Reduced RNA polymerase II transcription in extracts of Cockayne syndrome and xeroderma pigmentosum/Cockayne syndrome cells
Reduced RNA polymerase II transcription in extracts of Cockayne syndrome and xeroderma pigmentosum/Cockayne syndrome cellsGrigory L. Dianov1,2, Jean-François Houle1, Narayan Iyer1,+, Vilhelm A. Bohr2 and Errol C. Friedberg1,*
1Laboratory of Molecular Pathology, Department of Pathology, University of Texas Southwestern Medical Center, Dallas, TX 75235, USA and 2Laboratory of Molecular Genetics, National Institute on Aging, NIH, 4940 Eastern Avenue, Baltimore, MD 21224, USA
Received May 27, 1997;Revised and Accepted July 24, 1997
ABSTRACT
The hereditary disease Cockayne syndrome (CS) is a complex clinical syndrome characterized by arrested post-natal growth as well as neurological and other defects. The CSA and CSB genes are implicated in this disease. The clinical features of CS can also accompany the excision repair-defective hereditary disorder xeroderma pigmentosum (XP) from genetic complementation groups B, D or G. The XPB and XPD proteins are subunits of RNA polymerase II (RNAP II) transcription factor IIH (TFIIH). We show here that extracts of CS-A and CS-B cells, as well as those from XP-B/CS cells, support reduced levels of RNAP II transcription in vitro and that this feature is dependent on the state or quality of the template.
INTRODUCTION
Cockayne syndrome (CS) is a rare autosomal recessive human disease characterized primarily by arrested post-natal growth and development (1 ). Two genetic complementation groups (CS-A and CS-B) have been identified for the disease (2 ). Both the CSA and CSB genes have been isolated by functional cloning strategies and mutations have been identified in these genes in afflicted individuals (3 ,4 ). Little is known about the function(s) of the proteins encoded by the CSA and CSB genes. The predicted amino acid sequence of CSB protein places it in a superfamily of ATPases designated the SNF/SWI family (5 ,6 ) which, among multiple other subfamilies, includes the yeast Snf2 ATPase and its human and Drosophila homologs (4 ). Snf2 protein is a subunit of the multi-protein Swi-Snf complex, which is believed to function in chromatin remodeling associated with transcriptional activation in yeast (5 ,6 ). The predicted amino acid sequence of the human CSA gene reveals that it is a member of the WD repeat protein family (3 ). WD repeat proteins are involved in various aspects of cellular metabolism, including signal transduction, vesicular trafficking, regulation of cytoskeletal assembly, cell cycle regulation, RNA processing and gene regulation (reviewed in 7 ). Structural models suggest that WD repeat proteins have the potential for interacting with other proteins, consistent with the observation that many such proteins are components of multi-protein complexes (reviewed in 7 ).
The clinical features of CS are sometimes also observed in patients with xeroderma pigmentosum (XP), a hereditary disease which by itself manifests with a marked predisposition to skin cancer and defective nucleotide excision repair (NER) in both transcriptionally active and transcriptionally silent DNA when patients or their cells in culture are exposed to UV radiation (2 ). The association of the clinical features of both CS and XP is presently confined to three of the eight known XP genetic complementation groups, XP-B, XP-D and XP-G (2 ), and mutations in the XPB and XPD and XPG genes have been identified in XP-B/CS, XP-D/CS and XP-G/CS individuals (8 -10 ). Both the XPB and XPD genes encode subunits of RNA polymerase II (RNAP II) basal transcription factor IIH (TFIIH), which is essential for RNAP II transcription initiation and is also indispensable for NER (11 ,12 ). A role of the XPG gene in RNAP II transcription has not been documented. However, recent studies have shown that deletion of the XPG gene in mouse cells results in early post-natal death (Y.Harada and T.Shiomi, personal communication). Hence, it would appear that, like the XPB and XPD genes, the XPG gene has a second important (if not essential) function in addition to its role in NER.
