Overexpression of the XPA Repair Gene Increases Resistance to Ultraviolet Radiation in Human Cells by Selective Repair of DNA Damage'

Home , XPA

CANCER RESEARCH55. 6152-6160. December 5. I995J Overexpression of the XPA Repair Gene Increases Resistance to Ultraviolet Radiation in Human Cells by Selective Repair of DNA Damage'

James E. Cleaver,2 Wayne C. Charles, Mindy L. McDowell, Walt J. Sadinski, and David L. Mitchell

Laboratory of Radiobiology and Environmental Health. University of Cal(fornia. San Francisco, California 94143-0750 Ii. E. C., W. C. C., M. L M., W. J. S.], and University of Texas M. D. Anderson Cancer Center, Science Park, Research Division, Smithville, Texas 78957 fD. L M.J

ABSTRACT formed with the use of wild-type XPA and a form ofXPA with a single missense mutation that produces an arginine-to-glycine change in Overexpression of XPA genes, both wild type and a missense mutant, codon 207 (R207G; Ref. 9). This mutation seems to reduce the which code for a damage-specific, DNA-binding protein, increased the capacity for the repair of cyclobutane dimers in inactive regions of survival of repair-deficient and -competent human cells to levels above that of normal cells that did not overexpress XPA. The first 3 h after cells chromatin (10, 11). We wanted to investigate variations of XPA were damaged were most critical to achieving this increased survival. The expression in relation to repair and survival and to determine whether dose at which 37% of the irt-adiated population survives could be restored the wild-type and mutant forms of XPA could, at high levels of to about one-half that of normal cells, with no detectable genome-wide expression, contribute to pynmidine dimer excision in the overall repair of pyrimidine dimers or (6â€”4) photoproducts, suggesting that genome, and whether a second gene hypothesized for pyrimidine intermediate levels of XPA gene expression can direct repair to restricted dimers was, in fact, likely to exist (12â€”14).After constructing a vector critical regions of the genome. Current views of repair implicate tran in which the XPA cDNA was inserted downstream from an inducible scriptionally active genes as a major component of such critical regions. promoter, we demonstrated that repair-deficient human cells can be Consistent with this interpretation, the repair of a transfected, actively made at least as resistant as normal cells by overexpression of this expressed luciferase gene was higher than that of genomic DNA at inter gene. Pyrimidine dimer excision from the whole genome, however, is mediate and higher levels of XPA expression. High levels of XPA expres. sion resulted in increased repair at early times after irradiation and very difficult to achieve even at high levels of expression, but repair extensive repair of (6â€”4)photoproducts but little, if any, pyrimidine during the first 3 h after cells are damaged is especially critical for dimer repair in the whole genome. At the highest level of expression, some improved survival. clonal cell lines acquired resistance to radiation that corresponded to a dose at which 37% of the irradiated population survives that was about 1.5 to 2 times that of normal cells. The XPA gene product, therefore, can MATERIALS AND METHODS influence levels of DNA repair and radiation sensitivity quantitatively by Cell Culture and UV Survival. The parentalcell lines used in this study contributing to selective repair at certain sites in the genome. were all SV4O-transformed:GM637 was from a normal subject; XP12RO was from a patient with XPAhomozygous for TGA stop codons in position 207 of INTRODUCTION the cDNA (R207 stop mutations); and XP129 was a revertant derived from XP12RO that was heterozygous for the stop codon and a missense mutation DNA nucleotide excision repair in human cells is a finely con (arginine to glycine, an R207G mutation) at the corresponding position in the trolled, complex system involving recognition proteins that bind to other allele (9, 10). Cells were grown in MEM with 4.5 glliter glucose damaged DNA, components of transcription factors that deliver repair supplemented with 10% fetal bovine serum. The medium also contained 200 gene products to actively transcribed genes and other genomic re units/mi penicillin and 200 @xg/mlstreptomycin G sulfate (Gibco BRL, Gaith gions, and several nucleuses, polymerases, and ligases; as many as 30 ersburg, MD). The UV sensitivity of these cells, and of other clonal cell lines proteins can be involved (1, 2). Associated with this complex control, derived from them, was determined by colony formation and by rapid well there is considerable variation in the precise level of DNA repair, assays (15). Cells were irradiated by using a low-pressure mercury lamp (254 which can be related to cancer susceptibility in cells from the normal nm, 1.3 joules/m2/sec) at doses as great as 26 joules/m2. human population (3, 4). Therefore, the question arises how the level Expression Vector System. A new expression system based on the bac terial lad repressor, the LacSwitch (Stratagene, La Jolla, CA), was used of excision repair in human and other cells is controlled. Can DNA because of the apparent tight on/off control provided by the lad repressor and repair and cell survival in irradiated human cells be improved beyond the bacterial inducer LPTG.3Notall of the anticipated properties of this system that seen in normal cells, and why do some species repair so poorly proved effective or ideal, in part because of the sensitivity of the UV repair compared with humans (5)? process to low levels of expression of XPA, but it has proved to be a useful To answer this question, we chose to work with the XPA gene, inducible system to use as a â€œfirst-generationâ€•construction. The cDNAs for the which was cloned by Tanaka et a!. (6). The gene is on chromosome wild-type XPA gene and the R207G missense mutation were isolated previ 9q34.l and codes for a 273-amino acid photoproduct-specific, DNA ously (9) by means of RT-PCR and cloning into the TA PCR vector (Invitro binding protein that has high binding specificity for (6â€”4)photoprod gen, San Diego, CA), and their integrity was confirmed by sequencing. This ucts but weak specificity for pyrimidine dimers (7). Mutations in this cDNA was transferred into the LacSwitch expression vector pOPRSVI to gene are associated with severe forms of the human disease xero make the XPA expression vector we called pOPRSVI-C12OF.The construct was sequenced by means of the dideoxy method to ensure that the inserted derma pigmentosum, which shows high levels of sunlight-induced sequence was intact. In the process of inserting the XPA cDNA, we generated skin cancer and neurological dysfunction (8). This study was per both forward and antisense orientations, both of which were used. Cional Selection. XP12RO and GM637 cultures were both transfected by Received 12/8/94; accepted lO/l3/95. using lipofectin with the vector p3'SS, which expresses the lad repressor, and The costs of publication of this article were defrayed in part by the payment of page selection carried out in hygromycin (200 @xg/ml).Thesecultures were trans charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. fected subsequently with the forward constructions of XPA in XP12RO or the I This work was supported by the Office of Health and Environmental Research, antisense construction of R207G in GM637, and colonies were isolated in United States Department of Energy. contract DE-ACO3-76-SFOIO12, and by NIH Grant G418 (400â€”600@xg/ml).Single-cellcolonies were selected and recloned to CA04484-34, National Institute of Environmental Health Sciences Grant ES05914, and American Cancer Society Grant S19-14. 2 To whom requests for reprints should be addressed, at Laboratory of Radiobiology 3 The abbreviations used are: IPTG, isopropyl-f3-D-thiogalactoside; RT, reverse tran and Environmental Health, Box 0750, University of California, San Francisco, CA scription; D37, dose at which 37% of the irradiated population survives; Mda. million 94143-0750. Phone: (415) 476-4563; Fax: (415) 476-0721. daltons. 6152

