Yeast DNA Repair Protein RAD23 Promotes Complex Formation between Transcription Factor TFIIH and DNA Damage Recognition Factor RAD14
1995; Elsevier BV; Volume: 270; Issue: 15 Linguagem: Inglês
10.1074/jbc.270.15.8385
ISSN1083-351X
AutoresSami N. Guzder, Véronique Bailly, Patrick Sung, Louise Prakash, Satya Prakash,
Tópico(s)Bacterial Genetics and Biotechnology
ResumoIn Saccharomyces cerevisiae, the multisubunit RNA polymerase II general transcription factor TFIIH is indispensable for transcription initiation and some of its subunits are known to be required for nucleotide excision repair (NER) of DNA damaged by ultraviolet light. RAD3, a subunit of TFIIH, binds UV-damaged DNA in an ATP-dependent manner. It has, however, remained unclear how TFIIH is assembled with the other damage recognition component RAD14. Here, we demonstrate a higher order complex consisting of TFIIH, RAD14, and another NER protein RAD23, and complex formation between TFIIH and RAD14 is facilitated by the RAD23 protein. In Saccharomyces cerevisiae, the multisubunit RNA polymerase II general transcription factor TFIIH is indispensable for transcription initiation and some of its subunits are known to be required for nucleotide excision repair (NER) of DNA damaged by ultraviolet light. RAD3, a subunit of TFIIH, binds UV-damaged DNA in an ATP-dependent manner. It has, however, remained unclear how TFIIH is assembled with the other damage recognition component RAD14. Here, we demonstrate a higher order complex consisting of TFIIH, RAD14, and another NER protein RAD23, and complex formation between TFIIH and RAD14 is facilitated by the RAD23 protein. Extensive genetic studies in Saccharomyces cerevisiae have indicated the requirement of 11 genes, RAD1, RAD2, RAD3, RAD4, RAD7, RAD10, RAD14, RAD16, RAD23, RAD25, and MMS19 in nucleotide excision repair (NER). 1The abbreviations used are:NERnucleotide excision repairMOPS3-(N-morpholino)propanesulfonic acidBSAbovine serum albuminPAGEpolyacrylamide gel electrophoresis. 1The abbreviations used are:NERnucleotide excision repairMOPS3-(N-morpholino)propanesulfonic acidBSAbovine serum albuminPAGEpolyacrylamide gel electrophoresis. Among these genes, RAD3 and RAD25 are of particular interest because in addition to their role in NER, they are essential for cell viability (1Prakash S. Sung P. Prakash L. Annu. Rev. Genet. 1993; 27: 33-70Crossref PubMed Scopus (256) Google Scholar). Studies with temperature-sensitive conditional lethal mutations have indicated a direct and essential role of RAD3 and RAD25 in RNA polymerase II transcription (2Guzder S.N. Qiu H. Sommers C.H. Sung P. Prakash L. Prakash S. Nature. 1994; 367: 91-94Crossref PubMed Scopus (121) Google Scholar, 3Guzder S. Sung P. Bailly V. Prakash L. Prakash S. Nature. 1994; 369: 578-581Crossref PubMed Scopus (159) Google Scholar, 4Qiu H. Park E. Prakash L. Prakash S. Genes & Dev. 1993; 7: 2161-2171Crossref PubMed Scopus (73) Google Scholar). Both RAD3 and RAD25 proteins contain single-stranded DNA-dependent ATPase and DNA helicase activities (3Guzder S. Sung P. Bailly V. Prakash L. Prakash S. Nature. 1994; 369: 578-581Crossref PubMed Scopus (159) Google Scholar, 5Sung P. Prakash L. Matson S.W. Prakash S. Proc. Natl. Acad. Sci. U. S. A. 1987; 84: 8951-8955Crossref PubMed Scopus (167) Google Scholar, 6Sung P. Prakash L. Weber S. Prakash S. Proc. Natl. Acad. Sci. U. S. A. 1987; 84: 6045-6049Crossref PubMed Scopus (90) Google Scholar). The DNA helicase activity of RAD3 is required for NER but is dispensable for polymerase II transcription (7Sung P. Higgins D. Prakash L. Prakash S. EMBO J. 1988; 7: 3263-3269Crossref PubMed Scopus (221) Google Scholar). In contrast, the DNA helicase activity of RAD25 is essential for both transcription and DNA repair (3Guzder S. Sung P. Bailly V. Prakash L. Prakash S. Nature. 