Nucleosome structure and positioning modulate nucleotide excision repair in the non-transcribed strand of an active gene
1997; Springer Nature; Volume: 16; Issue: 16 Linguagem: Inglês
10.1093/emboj/16.16.5046
ISSN1460-2075
Autores Tópico(s)DNA Repair Mechanisms
ResumoArticle15 August 1997free access Nucleosome structure and positioning modulate nucleotide excision repair in the non-transcribed strand of an active gene Ralf Erik Wellinger Ralf Erik Wellinger Institut für Zellbiologie, ETH, Hönggerberg, CH-8093 Zürich, Switzerland Search for more papers by this author Fritz Thoma Corresponding Author Fritz Thoma Institut für Zellbiologie, ETH, Hönggerberg, CH-8093 Zürich, Switzerland Search for more papers by this author Ralf Erik Wellinger Ralf Erik Wellinger Institut für Zellbiologie, ETH, Hönggerberg, CH-8093 Zürich, Switzerland Search for more papers by this author Fritz Thoma Corresponding Author Fritz Thoma Institut für Zellbiologie, ETH, Hönggerberg, CH-8093 Zürich, Switzerland Search for more papers by this author Author Information Ralf Erik Wellinger1 and Fritz Thoma 1 1Institut für Zellbiologie, ETH, Hönggerberg, CH-8093 Zürich, Switzerland *Corresponding author. E-mail: [email protected] The EMBO Journal (1997)16:5046-5056https://doi.org/10.1093/emboj/16.16.5046 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Nucleotide excision repair (NER) is a major pathway to remove pyrimidine dimers (PDs), a class of DNA lesions generated by ultraviolet light. Since folding of DNA into nucleosomes restricts its accessibility and since transcription and DNA repair require access to DNA, nucleosome structure and positioning as well as the transcriptional state may affect DNA repair. We recently determined the chromatin structure of the yeast URA3 gene at high resolution and found multiple positions of nucleosomes as well as strand- and site-specific variation in DNA accessibility to DNase I (internal protected regions). Here, the same high-resolution primer extension technique was used to investigate NER of PDs in the URA3 gene of a minichromosome in vivo. In the non-transcribed strand (NTS), fast repair correlates with PD locations in linker DNA and towards the 5′ end of a positioned nucleosome. Slow repair correlates with the internal protected region of the nucleosome. This repair heterogeneity reflects a modulation of NER by positioned nucleosomes in the NTS. NER in the transcribed strand (TS) is fast, less heterogeneous and shows no correlation with chromatin structure. Apparently, transcription-coupled repair overrides chromatin modulation of NER in the TS. Heterogeneity in NER generated by chromatin structure on the NTS may contribute to heterogeneity in mutagenesis. Introduction Since packaging DNA in nucleosomes and higher-order chromatin structures affects its accessibility to proteins, all DNA-processing reactions including transcription and DNA repair must be intimately coupled to and could even be regulated by structural and dynamic properties of chromatin. Indeed, nucleosomes positioned in promoter regions play a significant role in regulation of transcription, and promoter activation may require remodelling of nucleosome structures, generating a nuclease-sensitive region (NSR; reviewed in Wallrath et al., 1994). Transcription elongation through nucleosomal templates requires a disruption of nucleosomes or a displacement of the histones by RNA polymerase (reviewed in Kornberg and Lorch, 1995). This may lead to a loss of nucleosomes from the transcribed gene (e.g. ribosomal RNA genes, references in Dammann et al., 1993) or to a rearrangement of nucleosome positions as shown for a gene transcribed from a GAL1 promoter in yeast (Cavalli and Thoma, 1993; Cavalli et al., 1996). Less frequently transcribed yeast genes such as URA3 or HIS3 show positioned nucleosomes (Thoma, 1986; Losa et al., 1990; Tanaka et al., 1996). This structure most likely reflects the inactive, major fraction of the genes and suggests that the disruption of the chromatin structure by RNA polymerase II is rapidly reversed and does not require replication. Irradiation of DNA with ultraviolet light (UV) generates two major classes of pyrimidine dimers (PDs), cyclobutane pyrimidine dimers (CPDs) and