- Email: [email protected]
Blockage of RNA polymerase II at a cyclobutane pyrimidine dimer and 6–4 photoproduct
BBRC Biochemical and Biophysical Research Communications 320 (2004) 1133–1138 www.elsevier.com/locate/ybbrc
Blockage of RNA polymerase II at a cyclobutane pyrimidine dimer and 6–4 photoproductq Joan Seah Mei Kwei,a,1 Isao Kuraoka,a,b,1 Katsuyoshi Horibata,a Manabu Ubukata,d Eiry Kobatake,d Shigenori Iwai,c Hiroshi Handa,d and Kiyoji Tanakaa,b,* a Graduate School of Frontier Biosciences, Osaka University, 1-3 Yamadaoka, Suita, Osaka 565-0871, Japan Core Research for Evolutional Science and Technology (CREST), Japan Science and Technology Corporation, 1-3 Yamadaoka, Suita, Osaka 565-0871, Japan c Graduate School of Engineering Science, Osaka University, 1-3 Machikanemaya, Toyonaka, Osaka 560-8531, Japan Faculty of Bioscience and Biotechnology and Frontier Collaborative Research Center, Tokyo Institute of Technology, Yokohama 226-8501, Japan b
d
Received 21 May 2004 Available online 25 June 2004
Abstract The blockage of transcription elongation by RNA polymerase II (pol II) at a DNA damage site on the transcribed strand triggers a transcription-coupled DNA repair (TCR), which rapidly removes DNA damage on the transcribed strand of the expressed gene and allows the resumption of transcription. To analyze the effect of UV-induced DNA damage on transcription elongation, an in vitro transcription elongation system using pol II and oligo(dC)-tailed templates containing a cyclobutane pyrimidine dimer (CPD) or 6–4 photoproduct (6–4PP) at a specific site was employed. The results showed that pol II incorporated nucleotides opposite the CPD and 6–4PP and then stalled. Pol II formed a stable ternary complex consisting of pol II, the DNA damage template, and the nascent transcript. Furthermore, atomic force microscopy imaging revealed that pol II stalled at the damaged region. These findings may provide the basis for analysis of the initiation step of TCR. Ó 2004 Elsevier Inc. All rights reserved. Keywords: RNA polymerase II; Cyclobutane pyrimidine dimer; 6–4 Photoproduct; Transcription-coupled DNA repair; Transcription elongation
Genomic DNA as carrier of genetic information in living cells is fundamentally vulnerable to ubiquitous DNA-damaging agents of endogenous and environmental origins [1]. Nucleotide excision repair (NER) eliminates a wide variety of bulky helix-distorting DNA lesions such as ultraviolet light (UV)-induced photolesions: cyclobutane pyrimidine dimers (CPD) and 6–4 photoproducts (6–4PP). NER can operate via two pathways: global genome repair (GGR) and transcription-coupled repair (TCR). GGR can repair DNA leq Abbreviations: NER, nucleotide excision repair; GGR, global genome repair; TCR, transcription-coupled DNA repair; pol II, RNA polymerase II; CPD, cyclobutane pyrimidine dimer; 6–4PP, 6–4 photoproduct; XP, xeroderma pigmentosum; CS, Cockayne syndrome; NTP, ribonucleoside triphosphate; RT, reverse transcriptase. * Corresponding author. Fax: +81-6-6877-9136. E-mail address: [email protected] (K. Tanaka). 1 Both authors contributed equally to this work.
0006-291X/$ - see front matter Ó 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2004.06.066
sions at any location in the whole genome, while TCR removes DNA lesions on the transcribed strands of expressed genes [2]. Xeroderma pigmentosum (XP) is an autosomal recessive disease characterized by a high incidence of skin cancer on sun-exposed areas while Cockayne syndrome (CS) is characterized by photosensitivity and several neuro-developmental abnormalities without a predisposition to skin cancer. So far, eight complementation groups have been identified in XP (XP-A through XP-G and XP-V), and two in CS (CS-A and CS-B). XP-A through XP-G have a defect in both NER pathways except for XP-C and XP-E, which have a defect in GGR alone, and XP-V, which has a proficient NER but is defective in translesion DNA synthesis. On the other hand, CS is defective in TCR alone. All the genes responsible for XP and CS have been cloned. The XPC protein that complexes with HR23B and the UV-DDB
1134
J.S.M. Kwei et al. / Biochemical and Biophysical Research Communications 320 (2004) 1133–1138
complex that comprises DDB1 and DDB2 proteins preferentially bind to various types of DNA damage and are involved in GGR-specific damage recognition. In TCR, elongating RNA polymerase II (pol II) that encounters a lesion is thought to play the corresponding role of XPC/HR23B and UV-DDB. For each subpathway-specific damage recognition event, NER factors including TFIIH, XPA, RPA, XPF/ERCC1, and XPG may be recruited to the lesion [3–5]. Although 6–4PP is less frequently produced than CPD, 6–4PP is more cytotoxic and mutagenic. XPCHR23B complex preferentially binds to 6–4PP rather than CPD, and GGR repairs 6–4PP more rapidly than CPD. In contrast, TCR can repair CPD and 6–4PP at the same rate in an XPC-independent manner [6]. Here, we showed that (1) pol II itself completely stalls at both CPD and 6–4PP lesions, that (2) RNA synthesis by pol II is stopped by incorporating one or two nucleotide(s) opposite the lesion, and that (3) pol II forms ternary complex with the nascent transcripts around the lesion. Our present data indicate that the stalling of pol II itself at the damage site might be a trigger for TCR and provide the reasons why it might recognize the lesion. Methods Enzymes. The preparation of human RNA polymerase II (pol II) was previously described [7]. Bacteriophage T4 pyrimidine dimer DNA glycosylase (T4 endonuclease V) and ultraviolet damage endonuclease (UVDE) of Schizosaccharomyces pombe were purchased from Trevigen. DNA template with single UV-damage at a defined position. To prepare a template with single DNA damage for transcription elongation, 30-mer oligo-deoxyribonucleotides containing a cyclobutane