1994; 369: 578-581Crossref PubMed Scopus (159) Google Scholar, 4Qiu H. Park E. Prakash L. Prakash S. Genes & Dev. 1993; 7: 2161-2171Crossref PubMed Scopus (73) Google Scholar). RAD3 and RAD25 are components of the yeast polymerase II general transcription factor TFIIH. In addition, TFIIH contains four other subunits of 75, 55, 50, and 38 kDa (8Feaver W.J. Svejstrup J.Q. Bardwell L. Bardwell A.J. Buratowski S. Gulyas K.D. Donahue T.F. Friedberg E.C. Kornberg R.D. Cell. 1993; 75: 1379-1387Abstract Full Text PDF PubMed Scopus (281) Google Scholar). TFB1 and SSL1 encode the 75- and 50-kDa subunits, whereas the genes for the 55- and 38-kDa subunits have not yet been identified. A role of TFIIH in NER has been inferred from the observation that TFIIH corrects the NER defect in rad3 and rad25 mutant extracts (9Wang Z. Svejstrup J.Q. Feaver W.J. Wu X. Kornberg R.D. Friedberg E.C. Nature. 1994; 368: 74-75Crossref PubMed Scopus (137) Google Scholar). nucleotide excision repair 3-(N-morpholino)propanesulfonic acid bovine serum albumin polyacrylamide gel electrophoresis. nucleotide excision repair 3-(N-morpholino)propanesulfonic acid bovine serum albumin polyacrylamide gel electrophoresis. Recognition of DNA damage represents the first crucial step in NER. We have previously shown that RAD14, a zinc metalloprotein, binds specifically to ultraviolet-damaged DNA (10Guzder S.N. Sung P. Prakash L. Prakash S. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 5433-5437Crossref PubMed Scopus (88) Google Scholar). Interestingly, RAD3 also binds preferentially to UV-damaged DNA in a manner dependent upon ATP and negative superhelicity (11Sung P. Watkins J.F. Prakash L. Prakash S. J. Biol. Chem. 1994; 269: 8303-8308Abstract Full Text PDF PubMed Google Scholar). The rad3 Arg-48 mutant protein defective in DNA helicase activity also binds UV damaged DNA like the wild type RAD3 protein, indicating that DNA helicase activity and damage binding are two distinct and separable functions in RAD3 (11Sung P. Watkins J.F. Prakash L. Prakash S. J. Biol. Chem. 1994; 269: 8303-8308Abstract Full Text PDF PubMed Google Scholar). Because RAD3 is a damage recognition protein, it is important to determine how TFIIH is assembled with RAD14 2We have observed that the RAD14 gene contains an intron, representing the first example of an intron in an S. cerevisiae DNA repair gene. RAD14 encodes a protein of 371 amino acids with a predicted size of 43 kDa, and RAD14 immunoprecipitated from RAD+ cells exhibits a size of 48 kDa in SDS-PAGE. The translation-initiating ATG codon in RAD14 is at position −456 in the previously reported sequence (19Bankmann M. Prakash L. Prakash S. Nature. 1992; 355: 555-558Crossref PubMed Scopus (97) Google Scholar), and an 84-base pair intron occurs between positions −429 and −346 (19Bankmann M. Prakash L. Prakash S. Nature. 1992; 355: 555-558Crossref PubMed Scopus (97) Google Scholar). The presence of the intron was confirmed by sequence analysis of reverse transcriptase-polymerase chain reaction product of poly(A)+ mRNA isolated from wild type and rna2-1 strains held at 25°C or at 37°C for 1 h. 2We have observed that the RAD14 gene contains an intron, representing the first example of an intron in an S. cerevisiae DNA repair gene. RAD14 encodes a protein of 371 amino acids with a predicted size of 43 kDa, and RAD14 immunoprecipitated from RAD+ cells exhibits a size of 48 kDa in SDS-PAGE. The translation-initiating ATG codon in RAD14 is at position −456 in the previously reported sequence (19Bankmann M. Prakash L. Prakash S. Nature. 1992; 355: 555-558Crossref PubMed Scopus (97) Google Scholar), and an 84-base pair intron occurs between positions −429 and −346 (19Bankmann M. Prakash L. Prakash S. Nature. 