p52 Mediates XPB Function within the Transcription/Repair Factor TFIIH* 210

To further our understanding of the transcription/DNA repair factor TFIIH, we investigated the role of its p52 subunit in TFIIH function. Using a completely reconstitutedin vitro transcription or nucleotide excision repair (NER) system, we show that deletion of the C-terminal region of p52 results in a dramatic reduction of TFIIH NER and transcription activities. This mutation prevents promoter opening and has no effect on the other enzymatic activities of TFIIH. Moreover, we demonstrate that intact p52 is needed to anchor the XPB helicase within TFIIH, providing an explanation for the transcription and NER defects observed with the mutant p52. We show that these two subunits physically interact and map domains involved in the interface. Taken together, our results show that the p52/Tfb2 subunit of TFIIH regulates the function of XPB through pair-wise interactions as described previously for p44 and XPD.

To further our understanding of the transcription/ DNA repair factor TFIIH, we investigated the role of its p52 subunit in TFIIH function. Using a completely reconstituted in vitro transcription or nucleotide excision repair (NER) system, we show that deletion of the Cterminal region of p52 results in a dramatic reduction of TFIIH NER and transcription activities. This mutation prevents promoter opening and has no effect on the other enzymatic activities of TFIIH. Moreover, we demonstrate that intact p52 is needed to anchor the XPB helicase within TFIIH, providing an explanation for the transcription and NER defects observed with the mutant p52. We show that these two subunits physically interact and map domains involved in the interface. Taken together, our results show that the p52/Tfb2 subunit of TFIIH regulates the function of XPB through pair-wise interactions as described previously for p44 and XPD.
Human TFIIH is a multiprotein complex composed of nine subunits ranging from 89 to 32 kDa. These subunits are assembled into two subcomplexes: the core TFIIH, which is composed of six subunits (XPB, XPD, p62, p52, p44, and p34) and the cdk-activating kinase (CAK), 1 which is composed of cdk7, cyclin H, and MAT1. Recently, the molecular structures of both human and yeast TFIIHs have been determined by electron microscopy and show similarities in size, shape, and architecture (1,2). Originally identified as a basal transcription factor, TFIIH was subsequently found to contain XPB (Rad 25) and XPD (Rad 3), two helicases involved in nucleotide excision repair (NER) (3)(4)(5)(6)(7)(8). Mutations in one of these subunits induce UV sensitivity in both human and yeast and are responsible for three rare human genetic disorders, xeroderma pigmentosum (XP), Cockayne syndrome, and trichothiodystrophy (9,10). Many XP patients suffer from a high incidence of skin cancer due to their inability to remove lesions from their DNA. Further studies in yeast have shown that additional subunits of TFIIH, including p62 (Tfb1), p52 (Tfb2), and p44 (Ssl1) also have an essential role in DNA repair (11).
In an effort to understand the function of TFIIH in the diverse fundamental cellular processes in which it participates (transcription and DNA repair but also cell cycle regulation through the CAK complex), efforts have been aimed at systematically characterizing all TFIIH subunits. Several functions have been determined for individual components: XPB and XPD are ATP-dependent helicases indispensable for opening the DNA around a promoter and/or a lesion (12)(13)(14)(15)(16). Cdk7 is a serine/threonine kinase, which is regulated by cyclin H and MATI and phosphorylates several substrates including the Cterminal domain (CTD) of RNA polymerase II (see Ref. 17 and references therein). Recently, the N-terminal part of p44, a subunit of core TFIIH, has been shown to positively regulate XPD helicase activity, whereas the C-terminal part is involved in promoter escape (18). The XPD helicase regulation is lacking in a majority of XP-D patients because mutations in the Cterminal end of XPD abolish the XPD/p44 interaction (19). This can explain the UV sensitivity and NER defect harbored by yeast deleted for Ssl1, the p44 yeast homolog (20 -23). The three remaining subunits, p62, p52, and p34, do not contain any known specific motif or enzymatic activities, and their function, as components of TFIIH structure, remains obscure.