pyrimidine dimer (CPD) or a pyrimidine–pyrimidone 6–4 photoproduct (6–4PP) were incorporated into covalently closed circular DNA as previously described [8]. The plasmid pBluescript II KS-UV (pBSII KS-UV) was constructed by modifying pBSII KS-GTG [7] with 30 synthetic oligonucleotides containing the damage (Fig. 1A). In vitro transcription elongation reactions using Pol II and an oligo(dC)-tailed template. For the in vitro transcription elongation assay [7], 20 ll reaction mixtures containing 50 ng of oligo(dC)-tailed template and 0.5 ll pol II were preincubated for 30 min at 30 °C. Elongation was started by adding 5 ll NTP and terminated by adding 100 ll stop buffer. The purified transcripts were resuspended in formamide loading dye and separated on a denatured 6% polyacrylamide gel. The dried gels were analyzed using a FUJIFILM BAS 2500 bioimage analyzer. Sequence analysis of nascent transcripts amplified by 30 -ligationmediated RT-PCR. For sequence analysis, the stalled transcripts were extracted from the elongation reaction mixture in the presence of a cold NTP mixture (50 lM each of ATP, CTP, GTP, and UTP) using a RNA purification kit (Qiagen) and then treated with T4 RNA ligase (TaKaRa) as recommended by the supplier in the presence of RNA primers (50 -p-CACCAAACGTGGCUUGCCAGCCC-NH2 -30 ). The ligated transcripts were used to synthesize PCR products with a One Step RT-PCR kit (Qiagen) using the primers (704–724) 50 GGAATTCGATATCAAGCTTA-30 and (anti-RNA primer) 50 GGGCTGGCAAGCCACGTTTGGTG-30 . PCR was performed again using anti-RNA linkers and the primer (751–770) 50 GCCCTGCTGCCATGCGCGG-30 . The PCR products were isolated with a TOPO TA Cloning kit (Invitrogen) and then sequenced using a BigDye Terminator kit (Applied Biosystems) and the PCR primer (751–770). Immunoprecipitation of pol II-ternary complex. Elongating pol II was enriched from mixtures of the transcription elongation reaction by immunoprecipitation using anti pol II antibody (C21)-conjugated agarose in the presence of cold NTP. The samples were washed five times with a buffer [20 mM Tris–HCl (pH 8), 500 mM NaCl, 2 mM EDTA, 1% Triton X-100, and 0.1% SDS] and resuspended with restriction buffer [10 mM Tris–HCl (pH 7.5), 10 mM MgCl2 , 50 mM NaCl, and 1 mM dithiothreitol] for digestion with EcoRV and PvuII. After the digestion, elongating pol II was enriched again by immunoprecipitation with anti-pol II antibody beads. PCR was then carried
Fig. 1. Preparation of DNA template with single UV-damage at a defined position. (A) Covalently closed circular duplex DNA containing a single lesion. Thirty-mer oligonucleotides containing a single DNA lesion and the plasmid pBSII KS-UV are shown diagrammatically. The TT dimer site is indicated as D. Unique PstI and SmaI restriction enzyme sites are also indicated. (B) One percentage of agarose gel electrophoresis of pBSIIKS-UV after digestion of each plasmid with T4 endonuclease V (T4 endo), demonstrating the presence of a single CPD site. (C) One percentage agarose gel after digestion of each plasmid with ultraviolet damage endonuclease (UVDE), demonstrating the presence of CPD and 6–4PP sites. The mobility of covalently closed circular (ccc) and nicked circular (nc) pBSII KS-UV is indicated by the arrows.
J.S.M. Kwei et al. / Biochemical and Biophysical Research Communications 320 (2004) 1133–1138 out with primers (751–770) and (1009–990) 50 -ACTCATTAGGC ACCCCAGGC-30 . In the case of RT-PCR, the immunoprecipitated samples were treated with RNase-free DNase I (TaKaRa) instead of the restriction digestion step. The RNA was purified by a RNA purification kit (Qiagen), and then 30 -ligation-mediated RT-PCR was performed. The PCR products (259 bp) and RT-PCR products (109 bp) were analyzed by electrophoresis on a 4% agarose gel. AFM imaging. For AFM (atomic force microscopy) imaging, 10 ll reaction mixtures containing 10 ng of oligo(dC)-tailed template and 0.5 ll pol II in a buffer [10 mM Hepes–KOH (pH 7.9), 50 mM KCl, 10% glycerol, 0.1 mM EDTA, 0.25 mM dithiothreitol, 6 mM MgCl2 , 50 lM NTP mixture, and 8 U RNase Inhibitor (Promega)] were incubated for 30 min at room temperature. The sample was deposited on the center of a freshly cleaved mica disk. After 1 min, the mica was washed with 100 ll distilled water and dried. AFM studies were performed using a NanoScope III system (Veeco Instruments).
Results Construction of DNA templates containing either CPD or 6–4PP at a defined position on the transcribed strand To verify that the DNA template contained a lesion, 3060-bp pBSII KS-UV (Fig. 1A) was incubated with either T4 endonuclease V or UVDE. Since T4 endonuclease V is specific for CPD, 97% of the DNA template containing CPD (CPD template) was cleaved to produce
1135
an open circular form (Fig. 1B, lane 4), but the DNA template with 6–4PP (6–4PP template) and the one without damage (control template) were not cleaved (Fig. 1B, lanes 2 and 6). However, 97% of the CPD template and 95% of the 6–4PP template were cleaved by UVDE, which is specific for both photoproducts (Fig. 1C, lanes 4 and 6), while the control template was not cleaved (Fig. 1C, lane 2). Pol II stalled at CPD and 6–4PP sites When pol II was incubated with control templates, it synthesized elongation transcripts, but did not do so in the presence of a-amanitin (Fig. 2A, lanes 1 and 2). When CPD or 6–4PP template was used, pol II produced a transcript of about 130 nt in size. This corresponded to the region from the transcription start site to the DNA damage site (Fig. 2A, lanes 4 and 6), suggesting that pol II stalled at the CPD and 6–4PP sites. To confirm that pol II stalled at both lesions, time course experiments on the transcription elongation reaction were carried out (Fig. 2B). Pol II produced elongation transcripts from the control template (Fig. 2B, lanes 1–3). In the case of CPD or 6–4PP template, pol II produced transcripts of approximately
Fig. 2. Transcription elongation by pol II using an oligo(dC)-tailed template containing a single TT dimer. (A) Autoradiograph after denatured 6% polyacrylamide gel electrophoresis of the nascent transcripts derived from a transcription elongation reaction in the presence or absence of a-amanitin (a-ama). Transcripts from the templates containing no lesions (lanes 1 and 2), CPD (lanes 3 and 4), and 6–4PP (lanes 5 and 6) lesions in the presence (lanes 1, 3, and 5) or absence of a-amanitin (lanes 2, 4, and 6). (B) Time course (2–16 min) reaction of transcription elongation with templates containing no lesion (lanes 1–3), CPD (lanes 4–7) or 6–4PP (lanes 8–11) lesions. Stalled and elongation transcripts are indicated by the arrow and bracket, respectively.
