TES

SDO

R

Rr(oAc(cAoamsmpACiceHhca

esarrDttv

1

Biochemical and Biophysical Research Communications 261, 859–863 (1999)

Article ID bbrc.1999.1125, available online at http://www.idealibrary.com on

ranscriptional Activation of Apurinic/Apyrimidinicndonuclease (Ape, Ref-1) by Oxidativetress Requires CREB

abine Grosch and Bernd Kaina1

ivision of Applied Toxicology, Institute of Toxicology, University of Mainz,bere Zahlbacher Strasse 67, D-55131 Mainz, Germany

eceived July 8, 1999

Myb, thereby affecting the transcriptionally activatingprrnhRsgoiips

itippohsThicemBbc1eiinis

Apurinic/apyrimidinic endonuclease (APE aliasef-1) is a multifunctional enzyme involved in DNA

epair and redox regulation of transcription factorse.g., AP-1). It also acts as a repressor of its own andther genes. Recently, it was shown that the level ofPE mRNA and protein is enhanced upon treatment ofells with oxidative agents, such as hydrogen peroxideH2O2), which gives rise to an adaptive response ofells to oxidative stress. Induction of APE is due toPE promoter activation. To elucidate the mechanismf transcriptional activation of APE by oxidativegents, we introduced mutations into the cloned hu-an APE promoter and checked its activity in tran-

ient transfection assays. Here we demonstrate thatutational inactivation of a CREB binding site (CRE)

resent within the promoter completely abolishedPE promoter activation by H2O2, indicating thatREB is required for APE induction. The CRE element

n the context of the APE promoter sequence binds-Jun and ATF-2, which was shown in gel retardationxperiments. Under conditions of induction of APE by

2O2, the expression of c-Jun was significantly en-anced, which supports the view that induction of-Jun is involved in signaling leading to APE promoterctivation by oxidative stress. © 1999 Academic Press

The multifunctional protein apurinic/apyrimidinicndonuclease (APE) is a key enzyme of the base exci-ion repair (BER) pathway (1). It cleaves apurinic/pyrimidinic sites in DNA and thus allowes them to beepaired by other enzymes involved in BER. Also, itemoves 39phosphoglycolate residues from brokenNA ends arising upon ionizing radiation and oxida-

ive treatments (2, 3). Besides these DNA repair func-ions, APE (alias Ref-1) acts as a redox regulator ofarious transcription factors such as AP-1, NF-kB and

1 To whom correspondence should be addressed. Fax: 0049-6131-7-3421. E-mail: kaina@mail.uni-mainz.de.

859

roperties of these proteins (4). APE also acts as aepressor by binding to the so-called negative calcium-esponsive element (nCaRE) and thus is involved inegative regulation of genes, such as the parathyroidormone gene and of its own (5, 6). Mice lacking APE/ef-1 are not viable indicating that the protein is es-ential for normal development and possibly also cellrowth (7). The multifunctional activities of APE, bothn the level of DNA repair and gene regulation, makest a prime candidate for functional and regulatory stud-es. It rises the question as to the importance of therotein for cellular defense upon exposure to genotoxictress as well as to the regulation of APE itself.That APE is subject to regulation by external factors

s indicated by the finding that the protein becomesranscriptionally stimulated by thyrotropin treatmentn rat thyroid FRTL-5 cells (8). Most interestingly, APEroved to be enhanced in its expression both on mRNA,rotein and activity level upon exposure of cells toxidative stress. Thus, hydrogen peroxide (H2O2) andypochloric acid caused induction of APE, which washown to be due to APE promoter activation (9, 10).his finding is of particular interest, since H2O2 andypochloric acid are endogenously formed during the

nflammatory response of macrophages and lympho-ytes. Endogenous oxygen radical burst may cause el-vation in the level of DNA damage which, for defense,ay require an increase in APE level thus enhancingER capacity. Indeed, induction of APE was found toe accompanied by an adaptive response of cells to theytotoxic and clastogenic activity of oxidative agents (9,0) indicating a physiological relevance of the phenom-non. Given this data, APE can be considered as annducible DNA repair gene in mammalian cells whichs subject to regulation by various external stimuli,otably oxidative stress. The cellular factors involved

n transcriptional activation of APE upon oxidativetress are unknown. Here we report that a functional

0006-291X/99 $30.00Copyright © 1999 by Academic PressAll rights of reproduction in any form reserved.