TabletheD37D37Cell I UVsensitivity of cell lines carrying wild-type or R207G XPA on then prepared. These mixtures were allowed to stand at ambient temperature LacSwitch vector, determined by rapid well assaycDNA for 15 mm before being added to 60-mm Petri dishes containing 1â€”2X106 cells/dish. Cultures were incubated overnight at 37Â°C (17â€”24 h) before being XPA (MEM) washed with PBS and harvested. Cells were scraped into PBS, washed once, (Joules/m2)GM637â€•wi-/'lineHost cell (genomic or cDNA) (Joules/m2)(IVFG) counted, and centrifuged. The pellets were lysed in 250 p.1cell lysis buffer 5.25.8XP129â€•R207G1R207(genomic (Promega) for 20 mm. For 20 p.1each lysate, 130 p.1 luciferase assay reagent 5.45.9XPI stop (genomic) (Promega) were added, and the mixture was placed immediately in a Mono .391.44XPA/AIXPI2RO2ROâ€•R2O7 stop (genomic) 1 0.50.5XPA/BIXPI2RO WT (cDNA) light 2010 Luminometer (Analytical Luminescence Laboratory, San Diego, 1.71.7XPAIB2XPI2RO Wi' (cDNA) CA) and measured for luminescence over 10s. Three aliquots were assayed for 8.210.3XPA/B3XPI2RO WT (cDNA) each lysate, and the luminescence per cell was calculated. 3.24.3XPA/CIXPI2RO WT (cDNA) 7.912.2XP199CXPI2RO WT (cDNA) 3.43.4XPI941EXPI2RO R207G (cDNA) RESULTS 2.74.6XPCI2OFXPI2RO R207G (cDNA) 5.110.3GM/XPAIGM637 R207G (cDNA) A series of single-cell clones that contained the sense constructs of 8.37.5GM/XPA2GM637 WT (cDNA) WT (cDNA) 6.312.1 XPA in a repair-competent (GM637) or repair-deficient (XP12RO) aUntransfectedcelllinescontainingonlyendogenousXPAgenomicsequences. background or the antisense construct in a repair-competent back I, WT, wild lype. ground(GM637)wasscreenedfor UV survivalwithor withoutIPTG induction by using a rapid well assay method (Table I; Ref. 15). To establish the general characteristics of this inducible system, we also generate clonal cell lines carrying both plasmids and expressing resistance to determined the kinetics of the induction and decline of UV resistance both drug selection markers. Rapid screening of these clones using the well survival assay ( 15) allowed us to choose a series of clonal lines that expressed by IPTG in one cell line that showed inducibility (XPC12OF; Fig. 1). the XPAgene and responded to IPTG induction. For one experiment, a clonal cell line expressing high levels of the lad repressor in XP12RO was selected first using antibodies to lad, and the XPA construct was cotransfected with additional p3'SS to generate higher levels of the lad repressor. The clonal 1-, I derivative (XPAICI), however, was not significantly different from the other clonal lines in its IPTG response. 10. Molecular Analysis. Total cellular RNA was isolated by means of the guanadinium isothiocyanate procedure (16). For discrimination of the trans fected from endogenous XPA sequences, allele-specific primers were used in S. which the 3 â€˜-terminal base of one primer was chosen to pair at the base that differs between plasmid and genomic DNA at codon 207 of the transfected 0-.. XPA cDNA, as described previously (9). 6. First-strand cDNA was generated in a 20-,xl RT reaction (GeneAmp RNA PCR kit; Perkin Elmer, Norwalk, CT). The reaction mixture contained 1 @xg 4. total RNA, 5 mMMgCI, solution, I X GeneAmp PCR buffer [500 mMKCI 0 and 100 mM Tris-HCI (pH 8.3)1, 4 mM dNTP mixture, 20 units RNase inhibitor, 50 units reverse transcriptase, and 2.5 ,xM random primers (hexam 2. A era). The reaction mixture was incubated at ambient temperature for 10 mm to anneal and extend primers and at 42Â°Cfor15 mm to extend primers further; the reaction was terminated by incubation at 99Â°Cfor 5 mm, followed by 5 mm 0 5 10 15 20 25 30 at 5Â°C. HOURS AFTER IPTG ADDITION OR REMOVAL To complete 20-pi RT reaction mixture was added MgCI2 to a final concentration of 2 mM, I X GeneAmp PCR buffer, 1.25 units AmpliTaq DNA polymerase, and 10 pmol each 3' and 5' oligomer in a total volume of 50 p.1. Mineral oil was added to prevent evaporation, and samples were denatured at 95Â°Cfor 2 mm, cycled 35 times for 1 mm each at 95Â°Cand 60Â°C,then incubated for 7 mm at 60Â°Cto extend primers fully and left to soak at 4Â°Cin a Perkin Elmer DNA thermal cycler. Although these PCR conditions were not optimized for precise quantification, the cycle time did not correspond to saturation of the reaction, and quantitative differences between cell lines were detected. PCR products were analyzed on 2% agarose, photographed, and analyzed by densitometry. DNA Repair Assays. Photoproductswereassayed at 0, 3, and 24 h after irradiation with UV light ( I3 joules/m2) by using specific antibodies for (6â€”4) photoproducts and pyrimidine dimers in a RIA (17, 18) or by digestion with dimer-specific T4 endonuclease V and analysis on alkaline sucrose gradients (19). Relative frequencies of excision repair sites per 200 MDa were deter mined by growing UV-irradiated cells in the chain-terminating inhibitors cytosmne arabinoside (2 X l0@ M) and hydroxyura (2 X lO@ M) for 6 h. Single-strand-break frequencies were quantified by alkaline sucrose gradient sedimentation (19). 0.1 1 Luciferase Assay. Plasmid pGL2, which expresses the luciferase gene IFI@G CONCENTRATION (mM) (Promega Corp., Madison, WI), was grown in competent bacteria, isolated, and purified in CsCl gradients. Plasmid preparations were irradiated in buffer Fig. 1. A, time course of induction and decline of UV resistance in XPCI2OF cells, ITns-HC1,(pH 7.4)1 at UV fluences as great as 2400 joules/m2. Transfection resulting from addition or removal of IPTG (I mat). â€¢,IPTG added at various times before irradiation. 0, IPTG removed for various intervals after 24 h induction and before mixtures consisting of pGL2 (10 p.g), 10â€”20 p.1 lipofectin (GIBCO BRL), and irradiation. B, UV resistance in XPCI2OF cells after induction for 24-h with various culture medium without serum or antibiotics to a final volume of 200 p.1were concentrations of IPTG. Bars, SE. 6153

