Article No. mb981949 J. Mol. Biol. (1998) 281, 211±218

COMMUNICATION

Interactions of the Transcription/DNA Repair FactorTFIIH and XP Repair Proteins with DNA Lesions in aCell-free Repair Assay

Ruo-Ya Li1, Patrick Calsou2*, Christopher J. Jones3 and Bernard Salles2

1SocieÂte FrancËaise deRecherches et d'Investissements(SFRI), Berganton, 33127Saint Jean d'Illac, France2Institut de Pharmacologie et deBiologie Structurale, CNRSUPR 9062, 205 route deNarbonne, 31077 ToulouseCedex, France3University of Wales College ofMedicine, Dept of PathologyHeath Park, Cardiff, CF4 4XNUK

0022±2836/98/320211±08 $30.00/0

We have studied the interactions between DNA damage and human pro-teins involved in the early steps of nucleotide excision repair (NER) reac-tion under in vitro conditions with human protein extracts. By using anew assay, we have detected a long-lived DNA/protein complex invol-ving XPA and TFIIH in the course of the NER process. The formation ofthis complex is exclusively limited to DNA lesions that are substrates ofthe human excinuclease. We show that, while XPA binding to damagedDNA is ATP-independent, stable association of TFIIH with DNA lesionsis promoted by ATP hydrolysis and is dependent on the integrity of XPAand XPC proteins in the cell extract. In addition, XPC is necessary topromote a stable binding of XPA to UV-irradiated DNA. Finally, theco-binding of XPA and TFIIH to DNA damage is correlated to a dose-dependent titration of TFIIH and not XPA from the free protein fraction.

# 1998 Academic Press

Keywords: DNA damage; DNA repair; XPC; XPA; TFIIH; excision repair*Corresponding author

The genetic integrity of living organismsdepends on their capacity to repair damage pro-duced by endogenous and exogenous agents.Among various repair mechanisms, nucleotideexcision repair (NER) plays a major role by repair-ing a wide range of lesions in DNA, including UV-photoproducts and base modi®cations by manycarcinogenic or chemotherapeutic agents (Ma et al.,1995; Sancar, 1996; Wood, 1996). The sequence ofthe NER process consists of two broad steps: (1)lesion recognition, strand incision and damagedoligonucleotide displacement, (2) gap ®lling byDNA polymerization and ligation. It is commonlyaccepted that the events preceding DNA resynth-esis, i.e. preincision and incision/excision, are lim-iting in the repair reaction; accordingly, the celllines of complementation groups A to G of thehuman cancer prone inherited disease xerodermapigmentosum (XP) exhibit a defect in NER whichrelies on mutations in genes encoding proteins

E-mail address of corresponding author:calsou@ipbs.fr

Abbreviations used: NER, nucleotide excisionrepair; XP, xeroderma pigmentosum; MMS,methyl methanesulphonate; CDDP, cis-diamminedichloroplatinum-II; TFIIHp62, p62 subunit ofTFIIH; Ab, antibody; mAb, monoclonal antibody.

involved in these early steps of the repair pathway(Ma et al., 1995; Sancar, 1996; Wood, 1996).

A major challenge has been to determine theproteins necessary for damage removal and repairsynthesis. Several authors have successfully recon-stituted the NER reaction with highly puri®edyeast (Guzder et al., 1995) or human protein com-ponents (Aboussekhra et al., 1995; Mu et al., 1995,1996). In the case of human proteins, 16 polypep-tides in six fractions are suf®cient for the incision/excision steps: XPA, replication protein A (RPA),the general transcription factor TFIIH, XPC-HHR23B, XPG and XPF-ERCC1. TFIIH contains atleast nine polypeptides (Marinoni et al., 1997)including the helicases XPB and XPD, known asexcision repair proteins (Drapkin & Reinberg, 1994;Schaeffer et al., 1993, 1994; van Vuuren et al., 1994).TFIIH is involved in both DNA repair and tran-scription and a mutant form of TFIIH can inhibitboth processes (Hwang et al., 1996).

