Affinity Purification of Human DNA Repair/Transcription Factor TFIIH Using Epitope-tagged Xeroderma Pigmentosum B Protein*

TFIIH is a high molecular weight complex with a remarkable dual function in nucleotide excision repair and initiation of RNA polymerase II transcription. Mutations in the largest subunits, the XPB and XPD helicases, are associated with three inherited disorders: xeroderma pigmentosum, Cockayne’s syndrome, and trichothiodystrophy. To facilitate the purification and biochemical characterization of this intricate complex, we generated a cell line stably expressing tagged XPB, allowing the immunopurification of the XPB protein and associated factors. Addition of two tags, a N-terminal hexameric histidine stretch and a C-terminal hemagglutinin epitope, to this highly conserved protein did not interfere with its functioning in repair and transcription. The hemagglutinin epitope allowed efficient TFIIH immunopurification to homogeneity from a fractionated whole cell extract in essentially one step. We conclude that the predominant active form of TFIIH is composed of nine subunits and that there is one molecule of XPB per TFIIH complex. The affinity-purified complex exhibits all expected TFIIH activities: DNA-dependent ATPase, helicase, C-terminal domain kinase, and participation inin vitro and in vivo nucleotide excision repair and in vitro transcription. The affinity purification procedure described here is fast and simple, does not require extensive chromatographic procedures, and yields highly purified, active TFIIH.

Nucleotide excision repair (NER) 1 is a versatile DNA repair mechanism that removes a wide variety of lesions, such as UV-induced lesions and numerous chemical adducts (1,2). The principal steps in the reaction mechanism of NER are recognition and demarcation of the lesion, probably involving chromatin remodelling and local helix opening, incision of the DNA on both sides of the lesion at some distance, removal of the damaged oligonucleotide, and, finally, repair DNA synthesis and ligation. In eukaryotes, this reaction requires about 30 poly-peptides and has been reconstituted with purified components, including XPA, XPC/HHR23B, replication protein A, the structure-specific nucleases ERCC1/XPF and XPG, and the multisubunit basal transcription factor TFIIH (3)(4)(5). At least two subpathways can be discerned in the NER system. One of these, transcription-coupled repair, preferentially removes DNA damage from the transcribed strand of active genes, whereas lesions in the rest of the genome are repaired more slowly and less efficiently by the global genome repair pathway. TFIIH appears to be a core component of both excision subpathways.
Mutations in the two largest subunits of TFIIH, the XPB and XPD helicases, are associated with the rare genetically heterogeneous disorders xeroderma pigmentosum (XP), Cockayne's syndrome (CS), and trichothiodystrophy (TTD) (6,7). Many complementation groups and considerable overlap have been established for these syndromes: seven complementation groups in XP (XP-A-XP-G), three of which include patients with combined XP and CS phenotypes (XP-B, XP-D and XP-G), two in the classical form of CS (CS-A and CS-B), and three in TTD (XP-B, XP-D, and TTD-A). The discovery of the dual function of XPB and XPD in both NER and transcription provides a rationale for the complex clinical features that are specifically associated with inherited defects in TFIIH subunits, such as seen in the combined XP/CS and the photosensitive form of TTD, that were difficult to explain solely on the basis of a NER defect. Thus, it was proposed that typical XP characteristics, such as UV-induced cutaneous abnormalities and predisposition to skin cancer, are due to inactivation of the NER function of TFIIH, whereas features typical for CS and/or TTD, such as neurodysmyelination, brittle hair, and growth defects, are due to a deficiency in the transcription function of TFIIH, possibly affecting only a subset of genes (8).
In addition to XPB and XPD, which exhibit DNA-dependent ATPase activities and are 3Ј-5Ј and 5Ј-3Ј DNA helicases, respectively (9,10), seven more TFIIH subunits have been identified to date. CDK7 was identified as the catalytic subunit of the kinase activity of TFIIH that is able to phosphorylate the C-terminal domain (CTD) of the largest subunit of RNA polymerase II (11). Interestingly, CDK7 also constitutes a separate trimeric kinase complex that is possibly involved in cell cycle regulation together with the cyclin H and MAT1 subunits of TFIIH (12,13). Furthermore, p44, the human homologue of yeast SSL1, and p34 contain zinc finger domains and possess putative DNA binding capacity (14). So far, no activity has been detected for the p62 and p52 subunits (15,16). Whether these nine proteins constitute the TFIIH complex and whether the composition of TFIIH differs during NER and transcription initiation are yet unresolved issues.