pyrimidine (6-4) pyrimidone dimers [(6-4)-PD)]. Unless repaired they may lead to transcription blockage, mutations, cell death and cancer (reviewed in Friedberg et al., 1995). The formation of CPDs and (6-4)-PD is dependent on DNA sequence, protein–DNA interactions and chromatin structure (for reviews see Smerdon, 1989; Tornaletti and Pfeifer, 1996). CPD formation is strongly modulated in nucleosomes due to the bending of DNA around the histone surface (Gale et al., 1987; Pehrson, 1989; Pehrson and Cohen, 1992; Schieferstein and Thoma, 1996), while (6-4)-PD are more evenly distributed within nucleosome cores (Gale and Smerdon, 1990). Nucleotide excision repair (NER) is a ubiquitous multistep pathway in which numerous proteins are involved to sequentially execute damage recognition, excision of an oligonucleotide with the pyrimidine dimer, and gap repair synthesis (reviewed in Friedberg, 1996). In genes transcribed by RNA polymerase II, the transcribed strand (TS) is faster repaired than the non-transcribed strand (NTS) in higher eukaryotes (e.g. Mellon et al., 1987), bacteria (e.g. Mellon and Hanawalt, 1989) as well as in yeast minichromosomes and chromosomes (Smerdon and Thoma, 1990; Leadon and Lawrence, 1992; Sweder and Hanawalt, 1992; Verhage et al., 1994). This preferential repair of the TS is frequently referred to as 'transcription-coupled repair' (TCR). NER and initiation of transcription share proteins of the general transcription factor TFIIH. However, the molecular mechanism which leads to preferential repair of the TS remains to be elucidated (discussed in Friedberg, 1996). The NTS is repaired more slowly, similar to the genome overall. Mutations in the yeast genes (RAD7, RAD16) that affect the repair of the NTS also affect the repair of inactive genes and the genome overall (Verhage et al., 1994, 1996; Mueller and Smerdon, 1995). These genetic observations suggest that the NTS could be repaired as a chromatin substrate. The molecular mechanism of NER in chromatin is barely understood. Early studies in cultured human cells showed that repair synthesis occurs in a rapid early phase and in a slow late phase. At the nucleosome level there is an early phase that may represent refolding of nucleosomes and a prolonged slow phase that may be explained by rearrangement or repositioning of nucleosomes (reviewed in Smerdon, 1989, 1991). Repair synthesis during the early phase showed a bias towards the 5′ ends of nucleosome core DNA (Jensen and Smerdon, 1990). This repair effect was accounted for by the preferential formation of CPDs in the 5′ ends of core DNA (Smerdon, 1991). Those studies were performed on bulk chromatin containing mixed sequences and did not account for the transcriptional state of the chromatin nor did they address repair of positioned nucleosomes. To address directly the question of whether and how NER is modulated by transcription and chromatin structure, we used an indirect endlabelling technique to compare CPDs along the DNA sequence with the chromatin structure assayed by nuclease digestion (Smerdon and Thoma, 1990; Bedoyan et al., 1992). These studies were done in a yeast strain containing the minichromosome (YRpTRURAP) with characterized chromatin structure as a model substrate (Thoma, 1986) (Figure 1). They showed fast repair on the TS consistent with TCR, and slow repair in the NTS. Repair was also efficient in a nuclease-sensitive promoter region of the URA3 gene, but slow in the nuclease-sensitive origin of replication (ARS1). Those studies, however, were limited in resolution and could not resolve differences in repair of linker DNA and nucleosomes nor differences of repair within positioned nucleosomes. Figure 1.Chromatin structure of the minichromosome YRpTRURAP and the URA3 gene. (A) YRpTRURAP contains the URA3 gene inserted in the TRP1 gene of the TRP1ARS1 circle (Thoma, 1986). Indicated are: approximate positions of nucleosomes (circles), genes (arrows) and their 5′ and 3′ ends, location of primers (numbered arrows), the origin of replication (ARS1), EcoRI restriction site (R), dots in 200 bp intervals. (B) Detailed structure of the URA3 gene (Tanaka et al., 1996). Indicated are: approximate positions of nucleosomes (ellipses 1–6), internal regions