Recent studies have shown that the SWI/SNF family of ATP-dependent chromatin-remodelling complexes play important roles in DNA repair as well as in transcription. The INO80 complex, the most recently described member of this family, has been shown in yeast to play direct role in DNA DSB (double-strand break) repair without affecting the expression of the genes involved in this process. However, whether this function of the INO80 complex is conserved in higher eukaryotes has not been investigated. In the present study, we found that knockdown of hINO80 (human INO80) confers DNA-damage hypersensitivity and inefficient DSB repair. Microarray analysis and other experiments have identified the Rad54B and XRCC3 (X-ray repair complementing defective repair in Chinese-hamster cells 3) genes, implicated in DSB repair, to be repressed by hINO80 deficiency. Chromatin immunoprecipitation studies have shown that hINO80 binds to the promoters of the Rad54B and XRCC3 genes. Re-expression of the Rad54B and XRCC3 genes rescues the DSB repair defect in hINO80-deficient cells. These results suggest that hINO80 assists DSB repair by positively regulating the expression of the Rad54B and XRCC3 genes. Therefore, unlike yeast INO80, hINO80 can contribute to DSB repair indirectly via gene expression, suggesting that the mechanistic role of this chromatin remodeller in DSB repair is evolutionarily diversified.
DNA DSBs (double-strand breaks) can be generated by exposure to genotoxic agents such as ionizing radiation, and also occur when replication forks collapse after encountering chemical adducts or DNA nicks. Accurate and efficient repair of DSBs are essential for maintaining the genome integrity and suppressing cancer development. Upon generation of DSBs, the cellular pathways of the DNA-damage response become activated and stop cell-cycle progression to allow for DNA repair. DSBs in mammalian cells are repaired by the two major pathways of HR (homologous recombination) and NHEJ (non-homologous end-joining). HR involves a number of cellular proteins including the Rad52 epistasis group such as Rad51and Rad54 for error-free DNA synthesis using an undamaged homologous sequence as template, and NHEJ carries out error-prone repair by joining broken DNA ends using a set of proteins including DNA-PK (DNA-dependent protein kinase) and XRCC (X-ray repair complementing defective repair in Chinese-hamster cells) 4/DNA ligase IV [1–3].
Genomic DNA in eukaryotes is packaged into nucleosomes and higher-order chromatin structure, which creates a barrier to protein access for the nuclear processes such as DNA repair. Histone modifications and ATP-dependent chromatin remodelling are the major mechanisms responsible for overcoming this barrier and are known to be important for DNA repair and damage response [4,5]. One of the early events that occurs in response to DSBs is the phosphorylation of the histone variant H2AX. H2AX is phosphorylated by ATM (ataxia telangiectasia mutated)/ATR (ATM- and Rad3-related) kinases immediately after DSB generation, and the phosphorylated H2AX (γ-H2AX) spreads into the megabase region of chromatin surrounding a DSB and forms microscopically visible nuclear foci. γ-H2AX plays important roles in DSB repair and checkpoint activation by providing a binding site for many proteins responsible for the DNA-damage response [6,7]. Several members of the SWI/SNF family of ATP-dependent chromatin-remodelling complexes, including INO80, SWR1, SWI/SNF and RSC, have recently been shown to be implicated directly in DSB repair and checkpoint response [8–22]. Evidence suggests that chromatin-remodelling complexes, sometimes in co-operation with histone-modifying enzymes, alter the nucleosome structure around a DSB in a way to increase the accessibility of the DNA lesion to damage-response proteins [11,14,20,23–25].
The INO80 complex was first purified and characterized from the budding yeast Saccharomyces cerevisiae. The yINO80 (yeast INO80) complex consists of at least 15 subunits, including the INO80 ATPase, Rvb1, Rvb2, actin, Arp (actin-related protein) 4, Arp5, Arp8, Nhp10 (non-histone protein 10), Anc1/Taf14, Ies (INO80 subunit) 1–6 [26,27]. A similar complex to yINO80 has been purified from HeLa cells, and the hINO80 (human INO80) complex comprises at least 13 principal subunits, including the orthologues of INO80, Rvb1, Rvb2, Arp4, Arp5, Arp8, Ies2 and Ies6 (termed INO80, Tip49a, Tip49b, Arp4/BAF53a, Arp5, Arp8, PAPA-1 and C18orf37 respectively in humans), as well as five unique subunits (YY1, NFRKB, Uch37, INO80D, INO80E, MCRS1 and Amida) [28–30]. Both yINO80 and hINO80 complexes, like other typical chromatin remodellers, have an ATPase activity that is stimulated by nucleosomes and DNA, exhibit chromatin-remodelling activity towards in vitro reconstituted nucleosomes, and have an important role in transcription [31,32].