The presence of two DNA helicases has implicated TFIIH in a helix-opening step during both transcription initiation and NER. It has been shown that such open-complex formation at the transcription start site depends on TFIIH and that the requirement for TFIIH is dependent on promoter topology and can be alleviated by premelted regions at the transcription start site (17)(18)(19). During NER, TFIIH is thought to convert a recognized damaged site into a substrate for the XPG and XPF/ERCC1 structure-specific nucleases by locally opening DNA around a lesion. The formation of an opened DNA conformation around a recognized lesion has been demonstrated; however, the direct involvement of TFIIH in this step has not been shown (20). Answers to these questions are hampered by the difficulty in obtaining large quantities of highly purified TFIIH. Therefore, we developed a procedure that facilitates the isolation of TFIIH using a human cell line expressing functional XPB provided with two tags. The affinity purification procedure described here is fast and simple, does not require extensive chromatographic procedures, and yields highly purified, active TFIIH.

EXPERIMENTAL PROCEDURES
General-Purification of nucleic acids, restriction enzyme digestion, gel electrophoresis of nucleic acids and proteins, immunoblotting, detection of proteins and nucleic acids were performed according to standard procedures (21).
Oligonucleotides and Plasmid DNA Construction-The coding sequence for the C-terminal HA epitope tag was added via PCR using oligonucleotide primer pairs p90 5Ј-CCCGGATCCTCAGCTAGCGTA-ATCTGGAACATCGTATGGGTATTTCCTAAAGCGCTTGAAG-3Ј (3Ј primer; underlined sequence encodes the HA epitope, and double underlined sequence indicates a BamHI restriction site) and p33 5Ј-GGATCCACCATGGGCAAAAGAGACCG-3Ј (5Ј primer). Likewise, the hexahistidine tag was added using oligonucleotides p123 5Ј-CGCGCG-GAATTCATGGGCAGCAGCCATCATCATCATCATCACAGCAGCGG-CCTGGTGCCGCGCGGCAGCCATATGGGCAAAAGAGACCG-3Ј (5Ј primer; underlined sequence encodes a hexahistidine stretch and thrombin cleavage site, and double underlined sequence indicates an EcoRI site) and p41 5Ј-CGGGAAGTGGAGGGCCCACC-3Ј (3Ј primer). PCR fragments were cloned, sequenced, and confirmed to be free of PCR-introduced sequence errors. Full-length XPB and double-tagged XPB cDNA (dtXPB cDNA) were subcloned as EcoRI-BamHI fragments in a modified pSG5 eukaryotic expression vector yielding plasmids pSHE3 and pSHE3HA, respectively. From pSHE3HA, the dtXPB cDNA was subcloned as a EcoRI-XhoI fragment in the eukaryotic expression vector pcDNA3 (Invitrogen) yielding plasmid pM300.
DNA Transfection and UV Survival Assay-Cells were cultured in a 1:1 mixture of Dulbecco's modified Eagle's medium and Ham's F-10 medium supplemented with 10% fetal calf serum and antibiotics. HeLa TK Ϫ cells were transfected with vector pM300 by electroporation, and after selection with 1.0 mg/ml G418, individual clones were selected for expression of dtXPB by immunoblot analysis using anti-XPB antibodies.
Cell lines XP-wt and XP-t3 were generated by transfection of XPCS2BASV (XP-B) SV40-immortalized fibroblasts with XPB and dtXPB cDNA, respectively. Expression vector pSHE3 or pSHE3HA and a neomycin-selectable marker, were cotransfected using Lipofectin reagent essentially as described by the manufacturer (Life Technologies, Inc.). Stably expressing mass populations were obtained by selection with 300 g/ml G418 and repeated UV irradiation at 1 day intervals (three doses of 4 J/m 2 ).