CREB binding site within the human APE promoter iselee

M

cfCccAASaoFCpCTTnsiCsm

tpwpD(p4tpt

aaTaCC5CGp0C7b0nimAt0

to a final concentration of 20% and aliquots were stored at 280°C. Thea

oc41datnpmas1aca

dhmDcgnHsptb2faTiHfat

R

hbpbaCfptadknw

imt

Vol. 261, No. 3, 1999 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

ssential for APE induction. We also show that theevel of Jun/ATF-2, which binds to the CRE element, islevated in cells upon exposure to oxidative agentsvoking the response.

ATERIALS AND METHODS

Cell culture, plasmids, and sited-directed mutagenesis. CHO-9ells were grown in F12-Dulbecco9s medium, supplemented with 5%etal calf serum (Life Technologies, Inc.) at 37°C in an atmosphere of 7%O2/air as previously described (9). The human APE promoter was

loned by PCR from HeLa genomic DNA using primers that werehosen according to the sequence previously published (9) 59-CACGCATGCTTAGGAAGATGGAAGGC-39 (primer 1) and 59-CACGTCGACCCACTCACATCTAATCC-39 (primer 2). A 900-bpphI/NruI fragment was cloned into pCATBasic (Promega) (thus cre-ting the plasmid hAPE SphI/NruI CATBasic). The AP-1 and CRE sitef the human APE promoter was mutated by site directed mutagenesis.or PCR, the following primers we used: 59-CATTGAGCAACTGA-TG-39 (AP-1 primer 1); 59-GCAAGCTTGAGTTGGGACCC-39 (AP-1rimer 2); 59-ACACGCATGCTTAGGAAGATGGAAGGC-39; 59-ACACG-ATGCTTAGGAAGATGGAAGGC-39 (CRE primer 1) 59-CGGACT-ACATAACGCCTCCT-39 (CRE primer 2) 59-TGTGCTGTCGACCC-ACTCGCGAGATCTGC-39 (CRE primer 3). The 59 primer termi-ated in a SphI site whereas the 39 primer terminated in a SalIite. The SphI/SalI PCR fragments was cloned into pCATBasic (lead-ng to the plasmid APE (AP-1mut) CATBasic and APE (CREmut)ATBasic). The “wild-type” and mutated sequences were checked byequencing of the promoters using the T7-sequencing kit from Phar-acia Biotech (Freiburg).

Transient transfection experiments and CAT assay. Cells wereransfected with human APE promoter-CAT constructs by calciumhosphate coprecipitation as described (11). In brief, 5 3 105 cellsere seeded per 10-cm dish and treated 24 h later with 1 ml of DNArecipitate containing 10 mg each of plasmid and salmon spermNA. After overnight incubation, cells were treated with DMSO

final concentration, 10%) for 10 min and incubated for another 24 heriod. Thereafter they were treated with the mutagen. After further8 h, cells were harvested for sonication. The amount of protein inotal cell extracts was determined (12), and the amount of CATrotein was measured by CAT-ELISA according to the manufac-urer9s protocol (Boehringer Mannheim).

Gel retardation assay. The gel retardation assays were performeds described (13). The following pairs of oligonucleotides were annealednd the double-stranded DNAs were labeled with g-(32P)dATP using4-polynukleotide kinase (Boehringer Mannheim). (1) APE-CRE:) 59-AGGAGGCGTGACGTAAGTCCG-39, b) 59-CGGACTTACGTCA-GCCTCCT-39; (2) APE-CREmut: a) 59-AGGAGGCGTTATGTAAGT-CG-39, b) 59-CGGACTTACATAACGCCTCCT-39; (3) CREB/ATF: a)9-AGAGATTGCCTGACGTCAGAG-39, b) 59-CTAGCTCTCTGACGT-AGGCAATCTCT-39; (4) Collagenase AP-1 (Col/AP-1): a) 59-AGT-GTGACTCATCACT-39, b) 59-AGTGATGAGTCACCACT-39. For pre-aring nuclear protein extracts, cells were not treated or treated with.3 mM H2O2 for different periods of time and harvested by scraping.ell pellets were resuspended in 1 ml lysis buffer I (10 mM Tris-Cl pH.4, 10 mM NaCl, 3 mM MgCl2, 1 mM PMSF, 2 mM DTT) and incu-ated for 10 min on ice. After addition of NP-40 (final concentration.5%) the solution was vortexed and centrifuged (400 g, 5 min). Theuclear pellet was washed with lysis buffer I. Pellets were resuspended

n 2 vol lysis buffer II (20 mM HEPES-KOH, pH 7.4, 600 mM KCl, 0.2M EDTA, 1 mM PMSF, 2 mM DTT) and incubated for 30 min on ice.fter centrifugation (10,000 g, 10 min) the supernatant was diluted by

he addition of 1 vol of lysis buffer III (20 mM HEPES-KOH, pH 7.4,.2 mM EDTA, 0.5 mM PMSF, 2 mM DTT). Glycerol was added