Concerning p52, in vitro transcription and DNA repair assays, as well as microinjection experiments using p52 antibody, have demonstrated that p52, the last subunit of TFIIH cloned (24), is involved in both transcription and NER. Moreover, deletion of the C-terminal region of Tfb2, the yeast counterpart of p52 (64% similar and 40% identical), is detrimental for NER activity of yeast TFIIH in vivo and in vitro (25). In the present paper we have introduced a similar mutation in the human p52 subunit. We demonstrate that the C-terminal region of p52 is required to anchor XPB (the helicase involved in the opening of the promoter) within the core TFIIH. This study defines p52 as a preferential and indispensable partner of XPB in the TFIIH complex and helps to define the function of the XPB helicase in class II genes transcription and NER.

Construction of Recombinant Baculoviruses-Baculoviruses express-
ing the TFIIH subunits XPB, His-XPB, XPD, p62, p52, His-p44, p44, p34, cdk7, His-cyclinH, and MAT1 were constructed as previously described (Ref. 13 and references therein). To obtain baculoviruses expressing p52(1-135), p52(137-296), p52(1-358), p52(381-462), * This work was supported by grants from the Institut National de la Recherche et de la Santé, the Centre National de la Recherche Scientifique, and the Association pour la Recherche sur le Cancer and Grant QLG1-CT-1999-0081 from the European Economic Community. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. □ S The on-line version of this article (available at http://www.jbc.org) contains sequence alignment of human p52 with its orthologs. Alignments are shown of Homo sapiens p52 sequence with its eukaryotic counterparts: Mus musculus, Drosophila melanogaster, Saccharomyces cerevisiae, Saccharomyces pombe, and Caenorhabditis elegans. Conserved and strictly conserved residues are drawn in yellow and blue, respectively. Boxes I, II, III, IV, and V correspond to stretches of highly conserved residues. The sequences are numbered according to the human protein.
XPB(44 -782), and XPB(208 -782), the corresponding cDNAs were amplified by PCR and inserted into a modified pVL1392 baculovirus transfer vector that adds a 6-histidine tag at the N terminus of the open reading frame. To obtain baculoviruses expressing p52, p52(1-304), and p52(305-462) with a FLAG peptide (MTKDDDDKH) fused to its N terminus, the corresponding cDNAs were cloned in the PSK277 vector using a NdeI and BamHI restricted site as described (26). The vectors were then recombined with baculovirus DNA (Baculogold, PharMingen), and the resulting viruses were plaque-purified and -amplified according to the manufacturer.
Alternatively, to purify recombinant TFIIH by FLAG strategy, a virus derived from the pSK277 transfer vector expressing a p34 subunit with a FLAG peptide (MTKDDDDKH) fused at its N terminus was used. This virus was used together with viruses expressing p44, p52, p62, His-XPB, XPD, cdk7, His-cyclinH, and MAT1. 48h hours after infection, 10 9 cells were collected, washed in 1ϫ phosphate-buffered saline, 30% glycerol and dounced in buffer B (20 mM Tris-HCl, pH 7.8, 10% glycerol, NaCl 250 mM, 2 mM ␤-mercaptoethanol). DNA and cell membranes were pulled down by centrifugation at 14,000 ϫ g during 30 min. The supernatants were incubated for 1 h at 4°C with 1/40 fraction volume of cobalt chelate affinity resin (Talon, CLONTECH). After a 10-resin volume wash with buffer B containing 5 mM imidazole, TFIIH complexes were eluted with the same buffer containing 250 mM imidazole. The fractions from the Talon column, which contained TFIIH complexes, were incubated for 4 h at 4°C with protein A-Sepharose beads cross-linked with anti-FLAG antibodies (FLAG-M2, Sigma). After three washes with buffer B containing 0.1% Nonidet P40, proteins were eluted in one bead volume of buffer B containing 1 mg/ml epitope peptide for 12 h.
