. (Fig.5C).5C). aah1 encodes adenine deaminase, which deaminates adenine to hypoxanthine, and ade1 is a gene involved in the de novo synthesis pathway (Fig. (Fig.3C).3C). The increase of Ino80 accumulation at the cdc22 promoter was less pronounced, consistent with our finding that Iec1 regulates cdc22 in phosphate-containing and phosphate-depleted media (Fig. (Fig.5C).5C). However, there was a notable increase in Ino80 binding to the suc22 promoter upon phosphate starvation (data not shown). Other genes involved in adenine metabolism exhibiting a substantial increase of Ino80 binding during phosphate starvation were ade3, ade4, and prs1 (ribose-phosphate pyrophosphokinase) (data not shown). In conclusion, phosphate starvation invokes a stress response upon which Ino80 binds to a large number of promoters, including those of genes that we have shown to be downregulated when iec1 is deleted.
We examined whether deletion of iec1 causes a decrease in the occupancy of the Ino80 complex at the pho1 gene. We addressed this by chromatin immunoprecipitation of FLAG-tagged Ino80 protein in wild-type and Δiec1 cells grown in normal and low-phosphate EMM, followed by quantitative PCR. In cells grown in minimal medium with normal phosphate levels, iec1 deletion did not significantly affect Ino80 occupancy at pho1; if anything, it led to some increased occupancy (Fig. (Fig.6A,6A, left). However, in low phosphate, Ino80 occupancy at pho1 was dependent on Iec1 (Fig. (Fig.6A,6A, right). We also examined the role of Iec1 for binding of Ino80 to the pho4, apt1, aah1, and ade1 promoters. In line with the notion that Iec1 is important for binding of Ino80 to these genes in low phosphate, we found that the occupancy of Ino80 decreased at these loci upon the deletion of iec1 but not at the promoter of the control ade10 gene upon phosphate starvation (Fig. (Fig.6B6B).
The presence of Ino80 at the cdc22 promoter decreased in iec1 mutant cells grown in normal and low phosphate, suggesting that the Iec1-Ino80 complex is required for correct cdc22 regulation. This is in line with our observation that Iec1 is important for cdc22 mRNA expression irrespective of phosphate levels in the medium (Fig. (Fig.6C).6C). These results are different from the results in budding yeast, where Ino80 does not appear to regulate RNR gene transcription (52).
Next, we asked if Iec1 is targeted to the phosphate and nucleotide metabolism genes and if it binds to specific sequence elements, e.g., over the promoters, by performing chromatin immunoprecipitation experiments for Iec1 at the pho1, pho4, apt1, and aah1 loci with cells grown in low-phosphate media. We found Iec1 broadly distributed over the bodies of the genes with no indication that it binds specific sequence elements (Fig. 7A to D).
We concluded that the Ino80 complex is recruited to genes involved in phosphate and adenine metabolism upon phosphate starvation, leading to transcription, and that Iec1 is required for efficient binding of the Ino80 complex at its target genes.
We explored whether Iec1 mediates nucleosome remodeling at target genes to facilitate transcription. We tested this by performing ChIP of histone H3, which reflects nucleosome occupancy, over the promoter regions and gene bodies of the Iec1 target genes, pho1, pho4, apt1, and aah1, in control and iec1 deletion strains grown in low-phosphate EMM. We found that in the absence of Iec1, histone H3 occupancy increased at each locus tested. This occurred not only in the upstream region of the genes, but also within the gene bodies themselves (Fig. 8A to D). Interestingly, Iec1 binding, histone occupancy, and the impact of iec1 deletion on histone occupancy are broadly distributed in the same way over these target genes (compare Fig. Fig.77 and and8).8). Deletion of Arp8, another component of the Ino80 complex, led to a similar increase in histone density over these genes (data not shown). These in vivo data are consistent with the in vitro activity of Ino80 that rendered chromatin more accessible (Fig. (Fig.1B)1B) and indicate that the presence of Ino80, mediated by Iec1, results in an overall reduction of nucleosomes over target loci.
