ATP-dependent chromatin remodeling facilitates nucleotide excision repair of UV-induced DNA lesions in synthetic dinucleosomes
2001; Springer Nature; Volume: 20; Issue: 8 Linguagem: Inglês
10.1093/emboj/20.8.2004
ISSN1460-2075
AutoresKiyoe Ura, Marito Araki, Hideaki Saeki, Chikahide Masutani, Takashi Ito, Shigenori Iwai, Toshimi Mizukoshi, Yasufumi Kaneda, Fumio Hanaoka,
Tópico(s)DNA and Nucleic Acid Chemistry
ResumoArticle17 April 2001free access ATP-dependent chromatin remodeling facilitates nucleotide excision repair of UV-induced DNA lesions in synthetic dinucleosomes Kiyoe Ura Corresponding Author Kiyoe Ura Division of Gene Therapy Science, Osaka University School of Medicine, 2-2 Yamada-oka, Suita, Japan Search for more papers by this author Marito Araki Marito Araki Institute for Molecular and Cellular Biology, Osaka University and CREST, JST, 1-3 Yamada-oka, Suita, Osaka, 565-0870 Japan Present address: Department of Genetics, Duke University Medical Center, Durham, NC, 27710 USA Search for more papers by this author Hideaki Saeki Hideaki Saeki Division of Gene Therapy Science, Osaka University School of Medicine, 2-2 Yamada-oka, Suita, Japan Search for more papers by this author Chikahide Masutani Chikahide Masutani Institute for Molecular and Cellular Biology, Osaka University and CREST, JST, 1-3 Yamada-oka, Suita, Osaka, 565-0870 Japan Search for more papers by this author Takashi Ito Takashi Ito Second Department of Biochemistry, Saitama Medical School, Moroyama, Iruma-gun, Saitama, 350-0495 USA Search for more papers by this author Shigenori Iwai Shigenori Iwai Biomolecular Engineering Research Institute, 6-2-3 Furuedai, Suita, Osaka, 565-0874 Japan Search for more papers by this author Toshimi Mizukoshi Toshimi Mizukoshi Biomolecular Engineering Research Institute, 6-2-3 Furuedai, Suita, Osaka, 565-0874 Japan Search for more papers by this author Yasufumi Kaneda Yasufumi Kaneda Division of Gene Therapy Science, Osaka University School of Medicine, 2-2 Yamada-oka, Suita, Japan Search for more papers by this author Fumio Hanaoka Fumio Hanaoka Institute for Molecular and Cellular Biology, Osaka University and CREST, JST, 1-3 Yamada-oka, Suita, Osaka, 565-0870 Japan Search for more papers by this author Kiyoe Ura Corresponding Author Kiyoe Ura Division of Gene Therapy Science, Osaka University School of Medicine, 2-2 Yamada-oka, Suita, Japan Search for more papers by this author Marito Araki Marito Araki Institute for Molecular and Cellular Biology, Osaka University and CREST, JST, 1-3 Yamada-oka, Suita, Osaka, 565-0870 Japan Present address: Department of Genetics, Duke University Medical Center, Durham, NC, 27710 USA Search for more papers by this author Hideaki Saeki Hideaki Saeki Division of Gene Therapy Science, Osaka University School of Medicine, 2-2 Yamada-oka, Suita, Japan Search for more papers by this author Chikahide Masutani Chikahide Masutani Institute for Molecular and Cellular Biology, Osaka University and CREST, JST, 1-3 Yamada-oka, Suita, Osaka, 565-0870 Japan Search for more papers by this author Takashi Ito Takashi Ito Second Department of Biochemistry, Saitama Medical School, Moroyama, Iruma-gun, Saitama, 350-0495 USA Search for more papers by this author Shigenori Iwai Shigenori Iwai Biomolecular Engineering Research Institute, 6-2-3 Furuedai, Suita, Osaka, 565-0874 Japan Search for more papers by this author Toshimi Mizukoshi Toshimi Mizukoshi Biomolecular Engineering Research Institute, 6-2-3 Furuedai, Suita, Osaka, 565-0874 Japan Search for more papers by this author Yasufumi Kaneda Yasufumi Kaneda Division of Gene Therapy Science, Osaka University School of Medicine, 2-2 Yamada-oka, Suita, Japan Search for more papers by this author Fumio Hanaoka Fumio Hanaoka Institute for Molecular and Cellular Biology, Osaka University and CREST, JST, 1-3 Yamada-oka, Suita, Osaka, 565-0870 Japan Search for more papers by this author Author Information Kiyoe Ura 1, Marito Araki2,3, Hideaki Saeki1, Chikahide Masutani2, Takashi Ito4, Shigenori Iwai5, Toshimi Mizukoshi5, Yasufumi Kaneda1 and Fumio Hanaoka2 1Division of Gene Therapy Science, Osaka University School of Medicine, 2-2 Yamada-oka, Suita, Japan 2Institute for Molecular and Cellular Biology, Osaka University and CREST, JST, 1-3 Yamada-oka, Suita, Osaka, 