Vol. 175, No. 3, 1991 March 29, 1991

BIOCHEMICAL

UV DAMAGE-SPECIFIC PIGMENTOSUM

AND BIOPHYSICAL RESEARCH COMMUNICATIONS Pages 1139-1143

DNA-BINDING PROTEIN COMPLEMENTATION

IN XERODERMA GROUP E

Hiroko Kataokaand YoshisadaFujiwara* Departmentof RadiationBiophysics,Kobe University Schoolof Medicine. Kusunoki-cho7-5 1, Chuo-ku. Kobe 650, Japan Received

February

14,

1991

The gel mobility shift assaymethodrevealed a specifically ultraviolet (UV) SUMMARY damagerecognizing,DNA-binding protein in nuclearextractsof normalhumancells. The resulted DNA/protein complexescausedthe two retardedmobility shifts. Four xerodermapigmentosum complementationgroup E (XPE) tibroblast strainsderived from unrelatedJapanesefamilieswere not deficient in sucha DNA damagerecognition/bindingprotein becauseof the normalcomplex formation and gel mobility shifts. although we confirmed the reported lack of the protein in the EuropeanXPE (XP2RO andXP3RO) cells. Thus.the absenceof this bindingprotein is not always commonlyobservedin all the XPE strains,andthe partially repair-deficientandintermediatelyUVhypersensitivephenotypeof XPE cellsaremuchsimilarwhetheror not they lack the protein. 0 1991 Academic

Press,

Inc.

Xeroderma pigmentosum (XP) is a multigenic DNA repair disorder consisting of eight complementationgroupsA through H andvariant. which presentsa high risk of sunlight-induced skin cancers] 11. The existenceof many humanXP androdent repair-deficientcomplementation groups[ 1, 21 indicatesthe requirementof multiple repair geneproducts for nucleotide excision repair in mammalianchromatin DNA. By the gel mobility shift assay,Chu andChang[3,4] and Hirshfeldet al. ]51have recently demonstratedthe presenceof a UV damage-specific DNA binding proteinin nuclearextractsof normalhumanandprimatecellsto resultin the mobility-retardedbands of UV-DNA/protein complexes,but the constitutive lack of this specificprotein andthe complexes in only XPE (XP2R0, XP3RO) cells, despitethe presencein the other XP groups[3]. Thus,this specificproteinhasbeendesignatedthe XPE factor actingasan importantUV damagerecognition protein for nucleotide excision repair [3.41. The normal humanDNA binding protein hasbeen suggestedto be an inactive, evolved humanhomologof yeastphotolyase[6]. On the other hand, the in V&Wexcisionrepairinterferenceassayhasshownthat normalhumancell-free extract inhibits photorepairof thymine dimer by E. coli photolyaseand excisionby E. coli (A)BC exinucleaseby competingfor the recognitionandbindingsites[7]. In this assay,the XPD (complementation group D) andXPE (XP2RO or CRL1259)extractsfail to inhibit the repair by E. coli (A)BC excinuclease [71.againsuggestingthat at leastthe XPE cells aredefectivein the UV damagerecognition/binding proteinof UV damageexinuclease. * To whomcorrespondence shouldbe addressed. Abbreviations: UV. ultraviolet light; XP, xerodermapigmentosum;XPE, XP complememation group E: Hepes,N-2-hydroxyethylpiperazine- N ‘-2-ethanesulfonicacid; DTT. dithiothreitol: EDTA, ethylenediaminetetmacetic acid: PMSF, phenylmethylsulfonylfluoride.

1139

0006-291X/91 $1.50 Copyright 0 1991 bv Academic Press, Inc. All rights of reproduction in any form reserved.

Vol.

175,

No.