protected against DNase I cutting on the top and bottom strand (bars in ellipses), multiple positions of nucleosomes (bars below nucleosomes) and position of linker DNA [numbers, in map units (base pairs from the EcoRI site)]. Download figure Download PowerPoint To study DNA repair within nucleosomes, high-resolution approaches are required for chromatin analysis and for DNA repair assays. We applied a primer extension technique (Axelrod and Majors, 1989) to characterize the chromatin structure of the URA3 gene at high resolution using MNase and DNase I digestions (Tanaka et al., 1996). The structure of the URA3 gene is similar in the minichromosome and in the genome. At low resolution, the URA3 gene showed six positioned nucleosomes flanked by NSRs (Figure 1B); Thoma, 1986). At high resolution, each position resolves into a complex pattern of multiple positions which suggests that nucleosome positions can change (nucleosome mobility). The high-resolution analysis provided important insight into the nucleosomal structure. Reduced DNase I cutting from ∼50 bp from the 5′ end towards the 3′ end was common to all nucleosome regions. This polarity reflects a structural property of nucleosomes (here referred to as 'internal protected region'). Since these 'internal protected regions' only partially overlap on both strands (Figure 1B), they demonstrate differential accessibility of DNA strands on the nucleosome surface. We think that these properties of nucleosome positioning and the internal structure of nucleosomes play an important role in DNA damage recognition and repair. Here, we have tested this hypothesis using the same primer extension protocol (Axelrod and Majors, 1989; Wellinger and Thoma, 1996) to analyse NER of PDs at high resolution and compared it with the local chromatin structure. We show that a pronounced repair heterogeneity on the NTS correlates with the positioning and the internal structure of nucleosomes. In contrast, repair on the TS is dominated by TCR. Results Efficient NER in a minichromosome FTY23 cells containing the minichromosome YRpTRURAP (Figure 1) were irradiated in suspension with 100 J/m2 and NER was allowed to occur for 0–50 min in the absence of photoreactivating light. To calculate CPD damage and repair on the whole minichromosome (overall repair), DNA was cut with EcoRI, treated with T4 endonuclease V (T4-endoV) which cleaves at CPDs, or mock treated, fractionated on alkaline agarose gels, blotted to nylon membranes and hybridized to a plasmid-specific probe (Figure 2). Unirradiated DNA (lane 2) and mock-treated samples (odd lanes 1–15) showed intact restriction fragments. In T4-endoV-cleaved samples (even lanes 4–16), the top band (region a) represents the fraction of undamaged DNA, while the smear (region b) shows the fraction of DNA cut at CPDs. With increasing repair time, the smear decreased and the proportion of intact fragment increased, demonstrating efficient repair. After irradiation with 100 J/m2, ∼48% of the DNA remained undamaged. This corresponds to an initial damage of 0.7 CPDs/plasmid or 0.27 CPDs/kb. After 2.5 h incubation at 30°C, ∼80% of the CPDs were repaired by NER (Figure 2B). Figure 2.Overall repair of CPDs in YRpTRURAP. (A) Yeast cells were irradiated with 100 J/m2 and repair was allowed for 0, 20, 40, 60, 90, 120 and 150 min in the dark (NER). DNA was extracted and cut with EcoRI and the fragments were separated by gel electrophoresis on alkaline agarose. Southern blot analysis shows the removal of CPDs after treatment of DNA with T4-endoV (T4-EndoV +, lanes 4, 6, 8, 10, 12, 14 and 16). Non-irradiated DNA was used as a control (lane 2). Mock-treated DNA is shown (T4-EndoV−, lanes 1, 3, 5, 7, 9, 11, 13 and 15). Undamaged and mock-treated DNA migrate on top of the gel (a), whereas DNA cut by T4-endoV at CPDs generates a smear (b). (B) The graph shows the time-dependent removal of CPDs. At time 0 min, 49% of plasmids were cut by T4-endoV which corresponds to an initial damage of 0.7 CPDs per plasmid. Data represent the average of three independent experiments. Download figure Download PowerPoint High-resolution repair analysis by primer extension Primer extension (Wellinger and Thoma, 1996) was used to characterize site-specific repair