Protein-Protein Interaction Assays-Pair-wise protein interactions were characterized by co-infection in Sf9 cells (2.5 ϫ 10 7 ) with the corresponding recombinant baculoviruses at a multiplicity of infection of 5, collected 48 h after infection, washed in 1ϫ phosphate-buffered saline, 30% glycerol and dounced in 2.5 ml of buffer A. Clarified lysates were obtained by centrifugation at 14,000 ϫ g for 30 min at 4°C. 50 l of clarified lysate was adsorbed on 20 l of protein G-Sepharose beads linked with the appropriate monoclonal antibody (1B3, which recognizes the ATP binding site of XPB, and 1D11, which recognizes the N terminus of p52) in buffer C (20 mM Tris-HCl, pH 7.5, 10% glycerol, 50 mM KCl, 0.1 mM EDTA). After 1 h of incubation at 4°C, the beads were washed extensively in buffer C containing 150 mM KCl and resuspended in Laemmli buffer. The proteins were resolved by SDS-PAGE (12.5% acrylamide) and revealed by Western blotting using the appropriate monoclonal antibodies.
Nucleotide Excision Repair-Dual Incision Assay-Circular DNA containing a single 1,3-intrastrand d(GpTpG) cisplatin-DNA cross-link (Pt-GTG) was prepared as described (27). Repair reactions were carried out in buffer containing 45 mM HEPES at 7.8, 70 mM KCl, 5 mM MgCl 2 , 1 mM dithiothreitol, 0.3 mM EDTA, 10% glycerol, 2.5 g of bovine serum albumin, and 2 mM ATP. Each reaction contained 50 ng of RPA, 22.5 ng of XPA, 10 ng of XPC-hHR23B, 50 ng of XPG, 20 ng of ERCC1-XPF NER factors, or 1.5 l of purified TFIIH fractions. Following pre-incubation for 10 min at 30°C, 50 ng of Pt-GTG was added and reactions were continued for 90 min at 30°C. Reactions were stopped by rapid freezing. 6 ng of an oligonucleotide complementary to the excised DNA fragment (and containing four extra G residues at the 5Ј end) was annealed to the excision products. Sequenase v.2.0 polymerase (0.1 unit) and 1 Ci of ( 32 P)dCTP were used to add four radiolabeled C residues to each excision product. The products were separated by electrophoresis on 14% polyacrylamide gel and visualized by autoradiography.
TFIIH Enzymatic Assays-The helicase substrate was obtained by annealing 5 ng of an oligonucleotide encompassing a sequence complementary to nucleotides 6219 -6255 of single-stranded M13mp18 DNA to 1 g of single-stranded M13mp18. The resulting heteroduplex was digested for 1 h at 37°C with EcoRI (New England Biolabs) and then HeLa TFIIH is used as a control. rIIH6wt and rIIH6/p52(1-358) were tested for their ability to respond to CAK stimulation when present in an in vitro transcription assay (C, right panel) extended to 21 and 20 bp, respectively, with the Klenow fragment (5 units) in the presence of 50 mM dTTP and 7 Ci of [␣-32 P]dATP (3000 Ci/mmol, Amersham Biosciences). The helicase assay was then performed as described (24).
ATP hydrolysis was monitored as previously described. Briefly, protein fractions were incubated for 2 h at 30°C in the presence of 1 Ci of [␥-32 P]ATP (7000 Ci/mmol, ICN Pharmaceuticals) in a 20-l reaction volume containing 20 mM Tris/HCl, pH 7.9, 4 mM MgCl 2 , 1 mM dithiothreitol, and 50 g/ml bovine serum albumin. Reactions were stopped by adding EDTA to 50 mM. The reactions were then diluted 5-fold, spotted onto polyethylenimine TLC plates (Merck), run in 0.5 M LiCl/ 1 M formic acid, and autoradiographed.