565-0870 Japan 3Present address: Department of Genetics, Duke University Medical Center, Durham, NC, 27710 USA 4Second Department of Biochemistry, Saitama Medical School, Moroyama, Iruma-gun, Saitama, 350-0495 USA 5Biomolecular Engineering Research Institute, 6-2-3 Furuedai, Suita, Osaka, 565-0874 Japan *Corresponding author. E-mail: [email protected] The EMBO Journal (2001)20:2004-2014https://doi.org/10.1093/emboj/20.8.2004 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info To investigate the relationship between chromatin dynamics and nucleotide excision repair (NER), we have examined the effect of chromatin structure on the formation of two major classes of UV-induced DNA lesions in reconstituted dinucleosomes. Furthermore, we have developed a model chromatin-NER system consisting of purified human NER factors and dinucleosome substrates that contain pyrimidine (6-4) pyrimidone photoproducts (6-4PPs) either at the center of the nucleosome or in the linker DNA. We have found that the two classes of UV-induced DNA lesions are formed efficiently at every location on dinucleosomes in a manner similar to that of naked DNA, even in the presence of histone H1. On the other hand, excision of 6-4PPs is strongly inhibited by dinucleosome assembly, even within the linker DNA region. These results provide direct evidence that the human NER machinery requires a space greater than the size of the linker DNA to excise UV lesions efficiently. Interestingly, NER dual incision in dinucleosomes is facilitated by recombinant ACF, an ATP-dependent chromatin remodeling factor. Our results indicate that there is a functional connection between chromatin remodeling and the initiation step of NER. Introduction DNA is frequently damaged by a variety of environmental and endogenous agents produced by physical or chemical reactions (Friedberg et al., 1995). Nucleotide excision repair (NER), one of the most important and well-studied pathways of DNA repair, is capable of eliminating a broad range of structurally unrelated lesions from DNA, including UV-induced damage (Sancar, 1996; de Laat et al., 1999; Batty and Wood, 2000). The NER process consists of four steps: (i) recognition of the damaged DNA; (ii) excision from the DNA of the 24–32 residues surrounding the damaged oligonucleotide; (iii) DNA synthesis to fill the gap; and (iv) ligation of the nick (Sancar, 1996; de Laat et al., 1999; Batty and Wood, 2000). Each process is carried out by multi-protein complexes and has been thoroughly analyzed using highly purified human proteins or recombinant polypeptides (Araki et al., 2000; Araujo et al., 2000). However, mechanisms of NER in the context of chromatin structure remain unclear, partly due to the lack of systematic analyses of NER in defined chromatin. Packaging of eukaryotic DNA into chromatin affects many of the dynamic processes of DNA metabolism, including transcription, replication, recombination and repair. This is because the assembly of nucleosomes, the basic unit of chromatin, changes the structure of the DNA and restricts the access of DNA-binding factors to their recognition sites (Luger et al., 1997; Wolffe, 2000). Recent studies have led to the purification of >10 large protein complexes that locally disrupt or alter the association of histones with DNA depending on ATP hydrolysis (LeRoy et al., 2000; Shen et al., 2000; Vignali et al., 2000). All of these ATP-dependent chromatin remodeling complexes contain the ATPase subunit of the SNF2 superfamily and fall into one of three distinct families of remodeling complexes: SWI/SNF2-like (e.g. SWI/SNF, RSC and BRM), ISWI-like (e.g. NURF, CHRAC, ACF, yISWI and RSF) and Mi-2-like (e.g. NURD) (Kingston and Narlikar, 1999; Vignali et al., 2000). Although functional analysis of these complexes has been restricted mainly to transcription, it is conceivable that similar activities may also assist NER in chromatin (Thoma, 1999). UV light induces two major classes of mutagenic DNA lesions: cyclobutane pyrimidine dimers (CPDs) and pyrimidine (6-4) pyrimidone photoproducts (6-4PPs). UV-induced damage formation has been shown to be affected by the chromatin environment (Pfeifer, 1997; Thoma, 1999). To investigate the relationship between chromatin dynamics and NER, we have used