A number of recent studies have shown that the yINO80 complex plays role in a wide range of non-transcriptional nuclear processes such as DNA repair and replication [31,32]. The yINO80 complex has been shown to be recruited to DSBs via interaction with γ-H2AX and to induce histone eviction around a DSB [8–10,23]. Evidence suggests that the yINO80 complex participates directly in both HR and NHEJ repair pathways [8,9,11,14]. Some subunits of the yINO80 complex have been demonstrated to be involved in the regulation of DNA-damage checkpoint responses . Analysis of the global gene-expression profiles shows that the yINO80 complex does not positively regulate the expression of genes implicated in DNA repair or in checkpoint responses, further supporting the direct role for the yINO80 complex in these processes [8,9,13].
Since most of the studies for the roles of the INO80 complex in DSB repair thus far have been focused on yeast, whether and to what extent these functions are conserved in higher eukaryotes has remained uncertain. In the present study, we investigated the hINO80 protein for its potential role in DSB repair. We find that hINO80 is important for efficient DSB repair and cell survival after DNA damage, emphasizing the evolutionary importance of this chromatin remodeller in DNA-damage response. However, in contrast with yINO80, our results suggest that hINO80 contributes to DSB repair to a large extent indirectly by regulating the expression of genes involved in DSB repair.
Plasmids and synthetic siRNA (short interfering RNA)
The vector expressing hINO80-specific siRNA has been described previously . The full-length human Rad54B and XRCC3 cDNA, pCMV-Sport6-Rad54B (clone ID, hMU000216) and pOTB7-XRCC3 (clone ID, hMU011345) respectively were purchased from 21C Frontier Human Gene Bank (Daejon, South Korea). To generate pCMV-Sport6-XRCC3, the XRCC3 cDNA was amplified by PCR using primers 5′-GCGGCCGCATGGATTTGGATCTACTGGACCTG-3′ and 5′-GTCGACTCAGTGGGACTGGGTCCCAGG-3′, and cloned into the NotI and SalI sites of pCMV-Sport6. The insert DNA sequences of all the plasmid constructs were verified by sequencing. The sequences of synthetic siRNA are as follows: hINO80, 5′-UUAAGAGUGUGAUUUCUCAUU-3′ and 5′-UGAGAAAUCACACUCUUAAUU-3′; non-specific control, 5′-CCUACGCCACCAAUUUCGUUU-3′ and 5′-ACGAAAUUGGUGGCGUAGGUU-3′. The vector expressing FLAG–hINO80R was generated by subjecting the FLAG–hINO80 vector  to site-directed mutagenesis using the following oligonucleotides: 5′-AAGCAGAAAGCTTCAGCCAGGAATCTGTTTCTCACC-3′ and 5′-GGTGAGAAACAGATTCCTGGCTGAAGCTTTCTGCTT-3′.
Cells, irradiation and transfection
Generation of HeLa-S3 cells expressing hINO80-specific siRNA has been described previously . Briefly, cells were transfected with the pSuperior.puro vector expressing the hINO80-specific siRNA or corresponding empty vector using calcium phosphate, and treated with puromycin at a concentration of 300 ng/ml. Cells that survived drug selection were amplified further to establish the stable clones and maintained in the medium containing puromycin. Two independent clones stably expressing hINO80-specific siRNA were named si-hINO80-2 and si-hINO80-3 cells, and the stable cells containing the empty vector was named ‘vector’ cells. Cells were irradiated by γ-ray using Cell Irradiation System GC 3000 Elan-Model (137Cs, MDS Nordion). Transfections with plasmids or synthetic siRNA were performed using Lipofectamine™ 2000 (Invitrogen) typically for 48 h.