In this study, we found that the fission yeast Ino80 complex is highly conserved through evolution (Table (Table4).4). S. pombe Ino80 interacts with Iec1, which contains zinc finger motifs similar to those found in YY1, an Ino80 complex component of higher eukaryotes (10, 41, 82). In budding yeast, there is no clear Iec1 or YY1 homologue. We identified an Iec1-Ino80 core complex that is similar to the Drosophila and mammalian YY1/PHO-Ino80 core complexes (Table (Table4)4) (10, 41, 82). We tested whether human YY1 can complement Iec1 in its roles in replication stress (HU), temperature stress, and growth without adenine and in low phosphate and found that human YY1 cannot substitute for Iec1 under these conditions, with the exception that it ameliorated the “pink-pigment” phenotype when grown in low phosphate (data not shown). YY1 binds to specific sequence motifs but has recently been shown to also bind structured DNA (four-way junction DNA) in a sequence-independent manner (82). We found no evidence that Iec1 binds specific sequence elements, since it was broadly distributed over target genes (Fig. (Fig.7).7). Its binding seemed to reflect nucleosome density and was consistent with its impact on nucleosome occupancy. It is tempting to speculate that Iec1 may also recognize a specific DNA structure, e.g., the DNA crossovers at the entry-exit sites of the DNA around nucleosomes.
YY1, Ino80, and the YY1-associated Ino80 complex have previously been shown to play roles in the DNA damage response, and our results suggest that this is also the case for the fission yeast Iec1-Ino80 complex (Fig. 2B to D) (1, 65, 82). More recently, the Ino80 complex in budding yeast has been shown to be recruited directly to replication forks (57, 69, 80). It is not known if the mammalian YY1-Ino80 complex has a role in DNA replication. However, our results suggest a role for the Iec1-Ino80 complex in replication, given the sensitivity of Iec1 and Ino80 mutants to the drug HU (Fig. 2B to D).
While Iec1 is required for Ino80 binding, the distributions of Ino80 and Iec1 do not completely mirror each other at target loci. We cannot exclude the possibility that this is the result of epitope exclusion of Iec1 in, for example, the promoter regions of genes. Another explanation could be that not all chromatin-bound Iec1 is in complex with Ino80, but binding of Iec1 to chromatin may be required to load Ino80 onto chromatin. It is possible that Iec1 recognizes a specific histone modification that is upregulated upon phosphate starvation or some other change in chromatin structure. In this respect, it is interesting that iec1 mRNA is upregulated severalfold upon phosphate starvation (Fig. (Fig.4C),4C), indicating that it is part of a concerted stress response involving Ino80.
One question that arises from these results is why Iec1 is present in the fission yeast but not in the budding yeast Ino80 complex. A major difference between budding yeast and fission yeast is the organization of their chromatin structures. Budding yeast exhibits very little higher-order chromatin, whereas fission yeast has extensive heterochromatic domains regulated by the RNA interference machinery, homologues of heterochromatin protein 1, and histone H3 lysine 9 methylation, all of which are absent in budding yeast. Therefore, we might speculate that fission yeast, like higher eukaryotes, requires the presence of a module for Ino80 to guide it through this more complex higher-order chromatin structure.
Interestingly, whereas the single deletions of the Ino80 complex subunit gene iec1 or ies2 resulted in loss of viability in low-phosphate medium, the mutant with a double deletion of iec1 and ies2 was almost as viable as the control cells under this condition. Null alleles in two subunits of the same complex usually show a nonadditive phenotype. One possible explanation is that the suppression phenotype observed in the double mutants is the result of partial complexes that are formed in single mutants and are inhibitory to cell function. An alternative explanation is that Iec1 is in a distinct complex (or functions in isolation) whose function is antagonized by an Ies2-containing complex. This explanation may be more attractive, given that the alleles are nulls, not partial loss-of-function alleles. Therefore, different Ino80 complex components may program opposing functions of Ino80.