synthetic, physiologically spaced and positioned dinucleosome substrates that contain 6-4PP lesions at specific sites. The reconstituted dinucleosome system was previously established to analyze the effect of chromatin structure on transcription (Ura et al., 1995, 1997), and it demonstrated structural changes in chromatin depending on the incorporation of linker histones (Ura et al., 1995; Sato et al., 1999). Here, we have found that the excision of 6-4PP lesions by highly purified human NER factors (Araki et al., 2000) is inhibited when the DNA is folded into chromatin, irrespective of the position of 6-4PP within the chromatin. In addition, we demonstrate that ATP-dependent chromatin remodeling by ATP-utilizing chromatin assembly and remodeling factor (ACF) (Ito et al., 1997) facilitates the excision of 6-4PP lesions by the NER factors, in particular those situated in the linker DNA. Our results indicate that NER clearly requires chromatin reconfiguration, which probably involves alterations in histone–DNA interactions by ATP-dependent chromatin remodeling factors other than the minimal components of NER complexes. Results Effects of chromatin structure on acquisition of UV-induced DNA damage CPDs and 6-4PPs are the two major classes of mutagenic DNA lesions induced by UV light, and they induce a DNA bend or kink of 7–9° and 44°, respectively (Wang and Taylor, 1991; Kim and Choi, 1995). Their frequency and distribution depend on the DNA sequence, the local DNA structure and the association of DNA with chromosomal proteins (Friedberg et al., 1995; Pfeifer, 1997; Thoma, 1999). To address the question of how the formation of UV-induced lesions is affected by histone octamers and linker histone H1, we used reconstituted dinucleosomes composed of two tandem 5S RNA genes (Figure 1A). The dinucleosome is the minimal unit of chromatin containing intact linker DNA. Purified 5′-end-labeled dinucleosome cores were incubated with histone H1, resulting in at least one H1 binding to each dinucleosome. The dinucleosomes with or without histone H1 were irradiated with a UV dose of 100 or 450 J/m2, as was the naked 5S DNA that served as a control. The CPD and 6-4PP UV-induced lesions were then mapped as described in Figure 1B. As shown in Figure 1C, the overall yields and positions of both CPDs (lanes 1–9) and 6-4PPs (lanes 10–18) were not significantly affected by the 5S nucleosome structure. On average, each dinucleosome irradiated with 450 J/m2 UV contained 0.50 CPD and 0.095 6-4PP lesions. Our previous atomic force microscopy images and biochemical analyses showed a structural change in the linker DNA of the dinucleosome, depending on the incorporation of histone H1 (Ura et al., 1995; Sato et al., 1999). However, we did not observe here that the association of H1 altered the formation of either of the two types of UV-induced lesion in dinucleosomes (Figure 1C, lanes 6 and 9 for CPD, 15 and 18 for 6-4PP). Careful comparison of CPD formation between naked DNA and nucleosomal DNA did, however, reveal inhibition of CPD formation at several sites, in particular where the minor groove faces the histone octamer surface (Figure 1C, arrowheads). This is in agreement with earlier work (Gale et al., 1987; Pehrson, 1995; Liu et al., 2000). In addition, CPDs in the dinucleosome were less frequent around the dyad axis of the nucleosome (Figure 1C, dots) compared with naked DNA. Our quantitative analysis shows that, compared with the naked DNA control, CPD formation was reduced up to 4-fold by nucleosome folding at specific sites in the nucleosome (Figure 1C). Figure 1.The effect of chromatin structure on CPD and 6-4PP formation in 5S dinucleosome DNA. (A) Structure of the 5S dinucleosome DNA (Ura et al., 1995). The 5S dinucleosome DNA (418 bp) contains two 197 bp tandem repeats of the Xenopus borealis somatic 5S RNA gene and associated upstream sequences, including the entire nucleosome positioning element. Thick arrows indicate the location and orientation of the 120 bp 5S RNA genes. The major nucleosome positions in the absence (core particle position) or presence (chromatosome position) of linker histone are based upon our previous micrococcal nuclease mapping results and are indicated by ovals. All positions are relative to the transcription