3, 1991

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

The specific XPE protein defect has been currently confined to only the European XPE subjects (XP2RO and XP3RO second cousins). We assigned several XPE patients from unrelated Japanese families [g-lo]. Those XPE fibroblast strains had the same partially repair-deficient phenotype ]8lo] as did XP2RO and XP3RO XPE cells Ill]. Such a situation prompted us to study by the gel mobility shift assay whether or not these XPE fibroblast strains may have a common or similar protein defect. The result will show that the four Japanese XPE strains tested have the approximately normal constitutive level of a W damage-specific protein in their nuclear extracts and its lack may not be the common XPE defect. MATERIALS

AND METHODS

Human fibroblasts Normals were SV40-transformed GM637 (purchased from Human Genetic Mutant Cell Repository. Camden, NJ) and TIG-1 diploid normal embryonic lung fibroblast (courtesy of Dr. M. Ohashi, Tokyo Metropolitan Institute of Gerontology). The XPE skin fibroblasts were XP24KO ]8]. XP26KO ]8], XPSlTO [9] and XP89TO [lo] from unrelated Japanese patients. and XP2RO and XP3RO second cousins [ 1 I] from Rotterdam, the Netherlands (courtesy of Drs. J. H. J. Hoejimakers, Erasmus University). Cells were grown in Eagle’s minimal essential medium supplemented with 1 x non-essential amino acids (Flow Laboratories, Inc.) and 10% fetal bovine serum (Hyclone). as described [8]. Extracts Nuclear and cytoplasmic extracts of 2-5 X 107 preconfluent cells were prepared by slight modifications of the method of Dignam et al. [ 121. Cytosol was the supematant after centrifugation of cell homogenate. Isolated nuclei were suspendend at a density of 5 X 107 nuclei per 1 ml of extraction buffer [20 mM Hepes-KOH (pH 7.7 at 4°C). 1.5 mM MgQ, 0.2 mM EDTA, 0.5 mM DTT, 0.42 M NaCl, 25% (v/v) glycerol and 0.5 mM PMSF], extracted for 40 min at 4°C on ice with occasional agitations. and centrifuged at 15.0 ‘pm for 10 min at 4’C using a Tomy MR-150 microcentrifuge. The supematant was dialyzed against a buffer [20 mM Hepes-KOH (pH 7.7 at 4°C) 0.1 M KCl. 0.2 mM EDTA. 0.5 mM DTT. 20% (v/v) glycerol, 0.5 mM PMSF] for 1216 h at 4°C to prepare the final nuclear extract. Cytosol was also dialyzed as above. Protein concentration was determined by a Bio Rad protein assay kit. Aliquots of nuclear extracts and cytosols were quickly frozen in dry ice-ethanol. and stored at -85 or -20°C. Probe DNA The probe used was a 90 bp synthetic and cloned UT8-2 sequence consisting of three tandem repeats of oligo(dT), io *oligo(dA)7- io in the center and two random sequencesat both 5’ and 3’ ends, which was prepared by cutting the pUT8-2 plasmid (provided by Dr. T. Todo, Osaka University) with Eco Rl and Hind III. The gel-purified probe was end-labeled with ]as*P]dCTP (sp. act., 111 TBq/ mmol, NEN) or [a-3 P]dATP (sp. act., 111 TBq/mmol, NEN) by incubating with DNA polymerase I large fragment The radiolabeled probe and unlabeled competitor probe at 5 pg/ml were &diated with 4 W/m2 254 nm W (- 3 dime&probe). Detection of UV-DNA/protein complex The gel mobility shift assay wascarried out by slight modifications of the method of Carthew et al. ] 131. The UV-irradiated or unirradiated 13*Plprobe (0.5 or 1.0 rig/assay) was incubated with nuclear extract or cytosol (O-5 ug/assay) in the presence of 1 pg carrier DNA (Hinf I-digested pUCl9 DNA) in a 10 pl final volume of the reaction buffer (12 mM Hepes-KOH, pH 7.9.60 mM KCl, 5 mM MgC12. 12% glycerol, 4 mM Tris-HCl, pH 8.0.0.6 mM EDTA, 0.6 mM DTT) ] 131 for 30 min at 30°C in the dark. The above carrier DNA was more effective for absorbing non-specific binding roteins than poly(dIdC)poly(dldC). To ascertain the W damage specificity of binding, 4 kJ/m 8-irradiated excessprobe DNA was added to the reaction mixture as competitor. The reaction products were electrophoresed in 4% nondenaturing polyacrylamide gel in TBE (89 mM Tris-borate, pH 8.0,2 mM EDTA), fixed on Whatmann 3MM paper and exposed to Fuji RX X-ray fihn at -80°C. RESULTS UV damage-specific