In order to dissect the potential interaction ofrepair proteins and damaged DNA in the preincision step of NER under conditions whichclosely resemble cellular repair, we have designeda new assay. Protein-DNA interactions take placeduring a NER reaction in a mixture of whole cellhuman protein extracts and damaged DNA. Here,we report the monitoring of repair proteins bind-

# 1998 Academic Press

212 XP Proteins and TFIIH Interactions with Damaged DNA

ing to damaged DNA in the course of the NERreaction, in parallel with the measurement of theextent of repair synthesis activity.

We took advantage of our published methodusing plasmid DNA adsorbed on sensitized micro-plates as substrate for in vitro repair (DamagedDNA Detection assay-3D assay; Salles et al., 1995).As shown in Figure 1 (arrow A), biotinylated-dUMP is incorporated during the DNA repair syn-thesis. The modi®ed nucleotide incorporation isquanti®ed in a chemiluminescent reaction after rec-ognition by an ExtrAvidin-peroxide conjugate. Asshown in Figure 2A, the level of incorporationincreased with the number of UV photoproductsper plasmid DNA circle and with the incubationtime. Under these repair conditions, the increasewas maximal with 450 J/m2 UV dose to plasmidDNA (about 15 photoproducts per circle; Woodet al., 1988) after three hours reaction time.

In parallel with the repair synthesis assay, pro-tein-DNA interactions are detected by an ELISA-like method coupled to chemiluminescence directlyin the microplate wells{ (Figure 1, arrow B). Inorder to focus on the preincision step of the NERprocess, XPA and TFIIHp62 were initially chosenas proteins representative of the damage recog-nition and the incision/excision steps of the NERprocess, respectively (Sancar, 1996; Wood, 1996).By using this approach, the binding of XPA andTFIIHp62 proteins to UV-irradiated plasmid DNAwas measured in an immunoreaction with speci®cantibodies (Figure 2B). Under the same conditionsas the repair synthesis assay (Figure 2A), theresults showed a UV-dose-dependent increase ofimmunoreactive proteins retained in the plasticwells suggesting that the proteins detected asspeci®cally bound to damaged DNA adsorbed inthe wells. XPA and TFIIHp62 exhibited a similarbinding pro®le. However, a direct quantitativecomparison could not be done due to differencesin epitope accessibility and Ab reactivity. Signi®-cant binding over the background was alreadydetected after 30 minutes incubation time withplasmid DNA irradiated with 300 J/m2 and thebinding increased up to two to three hours of reac-tion time (data not shown). Therefore, two hourswas chosen as standard incubation time in furtherexperiments. In addition, signi®cant binding ofTFIIHp62 to damaged DNA was detected with adose down to 5 J/m2 to plasmid DNA (data notshown).

An immunoblotting technique coupled to theDNA repair assay in microplates was also devel-oped for the detection of repair proteins eitherdetached from the damaged DNA immobilized onmicroplates (DNA-binding fraction) or present inthe supernatant (free fraction; Figure 1, arrow C).This allows us to validate the data obtained by

{ B. Salles, P. Calsou and R.-Y. Li, Procedure of DNAlesions detection by associated repair protein complexes,French Patent nb 9804064.

immunodetection of protein/DNA complexesdirectly in the microwells and also to avoid poten-tially false negative signals due to epitope accessi-bility. Taking into account that the desorption ofrepair proteins bound to DNA might not be exactlyquantitative, the results are broadly in accordancewith those found by the ELISA-like assay(Figure 2C). The amount of both XPA andTFIIHp62 proteins increased in the DNA-bindingfraction with the level of DNA damage. Inaddition, the amount of TFIIHp62 in the free frac-tion exhibited an inverse pro®le with a notabledecrease in the presence of plasmid DNA irra-diated with 450 J/m2. In contrast, the amount ofXPA in the free fraction remained identical what-ever the UV-dose to plasmid DNA. When themembranes were reblotted with a mAb directedagainst the XPB subunit of TFIIH, a binding pro®leidentical to that of TFIIHp62 was obtained(Figure 2C). Similar data were also obtained with amAb directed against XPD and p40 (CDK7 orMO15) subunits of TFIIH (data not shown). Thelatter results indicate that the data obtained on thebinding of TFIIHp62 to damaged DNA can be con-®dently extrapolated to the whole TFIIH complex.

