The xeroderma pigmentosum group C protein complex XPC-HR23B plays an important role in the recruitment of transcription factor IIH to damaged DNA.

The xeroderma pigmentosum group C protein complex XPC-HR23B was first isolated as a factor that complemented nucleotide excision repair defects of XP-C cell extracts in vitro. Recent studies have revealed that this protein complex plays an important role in the early steps of global genome nucleotide excision repair, especially in damage recognition, open complex formation, and repair protein complex formation. However, the precise function of XPC-HR23B in global genome repair is still unclear. Here we demonstrate that XPC-HR23B interacts with general transcription factor IIH (TFIIH) both in vivo and in vitro. This interaction is thought to be mediated through the specific affinity of XPC for the TFIIH subunits XPB and/or p62, which are essential for both basal transcription and nucleotide excision repair. Interestingly, association of TFIIH with DNA was observed in both wild-type and XP-A cell extracts but not in XP-C cell extracts, and XPC-HR23B could restore the association of TFIIH with DNA in XP-C cell extracts. Moreover, we found that XPC-HR23B was necessary for efficient association of TFIIH with damaged DNA in cell-free extracts. We conclude that the XPC-HR23B protein complex plays a crucial role in the recruitment of TFIIH to damaged DNA in global genome repair.

Nucleotide excision repair (NER) 1 is the primary pathway for removal of lesions from DNA and is conserved across a wide range of species and from prokaryotes to eukaryotes. Studies of prokaryotic NER have shown that the NER pathway is controlled by phased enzymatic reactions (1,2). Recently, analyses of NER-deficient yeast, rodent, and human mutant cells have resulted in the identification and characterization of the eu-karyotic proteins involved in NER. Xeroderma pigmentosum and Cockayne's syndrome are human autosomal recessive hereditary diseases that result from a deficiency in NER (3); to date, seven complementation groups (A to G) in xeroderma pigmentosum and two (A and B) in Cockayne's syndrome have been identified, and most of the corresponding NER genes have been cloned (4). Additional factors involved in NER have also been identified (5-7) using a cell-free NER assay (8). Based on these studies, including studies on reconstitution of NER using purified proteins (9 -11), various models of the mechanism of eukaryotic NER have been proposed (see Refs. 12 and 13 for review).
The xeroderma pigmentosum group C protein complex XPC-HR23B is a tightly associated complex of the products of the XPC and HR23B genes and was initially purified from HeLa cell nuclear extracts as a protein factor that complemented the DNA repair defects of XP-C whole cell extracts in a cell-free NER reaction (14,15). Although no enzymatic activity was observed in the purified complex, it has been reported that this complex is required for DNA repair prior to the excision step (15,16). During purification, XPC-HR23B exhibited a high affinity for single-stranded DNA (14,15). Moreover, using recombinant human XPC and HR23B, the high affinity of XPC-HR23B for both double-stranded and UV-irradiated DNA has also been demonstrated (17). 2 Several lines of evidence have been accumulated that indicate that this complex plays a role in NER before and after damage recognition. We recently demonstrated, using a novel DNA damage recognition-competition assay, that XPC-HR23B is the earliest acting of the damage detectors that initiate NER (18). XPC, together with TFIIH, is necessary for the initial opening reaction immediately around the lesion (19). The XPC-HR23B complex was characterized as a molecular matchmaker that participates in the assembly of the NER proteins on damaged DNA but is not present in the ultimate dual incision complex (20). These data prompted us to hypothesize that XPC-HR23B plays important roles in the assembly of NER proteins on the damaged DNA and in initiation of NER. However, the precise molecular mechanisms that follow damage detection by XPC remain unclear.
Here we report that XPC-HR23B interacts directly with TFIIH in vitro and that XPC-HR23B is absolutely required for association of TFIIH with damaged DNA in cell extracts. We conclude that XPC-HR23B is necessary for recruitment of TFIIH to damaged DNA.