Polyclonal anti-hINO80 antibodies were raised in rabbits against a synthetic peptide, RKQGKGTNPSGGR, corresponding to amino acids 1549–1561 of hINO80 (GenBank® accession number NM_017553), as described previously . The sources of other antibodies used were as follows: anti-γ-H2AX and anti-H2AX from Upstate Biotechnology; anti-Rad54B and anti-actin from Santa Cruz Biotechnology; anti-XRCC3 from Novus Biologicals; and anti-FLAG from Sigma.
RT (reverse transcription)–PCR
Total RNA was isolated from cells using an RNeasy mini kit (Qiagen) and was converted into cDNA using reverse transcriptase. The resulting cDNA was subjected to PCR amplification as follows: after an initial denaturation at 95 °C for 5 min, the reaction was subjected to 33–35 cycles of denaturation at 95 °C for 30 s, annealing at 55 °C for 30 s and extension at 72 °C for 60 s. PCR products were analysed on 1% agarose gels. The PCR primers used are as follows: hINO80, 5′-GACTATATCCAACGGTCTCT-3′ and 5′-CCTATGCCCTTGAGACTTGA-3′; actin, 5′-CTATCCCTGTACGCCTCTGGC-3′ and 5′-CACGATTTCCCGCTCGGCCGTG-3′; Rad54B, 5′-GCTGTGATTTATAATGACATCCT-3′ and 5′-AGAACAGGGTAGAATGACTCC-3′; XRCC3, 5′-GGTGGTATTCTGGTAGGGAT-3′ and 5′-GGTGGATATTTCAGCCTGT-3′; GAPDH (glyceraldehyde-3-phosphate dehydrogenase) 5′-CCCTTCATTGACCTCAACTAC-3′ and 5′-CCAAAGTTGTCATGGATGACC-3′.
Total RNA was isolated from vector and si-hINO80-2 cells using the RNeasy mini kit according to the manufacturer's instructions. Purified cRNA was labelled with biotin using GeneChip® IVT Labeling kit (Affymetrix). After being cleansed and fragmented, the biotinylated cRNA was hybridized on to an Affymetrix Human Genome U133 Plus 2.0 Array, which comprises more than 54000 probesets and 1300000 distinct oligonucleotide features that represent over 47000 transcripts and variants, including 38500 well-characterized human genes (http://www.affymetrix.com/analysis/index.affx). The hybridized array was washed and stained using GeneChip® Fluidics Station 450 (Affymetrix), and scanned at high resolution using the GeneChip® Scanner 3000. Data analysis was performed with GeneChip® Operating Software, the High-Resolution Scanning Patch and default Statistical Algorithm parameters. Sequence data were obtained from dbEST (NCBI, February 2003), GenBank® (NCBI, February 2003, Release 134), and RefSeq (NCBI, March 2003). Additionally, a draft assembly of the human genome (NCBI, November 2002, Build 31) was used to assess sequence orientation and quality. UniGene (NCBI, January 2003, Build 159) was then used to create initial clusters of cDNA sequences. Sequence-based subclustering was accomplished using StackPACK software (Electric Genetics). Biological data mining for gene ontology was performed by Gene Set Enrichment Analysis to investigate the functional relationships among the ≥2-fold differentially expressed genes using high-throughput GoMiner.