start site of the 5S RNA gene, which is denoted by +1. (B) Flow diagram showing how the UV-induced CPD and 6-4PP sites on UV-irradiated dinucleosomes were mapped. (C) Mapping of CPD and 6-4PP sites on UV-irradiated dinucleosomes. Naked DNA (lanes 1–3 and 10–12) and dinucleosomes without histone H1 (lanes 4–6 and 13–15) or with histone H1 (lanes 7–9 and 16–18) were irradiated with increasing UV doses (0, 100 or 450 J/m2) as shown at the top of the lanes. Purified DNA was digested with T4 endonuclease V to map CPD sites (lanes 1–9), or incubated with CPD photolyase followed by digestion with UVDE to map 6-4PP sites (lanes 10–18). The amount of DNA in each lane was adjusted according to the total radioactivity before performing denaturing PAGE. The position at which CPD formation is clearly inhibited by the nucleosome folding is shown by arrowheads. The rotational positioning of the 5S nucleosome was demonstrated by the hydroxyl radical cleavage pattern of dinucleosomes in lane 19. M, Maxam–Gilbert G tracks as markers. Dots indicate the the axis of dyad symmetry of the nucleosome. Vertical arrows show the location and orientation of the 5S RNA gene. Ovals indicate the predominant regions contacted by the nucleosome cores as shown in (A) (core particle position). Asterisks show the sites where synthetic 6-4PPs are introduced into the 5S dinucleosome for use as templates in the NER assay. Download figure Download PowerPoint In contrast to CPD formation, dinucleosome assembly did not seem to affect the distribution of 6-4PPs (Figure 1C). We found that the 6-4PP distribution in the dinucleosome was almost the same as that of naked DNA, even in the linker DNA, where preferential formation of 6-4PPs in chromatin was reported with mixed-sequence nucleosomes from UV-irradiated cells (Mitchell et al., 1990). On the whole, our results demonstrate that the chromatin structure of DNA does not restrict its acquisition of UV-induced lesions. 5S dinucleosome substrates containing specifically positioned synthetic 6-4PPs The chromatin structure restricts the access of proteins or protein complexes to their respective DNA binding sites (Wolffe, 2000). As we have shown above, however, UV-induced DNA lesions occur throughout chromatin DNA (Figure 1C). Thus, there must be molecular mechanisms that facilitate DNA repair within chromatin. To examine the initiation mechanisms of NER at the chromatin level, we made 5S dinucleosomes that contained 6-4PPs at specific positions on the non-transcribed strand of the 5S DNA: either at the center of two nucleosome cores (center-dinucleosome) or in the linker DNA (linker-dinucleosome). We chose 6-4PP as the model UV lesion because this lesion is removed rapidly by global genome repair (GG-NER), which is the transcription-independent NER subpathway. In contrast to 6-4PPs, CPDs are repaired only very slowly by GG-NER. The other NER subpathway, transcription-coupled repair (TC-NER), is instead responsible for the efficient removal of CPDs from the transcribed strand of expressed genes (de Laat et al., 1999; Thoma, 1999). In order to maintain 5S nucleosome positioning in damaged 5S dinucleosomes, the 6-4PP at the center of the nucleosome was generated at the 5′-TC (−2) site of the 5S gene. The TC sequence that faces out from the nucleosome surface at this site allowed efficient UV-induced 6-4PP formation (Figure 1C, asterisks). The 6-4PP site in the linker DNA was positioned at 5′-TC (+80). The DNA fragments bearing 6-4PPs at either of the two sites were prepared using chemically synthesized 6-4PP-containing oligonucleotides (Figure 2A). 5′-end-radiolabeled dinucleosomes were reconstituted by salt dialysis. The efficiency with which the dinucleosomes were reconstituted was not significantly affected by the presence of either or both of the 6-4PPs. Figure 2.Preparation and characterization of dinucleosome substrates containing synthetic 6-4PPs. (A) Schematic representation of dinucleosomes that contain 6-4PPs either in the center of the nucleosome (center-dinucleosome) or in the linker DNA (linker-dinucleosome). Asterisks indicate 6-4PPs. (B) DNase I footprinting of 6-4PP-containing dinucleosomes. Lanes 1–3, undigested DNA; lanes 4–6, digested naked DNA; lanes 7–9, digested dinucleosomes. Lane 10 shows a