Fig. 2. Relative colony formation in UV-irra diated cell lines expressing XPAcDNA. In some panels, survival curves are shown for the parental cell lines GM637 (V), XPI29 (A), and XP12RO (â€¢).Each panel shows survival for the desig nated cell line growth with (â€¢)or without (0) IPTG (1 msi). A, GM637 cells transfected with wild-type XPA cDNA (GM/XPA2). B, XPI2RO cells transfected with wild-type XPA cDNA (XPA/B3). C, XPI2RO cells transfected with R207G XPAcDNA (XP199C). D, XPI2RO cells 10@ transfected with R207G XPAcDNA (XPC12OF).

ULTRAVIOLET DOSE (J/m@)

Table 2 UV sensitivity determined by colonyformationdifferingD37 for c lines with Similar levels of UV resistance with or without induction were expression of the XPA geneeli achieved with both the wild-type and mutant forms of XPA (Table 1), (MEM) and no differences in response between the two forms of XPA were (Joules/m2)GM637Cell line (Joules/m2)D37 obvious. 10.4 GM637 (antisense XPA) 5.1 5.1 XP129 8.5 8.5 XPI2RO 0.25 0.25 XPAIB3 12.3 14.6 XPA/CI 11.1 13.5 XPI99C 1.4 2.5 XP194IE 2.65 4.6 XPCI2OF 7.5 15.0 GMXPA1 1I.0 NDa 13.0aGMXPA2 10.710.4 ND, not determined.

Several features of these initial results were distinctive. All cell lines derived from a repair-deficient host cell exhibited an uninduced level of UV resistance that was higher than that of the parent XP12RO and an induced level that was increased by only a factor of 2 at best. Very few clonal cell lines showed high UV sensitivity, suggesting that even low levels of expression of XPA confer significant UV resistance. Induction was rapid, and full UV resistance was achieved when IPTG was added immediately after irradiation without a detectable lag time 0 2 4 6 S 10 12 14 16 15 20 (Fig. 1). Maximal induction occurred at 1 mrviIPTG. The decline in UV resistance after removal of IPTG took at least 24 h, which is ULTRAVIOLET DOSE (JIma) consistent with earlier work on the persistence of UV repair after Fig. 3. Relative colony formation in UV-irradiated normal cells (GM637, V) and inhibition of protein synthesis (20) and indicates that the missense normal cells carrying antisense XPA cDNA (0, without IPTG; â€¢,withI mMIPTG, for mutation did not confer extreme instability to the repair process. 24 h before irradiation). For other symbols. see legend to Fig. 2. 6154

OVEREXPRESSIONOF A REPAIR GENE AND UV RESISTANCE

Several cloned cell lines carrying either of the XPA cDNA con Table 3 Plating efficiencyof unirradiated cell clones expressing XPA cDNA. with or structs were then analyzed for UV sensitivity by colony formation without JPTGinductionPlating (Fig. 2). The rapid well assay, although satisfactory for monitoring efficiency (%)Ratio Uv sensitivity,isinfluencedbydifferencesingrowthrates;therefore, induced/Cell colony formation is more reliable for quantification (15). Survival of +IPTGuninducedGM637I13.4lineExperimentâ€”LPTG repair-deficient or -competent host cells transfected with wild-type 12.20.91XPI29I 3.1 XPA cDNA was close to normal when uninduced but could be 0.49XPA/C3I9.9 22.8 7.8 3.81.11 induced to higher levels by IPTG (Fig. 2). The expression of wild-type 7.350.74XPA/CI21.4 XPA in a normal background (Fig. 2A) was less effective than that in 2.41.71GMXPA II 9. 1 a repair-deficient background (Fig. 2B). One cell line containing the 0.56GMXPA2I 220.6 34.7 19.50.44 mutant construct (XP199C) had UV resistance that was intermediate 24.5 1.18XPCI2OFI 246.9 35.1 41.30.52 between that of XP12RO and normal (GM637) but had no response to 18.6 IVfG induction (Fig. 2C). Another line containing the mutant con 2 51.4 17.0 0.33 struct (XPC12OF) showed near-normal resistance to UV and a strong 323.1 11.8 11.90.81 1.01 response to induction by IPTG that resulted in survival that was higher than that in normal cells (Fig. 2D and Table 2). At UV fluences as