in distinct chromatin regions, namely in two nucleosomal regions of the coding part of URA3 and in the nuclease-sensitive promoter. EcoRI-linearized DNA was treated with T4-endoV to prevent premature stops of Taq polymerase at CPDs (Wellinger and Thoma, 1996). The DNA was denatured and annealed to an endlabelled primer which was extended by Taq polymerase. Taq polymerase is efficiently blocked at CPDs and (6-4)-PD (Wellinger and Thoma, 1996) producing bands which map at pyrimidine dimers or pyrimidine tracts (compare sequencing lanes with lanes of UV-irradiated samples; Figures 3, 5 and 6). In a few cases, Taq polymerase was blocked at sites containing CA (TACACA 1019, Figure 3A; TCA 1331, TCA 1340, Figure 5A), possibly by purine photoproducts (Gallagher and Duker, 1986; Bourre et al., 1987). The primer extension technique used here does not discriminate between CPDs and (6-4)-PD. However, since (6-4)-PD are generated in much lower yields than CPDs (Sage, 1993), the results of this study preferentially relate to CPD repair. Indeed, reversal of CPDs by photolyase prior to primer extension revealed only very weak signals originating from (6–4)-PD. These signals were too weak for a repair analysis (not shown). Non-irradiated DNA yields a very weak background. Stops of the Taq polymerase are observed towards the top of the lanes. Hence, accurate repair analysis can be done in the lower part of the gel, ∼25–250 bp from the primer. The intensities of bands reflect the frequency of lesions at individual sites or in polypyrimidine tracts. The signal in the whole lane represents the total amount of amplified DNA. Site-specific damage was calculated as the fraction of signal at a lesion divided by the signal of the whole lane after appropriate subtraction of background. After irradiation of cells with 100 J/m2, the fraction of DNA damaged at individual sites scattered between 0.001% for weak bands (e.g. CCT1079) and 0.1% for strong bands (e.g TTTATC 902) (Figure 3A). Figure 3.Repair of PDs in the region of nucleosome U2 displayed by primer extension. (A) Top strand (non-transcribed strand, NTS) using primer No. 897 (Figure 1). (B) Bottom strand (transcribed strand, TS) using primer No. 896. Indicated are: nucleosome positions (ovals U2, U1, U3) with internal protected regions (bars) according to Tanaka et al. (1996); repair times (min, 0 to 150); T4-endoV treatment of DNA (+); UV irradiation conditions for DNA (40 J/m2, lane 5) and for cells (chromatin; 100 J/m2, lanes 7–13); DNA sequence of the template strand (A, G, C, T); sequence elements with DNA lesions that were quantified (dark boxes, numbers refer to the 5′ nucleotide in the YRpTRURAP sequence); sequence elements with DNA lesions that were not quantified due to background or low signal intensities (white boxes); strong background signals (stars). Download figure Download PowerPoint Figure 4.Repair curves for lesions in the top strand (Figure 3A) and bottom strand (Figure 3B) of the U2 region respectively. DNA damage for each cluster was set to 100% at time 0 [lane 6 in (A) and (B)]. The average and standard deviation of three experiments (circles) are shown. In a few cases data of two experiments could be used (triangles). The curves represent exponential fits. Download figure Download PowerPoint Figure 5.Repair of PDs in the region of nucleosome U4. (A) Top strand (non-transcribed strand, NTS) using primer No. 1119. (B) Bottom strand (transcribed strand, TS) using primer No. 1118. Description is as in Figure 3. Download figure Download PowerPoint Figure 6.Repair of PDs in the URA3 promoter region. (A) Top strand (non-transcribed strand, NTS) using primer No. 1014. (B) Bottom strand (transcribed strand, TS) using primer No. 1027. Download figure Download PowerPoint DNA repair was obvious as the decrease in band intensities with increasing repair time (Figures 3, 5 and 6). There is, however, substantial strand- and site-specific heterogeneity (see below). Some sites appear to be repaired quickly, while others are much more slowly repaired. To generate repair curves, the damage at each site and at each repair time was measured and normalized with respect to the initial damage (initial damage at time 0 min was set as 100%). Repair curves