1136
J.S.M. Kwei et al. / Biochemical and Biophysical Research Communications 320 (2004) 1133–1138
130 nt in size and elongation transcripts in a time-dependent manner (Fig. 2, lanes 4–7 and lanes 8–11). To examine whether pol II can bypass the lesions, the stalled transcripts (130 nt) and elongation transcripts were quantified. The results showed that the amount of elongation transcripts was 3.1% and 5.3% of the total transcripts on the CDP and 6–4PP template, respectively (data not shown). Sequencing of the elongation transcripts derived from the CPD and 6–4PP templates revealed no mutations (data not shown). Since the purity of the CDP and 6–4PP templates was 97% and 95%, respectively, this suggests that the elongation transcripts were produced from undamaged templates contaminated in the CPD and 6–4PP templates and that pol II completely stalled at the CPD and 6–4PP sites. Two other RNA polymerases (Escherichia coli and T7 RNA polymerase) stalled at the CPD and 6–4PP sites as well (data not shown). Analysis of 30 end stalled transcripts on CPD and 6–4PP templates To examine whether pol II has an ability to incorporate nucleotides when it encounters a lesion, the stalled transcripts were isolated and sequenced. We designed 50 phosphate RNA linkers that can be ligated to the 30 -OH ends of the stalled transcripts in the presence of T4 RNA ligase. Using a complementary primer that anneals to the RNA linkers, RT-PCR, and nested PCR were performed. This step helped to eliminate the pos-
sibility of cloning natural pausing products and other transcripts. The RT-PCR products were subcloned into the TA-cloning plasmid. An outline of this strategy is shown in Fig. 3A. The sequencing data (Fig. 3B) indicated that pol II incorporated nucleotide(s) opposite the lesion on the CPD and 6–4PP templates at frequencies of 92.5% and 86.8%, respectively. The frequency of misincorporation (rCMP and UMP) opposite the CPD site was 57.5%, while most of the nucleotide(s) incorporated opposite the 6–4PP site were correct. These findings suggest that the promiscuous behavior of pol II still fails to bypass these strong blocks even when ideal incorporation occurs. Ternary complex of pol II/transcripts/DNA template Stalled pol II forms a stable ternary complex of pol II/transcripts/DNA template in in vitro transcription reactions using general transcription factors and the adenovirus major late promoter [9–11]. We examined whether pol II can form a ternary complex at the DNA damage site in our transcription assay as well. As shown in Fig. 4A, the PCR products (259 bp) located around the CPD and 6–4PP sites were amplified in the anti-pol II immunoprecipitants of the transcription reaction mixtures including CPD and 6–4PP templates, but not in the immunoprecipitants including control template. In addition, RT-PCR products of the stalled transcript were amplified only in the anti-pol II immunoprecipitants of the transcription reaction mixtures including
Fig. 3. Analysis of 30 end of stalled transcripts on CPD and 6–4PP templates. (A) Flowchart showing how the stalled transcripts were isolated, ligated with RNA linker at the 30 end, amplified by RT-PCR and nested PCR, and subcloned into plasmid for DNA sequencing. (B) Nucleotide sequences of the 30 end of stalled transcripts opposite the CPD and 6–4PP sites. The frequency of stalled transcripts containing each type of the 30 end was expressed as the colony number of the TA-cloning plasmids and their percentages.
J.S.M. Kwei et al. / Biochemical and Biophysical Research Communications 320 (2004) 1133–1138
1137
Fig. 4. Stable ternary complex of stalled Pol II/transcripts/DNA template. (A) Four percentage agarose gel electrophoresis of PCR products derived from the in vitro transcription reaction mixture with UV-damage or control DNA template that was anti-pol II coimmunoprecipitated (after IP) or nonimmunoprecipitated (before IP). The PCR products (259 bp) are indicated by the arrow. (B) Four percentage agarose gel electrophoresis of the 30 ligation-mediated RT-PCR products derived from stalled transcripts that were coimmunoprecipitated by anti-pol II. RT-PCR products (109 bp) with (+) or without ()) reverse transcriptase (RT) are indicated by the arrow. (C–E) Atomic force microscopy (AFM) image of a field of molecules deposited from the pol II reaction mixture. Each panel indicates the protein bound to DNA. Pol II bound at the end of the control DNA (C), bound at the expected site of CPD (D), and 6–4PP (E). The scan area is 1lm 1lm for the upper panels, and the lower panels indicate the enlarged images (0.25 lm 0.25 lm). The height (0–5.0 nm) is indicated along the side of the (E). Stalled pol II is indicated by the arrow.
DNA damaged templates (Fig. 4B). These results again suggested that pol II stalled at the DNA damage sites with nascent transcripts. AFM imaging of the stalled pol II To visualize the pol II stalled at DNA damage site, AFM (atomic force microscopy) was employed. The in vitro transcription elongation reaction using pol II and oligo(dC)-tailed template with or without UV damage was performed and then deposited onto mica for AFM imaging. Although for our experimental AFM imaging conditions, transcripts and oligo(dC) tails were not detected. However, pol II (0.5 MDa) and double stranded templates (1 lm) were visualized. The AFM data revealed that pol II bound to the end of the control template, implying that the pol II completely reads through the template (Fig. 4C) or nonspecifically binds to its end. In the case of CPD and 6–4PP templates (Figs. 4D and E), pol II bound to the middle of DNA. The position of pol II along the DNA was measured. The length of the DNA from the pol II to the nearest end was divided by the full-length of the DNA template. According to the proportion of the DNA fragment length, it was estimated that pol II approximately bound at the damaged sites which were about 130 nt away from the nearest end of the DNA template. Pol II also bound to the other end of the CPD template. This binding may have been nonspecific due to our experimental conditions.