UV survival was assayed by [methyl-3 H]thymidine incorporation 3 days after irradiation as described (22). Proliferating cells were pulselabeled with [methyl-3 H]thymidine (3 h) and then chased (1 h). Cell lysates were transferred to scintillation-counting vials, and survival was calculated as the average ratio of incorporated radiolabel in treated triplicates to that in four untreated control dishes.
Immunopurification of dtXPB Protein and Associated Factors-XP-t3 cells were cultured in suspension in RPMI 1640 medium supplemented with 10% fetal calf serum and antibiotics. Whole cell extracts (WCEs) (total, 530 mg) were prepared from frozen cell pellets (total, 50 ml packed cell volume) as described (23). Subsequently, WCEs were fractionated on heparin-Ultrogel (IBF, France) equilibrated in Buffer A (10 mM Tris-HCl, pH 7.8, 17% glycerol (v/v), 0.5 mM dithiothreitol, 5 mM MgCl 2 ) containing 0.1 M KCl and eluted with Buffer A containing 0.22, 0.4, and 1.0 M KCl as described (24). All of the dtXPB protein was present in the 0.4 M KCl fraction (designated Hep0.4; 175 mg of protein) as judged by immunoblot analysis similar to TFIIH. Typically, 10 -12 ml (7.5-9.0 mg of protein) of the Hep0.4 fraction was successively incubated with 400 g of purified 12CA5 anti-HA monoclonal antibody bound to 400 l of protein G-Sepharose (Pharmacia Biotech Inc.) overnight at 4°C. The resin was washed three times with 10 volumes of ice-cold buffer T (25 mM Tris-HCl, pH 7.9, 17% glycerol (v/v), 0.5 mM EDTA, 0.2 mM dithiothreitol, 5 mM MgCl 2 ) containing 0.4 M KCl and 0.1% Nonidet P-40 and twice with buffer T/0.1 M KCl containing 0.01% Nonidet P-40. Bound material was eluted for 1 h at 30°C in 400 l of buffer T/0.1 M KCl containing 0.01% Nonidet P-40, 0.2 mg/ml insulin, 2.0 mg/ml synthetic peptide corresponding to the HA epitope (sequence YPYDVPDYA), and 1.0 g/ml aprotinin. This step was repeated once or twice. Routinely, nearly all dtXPB and associated factors were present in the first eluate as detected by immunoblotting. Purified proteins were stored at Ϫ80°C.
The same procedure was used for immunoprecipitation of dtXPB directly from a WCE (15 mg) to analyze protein-protein interactions. However, in this case, bound material was washed five times with buffer containing 0.1 M KCl, 0.1% Nonidet P-40 and analyzed by immunoblotting after overnight elution at 4°C with buffer containing 1.0 mg/ml peptide and 0.1 mg/ml bovine serum albumin (Sigma).
Microneedle Injection-Prior to microinjection, human primary fibroblasts were fused by inactivated Sendai virus as described earlier (8). Protein fractions were microinjected into the cytoplasm of XP polykaryons, and NER activity was measured by pulse labeling with [methyl-3 H]thymidine and in situ autoradiography as described (8). Repair activity was quantified by counting autoradiographic grains above at least 50 non-S phase nuclei. Primary cell lines used were XPCS1BA (XP-B), XP6BE (XP-D), XP126LO (XP-F), XP3BR (XP-G), TTD1BR (TTD-A), and C5RO (wild type).
In Vitro NER Assay-WCEs were prepared from repair-proficient HeLa and XP-t3 cells and repair-deficient SV40-immortalized XPCS2BASV (XP-B) cells (23). NER reactions (50 l) contained 250 ng of plasmid DNA randomly damaged with N-acetoxy-2-acetylaminofluorene and as an internal control an equal amount of undamaged control plasmid of a different size, the indicated amount of extract, and purified proteins; the reactions were incubated for 3 h at 30°C. DNA was purified, linearized with BamHI, and analyzed on a 0.8% agarose gel (25,26). Antibody depletion of extracts was performed as follows: for each reaction, 100 g of WCE was incubated with 0.5 l of anti-p62 ascites and 5 l of protein G-agarose overnight at 4°C. For the experiment shown in Fig. 6A, plasmid DNA was randomly damaged with cis-diaminedichloro-platinum(II) (16).