G-specific cleavage reaction as a marker. N, undamaged DNA (lanes 1, 4 and 7); C, center-dinucleosome DNA (lanes 2, 5 and 8); L, linker-dinucleosome DNA (lanes 3, 6 and 9). Dots indicate the axis of dyad symmetry of the nucleosome. The 10 bp repeat pattern, common to nucleosomal footprinting, is shown by arrowheads. As a control, we show that DNA fragments containing 6-4PPs are frequently slightly nicked at damage sites even without DNase I treatment (lanes 2 and 3) (Franklin et al., 1982). Dots, vertical arrows, ovals and asterisks indicate the same items described in Figure 1A. (C) Micrococcal nuclease mapping of core positions on dinucleosomes containing 6-4PPs. Dinucleosomes were digested with micrococcal nuclease. 5′-end-labeled core particle DNA (147 bp) was digested with EcoRV. There is no EcoRV site in the non-specific DNA that was used for nucleosome reconstitution with dinucleosome DNA. Lanes 1–3, undigested core particle DNA; lanes 4–6, DNA digested with EcoRV. Lane 7 shows a DNA maker produced by the hydroxyl radical cleavage reaction. N, undamaged DNA; C, center-dinucleosome DNA; L, linker-dinucleosome DNA; CP, core particle DNA products of digestion. Download figure Download PowerPoint As 6-4PPs can cause marked DNA distortions, it is still possible that nucleosome positioning in 5S DNA is altered by the introduction of 6-4PPs. We thus investigated nucleosome positioning in reconstituted 6-4PP-containing dinucleosomes by using two independent nuclease mapping methodologies, namely, DNase I footprinting and micrococcal nuclease mapping. The DNase I cleavage pattern of chromatin shows the rotational positioning of nucleosome cores. Well-positioned nucleosome cores like undamaged 5S dinucleosomes give typical 10 nucleotide cleavage ladders because the nuclease cuts preferentially where the DNA is maximally exposed on the nucleosome surface (Ura et al., 1995). DNase I footprinting patterns of dinucleosomes that contain 6-4PPs at the center of nucleosomes (center-dinucleosomes) or in linker DNA (linker-dinucleosomes) were exactly the same as that of undamaged dinucleosomes, indicating that rotational positioning of 5S dinucleosomes was not changed by the introduction of 6-4PPs (Figure 2B). Furthermore, we performed micrococcal nuclease mapping of nucleosome core positions on each 5S dinucleosomes. Each dinucleosome was trimmed with micrococcal nuclease and digestion products were end labeled. Nucleosome core particle fragments (147 bp) were isolated and subjected to EcoRV restriction enzyme digestion, which gave cuts around the middle of the positioned 5S nucleosome cores (Figure 1A). Carrier DNA that was used for reconstitution of dinucleosomes had no EcoRV site and remained at the position of core particles (Figure 2C, CP). The EcoRV restriction enzyme cleavage patterns of core particle fragments derived from each dinucleosome containing 6-4PPs at the center of nucleosomes or in linker DNA were almost the same as that from undamaged dinucleosomes (Figure 2C). These results revealed that not only rotational positioning but also two major translational positions of 5S nucleosome cores (Figure 1A) remained unchanged after the introduction of positioned 6-4PPs. Excision of 6-4PP from dinucleosomes by purified human NER factors We assessed the ability of NER to remove the 6-4PP lesions from our dinucleosome substrates by adding the purified human NER factors RPA, XPA, XPC-hHR23B, XPG, XPF-ERCC1 and TFIIH. Excision reactions were monitored by following the appearance of internally labeled dual incision products. The predicted products generated by dual or uncoupled incision of the substrates are shown in Figure 3A. When the naked dinucleosome DNA template was used, the mixture of purified NER factors efficiently excised 6-4 PPs as 24–28 nucleotide oligomers. Approximately 30% of the total substrate signal was thus processed in 30 min (Figure 3B, lanes 2 and 5). Dual incision of both positioned 6-4PPs was repressed by increasing the number of histone octamers on both types of 5S dinucleosome DNA. Surprisingly, strong repression of NER in physiologically spaced dinucleosome templates was observed even when the 6-4PP lesion was located in the linker DNA (Figure 3B, lane 7). The excision activity in dinucleosomes, either