EXPERIMENTAL PROCEDURES
Cell Culture and Preparation of Whole Cell Extracts-Human 293, XP7CASV (group A), and XP4PASV (group C) cells were grown at 37°C in Dulbecco's modified Eagle's medium (Nissui) supplemented with 10% fetal bovine serum. HeLa cells were grown in suspension at 37°C in Dulbecco's modified Eagle's medium supplemented with 10% bovine serum. Whole cell extracts were prepared as described previously (21).
Preparation of NER Proteins-Untagged (rHR23B), His 6 -tagged (rHR23B-His 6 ), and GST-tagged (GST-rHR23B) HR23B constructs were expressed in Escherichia coli and purified as described previously (22,23). Recombinant XPC (rXPC) was expressed using a baculovirus expression system and purified as described previously (23). Eight subunits of TFIIH (XPB, XPD, p62, p44, p34, MO15, cyclin H, and MAT1) were expressed in E. coli as GST fusion proteins. The coding regions of these subunits were cloned using PCR technique. PCR primers designed to introduce an NdeI restriction site at the first methionine codon of XPB, XPD, p62, p44, p34, MO15, and cyclin H or an NcoI site at the first methionine codon of MAT1 were used in conjunction with antisense primers containing an appropriate restriction site: a BamHI restriction site for XPB, p62, p44, MO15, cyclin H, and MAT1, a HindIII restriction site for XPD, and an EcoRI restriction site for p34. PCR products were digested with the appropriate restriction enzymes, subcloned into the multicloning site of the E. coli expression vector pGEX-2TL(ϩ) (24), and expressed as described previously (25). Recombinant PCNA (rPCNA) and recombinant RPA (rRPA) were expressed and purified as described previously (26). Purification of the XPC-HR23B complex (14) and TFIIH (27) from HeLa cells was also carried out as described previously.
In Vitro Assay of TFIIH Binding to XPC-rHR23B-His 6 (2.2 mg) was dialyzed against buffer containing 20 mM sodium phosphate (pH 6.8) and 0.3 M NaCl and then incubated with activated CH-Sepharose 4B beads (Amersham Pharmacia Biotech) at 4°C overnight in the same buffer. Unreacted groups were quenched by washing with excess buffer containing 0.1 M Tris-HCl (pH 8.0) and 0.5 M NaCl. The rHR23B-His 6 -Sepharose beads (10 l) were preincubated with or without rXPC (90 ng) in buffer C for 30 min on ice, mixed with purified TFIIH (200 ng), and further incubated at 4°C for 1 h with gentle rotation. After washing three times with buffer C, bound proteins were extracted by boiling in SDS sample buffer, separated by 10% SDS-PAGE, and analyzed by immunoblotting with anti-p62 antibody. Alternatively, GST-tagged HR23B, which had been incubated with or without rXPC for 30 min on ice, was incubated with purified TFIIH (200 ng), and incubated at 4°C for 1 h with gentle rotation. The mixture was adsorbed to glutathione-Sepharose beads (Amersham Pharmacia Biotech) and then washed three times with buffer C. The bound proteins were extracted, separated, and analyzed as described above.
In Vitro Assay of XPC-HR23B Binding to GST-tagged TFIIH Subunits-GST-tagged TFIIH subunits (100 ng) were adsorbed to glutathione-Sepharose beads in 100 l of buffer D (20 mM Tris-HCl (pH 7.5), 10% glycerol, 0.5 M NaCl, 2 mM dithiothrietol, 0.5 mM phenylmethylsulfonyl fluoride, 0.6 g/ml antipain, 0.6 g/ml aprotinin, 0.3 g/ml leupeptin, 0.24 g/ml pepstatin, 150 M EGTA) at 4°C for 1 h with rotation. After washing three times with buffer C, the beads were incubated in the same buffer at 4°C for 1 h with rotation in the presence of purified XPC-HR23B complex (50 ng), rXPC (60 ng), rHR23B (60 ng), rRPA (200 ng), or rPCNA (200 ng). After washing three times with buffer C, bound proteins were eluted by boiling in SDS sample buffer, separated by 8% (for detection of XPC and HR23B) or 12% (for detection of RPA and PCNA) SDS-PAGE, and analyzed by immunoblotting with anti-XPC, HR23B, RPA32, or PCNA antibodies. In advance, we confirmed that full-length GST-tagged TFIIH subunits were expressed in E. coli by immunoblotting with anti-GST antibody (data not shown).