In Vitro Transcription Assay-Purified TFIIH was incubated with recombinant human TBP, TFIIB, and TFIIE and purified TFIIA, TFIIF, and RNA polymerase II as described earlier (24). After 15 min of preincubation at 25°C with 70 ng of linearized template DNA containing the adenovirus 2 major late promoter, nucleotides were added, and transcription was allowed to proceed in a final reaction volume of 25 l for 45 min at 25°C. The 309-nucleotide runoff transcripts were analyzed by electrophoresis through a 5% acrylamide/50% urea gel.

Generation of a Cell Line Expressing
Tagged XPB-To analyze interactions of the XPB protein with other proteins, including TFIIH under physiological conditions, and to facilitate the purification of active TFIIH, we decided to generate a human cell line stably expressing a tagged version of XPB (dtXPB) cDNA. To permit isolation of full-length XPB protein and allow purification of XPB on the basis of different reversible affinity purification steps, we chose to add two different types of tags, one on each end of the protein. Thus, coding sequences for a N-terminal hexahistidine stretch followed by a thrombin cleavage site and a C-terminal HA epitope tag (28,30) were added to XPB cDNA fragments, and a full-length double-tagged XPB cDNA was constructed (Fig. 1A) and subcloned in eukaryotic expression vectors. To obtain cell lines stably expressing dtXPB, the cDNA vectors were transfected to two human cell lines. First, dtXPB cDNA and a neomycinselectable marker were transfected into HeLa cells. After selection with G418, individual HeLa clones were obtained and analyzed by immunoblotting for the level of the dtXPB protein using anti-XPB antibodies. As shown in Fig. 1B, lanes 1-5, the double-tagged XPB can be conveniently discerned from the endogenous wild type XPB protein because of its increased size (the predicted molecular mass increases from 89,279 Da to 92,690 Da). In the 48 clones analyzed, various levels of dtXPB protein were detected in WCEs, with many clones expressing no or hardly detectable dtXPB, despite the fact that the dtXPB cDNA was under the control of the strong cytomegalovirus promoter, and multiple copies are expected to be integrated in the genome. Interestingly, in neither case did we observe a large overexpression, and clones expressing the highest level of dtXPB clearly showed decreased levels of the wild type endogenous protein as compared with the p62 core subunit of TFIIH (e.g. Fig. 1B, compare clone 1-19 with clones 1-6, 1-2, and 1-7). These findings suggest that the cellular content of XPB is kept within narrow concentration ranges by degrading excess protein and that there is competition between the endogenous wild type and exogenous tagged protein. Secondly, XP-B UV-sensitive cells were transfected, and after selection with G418 and repeated UV irradiation, a stably expressing mass population was established, designated XP-t3. As with the HeLa transfectants, immunoblot analysis of XP-t3 cell extracts indicated that dtXPB protein levels were comparable with the endogenous (mutant) XPB levels (Fig. 1B, lanes 6 -8). Because the relatively high dtXPB protein levels in HeLa clone 1-19 varied during culturing and appeared to be more stable in the transfected XP-B cells, the XP-t3 cell line was further characterized and used for all experiments described here.
To characterize the functionality of the dtXPB protein in vivo, an UV survival experiment was carried out. Fig. 1C shows that dtXPB is able to reverse the UV sensitivity of XP-B cells to the same extent as wild type XPB cDNA, although both not to the same level as the wild type MRC5 transformed cell line used as a repair-proficient control. The NER activity was further analyzed using an in vitro assay. WCEs were prepared and incubated in the presence of both damaged and undamaged plasmid DNA and labeled nucleotides. The resulting incorporation of labeled nucleotides in the damaged plasmid DNA is due to repair DNA synthesis and a measure for NER activity. Fig. 1D shows that extracts prepared from XP-t3 cells were able to repair N-acetoxy-2-acetylaminofluorene-damaged DNA in vitro, whereas extracts prepared from XP-B cells display a strongly reduced repair activity. The above findings indicate that the dtXPB protein is functional in NER, implying that it most likely is incorporated in the TFIIH complex. Finally, the fact that these cells grow normally strongly suggests that the dtXPB protein is not interfering with the basal transcription initiation function of the complex.