Kinase assays (25) were carried out in a 20-l reaction volume containing 20 mM HEPES, pH 7.9, 20 mM Tris/HCl, pH 7.9, 7 mM MgCl 2 , 0.5 mg/ml bovine serum albumin, 30 mM KCl, 1 g of ctd4 (a synthetic tetrapeptide of SPTSPSY), and 2.5 Ci of [␥-32 P]dATP. Samples were incubated 30 min at 25°C, and reactions were stopped by the addition of 5 l of loading buffer. After SDS-PAGE (15%), the gel was fixed and dried on Whatman filter paper. The phosphorylated ctd4 was visualized by autoradiography.
KMnO 4 Footprinting Assay-AdMLP template (20 ng) was incubated at 25°C for 30 min with recombinant TBP, TFIIB, TFIIF, TFIIE, highly pure pol II, and TFIIH as indicated in a 20-l reaction that contained 50 mM Tris/HCl, pH 7.9, 10% glycerol, 1 mM EDTA, 0.5 mM dithiothreitol, and 5 mM MgCl 2 . ATP and CTP (200 M) were added for the last 5 min. Two l of 160 mM KMnO 4 was added for 2 min, after which the reaction was stopped by addition of 2 l of 14.4 M ␤-mercaptoethanol. After phenol-chloroform extraction, DNA was recovered by ethanol precipitation, redissolved in water, and subjected to 30 cycles of primer extension using an end-labeled primer. After phenol-chloroform extraction, ethanol precipitation, and wash, the sample was loaded onto a 6% sequencing gel. The gel was dried and autoradiographed (12).
Proteolysis Assay-For the proteolysis experiments, 30 l of recombinant TFIIH purified by the FLAG strategy were incubated with 10 l of chymotrypsin at various concentrations (0 -0.25 mg/ml) and incubated 1 h at room temperature. The reaction was stopped by addition of 5 l of 100 mM Pefablock (Roche Diagnostics), and the hydrolyzed peptides were analyzed by SDS-PAGE followed by Western blotting using monoclonal antibodies directed against either the N-terminal (1D11) or the C-terminal (5D6) part of p52.

RESULTS
The C-terminal Deletion of p52 Is Detrimental for TFIIH Activity in Both DNA Repair and Transcription-By taking advantage of recombinant technology, we have generated TFIIH complexes in insect cells. One of them (rIIH9/p52 (1-358)) contains a C terminus-truncated p52. Recombinant TFIIHs, (rIIH9wt and rIIH9/p52(1-358)) were produced in High-Five insect cells co-infected with baculoviruses expressing either the wild type or truncated forms of p52 as well as the eight other TFIIH subunits (13). The insect cell extracts were then loaded on a heparin-Ultrogel column, and the protein was eluted with 0.5 M KCl including rIIH9wt and rIIH9/p52(1-358), respectively, which were applied onto a cobalt chelate affinity resin. After extensive washing of the column under mild salt conditions (50 mM KCl), both rIIH9s were eluted with the same buffer containing 50 mM EDTA. As judged by immunoblotting, the stoichiometry of both rIIH9 complexes are almost identical (Fig. 1A). The mutated TFIIH differs only from the wild type complex by the presence of the truncated p52(1-358) protein (lane 2); p52 co-migrates with histidine-tagged p44 (lane 1).
Both recombinant TFIIHs were then tested for their DNA repair and transcription activities. As expected (25), the recombinant human TFIIH (rIIH9/p52(1-358)) was inactive in incision/excision (one of the first steps of NER) of cisplatin-damaged DNA when added to an in vitro assay (28) containing highly purified recombinant XPC/HR23b, XPA, RPA, ERCC1/ XPF, and XPG (Fig. 1B). In addition, the transcription activity of rIIH9/p52(1-358) was strongly reduced (80%) relative to wild type (Fig. 1C, left panel) in an in vitro transcription assay containing the adenovirus major late MLP template, RNA pol II, and purified basal transcription factors except TFIIH.