860

mount of protein was determined as described above.For gel retardation analysis (“bandshifts”) 2.5 mg or 1 mg (in case

f competition experiments) of protein of nuclear extracts were in-ubated in 10% glycerol, 10 mM HEPES-KOH, pH 7.9, 50 mM KCl,mM MgCl2, 4 mM Tris-Cl, 0.5 mM DTT, 0.5 mM EDTA, 1 mg BSA,mg poly (dI-dC) together with 25 fmol of radioactively labeled DNAuplex for 30 min at room temperature. For competition, a 5-, 10-nd 20-fold molar excess of non-labeled oligo nucleotids was added tohe incubation mixture. For supershift experiments, 5 mg of proteinuclear extract were mixed with 10% glycerol, 10 mM HEPES-KOH,H 7.9, 50 mM KCl, 4 mM MgCl2, 4 mM Tris-Cl, 0.5 mM DTT, 0.5M EDTA, 1 mg BSA, 1 mg poly (dI-dC) and, additionally, 5 mg of

ntibody (c-Jun, ATF-2 or ATF-3) which were purchased for super-hift analyses (from SANTA CRUZ). The mixture was incubated forh at 4°C. Thereafter 25 fmol of radioactively labeled duplex was

dded and incubated for 30 min at room temperature. DNA-proteinomplexes were separated on a 4% PAA gel using 0.25 3 TBE buffert 100 V.

Western blot analysis. Cells were treated with 0.3 mM H2O2 forifferent times. For preparing nuclear protein extracts, cells werearvested and cell pellets were resuspended in 1 ml lysis buffer (10M Tris-Cl pH 7.4, 10 mM NaCl, 3 mM MgCl2, 1 mM PMSF, 2 mMTT) and incubated for 10 min on ice. After addition of NP-40 (final

oncentration 0.5%) the solution was vortexed and centrifuged (400, 5 min) and the pellet was washed with lysis buffer. Thereafter,uclear pellets were resuspended in sonication buffer (20 mM Tris-Cl pH 8.5; 1 mM EDTA; 1 mM b-mercaptoethanol; 5% glycerol) and

onicated. Aliquots of 40 mg of nuclear protein extracts were electro-horetically separated onto a 10% SDS-PAGE and then electroblot-ed onto nitrocellulose (Schleicher and Schuell, Germany). The mem-rane was incubated overnight in 5% non-fat dry milk, 0.2% Tween-0, PBS. Then the filter was incubated with polyclonal antiserumrom rabbit rised against c-Jun and ATF-2 protein (SANTA CRUZ) insolution containing 5% dry milk, 0.2% Tween-20 and PBS for 2 h.he filters were extensively rinsed with 0.2% Tween-20 in PBS and

ncubated with peroxidase-conjugated goat anti-rabbit IgG (Dianova,amburg) diluted 1:5000 in 5% dry milk, 0.2% Tween-20 and PBS,

or 1 h. After extensive rinsing in 0.2% Tween-20 in PBS, protein-ntibody complexes were visualised by ECL (Amersham) accordingo the manufacturers protocol.

ESULTS AND DISCUSSION

To analyze APE regulation, a 900 bp fragment of theuman APE promoter was cloned by PCR, confirmedy sequencing, and cloned in front of the CAT gene ofCATBasic. A search for putative transcription factorinding sites revealed, among others, the presence ofn AP-1 site (39 of the transcription start site) andREB binding site (designated as CRE) within the

ragment (Fig. 1A). The inducibility of the APEromoter-CAT construct was studied in transientransfection assays in CHO cells treated with differentgents. There was no induction upon treatment withexamethasone, the tumor promoter TPA and the al-ylating agent methyl methanesulfonate (MMS). Sig-ificant stimulation of the APE promoter was observedith the oxidative agents H2O2 and NaOCl (Table 1).To elucidate wether AP-1 site is required for APE

nduction by oxidative stress, an APE promoter frag-ent was constructed by introducing base exchanges

hat prevented AP-1 binding, which was shown by gel

rtbTft

tmmciCiHfe

cc

ttoDdb

upmpu(ccwwafD

TABLE 1

paCoca

Vol. 261, No. 3, 1999 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

etardation assays in a previous work (14). The mu-ated promoter construct, however, was still inducibley H2O2 to an extent similar to the control (Fig. 1B).his indicates that the AP-1 site within the cloned APE

ragment is not essential for APE induction by oxida-ive stress.

