(CANCER RESEARCH57, 1405-1411. April 15, 1997) Perspectives in Cancer Research The Overexpression of Bcl-2 Inhibits UVA-mediated Immediate Apoptosis in Rat 6 Fibroblasts: Evidence for the Involvement of Bcl-2 as an Antioxidant' Charareh Pourzand, Gregoire Rossier, Olivier Reelfs, Christoph Bonier, and Rex Michael Tyrrell2 School of Phannacy and Phannacology. University of Balk Bath BA2 7AY, United Kingdom (C. P., R. M. TI; Swiss Institute for Experimental Epalinges, Switzerland fG. R., 0. RI; and Biochemistry Institute, University of Fribourg, CH-l7@ZPFribourg. Switzerland (C. B.J Abstract We examined the effect of broad spectrum UVA (320-380 nm) and UvB (290-320 nm) radiation on the induction of apoptosis In the rat 6 fibroblast cell line (R6). UVA, but not UVB, Induces apoptosis In this cell line. The morphological changes and DNA ladders associated with apop tosis occurred within the first 4 h after UVA Irradiation, a phenomenon referred to as “immediate― apoptosis. From previous studies, It is known that Bcl-2 inhibits most types of apoptotic cell death. Overexpression of mouse Bcl-2 in the R6 flbroblasts Inhibited the UVA.induced immediate apoptosis. The induction of the heme oxygenase 1 (HO-i) gene by UVA is a general response to oxidative stress. As a marker of oxklative stress, we monitored the effect of Bcl-2 overexpression on the level of HO-i mRNA accumulation after UVA irradiation The results showed that the overex pression of Bcl-2 in the R6 fibroblasts strongly reduces the level of HO-i Induction from 123- to 4.9-fold. We propose that Bcl-2 expression inhibits UVA-induced immediate apoptosis via an antioxidant pathway, suppress lug either the generation or effects of specific UVA-mediated reactive oxygen species. Introduction PCD,3 also known as apoptosis, is a morphologically distinct form of cell death characterized by cell shrinkage, chromatin condensation, and cell surface blebbing that splits into apoptotic bodies (1). One hallmark of apoptosis is the rapid cleavage of DNA by an unidentified Cancer Research, CH-I@.6 antioxidant function for Bcl-2. An argument against this hypothesis is that the expression of Bcl-2 prevents apoptotic cell death under hypoxia, where ROS involvement is unlikely. Furthermore, neither ROS scavengers nor inhibitors of ROS scavengers affect cell death under these conditions,regardlessof the expressionof Bcl-2. These observations suggest that Bcl-2 exerts an anti-cell death function by a mechanismother than the modulationof ROS (15, 16). Godar et al. (17) and Godar and Lucas (18) investigated cell death mechanisms with different waveband regions of UV (i.e., UVA, 340—380 nm; UVB, 290—320 am; and UVC 200—290 nm) on L5178Y-R murine lymphoma cells. They have found that equitoxic fluences (90% reproductive death) of all waveband regions of UV radiation cause apoptosis. However, UVA induced immediate (0—4h) and delayed apoptosis, whereas UVB or UVC induced only delayed apoptosis (>20 h). Because UV radiation induces apoptosis and generates ROS (19), we investigated whether the expression of Bcl-2 will protect the cells against both the oxidizing UVA component of the sunlight and the primarily direct-acting, DNA-damaging UVB component of sunlight. R6 fibroblasts, transfected either with mouse bcl-2 (R6-Bcl-2) or control vector (R6-MCV), were used to compare the effects of irradiation on the induction of apoptosis in control cells (R6-MCV)versus Bcl-2-expressingcells (R6-Bcl-2)with different endonuclease(2), whichoccursinitiallyat nuclease-sensitivesites (3, 