Because CAK is anchored to the core TFIIH, we wondered if the p52 C-terminal deletion would have affected the stimulatory effect of CAK. Previous work has shown that for certain promoters such as MLP, dependence on the CAK subunits is not a function of the kinase activity of cdk7 but is rather due to the contribution of CAK to an optimal positioning of TFIIH within the transcription preinitiation complex (29,30). Recombinant rIIH6/p52(1-358) and rIIH6wt were produced in Sf9 cells and purified as described above. Wild type CAK was then added to the reconstituted transcription assay containing either the rIIH6 or the rIIH6/p52(1-358) subcomplex (Fig. 1C,  right panel). In both cases CAK stimulates rIIH6/p52(1-358) and rIIH6wt transcription activity by about 4 -5-fold (Fig. 1C,  right panel, compare lanes 2 and 4 with lanes 6 and 8). These results clearly indicate that the transcriptional defect associated with rIIH6/p52(1-358) is independent of CAK transcription stimulation and is instead intrinsic to the core structure of TFIIH.
p52 Mutations Affect Promoter Opening by XPB-We then wondered whether p52 mutations would affect the enzymatic activities of TFIIH that participate in transcription and DNA repair events. We then investigated the phosphorylation by the cdk7 subunit of TFIIH of peptide substrate (ctd4) that contains four copies of the hepta repeat found in the C-terminal domain of the largest RNA pol II subunit. Both TFIIHs phosphorylated ctd4 ( Fig. 2A), showing that the cdk7 activity of CAK is not affected by the p52 mutation. In another set of experiments, we measured the overall ATPase activity of both rIIH9wt and rIIH9/p52(1-358). Neither the XPB nor the XPD ATPase activities were affected by the p52 mutation (Fig. 2B). Then, we asked whether mutation of p52 would impair the XPD or XPB unwinding activities. We found that the ATP-dependent XPD helicase activity of rIIH9/p52(1-358) is similar to rIIH9 wild type (Fig. 2C). As the XPB helicase activity cannot be measured using a standard strand displacement assay (31) (Fig. 2C) and XPB has been shown to be the helicase involved in promoter opening, we investigated whether rIIH9 retains its ability to direct promoter opening using a KMnO 4 footprinting assay (27). In the presence of saturating amounts of basal transcription factors including HeLa TFIIH and RNA pol II, addition of both ATP and CTP (allowing the formation of the first phosphodiester bond) leads to an enhancement of the sensitivity of the thymidines around the AdMLP promoter (Fig. 2D, lanes 1  and 2, positions ϩ3, ϩ5, ϩ7, and ϩ8), indicating promoter opening around the transcription initiation site. When rIIH9/ p52(1-358) is added instead of rIIH9wt, promoter opening is significantly decreased (Fig. 2D, lanes 2 and 3). This inhibition strongly correlates with the decrease in run-off transcription, indicating that this impairment of transcription can be directly linked to an impairment of promoter opening. The Subunit p52 Is Required for the Integration of XPB in TFIIH-Since promoter opening is impaired and XPB has been shown to be the helicase involved in the transcription initiation step (31), we then asked whether p52 mutation affects association of XPB with the TFIIH multiprotein complex. Antibodies raised against p44 were used to immunoprecipitate TFIIHs produced in insect cells co-infected with baculoviruses overexpressing wild type p52 or p52(1-358) together with either all of the eight TFIIH subunits or only with XPB, p62, p44, and p34. When performed at higher salt concentration (250 mM KCl), XPB did not co-precipitate with other TFIIH subunits when p52(1-358) is present (Fig. 3A, lanes 4 -6). Indeed, although treatment with 250 mM KCl does not dissociate TFIIH subcomplexes (compare lanes 1-3), we notice that p52(1-358) prevents anchoring of XPB in the core TFIIH (Fig. 3B, lane 2).
In another set of experiments, we demonstrate that p52 is required for the presence of XPB in the core TFIIH. Indeed XPB could be co-immunoprecipitated from infected insect cells with p44, p62, and p34 only in the presence of p52 (Fig. 3C, compare  lanes 2 and 4). Together these results show that the C-terminal domain of p52 determines the strength of the binding of XPB within TFIIH whether or not CAK/XPD is present.