DNA binding

protein

in normal human cells

First, Figure 1A displays the autoradiographed signals of gel mobility shit&J bands as a result of formation of the UV-DNA/protein complexes, when 4 kJ/niL W-irradiated 90 bp ~*PlprObe was incubated with GM637 nuclear extract. Under the optimal conditions, a -lOOO-fold excess of carrier DNA of Hinf I-cut pUCl9 DNA eliminated completely the non-specific protein biding to the 1140

Vol.

175,

No.

3, 1991

A

BIOCHEMICAL

Nuclear

extract

AND

B

0

1

UV

-+++++++-+

2.5

5

Probe

extracts

Cytosol II 1

GM637 I

5

5

5

5

5

COMMUNICATIONS

I

GM637 I (pg)

RESEARCH

Nuclear

Cytoso1 an

I

BIOPHYSICAL

XP3RO

-II (pg)O

5

4

4

4

Probe

DNA

4

4

4

4

4

DNA

uv+++++++-+ Competitor

DNA

--++-++-wJv)(+w

Lanes

1

2

3

4

5

6

7

6

9

10

Lanes

1

vJv)(+w

23456789

Figure 1. Gel mobili? shift assay for UV damage-specificDNA bindingprotein in GM637 normal cells. A. UV(4 kJ/m )-irradiated or u&radiated 90 bp [s*P]probe (0.5 @assay) was incubated with nuclear extract (0 to 5 pgIassay) or cytosol(5 pg/assay) in a lO-p.lmixture. UV-DNA/protein complexes are indicated by Bl and B2 bands of shifted mobility apart from free probe. Lane 6, pretreatmentof extract with pronase(1.5 mg/ml~plus-proteina K (1.5 mg/ml) for 15 min at 37°C; lane 7, preincubation without proteases; lane 8, absenceof MgCl* during incubation. Conditions of other lanes are indicated. B . Binding activity of XP3RO extracts (4 ug/assay) to 90 bp [32Plprohe(0.5 @assay) compared with GM637 extract. Lane 1, brank; lane 2. no competitor, lanes 3 and 6, addition of unlabeled.undamaged(-UV) competitor (20 rig/assay); lanes 4 and 7. addition of unlabeled, UV-irradiated (+UV) competitor (20 @assay); lane 8, unirradiated [32Plprobe: lane 9. XP3RO cytosol. NB indicates non-specific binding. unirradiated probe (Fig. 1A. lanes 1 and 9). With the UVdamaged probe, the two mobility shifted bands, fast-migrating B 1 and slow-migrating B2, appeared apart from the unbound free probe (Fig. 1A, lanes 3-5). B 1 was far more intense than B2. Such a binding activity of nuclear extract was UV damage-specific. since the unirradiated probe remained free (Fig. IA. lane 1) and a a-fold excess of UV-irradiated unlabeled competitor probe DNA abolished completely the binding (Fig. lB, lane 4 versus lane 3). This binding activity was lost by pretreatment of nuclear extract with prom&e-plus-proteinase K (Fig. 1A. lanes 6 and 7). but not by RNase A present during reaction (data not shown), as described [ 31. Thus. the B 1 and B2 bands represent the UV-DNA/protein complexes. The Bl and 82 intensities increased as a function of quantity of nuclear extract on the 0.5 rig/assay basis of 4 kJ/m%mdiated probe (Fig. 1A, lanes 2-5). The absence (Fig. lA, lane 8) or presence of MgCl, (Fig. 1A. lane 5) during the reaction did not affect the binding. The various conditions as such did not alter distances of the two mobility shifts. In addition. cytosol at the present maximum of 5 pg/assay had no detectable damage-specific binding activity (Fig. 1A. lane 10). DNA binding protein in XPE cells Figure 1B shows the results of XP3RO XFE cells in comparison with GM637 normal. Nuclear extract and cytosol of XP3RO cells formed no normal B 1 and B2 complexes. but a new NB band (Fig. lB, lanes 5-9), which migrated slightly faster than did the normal damage-specific Bl complex with the GM637 protein (Fig. 1B lanes 2 and 3). The NB intensity in XP3RO was not attenuated by challenge with the undamaged or W-irradiated competitor (Fig. lB, lanes 6 and 7). The NB complex with XP3RO cytosol was less intense (Fig. 1B. lane 9). We obtained the similar 1141

Vol.