Analysis of TFIIH Factors Associated with dtXPB Pro- tein-To identify proteins interacting with XPB, an immunoprecipitation experiment was carried out under physiological salt conditions (0.1 M KCl) using a repair-and transcriptioncompetent XP-t3 WCE. As shown in Fig. 2, the dtXPB protein could be specifically and quantitatively immunoprecipitated using anti-HA monoclonal antibodies, and in addition to dtXPB, we could detect several TFIIH subunits in the bound fraction (XPD, p62, CDK7, and cyclin H). This confirms that dtXPB was incorporated in TFIIH. Furthermore, the fact that the relative intensities of XPB versus p62 are not significantly altered in the load, unbound, and the bound fractions indicates that the established XP-t3 cells harbor dtXPB in the majority of the TFIIH complexes. Interestingly, although all dtXPB protein was depleted from the WCE, none of the endogenous (mutant) XPB was detected in the bound fraction. This demonstrates that only one XPB subunit is present per complex: if the complex contained more than one XPB molecule, complexes with both the endogenous (mutant) subunit and the tagged protein would be expected to be present in the bound material.
Immunopurification of XPB and Associated Factors-Because a number of TFIIH subunits specifically co-immunoprecipitated with dtXPB protein using anti-HA antibodies, we set out to purify TFIIH on this basis and analyze the composition of the TFIIH complex (Fig. 3A). First, WCEs from XP-t3 cells were fractionated on heparin-Ultrogel as described (24). All dtXPB protein, as well as XPB and other known TFIIH subunits, were present in the 0.4 M KCl fraction, designated Hep0.4. Portions of this fraction were directly incubated without further fractionation with anti-HA resin to purify dtXPB and associated proteins. After incubation, the anti-HA resin was extensively washed, and bound material was eluted by competition with excess HA peptide. Subsequently, the compo-sition of the eluate was analyzed by SDS-polyacrylamide gel electrophoresis and staining with silver nitrate. In addition to dtXPB, we identified only eight polypeptides ranging in molecular mass from 80 to 34 kDa that specifically and consistently immunoprecipitated with dtXPB and are thus XPB-associated factors (Fig. 3B). These associated proteins were all identified as known TFIIH subunits by two criteria: (i) reactivity with monoclonal antibodies specifically recognizing TFIIH subunits; and (ii) exact co-migration with known TFIIH subunits in SDSpolyacrylamide gel electrophoresis (data not shown). Finally, the staining intensity of the dtXPB subunit, compared with the other subunits, suggests that the dtXPB protein was predominantly present in stoichiometric amounts and not in a free form.
Presence of Additional NER and Transcription Factors in the Affinity-purified TFIIH Fraction-A number of NER and basal transcription factors (TFIIE, CSA, CSB, and XPG, among others) have been described to interact with the TFIIH complex, either as part of a RNA polymerase II holoenzyme (31, 32) or using isolated proteins (33,34). Therefore, it was unexpected to find only nine polypeptides stained by silver nitrate in the immunopurified fraction. The silver-stained protein profile of affinity-purified TFIIH (Fig. 3B) does not exclude the possibility that substoichiometric amounts of other NER or transcription factors are present. Furthermore, the heparin fractionation might have disrupted salt-sensitive interactions. Therefore, the TFIIH-containing fraction that was immunoprecipitated directly from a WCE using low salt conditions was tested by immunoblot analysis for the presence of additional NER and transcription factors (Fig. 4). However, we were not able to detect the presence of any of the NER factors ERCC1, XPC, HHR23B, XPG, and CSB, RNA polymerase II, or any significant levels of the basal transcription factors tested. The absence of a stable interaction with CSB was confirmed by the reverse experiment, in which tagged CSB was immunoprecipitated and analyzed for the presence of TFIIH subunits (46). As a positive control, we detected the presence of the human homologue of yeast SUG1, a protein that we recently identified to interact with the XPB subunit of TFIIH (35). Notably, human SUG1 was below immunodetection level in the immunopurified fraction shown in Fig. 3B, indicating that this interaction was salt-sensitive under the conditions used. Similar results were obtained using fractionation of the WCE on a Ni-NTA column that has a high affinity for the hexahistidine stretch or by using a nuclear extract preparation (data not shown). The above results indicate that the interactions of TFIIH with the various NER and transcription factors are, at least under the various conditions we used, not stable.