Unlike XP, CS is not related to any known environmental exposure. However, when CS cells in culture are treated with various DNA damaging agents they manifest phenotypes consistent with a defect(s) in DNA repair. The cells are abnormally sensitive to UV radiation and to a number of UV-mimetic chemicals (2 ). Following exposure to UV radiation CS cells are defective in strand-specific NER (so-called transcription coupled repair), during which the rate of repair of the template (transcribed) strand of transcriptionally active genes is enhanced relative to the coding (non-transcribed) strand (13 ,14 ). In contrast to normal cells, CS-B cells as well as cells from the XP-G/CS complex also fail to show a kinetic preference for repair of the transcribed strand after exposure to ionizing radiation (15 ) or hydrogen peroxide (16 ).
The elucidation of a role for the XPB and XPD proteins in RNAP II transcription has prompted the suggestion that the CS phenotype in XP/CS patients primarily results from abnormalities in the transcription function of these proteins (17 -19 ). A logical extension of this hypothesis predicts that cells from patients belonging to the CS-A and CS-B complementation groups (who do not simultaneously suffer from XP) might also fail to support normal RNAP II-mediated transcription. Indeed, reduced levels of RNAP II transcription in vivo have recently been documented in CS-B cells (20 ). Defective repair that is coupled to transcription may then be a secondary phenotype observed when CS or XP/CS cells are exposed to various DNA damaging agents (21 ).
In the present study we have examined the capacity of CS cells to carry out transcription by RNAP II in vitro. We have observed reduced levels of transcription driven by plasmid-borne RNAP II promoters in extracts of various CS-A and CS-B cells. Reduced transcription was corrected by mixing specific fractions of CS and normal cell extracts. Additionally, when CS-A or CS-B cells were transfected with the corresponding wild-type gene, transcription was restored to normal levels. Reduced levels of RNAP II transcription were also observed in extracts of cells from an individual with the XP-B/CS syndrome.
MATERIALS AND METHODS
Cell lines
Human lymphoid cell lines and SV40-transformed human fibroblast cell lines were obtained from the Human Genetic Mutant Cell Repository (Camden, NJ) and were shown to be free of Mycoplasma. Cells were grown in medium recommended by the Repository. The following cell lines were used: normal human lymphoblasts GM3798, GM1953, GM3714 and GM3201; CS-A lymphoblasts GM1857; CS-B lymphoblasts GM1712; XP-B/CS lymphoblasts GM2252; SV40-transformed human fibroblasts CS3BE (CS-A) and CS/1AN (CS-B); primary fibroblasts MpSI (CS-A). This cell line was kindly provided by Dr M.Yamaizumi (Kumamoto University, Japan). Stable transfectants were derived from the CS3BE (CS-A) and CS/1AN (CS-B) cell lines by transfection of the CSA and CSB genes respectively, as described previously (3 ).
Preparation and fractionation of whole cell extracts
Whole cell extracts were prepared essentially by the method of Manley et al. (22 ). Extracts for fractionation on phosphocellulose were dialyzed overnight at 4oC against buffer A [25 mM HEPES-KOH, pH 7.9, 1 mM dithiothreitol (DTT), 1 mM EDTA, 10% glycerol, 0.01% Nonidet-P40, 0.1 mM phenylmethylsulphonyl fluoride] containing 0.1 M KCl. The dialysate was centrifuged for 5 min in a microfuge and the supernatant loaded onto a phosphocellulose column (Whatman P11, 10 mg protein/ml bed column vol.) equilibrated in buffer A with 0.1 M KCl. The column was washed with 5 column vol. buffer A containing 0.1 M KCl. The bound protein was sequentially eluted with 10 vol. buffer A containing 0.35 M KCl and then with 5 vol. buffer A containing 1 M KCl. Peak fractions from the flow-through (PC-FI) and 1 M eluate (PC-FII) were concentrated by centrifugation on Centriprep 10 (Amicon), dialyzed against 25 mM HEPES-KOH, pH 7.9, 1 mM DTT, 1 mM EDTA, 17% glycerol, 12 mM MgCl2, 0.1 M KCl, aliquoted and stored at -80oC.