1992; 355: 555-558Crossref PubMed Scopus (97) Google Scholar). The presence of the intron was confirmed by sequence analysis of reverse transcriptase-polymerase chain reaction product of poly(A)+ mRNA isolated from wild type and rna2-1 strains held at 25°C or at 37°C for 1 h. that also functions in damage recognition. Here, we show that TFIIH is complexed with RAD14 via the RAD23 protein. We discuss the possible role of this complex in NER. Antibodies to RAD3 (6Sung P. Prakash L. Weber S. Prakash S. Proc. Natl. Acad. Sci. U. S. A. 1987; 84: 6045-6049Crossref PubMed Scopus (90) Google Scholar), RAD14 (10Guzder S.N. Sung P. Prakash L. Prakash S. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 5433-5437Crossref PubMed Scopus (88) Google Scholar), RAD23 (12Watkins J.F. Sung P. Prakash L. Prakash S. Mol. Cell. Biol. 1993; 13: 7757-7765Crossref PubMed Scopus (215) Google Scholar), and RAD25 (3Guzder S. Sung P. Bailly V. Prakash L. Prakash S. Nature. 1994; 369: 578-581Crossref PubMed Scopus (159) Google Scholar) were raised in rabbits as described. Antibodies to TFB1 were raised against a GST-TFB1 hybrid protein (13Gileadi O. Feaver W.J. Kornberg R.D. Science. 1992; 257: 1389-1392Crossref PubMed Scopus (53) Google Scholar) that was expressed in Escherichia coli and purified from inclusion bodies by preparative denaturing polyacrylamide gel electrophoresis. All the antibodies were purified by affinity chromatography as described (14Bailly V. Sommers C.H. Sung P. Prakash L. Prakash S. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 8273-8277Crossref PubMed Scopus (56) Google Scholar). Yeast extracts were prepared in buffer B (50 m M Tris-HCl, pH 7.5, 50 m M NaCl, 0.2% Triton X-100, and protease inhibitors) using a French press as described previously (14Bailly V. Sommers C.H. Sung P. Prakash L. Prakash S. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 8273-8277Crossref PubMed Scopus (56) Google Scholar). Extract from 0.7 g of cells was mixed for 60 min at 25°C with 30 μl of protein A-agarose beads containing 3 mg/ml of covalently conjugated anti-RAD14 antibodies or non-immune IgG. After being washed twice with 600 μl of buffer B, bound proteins were eluted from the immunoprecipitate by a 5-min treatment with 70 μl of 2% SDS at 42°C, and 10 μl of the SDS eluates were analyzed by immunoblotting. RAD23 protein (2 mg) was dialyzed against 1 liter of coupling buffer (0.1 M MOPS, pH 7.5) at 4°C for 12 h. Affi-Gel 15 matrix (Bio-Rad), 0.5 ml, previously washed with cold water, was mixed with purified RAD23 protein in a final volume of 1 ml overnight at 4°C. The unreacted active groups on the matrix were blocked by incubation with 1 M ethanolamine, pH 8.0, for 4 h at 4°C. Bovine serum albumin (4 mg) was coupled to Affi-Gel 15 using the same procedure. 35S-Labeled RAD3, RAD14, RAD25, and TFB1 proteins were obtained by coupled in vitro transcription and translation in 50-μl reactions containing 40 μCi of [35S]methionine with the use of the TNT T7 reticulocyte lysate system (Promega). The 35S-labeled translation products were partially purified by precipitation with 50% ammonium sulfate (0.31 g/ml). The protein pellet was dissolved in 50 μl buffer A (20 m M HEPES-KOH, pH 7.5, 70 m M KCl, 5 m M sodium bisulfite, 4 m M MgCl2, 0.5 m M EDTA, 1 m M dithiothreitol, 0.1% Tween 20, and 2.5% glycerol). A 10-μl aliquot of the protein solution was mixed with 90 μl of buffer A and 10 μl of BSA-Affi-Gel or RAD23-Affi-Gel beads for 30 min at 25°C. After washing with 500 μl of ice-cold buffer A, bound proteins were eluted from the beads with the use of 2% SDS and fractionated in 9% denaturing polyacrylamide gels. The gels were dried onto Whatman 3MM paper and subjected to fluorography. Extract was prepared from 80 g of the rad23Δ yeast strain JWY36 with the use of a French press and passed through a Bio-Rex 70 column (1.6 × 8 cm) as described (15Sayer M.H. Tschochner H. Kornberg R.D. J. Biol. Chem. 1992; 267: 23376-23382Abstract Full Text PDF PubMed Google Scholar). After washing the Bio-Rex 70 column with buffer A containing 300 m M KOAc, TFIIH and other proteins were eluted with 600 m M KOAc (15Sayer M.H. Tschochner H. Kornberg R.D. J. Biol. Chem. 1992; 267: 23376-23382Abstract Full Text PDF PubMed Google Scholar). Fractions corresponding to the protein peak were pooled (5 ml), concentrated to 2 ml in a Centricon-30 (Amicon), and dialyzed overnight against 1 liter of buffer C (50 m M Tris-HCl, pH 7.5, 0.1 m M EDTA, 5% glycerol, 1 m M dithiothreitol, and protease inhibitors containing 50 m M KOAc). RAD23-Affi-Gel and BSA-Affi-Gel, 0.15 ml each, were packed in a 1-ml pipette tip plugged with glass wool and 1 ml of the dialyzed TFIIH pool was passed through the RAD23 affinity column or the BSA control column twice at 25°C. The columns were eluted with 0.45 ml of 0.2 M, 0.5 M, 1 M, and 2 M KOAc in buffer C, collecting two 0.225-ml fractions during each elution step. Five μl of the starting Bio-Rex 70 TFIIH pool of the flow-through fraction from the RAD23-Affi-Gel and BSA-Affi-Gel columns and of the various salt eluates were examined for their content of the RAD25, RAD3, and the TFB1 proteins. Extracts for transcription were prepared as described previously (16Woontner M. Wade P.A. Bonner J. Jaehning J.A. Mol. Cell. Biol. 1991; 11: 4555-4560Crossref PubMed Scopus (79) Google Scholar). The template for transcription, pSL187, contains the promoter of the yeast CYC1 gene and yields transcripts of 375 and 350 nucleotides (17Koleske A. Buratowski S. Nonet M. Young R.A. Cell. 1992; 69: 883-894Abstract Full Text PDF PubMed Scopus (139) Google Scholar). Reaction mixtures were assembled and processed as described (2Guzder S.N. Qiu H. Sommers C.H. Sung P. Prakash L. Prakash S. Nature. 1994; 367: 91-94Crossref PubMed Scopus (121) Google Scholar, 3Guzder S. Sung P. Bailly V. Prakash L. Prakash S. Nature. 1994; 369: 578-581Crossref PubMed Scopus (159) Google Scholar). To investigate whether RAD14 protein forms a complex with TFIIH, affinity-purified antibodies to RAD14 (10Guzder S.N. Sung P. Prakash L. Prakash S. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 5433-5437Crossref PubMed Scopus (88) Google Scholar) were covalently conjugated to protein A-agarose, and the resulting immunobeads were mixed with wild type RAD+ yeast extract at 25°C for 60 min to bind RAD14 and proteins that are associated with RAD14. As control, the extract was also incubated with protein A-agarose beads containing non-immune IgG. After washing the immunoprecipitates with a large volume of buffer, bound proteins were eluted with SDS and analyzed by immunoblotting with the appropriate antibodies for RAD14 and the TFIIH components RAD3, RAD25, and TFB1. We found that the amount of the TFIIH components associated with the anti-RAD14 immunobeads was ≈6-fold higher than the background level of these components in the control (Fig. 1, compare lanes 3 and 4). Interestingly, another NER protein, RAD23 (1Prakash S. Sung P. Prakash L. Annu. Rev. Genet. 1993; 27: 33-70Crossref PubMed Scopus (256) Google Scholar, 12Watkins J.F. Sung P. Prakash L. Prakash S. Mol. Cell. Biol. 1993; 13: 7757-7765Crossref PubMed Scopus (215) Google Scholar), also co-precipitated specifically with RAD14, as the level of RAD23 in the anti-RAD14 immunoprecipitate was ≈6-fold higher than the control (Fig. 1, compare lanes 3 and 4). These results suggest the existence of a higher order protein complex in wild type yeast cells consisting of TFIIH, RAD14, and RAD23. Mutations in the RAD23 gene greatly compromise the efficiency of NER (18Miller R.D. Prakash L. Prakash S. Mol. & Gen. Genet. 1982; 188: 235-239Crossref PubMed Scopus (49) Google Scholar), and we have previously suggested that RAD23 protein is a non-catalytic NER component that could act in the assembly of a functional nucleotide excision repair complex (1Prakash S. Sung P. Prakash L. Annu. Rev. Genet. 