Earlier work showed that Ino80 mediates nucleosome eviction at DNA double-strand breaks and that this eviction is important for subsequent DNA repair (52, 56, 74-76). Ino80 has been shown to cooperate with SWI/SNF in nucleosome remodeling at the INO1 and PHO5 promoters in budding yeast (5, 23). However, little is known about how Ino80 remodels chromatin for gene expression. Our results suggest that fission yeast Iec1 and Ino80 cooperate in gene regulation by mediating gene-specific chromatin remodeling. We showed for the first time that Ino80 is involved in the loss of nucleosome density within the promoter regions and bodies of genes (Fig. (Fig.8).8). It is tempting to speculate that Ino80-mediated nucleosome loss over genes underlies the role of Ino80 in transcription, e.g., by facilitating preinitiation complex assembly, RNA polymerase promoter escape, or transcription elongation.
We found that the Δiec1 mutants and mutants of other components of the Ino80 complex are sensitive to low levels of phosphate, and under phosphate starvation, the Δiec1 mutants appear to activate the de novo synthesis pathway of adenine metabolism (Fig. (Fig.3B).3B). This was confirmed by the finding that Δiec1 and Ino80 complex mutants displayed an altered phenotype when grown on low-adenine medium (Fig. (Fig.3B).3B). The mutants also displayed defects in sporulation when grown under either phosphate- or adenine-limiting conditions (Fig. (Fig.3D).3D). Taken together, these findings suggest that there is cross talk between phosphate and adenine metabolic pathways. The requirement for this cross talk within the cell can be explained by the fact that nucleotides are abundant biomolecules and their synthesis is associated with significant phosphate consumption. It therefore seems imperative for cells to evolve regulatory mechanisms to coordinate phosphate utilization and nucleotide synthesis. For example, in budding yeast, the Pho2p transcription factor is required for purine de novo biosynthesis and phosphate utilization pathways (14, 58). Ado1, which encodes an adenosine kinase, negatively regulates PHO5 expression, and the deletion of adenylate kinase (Adk1p) strongly induces the expression of the PHO and ADE genes involved in phosphate utilization and AMP de novo biosynthesis, respectively (27, 33). Our results suggest that a functional Iec1-Ino80 complex is required to ensure this coordinated cross talk within the cell.
We showed that Iec1 regulates the transcription of genes involved in phosphate and adenine metabolism and the recruitment of the Ino80 complex to these genes upon phosphate starvation (Fig. (Fig.4C,4C, ,5C,5C, and and6C).6C). This indicates that the Iec1-Ino80 complex is involved in transcriptional regulation. Ino80 has also been shown by several laboratories to have a direct role in repair in budding yeast (52, 56, 74-76). Therefore, we do not believe that the impaired DNA damage response in cells with an impaired Ino80 complex is simply due to altered gene expression in S. pombe. Since phosphate and nucleotides are essential constituents of many essential biomolecules, their misregulation may influence cellular processes, such as replication and repair. We propose that by regulating phosphate and nucleotide metabolism, the Ino80 complex integrates its role in transcription with its roles in replication and repair. This ensures an integrated response when exposed to cellular stress.
We thank Jacob Dalgaard for yeast strains, initial guidance in fission yeast techniques, and discussion; Robin Allshire and Susan Forsburg for yeast strains; and Jürgen Kohli and Yang Shi for plasmids. We thank one anonymous referee for help with the explanation of the iec1 ies2 double-deletion phenotype; Colin Goding for providing bench space to S.A.; and Peter Fraser, Cameron Osbourne, and Jacqueline Mermoud for comments that improved the manuscript.
This work was supported by a BBSRC studentship (to C.J.H.), a Marie Curie Cancer Care studentship (to S.A.), and a BBSRC grant to the P.D.V.-W. laboratory (BB/F020236/1). Work in the P.D.V.-W. and K.E. laboratories was supported by the EU through the Epigenome Network of Excellence. Work in the K.E. laboratory was also supported by the Swedish Cancer Fund and the Swedish Research Council. Work in the C.K.M. laboratory was supported by grant no. 49557LS from the U.S. Army Research Office. S.E. is supported by a Wellcome Trust Career Development Award.
Published ahead of print on 23 November 2009.