DNA template for transcription
Plasmid pML(C2AT) was provided by Dr Roger Kornberg (Stanford University, Stanford, CA). Briefly, plasmid preparations designated as DNA I were prepared by lysozyme lysis of cells followed by clearing of the lysate. The cleared lysate was loaded onto a Qiagen column and the plasmid eluted utilizing conditions recommended by the manufacturer (Qiagen). The plasmid was purified by equilibrium centrifugation in a CsCl/ethidium bromide gradient under conditions that strictly avoided exposure of the gradients to visible light. The DNA was further purified by sucrose gradient centrifugation (23 ). Plasmid preparations designated as DNA II were isolated by the SDS/alkaline lysis method and purified by equilibrium centrifugation in a CsCl/ethidium bromide gradient (24 ) followed by phenol/chloroform treatment and ethanol precipitation.
RNAP II transcription in vitro
Transcription reactions (50 [mu]l) contained 2 [mu]g supercoiled plasmid pML(C2AT) containing the adenovirus major late promoter (AdML) plus a G-less cassette (25 ), 37.5 mM HEPES-KOH, pH 7.9, 1.5 mM DTT, 0.5 mM EDTA, 8.5% glycerol, 8.5 mM MgCl2, 50 mM KCl, 8 mM phosphocreatine, 2.5 [mu]g creatine phosphokinase (Type I; Sigma), 500 [mu]M ATP and CTP, 5 [mu]M UTP, 20 [mu]Ci [32P]UTP (3000 Ci/mmol; Amersham), 20 U RNase inhibitor (rRNasin; Promega) and varying amounts of PC-FI and PC-FII to give a total of 40 [mu]g as protein. When noted, reactions contained 40-100 [mu]g protein of unfractionated whole cell extract instead. Reactions were carried out at 30oC for 1 h, followed by treatment with RNase T1 (10-20 U in 200 [mu]l 10 mM Tris-HCl, pH 7.5, 0.3 M NaCl and 5 mM EDTA) for 10 min at room temperature. SDS (12.5 [mu]l of a 10% solution) and 10 [mu]l proteinase K (5 mg/ml) were added and reactions were incubated at 30oC for 20 min, followed by addition of 2 [mu]l tRNA (10 mg/ml) and 0.75 ml ethanol. After 20 min at -20oC RNA was pelleted by centrifugation for 10 min in a microfuge and washed with 0.5 ml 75% ethanol. The pellet was dried under vacuum, dissolved in 20 [mu]l formamide dye solution, heated for 5 min at 95oC and 10 [mu]l was loaded onto a 5% polyacrylamide-7 M urea gel. Electrophoresis was carried out in Tris-borate-EDTA buffer at 45 mA and gels were washed for 20 min in distilled water, dried and subjected to autoradiography. The expected product has a length of ~400 nt and no transcript is produced when [alpha]-amanitin is introduced into the reaction mixture.
Detection of RNAP II transcription factors by immunoblotting
Whole cell extract and phosphocellulose fractions I and II (50 [mu]g protein) from wild-type (GM3798 or GM3201) and CS-A (GM1857) cells were resolved by SDS-PAGE and transferred to nitrocellulose membranes. Blots were probed with polyclonal rabbit antibodies (Santa Cruz Biotechnology) directed against TFIIE-[alpha], TFIIE-[beta], TFIIF-RAP74, TFIIF-RAP30 and p89 of TFIIH. Horseradish peroxidase-conjugated polyclonal donkey antibody to rabbit immunoglobulin (Amersham) was used as secondary antibody and immunocomplexes were detected by enhanced chemiluminescence (ECL; Amersham).
Measurement of DNA repair synthesis
This assay was performed as previously described (26 ).