GMXPA2 GMXPA1 XPAC1 GM637 XP12RO M +

â€” @- â€”- â€” â€”

Fig. 4. PCR products synthesized by using al Ide-specific primers for the transfected XPA cDNA. M. marker lane; +, growth in IPTG for 24 h; â€”,growth without IPTG. Top, cell lines trans fected with wild-type XPAcDNA vectors. Bottom, cell lines transfected with R207G XPA cDNA vectors. M XP129 XPCI2OF XPI99C XPI2RO IPTG + - + - + - +

E@ 256â€” ?â€¢

6155

Fig. 5. Pyrimidine dimers and (6â€”4)photo products remaining in total genomic DNA as a function of time after irradiation with I3 joules/m2 UV light. A, pyrimidine dimers in normal GM637 cells (V) and GM637 cells car rying wild-type XPA cDNA (GM/XPA2) with out IVFG (0, MEM) or with IVFG (â€¢).B, pyrimidine dimers in XP12RO cells carrying B wild-type XPA cDNA: XPA/CI without IPTG (0), XPA/CI with IPTG (I), XPA/B3 without IPTG (0). and XPA/B3 with IPTG (â€¢). C, 10. photoproducts in XPI99C (XPI2RO carrying the R207G XPA cDNA). 0 and U, pyrimidine 0 5 10 15 20 25 dimers in cells grown without or with IPTG, respectively. 0 and â€¢,(6â€”4)photoproductsin cells grown without or with IPTG, respectively. D, photoproducts in XPC12OF(XPI2RO carry aM ing the R207G XPA cDNA). U and â€¢,pyrimi dine dimers in XPCI2OF cells grown without or with IPTG, respectively. 0 and â€¢,(6â€”4)pho toproducts in XPCI2OF cells grown without or with IPTG, respectively. V. pyrimidine dimers in GM637 cells. A, (6â€”4)photoproducts in XP129 cells. All data points are the average of duplicate measurements. Data for pyrimidine dimers in GM637 and for (6â€”4)photoproducts in XP129, shown for comparison, are taken from previous reports (10. 21).

HOURS AFrER IRRADIATION great as about S joules/m2, XPC12OF showed 100% survival. High on the cells by measuring the plating efficiency of the unirradiated survival levels were also achieved with the wild-type gene (Tables 1 cells (Table 3). After transfection, we detected a decrease in plating and 2 and Fig. 2B). efficiency in several of the cell lines. But this was also evident in some An antisense sequence in GM637 cells reduced the D37 to about experiments with cells that carried no transfected genes; therefore, it one-half that of normal cells, but there was no further decrease due to may not be related specifically to XPA expression (Table 3). IPTG induction (Fig. 3 and Table 2). Although this antisense construct The variations in survival correlated, approximately, with expres seemed to be partially effective at increasing the sensitivity of GM637 sion of the XPA gene as detected by RT-PCR (Fig. 4). With primers cells to UV light, we were not able to detect a concomitant change in specific for the transfected XPA gene, no amplification signal was excision repair [dimer excision, (6â€”4)photoproduct excision, or total detected in XP1 2RO cells, low levels were detected in XP199C cells frequency of repair patchesi; therefore, we discontinued work with independent of IPTG, and higher inducible levels were detected in cell lines carrying full-length cDNA antisense constructs. XPC12OF cells (Fig. 4). The expression of XPA in XPC12OF cells We determined whether overexpression of XPA had a toxic effect after IPTG induction seemed to exceed that in the haploid XP129 cell

Table 4andCell Excision repair sites (single-strand breaks/200 MDa) in cells irradiated with 13 joules/m2 UV light grown for 6 h withcvtosine arabinoside (2 x i0@ M) hydroxyurea (2 x i0@ M)and

frequency'@Without

IPTGGM637'@WTlineHost cellXPA (genomic or cDNA)Break IPTGWith 0.01XPI29Â°R207G/R207 (genomic)I .04 Â±0.490.85 Â± 0.16xpl2ROaR207 stop (genomic)0.89 Â±0.090.87 Â± 0.26XPA/A1XP12ROWT stop (genomic)0.35 Â±0.210.39 Â± (cDNA)0.911.16XPA/BlXP12ROWT (cDNA)0.760.75XPA/B2XPI2ROWT 0.22XPA/B3XPI2ROWT (cDNA)0.69 Â±0.221 .34 Â± (cDNA)1.231.40XPA/C1XPI2ROWT (cDNA)3.223.49XPI99CXP12ROR207G 0.10XPI94IEXPI2ROR207G (cDNA)0.21 Â±0.060.22 Â± 0.17XPCI2OFXP12ROR207G (cDNA)0.46 Â±0.120.49 Â± 0.34GM/XPAIGM637WT (cDNA)1.50 Â±0.163.37 Â± 0.43GM/XPA2GM637Wi' (cDNA)1.01 Â±0.061.28 Â± (cDNA)0.94 Â±0.020.97 Â±0.14

a Untransfected cell lines containing only endogenous XPA genomic sequences. b Mean and range for duplicate measurements. No errors given for single determinations. 6156