Although numerous lesions are repaired byNER, some damage, like methylated DNA bases,are not repaired by this pathway (Wood, 1996).Thus, it was of interest to assess the substratespeci®city of XPA and TFIIHp62. DNA damageinduced by treatments of plasmid DNA with UV-C, CDDP or MMS were used as repair substrates.As shown in Figure 3A, a signi®cant DNA repairactivity of HeLa extracts was observed in the pre-sence of the three types of DNA lesions tested ascompared with control reactions carried out with-out DNA or with non-damaged DNA. Then, XPAand TFIIHp62 binding to these DNA lesions wastested by immunoblotting (Figure 3B). AlthoughXPA and TFIIHp62 proteins were detected onDNA damaged with UV-C or CDDP, no variationover the control background was observed in thepresence of MMS-damaged DNA. Similar resultswere obtained by immunoblotting with mAbdirected against XPB and XPD subunits of TFIIH(data not shown). These results indicate that bothXPA and TFIIH do not recognize MMS-damagedDNA as substrate in agreement with the knownpattern of DNA lesion recognition by NER. Inaddition, since under these repair conditions,MMS-damaged DNA stimulated repair synthesisactivity, this indicates that the corresponding DNAdamages are repaired by an ef®cient alternativepathway operating in cell extracts, namely baseexcision repair as already observed (Satoh et al.,1993).

Then, we have investigated whether formationof the preincision repair complex including XPAand TFIIH requires ATP and/or ATP hydrolysis.Plasmid DNA irradiated with UV-C light wasincubated with the reaction mixtures in the absenceof ATP or containing ATP or ATPgS. While theDNA repair synthesis increased with the ATP con-

Figure 1. Design of the experimental procedure. A, Chemiluminescent DNA repair synthesis assay in microplates.The repair synthesis assay was performed as described (Salles et al., 1995) with modi®cations: brie¯y, 50 ng of plas-mid DNA (pBS-pBluescript KS�; Stratagene) exposed to DNA damaging agents or untreated were adsorbed on sensi-tized 96-well microtiter plates (Microlite II, Dynatech) in 10 mM phosphate buffer (pH 7) for 20 minutes at 30�C. Therepair reaction was carried out in 50 ml and contained 150 mg protein extract from HeLa cells obtained as described(Manley et al., 1983) in 70 mM KCl, 40 mM Hepes-KOH (pH 7.6), 7 mM MgCl2, 2 mM ATP, 0.5 mM dithiothreitol,10 mM phosphocreatine, 2.5 mg of creatine phosphokinase, 2 mM EGTA, 18 mg of bovine serum albumin (BSA),0.4 mM each dGTP, dCTP, dATP and biotin-21-dUTP (Clontech). After a three hour incubation time at 30�C, the wellswere washed three times with phosphate buffered saline solution pH 7.4 (PBS) plus 0.01% (w/v) Tween-20 (PBST).Biotin-21-dUTP incorporated into DNA was detected by incubation with ExtraAvidin conjugated to peroxidase(Sigma) diluted 1/10,000 in PBS plus 0.025% (w/v) acetylated BSA and 0.05% (w/v) Igpal CA-630 (Sigma) for 15minutes at 30�C. Wells were washed three times with PBST buffer and 50 ml of chemiluminescent substrate mixture(10 mM Tris-HCl (pH 8.5), 1.28 mM luminol and 145 mM 4-iodophenol (both from Aldrich), 30% (w/w) H2O2 diluted1/25,000) was added for ®ve minutes at 30�C. Emitted light was measured with a luminometer (Lumax 2, SFRI,France) and expressed as a relative light unit (RLU). The DNA repair synthesis activity was expressed as the ratio ofRLU in treated versus untreated plasmid DNA. B, Microplate repair assay coupled to chemiluminescent immuno-detection of repair proteins bound to DNA. DNA repair reactions were performed at 30�C for two hours as describedabove. Wells were washed three times with PBST, and individual polypeptides bound to DNA were detected in thewells by incubation with a speci®c monoclonal or polyclonal antibody at 30�C for 30 minutes in PBS plus 0.025%acetylated BSA and 0.05% Igpal (mAbs anti-TFIIHp62 (3C9), XPB (1B3) and XPD (2F6), kind gifts from Dr J. M. Egly(Strasboug, France), polyclonal antibody anti-XPA obtained against a denatured His-tagged XPA protein (Jones &Wood, 1993) by rabbit immunization). The wells were then washed three times with PBST, secondary antibody conju-gated to horseradish peroxidase (Sigma) added to each microwell, incubated at 30�C for 15 minutes and then washedagain six times with PBST. The immuno-adsorbed complex was detected by 50 ml of chemilumininescent substratemixture as above for ®ve minutes at 30�C. Then the emitted light was measured with a luminometer and expressedas relative light units (RLU). The antibody reactivity was expressed as the ratio of RLU in treated versus untreatedplasmid DNA. C, Immuno-blotting analysis of the repair proteins either present in the supernatant or dissociatedfrom immobilized DNA. On completion of repair reaction, the free proteins fraction was removed from the micro-wells and boiled in Laemmli sample buffer. The protein fraction bound to DNA was analysed after four washes withPBST by elution with 30 ml of sample buffer (62.5 mM Tris-HCl (pH 6.8), 4 M urea, 10% (v/v) glycerol, 2% (w/v)SDS, 5% (v/v) b-mercaptoethanol, 0.003% (w/v) bromophenol blue; Shah et al., 1995) for 30 minutes at 30�C withshaking followed by 15 minutes denaturation at 85�C. Both free and DNA-binding fractions were resolved by SDS-PAGE, transferred to polyvinylidene di¯uoride membrane (Schleicher and Schuell) and analysed by standard immu-noblotting technique including an enhanced chemiluminescence detection system (ECL. Amersham). The ®lms werescanned and processed with adobe Photoshop 3.0 software.