The p52 Subunit Directly Interacts with XPB-The above data suggest a role for p52 in promoter opening and for the integration of XPB into TFIIH. We therefore wondered if some connection between these two subunits could be identified. Sf9 insect cells were co-infected with two baculoviruses expressing p52 and one of the five subunits of the core TFIIH (XPB, XPD, p62, p44, and p34). Recombinant proteins of the core TFIIH contained in cell lysates were then immunoprecipitated with the corresponding antibody. After extensive washing at 250 mM KCl, the immunoprecipitates were analyzed by Western blotting. Under these conditions, only XPB (Fig. 4B, compare lanes  1 and 2) and to a lesser extent p62 (data not shown) were able to significantly interact with p52. These results argue that XPB binds preferential to p52 within TFIIH. FIG. 4. p52 interacts directly with XPB. Co-infected baculovirus extracts expressing wild type p52 or truncated forms and XPB were analyzed by Western blotting (A) and tested for ability to form a complex by immunopurification using an anti-XPB antibody; L, heavy chain (B). Recombinant TFIIH was digested with increasing amounts of chymotrypsin (0.02, 0.1, 0.5, and 2.5 g), subjected to SDS-PAGE (12.5% acrylamide) and analyzed by Western blotting using monoclonal antibodies directed against the N-terminal or C-terminal end of p52. A 43-kD protease-resistant fragment that contains the p52 N-terminal epitope accumulates, whereas the p52 C-terminal epitope is rapidly degraded (C). Schematic drawings of p52. The stretches of highly conserved residues in eukaryotes are indicated in black, and the regions involved in XPB binding are delimited (D).
The above results show that p52 contains two distinct XPB binding regions that correspond to residues 1-135 and 304 -381 (Fig. 4D) and raise questions concerning the role, if any, of the extreme C-terminal end of p52 in XPB interaction. To address this point, recombinant TFIIH containing the nine subunits produced in baculovirus-infected Sf9 cells was subjected to limited proteolysis. At low protease concentration, a substrate protein would be expected to be cleaved preferentially at sites that are exposed to the solvent. It is generally accepted that protease-resistant polypeptides are involved in the formation of structural entities. Recombinant TFIIH was incubated with increasing amounts of chymotrypsin and analyzed by SDS-PAGE followed by immunoblotting. Using an antibody directed against the N terminus of p52, we noticed that p52 was proteolysed to result in a fragment of 43 kDa that corresponds approximately to 390 -400 amino acids (Fig. 4C, left panel). Under these conditions, the C-terminal part of p52 is presumably degraded to very small fragments as the antibody directed against the C terminus of p52 does not detect any low molecular weight peptides (right panel). This experiment shows that the extreme C terminus of the protein is accessible to the protease and is in agreement with the observation that p52(381-462) does not interact with XPB.

DISCUSSION
Gene expression is regulated by the various transcription factors that are involved in RNA synthesis. Some of these transcription factors possess enzymatic activities that orchestrate the various events of the transcription reaction, such as chromatin remodeling, promoter opening, and RNA pol II phospho-/dephosphorylation.
It has been known for some years that the enzymatic activities of the cdk7 and XPD subunits of TFIIH are regulated. The activity of cdk7 kinase depends on the presence of cyclinH and MAT1, two other TFIIH subunits, as well as on the phosphorylation state of its T-loop (17). Similarly, upon interaction with p44, XPD helicase activity increases (19). Mutations in the C-terminal domain of XPD modify the contact with p44, explaining the NER defect associated with most of the XP-D patients (32). In addition, mutations in XPD or in p44 that modify the XPD-p44 interaction affect the composition of TFIIH by decreasing the amount of XPD and CAK subunits associated with the core and/or weakening the anchoring of FIG. 5. The N-terminal part of XPB is required for the interaction with p52 and the anchoring to core TFIIH. Co-infected baculovirus expressing combinations of wild type or mutated XPB and p52 (A) were immunoprecipitated with Ab-XPB or Ab-p52 antibodies (IP-Ab XPB, IP-Ab p52). After extensive washing, the immunoadsorbed proteins as well as the loading material were resolved on a 12.5% SDS-PAGE and analyzed by Western blotting using a mixture of antibodies generated toward XPB, the N terminus, and the C terminus of p52. H and L correspond to the heavy chains and the light chains of the anti-XPB antibody. Sf9 cells were co-infected with baculoviruses expressing p34F, p44, p52, p62, and XPB, XPB(44 -782), or XPB(208 -782), immunoprecipitated using an anti-FLAG antibody as indicated at the top of the panel, and analyzed by Western blotting (B). C, schematic drawing of XPB. The nuclear localization signal (NLS), the putative DNA binding domain (DBD), and the conserved helicase motifs (I, Ia, II, III, IV, V, and VI) of XPB are indicated.