essentially the same repair as did normal untransfected cells (GM637), except for XPA/B3 and XPA/C1 cells, which exceeded normal levels (Table 4). XP199C cells showed no accumulation ofbreaks above that of the repair-deficient XP12RO cells after UV irradiation (Table 4), aM consistent with the absence of (6â€”4)photoproducts and pyrimidine dimer excision (Fig. SC). XPC12OF cells showed a break frequency LI about twice that of normal cells without IPTG induction and a large 0 increase in breaks with induction. This indicates that the rate of total z genomic repair could increase in cells carrying either wild-type (XPA/B3 and XPA/C1) or mutant (XPC12OF) genes at the earliest times after irradiation, and that this increase is not resolved easily by 0 the excision data and is not associated with increased pyrimidine I- dimer repair from the whole genome (Fig. 5). A closer examination of the amount of total genomic repair carried out by XPCI2OF cells during the first 6 h after irradiation indicated that IPTG stimulated 1 2 3 4 S repair over the whole time period (Fig. 6). In the absence of IPTG HOURS AFTER IRRADIATION induction, repair seemed to slow at about 2 h after irradiation. Fig. 6. Total genomic repair as measured by the number of single-strand breaks per 200 An irradiated vector carrying the luciferase gene was used to assay MDa in XPCI2OF cells grown in cytosine arabinoside (2 X l0@ M)and hydroxyurea the capacity of cells to repair UV damage in the actively expressed (2 X lO@ M) for various times after irradiation with 13 joules/m2 UV light. 0, grown in episomal luciferase gene (Fig. 7). When compared with the normal MEM; â€¢,grown in IPTG for 24 h before irradiation and continuously afterward. Bars, SE. untransfected cell line (GM637), repair of the luciferase gene was much more effective in cells expressing the XPA cDNA (Fig. 7). Even line (Fig. 4), consistent with the high resistance to UV light observed XP199C cells, which were more UV sensitive than GM637 cells (Fig. in this cell line (Fig. 2D). In parallel experiments using internal 2C), repaired the luciferase gene to a similar extent, as did GM637 standards and the mimic procedure (Clontech), the basal expression cells (Fig. 7C). These results imply that the overexpressed XPA level of XPA in normal cells corresponded to only two to four mRNA cDNA, either wild type or missense, can be more effective in repair molecules/cell.4 ing an actively expressed episomal gene than the whole genome. To define the events associated with DNA damage and repair that The survival and repair experiments described in Figs. 2â€”7were could be correlated with survival after irradiation, we determined the carried out after 24 h induction with IPTG, followed by continuous pyrimidine dimer and (6â€”4)photoproduct frequencies in cells carry growth in IPTG. We noted, however, that the IPTG-inducible vector ing the wild-type XPA cDNA (XPA/B3, XPA/C 1, and GM/XPA2) or was rapidly responsive, and that full expression of UV resistance in the R207G missense XPA cDNA (XPI99C and XPC12OF) with or XPC12OF cells was achieved when IPTG was added immediately without IPTG induction (Fig. 5) and compared them with previous after irradiation (Fig. 1). We used this property to define the time after results obtained with GM637 and XP129 cells (18, 21). In XPAIB3, irradiation during which IPTG induction of the XPA gene could XPA/C1, and XPI99C cells, despite the clear increases in UV resist increase cell survival. By leaving variable times between irradiation ance and XPA expression, there was no detectable excision (Fig. 5, B and induction, we found that the ability of increased XPA expression and C). In XPC12OF cells, expression of the gene without IPTG to enhance survival was lost rapidly, and that induction by IPTG was induction permitted (6â€”4)photoproduct excision but not pyrimidine ineffective after 3 h (Fig. 8). Critical events that can contribute to cell dimer excision from total genomic DNA (Fig. SD); after IPTG induc survival, therefore, must occur in the first few hours after UV expo tion, there was a slight acceleration of (6â€”4)photoproduct excision sure. Once this time is past, cells are irreversibly committed, and the and limited excision of pyrimidine dimers at 24 h (Fig. SD). At 3 h, XPA gene product cannot alter their fate. without induction, â€”43%of the (6â€”4)photoproducts in XPC12OF cells remained, and after induction, â€”35%remained, compared with DISCUSSION >50% remaining in XP129 cells (Fig. SD; Ref. 10). In cells carrying the wild-type gene against a repair-competent or -deficient host, Our results show that manipulation of the XPA gene, which is likely pyrimidine dimer excision did not seem to increase (Fig. 5, A and B) to code for one of the gene products that initiate a cascade of events despite high expression levels; (6â€”4) photoproduct measurements leading to excision and repair replication (1, 2, 24, 25), can produce were not made for these cells. quantifiable changes in repair and survival. This was not necessarily The relative number of repair patches was determined from the expected, because excision repair requires the activity of such an frequency of strand breaks at repair sites introduced by the inhibitor intricatecombinationof gene productsthat a variationof any one of cytosine arabinoside (Table 4). This method inhibits gap sealing them might not influence the whole process (2, 8, 24, 25). We found, during excision repair, and most breaks are accumulated in the first 2 however, that XPA seems to act as a modulator of the overall process: h after irradiation, even when the inhibitor is present for longer times low levels of expression gave rise to intermediate levels of survival (22). The method, therefore, highlights early excision of (6â€”4)pho but no detectable overall repair; and higher levels of expression gave toproducts ( 10). Because the transformed cells used in these studies rise to high levels of survival and repair of (6â€”4)photoproducts but have large endogenous nucleotide pools, competition between cyto rarely of dimers in the overall genome. Overexpression of XPA by sine arabinoside and endogenous deoxycytidine triphosphate reduces induction with IP'FG even led to UV resistance in several clonal cell the absolute number of repair sites detectable by this method com lines that was significantly higher than that in normal cells (Tables 1 pared with the numbers observed in primary fibroblasts, but the and 2 and Fig. 2). This indicates that the natural UV resistance of method is satisfactory for comparative studies (23). Cells carrying the human cells is not as great as it could be and that it is limited to some wild-type cDNA on a normal or repair-deficient background had extent by XPA expression. The increase in survival of XPC12OF cells, for example, was achieved by an expression level of the XPA gene that 4 S. K. Layher and J. E. Cleaver. unpublished observations. was about twice that in cells with one chromosomal gene (Fig. 4). 6157

Fig. 7. Relative luciferase activity in cells transfected with the UV-damaged lucifera.se vector pGL2 (Promega Corp.). A, GM637 (V) and XP12RO (â€¢)cells.B, XPI2RO cells carry ing wild-type XPA cDNA (XPA/C 1: without IPTG, 0; with IVFG, â€¢).The single low value at 600 joules/m2 for cells in IPTG is considered an outlier and is not included in the lines con necting data points. C, XPI2RO cells carrying R207G XPAcDNA (XPI99C: without IPTG, 0; with IPTG. â€¢). D, XP12RO cells carrying R207G XPA cDNA (XPCI2OF: without IPTG, 0; with IPTG, â€¢).InC and D, GM637 (V) and XPI2RO (U) curves, and in B the GM637 (V) curve, are shown for comparison (dashed lines). Bars, SE.