ChIP (chromatin immunoprecipitation)
Approx. 4×107 HEK (human embryonic kidney)-293T cells were cross-linked with 1% formaldehyde at room temperature (25 °C) for 10 min followed by incubation with 125 mM glycine for 5 min at room temperature. After rinsing with ice-cold PBS, cells were harvested and resuspended in FA lysis buffer (50 mM Hepes/KOH, pH 7.5, 140 mM NaCl, 1 mM EDTA, 1% Triton X-100, 0.1% sodium deoxycholate, 0.1% SDS and protease inhibitors). Cell suspension was sonicated six times for 10 s at 30% amplitude using a Cole-Parmer Ultrasonic Homogenizer. The cell lysate was then clarified by centrifugation at 8000 g for 30 s, and the supernatant was transferred into a new tube from which a 50 μl aliquot was taken for the analysis as input. The remaining supernatant was added to BSA (1.5 mg/ml final concentration) and 20 μl of Protein G–Sepharose followed by incubation at 4 °C for 1 h then by centrifugation at 2000 g for 1 min. The supernatant was taken and incubated with 3 μg of the anti-hINO80 antibodies or IgG at 4 °C for overnight, and then incubated with 40 μl of Protein G–Sepharose for 3 h at 4 °C. The Sepharose beads were centrifuged at 2000 g for 1 min and washed five times with wash buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 150 mM NaCl and 20 mM Tris/HCl, pH 8) and twice with final wash buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 500 mM NaCl and 20 mM Tris/HCl, pH 8). Washed beads were eluted with elution buffer (1% SDS and 100 mM NaHCO3) for 15 min at 30 °C, proteinase K (2 mg/ml) was added, and the mixture was incubated at 65 °C for 4 h to reverse cross-link. DNA was purified by phenol/chloroform extraction. A 50 ng amount of immunoprecipitated DNA per sample was subjected to real-time PCR in triplicates using KAPA SYBR® FAST ABI Prism™ (2×) qPCR master mix (KAPA Biosystems) with a 7300 real-time PCR detection system (Applied Biosystems). The primers used in the PCR are as follows: Rad54B promoter, 5′-AAGGATCCCGCCTCGGCGAAGCCAATCGCG-3′ and 5′-TCCTACTTAACATGCTTTAATGCTTTTCCA-3′; Rad54B intron 9, 5′-CCTACACCTATATCCTCCTGTAACCTCTAG-3′ and 5′-GTCACATTGGGGGTTAGAGCTTCAACATAG-3′; XRCC3 promoter, 5′-CTCGGCCTTCTCCTCCAATAGGCGCCTAGC-3′ and 5′-ACACCTCGGAGGACATGACGCCGCCTCAGC-3′; XRCC3 intron 7, 5′-TGGAGCTTGCAGTGAGCTGAGATTGCACCA-3′ and 5′-GAGGGACCTATGGCAGGGCCTTGCAGAGAG-3′.
Neutral comet assays, colony formation assays, immunoblot analysis of histones and cellular proteins, and immunofluorescence microscopy were performed as described previously .
hINO80 knockdown confers DNA-damage sensitivity and inefficient DSB repair
To investigate the role of hINO80 in DNA repair, we employed two independent stable clones expressing hINO80-specific siRNAs, named si-hINO80-2 and si-hINO80-3, and verified that these cells show approx. 70% knockdown of hINO80 protein expression (Figure 1A) . First, we examined the effects of hINO80 knockdown on cell survival after DNA damage using colony-formation assays. si-hINO80-2/3 cells were both hypersensitive to exposure to ionizing radiation as compared with parent or vector cells (Figure 1B). To determine whether the DNA-damage hypersensitivity of si-hINO80-2/3 cells is due to defective DNA repair, we performed comet assays under the neutral conditions that specifically measure DSBs. As shown in Figures 1(C) and 1(D), the efficiency of DBS repair of si-hINO80-2/3 cells was much lower than that of control cells, indicating that hINO80 is required for optimal DSB repair. To confirm these results further, we transiently transfected HeLa cells with either non-specific or hINO80-specific synthetic siRNAs. The expression of hINO80 mRNA and proteins was efficiently down-regulated by transfection with hINO80-specific siRNA (Figure 1E). Similar analysis of these cells as described above showed that hINO80 knockdown renders cells hypersensitive to DNA damage (Figure 1F) and defective in DSB repair (Figure 1G). These results suggest that hINO80 plays an important role in DSB repair. It should be noted that the plasmid form of siRNA for stable cells and the synthetic siRNA for transient transfection were designed to target different sites of the hINO80 sequences, indicating that the phenotypes of hINO80 knockdown are not likely to be due to off-target effects (see below for further control experiments).
hINO80 knockdown confers DNA-damage sensitivity and inefficient DSB repair
hINO80 has little effect on γ-H2AX induction following DNA damage
The yINO80 complex has been shown to interact with γ-H2AX [8–10] and be required for the high-level induction of γ-H2AX following exposure to DNA-damaging agents, such as MMS (methyl methanesulfonate), camptothecin and phleomycin . Mammalian SWI/SNF complexes also have been shown to interact with γ-H2AX and promote the formation of γ-H2AX after exposure to a wide range of ionizing radiation . We therefore asked whether the hINO80 complex has an effect on γ-H2AX induction following DNA damage. First, immunoblot analysis showed that the levels of γ-H2AX after exposure to 5–50 Gy of ionizing radiation were not significantly different between si-hINO80-2/3 and control cells (Figures 2A and 2B). There was little difference in the kinetics of ionizing radiation-induced γ-H2AX formation between si-hINO80-2/3 and control cells (results not shown). Next, we performed immunofluorescence microscopy to examine the effects of hINO80 knockdown on γ-H2AX induction after exposure to a low dose (2 Gy) of ionizing radiation. The average number of γ-H2AX foci at 30 min after exposure to 2 Gy of ionizing radiation was not significantly different between si-hINO80-2/3 and control cells (Figures 2C and 2D, 0.5-h time point). We obtained the same results when we analysed HEK-293T and HCT116 cells after transfecting them with hINO80-specific siRNA (results not shown). Thus, unlike yINO80 or mammalian SWI/SNF, hINO80 is not likely to regulate the induction of γ-H2AX following DNA damage.