Functional Characterization of Affinity-purified TFIIH-To determine whether the nine polypeptides identified in the silver-stained gel represent the active form of TFIIH, we tested whether the enzymatic DNA-dependent ATPase, DNA helicase, and CTD kinase activities that are associated with TFIIH were present in our purified preparations (27,36). As shown in   5A, the ATPase activity detected was dependent on the presence of DNA and strongly stimulated by either circular M13 single-stranded or double-stranded supercoiled plasmid DNA. In addition, the DNA helicase and CTD kinase activities were readily detected (Fig. 5, B and C). In contrast, we were not able to detect any DNA nicking or exonuclease activity in the anti-HA fraction (data not shown).
Next, we tested the transcriptional activity of the anti-HA eluate (Fig. 5D). Addition of the purified complex was absolutely required in a fully defined reconstituted RNA polymerase II transcriptional reaction using the adenovirus 2 major late promoter. The anti-HA eluate contained only detectable TFIIH activity because omitting either TBP, TFIIB, TFIIE, or TFIIF abolished transcription completely, whereas omitting TFIIA resulted in a strongly decreased signal, in agreement with a stimulatory role for TFIIA in defined transcription re-actions (Fig. 5D). These findings indicate the absence of detectable contaminating transcriptional activities in the TFIIH preparation. A quantitation of the enzymatic and transcriptional assays is presented in Fig. 5, E and F. Finally, we characterized the activity of the affinity-purified TFIIH in in vitro NER assays. As shown in Fig. 6A, the affinitypurified complex was able to complement the NER defect in XP-B extracts. Quantitation of the incorporated labeled nucleotides in the damaged DNA allowed us to estimate the yield of active TFIIH, which was calculated to be ϳ12% (Table I), indicating that the anti-HA elution is rather efficient. Theoretically, the in vitro complementation could also be due to exchange of XPB with other TFIIH subunits. Therefore, we performed an antibody depletion experiment. Using anti-p62 antibodies, a HeLa WCE was depleted of TFIIH, which resulted in a co-depletion of NER activity (Fig. 6B, compare lane 1 with  lanes 6 and 7). As shown (Fig. 6B, lanes 2-5), the anti-HA eluate restored NER activity in the depleted HeLa WCE up to the level of the nondepleted extract, indicating that TFIIH complex formation involving XPB and p62 is essential for NER activity (25).
To analyze in more detail the NER activities of the purified TFIIH in vivo, a microneedle injection experiment was carried out. As anticipated, a strong stimulation of unscheduled DNA synthesis was readily observed directly after microinjection in XP-B and XP-D polykaryons, which was specific because it was not seen in XP-F and XP-G cells (Fig. 7, A and C). Interestingly, correction was also observed in TTD-A cells (Fig. 7, B and C) that contain a mutation in an as yet unidentified NER factor (37), in agreement with our earlier findings using highly purified TFIIH from HeLa cells by classical purification (8). These experiments indicate that TTDA either is an intrinsic component of TFIIH or is required for a modification of TFIIH, enabling it to function in repair (see "Discussion").
In conclusion, these experiments show that the immunopurification of dtXPB and associated proteins result in the rapid and efficient purification of TFIIH, which is active in NER and transcription.

FIG. 4. Association of TFIIH with NER and transcription factors.
None of the tested NER and/or basal transcription factors showed detectable association with TFIIH immunoprecipitated using HA-antibodies. As a positive control, human SUG1 was identified in the bound fraction. Antibodies used to identify multisubunit complexes recognized the p62 subunit of TFIIH, the RPB1 subunit of RNA polymerase II, the 34-kDa ␤-subunit of TFIIE, and the RAP74 subunit of TFIIF. Indicated are the load (WCE) and the bound fraction (eluted with excess HA peptide).