ULTRAVIOLET DOSE TO PLASMID (JIma)

Therefore, the changes needed to produce significant effects on cell seem to be toxic to cells to an extent separable from effects of IPTG survival are quite modest, and the overall colony formation seems a alone (Table 3). more sensitive measure of changes resulting from IPTG induction Several general comments should be made about the LacSwitch than the other assays used. High expression of the XPA gene did not system itself. These will be important with reference to our use of XPA expression as a basis for comparison between cell survival, repair of the overall genome and the episomal luciferase gene, and the functional activity of the R207G missense mutation. The LacSwitch induction system has provided us with useful in formation, particularly because induction causes specific increases in the expression of XPA cDNA that then can be related to defined biological end points. Conversely, the weakness of the LacSwitch system seems to be its limited range of induction and its leakiness as used here. Considerable UV resistance was observed even when the gene was supposedly switched off by the lad repressor. This resist ance is unlikely to be due to the lad protein itself, because it has no natural binding sites in mammalian DNA and, therefore, should be neutral. There was wide variation in the degree of correction of UV sensitivity in clonal cell lines, and the induction of UV resistance was rarely more than a 2-fold increase in D37. This range was reflected in the signal intensities of specific PCR products (Fig. 4), although detailed quantification of the system is reserved for a later report. When we transfected XPA cDNA by electroporation, we observed HOURS AFTER IRRADIATION very low degrees of correction of UV sensitivity and again only a Fig. 8. Recoverable fraction of cell survival in XPCI2OF obtained by adding IPTG (1 mM) at various times after irradiation. The recoverable fraction is defined as (D,â€”D2)/ 2-fold induction (data not shown). With this method of transfection, (D3â€”D2),whereD is the 37% surviving dose, D1 is the value obtained by adding LP'l'G however, growth in G418 to maintain the XPA cDNA, which is linked at designated times, D2 is the value without IPTG induction, and D3 is the value obtained with full induction by adding IPTG at the time of irradiation or up to 24 h earlier. Bars, covalently to the neomycin resistance gene in the vector, resulted in a SE. gradual increase in UV resistance and DNA repair with passage in 6158

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 1995 American Association for Cancer Research. OVEREXPRESSIONOF A REPAIR GENE AND UV RESISTANCE culture. To avoid such an unstable system, we confined our detailed we demonstrated that the ability of IPTG to enhance survival was investigation to stable clones made by lipofectin. confined to the first few hours after irradiation (Fig. 8). During this We chose to monitor XPA cDNA expression by PCR because of its time, (6â€”4)photoproducts were excised, and the measurements of ability to distinguish the transfected from endogenous transcripts. overall repair indicated a large increase in the total number of repair Although this method should give values that correlate with amounts patches over that in untransfected cells (from 0.5 to 3.00/200 MDa) at of protein, the correlation is not necessarily linear, nor can it account these early times (Fig. 6 and Table 4). The increased survival on for posttranslational modification. We used peptide antibodies to XPA induction by IPTG could be correlated with increased repair within protein previously, but cross-reaction with other proteins made this approximately the first 3 h after irradiation (Fig. 6). Although pyrim method unsuitable for quantification (26). idine dimer excision did occur at 24 h after irradiation, it was too late The general observation that results from this study is that UV to affect cell survival significantly. It is possible that these early survival is related closely to levels ofXPA expression. Overexpression events also could involve excision of dimers from actively transcribed of both wild-type and mutant cDNAs can raise survival above that in genes. But even for rapidly repaired genes such as p53, only about normal cells, especially for cell lines such as XPA/C1, which is 20% of the damage is repaired by 3 h (29), compared with >60% of constitutively high, and XPCI2OF, which induces to high UV resist (6â€”4)photoproducts from the whole genome (Fig. SD). (6â€”4)pho ance (Fig. 2, B and D). Because of the leakiness of the inducible toproduct excision, therefore, may be a major contributor to cell vector and our observation that normal sensitivity is associated with survival under these conditions. It is of interest that the early critical only one or two mRNA transcripts per cell,5 it was more difficult to period identified by these experiments (Fig. 8) correlates not only with study low levels of expression than overexpression, which is why we the periods of (6â€”4)photoproduct excision (Figs. S and 6) but also concentrated on high expression in this study. with the time needed for recovery of transcription and replication (30). No obvious distinction could be observed between clones derived The properties of most of our transfected clones resemble those of from wild-type and missense mutant cDNA plasmids. All clones mouse and hamster cells. They are UV