hINO80 has little effect on the γ-H2AX induction following DNA damage
Formation of γ-H2AX foci is an indicator of the presence of DSBs within the cells, and disappearance kinetics of γ-H2AX foci can be used for evaluating an efficiency of DSB repair . Having found that γ-H2AX induction is independent of hINO80, we were able to examine the effects of hINO80 on the repair of a low level of DSBs, which would not be possible with comet assays. When recovered for DNA repair up to 10 h after exposure to 2 Gy of ionizing radiation, the parent and vector cells showed a rapid decrease in γ-H2AX foci, indicative of an efficient DSB repair. In parallel experiments, si-hINO80-2/3 cells also showed a decrease in γ-H2AX foci; however, the kinetics were much slower compared with the control cells (Figures 2C and 2D, 1-, 5- and 10-h time points). These results show that hINO80 is important for DSB repair in response to a wide range of ionizing radiation doses.
The Rad54B and XRCC3 genes are repressed in hINO80-deficient cells
To gain insights into whether hINO80 is involved in DSB repair directly or indirectly via gene expression, we analysed the genes that are differentially expressed between vector and si-hINO80-2 cells using the Affymetrix GeneChip® microarray containing 54000 probesets, representing approx. 39000 well-substantiated human genes. We found that more than 2000 genes are repressed or induced at least 2-fold by hINO80 knockdown (see Supplementary Table S1 at http://www.BiochemJ.org/bj/431/bj4310179add.htm). Some 17 genes among those listed are implicated in various pathways of DNA repair, among which the Rad54B and XRCC3 genes are 6- and 2.7-fold repressed by hINO80 knockdown respectively (Figure 3A). Rad54B and XRCC3 are the paralogues of Rad54 and Rad51 respectively and have been shown to play important roles in HR repair [34–38]. To verify the microarray data, we performed RT–PCR and immunoblot analysis and found that the Rad54B and XRCC3 genes in si-hINO80-2/3 cells were significantly reduced in their mRNA and protein levels (Figures 3B and 3C). Confirming these results further, the expression of the Rad54B and XRCC3 genes were also significantly decreased when hINO80 was transiently knocked down in HEK-293T cells (Figure 3D). These results suggest that hINO80 positively regulates the expression of the Rad54B and XRCC3 genes. The hINO80 regulation of Rad54B and XRCC3 gene expression appears to occur constitutively since neither hINO80 nor Rad54B/XRCC3 genes are up-regulated by DNA damage (results not shown).
The Rad54B and XRCC3 genes are repressed in hINO80-deficient cells
hINO80 binds to the promoters of the Rad54B and XRCC3 genes
To investigate whether hINO80 regulates the Rad54B and XRCC3 genes directly, we carried out ChIP assays. Since the transcriptional regulatory sequence elements of the Rad54B and XRCC3 genes are not well defined, we designed the primer sets spanning the 500 and 320 bp regions located just upstream of the first exon of each gene respectively, which we considered to be promoters. We also prepared the primer sets at a region within intron 9 of Rad54B or at a region within intron 7 of XRCC3. These intronic sequences are located approx. 72 kb and 14 kb downstream of the Rad54B and XRCC3 promoters respectively, and are therefore expected to not be bound by hINO80 (Figure 4A). Chromatin was prepared from vector and si-hINO80-2 stable cells and subjected to ChIP using either the anti-hINO80 antibody or pre-immune IgG as control. Immunoprecipitated DNA was analysed by real-time PCR with the aforementioned primer sets. hINO80 was found to be approx. 5- and 8-fold enriched over background on the promoters of the Rad54B and XRCC3 genes in vector cells respectively, indicating that hINO80 binds to these promoters (Figure 4B). Importantly, the fold enrichment of hINO80 on both promoters was significantly reduced in si-hINO80-2 cells, verifying the specificity of the anti-hINO80 antibody in the ChIP experiments (Figure 4B). hINO80 appears to bind specifically to the promoter regions along the length of the Rad54B and XRCC3 genes since it was not enriched on the intronic DNA of either genes (Figure 4B). These results suggest that hINO80 regulates the expression of the Rad54B and XRCC3 genes directly, raising the possibility that the hINO80 complex assists DSB repair via the expression of these genes.