A further possible candidate triggering APE induc-ion by oxidative agents is the cAMP responsive ele-ent (CRE). To analyze its role in APE induction,utations were introduced into the CRE site and

hecked for effect on promoter-CAT expression. Exper-ments with the original and mutated CRE promoter-AT constructs revealed a significantly reduced capac-

ty of the mutated promoter to become activated by2O2 (Fig. 1B). The basal activity of the APE promoter

ragment was not changed by introducing mutationsither into the AP-1 or CRE site (data not shown).

FIG. 1. Transcriptional activity of APE promoter constructspon treatment of cells with H2O2. (A) The cloned human APEromoter fragment spans from 2768 to 1114 bp (SphI–NruI frag-ent). The sequence (not shown) was identical to one previously

ublished (19). Putative transcription factor binding sites (search forsing computer program “signal scan”) were schematically indicated.B) Transcriptional activation of the APE (SphI/NruI) CATBasiconstruct, APE (AP-1mut) CATBasic, and APE (CREmut) CATBasiconstruct by oxidative treatment. The APE promoter-CAT constructsere transfected into CHO cells, which were subsequently treatedith H2O2 for 48 h with the doses indicated. Cells were harvestednd CAT protein was measured by CAT-ELISA assay. Inductionactors were given in relation to untreated control transfections.ata are the mean of three independent experiments 6 S.D.

861

In gel retardation experiments the protein bindingapacity of the CRE element was analyzed. The basehanges introduced into the CRE sequence of the APE

FIG. 2. DNA binding activity of nuclear extract from CHO con-rol cells to the 20-mer APE-CRE and APE-CREmut oligonucleo-ides. The radioactively labeled oligonucleotides were incubated withr without 2.5 mg nuclear extract for 30 min at room temperature.NA protein complexes were separated by electrophoresis on a non-enaturing 4% PAA gel. Arrow indicates the specific DNA proteininding complex.

Induction of APE Promoter in Cells Treatedwith Various Agents

Treatment Induction factor

NaOCl (0.1 mM) 2.4 6 1.5NaOCl (0.5 mM) 4.3 6 3.3H2O2 (0.3 mM) 3.4 6 0.8H2O2 (0.5 mM) 4.0 6 2.2Dexamethasone (100 nM) 1.2 6 0.2TPA (100 ng/ml) 1 6 0MMS (1 mM) 1.6 6 0.3

Note. CHO cells were transient transfected with the human APEromoter-CAT construct (SphI/NruI) and treated 24 h later with thegents indicated for another 48 h. Thereafter cells were harvested andAT protein was determined by CAT-ELISA (see Materials and Meth-ds). For calculation of the induction factor, CAT protein level fromontrol cells (not-treated) was set to 1. Induction factors are the mean oft least three independent experiments 6 standard deviation.

paeirnpBotttgwiqwAppt

To investigate the binding proteins in more detail,gtFsnaA(C4ba

iidHtyi

CitTabA(c

CACotDPAAodAms

Vol. 261, No. 3, 1999 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

romoter clearly reduced protein binding, as shownfter incubation of the oligonucleotide with nuclearxtract (Fig. 2). In competition experiments (see Fig. 3)t is shown that this binding is specific, as detected byeduced binding if competition was done with an oligo-ucleotide containing the CRE sequence of the APEromoter (in the following designated as APE-CRE).oth a collagenase AP-1 and a CREB/ATF sequenceligonucleotide were able to compete for the protein(s)hat bind to the APE-CRE sequence, whereas the mu-ated APE-CRE oligonucleotide did not display compe-ition effect (Fig. 3A). In a control experiment, an oli-onucleotide with the collagenase AP-1 binding siteas radioactively labeled and tested for protein bind-

ng. In this assay, also the CRE within the APE se-uence context was able to compete for protein binding,hich was not observed for the mutated CRE of thePE promoter (Fig. 3B). These data indicate that therotein complex that binds to the CRE of the APEromoter belongs to the AP-1 and CREB family ofranscription factors.