resistant, excise (6â€”4)pho showed correction of UV sensitivity; all clones also showed increased toproducts but not dimers from the overall genome, and can repair (6â€”4)photoproduct repair, as judged either by direct measurement or actively transcribed genes (10, 11, 31, 32). A fundamental difference indirectly by the number of repair patches inhibitable by cytosine in the way mouse and normal human cells repair UV damage was arabinoside. Excision of pynmidine dimers, however, was very dif recognized very early in DNA repair studies (5). A possible explana tion for this difference, therefore, could lie in differences in the ficult to achieve. Only one cell line, XPCI2OF, which carried a expression level and affinity constants of the human and mouse XPA missense cDNA, achieved limited dimer excision, and this was asso gene products. ciated with enhanced repair by other assays. Our results are similar to The most perplexing aspect of these studies is the inability to those obtained with the transfer of a whole chromosome 9, which achieve normal levels of pyrimidine dimer excision in the whole resulted in repair of (6â€”4)photoproducts only (27). genome, even in cells expressing high levels of XPA cDNA. The The intermediate level ofXPA expression and survival in XP199C, levels of mRNA detected by RT-PCR in some clones were clearly for example, was not associated with detectable excision of either greater than those in normal cells for both wild-type and missense dimer or (6â€”4)photoproducts or increased numbers of repair patches XPA cDNA (Fig. 4). When expressed from genomic sequences, re in the overall genome. And most of the other cell lines failed to duced mRNA expression was associated with reduced levels of XPA remove dimers even at high levels of survival. Most likely, therefore, protein (26). Overexpression from a transfected cDNA, however, may the increased survival in these cells was due to repair of pynmidine present distinct problems of transport, translation, and assembly into dimers in restricted, critical regions of the genome or of (6â€”4) macromolecular complexes. Therefore, we need to evaluate ways to photoproducts in restricted regions and in the overall genome at quantify the actual level of the XPA gene product under these condi higher levels of XPA expression. The haploid expression level of the tions. One previous suggestion was that there is a second gene that is missense mutant XPA gene in XP129 cells is associated with repair of specific for pyrimidine dimers (12â€”14),but this is difficult to sustain, dimers in active genes ( 11) and (6â€”4)photoproducts overall (10). because the initial evidence for transfection of such a gene was Consequently, we would expect that the intermediate UV resistance of confounded by reversion of the endogenous XPA gene (33). There a cell line such as XP199C is due to low levels of the XPA gene fore, the mechanism by which excision repair acts on pyrimidine product that direct repair of dimers and (6â€”4)photoproducts exclu dimers in untranscribed regions of the genome remains unexplained, sively to actively transcribed genes, and that other clones that repair because even expression of a transfected XPA gene at high levels (6â€”4) photoproducts also can repair dimers in active genes. Our could not restore dimer excision to normal levels. results with the actively expressed luciferase gene support this inter The enhanced survival we observed on IPTG induction seemed to pretation, because luciferase activity was repaired to high levels in all be associated with an acceleration of events within the first few hours clonal lines, even those, such as XP199C, that were repair deficient after irradiation (Figs. 6 and 8). This observation focuses attention on and UV sensitive (Fig. 7). The results obtained with the luciferase the possibility that the eventual outcome of DNA damage in terms of gene, in an episomal configuration, to a first approximation may be cell death or survival depends on competition between potentially consistent with the hypothesis that XPA directs repair preferentially to lethal and DNA repair processes within the first few hours after cells actively expressed genes. XPA, therefore, may exercise an indirect are damaged. Small differences in repair within this early time period role in stimulating the coupling of repair with transcription that is that are associated with increased expression of XPA, therefore, can more clearly associated with the components of transcription factor have disproportionate effects on cell survival. We are concentrating IIH (2, 25), and recent work has shown that it actually recruits our attention on this time frame in our further studies. The commit transcription factor IIH to sites of damage (28). ment of cells to processes that are potentially lethal, e.g., apoptosis, XPC12OF, especially after induction with IPTG, showed high sur seems to become irreversible within a few hours of DNA damage. vival, excision of (6â€”4)photoproducts, and a small amount of exci Improvement and exploitation of the inducible cellular system de sion of pyrimidine dimers from the overall genome. In this cell line, scnbed here should help elucidate molecular details of the balance between repair and toxicity. Our results also indicate that comparable 5 5. K. Layher and J. E. Cleaver. unpublished observation. overexpression of genes such as XPA in vivo could, if achieved, 6159