4). This is followed by intemucleosomal DNA cleavage, resulting in oligonucleosome chains of 180—200bp multiples. The latter phase is visible in agarose gels as a DNA ladder (5). The bcl-2 gene was initially discovered as a gene associated with a frequent translocation breakpoint in certain B-cell leukemias and was later found to suppress PCD (6) and, therefore, be involved in its regulation. The mechanism of the antiapoptotic action of Bcl-2 is still unknown.It has been shown, however,that the expressionof Bcl-2 protein decreases lipid peroxidationand prevents the induction of apoptosis by a variety of oxidative stresses, including ionizing radi ation, heat shock, exogenous hydrogen peroxide, or inhibition of glutathione synthesis (6—8).These findings suggest that Bcl-2 has antioxidant properties and inhibits PCD by suppressing the formation or effects of ROS (8, 9). Furthermore, Bcl-2 is localized in organelles known to participate in redox reactions, including those involved in the formation of ROS (10—14), supporting the hypothesis of an doses of UVA (320—380nm) and UVB (290—320nm) radiation. The involvement of ROS in UV-induced apoptosis in both cell lines was monitored by using the expression of the HO-i gene as a marker of oxidative stress (20, 21). Materials and Methods All biochemicalswerefromSigmaChemicalCo. (Poole, UnitedKingdom), except where indicated. Construction of R6 Embryo Fibroblasts Overexpressing Bcl-2. CsC1purified pMV12 plasmids in the sense orientation alone or carrying full-length wild-type mouse bcl-2 cDNA (10 @g) were transfected into subconfluent *2 cells with 25 i.@gof Lipofectin (Life Technologies, Inc., Paisley, Scotland) to producerecombinantmurine Moloneyretroviruses.After 48 h, the culture medium containing viruses was collected, filtered, and stored at —70°C. Five X 10@R6 cells (subconfluent) were infected with the virus-containing medium in the presence of 8 p@g/mlPolybrene. Forty-eight h later, the cells were trypsinized and replated into medium containing 400 @tg/mlof hygro mycin B (Calbiochem, San Diego, CA). Resistant clones of R6 cells were picked by ring isolation after 2 weeks of selection. Clones and mixed cell populationswere expanded and analyzed for Bcl-2 expression by Western blots using the a-mouse Bcl-2 antibody (results not shown). A vector control Cancer Research (United Kingdom), as well as grants from the League Against Cancer of cell line (referred as R6-MCV) and a cell line overexpressing Bcl-2 (referred as R6-Bcl-2) were used for additional studies. Cell Culture. Monolayers of the transfected rat fibroblast cell lines (R6MCV and R6-Bcl-2) were grown in 10-cm dishes for 3 days to reach 80% confluence at 37°C/5% CO2 in DMEM (Life Technologies, Inc.) supplemented Central Switzerland, the Neuchatcloisc Science Foundation. The fibroblastswere trypsinizedfor dilutiononce a week. Received 10/14/96; accepted 2/17/97. The costsof publicationof this articleweredefrayedin partby the paymentof page charges. This article must therefore be hereby marked advertisement 18 U.S.C. Section 1734 solely to indicate this fact. I This 2 To work whom was supported requests for by a core reprints grant from the in accordance with Association for League Against Cancer, and the Swiss National should be addressed, at School of Pharmacology,Universityof Bath,BathBA27AY,UnitedKingdom. 