1993; 27: 33-70Crossref PubMed Scopus (256) Google Scholar, 12Watkins J.F. Sung P. Prakash L. Prakash S. Mol. Cell. Biol. 1993; 13: 7757-7765Crossref PubMed Scopus (215) Google Scholar). To directly test this possibility, we carried out anti-RAD14 immunoprecipitation using extracts prepared from a yeast strain lacking the genomic RAD23 gene. In the absence of RAD23 protein, as in the case of rad23Δ extract, the amount of TFB1, RAD3, and RAD25 proteins bound in the anti-RAD14 immunoprecipitate was only slightly higher (20-40%) than the background level of these proteins that were associated nonspecifically with beads containing the non-immune IgG (Fig. 1, compare lanes 1 and 2). Thus, co-precipitation of TFIIH with RAD14 protein is strongly dependent on the RAD23 protein, lending support to the notion that efficient assembly of the complex of NER proteins requires the RAD23 protein. To further establish the role of RAD23 in complex formation, we purified this protein from yeast cells. To facilitate the purification, the RAD23 gene was joined to the highly expressed yeast ADC1 promoter to yield the multicopy plasmid pJW112 (2μ,ADC1-RAD23). Purification of RAD23 from the protease-deficient yeast strain LP2749-9B harboring pJW112 was achieved by a combination of ammonium sulfate precipitation and chromatographic fractionation in columns of Q-Sepharose, hydroxylapatite, and Mono Q. The purified RAD23 protein was analyzed by SDS-PAGE and staining with Coomassie Blue, which revealed that the protein preparation was nearly homogeneous (Fig. 2A). We obtained 5 mg of RAD23 protein from 200 g of starting yeast paste. With the use of nitrocellulose filter DNA binding assay and DNA mobility shift assay in agarose gels, using a wide pH range, we found no interaction of RAD23 protein with DNA. We also found no ATPase or nuclease activity in RAD23. Our immunoprecipitation studies indicated that RAD23 is part of a higher order complex of excision repair proteins and TFIIH and that it in fact promotes the assembly of this protein complex. To determine whether RAD23 contacts TFIIH and RAD14 directly or does so via some other protein component(s), we covalently conjugated purified RAD23 protein to Affi-Gel-15 and used the resulting RAD23 matrix as affinity beads for binding 35S-labeled TFB1, RAD3, RAD25, and RAD14 proteins. To obtain radiolabeled proteins for this work, the protein coding frames of the TFB1, RAD3, RAD25, and RAD14 genes were placed under the bacteriophage T7 promoter, and the resulting constructs were transcribed in vitro to obtain mRNAs that code for these proteins, followed by translation of the mRNAs in rabbit reticulocyte lysate in the presence of [35S]methionine. The radiolabeled proteins thus obtained were partially purified by ammonium sulfate precipitation, dissolved in reaction buffer, and mixed with the RAD23 Affi-Gel-15 beads. Affinity binding to the RAD23 matrix was allowed to proceed at 25°C for 30 min. After washing with binding buffer, the bound 35S-labeled proteins were eluted from the RAD23 Affi-Gel beads with the use of SDS and revealed by fluorography after denaturing polyacrylamide gel electrophoresis. As a control in these experiments, we also mixed the 35S-labeled proteins with Affi-Gel-15 beads containing per unit volume of matrix an amount of BSA twice that of RAD23 used. The BSA beads were treated with SDS, and the eluates were run in polyacrylamide gels alongside the SDS eluates from the RAD23 beads. As shown in Fig. 2 B, the level of the 35S-labeled TFB1 and RAD14 proteins in the SDS eluates from the RAD23 affinity beads was ∼10- and ∼7-fold, respectively, of that in the BSA control, indicating a specific and direct interaction of TFB1 and RAD14 with RAD23. The amount of 35S-labeled RAD25 protein bound to the RAD23 beads was about 3-fold higher than to the BSA beads, suggesting an interaction between RAD25 and RAD23 proteins as well (Fig. 2B). Reproducibly, in three separate experiments (Fig. 2B and data not shown), the level of RAD3 bound to the RAD23 beads was the