hINO80 binds to the promoters of the Rad54B and XRCC3 genes
Re-expression of Rad54B and XRCC3 rescues the DSB repair defect in hINO80-deficient cells
To test the possibility that hINO80 contributes to DSB repair via the expression of the Rad54B and XRCC3 genes, we examined whether re-expression of these genes would rescue the repair defect in hINO80-deficient cells. HEK-293T cells were co-transfected with non-specific or hINO80-specific siRNAs, in combination with the expression vectors for Rad54B or XRCC3, or with both vectors together. The Rad54B and XRCC3 genes were largely down-regulated by transfection with hINO80 siRNA as seen previously (Figure 5A, compare lanes 1 and 2), and were re-expressed at levels similar to those of their endogenous counterparts after transfection with the corresponding expression vectors (Figure 5A, lanes 3–5). These cells were then evaluated for their DSB repair efficiency using neutral comet assays. After 10 h of recovery following exposure to 30 Gy of ionizing radiation, whereas the hINO80-knockdown cells exhibited a large defect in DBS repair as expected, the ectopic expression of either Rad54B or XRCC3 in these cells rescued the repair defect to a large extent, and the repair defect was further, but not completely, rescued when both Rad54B and XRCC3 were expressed together (Figures 5B and 5C). To confirm these results, we performed similar rescue experiments and evaluated repair efficiency by monitoring γ-H2AX foci after exposure to low-dose ionizing radiation (2 Gy). As shown in Figure 5(D) and 5(E), the results were virtually the same as those obtained from the experiments employing high-dose ionizing radiation and comet assays. These data demonstrate that hINO80 contributes to DSB repair via the expression of the Rad54B and XRCC3 genes.
Re-expression of Rad54B and XRCC3 rescues the DSB repair defect in hINO80-deficient cells
Finally, to exclude a possible off-target effect of hINO80 siRNA in down-regulating Rad54 and XRCC3 expression, we generated an expression vector for a silent mutant form of the FLAG-tagged version of hINO80 whose expression is resistant to the hINO80 siRNA (FLAG–hINO80R). When this vector was transfected into hINO80-knockdown cells, FLAG-tagged proteins were expressed at the similar levels as the endogenous hINO80 in wild-type control cells. Importantly, Rad54B and XRCC3 were both fully re-expressed in those FLAG-tagged protein-expressing cells (Figure 5F). Given that re-expression of Rad54B and XRCC3 rescues the DSB repair defect in hINO80-knockdown cells, these results not only verify the specificity of hINO80 siRNA, but also demonstrate further that it is via the expression of Rad54B and XRCC3 that hINO80 exerts its role of stimulating DSB repair.
Recent studies showed that the SWI/SNF family chromatin-remodelling complexes such as INO80, SWR, SWI/SNF and RSC participate directly in DSB repair and checkpoint response in yeast, and that these complexes do not significantly affect the expression of genes implicated in such DNA-damage responses [8–15,19–22]. Mammalian SWI/SNF complexes were also shown to play a direct role in DSB repair without affecting the expression of genes responsible for this process [16–18]. In the present study, however, we show that hINO80 contributes to DSB repair by positively regulating the expression of the Rad54B and XRCC3 genes, providing the first example that a chromatin remodeller can also assist DNA repair indirectly via gene expression. Our data, showing that re-expression of Rad54B and XRCC3 rescues, to a large extent, the repair defect in hINO80-deficient cells, suggest that this indirect mechanism constitutes an important, albeit not exclusive, pathway by which INO80 regulates DSB repair in human cells. Therefore the role of the chromatin remodeller in DSB repair appears to be diversified between yeast and humans, as well as among different types of remodelling complexes.