3 The abbreviations used are: PCD, International programmed cell death; ROS. Pharmacy reactive and oxygen species; HO-I, heme oxygenase 1; GAPDH, glyceraldehyde phosphate dehydrogenase. with 10% FCS (Seromed, Germany) and 50 IU/mg/ml penicillin/streptomycin. Uv Irradiation. ForUVAirradiation,fibroblastswereirradiatedwith 100, 250, 300, and 500 kJ/m2 ofbroad spectrum UVA light using a UVASUN 3000 lamp (Mutzhas, Germany). For the UVB irradiation, cells were irradiated with 1405 Downloaded from cancerres.aacrjournals.org on January 20, 2022. © 1997 American Association for Cancer Research. Bcl-2 OVEREXPRESSIONIN UVA4NDUCED APOPTOSIS 1, 2.5, 4, and 5 kJ/m2 provided by a set of six UVB lamps (TLO1; Philips, Amsterdam, Holland). Irradiations were performed at 25°C.The UV fluences were measuredusing an 1L1700 radiometer(InternationalLight, Newbury, MA). Prior to irradiation, medium was removed and retained, and the fibro blasts were covered with Ca2@/Mg@ (0.01% each) enriched PBS (Oxoid, Basingstoke, United Kingdom). After irradiation, the original media were added back to the fibroblasts and incubated at 37°C/5%CO2. Control fibro blasts were treated in the same manner, except that they were not irradiated. Clone-Forming Ability Assay. The transfectedfibroblastswere grown to 80% confluence in 10-cm plates as described above. After UV irradiation (UVA or UVB), cells were immediately trypsinized, electronically counted (Al cell counter 134; Analysis Instrument, Stockholm, Sweden), and diluted with fresh media to 400 and 1600 cells in 10-cm plates (at least three per condition and dilution) and incubated for 2 weeks at 37°C,5% CO2. Cells were then stained with crystal-violet in 1% methanol, and the colonies were counted. The fraction of surviving cells was determined relative to the sham-exposed values, which were normalized to 100%. The plating efficiency of the sham-exposed population of the two strains was 65 ±7% using these conditions. DNA Extraction. After 4 or 24 h postirradiationtime (for time course experiments,the incubationtimes were from2 to 24 h), approximately4 X 106 cells were lysedfor 1 h in lysis buffer [100m@iNaCl, 10 mt@i Ins (pH 8.0), 25 mi.iEDTA,0.5% SDS, and 0.4 mg/mIproteinaseK] at 50°C. The lysate was extracted with salt-saturated phenol:chloroform:isoamyl alcohol (25:24: 1). The RNA was digested for 1 h with 5 @g/m1 RNase at 37°C,followed by the phenol/chloroformextractionprocedure.Six to 9 @g of DNA (measured with a Uvikon 930 spectrophotometer; Kontron, Basel, Switzerland) were loaded on a 2% agarose gel supplemented with 0.05 pg/ad ethidium bromide. The DNA was visualized under a UV lamp (Bioblock Scientific, Paris, France). Apoptag Kit. The percentages ofapoptotic cells were determined using the “Oncor Apoptag Plus in Situ Apoptosis Detection Kit Fluorescein―(Appli gene, Basel, Switzerland)according to the manufacturer'sinstructions.The slides were observed under a fluorescence microscope (Axiovert 25; Zeiss, Jena, Germany).The results were scored by counting apoptoticcells (green) and viable cells (red) randomly in various fields. For each condition, 550 to 2000 cells were counted. As a negative control, some reactions were performed without the terminal deoxynucleotidyl transferase enzyme. Northern Analysis. Total RNA was isolated by the guanidinium thiocya nate-phenol-chloroform @ method (22) 30 mm, I, 2, 3, 4, and 6 h after irradiation with equitoxic fluences of both UVA (100 Id/rn2) and UVB (1 kS/rn2)sources. Fifteen of RNA was then electrophoresed in a 3-(N-morpholino)-propane sulfonic acid/formaldehyde [MOPS/HCHO] 1.3% agarose gel (23), transferred onto a ZetaProbenylon membrane(Bio-Rad,Munich,Germany),andhybrid ized either with the 32P-labeled745-bp PstI fragment ofthe