XP Proteins and TFIIH Interactions with Damaged DNA 213

Figure 2. Dose-response dependent binding of XPA and TFIIH subunits to UV-C photoproducts. A standard repairreaction in microplate was performed at 30�C with plasmid DNA irradiated with UV-C-light (peak wave-length254 nm) at the indicated dose under conditions where 100 J/m2 produced an average of 2.7 cyclobutane pyrimidinedimers per molecule as described (Wood et al., 1988). A, The repair ratio was determined after one, two and threehours of incubation. Each value was the mean � S.D. of at least three experiments. B, TFIIHp62 and XPA repair pro-teins were immuno-detected in the microplate after two hour repair reactions. Each value was the mean � S.D. of atleast three experiments. C, XPA, XPB and TFIIHp62 binding to DNA was determined by Western-blotting analysisafter two hours repair reaction. The protein fraction retained in the microplate wells (left panel) and the protein frac-tion in the supernatant (right panel) were analysed in parallel.

214 XP Proteins and TFIIH Interactions with Damaged DNA

centration and reached a maximum at 500 mMATP, it did not occur in the absence of ATP or inthe presence of ATPgS (data not shown). Whendamaged DNA associated proteins were examined,XPA protein was equally immunodetected in theDNA binding fraction in the absence of ATP or inthe presence of ATPgS (Figure 4), suggesting thatneither ATP nor ATP hydrolysis were required forthe XPA function as damage recognition factor, asexpected (Jones & Wood, 1993). In contrast, anincrease in the amount of TFIIHp62 with ATP con-centration was observed and the non-hydrolysableATP analogue ATPgS did not promote p62 associ-ation with damaged DNA (Figure 4). This indicatesthat p62 binding to damaged DNA is promoted byATP hydrolysis. Detectable TFIIHp62 in the DNAbinding fraction in the absence of added ATPresulted most probably from the presence of con-taminating ATP in the cell-free extract, since p62DNA binding decreased signi®cantly uponaddition of 1 mM ATPgS (Figure 4). Stable TFIIHbinding to the site of lesion does not rely on con-formational change upon ATP binding, but mostprobably requires DNA unwinding by the ATP-dependent XPB/XPD helicase activities in order toposition the nuclease subunits properly (Evanset al., 1997). In addition, XPA binding to UV-

damaged DNA without TFIIH in the presence ofATPgS argues against the existence of preas-sembled XPA-TFIIH complex in the absence ofactive NER, as also ruled out recently in yeast(Guzder et al., 1996).