CAK to the core TFIIH (23, 31) with a consequent change in the rate of some hormone-responsive genes (33).
In the present study, we show that the p52 subunit is the privileged partner of the XPB helicase. Each of the enzymatic activities found in TFIIH now possess its own regulatory partner within the protein complex. Using a recombinant TFIIH, with which the C-terminal domain of p52 was deleted, we have demonstrated that the p52 mutation is detrimental for both TFIIH-DNA repair and transcription activities. Indeed, several experiments, including permanganate footprinting, allow us to assign this transcriptional defect to an impairment in the opening of the DNA around the promoter. Having previously demonstrated that the XPB helicase was crucial in the promoter opening (12,31), we further investigated the role of p52 in anchoring of XPB to core TFIIH. We first show that the integrity of p52 is a prerequisite for the presence of XPB in core TFIIH. Indeed, the ternary p34/p44/p62 complex does not interact with XPB in the absence of p52. In fact XPB anchoring is mediated by p52 and, as also observed in yeast (34), the two proteins interact directly. XPB binding involves two independent regions of p52: the first is located within residues 1-135 and contains a few residues strictly conserved among p52 orthologs. The second binding region includes residues 303-381, as p52(303-462) interacts with XPB, whereas p52(381-462) does not. This second binding region is highly conserved from human to yeast ( Fig. 4 and supplemental material at http://www.jbc.org).
The fact that p52(381-462) does not interact with XPB correlates with the observation that the C terminus of p52 is rapidly digested in a mild proteolysis assay (Fig. 4D). It seems, however, that this domain is crucial for the architecture and/or the stability of TFIIH, as a deletion of residues 358 -462 weakens the binding of XPB within the core TFIIH. This instability can explain the NER defect and transcriptional impairment as well as the phenotype observed in yeast cells containing a similar deletion. We propose that the C-terminal domain of p52 might possess a three-dimensional structure that locks up XPB inside TFIIH. The consequence of the p52 deletion would be a weaker and/or inappropriate positioning of the XPB helicase first within TFIIH itself and second within the transcription and/or DNA repair machinery. This defective association of XPB with TFIIH would result in a defect in one of the essential function of TFIIH, i.e. DNA opening.
Unlike the XPB and XPD genes, no human genetic disorders have been associated with the gene encoding p52. However, a change in the last 40 amino acids of XPB from XP-B/CS patients results in a similar defect in promoter opening and consequently RNA synthesis (31). Moreover, using photo-crosslinking experiments, we showed that the positioning of the mutated XPB helicase of TFIIH onto promoter DNA was impaired (35). Sequence analysis as well as deletion experiments suggest that XPB is composed of three modules: a N-terminal domain (also referred to as DNA binding domain), the catalytic core domain, which contains the helicase motifs, and the Cterminal domain (36). We have shown that the region of XPB that contacts p52 corresponds to the N-terminal putative DNA binding domain of this helicase (37,38). It is likely that the C-terminal moiety of p52 allows trapping and stabilization of XPB within TFIIH. Whether deletion of the C-terminal region of p52 affects the accuracy of positioning and/or the association of TFIIH within the transcription preinitiation and NER complexes remains to be further investigated.