Antibodies-Antibodies raised against XPC and HR23B were obtained as described previously (23). Antibodies raised against XPB, p62, cyclin H, RPA32, PCNA, and GST were prepared by Medical and Biological Laboratories Co., Ltd. Anti-XPA polyclonal antibody was kindly provided by Kiyoji Tanaka.
Precipitation of Proteins with DNA-Cellulose-To prepare three kinds of cellulose solution for the precipitation experiments, ssDNAcellulose and dsDNA-cellulose (both from Sigma-Aldrich) or cellulose CF11 (Whatman) was mixed with equal volumes of Sepharose CL4B (Amersham Pharmacia Biotech) and suspended (50% v/v) in buffer C. The dsDNA cellulose-Sepharose solution contained 7 g of DNA in 10 l. Whole cell extracts (125 g) from 293 or xeroderma pigmentosum cells were incubated with 10 l of the cellulose-Sepharose suspensions in buffer C at 4°C for 1 h with gentle rotation. After washing three times with buffer C, the bound proteins were extracted, separated by 10 -14% gradient SDS-PAGE, and analyzed by immunoblotting with anti-XPC, p62, XPA, or RPA32 antibodies. Buffer C without ATP or with 2 mM AMP-PNP was used in separate experiments designed to test the ATP dependence of binding.
Precipitation of NER Complex with Damaged or Undamaged DNA Substrate-PCR reactions were performed with T3 and biotin-labeled T7 primers, using a partial (nucleotides 8330 -9079) human Rev3 cDNA (28) subcloned into the EcoRI site of pBS as template. PCR reactions (35 cycles) were carried out in a total volume of 50 l containing 10 ng of the template DNA, 10 mM KCl, 20 mM Tris-HCl (pH 8.8), 2 mM MgSO 4 , 10 mM (NH 4 ) 2 SO 4 , 0.1% Triton X-100, 100 g/ml bovine serum albumin, 200 M dNTPs, 20 pmol of each oligonucleotide primer, and 2.5 units of Pfu DNA polymerase (Stratagene), using a GeneAmp PCR System 9700 thermal cycler (Perkin-Elmer Applied Biosystems). Each cycle consisted of 30 s at 94°C, 30 s at 55°C, and 1 min at 72°C. After removal of excess primer by ethanol precipitation, the PCR products were adsorbed to Streptavidin-agarose (Life Technologies, Inc.) in TE (pH 8.0) containing 0.1 M NaCl at room temperature for 1 h with gentle agitation. Damaged DNA substrate was prepared by treating PCR products in TE (10 mM Tris-HCl, 1 mM EDTA) (pH 7.5) containing 20% ethanol (0.2 g/ml) with 0.15 mM N-AAAF at 37°C for varying times followed by di-ethyl-ether extraction, chloroform extraction, and ethanol precipitation. Damaged or undamaged DNA substrate (2 g) dissolved in TE (pH 8.0) was incubated with 20 l of a 2-fold suspension of Streptavidinagarose in TE (pH 8.0) containing 0.1 M NaCl at room temperature for 1 h with gentle agitation. After washing with the same buffer, whole cell extracts (125 g) was incubated with DNA bound-agarose in 100 l of buffer C at 4°C for 1 h. After washing three times with buffer C, the bound proteins were extracted, separated by 10 -14% gradient SDS-PAGE, and analyzed by immunoblotting with anti-XPC, p62, XPA, or RPA32 antibodies. The ATP dependence was examined as described above.