were calculated for DNA lesions in pyrimidine clusters, rather than for individual dipyrimidines. This is justified, since CPDs in individual clusters were generally repaired with similar rates. The data of three independent repair experiments were averaged and displayed as exponential curves (the curves are similar in all regions and therefore shown for the U2 region only, Figure 4). The t50%, the time at which 50% of the lesions were repaired, was graphically summarized and compared with the local chromatin structure as previously determined (Tanaka et al., 1996) (Figure 7). It is pointed out that quantification of lesions as a fraction of the total DNA in a lane corrects for different amounts of DNA used in different lanes (e.g. lanes 10 and 11 in Figure 5) and allows one to obtain accurate repair data. Moreover, visual inspection of bands may suggest differential repair, but quantification yields similar repair rates (e.g. compare TTTT 996, TTTT 1010, Figures 3B and 4B). This is due to different yields of damage in TTTT 996 (0.016%) and TTTTT 1010 (0.012%). Finally, a number of sites were excluded from the quantitative analysis (open boxes in Figures 3, 5 and 6), since the background was too high or the data too scattered. Figure 7.Summary of repair rates and correlation between DNA-repair and chromatin structure. (A) The URA3 promoter. (B) Nucleosome U2. (C) Nucleosome U4. The t50% for each PD cluster is shown as vertical bars for the top and bottom strands respectively. Sites with no repair are shown as bars with dotted ends (this includes two sites possibly containing purine photoproducts, TCA1331, TCA1340). Indicated are the major nucleosome positions (ovals) including the internal protected regions (bars in ovals), promoter elements (white boxes) including the TATA element (dark box). Data are averages of three (B) and two experiments (A and C). Download figure Download PowerPoint Repair in the transcribed region To address how repair occurs in the transcribed region and whether it correlates with chromatin structure or transcription, PD removal was mapped in and around the second and the fourth nucleosomes (U2 and U4) (Figure 1). Nucleosome U2 is best suited for this analysis. It is positioned within a few base pairs, it shows clear internal structure ('internal protected regions') and it contains sufficient pyrimidine dimers and pyrimidine tracts distributed over the whole region on both strands (Figure 3). U4 is more complex, since it is tightly packed together with U3 and U5, and it may occupy multiple positions (Figure 1B). Nucleosomes U3, U5 and U6 are less suited due to the pyrimidine dimer distribution and multiple positioning. Nucleosome U1 lacks pyrimidine tracts on the top strand (Figures 6 and 7 Fast repair with low heterogeneity on the transcribed strand In general, repair was much faster on the TS (bottom strand) than on the NTS (top strand). This is obvious from visual inspection of the autoradiographs (Figures 3 and 5), from repair curves (Figure 4) and from the t50% values (Figure 7). The average repair of 33 lesions measured on the TS in U2, U4 and in U1 showed a t50% of 32 ± 8 min. In contrast, the average repair of 21 lesions in the NTS showed a t50% of 80 ± 41 min [lesions with no repair (Figure 7C) are not included]. Hence, repair in the TS was ∼2.5-fold faster. This result is consistent with TCR of the TS. Moreover, average repair in the TS in the U2 and U4 regions is 30 ± 6 min and 31 ± 8 min respectively. Hence, there is no measurable gradient in repair efficiency along the URA3 gene. However, a detailed inspection at individual sites and clusters on the TS reveals a moderate heterogeneity of repair rates (Figures 4B and 7). For example, in U2, repair of lesions in CTTTTC 915 (t50% = 23 min) was ∼2-fold faster than repair of lesions in TTTTT 1010 (t50% = 43 min). In U4, repair of lesions in TTTCCTT 1259 (t50% = 19 min) was ∼2-fold faster than repair in CCTCTT 1320 (t50% = 38 min). This modulation in the repair rates does not obviously correlate with chromatin structure; in particular, it does not show slow repair in nucleosomes and fast repair in linker DNA nor does it show remarkably slower repair in the 'internal protected region' of U2 and U4 (Figure 7B and C). Hence, the local differences in repair rates could be related to a sequence specificity of the repair process. Repair heterogeneity on the non-transcribed strand correlates with chromatin structure In contrast to the TS, a much more pronounced repair heterogeneity was observed on the NTS (top strand, Figures 3A, 4A and 7). For example, repair was fast in TTC 931 (t50% = 30 min), CCT 1079 (t50% = 32 min), TTTATC 902 (t50% = 56 min) and CTTC 940 (t50% = 57 min). Slow repair was observed, e.g. in CCTTTT 1042 (t50% = 103 min). A comparison with the chromatin structure (Figure 7B) shows that this repair heterogeneity correlates well with chromatin structure according to two criteria. First, repair was fast in sites that map in linker DNA, e.g. in linkers between U1 and U2 (Figure 3, site TTTATC 902), and between U2 and U3 (Figure 3, sites TT 1075, CCT 1079). Second, in U2, repair correlates well with the internal structure described by the accessibility of DNA to DNase I. Fast repair was observed towards the 5′ end (sites between TTC 931 and CTT 951, Figures 3 and 7B) where DNase I accessibility is high. Slow repair was recorded in the region towards the 3′ end (sites between TT 966 and CCTTTTT 1042, including a purine–pyrimidine region TACACA 1019), where DNase I cutting is inhibited ('internal protection', bar in Figures 3A and 7B). In the NTS (top strand) of the nucleosome U4 region, the correlation between DNA repair (t50%) and chromatin structure was more complex, since nucleosomes U3, U4 and U5 are tightly packed leaving only limited space for linker DNA (Figure 1B). Despite that, lesions observed in the top strand of U4 (sites between TTTTCC 1279 and TT 1373 including purine photoproducts TCA 1331 and TCA 1340) are more slowly repaired than TTTC 1382, TT 1385 and CTC 1389 which map in the contact region of U4 and U5. PDs in the contact region between U3 and U4 (TTGTT 1226) are relatively quickly repaired also (Figures 5A and 7C). In summary, these results show that repair on the NTS is modulated by chromatin structure, in particular by the internal structure of nucleosomes and nucleosome positioning. Fast repair in the URA3 promoter region The 5′ end of the URA3 gene is characterized by a nuclease-sensitive promoter region and a nucleosome (U1) which may occupy two major and possibly some minor positions (Figures 1 and 7A) (Thoma, 1986; Tanaka et al., 1996). Hence, it is possible that DNA lesions in the NSR could be recognized and repaired more efficiently. The promoter region contains, on both strands, several pyrimidine clusters and in particular long T-tracts on the top strand (Figure 6) (Roy et al., 1990). These T-tracts are hotspots of PD formation (Figure 6A). All pyrimidine clusters were repaired with t50% between 30 and 60 min. In contrast to the transcribed region, no obvious differences in repair rates were observed in the top and bottom strand of the promoter region. The repair rates were faster than the average rates in the NTS of the URA3 gene (t50% = 80 min), which is consistent with an enhanced accessibilty of lesions in a NSR. There is a moderate heterogeneity of repair rates on both strands. Previous reports showed enhanced repair rates close to the transcription initiation site of the human JUN gene, a property which could be explained by the enhanced concentration of factors (TFIIH) which are involved in repair and transcription initiation (Tu et al., 1996). However, in the URA3 gene, there are no dramatic DNA repair differences between sites upstream and downstream of the transcription initiation (Figure 7A). Discussion Chromatin modulates repair in the non-transcribed strand but not in the transcribed strand Previous repair studies using indirect endlabelling were restricted to a resolution of approximately ±20 bp and, hence, did not resolve differential repair in linker DNA and within nucleosomes (Smerdon and Thoma, 1990; Bedoyan et al., 1992; Mueller and Smerdon, 1995, 1996). The primer extension approach used in this work allowed us to determine the location of a lesion at the sequence level and to correlate it with the high-resolution chromatin structure as determined by nuclease digestion (Tanaka et al., 1996). As shown for the U1, U2 and