Discussion Our present study indicates that pol II completely stalled at the CPD and 6–4PP sites. Mammalian RNA
polymerase II stall at bulky adducts such as CPD, aminofluorene, acetylaminofluorene, and cisplatin in an in vitro transcription elongation system using the adenovirus major late promoter and general transcriptional factors [10,12]. We only used pol II and an oligo(dC)-tailed template for transcription elongation, indicating that pol II itself can recognize the damaged site where it stalls. Although TCR and GGR contribute to the removal of CPD and 6–4PP, the efficiency of removal varies. GGR can remove 6–4PP from the genome much faster than CPD in vivo. Consistent with this, XPC-HR23B complex, which is involved in the damage recognition step of GGR, has a higher affinity for 6–4PP than CPD in vitro. On the other hand, TCR can remove both lesions with equal efficiency in vivo. Current models indicate that TCR is triggered by stalled RNA pol II. Damage recognition by TCR might be dependent on how pol II stalls at the lesion. Thus, our finding that pol II stalled at CPD and 6–4PP lesions and forms a stable ternary complex in similar manner may account for the same efficiency of the TCR of CPD and 6–4PP in vivo. Sequencing analysis of the stalled transcripts revealed that pol II stalled after incorporating correct or incorrect nucleotide(s) opposite the CPD and 6–4PP sites. We also showed that pol II forms a stable ternary complex with the DNA lesion and stalled transcript. It was thought that pol II stalls when the 30 -OH end of the nascent transcript loses its ideal base pair contacts with the DNA template and hence is extruded from the pol II active site [13,14]. Previously, we reported that pol II stalled at 8-oxoguanine (8-oxoG), which is neither a bulky nor a distorting oxidative lesion, only when pol II makes misincorporation with 8-oxoG [7]. Considering this, we hypothesize that the incorporation of nucleotides opposite UV lesions at the 30 -OH end of a nascent
1138
J.S.M. Kwei et al. / Biochemical and Biophysical Research Communications 320 (2004) 1133–1138
transcript, even if the pairing is correct, may cause instability in the base pair contact at the lesions, leading pol II to stall. We showed that the stalled pol II securely attaches to the DNA damage site and the stable ternary complex is formed at the damaged site. Therefore, the lesion is most probably covered with pol II. It remains unresolved how the DNA repair machinery gains access into the lesion. However, the incubation of transcriptional elongation factor SII with the pol II ternary complex stalled at CPD produced a 10–30-nt cleavage profile of nascent transcripts [11,15], suggesting that SII enables stalled pol II to backtrack and make an opening for DNA repair proteins. Thus, we believe that the SII function is important for the initial step of TCR. In the normal transcription, SII stimulated the intrinsic nuclease activity of pol II and the misincorporated nucleotides are removed from nascent RNA to allow the resumption of transcription elongation [13,14]. It is thought that SII is recruited to pol II ternary complexes by naturally occurring misincorporation of pol II. Taken together, it is suggested that the promiscuous incorporation of pol II that we observed at UV damaged sites might recruit SII to the stalling pol II complexes at the damage sites in TCR.
Acknowledgments We thank Mika Hayashida for her technical assistance. J.S.M.K was the recipient of an Advanced Biotechnology Scholarship from Ishihara Sangyo Kaisha, Ltd., in collaboration with the Singapore Economic Development Board. This work was supported by a Grantin-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan, and a Core Research for Evolutional Science and Technology (CREST) of Japan Science and Technology Corporation.
References [1] T. Lindahl, R.D. Wood, Quality control by DNA repair, Science 286 (1999) 1897–1905.
[2] P.C. Hanawalt, D.J. Crowley, J.M. Ford, A.K. Ganesan, D.R. Lloyd, T. Nouspikel, C.A. Smith, G. Spivak, S. Tornaletti, Regulation of nucleotide excision repair in bacteria and mammalian cells, Cold Spring Harb. Symp. Quant. Biol. 65 (2000) 183– 191. [3] M.A. Nance, S.A. Berry, Cockayne syndrome: review of 140 cases, Am. J. Med. Genet. 42 (1992) 68–84. [4] K. Sugasawa, T. Okamoto, Y. Shimizu, C. Masutani, S. Iwai, F. Hanaoka, A multistep damage recognition mechanism for global genomic nucleotide excision repair, Genes Dev. 15 (2001) 507–521. [5] J.Q. Svejstrup, Mechanisms of transcription-coupled DNA repair, Nat. Rev. Mol. Cell Biol. 3 (2002) 21–29. [6] A. van Hoffen, J. Venema, R. Meschini, A.A. van Zeeland, L.H. Mullenders, Transcription-coupled repair removes both cyclobutane pyrimidine dimers and 6–4 photoproducts with equal efficiency and in a sequential way from transcribed DNA in xeroderma pigmentosum group C fibroblasts, EMBO J. 14 (1995) 360–367. [7] I. Kuraoka, M. Endou, Y. Yamaguchi, T. Wada, H. Handa, K. Tanaka, Effects of endogenous DNA base lesions on transcription elongation by mammalian RNA polymerase II. Implications for transcription-coupled DNA repair and transcriptional mutagenesis, J. Biol. Chem. 278 (2003) 7294–7299. [8] M.K. Shivji, J.G. Moggs, I. Kuraoka, R.D. Wood, Dual-incision assays for nucleotide excision repair using DNA with a lesion at a specific site, Methods Mol. Biol. 113 (1999) 373–392. [9] C.P. Selby, R. Drapkin, D. Reinberg, A. Sancar, RNA polymerase II stalled at a thymine dimer: footprint and effect on excision repair, Nucleic Acids Res. 25 (1997) 787–793. [10] S. Tornaletti, P.C. Hanawalt, Effect of DNA lesions on transcription elongation, Biochimie 81 (1999) 139–146. [11] S. Tornaletti, D. Reines, P.C. Hanawalt, Structural characterization of RNA polymerase II complexes arrested by a cyclobutane pyrimidine dimer in the transcribed strand of template DNA, J. Biol. Chem. 274 (1999) 24124–24130. [12] S. Tornaletti, S.M. Patrick, J.J. Turchi, P.C. Hanawalt, Behavior of T7 RNA polymerase and mammalian RNA polymerase II at site-specific cisplatin adducts in the template DNA, J. Biol. Chem. 278 (2003) 35791–35797. [13] H. Kettenberger, K.J. Armache, P. Cramer, Architecture of the RNA polymerase II-TFIIS complex and implications for mRNA cleavage, Cell 114 (2003) 347–357. [14] M.J. Thomas, A.A. Platas, D.K. Hawley, Transcriptional fidelity and proofreading by RNA polymerase II, Cell 93 (1998) 627– 637. [15] C.P. Selby, A. Sancar, Cockayne syndrome group B protein enhances elongation by RNA polymerase II, Proc. Natl. Acad. Sci. USA 94 (1997) 11205–11209.