same as that found in the control, indicating that these two proteins do not interact directly. To obtain further evidence that RAD23 and TFIIH interact physically, a Bio-Rex 70 column fraction derived from rad23Δ extract that was enriched in TFIIH was passed through a column of RAD23 Affi-Gel-15, and BSA Affi-Gel-15 was used as a control. The RAD23 column and control BSA column were washed with buffer and then eluted with 0.2 M, 0.5 M, 1 M, and 2 M potassium acetate, and the content of the TFIIH components RAD3, RAD25, and TFB1 in the various salt washes was examined by immunoblotting. The results from this analysis indicate that a sizable proportion (>70%) of TFB1, RAD3, and RAD25 proteins were retained on the RAD23-Affi-Gel column and that these proteins are eluted from the RAD23 column from 0.2-2 M acetate (Fig. 3A). Only a trace amount of these proteins bound nonselectively to the control BSA column, and all of the retained proteins were readily eluted by 0.2 M acetate (Fig. 3A). We have previously described conditional lethal mutations of RAD3 and RAD25 which result in defective RNA polymerase II transcription at the restrictive temperature both in vivo and in vitro (2Guzder S.N. Qiu H. Sommers C.H. Sung P. Prakash L. Prakash S. Nature. 1994; 367: 91-94Crossref PubMed Scopus (121) Google Scholar, 3Guzder S. Sung P. Bailly V. Prakash L. Prakash S. Nature. 1994; 369: 578-581Crossref PubMed Scopus (159) Google Scholar–4Qiu H. Park E. Prakash L. Prakash S. Genes & Dev. 1993; 7: 2161-2171Crossref PubMed Scopus (73) Google Scholar). As shown in Fig. 3B, the transcriptional defect in the rad3-ts14 and the rad25-ts24 extracts can be complemented specifically by the eluate from the RAD23 affinity column. These results are again consistent with interaction of TFIIH with the RAD23 protein. The cloning of the RAD14 and RAD23 genes allowed us to determine whether they are essential for cell viability besides their known role in nucleotide excision repair. Yeast strains bearing genomic deletions of these genes show no notable growth deficiency at 30°C (12Watkins J.F. Sung P. Prakash L. Prakash S. Mol. Cell. Biol. 1993; 13: 7757-7765Crossref PubMed Scopus (215) Google Scholar, 19Bankmann M. Prakash L. Prakash S. Nature. 1992; 355: 555-558Crossref PubMed Scopus (97) Google Scholar) or at 37°C, 3S. N. Guzder, V. Bailly, P. Sung, L. Prakash, and S. Prakash, unpublished observations. indicating that they are likely not required for RNA polymerase II transcription. In agreement with the genetic data, we found that extracts prepared from the rad14Δ and the rad23Δ strains are as proficient in RNA polymerase II transcription as the wild type extract, regardless of whether the mutant extracts were subjected to high temperature treatment at 37°C for 5 min prior to the transcription reaction (Fig. 4). The requirement for RAD14 and RAD23 proteins in NER, but not in RNA polymerase II transcription, stands in contrast with RAD3 and RAD25 proteins, which we have shown to be indispensable for either process (2Guzder S.N. Qiu H. Sommers C.H. Sung P. Prakash L. Prakash S. Nature. 1994; 367: 91-94Crossref PubMed Scopus (121) Google Scholar, 3Guzder S. Sung P. Bailly V. Prakash L. Prakash S. Nature. 1994; 369: 578-581Crossref PubMed Scopus (159) Google Scholar, 4Qiu H. Park E. Prakash L. Prakash S. Genes & Dev. 1993; 7: 2161-2171Crossref PubMed Scopus (73) Google Scholar). Yeast TFIIH consists of RAD3, RAD25, TFB1, SSL1, and two other as yet uncharacterized proteins with molecular sizes of 38 and 55 kDa (8Feaver W.J. Svejstrup J.Q. Bardwell L. Bardwell A.J. Buratowski S. Gulyas K.D. Donahue T.F. Friedberg E.C. Kornberg R.D. Cell. 1993; 75: 1379-1387Abstract Full Text PDF PubMed Scopus (281) Google Scholar). Genetic and biochemical studies have indicated a direct role of RAD3 and RAD25 proteins in both transcription and nucleotide excision repair, and multiple rad3 and rad25 mutant alleles that are differentially inactivated for either their repair or transcriptional function have been isolated (1Prakash S. Sung P. Prakash L. Annu. Rev. Genet. 