DISCUSSION
TFIIH was originally purified as a basal transcription factor from rat, yeast, and human (24, 38 -40) and was first shown by Schaeffer et al. (9) to be involved in NER; this involvement was subsequently demonstrated by others as well (25,33,41). By immunopurifying the XPB protein using a cell line expressing functional tagged XPB, we describe an improved and facilitated purification for TFIIH free of contaminating NER and transcriptional activities that is an efficient, essentially one-step, procedure utilizing physiological elution conditions. Utilizing this protocol, which avoids the high salt and hydrophobic chromatography conditions of the classical purification procedure, we identify TFIIH as a nine-subunit complex. In addition, we show that each complex contains only one molecule of the XPB helicase. The intensity of protein staining of the XPD subunit in the purified complex compared with XPB is consistent with the idea that the XPD helicase, also, is present on a molar basis in the complex. The occurrence of two helicase molecules per TFIIH complex is in agreement with the concept that the functional forms of a number of helicases are oligomers, generally dimeric or hexameric (42). As reported previously, CDK7 and cyclin H, together with MAT1, are part of both TFIIH and a separate trimeric complex in the cell (13). This is consistent with our findings in the immunoprecipitation experiment that CDK7 and cyclin H are relatively more abundant in the WCE and unbound fraction as compared with the XPB-associated fraction (see Fig. 2).
During the past years, all eight XPB-associated factors were identified and cloned as TFIIH subunits purified from both HeLa cells and from the budding yeast Saccharomyces cerevisiae (16,43), suggesting that in solution in repair-and transcription-competent WCEs, at least the major form of TFIIH is composed of nine subunits and is active in NER, as well as transcription. However, the possibility cannot be excluded that substoichiometric and/or poorly stained subunits are essential for, for example, NER functioning, and therefore, definite proof that both the NER and transcriptional activity of TFIIH resides with the nine identified and cloned subunits awaits re-  a One unit of activity was defined as the amount of protein used to increase the incorporation 2-fold relative to the receiving XP-B extract (45).
b Activity of the WCE could not be determined. c Protein concentrations were determined using the BCA protein assay (Pierce) and could not be reliably determined for the anti-HA fraction.
constitution of TFIIH from recombinant proteins.
One of the TFIIH factors that is not yet assigned to a subunit is TTDA. We are presently investigating whether any of the known TFIIH genes are mutated in TTD-A cells. However, it is theoretically also possible that TTDA is not a subunit of the TFIIH complex itself but is implicated in TFIIH modification as part of its function. Recently, we have identified human SUG1 as a protein interacting with the XPB subunit of TFIIH (35). Little is known about posttranslational regulation of TFIIH function and the role of factors like SUG1 that are thought to unfold or refold proteins in the context of several processes, including regulated proteolysis. Like SUG1, TTDA could play a role in TFIIH modification without being part of the complex. The inability to generate high levels of dtXPB protein, even when the cDNA was expressed under control of strong promoters, suggests an autoregulatory mechanism of XPB protein levels. For example, a similar observation was made in the case of overexpression of the NER protein ERCC1, which forms a complex with XPF, and TBP, which is part of the basal transcription factor TFIID (22,44).
Using NER-and transcription-competent WCEs, physiological washing conditions, and nonoverexpressed functional dtXPB protein, we failed to observe, within our limits of detection, interactions with any NER and/or transcription factor tested, although some of them were reported previously by others (31)(32)(33)(34). Several explanations can be put forward for this apparent discrepancy. Many methods for identification of protein-protein interactions use overexpressed, in vitro synthesized or purified proteins often involving heterologous expression systems. When the protein normally resides in a complex and has multiple interaction domains, it may exhibit promiscuous association behavior when studied in isolation because of the lack of its natural partners, improper folding, or lack of posttranslational modification. Alternatively or in addition, the interactions observed were not stable during our extract preparations and/or were transient or induced upon DNA binding.
The procedure described here greatly facilitates the isolation of active TFIIH. It is simple, fast, and reproducible and does not require extensive chromatographic procedures. In combination with specific procedures for extract preparation, it may be exploited further for the purification of holo-complexes involved in NER and/or transcription. Furthermore, this procedure may allow the isolation of TFIIH with mutated XPB subunits for biochemical analyses to obtain more insight into the requirements for the XPB helicase-mediated function in NER and transcription initiation. Because the reconstitution of TFIIH from recombinant source is lacking at this moment, it would also be of interest to add (epitope) tags to other TFIIH subunits for functional analysis of TFIIH with mutations in, for example, the second helicase subunit, XPD. These experiments are in progress.