FIG. 3. Competition experiments of radioactively labeled APE-RE oligonucleotide with nonlabeled APE-CRE, APE-CREmut, Col/P-1, and CREB/ATF oligonucleotides. (A) 1 mg nuclear extract fromHO control cells was incubated with 2.5 fmol labeled APE-CREligonucleotide and different molar excess nonlabeled oligonucleo-ides as indicated. After 30 min of incubation at room temperature,NA-protein complexes were separated by electrophoresis on a 4%AA gel. (B) Competition experiments of radioactively labeled Col/P-1 olignonucleotide with nonlabeled Col/AP-1, APE-CRE, andPE-CREmut oligonucleotides. 2.5 fmol labeled Col/AP-1 oligonucle-tide was incubated with 1 mg nuclear extract from CHO cells andifferent molar excess of nonlabeled oligonucleotides as indicated.fter 30 min of incubation at room temperature, the incubationixture was separated on a 4% PAA gel. The autoradiograms are

hown.

862

el retardation experiments were performed using an-ibodies against c-Jun, ATF-2 and ATF-3. As shown inig. 4A, c-Jun as well as ATF-2 antibody caused aupershift with APE-CRE oligonucleotide, which wasot observed using ATF-3 antibody. A supershift waslso observed with the control oligonucleotide, Col/P-1, and c-Jun antibody but not with ATF-2 antibody

Fig. 4B). Also, no supershift could be detected with theREB/ATF oligonucleotide and c-Jun antibody (Fig.C). This indicates that the protein complex whichinds to the APE-CRE is majorly composed of c-Junnd ATF-2.Next we studied, whether oxidative treatment lead-

ng to induction of the APE promoter gives rise toncrease in the CRE-protein complex formation. In-eed, as shown in Fig. 5A, treatment of CHO cells with2O2 caused a clear increase in APE-CRE binding ac-

ivity. Furthermore, as revealed by Western blot anal-sis, treatment of cells with H2O2 under conditionsnducing both APE promoter activity (Fig. 1B) and

FIG. 4. Supershift experiments with APE-CRE, Col/AP-1, andREB/ATF oligonucleotides and c-Jun, ATF-2, and ATF-3 antibod-

es. (A) 5 mg of CHO nuclear extract from H2O2 treated cells (4 hreatment) was incubated with each of 5 mg antibody at 4°C for 1 h.hereafter radioactively labeled APE-CRE oligonucleotide wasdded and further incubated for 30 min at room temperature. Incu-ation mixtures were separated by electrophoresis on a 4% PAA gel.rrows indicates DNA-protein-antibody complexes (supershift, SS).

B) The same experiment was performed with radioactively labeledollagenase AP-1 (Col/AP-1) and (C) CREB/ATF oligonucleotide.

Adl5ticort(snb

tvoiOttapwrdticpsiAs

sponse against alkylation-induced genotoxicity awaitstvmtAilAggta

R

1

1

11

1

1

1

1

1

1

ACitlaW0

Vol. 261, No. 3, 1999 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

PE-CRE protein binding resulted in a time-ependent increase in the level of c-Jun protein. Theevel of ATF-2 was only very slightly enhanced (Fig.B). Increase in the amount of c-Jun occurred 4 h afterreatment with H2O2 and thus proceeds APE mRNAnduction, which was detected 4–6 h after exposure ofells to H2O2 (9). Thus, the time sequence of inductionf c-Jun after oxidative treatment appears clearly to beelated to the induction of APE mRNA. We should notehat the basal level of ATF-2 was already quite highFig. 5B). Therefore, ATF-2 may be present in non-aturating amounts and further enhancement wouldot be required, whereas the induction of c-Jun woulde limiting for APE activation.Overall, the data show that transcriptional activa-

ion of the APE gene requires CREB and is mediatedia activation of c-Jun/ATF-2. In mammalian cells,nly few DNA repair genes have been shown to benducible by genotoxic stress. Thus, the repair protein

6-methylguanine-DNA methyltransferase (MGMT) isranscriptionally activated by different kinds of geno-oxic agents inducing DNA strand breaks, includinglkylating drugs and ionizing radiation (15, 16). Theromoter of MGMT harbours two AP-1 binding siteshich have been shown to play a role in MGMT basal

egulation (14). The elements required for MGMT in-uction and the signal chains involved are still uniden-ified. For MGMT, protection against O6-alkylguaninenducing agents has clearly been shown to occur underonditions of MGMT induction (15). Another DNA re-air gene, DNA polymerase b (Pol-b), is induced byimple methylating agents, such as MNNG (17). Thisnduction is mediated by a functional CREB site (18).lthough Pol-b can be activated by alkylating agents, apecific protective effect elicited by this cellular re-