In order to provide further evidence for inter-actions between DNA damage, XPA and TFIIH,the DNA repair activity and TFIIHp62-DNA bind-ing activity of cell-free extracts from control andXP-A cells were analyzed in parallel in our micro-plate assay. As shown in Figure 5A, both extractsexhibited similar DNA repair synthesis activity onMMS damaged-DNA substrate. This result con-®rms that repair of MMS DNA lesions in vitro isnot due to NER, as shown above. In contrast andas expected, extracts from XP-A cells failed to carryout DNA repair synthesis on UV-C-damaged DNA(Figure 5B). This de®ciency relies on the absence offunctional XPA protein in these extracts and can befully restored by addition of puri®ed XPA protein(Robins et al., 1991). This de®ciency in UV-C DNAdamage repair paralleled a very low immunode-tected p62-binding to UV-irradiated DNA with XP-A extracts (Figure 5C). Although the puri®ed yeastXPD homologue has been shown to bind damagedDNA (Sung et al., 1993) and TFIIH has beenreported to exhibit some band shift with damaged

Figure 4. Effect of ATP on XPA and TFIIHp62 bindingto UV damage. A repair reaction in the microplate wasperformed for two hours at 30�C in the presence of plas-mid DNA damaged with UV-C light (300 J/m2) underconditions as indicated. The binding of XPA andTFIIHp62 to DNA was determined by Western-blottinganalysis of the protein fraction retained in the micro-plate wells.

Figure 3. Speci®city of XPA and TFIIHp62 binding tovarious DNA damage. A standard repair reaction in themicroplate was performed for two hours at 30�C with-out DNA or with plasmid DNA as indicated, eitherundamaged or damaged with UV-C light (300 J/m2),CDDP (Hansson & Wood, 1989, �ten platinum adductsper plasmid molecule) or MMS (MMS at 5 mM in10 mM phosphate buffer (pH 7.4) and 1 mg/ml pBSplasmid DNA for 30 minutes at 37�C; the mixture wasthen added to microplate wells under DNA bindingconditions and wells were rinsed twice to remove freeMMS). A, DNA repair synthesis activity was determinedby chemiluminescent reaction. Light units were dividedby the value obtained with undamaged DNA in orderto calculate the repair ratio. B, XPA and TFIIHp62 in theprotein fraction retained in the microplate wells weredetected by Western-blotting analysis.

XP Proteins and TFIIH Interactions with Damaged DNA 215

DNA (Park et al., 1995), these data demonstratethat under our conditions of whole cell extractsoperating in vitro on damaged DNA, XPA isrequired for the stable binding of TFIIH to UVlesions.

There are controversial reports about the exist-ence of XPA-TFIIH/damaged DNA complexes. Anattempt to detect such a complex by gel mobilityshift assay with puri®ed components was unsuc-cessful (Park et al., 1995). However, by using pulldown experiments with GST-XPA beads andhighly UV-irradiated DNA, some authors reportedrecently the recruitment of TFIIH to XPA-damagedDNA complexes (Nocentini et al., 1997). However,other repair proteins may be required for the for-mation of a long-lasting DNA-protein complexinvolving XPA and TFIIH. For example, an associ-ation of XPC and TFIIH may exist (Drapkin &

Reinberg, 1994), although this has not beendetected in another report (van der Spek et al.,1996). Then we have analyzed XPA and TFIIHbinding to damaged DNA in nuclear extracts fromXP-C ®broblasts (Figure 6). As expected, XP-Cextracts were pro®cient for the repair of MMS-damaged DNA but exhibited a de®ciency in UV-CDNA damage repair (Figure 6A). Notably, thenuclear extracts from control ®broblasts showedthe same XPA and TFIIHp62 speci®c bindingactivity to UV-damaged DNA as HeLa extracts(Figures 2A, 6A). In sharp contrast, XPC defectiveextracts exhibited a complete defect in the bindingof both XPA and TFIIHp62 to UV-irradiated DNA(Figure 6B). This dramatic effect could not rely onthe expression of these proteins in XP-C extracts,since the control and XPC-defective extracts con-tained identical amounts of XPA and TFIIHp62(Figure 6B). XPC is dispensable for the repair oflesions on partially unwound DNA (Mu et al.,1996; Mu & Sancar, 1997). A recent report estab-lished that XPC, in addition to TFIIH helicaseactivity is absolutely required for the initial step inthe open complex formation around the lesion(Evans et al., 1997). The present data demonstratethat XPC is a key component in the earliest stepsin the NER mechanism which is indispensable forstabilizing both XPA and TFIIH interactions withdamaged DNA. Since we have repeatedly foundthat the p34 subunit of RPA was also present withXPA and TFIIH in this protein-DNA complex (ourunpublished results), we can reasonably infer thatRPA, XPA, TFIIH, and XPC (possibly associatedwith HHR23B) are necessary to form the preinci-sion complex on DNA lesions during the regularNER process, in agreement with proposed models(Sancar, 1996; Wood, 1996).