Other Methods-SDS-PAGE were performed as described previously (29). For immunoblotting, proteins separated on SDS gels were electrotransferred onto polyvinylidene difluoride membranes (Immobilon-P; Millipore) at 8 V/cm 15 h in ice-cold transfer buffer (50 mM Tris, 38.4 mM glycine, 0.01% SDS, and 15% methanol). The membranes were successively incubated in blocking buffer (1% Blocking reagent (Roche Molecular Biochemicals) in 0.1 M maleic acid (pH 7.5), 150 mM NaCl), antibody in blocking buffer, and finally anti-rabbit or anti-mouse F(abЈ) 2 antibody conjugated with horseradish peroxidase (Amersham Pharmacia Biotech). Detection was carried out with SuperSignal Substrate (Pierce) according to the manufacturer's instructions. Protein concentrations were measured as described previously (30) using the Bio-Rad Protein Assay reagent and bovine serum albumin as a standard.

RESULTS
Direct Interaction between XPC and TFIIH-Of the mammalian NER factors identified to date, only TFIIH has been suggested to interact with XPC (31-33), although conclusive evidence of such interaction has been lacking. We performed coimmunoprecipitation experiments with NER-proficient human cell extracts and an antibody raised against cyclin H, which is one of the components of TFIIH. As shown in Fig. 1, not only XPB and p62, two other subunits of TFIIH, but also XPC were detected in the fraction precipitated by anti-cyclin H antibody (lane 3) but not with the control antibody (lane 2). On the other hand, no clear evidence to imply the interaction between TFIIH and XPA nor RPA was obtained. This result indicated that XPC interacted with TFIIH in the whole cell extracts; to examine whether the interaction was direct or mediated by other proteins or DNA present in the cell extracts, co-precipitation experiments were performed using purified proteins. TFIIH purified from HeLa nuclear extracts (Fig. 2C) was incubated with glutathione-Sepharose beads bound to GST-rHR23B fusion protein that had been preincubated with or without rXPC. Co-precipitation of TFIIH with the Sepharose beads was assessed by immunoblotting with anti-p62 antibody. As shown in Fig. 2A, p62 was detected in the precipitate fraction in an XPC-dependent manner (compare lane 4 with lanes 2 and 3), suggesting that XPC alone or the XPC-HR23B complex interacted directly with purified TFIIH. This XPC-dependent interaction was also observed using rHR23B-conjugated Sepharose (Fig. 2B, lane 3). Therefore, we conclude that the XPC-HR23B complex interacts directly with TFIIH in vitro and that XPC is indispensable for this specific interaction. Next, a series of pull-down experiments were performed using the GST-tagged TFIIH subunits to assess which subunit of TFIIH interacted with the XPC-HR23B complex and whether HR23B was necessary for the interaction. As shown in Fig. 3, both the purified XPC-HR23B complex (A) and free rXPC (B) were found to bind XPB and p62. In contrast, rHR23B, rRPA, and rPCNA were not co-precipitated with any of the TFIIH subunits (C-E). These results indicate that the interaction between the XPC-HR23B complex and TFIIH is mediated through specific protein-protein binding (XPC-XPB and/or XPC-p62) and that HR23B is not essential for this interaction.