FIG. 5. Expression of APE-CRE binding protein, c-Jun, andTF-2 in CHO cells not treated (control) and treated with H2O2. (A)HO cells were treated with 0.3 mM H2O2 for the time periods

ndicated. Thereafter they were harvested and nuclear protein ex-ract was prepared. Extracts were incubated with the radioactivelyabeled APE-CRE oligonucleotide for 30 min at room temperaturend subjected to electrophoresis. The autoradiogram is shown. (B)estern blot analysis of c-Jun and ATF-2 protein in cells exposed to

.3 mM H2O2 for 4 and 8 h, respectively.

863

o be demonstrated. For APE, evidence has been pro-ided that induction of the gene by oxidative agentsakes cells more resistant to the same agent inducing

he gene (9, 10). Therefore, the response mediated byPE appears to be most specific. It would be interest-

ng to see whether there is an overlap in signalingeading to activation of mammalian DNA repair genes.lso, it will be seen whether batteries of DNA repairenes do exist in mammalian cells which are, analo-ous to bacteria, under coordinated control thus pro-ecting cells effectively from the genotoxic effects ofgents inducing them.

EFERENCES

1. Friedberg, E. C., Walker, G. C., and Siede, W. (1995) DNA Repairand Mutagenesis, ASM Press, Washington, DC.

2. Robson, C. N., Milne, A. M., Pappin, D. J. C., and Hickson, I. D.(1991) Nucleic Acids Res. 19, 1087–1092.

3. Winters, T. A., Henner, W. D., Russell, P. S., McCullough, A.,and Jorgensen, T. J. (1994) Nucl. Acid Res. 22, 186–1873.

4. Xanthoudakis, S., Miao, G., Wang, F., Pan, Y.-C. E., and Curran,T. (1992) EMBO J. 11, 3323–3335.

5. Okazaki, T., Chung, U.-I., Nishishita, T., Ebisu, S., Usuda, S.,Mishiro, S., Xanthoudakis, S., Igarashi, T., and Ogata, E. (1994)J. Biol. Chem. 269, 27855–27862.

6. Izumi, T., Henner, W. D., and Mitra, S. (1996) Biochemistry 35,14679–14683.

7. Xanthoudakis, S., Smeyne, R. J., Wallace, J. D., and Curran, T.(1996) Proc. Natl. Acad. Sci. USA 93, 8919–8923.

8. Asai, T., Kambe, F., Kikumori, T., and Seo, H. (1997) Biochem.Biophys. Res. Com. 236, 71–74.

9. Grosch, S., and Kaina, B. (1998) Cancer Res. 58, 4410–4416.0. Ramana, C. V., Boldogh, I., Izumi, T., and Mitra, S. (1998) Proc.

Natl. Acad. Sci. 95, 5061–5066.1. Tomicic, M., Eschbach, E., and Kaina, B. (1997) Mutation Res.

383, 155–165.2. Bradford, M. M. (1976) Anal. Biochem. 72, 248–254.3. Stein, B., Rahmsdorf, H. J., Steffen, A., Litfin, M., and Herrlich,

P. (1989) Mol. Cell. Biol. 9, 5169–5181.4. Boldogh, I., Ramana, C. V., Chen, Z., Biswas, T., Hazra, T. K.,

Grosch, S., Grombacher, T., Mitra, S., and Kaina, B. (1998)Cancer Res. 58, 3950–3956.

5. Fritz, G., Tano, K., Mitra, S., and Kaina, B. (1991) Mol. Cell.Biol. 11, 4660–4668.

6. Grombacher, T., Mitra, S., and Kaina, B. (1996) Carcinogenesis17, 2329–2336.

7. Fornace, J. A. J., Zmudzka, B., Hollander, M. C., and Wilson,S. H. (1989) Mol. Cell. Biol. 9, 851–853.

8. Kedar, P. S., Widen, S. G., Englander, E. W., Fornace, A. J., andWilson, S. H. (1991) Proc. Natl. Acad. Sci. USA 88, 3729–3732.

9. Harrison, L., Ascoine, A. G., Wilson III, D. M., and Demple, B.(1995) J. Biol. Chem. 270, 5556–5564.