In the course of the NER reaction in vitro, a par-allel was observed between time-dependentincrease in repair synthesis and repair proteins

Figure 5. XPA-dependent binding TFIIHp62 to UV damage. A standard repair reaction in the microplate was per-formed for two hours at 30�C with protein extracts from HeLa or XP-A lymphoblastoid cell line GM02345B (Japanesepatient XP2OS, Coriell Institute for Medical Research, Camden, NJ, USA) on plasmid DNA treated with MMS or UV-C light. A, B, DNA repair synthesis activity was determined by chemiluminescent reaction. Light units were dividedby the value obtained with undamaged DNA in order to calculate the repair ratio. C, TFIIHp62 was immuno-detected in the microplate. Light units were divided by the value obtained with undamaged DNA in order to calcu-late the antibody reactivity.

Figure 6. XPC-dependent binding of XPA and TFIIHp62to UV damage. A standard repair reaction in the micro-plate was performed for two hours at 30�C with 100 mgof nuclear protein extracts as described (Muller et al.,1998) from MRC5SV control or XP-C (XP4PASV) (Daya-Grosjean et al., 1987) ®broblastic cell lines (obtainedfrom Dr M, Mezzina, Villejuif, France) on plasmid DNAtreated with MMS (2 mM) or UV-C light (450 J/m2). A,DNA repair synthesis activity was determined by che-miluminescent reaction. Light units were divided by thevalue obtained with undamaged DNA in order to calcu-late the repair ratio. B, XPA and TFIIHp62 in the nuclearproteins extracts and in the protein fraction retained inthe microplate wells were detected by Western-blottinganalysis.

216 XP Proteins and TFIIH Interactions with Damaged DNA

binding to damaged DNA (Figure 2). This mayindicate that during the time interval when a sub-set of preincision complexes go to completion ofthe reaction and lead to repair synthesis, most ofthe preincision complexes accumulate as stableintermediates involving TFIIH among other repairproteins. These results in vitro would suggest that,under a certain ratio of DNA lesions to free TFIIHwithin the cell, TFIIH could be titrated out byDNA damage. Recruitment of TFIIH to damagesites may be involved in p53-mediated cell cyclearrest at G1/S (Jones & Wynford-Thomas, 1995).TFIIH is limiting for RNA pol II transcriptionelongation (Friedberg, 1996; Svejstrup et al., 1996).Since during regular NER, TFIIH is engaged in apreincision complex that appears long-lasting asshown above, there may be a massive recruitmentof TFIIH at damaged DNA sites in the nucleus.This phenomenon could contribute to DNAdamage-induced transcription shut-down, whichcould also rely non-exclusively on transcriptionarrest at sites of base damage (van Hoffen et al.,1995) and hijacking of the TATA-box binding pro-tein TBP by some classes of DNA lesions (Vichiet al., 1997).

Acknowledgments

We are grateful to Dr J. M. Egly for the gift of anti-bodies against TFIIH subunits and Drs M. Mezzina andA. Sarasin for the gift of cell lines. This study was sup-ported by grants from the Association pour la Recherchesur le Cancer, the Ligue Nationale Contre le Cancer andthe MinisteÁre de l'Enseignement SupeÂrieur et de laRecherche (ACCSV no. 14). C.J.J. was supported by agrant from the Nuf®eld Foundation, UK.

XP Proteins and TFIIH Interactions with Damaged DNA 217

References

Aboussekhra, A., Biggerstaff, M., Shivji, M. K. K., Vilpo,J. A., Moncollin, V., Podust, V. N., Protic, M.,Hubscher, U., Egly, J. M. & Wood, R. D. (1995).Mammalian DNA nucleotide excision repair recon-stituted with puri®ed protein components. Cell, 80,859±868.

Daya-Grosjean, L., James, M. R., Drougard, C. &Sarasin, A. (1987). An immortalized xeroderma pig-mentosum group C cell line which replicates SV40shuttle vectors. Mutat. Res. 183, 185±196.