XPC-HR23B Is Necessary to Precipitate TFIIH Efficiently with dsDNA-To examine the relevance of the interaction between XPC-HR23B and TFIIH, several precipitation experiments were performed. The XPC-HR23B complex has a high affinity for both ssDNA and dsDNA (14,17,23). Thus, not only XPC-HR23B but also those proteins that interact with this complex or which associate with DNA might be expected to be precipitated from whole cell extracts by DNA-cellulose. We therefore examined the presence of not only XPC and TFIIH but also XPA and RPA, which have been characterized as DNA binding NER factors, in DNA-cellulose bound fractions. Because it has been reported that ATP is necessary to form the repair protein complex on DNA (19, 20, 34), we also examined the ATP dependence of the precipitation of TFIIH by DNAcellulose. XPC, p62, XPA, and RPA32 were detected in the dsDNA-cellulose bound fraction of NER-proficient 293 whole cell extracts, as expected (Fig. 4, lanes 3 and 5). Interestingly, ATP was not required for efficient precipitation of TFIIH by DNA-cellulose from whole cell extracts (lane 3). Moreover, the presence of the nonhydrolyzable ATP analog AMP-PNP did not alter the amount of any precipitates (lane 6). Evans et al. (19) reported that ATP hydrolysis is necessary for the open complex formation immediately around the lesion. We therefore propose that the initial NER protein complex on DNA may be formed without reliance on ATP, but the initial opening reaction re- quires ATP. TFIIH has been shown to have little affinity for DNA by itself (20,35). Therefore, the observed precipitation of TFIIH by dsDNA may be due to specific interaction with other protein(s) bound to DNA or dependent upon conformational changes in DNA induced by DNA-binding proteins. To test the possibility that other NER proteins mediate the interaction between TFIIH and DNA, several precipitation experiments were performed using whole cell extracts derived from NERdeficient cell lines (Fig. 5). XPC, p62, and RPA32 were precipitated from XP-A and 293 whole cell extracts (A) with ssDNAcellulose (lanes 3 and 11) or dsDNA-cellulose (lanes 4 and 12). In contrast, p62 was not recovered in the precipitates from XP-C whole cell extracts, whereas binding of XPA and RPA32 to both ssDNA-and dsDNA-cellulose were unaffected (compare lanes 3 and 4 with 7 and 8, respectively). The same results were obtained with extracts from XP-B (GM2252A), XP-D (XP6BESV), XP-G (XP3BRSV), CS-A (CS2OSSV), and CS-B (GM1629SV) cells, all of which express the XPC protein, as with 293 whole cell extracts (data not shown). Intriguingly, ssDNA was insufficient to associate with TFIIH even though XPC binds to ssDNA as well as to dsDNA (Fig. 5A, compare  lanes 3, 4, 11, and 12). Furthermore, when the XP-C cell extracts were supplemented with purified XPC-HR23B complex, TFIIH was detected in the precipitate fraction of dsDNA-cellulose (Fig. 5B). Thus TFIIH stably associated with dsDNA, either directly or indirectly, only in the presence of XPC. It should be noted that extra signals were observed above the bands corresponding to XPA. These extra bands probably represent modified forms of the target proteins, because specific antibodies were used. We should also mention here that the amounts of precipitated proteins by ssDNA-or dsDNA-cellulose in each experiments could be altered when the different batch, but the same cell line, of cell extracts was employed (data not shown).
TFIIH Was Precipitated Efficiently by Damaged DNA as Well as XPC-To assess the relevance of the XPC-dependent association of TFIIH with DNA, similar precipitation experiments were performed with dsDNA fragments that had been treated with N-AAAF and immobilized on Sepharose beads. As shown in Fig. 6A, the amounts of XPC and TFIIH detected in the precipitates from extracts of (repair-proficient) 293 cells increased with increasing time of N-AAAF treatment of the DNA (lanes 3-6). On the other hand, a slight increase was observed in the amount of XPA in damaged DNA bound fractions (compare lane 3 with 4 -6), and no significant increase was observed in the amounts of RPA in any precipitates (lanes 3-6). Intriguingly, under the same conditions, TFIIH was found to