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 1995 American Association for Cancer Research. OVEREXPRESSIONOF A REPAIR GENE AND UV RESISTANCE contribute to resistance to carcinogenic effects of UVB from sunlight 16. Chirgwin, J. M., Przybyla, A. E., MacDonald, R. J., and Rutter, W. J. Isolation of and other carcinogens recognized by the nucleotide excision repair biologically active ribonucleic acid from sources enriched in ribonuclease. Biochem istry, 18: 5294â€”5299, 1979. system. This could have practical importance in protecting humans 17. Mitchell, D. L., Haipek, C. A., and Clarkson, J. M. (6â€”4)Photoproducts are removed from adverse exposure to carcinogens of various kinds. from the DNA of UV-irradiated mammalian cells more efficiently than cyclobutane pynmidine dimers. Mutat. Res., 143: 109â€”112, 1985. 18. Mitchell, D. L., Haipek, C. A., and Clarkson, J. M. Further characterization of a REFERENCES polyclonal antiserum for DNA photoproducts: the use of different labelled antigens to I. Cleaver, J. E. It was a very good year for DNA repair. Ce'l, 76: 1â€”4,1994. control its specificity. Mutat. Res., 146: 129â€”133,1985. 2. Aboussekhra, A., Biggerstaff, M., Shivji, M. K. K., Vilpo, J. A., Moncollin, V., 19. Cleaver, J. E. Methods for studying repair of DNA damaged by physical and chemical Podust, V. N., Protic, M., Hubscher, U., Egly, J-M., and Wood, R. D. Mammalian carcinogens. Methods Cancer Res., 11: 123â€”165,1975. DNA nucleotide excision repair reconstituted with purified protein components. Cell, 20. Gautschi, J. R., Young, B. R., and Cleaver, J. E. Repair of damaged DNA in the 80: 859â€”868,1995. absence of protein synthesis in mammalian cells. Exp. Cell Res., 76: 87â€”94,1973. 3. Wei, Q., Matanoski. 0. M., Farmer, E. R., Hedayati, M. A., and Grossman, L. DNA 21. Cleaver, J. E. DNA damage and repair in normal, xeroderma pigmentosum and XP repair and aging in basal cell carcinoma: a molecular epidemiology study. Proc. Nat!. revertant cells analyzed by gel electrophoresis: excision of cyclobutane dimers from Acad. Sci. USA, 90: 1614â€”1618, 1993. the whole genome is not necessary for cell survival. Carcinogenesis (Lond.), 10: 4. Wei, Q., Matanoski, G. M., Farmer, E. R., Hedayati, M. A., and Grossman, L. DNA 1691â€”1696,1989. repair related to multiple skin cancers and drug use. Cancer Res., 54: 437â€”440,1994. 22. Hashem, N., Bootsma, D., Keijzer, W., Greene, A., Coriell, L., Thomas, G., and 5. Painter, R. B., and Cleaver, J. E. Repair replication, unscheduled DNA synthesis, and Cleaver, J. E. Clinical characteristics, DNA repair, and complementation groups in the repair of mammalian DNA. Radiat. Res., 37: 451â€”466,1969. xeroderma pigmentosum patients from Egypt. Cancer Res., 40: 13â€”18,1980. 6. Tanaka, K., Miura, N., Satokata, I., Miyamoto, I., Yoshida, M. C., Satoh, Y., Kondo, 23. Cleaver, J. E., Charles, W. C., and Kong, S. H. Efficiency of repair of pyrimidine S., Yasui, A., Okayama, H., and Okada, Y. Analysis of a human DNA excision repair dimers and psoralen monoadducts in normal and xeroderma pigmentosum human gene involved in group A xeroderma pigmentosum and containing a zinc-finger cells. Photochem. Photobiol., 40: 621â€”629,1984. domain. Nature (Land.), 348: 73â€”76,1990. 24. Bootsma, D., and Hoeijmakers, J. H. The molecular basis of nucleotide excision 7. Jones, C. J., and Wood, R. D. Preferential binding of the xeroderma pigmentosum group repair syndromes. Mutat. Res., 307: 15-23, 1994. A complementingprotein to damaged DNA. Biochemistry,32: 12096â€”12104,1993. 25. Drapkin, R., Sancar, A., and Reinberg, D. Where transcription meets repair. Cell. 77: 8. Cleaver, J. E., and Kraemer, K. H. Xeroderma pigmentosum. In: C. R. Scriver, A. L. 9â€”12,1994. Beaudet, W. S. Sly. and D. Valle (eds.), The Metabolic Bases of Inherited Disease, 26. Cleaver, J. E., McDowell, M., Jones, C., Wood, R., and Karentz, D. Mutation and Ed. 7, Vol. 3, pp. 4393â€”4419. New York: McGraw-Hill, Inc., 1995. expression of the XPA gene in revertants and hybrids of xeroderma pigmentosum cell 9. McDowell, M. L., Nguyen, T., and Cleaver, J. E. A single-site mutation in the XPAC line. Somatic Cell Mol. Genet., 20: 327â€”337,1994. gene alters photoproduct recognition. Mutagenesis, 8: 155â€”161,1993. 27. Ishizaki, K., Matsunaga, T., Kato, M., Nikaido, 0., and Ikenaga, M. Repair of 10. Cleaver, J. E., Cortes, F., Lutze, L. H., Morgan, W. F., Player, A. N., and Mitchell, thymine dimers and (6â€”4)photoproductsin group A xeroderma pigmentosum cell D. L. Unique DNA repair properties of a xeroderma pigmentosum revertant. Mol. lines harboring a transferred normal chromosome 9. Photochem. Photobiol., 56: Cell. Biol., 7: 3353â€”3357,1987. 365â€”369,1992. 11. Lommel, L., and Hanawalt, P. C. Increased UV resistance of a xeroderma pigmen 28. Park, C. H., Mu, D., Reardon, J. T., and Sancar, A. The general transcription-repair tosum revertant cell line is correlated with selective repair of the transcribed strand of an expressed gene. Mol. Cell. Biol., 13: 970â€”976,1993. factor TFIIH is recruited to the excision repair complex by the XPA protein mdc 12. Mori, T., Rinaldy, T. L., Athwal, R. S., Kaur, G. P., Nikaido, 0., Lloyd, R. S., and pendent of the TFIIE transcription factor. J. Biol. Chem., 270: 4896â€”4902,1995. Rinaldy, A. A xeroderma pigmentosum complementation group A related gene: 29. Evans, M. K., Taffe, B. G., Harris, C. C., and Bohr, V. A. DNA strand bias in the confirmation using monoclonal antibodies against the cyclobutane dimer and (6â€”4) repair of the p53 gene in normal human and xeroderma pigmentosum group C photoproduct. Mutat. Res., 293: 143â€”150,1993. fibroblasts. Cancer Res., 53: 5377â€”5381,1993. 13. Kaur, G. P., Rinaldy, A., Lloyd, R. S., and Athwal, R. S. A gene that partially 30. Cleaver, J. E. DNA repair and its coupling to DNA replication in eukaryotic cells. complements xeroderma pigmentosum group A cells maps to human chromosome 8. Biochim. Biophys. Acta, 516: 489â€”516,1978. Somatic Cell Mol. Genet., 18: 371â€”379,1992. 31. Bohr, V. A. Gene specific DNA repair. Carcinogenesis (Lond.), 12: 1983â€”1992, 14. Rinaldy, A., Bellew, T., Egli, E., and Lloyd, R. S. Increased UV resistance in 1991. xeroderma pigmentosum group A cells after transformation with a human genomic 32. Terleth, C., van de Putte, P., and Brouwer, J. New insights in DNA repair: preferential DNA clone. Proc. NatI. Acad. Sci. USA, 87: 6818â€”6822,1990. repair of transcriptionally active DNA. Mutagenesis, 6: 103â€”111,1991. 15. Cleaver, J. E., and Thomas, G. H. Rapid diagnosis of sensitivity to ultraviolet light in 33. Jones, C. J., Lloyd, R. S., and Wood, R. D. Analysis of cells harboring a putative fibroblasts from dermatologic disorders, with particular reference to xeroderma DNA repair gene reveals a lack of evidence for a second independent xeroderma pigmentosum. J. Invest. Dermatol.. 90: 467â€”471. 1988. pigmentosum group A correcting gene. Mutat. Res., 324: 159â€”164.1994.

6160

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 1995 American Association for Cancer Research. Overexpression of the XPA Repair Gene Increases Resistance to Ultraviolet Radiation in Human Cells by Selective Repair of DNA Damage

James E. Cleaver, Wayne C. Charles, Mindy L. McDowell, et al.

Cancer Res 1995;55:6152-6160.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/55/24/6152

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/55/24/6152. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.