RESULTS
Reduced levels of RNAP II transcription in crude extracts
The XPB and XPD genes, which encode subunits of the RNAP II basal transcription factor TFIIH, have been implicated in the XP/CS complex (11 ,12 ,17 -19 ). A recent study demonstrated reduced levels of basal transcription in an in vitro reconstituted transcription assay supplemented with TFIIH isolated from XP-B/CS cells compared with assays supplemented with TFIIH isolated from normal cells (27 ). We examined the levels of RNAP II transcription in CS and XP-B/CS cells in vitro. In initial experiments performed with crude extracts of CS-A, CS-B and XP-B/CS cells we observed reduced levels of transcription from a supercoiled plasmid carrying the AdML RNAP II promoter (Fig. 1 ). More refined analysis led to the observation that reduced transcription with extracts of CS cells correlated with the particular protocol used for isolation and purification of the plasmid DNA used as the transcription template. When plasmid DNA was purified by lysis of Escherichia coli cells with lysozyme followed by CsCl/ethidium bromide gradient centrifugation under conditions that avoided exposure of the gradients to visible light and then by sucrose gradient centrifugation, we observed equivalent levels of transcription in normal, CS-A, CS-B and XP-B/CS cells (Fig. 2 A, DNA I). We designate this DNA I. In contrast, when the plasmid DNA was isolated by SDS/alkaline lysis and purified by CsCl/ethidium bromide gradient centrifugation (without regard to exposure of the DNA to light) followed by phenol/chloroform treatment and ethanol precipitation, we consistently observed reduced levels of transcription in CS and XP-B/CS cells (Fig. 2 A, DNA II). We designate this DNA II.
Reduced levels of RNAP II transcription in fractionated extracts
RNAP II transcription in extracts of XP-B/CS cells
The combined XP/CS clinical complex is most consistently observed in patients from XP complementation group B (2 ). As was the case with CS-A and CS-B cells, whole cell extracts (Figs. 1 and 2 A) as well as reconstituted phosphocellulose fractions of extracts from XP-B/CS lymphoblastoid cells (Fig. 4 A) showed reduced levels of transcription from the AdML promoter (DNA II). This defect was not corrected by mixing PC-FII from XP-B/CS cells with PC-FI from normal cells (Fig. 4 B, lane 2) or CS-A cells (Fig. 4 B, lane 4). However, the defect was corrected by mixing PC-FI from XP-B/CS cells with PC-FII from either normal (Fig. 4 B, lane 3) or CS-A cells (Fig. 4 B, lane 5). Hence, the proteins defective in XP-B/CS cells on the one hand and in CS-A and CS-B cells on the other reside in different phosphocellulose fractions. In fact, we have detected the p89 subunit of TFIIH in PC-FII by Western analysis, in accordance with previously published data (Fig. 5 ; 30 ). In addition, we have detected CSA protein in PC-FI by immunoblotting with an antibody prepared against CSA protein (J.-F.Houle and E.C.Friedberg, unpublished observations) (data not shown).
Specificity of reduced RNAP II transcription
In order to eliminate the possibility that the results described above reflect aberrant chromatography of CS extracts on phosphocellulose we compared the distribution of several well-characterized RNAP II transcription factors in PC-FI and PC-FII from normal and CS cells by Western analysis. We detected equivalent amounts of TFIIE-[alpha], TFIIE-[beta], TFIIF-RAP74, TFIIF-RAP30 and TFIIH-p89 in PC-FII of normal and CS cells (Fig. 5 ). We also observed equivalent (smaller) amounts of TFIIE-[alpha] and TFIIF-RAP30 in PC-FI of both normal and CS cells. Equivalent levels of TFIIE-[alpha], TFIIE-[beta], TFIIF-RAP 74 and TFIIF-RAP 30 were also detected in the discarded 0.35 M phosphocellulose fraction of normal and CS cells (data not shown).