Blockage of RNA polymerase II at a cyclobutane pyrimidine dimer and 6–4 photoproductq Joan Seah Mei Kwei,a,1 Isao Kuraoka,a,b,1 Katsuyoshi Horibata,a Manabu Ubukata,d Eiry Kobatake,d Shigenori Iwai,c Hiroshi Handa,d and Kiyoji Tanakaa,b,* a Graduate School of Frontier Biosciences, Osaka University, 1-3 Yamadaoka, Suita, Osaka 565-0871, Japan Core Research for Evolutional Science and Technology (CREST), Japan Science and Technology Corporation, 1-3 Yamadaoka, Suita, Osaka 565-0871, Japan c Graduate School of Engineering Science, Osaka University, 1-3 Machikanemaya, Toyonaka, Osaka 560-8531, Japan Faculty of Bioscience and Biotechnology and Frontier Collaborative Research Center, Tokyo Institute of Technology, Yokohama 226-8501, Japan b
d
Received 21 May 2004 Available online 25 June 2004
Abstract The blockage of transcription elongation by RNA polymerase II (pol II) at a DNA damage site on the transcribed strand triggers a transcription-coupled DNA repair (TCR), which rapidly removes DNA damage on the transcribed strand of the expressed gene and allows the resumption of transcription. To analyze the effect of UV-induced DNA damage on transcription elongation, an in vitro transcription elongation system using pol II and oligo(dC)-tailed templates containing a cyclobutane pyrimidine dimer (CPD) or 6–4 photoproduct (6–4PP) at a specific site was employed. The results showed that pol II incorporated nucleotides opposite the CPD and 6–4PP and then stalled. Pol II formed a stable ternary complex consisting of pol II, the DNA damage template, and the nascent transcript. Furthermore, atomic force microscopy imaging revealed that pol II stalled at the damaged region. These findings may provide the basis for analysis of the initiation step of TCR. Ó 2004 Elsevier Inc. All rights reserved. Keywords: RNA polymerase II; Cyclobutane pyrimidine dimer; 6–4 Photoproduct; Transcription-coupled DNA repair; Transcription elongation
Genomic DNA as carrier of genetic information in living cells is fundamentally vulnerable to ubiquitous DNA-damaging agents of endogenous and environmental origins [1]. Nucleotide excision repair (NER) eliminates a wide variety of bulky helix-distorting DNA lesions such as ultraviolet light (UV)-induced photolesions: cyclobutane pyrimidine dimers (CPD) and 6–4 photoproducts (6–4PP). NER can operate via two pathways: global genome repair (GGR) and transcription-coupled repair (TCR). GGR can repair DNA leq Abbreviations: NER, nucleotide excision repair; GGR, global genome repair; TCR, transcription-coupled DNA repair; pol II, RNA polymerase II; CPD, cyclobutane pyrimidine dimer; 6–4PP, 6–4 photoproduct; XP, xeroderma pigmentosum; CS, Cockayne syndrome; NTP, ribonucleoside triphosphate; RT, reverse transcriptase. * Corresponding author. Fax: +81-6-6877-9136. E-mail address: [email protected] (K. Tanaka). 1 Both authors contributed equally to this work.
0006-291X/$ - see front matter Ó 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2004.06.066
sions at any location in the whole genome, while TCR removes DNA lesions on the transcribed strands of expressed genes [2]. Xeroderma pigmentosum (XP) is an autosomal recessive disease characterized by a high incidence of skin cancer on sun-exposed areas while Cockayne syndrome (CS) is characterized by photosensitivity and several neuro-developmental abnormalities without a predisposition to skin cancer. So far, eight complementation groups have been identified in XP (XP-A through XP-G and XP-V), and two in CS (CS-A and CS-B). XP-A through XP-G have a defect in both NER pathways except for XP-C and XP-E, which have a defect in GGR alone, and XP-V, which has a proficient NER but is defective in translesion DNA synthesis. On the other hand, CS is defective in TCR alone. All the genes responsible for XP and CS have been cloned. The XPC protein that complexes with HR23B and the UV-DDB
1134
J.S.M. Kwei et al. / Biochemical and Biophysical Research Communications 320 (2004) 1133–1138
complex that comprises DDB1 and DDB2 proteins preferentially bind to various types of DNA damage and are involved in GGR-specific damage recognition. In TCR, elongating RNA polymerase II (pol II) that encounters a lesion is thought to play the corresponding role of XPC/HR23B and UV-DDB. For each subpathway-specific damage recognition event, NER factors including TFIIH, XPA, RPA, XPF/ERCC1, and XPG may be recruited to the lesion [3–5]. Although 6–4PP is less frequently produced than CPD, 6–4PP is more cytotoxic and mutagenic. XPCHR23B complex preferentially binds to 6–4PP rather than CPD, and GGR repairs 6–4PP more rapidly than CPD. In contrast, TCR can repair CPD and 6–4PP at the same rate in an XPC-independent manner [6]. Here, we showed that (1) pol II itself completely stalls at both CPD and 6–4PP lesions, that (2) RNA synthesis by pol II is stopped by incorporating one or two nucleotide(s) opposite the lesion, and that (3) pol II forms ternary complex with the nascent transcripts around the lesion. Our present data indicate that the stalling of pol II itself at the damage site might be a trigger for TCR and provide the reasons why it might recognize the lesion. Methods Enzymes. The preparation of human RNA polymerase II (pol II) was previously described [7]. Bacteriophage T4 pyrimidine dimer DNA glycosylase (T4 endonuclease V) and ultraviolet damage endonuclease (UVDE) of Schizosaccharomyces pombe were purchased from Trevigen. DNA template with single UV-damage at a defined position. To prepare a template with single DNA damage for transcription elongation, 30-mer oligo-deoxyribonucleotides containing a cyclobutane