co-pre-cipitate with N-AAAF-treated DNA, even though TFIIH displays little affinity for dsDNA, UV-irradiated or untreated, by itself (20,35). Again, we examined whether this binding of TFIIH to damaged DNA was dependent on the presence of ATP or XPC as well as previous experiments. Damaged DNA was incubated with 293 whole cell extracts in the absence or presence of ATP or AMP-PNP, and the precipitates were examined by immunoblotting. As shown in Fig. 6B, the examined proteins were equally immunodetected in the DNA binding fraction in the absence of ATP (lane 3) or in the presence of ATP (lane 5) or AMP-PNP (lane 6), suggesting that neither ATP nor ATP hydrolysis was required for early NER complex formation on damaged DNA. In the XP-C cell extracts, little p62 was found in the precipitate with undamaged DNA (Fig. 6C, lane 3), whereas addition of the XPC-HR23B complex resulted in a slight increase in the amount of bound p62 (Fig. 6C, lane 4). On the other hand, despite the absence of XPC, the amount of p62 in the precipitate was greater with the damaged DNA substrate (compare lane 3 with 5). A slight increase was also observed in the amounts of XPA, but no significant increase was observed in the amounts of RPA in the same precipitates (compare lane 3 with 5). One can presume that part of TFIIH will be recruited on damaged DNA with XPA. On the other hand, it has been reported that conformational changes in DNA can alter the cellular requirement for XPC in repair complex formation and initiation of NER (36). The minor increase in the amount of p62 precipitated by damaged DNA may have been due to modification of DNA structure by N-AAAF treatment. Importantly, the XPC-dependent binding of TFIIH was much more pronounced with the N-AAAF-treated DNA (compare  2, 6, and 10), ssDNA-cellulose (lanes 3, 7, and 11), or dsDNA-cellulose (lanes 4, 8, and  12) beads. The bound protein fractions were examined for the presence of XPC, TFIIH (p62), XPA, and RPA (p32) by immunoblotting with specific antibodies. Lanes 1, 5, and 9 contain 10% of the extracts included in the binding reactions. B, similar binding experiments were carried out with the XP-C cell extracts supplemented with purified XPC-HR23B complex (lanes 7-9). Lanes 1, 4, and 7, 10% of the input extracts; lanes 2, 5, and 8, cellulose-bound fractions: lanes 3, 6, and 9, dsDNA-cellulose-bound fractions.
lane 5 with lane 6). These results strongly indicate that TFIIH can be recruited to sites of NER (damaged DNA) via interaction with XPC. DISCUSSION XPC Binds to TFIIH through Interactions with XPB and/or p62-Previous studies have suggested that the XPC-HR23B complex and TFIIH may interact in some manner (31,32). Moreover, both RAD4, a putative yeast homolog of XPC (33,37), and RAD23, the yeast counterpart of HR23B (39), may interact with yeast TFIIH. However, evidence supporting direct interaction has been lacking in both human and yeast (39 -41). Here we show by co-immunoprecipitation that the XPC-HR23B complex interacts with TFIIH in cell extracts (Fig.  1). Although it is difficult to estimate exact percentages from the blot in Fig. 1, only small proportions of the total XPC and TFIIH in the extracts associated with each other, indicating that the interaction is rather unstable. Pull-down assays using purified proteins further demonstrated a direct interaction (Fig. 2), showing the specific affinity of XPC for XPB as well as for p62 (Fig. 3, A and B). In contrast, HR23B itself showed little affinity for any TFIIH subunits (Fig. 3C). This may be inconsistent with the observed behavior of its yeast counterpart (38). It has been reported that RAD23 interacts with in vitro translated RAD25 and TFB1, the yeast counterparts of human XPB and p62; however, the yeast TFIIH subunits used in that study were translated in rabbit reticulocyte lysates and not highly purified. Thus direct interaction between RAD23 and TFIIH subunits in yeast has yet to be determined. On the other hand, we cannot exclude the possibility that HR23B may somehow affect the interaction of XPC and TFIIH, because both HR23B and HR23A (another human homolog of RAD23) have been shown to stimulate the repair activity of XPC in cell-free NER reactions (22,26).