U4 region, this correlation is limited by the pyrimidine distribution in the DNA sequence as well as by the precision and dynamics of nucleosome positioning. The results on the NTS of the U2 region show a strong correlation between the rate by which pyrimidine dimers are removed and the accessibility of the sequence to micrococcal nuclease and DNase I (Figure 7). Lesions which map in linker DNA and towards the 5′ end of nucleosomes were repaired more quickly than lesions that map in the 'internal protected region' of the nucleosome. This correlation holds for the U4 region too, although the nucleosome arrangement is more complex with respect to the tight packaging of nucleosomes U3, U4 and U5 and with respect to multiple positions (Thoma, 1986; Tanaka et al., 1996). Lesions which map in the U3/U4 and U4/U5 boundary are repaired more quickly than lesions which map within nucleosome U4. In contrast, repair on the TS does not correlate with chromatin structure and shows fast repair and moderate repair heterogeneity. Hence, these data strongly suggest that NER on the NTS is modulated by chromatin structure. To our knowledge, this is the first evidence which shows a direct correlation between chromatin structure and DNA repair in the NTS, as well as the first set of data which correlates repair with the internal structure of a positioned nucleosome. To understand why only repair in the NTS is modulated by chromatin structure, we need to consider the transcriptional properties of the URA3 gene. Transcription disturbs chromatin structure (e.g. Thoma, 1991; Felsenfeld, 1996). Recent studies of a gene heavily transcribed from the yeast GAL1 promoter showed that nucleosomes were not lost, but nucleosome positioning was disturbed. This was consistent with a local disruption of nucleosomes at the site of the RNA polymerase and their rapid reformation behind the polymerase. After gene repression by glucose, nucleosomes were repositioned within a few minutes, demonstrating that nucleosome rearrangement can occur rapidly and without DNA replication (Cavalli and Thoma, 1993; Cavalli et al., 1996). Compared with transcription driven from the GAL1 promoter, the URA3 gene in the multicopy YRpTRURAP is rarely transcribed (Smerdon and Thoma, 1990; Bedoyan et al., 1992) and therefore most of the time transcriptionally silent. The presence of positioned nucleosomes (Thoma, 1986; Tanaka et al., 1996) means that this chromatin structure most likely represents the major fraction of inactive genes and that this structure might be only transiently disturbed by transcription. Consequently, modulation of NER in the NTS reflects NER in the inactive gene. In contrast to the NTS, we found no correlation between NER heterogeneity and chromatin structure in the TS. Current views on NER of transcribed genes agree that RNA polymerase II is blocked at pyrimidine dimers and somehow promotes assembly of the repair complex which results in preferential repair of the TS (TCR, reviewed in Wood, 1996; Friedberg, 1996; Sancar, 1996a). Hence, our data suggest that TCR in the TS overrides modulation of NER by chromatin structure. DNA repair in nucleosomes The central topic of this work is the repair of lesions in nucleosomal DNA. In the nucleosome U2 core, DNA repair seems to slow down gradually from the 5′ end towards the region of the 'DNase I protection'. This is the first time that such a correlation could be made in a positioned nucleosome in vivo. However, similar chromatin and repair studies on other precisely positioned nucleosomes will be required to establish whether the repair properties observed in U2 can be generalized. A previous study analysed the removal of CPDs in mixed sequence genomic nucleosome cores, which were isolated from human cultured fibroblasts after different repair times. In contrast to our observation, CPDs were removed at equal rates from nucleosome core subdomains with no indication for preferential repair towards the 5′ end (Jensen and Smerdon, 1990). One explanation could be that nucleosomes and, hence, damage distribution might rearrange as a consequence of the nucleosome isolation procedure. Alternatively, due to the small linker length and small gene size in yeast, yeast nucleosom
Referência(s)