175,

No.

BIOCHEMICAL

3, 1991

A

Nuclear

extracts

I (pg)

I

GM637 -II 1.2

uv

1.2

+

AND

B

BIOPHYSICAL

COMMUNICATIONS

RESEARCH

Nuclear

extracts I

I

XP24KO 1.2 1.2 2.44.8

Probe -+++++-

+

4.8 4.0

4.8

2

(pg)0

2

2

2

2

Probe

DNA

0

2

2

2

2

2

DNA

uv------++++++

.

,.I

,(

I. ‘.

,

, , ‘j

-82 -Bl

-Free

Lanes

I

2

3

4

5

6

7

8

9

Lanes

1

2

3

4

5 6

7 8

9 10 11 12

Figure 2. Comparison of the constitutive binding activity between normal and XPE nuclear extracts. A. GM637normal(lanesl-3) versus XP24KOXPE (lanes4-9). Amountsof extract usedwereindicated. 90 bp [3*P]probe(1 rig/assay):-, unirradiated:+, UV-irradiatedwith 4 kJ/m*. CompetitorDNA (20 rig/assay):-. none; +(-UV), additionof undamaged competitor; +(+UV), additionof UV-irradiatedcompetitor.B1 andB2 indicatesshiftedbands.Notethatthe XP24KOextractcaused morebreakdownof the probeandfainterbandintensitywith increasing amount(lanes4-6). B. Comparison betweenTIG-1 (normal)andfour XPE strains(XP24K0, XP26K0, XP81T0,XP89TO). Lanes1 and7, bra& lanes1-6,unirradiated 90bp[3*P]probe (1 rig/assay);lanes7-12, 4 kJ/m*UV-irradiated 90bp [QP]probe(1 rig/assay).Amountsof extracts usedareindicated.Bl andB2 arespecificUV-DNA/proteincomplexes. resultswith XP2RO XPE cells (datanot shown). Therefore,XP3RO andXP2RO cells aremissing the UV damage-specificbinding protein that forms the normalBl and B2 complexes,but possess an unusualnon-specific binding protein that forms the NB complex under the presentbinding conditions. On the contrary, Figure 2A demonstrates that the nuclearextract of XP24KO (XPE) cellsformed the two UV-DNA/protein complexesat the normal B1 andB2 positions(lanes4-7). asfound with the GM637 nuclearextract (lane 1). Again. the GM637 andXP24KO extracts hadno nonspecific binding activity to the unirradiated probe (Fig. 2A, lanes 3 and 9). The XP24KO binding protein/W-DNA complexeswere abolishedby the UV-irradiated competitor(Fig. 2A, lane 8). as were the GM637 protein/W-DNA complexes(Fig. 2A. lane 2). Figure 2B showsthat the four unrelatedJapaneseXPE fibroblast strains(XP24K0, XP26K0, XPSlTO andXP89TO) produced the normal Bl and B2 complexes. thus indicating the approximately normal level of UV-DNA binding activity in their nuclear extracts (lanes9-12). asdid TIG-1 normal cells (lane 8). The mobility shiftsand signalintensitieswerevery similarbetweenthe normalandtheseJapanese XPE strains(Fig. 2B, lanes8-12). As in normal. theseXPE nuclearextracts hadno nonspecificbinding activity (Fig. 2B. lanes2-6). Therefore the four XPE strainstested,unlike XP2RO and XP3RO. areapparentlynormalin the specificbindingactivity.