Studies suggest that γ-H2AX is important for the direct role of some chromatin remodellers in DSB repair. For example, the yINO80 complex has been shown to be recruited to a DSB via interaction with γ-H2AX, and the mammalian SWI/SNF complex has been shown to interact with γ-H2AX-containing nucleosomes and also to stimulate the formation of γ-H2AX following DNA damage [8,9,16–18]. Some of our experimental data suggest that hINO80 does not use the mechanism involving γ-H2AX for DSB repair. First, hINO80 deficiency does not lead to a significant defect in the induction of γ-H2AX after DNA damage (Figure 2). In addition, hINO80 did not interact with γ-H2AX in the nucleosome immunoprecipitation experiments under the conditions permitting detection of the interaction between SWI/SNF and γ-H2AX (, and results not shown). Furthermore, in immunofluorescence assays, whereas hINO80 does not form repair foci in response to DNA damage, it does form nuclear foci at DNA replication forks and these hINO80 foci do not overlap with γ-H2AX induced by spontaneously occurring DSBs around the replication forks . Thus it appears that hINO80, unlike yINO80 or mammalian SWI/SNF, neither interacts with γ-H2AX nor promotes its induction following DNA damage. A reason for this might be that the hINO80 complex does not contain the Nhp10 and the Ies3 subunits, which have been shown to be important for the interaction between the yINO80 complex and γ-H2AX [9,28].
It has been shown that genetic ablation of either Rad54B or XRCC3 confers inefficient DSB repair and hypersensitivity to DNA-damaging agents such as ionizing radiation, suggesting that each of these genes is independently important for DSB repair [34–38]. In agreement with these studies, defect of DSB repair in hINO80 knockdown cells is better rescued by simultaneous expression of Rad54B and XRCC3 than by individual expression of each gene (Figure 5). Interestingly, however, expression of either gene alone can rescue the repair defect of hINO80-knockdown cells to a significant level, seemingly conflicting with the results of the aforementioned studies. Although further studies will be necessary for clear explanation of these results, one possible reason could be because the Rad54B and XRCC3 genes are not completely repressed in hINO80-knockdown cells. In the presence of some suboptimal levels of Rad54B and XRCC3 proteins, the defect of DNA repair in hINO80-knockdown cells, unlike in the situation of Rad54B- or XRCC3-null mutant cells, could be compensated for, to some extent, by ectopic expression of either protein only. In support of this explanation, members of the Rad52 epistasis group often exhibit genetic interaction in which the effects of one gene are modified by one or several other genes . For example, ablation of Rad54B in mouse embryonic stem cells has an effect on HR efficiency in the absence of Rad54, but has little in the presence of Rad54 .
Our data, showing that the repair defect of hINO80-deficient cells is not completely rescued by re-expression of both the Rad54B and XRCC3 genes, indicate that hINO80 also contributes to DSB repair through mechanisms other than by means of regulating these two genes. The possibility cannot be formally excluded that hINO80 may regulate genes that have yet to be identified to be involved in DSB repair. Alternatively, and more preferably, hINO80 might also contribute to DSB repair by participating directly in the process of DSB repair. Indeed, it has been recently shown that the mammalian INO80 complex interacts with the YY-1 polycomb group transcription factor and that YY-1 preferentially binds to a recombination-intermediate DNA structure for HR repair, strongly suggesting a direct role for INO80 in DSB repair . The INO80 complex of mammals might therefore have evolved to adopt both direct and indirect mechanisms to assist DSB repair, which would certainly be beneficial to protecting their genome from the life-threatening DNA damage such as DSBs.
ataxia telangiectasia mutated
human embryonic kidney
non-histone protein 10
short interfering RNA
X-ray repair complementing defective repair in Chinese-hamster cells
Eun-Jung Park and Shin-Kyoung Hur essentially through equal contribution designed and performed the experiments, and analysed and interpreted the data. Jongbum Kwon devised the overall project, interpreted the data and wrote the paper.
This work was supported by the Research Program for New Drug Target Discovery [grant number M10748000334-08N4800-33410 (to J.K.)] and the Mid-Career Researcher Program [grant number R01-2007-000-10571-0 (to J.K.)] through the National Research Foundation of Korea (NRF) funded by the Korea Ministry of Education, Science and Technology (MEST).
These authors equally contributed to this work.