at the center of nucleosomes or in the linker DNA, was reduced to <20% of that of naked DNA (Figure 3C). These results suggest that the space that human NER complex requires to excise UV lesions in chromatin is greater than the general length of linker DNA. In the mononucleosome, a histone octamer is located at one of the two 5S nucleosome positioning elements of the dinucleosome DNA. The 50% loss of repression on center-mononucleosomes compared with naked DNA suggests that displacement of only one histone octamer from a damaged DNA site facilitates NER in chromatin (Figure 3C). We concluded that six human NER factors: RPA, XPA, XPC-hHR23B, XPG, ERCC1-XPF and TFIIH, are insufficient to overcome the structural barriers that chromatin poses to the removal of DNA damage, even when the lesion is located within the linker DNA region. Figure 3.Effects of chromatin structure on NER. (A) Schematic representation of the predicted cleavage products after incision of the two internally labeled 6-4PP substrates, center-dinucleosome DNA (a) and linker-dinucleosome DNA (b). The labeled products of the excision assay are represented by thin bidirectional arrows along with their lengths in nucleotides (nt). Thick arrows indicate the location and orientation of the 120 bp 5S RNA gene. The 5′-internal promoter of two repeats of the 5S RNA gene was mutated by base substitution, and is indicated by the white box in plasmid pXSP5S. Internally radiolabeled sites of center- and linker-dinucleosome DNA are indicated by black and white asterisks, respectively. The major nucleosome core positions are shown by ovals. (B) Excision activity of NER factors on synthetic chromatin templates. Chromatin templates were assembled on internally labeled DNA fragments containing either no 6-4PP (lane 1), 6-4PPs at the nucleosome centers (lanes 2–4) or a 6-4PP in the linker DNA (lanes 5–7). These were used as substrates in the excision assay. Naked DNA (lanes 1, 2 and 5), mononucleosomes (lanes 3 and 6) or dinucleosomes (lanes 4 and 7) were incubated with purified NER factors for 30 min at 30°C. Cleavage products were analyzed by 12% denaturing PAGE. Each value obtained from the excision products is expressed as a percentage of the total signal in each lane, which is shown at the bottom of the lane. This figure is representative of data obtained from repeated trials of independently conducted experiments. (C) The quantitative data from (B) shown as a bar graph. The relative excision activity of dinucleosomes containing 6-4PP at the center of nucleosomes (black bars) or in linker DNA (white bars) was calculated by taking the value of excision on each naked DNA template as 100%. Download figure Download PowerPoint ATP-dependent chromatin remodeling by ACF facilitates NER of damage in linker DNA The strong inhibition of NER in the reconstituted dinucleosome demonstrates the dependence of repair on chromatin dynamics, as can be inferred from previous in vitro studies using bacterial repair enzymes (Schieferstein and Thoma, 1998). Recent studies indicate that chromatin modification, including ATP-dependent chromatin remodeling and histone modification, facilitates transcription of chromatinized genes (Kingston and Narlikar, 1999; Kornberg and Lorch, 1999; Vignali et al., 2000). To unravel the possible molecular mechanisms of NER in chromatin, we investigated the effect of ATP-dependent chromatin remodeling on NER of the 5S dinucleosomes using recombinant ACF. ACF is an ATP-dependent chromatin remodeling complex derived from Drosophila, consisting of ISWI, a member of the SWI2/SNF2 family containing an ATPase/helicase domain, and Acf1, in addition to two other polypeptides (Ito et al., 1997, 1999). Recently, purified recombinant ACF (rACF) that consists only of Acf1 and ISWI has been shown to function in ATP-dependent chromatin assembly and remodeling (Ito et al., 1999, 2000). To examine whether ACF can influence NER of damaged chromatin, we compared the excision activity of purified NER factors in the presence or absence of rACF. Recombinant ACF complex (Acf1 and ISWI) was expressed in insect cells and purified. As expected, there was no detectable change in NER activity on naked DNA templates upon the addition of ACF (Figure 4A, lanes 1, 2, 5 and 6). In contrast, in