pyrimidine dimer (CPD) or a pyrimidine–pyrimidone 6–4 photoproduct (6–4PP) were incorporated into covalently closed circular DNA as previously described [8]. The plasmid pBluescript II KS-UV (pBSII KS-UV) was constructed by modifying pBSII KS-GTG [7] with 30 synthetic oligonucleotides containing the damage (Fig. 1A). In vitro transcription elongation reactions using Pol II and an oligo(dC)-tailed template. For the in vitro transcription elongation assay [7], 20 ll reaction mixtures containing 50 ng of oligo(dC)-tailed template and 0.5 ll pol II were preincubated for 30 min at 30 °C. Elongation was started by adding 5 ll NTP and terminated by adding 100 ll stop buffer. The purified transcripts were resuspended in formamide loading dye and separated on a denatured 6% polyacrylamide gel. The dried gels were analyzed using a FUJIFILM BAS 2500 bioimage analyzer. Sequence analysis of nascent transcripts amplified by 30 -ligationmediated RT-PCR. For sequence analysis, the stalled transcripts were extracted from the elongation reaction mixture in the presence of a cold NTP mixture (50 lM each of ATP, CTP, GTP, and UTP) using a RNA purification kit (Qiagen) and then treated with T4 RNA ligase (TaKaRa) as recommended by the supplier in the presence of RNA primers (50 -p-CACCAAACGTGGCUUGCCAGCCC-NH2 -30 ). The ligated transcripts were used to synthesize PCR products with a One Step RT-PCR kit (Qiagen) using the primers (704–724) 50 GGAATTCGATATCAAGCTTA-30 and (anti-RNA primer) 50 GGGCTGGCAAGCCACGTTTGGTG-30 . PCR was performed again using anti-RNA linkers and the primer (751–770) 50 GCCCTGCTGCCATGCGCGG-30 . The PCR products were isolated with a TOPO TA Cloning kit (Invitrogen) and then sequenced using a BigDye Terminator kit (Applied Biosystems) and the PCR primer (751–770). Immunoprecipitation of pol II-ternary complex. Elongating pol II was enriched from mixtures of the transcription elongation reaction by immunoprecipitation using anti pol II antibody (C21)-conjugated agarose in the presence of cold NTP. The samples were washed five times with a buffer [20 mM Tris–HCl (pH 8), 500 mM NaCl, 2 mM EDTA, 1% Triton X-100, and 0.1% SDS] and resuspended with restriction buffer [10 mM Tris–HCl (pH 7.5), 10 mM MgCl2 , 50 mM NaCl, and 1 mM dithiothreitol] for digestion with EcoRV and PvuII. After the digestion, elongating pol II was enriched again by immunoprecipitation with anti-pol II antibody beads. PCR was then carried
Fig. 1. Preparation of DNA template with single UV-damage at a defined position. (A) Covalently closed circular duplex DNA containing a single lesion. Thirty-mer oligonucleotides containing a single DNA lesion and the plasmid pBSII KS-UV are shown diagrammatically. The TT dimer site is indicated as D. Unique PstI and SmaI restriction enzyme sites are also indicated. (B) One percentage of agarose gel electrophoresis of pBSIIKS-UV after digestion of each plasmid with T4 endonuclease V (T4 endo), demonstrating the presence of a single CPD site. (C) One percentage agarose gel after digestion of each plasmid with ultraviolet damage endonuclease (UVDE), demonstrating the presence of CPD and 6–4PP sites. The mobility of covalently closed circular (ccc) and nicked circular (nc) pBSII KS-UV is indicated by the arrows.
J.S.M. Kwei et al. / Biochemical and Biophysical Research Communications 320 (2004) 1133–1138 out with primers (751–770) and (1009–990) 50 -ACTCATTAGGC ACCCCAGGC-30 . In the case of RT-PCR, the immunoprecipitated samples were treated with RNase-free DNase I (TaKaRa) instead of the restriction digestion step. The RNA was purified by a RNA purification kit (Qiagen), and then 30 -ligation-mediated RT-PCR was performed. The PCR products (259 bp) and RT-PCR products (109 bp) were analyzed by electrophoresis on a 4% agarose gel. AFM imaging. For AFM (atomic force microscopy) imaging, 10 ll reaction mixtures containing 10 ng of oligo(dC)-tailed template and 0.5 ll pol II in a buffer [10 mM Hepes–KOH (pH 7.9), 50 mM KCl, 10% glycerol, 0.1 mM EDTA, 0.25 mM dithiothreitol, 6 mM MgCl2 , 50 lM NTP mixture, and 8 U RNase Inhibitor (Promega)] were incubated for 30 min at room temperature. The sample was deposited on the center of a freshly cleaved mica disk. After 1 min, the mica was washed with 100 ll distilled water and dried. AFM studies were performed using a NanoScope III system (Veeco Instruments).
Results Construction of DNA templates containing either CPD or 6–4PP at a defined position on the transcribed strand To verify that the DNA template contained a lesion, 3060-bp pBSII KS-UV (Fig. 1A) was incubated with either T4 endonuclease V or UVDE. Since T4 endonuclease V is specific for CPD, 97% of the DNA template containing CPD (CPD template) was cleaved to produce
1135
an open circular form (Fig. 1B, lane 4), but the DNA template with 6–4PP (6–4PP template) and the one without damage (control template) were not cleaved (Fig. 1B, lanes 2 and 6). However, 97% of the CPD template and 95% of the 6–4PP template were cleaved by UVDE, which is specific for both photoproducts (Fig. 1C, lanes 4 and 6), while the control template was not cleaved (Fig. 1C, lane 2). Pol II stalled at CPD and 6–4PP sites When pol II was incubated with control templates, it synthesized elongation transcripts, but did not do so in the presence of a-amanitin (Fig. 2A, lanes 1 and 2). When CPD or 6–4PP template was used, pol II produced a transcript of about 130 nt in size. This corresponded to the region from the transcription start site to the DNA damage site (Fig. 2A, lanes 4 and 6), suggesting that pol II stalled at the CPD and 6–4PP sites. To confirm that pol II stalled at both lesions, time course experiments on the transcription elongation reaction were carried out (Fig. 2B). Pol II produced elongation transcripts from the control template (Fig. 2B, lanes 1–3). In the case of CPD or 6–4PP template, pol II produced transcripts of approximately
Fig. 2. Transcription elongation by pol II using an oligo(dC)-tailed template containing a single TT dimer. (A) Autoradiograph after denatured 6% polyacrylamide gel electrophoresis of the nascent transcripts derived from a transcription elongation reaction in the presence or absence of a-amanitin (a-ama). Transcripts from the templates containing no lesions (lanes 1 and 2), CPD (lanes 3 and 4), and 6–4PP (lanes 5 and 6) lesions in the presence (lanes 1, 3, and 5) or absence of a-amanitin (lanes 2, 4, and 6). (B) Time course (2–16 min) reaction of transcription elongation with templates containing no lesion (lanes 1–3), CPD (lanes 4–7) or 6–4PP (lanes 8–11) lesions. Stalled and elongation transcripts are indicated by the arrow and bracket, respectively.