Drapkin, R. & Reinberg, D. (1994). The multifunctionalTFIIH complex and transcriptional control. TrendsBiochem. Sci. 19, 504±508.

Evans, E., Fellows, J., Coffer, A. & Wood, R. D. (1997).Open complex formation around a lesion duringnucleotide excision repair provides a structure forcleavage by human XPG protein. EMBO J. 16, 625±638.

Friedberg, E. C. (1996). Relationship between DNArepair and transcription. Annu. Rev. Biochem. 65,15±42.

Guzder, S. N., Habraken, Y., Sung, P., Prakash, L. &Prakash, S. (1995). Reconstitution of yeast nucleo-tide excision repair and puri®ed Rad proteins, repli-cation protein A, and transcription factor TFIIH.J. Biol. Chem. 270, 12973±12976.

Guzder, S. N., Sung, P., Prakash, L. & Prakash, S.(1996). Nucleotide excision repair in yeast ismediated by sequential assembly of repair factorsand not by a pre-assembled repairosome. J. Biol.Chem. 271, 8903±8910.

Hansson, J. & Wood, R. D. (1989). Repair synthesis byhuman cell extracts in DNA damaged by cis- andtrans-diamminedichloroplatinum(II). Nucl. Acids Res.17, 8073±8091.

Hwang, J. R., Moncollin, V., Vermeulen, W., Seroz, T.,Vanvuuren, H., Hoeijmakers, J. H. J. & Egly, J. M.(1996). A 30- > 50XPB helicase defect in repair/tran-scription factor TFIIH of xeroderma pigmentosumgroup B affects both DNA repair and transcription.J. Biol. Chem. 271, 15898±15904.

Jones, C. J. & Wood, R. D. (1993). Preferential binding ofxeroderma pigmentosum groups A complementingprotein to damaged DNA. Biochemistry, 32, 12096±12104.

Jones, C. J. & Wynford-Thomas, D. (1995). Is TFIIH anactivator of the p53-mediated G(1)/S checkpoint?Trends Genet. 11, 165±166.

Ma, L. B., Hoeijmakers, J. H. J. & van der Eb, A. J.(1995). Mammalian nucleotide excision repair. Bio-chim. Biophys. Acta, 1242, 137±163.

Manley, J. L., Fire, A., Samuels, M. & Sharp, P. A.(1983). In vitro transcription: whole-cell extract.Methods Enzymol. 101, 568±582.

Marinoni, J. C., Roy, R., Vermeulen, W., Miniou, P.,Lutz, Y., Weeda, G., Seroz, T., Gomez, D. M.,Hoeijmakers, J. H. J. & Egly, J. M. (1997). Cloningand characterization of p52, the ®fth subunit of thecore of the transcription/DNA repair factor TFIIH.EMBO J. 16, 1093±1102.

Mu, D. & Sancar, A. (1997). Model for XPC-independenttranscription-coupled repair of pyrimidine dimersin humans. J. Biol. Chem. 272, 7570±7573.

Mu, D., Park, C. H., Matsunaga, T., Hsu, D. S., Reardon,J. T. & Sancar, A. (1995). Reconstitution of humanDNA repair excision nuclease in a highly de®nedsystem. J. Biol. Chem. 270, 2415±2418.

Mu, D., Hsu, D. S. & Sancar, A. (1996). Reaction mech-anism of human DNA repair excision nuclease.J. Biol. Chem. 271, 8285±8294.

Muller, C., Calsou, P., Frit, P., Cayrol, C., Carter, T. &Salles, B. (1998). UV sensitivity and impairednucleotide excision repair in DNA-dependent pro-tein kinase mutant cells. Nucl. Acids Res. 26, 1382±1389.

Nocentini, S., Coin, F., Saijo, M., Tanaka, K. & Egly, J. M.(1997). DNA damage recognition by XPA proteinpromotes ef®cient recruitment of transcription fac-tor II H. J. Biol. Chem. 272, 22991±22994.

Park, C. H., Mu, D., Reardon, J. T. & Sancar, A. (1995).The general transcription-repair factor TFIIH isrecruited to the excision repair complex by the XPAprotein independent of the TFIIE transcription fac-tor. J. Biol. Chem. 270, 4896±4902.