the presence of ACF, significant activation of NER was observed in linker-dinucleosome templates (Figure 4A, lanes 3 and 4). Interestingly, ACF had no effect on NER in 6-4PP-containing dinucleosomes when the lesions were located at the center of the nucleosomes (Figure 4A, lanes 7 and 8). Specific activation of NER in the linker DNA of dinucleosomes by ACF indicates that the addition of rACF did not cause artificial displacement or disruption of histone octamers from the short chromatin templates. In order to confirm the effect of ACF on NER in chromatin, we added a polyclonal antiserum against the N-terminal region of ISWI to the NER reaction mixtures. NER of the naked DNA template was not influenced by the addition of the antiserum (Figure 4A, lanes 1 and 10). In contrast to the naked DNA template, ACF-dependent activation of NER on linker-dinucleosome templates was strongly inhibited by the antiserum against ISWI (Figure 4A, lanes 3, 4 and 9). Curiously, NER activity on the linker-dinucleosome in the presence of ACF and anti-ISWI serum was only 7.1% of that of the naked DNA control. This is lower than the relative excision activities observed in dinucleosomes in the absence of ACF (Figure 4B). This effect of ISWI antiserum may be due to contamination of our partially purified TFIIH complex with ISWI-containing chromatin remodeling complexes. Indeed, the relative excision activities in dinucleosomes compared with control naked templates varied depending on the preparation of the TFIIH complex (data not shown). Our results demonstrate that ISWI complexes enhance NER dual incision activity by 3.8-fold, and that this can specifically occur in the linker DNA of structures as small as a dinucleosome (Figure 4A, lanes 4 and 9). Although the repressive effect of chromatin was not completely relieved by ACF, these results demonstrate for the first time a functional connection between ATP-dependent chromatin remodeling and NER. Figure 4.Effects of rACF on NER in vitro. (A) Excision in synthetic chromatin templates. The excision reaction of NER was carried out either in the presence or absence of ACF (300 fmol) as indicated at the top of each lane. DNA fragments containing 6-4PPs in linker DNA (lanes 1–4, 9 and 10) or at the nucleosome centers (lanes 5–8) were used as substrates for the excision assay. Naked DNA (lanes 1, 2, 5, 6 and 10) or dinucleosomes (lanes 3, 4, 7, 8 and 9) were incubated with purified NER proteins in the absence (lanes 1–8) or presence (lanes 9 and 10) of antiserum against ISWI for 45 min at 30°C. Each value obtained from the excision products is expressed as a percentage of the total signal in each lane, which is shown at the bottom of the lane. This figure is representative of data obtained from repeated trials of independently conducted experiments. (B) The quantitative data from (A) shown as a bar graph. Relative excision activity of center-dinucleosomes (black bars) or linker-dinucleosomes (white bars) was calculated by taking the value of excision from each naked DNA template as 100%. Download figure Download PowerPoint To investigate the mechanism by which ACF stimulates NER activity on the 5S dinucleosomes, we compared the accessibility of dinucleosomal DNA to cleavage by restriction enzymes under conditions similar to those employed for the NER reaction. Both 6-4PP-bearing and undamaged 5′-end-labeled dinucleosomes were subjected to digestion by either EaeI or RsaI. Based on the restriction enzyme sites in 5S dinucleosome DNA (Figure 5A), EaeI digestion should cut the DNA 6 bp downstream from the site where the 6-4PP lesion is placed at the nucleosomal center, while RsaI should cleave dinucleosome DNA 2 bp upstream from the site where the 6-4PP lesion in the linker DNA is situated. In the absence of ACF, dinucleosomes that are undamaged or have the 6-4PP lesion in the linker DNA are digested by EaeI much more poorly than by RsaI. In contrast, the center-dinucleosomes are digested by EaeI more efficiently, indicating that the 6-4PP lesion at the nucleosomal centers may cause local DNA distortion that improves the accessibility of nucleosomal DNA to EaeI (Figure 5B, top, lanes 7–9). When ACF and ATP were present, the cleavage of both damaged and undamaged dinucleosomes at the 5′
Referência(s)