1136
J.S.M. Kwei et al. / Biochemical and Biophysical Research Communications 320 (2004) 1133–1138
130 nt in size and elongation transcripts in a time-dependent manner (Fig. 2, lanes 4–7 and lanes 8–11). To examine whether pol II can bypass the lesions, the stalled transcripts (130 nt) and elongation transcripts were quantified. The results showed that the amount of elongation transcripts was 3.1% and 5.3% of the total transcripts on the CDP and 6–4PP template, respectively (data not shown). Sequencing of the elongation transcripts derived from the CPD and 6–4PP templates revealed no mutations (data not shown). Since the purity of the CDP and 6–4PP templates was 97% and 95%, respectively, this suggests that the elongation transcripts were produced from undamaged templates contaminated in the CPD and 6–4PP templates and that pol II completely stalled at the CPD and 6–4PP sites. Two other RNA polymerases (Escherichia coli and T7 RNA polymerase) stalled at the CPD and 6–4PP sites as well (data not shown). Analysis of 30 end stalled transcripts on CPD and 6–4PP templates To examine whether pol II has an ability to incorporate nucleotides when it encounters a lesion, the stalled transcripts were isolated and sequenced. We designed 50 phosphate RNA linkers that can be ligated to the 30 -OH ends of the stalled transcripts in the presence of T4 RNA ligase. Using a complementary primer that anneals to the RNA linkers, RT-PCR, and nested PCR were performed. This step helped to eliminate the pos-
sibility of cloning natural pausing products and other transcripts. The RT-PCR products were subcloned into the TA-cloning plasmid. An outline of this strategy is shown in Fig. 3A. The sequencing data (Fig. 3B) indicated that pol II incorporated nucleotide(s) opposite the lesion on the CPD and 6–4PP templates at frequencies of 92.5% and 86.8%, respectively. The frequency of misincorporation (rCMP and UMP) opposite the CPD site was 57.5%, while most of the nucleotide(s) incorporated opposite the 6–4PP site were correct. These findings suggest that the promiscuous behavior of pol II still fails to bypass these strong blocks even when ideal incorporation occurs. Ternary complex of pol II/transcripts/DNA template Stalled pol II forms a stable ternary complex of pol II/transcripts/DNA template in in vitro transcription reactions using general transcription factors and the adenovirus major late promoter [9–11]. We examined whether pol II can form a ternary complex at the DNA damage site in our transcription assay as well. As shown in Fig. 4A, the PCR products (259 bp) located around the CPD and 6–4PP sites were amplified in the anti-pol II immunoprecipitants of the transcription reaction mixtures including CPD and 6–4PP templates, but not in the immunoprecipitants including control template. In addition, RT-PCR products of the stalled transcript were amplified only in the anti-pol II immunoprecipitants of the transcription reaction mixtures including
Fig. 3. Analysis of 30 end of stalled transcripts on CPD and 6–4PP templates. (A) Flowchart showing how the stalled transcripts were isolated, ligated with RNA linker at the 30 end, amplified by RT-PCR and nested PCR, and subcloned into plasmid for DNA sequencing. (B) Nucleotide sequences of the 30 end of stalled transcripts opposite the CPD and 6–4PP sites. The frequency of stalled transcripts containing each type of the 30 end was expressed as the colony number of the TA-cloning plasmids and their percentages.
J.S.M. Kwei et al. / Biochemical and Biophysical Research Communications 320 (2004) 1133–1138
1137
Fig. 4. Stable ternary complex of stalled Pol II/transcripts/DNA template. (A) Four percentage agarose gel electrophoresis of PCR products derived from the in vitro transcription reaction mixture with UV-damage or control DNA template that was anti-pol II coimmunoprecipitated (after IP) or nonimmunoprecipitated (before IP). The PCR products (259 bp) are indicated by the arrow. (B) Four percentage agarose gel electrophoresis of the 30 ligation-mediated RT-PCR products derived from stalled transcripts that were coimmunoprecipitated by anti-pol II. RT-PCR products (109 bp) with (+) or without ()) reverse transcriptase (RT) are indicated by the arrow. (C–E) Atomic force microscopy (AFM) image of a field of molecules deposited from the pol II reaction mixture. Each panel indicates the protein bound to DNA. Pol II bound at the end of the control DNA (C), bound at the expected site of CPD (D), and 6–4PP (E). The scan area is 1lm 1lm for the upper panels, and the lower panels indicate the enlarged images (0.25 lm 0.25 lm). The height (0–5.0 nm) is indicated along the side of the (E). Stalled pol II is indicated by the arrow.
DNA damaged templates (Fig. 4B). These results again suggested that pol II stalled at the DNA damage sites with nascent transcripts. AFM imaging of the stalled pol II To visualize the pol II stalled at DNA damage site, AFM (atomic force microscopy) was employed. The in vitro transcription elongation reaction using pol II and oligo(dC)-tailed template with or without UV damage was performed and then deposited onto mica for AFM imaging. Although for our experimental AFM imaging conditions, transcripts and oligo(dC) tails were not detected. However, pol II (0.5 MDa) and double stranded templates (1 lm) were visualized. The AFM data revealed that pol II bound to the end of the control template, implying that the pol II completely reads through the template (Fig. 4C) or nonspecifically binds to its end. In the case of CPD and 6–4PP templates (Figs. 4D and E), pol II bound to the middle of DNA. The position of pol II along the DNA was measured. The length of the DNA from the pol II to the nearest end was divided by the full-length of the DNA template. According to the proportion of the DNA fragment length, it was estimated that pol II approximately bound at the damaged sites which were about 130 nt away from the nearest end of the DNA template. Pol II also bound to the other end of the CPD template. This binding may have been nonspecific due to our experimental conditions.