1993; 27: 33-70Crossref PubMed Scopus (256) Google Scholar, 2Guzder S.N. Qiu H. Sommers C.H. Sung P. Prakash L. Prakash S. Nature. 1994; 367: 91-94Crossref PubMed Scopus (121) Google Scholar, 3Guzder S. Sung P. Bailly V. Prakash L. Prakash S. Nature. 1994; 369: 578-581Crossref PubMed Scopus (159) Google Scholar, 4Qiu H. Park E. Prakash L. Prakash S. Genes & Dev. 1993; 7: 2161-2171Crossref PubMed Scopus (73) Google Scholar). In this study, we demonstrate that the NER proteins RAD14 and RAD23 are associated with TFIIH as indicated by their co-immunoprecipitation from wild type yeast extracts and that formation of this complex is modulated by RAD23. To examine the role of RAD23 in mediating complex formation, we coupled purified RAD23 to Affi-Gel-15 and used it as affinity matrix for binding in vitro translated TFIIH subunits and RAD14. We found that RAD23 interacts directly with the TFIIH subunits TFB1 and RAD25 and with RAD14. The RAD14 protein functions in damage recognition (10Guzder S.N. Sung P. Prakash L. Prakash S. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 5433-5437Crossref PubMed Scopus (88) Google Scholar). More recently, we have shown that RAD3 binds UV-damaged DNA in an ATP-dependent manner (11Sung P. Watkins J.F. Prakash L. Prakash S. J. Biol. Chem. 1994; 269: 8303-8308Abstract Full Text PDF PubMed Google Scholar). Our study identifies RAD23 protein as an intermediary that promotes association of TFIIH with RAD14. It is possible that the TFIIH-RAD23-RAD14 complex has a higher affinity for UV damaged DNA than can be achieved by the individual components. Both the RAD3 (5Sung P. Prakash L. Matson S.W. Prakash S. Proc. Natl. Acad. Sci. U. S. A. 1987; 84: 8951-8955Crossref PubMed Scopus (167) Google Scholar) and RAD25 (3Guzder S. Sung P. Bailly V. Prakash L. Prakash S. Nature. 1994; 369: 578-581Crossref PubMed Scopus (159) Google Scholar) subunits of TFIIH possess a DNA helicase activity that may be utilized for effecting an open conformation of the damaged helix for dual incision by the RAD1-RAD10 and RAD2 endonucleases (20Sung P. Reynolds P. Prakash L. Prakash S. J. Biol. Chem. 1993; 268: 26391-26399Abstract Full Text PDF PubMed Google Scholar, 21Habraken Y. Sung P. Prakash L. Prakash S. Nature. 1993; 366: 365-368Crossref PubMed Scopus (114) Google Scholar, 22Tomkinson A.E. Bardwell A.J. Bardwell L. Tappe N.J. Friedberg E.C. Nature. 1993; 362: 860-862Crossref PubMed Scopus (155) Google Scholar). In addition, the combined helicase action of RAD3 and RAD25 proteins may be essential for post-incision turnover of the NER complex and the damage containing DNA fragment, as our previous studies have suggested a role of RAD3 helicase in the post-incision step (7Sung P. Higgins D. Prakash L. Prakash S. EMBO J. 1988; 7: 3263-3269Crossref PubMed Scopus (221) Google Scholar). Although RAD23 has no known catalytic function and does not bind DNA, via its role as an assembly factor, it could facilitate the efficient recognition of the DNA lesion and perhaps influence other phases of NER. We thank J. Watkins for the initial work on RAD23 protein; T. Wood, D. Prusak, and C. Kodira for reverse transcriptase-polymerase chain reaction and sequence determination; W. J. Feaver and R. D. Kornberg for the plasmid that expresses the GST-TFB1 hybrid protein; and J. Woolford for the rna2-1 strain. Addendum-After the preparation of this manuscript, it was reported that several NER proteins associate with TFIIH because they were present in chromatographic column fractions that contained TFIIH (23Svejstrup J.Q. Wang Z. Feaver W.J. Wu X. Bushnell D.A. Donahue T.F. Friedberg E.C. Kornberg R.D. Cell. 1995; 80: 21-28Abstract Full Text PDF PubMed Scopus (238) Google Scholar).
Referência(s)