XPC-HR23B Recruits TFIIH to Damaged DNA-We recently reported that the XPC-HR23B complex specifically binds various NER lesions and functions as a damage detector initiating the repair reaction in the global genome repair pathway (18). Based on these findings, it is conceivable that XPC-HR23B bound to the lesion could recruit other NER factors, possibly via local conformational changes in the DNA as well as proteinprotein interactions. Importantly, TFIIH is the only NER factor that has been shown to interact directly with XPC; neither purified XPA nor RPA was co-precipitated with the XPC-HR23B beads (data not shown). We show in this report that TFIIH, like XPA, XPC, and RPA, can be precipitated from crude cell extracts with dsDNA immobilized on cellulose beads (Fig. 4), even though TFIIH exhibits little affinity for DNA by itself (20,35). Our observations indicate that the discrimination between damaged and undamaged DNA is modest among XPA, XPC, and RPA. We should point out, however, that the commercially available DNA-cellulose was used in this experiment, and so we cannot exclude the possibility that conformational changes in the DNA that were generated during the production of DNA-cellulose caused the binding of these NER proteins. It is also interesting to note that these NER proteins could be precipitated by dsDNA-cellulose or the damaged DNA substrate immobilized on Sepharose beads in an ATP-independent manner (Figs. 4 and 6B). It has been reported that ATP is necessary for formation of the stable excision nucleasedamaged DNA complex and for the dual incision reaction (19,20,34,36). However, some reports suggest that ATP hydrolysis is not required for formation of the DNA-NER protein complex (35,42). Our method may be sufficiently sensitive to detect relatively weak protein-protein and/or protein-DNA interactions formed in the absence of ATP. Importantly, the association of TFIIH with DNA was markedly stimulated by the presence of XPC ( Fig. 5) but not by XPA. Previous studies have reported that recombinant XPA may also be able to interact with purified TFIIH and possibly recruit it to damaged DNA (35,43). In contrast, our results strongly indicate that the protein-protein interaction between XPC and TFIIH is required for recruitment of TFIIH to damaged DNA in whole cell extracts. Furthermore, introduction of bulky adducts to dsDNA by N-AAAF treatment resulted in marked enhancement of both XPC and TFIIH binding to DNA (Fig. 6, A and C). Taken together, these observations indicate that TFIIH is recruited to the NER lesion via interaction with prebound XPC. These results, however, do not exclude the possibility that XPA is required to initiate the NER reaction. Wakasugi and Sancar (44) have proposed a model for the assembly of human NER factors on damaged DNA. Based on measurements of binding affinity to DNA containing a single (6 -4) photoproduct and a series of kinetic experiments using a reconstituted DNA repair system, they concluded that an RPA⅐XPA⅐DNA ternary complex forms first at the site of DNA lesions to initiate NER. In fact, the incision reaction has never been observed without XPA in the reconstituted excision reaction, probably because of the incomplete open complex formation (19,34,36). In contrast, our finding is consistent with the model proposed by Evans et al. (19), in which initial opening of dsDNA around the lesion may be conducted by XPC-HR23B and TFIIH. Li et al. (45) also demonstrated that XPC is an indispensable component of the initial step of NER that stabilizes the interaction of TFIIH with damaged DNA. Moreover, using gel mobility shift assays, Wakasugi and Sancar (20,44) demonstrated that the first stable NER protein complex specifically assembled on the lesion contains XPC-HR23B and TFIIH, although the complex also involves XPA, RPA, and XPG. The stability of the ternary complex consisting of XPC-HR23B, TFIIH, and damaged DNA may be insufficient to allow its detection by gel mobility shift assays, although we do not exclude the possibility that XPA, RPA, and XPG could be recruited concomitantly with TFIIH, because a network of interactions among these factors has been discerned (see Ref. 46 for review).
In conclusion, our current data clearly show that the XPC-HR23B complex, a detector of DNA lesions in global genome repair, plays a crucial role in the recruitment of TFIIH to damaged DNA. Our data correspond well with the recent molecular model for early global genome repair in which TFIIH is recruited to the site of repair in an XPC-HR23B complex-dependent manner following damage recognition by XPC-HR23B (46). However, it remains to be determined whether XPC has any effect on ATPase and/or helicase activities associated with TFIIH and whether HR23B is necessary for the recruitment of TFIIH by XPC. These issues are now under investigation in our laboratory. Our future studies are aimed at accurately determining the mechanism of early events in global genome repair.