DISCUSSION The presentstudy hasindicatedthat the four unrelatedXPE strainsfrom Japanarenot defective in the constitutive level of the UV-DNA specific binding protein and the approximately normal binding activity (Fig. 2). However, the fibroblastsof the relatedXP3RO andXP2RO XPE patients 1142

Vol.

175,

No.

3, 1991

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

are missing it (Fig. lB), as reconciled with the previous report [3,4]. Thus. the defect in the UV damage-specific binding protein is not commonly observed in all XPE strains tested. Figure IB indicated a single NB band produced by an abnormal nonspecific DNA binding protein in both XP3RO nuclear extract and cytosol under the present binding conditions. We found the same NB complex in XP2RO (or CRL1259). It is as yet unknown whether or not this abnormal XP2RO and XP3RO activity described above may arise by a mutation in the UV damage-specific DNA binding protein. In view of the previous XP2RO protein defect, Chu and Chang 13.41, Patterson and Chu 161,Hirshfeld et al. IS] and Sibghat-Ullah and Sancar [7] suggested that the XPE-binding protein plays an important role in the initial damage recognition of human excision repair. Nonetheless, all the XPE strains used here manifest much the same cellular phenotypes of 40% reduced excision repair and a -2 fold UV hypersensitivity on the basis of mean lethal UV dose 18-l 11, whether or not the present specific DNA binding protein is absent (Figs. 1B and 2). Thus, it is very likely at a moment that the altered UV damage-sepcific DNA binding protein may not be the key mutation leading to the partial XPE repair defect. An alternative possibility may not be excluded that if the binding protein is the real XPE gene product, the Japanese XPE strains tested might have a mutation in a different functional domain of the XPE protein. We must await cloning of the binding protein gene and further protein investigation to solve the role of its DNA bindiig in DNA excision repair. This work was supported in part from Grants-in-Aid for Scientific Acknowledgments Research and Cancer Research from the Mini&v of Education. Science and Culture. JaDan. We are grateful to Drs. T. Todo, M. Ohashi and J. H.J. Hoejimakers for generous supplied of pUT8-2 plasmid, TIG-1 cells, and XP2RO and XP3RO cells, respectively. REFERENCES 1. 2. 1. 5: ;: 8. 9. 10. ::: 13.

Cleaver, J. E. and Kraemer. K. H. (1989) In Metabolic Basis of Inherired Disease, 6th edition, (Striver, C. R.. Beaudet, A. L.. Sly. W. S. and Valle, D. Eds.). pp 2949-2974. McGraw-Hill, New York. Thompson. L. H.. Shiomi, T.. Salazar, E. P. and Stewart, S. A. (1988) Somat. Cell Mol. Genet. 14.605612. Chu, G. and Chang, E. (1988) Science 242.564-567. Chu, G. and Chang, E. (1990) Proc. Natl. Acad. Sci. USA. 87.3324-3327 Hirshfeld. S.. levine. A. S.. Ozato. K. and Protic, M. (1990) Mol. Cell. Biol. 10,20412048. Patterson, M. amd Chu. G. (1989) Mol. Cell. Biol. 9,5105-5112. Sibghat-Ullah and Sancar, A. (1990) Biochemistry 29,5711-5718. Fujiwara, Y., Uehara, Y.. Ichihashi, M.. Yamamoto. Y. and Nishioka, K. (1985) Mutation Res. 145.55-61. Kondo, S., Fukuro, S., Mamada, A., Kawada, A., Satoh, Y. and Fujiwara. Y. (1988) J. Invest. Dermatol. 90, 152-157. Kondo. S., Mamada, A., Miyamoto, C., Keong, C-H., Satoh, Y. and Fujiwara, Y. (1989) Photodermatology 6,89-95. De Weerd-Kastelein, E. A., Keijzer, W. and Bootsma, D. (1974) Mutation Res. 22,87-9 1. Dignam, J. D., Lebovitz, R. M. and Roeder, R. G. (1983) Nucleic Acids Res. 11,14751489. Carthew, R. W.. Chodosh, L. A. and Sharp, P. A. (1985) Cell 43.439-448.

1143