Discussion Our present study indicates that pol II completely stalled at the CPD and 6–4PP sites. Mammalian RNA
polymerase II stall at bulky adducts such as CPD, aminofluorene, acetylaminofluorene, and cisplatin in an in vitro transcription elongation system using the adenovirus major late promoter and general transcriptional factors [10,12]. We only used pol II and an oligo(dC)-tailed template for transcription elongation, indicating that pol II itself can recognize the damaged site where it stalls. Although TCR and GGR contribute to the removal of CPD and 6–4PP, the efficiency of removal varies. GGR can remove 6–4PP from the genome much faster than CPD in vivo. Consistent with this, XPC-HR23B complex, which is involved in the damage recognition step of GGR, has a higher affinity for 6–4PP than CPD in vitro. On the other hand, TCR can remove both lesions with equal efficiency in vivo. Current models indicate that TCR is triggered by stalled RNA pol II. Damage recognition by TCR might be dependent on how pol II stalls at the lesion. Thus, our finding that pol II stalled at CPD and 6–4PP lesions and forms a stable ternary complex in similar manner may account for the same efficiency of the TCR of CPD and 6–4PP in vivo. Sequencing analysis of the stalled transcripts revealed that pol II stalled after incorporating correct or incorrect nucleotide(s) opposite the CPD and 6–4PP sites. We also showed that pol II forms a stable ternary complex with the DNA lesion and stalled transcript. It was thought that pol II stalls when the 30 -OH end of the nascent transcript loses its ideal base pair contacts with the DNA template and hence is extruded from the pol II active site [13,14]. Previously, we reported that pol II stalled at 8-oxoguanine (8-oxoG), which is neither a bulky nor a distorting oxidative lesion, only when pol II makes misincorporation with 8-oxoG [7]. Considering this, we hypothesize that the incorporation of nucleotides opposite UV lesions at the 30 -OH end of a nascent
1138
J.S.M. Kwei et al. / Biochemical and Biophysical Research Communications 320 (2004) 1133–1138
transcript, even if the pairing is correct, may cause instability in the base pair contact at the lesions, leading pol II to stall. We showed that the stalled pol II securely attaches to the DNA damage site and the stable ternary complex is formed at the damaged site. Therefore, the lesion is most probably covered with pol II. It remains unresolved how the DNA repair machinery gains access into the lesion. However, the incubation of transcriptional elongation factor SII with the pol II ternary complex stalled at CPD produced a 10–30-nt cleavage profile of nascent transcripts [11,15], suggesting that SII enables stalled pol II to backtrack and make an opening for DNA repair proteins. Thus, we believe that the SII function is important for the initial step of TCR. In the normal transcription, SII stimulated the intrinsic nuclease activity of pol II and the misincorporated nucleotides are removed from nascent RNA to allow the resumption of transcription elongation [13,14]. It is thought that SII is recruited to pol II ternary complexes by naturally occurring misincorporation of pol II. Taken together, it is suggested that the promiscuous incorporation of pol II that we observed at UV damaged sites might recruit SII to the stalling pol II complexes at the damage sites in TCR.
Acknowledgments We thank Mika Hayashida for her technical assistance. J.S.M.K was the recipient of an Advanced Biotechnology Scholarship from Ishihara Sangyo Kaisha, Ltd., in collaboration with the Singapore Economic Development Board. This work was supported by a Grantin-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan, and a Core Research for Evolutional Science and Technology (CREST) of Japan Science and Technology Corporation.
References [1] T. Lindahl, R.D. Wood, Quality control by DNA repair, Science 286 (1999) 1897–1905.
[2] P.C. Hanawalt, D.J. Crowley, J.M. Ford, A.K. Ganesan, D.R. Lloyd, T. Nouspikel, C.A. Smith, G. Spivak, S. Tornaletti, Regulation of nucleotide excision repair in bacteria and mammalian cells, Cold Spring Harb. Symp. Quant. Biol. 65 (2000) 183– 191. [3] M.A. Nance, S.A. Berry, Cockayne syndrome: review of 140 cases, Am. J. Med. Genet. 42 (1992) 68–84. [4] K. Sugasawa, T. Okamoto, Y. Shimizu, C. Masutani, S. Iwai, F. Hanaoka, A multistep damage recognition mechanism for global genomic nucleotide excision repair, Genes Dev. 15 (2001) 507–521. [5] J.Q. Svejstrup, Mechanisms of transcription-coupled DNA repair, Nat. Rev. Mol. Cell Biol. 3 (2002) 21–29. [6] A. van Hoffen, J. Venema, R. Meschini, A.A. van Zeeland, L.H. Mullenders, Transcription-coupled repair removes both cyclobutane pyrimidine dimers and 6–4 photoproducts with equal efficiency and in a sequential way from transcribed DNA in xeroderma pigmentosum group C fibroblasts, EMBO J. 14 (1995) 360–367. [7] I. Kuraoka, M. Endou, Y. Yamaguchi, T. Wada, H. Handa, K. Tanaka, Effects of endogenous DNA base lesions on transcription elongation by mammalian RNA polymerase II. Implications for transcription-coupled DNA repair and transcriptional mutagenesis, J. Biol. Chem. 278 (2003) 7294–7299. [8] M.K. Shivji, J.G. Moggs, I. Kuraoka, R.D. Wood, Dual-incision assays for nucleotide excision repair using DNA with a lesion at a specific site, Methods Mol. Biol. 113 (1999) 373–392. [9] C.P. Selby, R. Drapkin, D. Reinberg, A. Sancar, RNA polymerase II stalled at a thymine dimer: footprint and effect on excision repair, Nucleic Acids Res. 25 (1997) 787–793. [10] S. Tornaletti, P.C. Hanawalt, Effect of DNA lesions on transcription elongation, Biochimie 81 (1999) 139–146. [11] S. Tornaletti, D. Reines, P.C. Hanawalt, Structural characterization of RNA polymerase II complexes arrested by a cyclobutane pyrimidine dimer in the transcribed strand of template DNA, J. Biol. Chem. 274 (1999) 24124–24130. [12] S. Tornaletti, S.M. Patrick, J.J. Turchi, P.C. Hanawalt, Behavior of T7 RNA polymerase and mammalian RNA polymerase II at site-specific cisplatin adducts in the template DNA, J. Biol. Chem. 278 (2003) 35791–35797. [13] H. Kettenberger, K.J. Armache, P. Cramer, Architecture of the RNA polymerase II-TFIIS complex and implications for mRNA cleavage, Cell 114 (2003) 347–357. [14] M.J. Thomas, A.A. Platas, D.K. Hawley, Transcriptional fidelity and proofreading by RNA polymerase II, Cell 93 (1998) 627– 637. [15] C.P. Selby, A. Sancar, Cockayne syndrome group B protein enhances elongation by RNA polymerase II, Proc. Natl. Acad. Sci. USA 94 (1997) 11205–11209.