We carried out several experiments to show that extracts of CS cells were not non-specifically inactivated during their preparation or fractionation. First, as previously documented (13 ), extracts of both CS-A and CS-B cells supported normal levels of NER of UV-irradiated plasmid DNA in vitro (data not shown). Second, we examined in vitro complementation of reduced transcription by mixing different phosphocellulose fractions from CS and normal cells. Efficient transcription from the AdML promoter was restored when PC-FI from normal cells was mixed with PC-FII from CS-A (Fig. 6 A, right) or CS-B (Fig. 6 B, right) cells. In contrast, the reverse combination of PC-FI from CS-A (Fig. 6 A, left) or CS-B cells and PC-FII from normal cells (Fig. 6 B, left) did not restore normal transcription activity. When equal amounts of PC-FI from both CS-A and CS-B extracts were mixed with PC-FII from normal cells defective transcription persisted (Fig. 7 A). These results indicate that factors present in PC-FI (presumably CSA and CSB proteins) which influence RNAP II transcription in vitro are defective in both CS-A and CS-B cells. When equal amounts of PC-FI from normal and CS-A or CS-B cells were mixed with PC-FII from wild-type cells normal levels of transcription were observed from the AdML promoter (Fig. 7 B, compare lanes 1, 2 and 3). Hence, CS cells do not contain an inhibitor of in vitro RNAP II transcription.
Complementation of reduced RNAP II transcription in vitro and in vivo
In order to establish that defective RNAP II transcription in extracts of CS-A and CS-B cells results from a defect in the corresponding gene, we compared RNAP II transcription in extracts of CS cell lines transfected with a plasmid vector or with the vector carrying the cloned CSA or CSB cDNA. Transcription from the AdML promoter was fully complemented in extracts prepared from CS cells carrying the appropriate gene (Fig. 8 A and B).
DISCUSSION
There is no experimental evidence of which we are aware that directly implicates the CSA and CSB proteins in RNAP II transcription in human cells. Hence, our observations of reduced levels of RNAP II transcription in both crude and fractionated extracts of CS-A and CS-B cells are provocative. Similar observations have recently been made using techniques that measure RNAP II transcription in vivo in CS-B cells (20 ).
The cellular phenotypes of sensitivity to UV radiation and of defective strand-specific NER after exposure to UV light have prompted speculation that defective strand-specific repair (transcriptionally coupled repair) is a primary defect in this disease (31 ). It has been suggested that spontaneous base damage associated with oxidative metabolism may be concentrated in cells with unusually high metabolic rates which proliferate rapidly during development, leading to the organ-specific defects observed in the disease (31 ). In support of this hypothesis, both CS-B cells (15 ) and more recently cells from individuals with the combined XP-G/CS complex (16 ) have been shown to be defective in strand-specific repair following exposure to ionizing radiation. Using a monoclonal antibody that specifically recognizes the oxidatively damaged base thymine glycol, it has been observed that CS-B cells remove thymine glycols from the overall genome as efficiently as normal cells (16 ). However, whereas thymine glycols are preferentially lost from the transcribed strand of the transcriptionally active MTIA and MTIIA genes in normal and XP-A cells, CS-B cells and cells from the combined XP-G/CS complex are defective in this process (16 ). These observations have led to the suggestion of a unique transcription-coupled excision repair process which requires at least the CSB and XP-G proteins (16 ).
ACKNOWLEDGEMENTS
We thank E.Lipinski and M.Sunesen for help with the EndoIII and in vitro DNA repair experiments. We also thank Nancy Tappe and Marzi Ranjbaran for expert technical assistance, Drs Roger Kornberg and Robert Tjian for various promoter plasmid constructs and our laboratory colleagues for general discussions and for critical reading of the manuscript.
13 van Hoffen,A., Natarajan,A.T., Mayne,L.V., van Zeeland,A.A., Mullenders,L.H.F., and Venema,J. (1993) Nucleic Acids Res., 21, 5890-5895.MEDLINE Abstract
14 Mullenders,L.H.F., Sakker,R.J., van Hoffen,A., Venema,J., Natarajan,A.T. and van Zeeland,A.A. (1992) DNA Repair Mechanisms. Munksgaard, Copenhagen, Denmark.
15 Leadon,S.A. and Cooper,P.K. (1993) Proc. Natl. Acad. Sci. USA, 90, 10499-10503.MEDLINE Abstract