Robins, P., Jones, C. J., Biggerstaff, M., Lindahl, T. &Wood, R. D. (1991). Complementation of DNArepair in xeroderma pigmentosum group A cellextracts by a protein with af®nity for damagedDNA. EMBO J. 10, 3913±3921.

Salles, B., Provot, C., Calsou, P., Hennebelle, I., Gosset,I. & Fournie, G. J. (1995). A chemiluminescentmicroplate assay to detect DNA damage inducedby genotoxic treatments. Anal. Biochem. 232, 37±42.

Sancar, A. (1996). DNA excision repair. Annu. Rev. Bio-chem. 65, 43±81.

Satoh, M. S., Poirier, G. G. & Lindahl, T. (1993). NAD�-dependent repair of damaged DNA by human cellextracts. J. Biol. Chem. 268, 5480±5487.

Schaeffer, L., Roy, R., Humbert, S., Moncollin, V.,Vermeulen, W., Hoeijmakers, J. H. J., Chambon, P.& Egly, J.-M. (1993). DNA repair helicase: a com-ponent of BTF2 (TFIIH) basic transcription factor.Science, 260, 58±63.

Schaeffer, L., Moncollin, V., Roy, R., Staub, A., Mezzina,M., Sarasin, A., Weeda, G., Hoeijmakers, J. H. J. &Egly, J. M. (1994). The ERCC2/DNA repair proteinis associated with the class II BTF2/TFIIH transcrip-tion factor. EMBO J. 13, 2388±2392.

Shah, G. M., Poirier, D., Duchaine, C., Brochu, G.,Desmoyers, S., Lagueux, J., Verreault, A., Ho¯ack,J.-C., Kirkland, J. B. & Poirier, G. (1995). Methodsfor biochemical study of poly(ADP-ribose) metab-olism in vitro and in vivo. Anal. Biochem. 227, 1±13.

Sung, P., Watkins, J. F., Prakash, L. & Prakash, S. (1993).Negative superhelicity promotes ATP-dependentbinding of yeast RAD3 protein to ultraviolet-damaged DNA. J. Biol. Chem. 269, 8303±8308.

Svejstrup, J. Q., Vichi, P. & Egly, J. M. (1996). The mul-tiple roles of transcription/repair factor TFIIH.Trends Biochem. Sci. 21, 346±350.

van der Spek, P. J., Eker, A., Radmakers, S., Visser, C.,Sugasawa, K., Masutani, C., Hanaoka, F., Bootsma,D. & Hoeijmakers, J. H. J. (1996). XPC and humanhomologs of RAD23: intracellular localization andrelationship to other nucleotide excision repair com-plexes. Nucl. Acids Res. 24, 2551±2559.

van Hoffen, A., Venema, J., Meschini, R., van Zeeland,A. A. & Mullenders, L. H. F. (1995). Transcription-coupled repair removes both cyclobutane pyrimi-dine dimers and 6-4 photoproducts with equal ef®-ciency and in a sequential way from transcribedDNA in xeroderma pigmentosum group C ®bro-blasts. EMBO J. 14, 360±367.

van Vuuren, A. J., Vermulen, W., Ma, L., Weeda, G.,Appeldoorn, E., Jaspers, N. G. J., van der Eb, A. J.,Bootsma, D., Hoeimakers, J. H. J., Humbert, S. L. S.

218 XP Proteins and TFIIH Interactions with Damaged DNA

& Egly, J. M. (1994). Correction of xeroderma pig-

mentosum repair defect by basal transcription factor

BTF2(TFIIH). EMBO J. 13, 1645±1653.

Vichi, P., Coin, F., Renaud, J.-P., Vermulen, W.,

Hoeijmakers, J. H. J., Moras, D. & Egly, J.-M. (1997).

Cisplatin- and uv-damaged DNA lure the basal

transcription factor TFIID/TBP. EMBO J. 16, 7444±7456.

Wood., R. D. (1996). DNA repair in eukaryotes. Annu.Rev. Biochem. 65, 135±167.

Wood, R. D., Robins, P. & Lindahl, T. (1988). Comple-mentation of the xeroderma pigmentosum DNArepair defect in cell-free extracts. Cell, 53, 97±106.

Edited by M. Yaniv

(Received 9 February 1998; received in revised form 7 May 1998; accepted 12 May 1998)