mouse bcl-2 cDNA and the 32P-labeled 1000-bp EcoRl fragment of the human heme oxygenase cDNA clone 2110 (21) or with the 32P-labeledPsi! l300-bp fragmentof rat GAPDH cDNA used as an internal standard. The RNA signals on the screen were scanned with a Phosphorlmager and quantified with the software Image Quant 3.3. (Molecular Dynamics, Sunnyvale, CA). The ratio of HO-1/GAPDH andbcl-2/GAPDHdeterminedtheconcentrationofHO-1 andbcl-2 transcripts. Results Expression of Bcl-2 Modulates the Fraction of Cells Surviving after UVA but not UVB Irradiation. UVA and UVB irradiations inactivate fibroblast cells in a fluence-dependent manner. To investi gate whether the expression of Bcl-2 protects the cells against these lethaleffects,we examinedthe fractionof survivalof a population2 weeks after UVA and UVB treatment. For UVA irradiation (Fig. 1A), almost complete survival of both strains was seen in low fluence range (<100 kJ/m2), but the survival curves diverged at higher fluences, such that Bcl-2-expressing cells (R6-Bcl-2) showed a higher survival comparedwith controls(R6-MCV).R6-Bcl-2cells were significantly protected from the lethal effects of UVA irradiation at the fluences of 250 and 300 kJ/m2, when compared to R6-MCV cells. However, at high fluences (500 kJ/m2), the percentage of survival was quite low in both strains.The rather large errors may reflectdifferencesin resist ance of the cells, due to variations in confluence at the moment of irradiation(between80 and 95%). For UVB irradiation (Fig. 1B), both R6-MCV and R6-Bcl-2 strains were resistant to fluences up to 1 kJ/m2. At higher doses, survival decreased, but no significant difference in surviving fraction was observed between Bcl-2-expressing and control cells. Effect ofBcl-2 Expression on the Formation ofDNA Ladders as a Marker of Apoptosis in the R6 Fibroblast Cell Line. The for mationof DNAladderswas used in our experimentsas a biochemical hallmark of apoptosis to determine and R6-Bcl-2 strains. After 2, 4, 8, 16, and 24 h after irradiation, the DNA of the cells was isolated, and the formation of ladders was B.UVB —.—- 100• Fig. I. Inactivation of clone-forming ability of MCV —0@— Bd-2 0 10 0 0 a) ataseriesofUVA(A. 0 0 0 C 0 U 0 a) C a) 313 nm) fluences. These results represent the mean of three independent experiments in which three plates per dilution and per fluence were counted. a. —U—MCV —0—Bc$-2 100@ R6-MCV(MCV)and R6-Bcl-2(Bcl-2)fibroblasts 364nm)andUVB(B. UVA or UVB radiations differentfluencesof eitherUVAor UVBradiationsin both R6-MCV A.UVA 0 > whether induce apoptosis in a fluence-dependent manner in R6-MCV and R6-Bcl-2 cell lines. Moreover, we investigated whether the expression of Bcl-2 protects the cells against the expected induction of PCD by UV. A series of time course experiments were carried out using 10@ a) 2 ai o un @m aoo 4oo sco UVAfluences[kJ/m@] 0 1 2 3 4 UVBfluences[kJ/m@] 1406 Downloaded from cancerres.aacrjournals.org on January 20, 2022. © 1997 American Association for Cancer Research. 5 @ : j:@ @j:@ BcI-2 OVEREXPRESSION IN UVA-INDUCED APOPTOSIS A. UVA B. UVB a.2h b.4h R6-MCV C. 8h R6-MCV a. 4 h R6-BcI-2 R6-MCV R6-Bcl-2 d. 16h R6-8C12 R6-MCV R6-8c12 b. 24 h R6-MCV R6-Bcl-2 e. 24 h R6-MCV Fig.2. Detectionof DNAladdersin a 2%agarosegel afterirradiationof R6-MCVand R6-Bcl-2cellswithdifferentfluencesof UVA(A)and UVB(B)radiations.Thegenomic DNAwasextracted2 (Aa),4 (Ab),8 (Ac),16(Ad),and24 h (Ac)after100—500 kJ/m2of UVAradiation.ForUVBirradiationusingfluencesbetweenI and4 Id/rn2,onlytheagarose gelsof4 (Ba)and24 h (Bb)afterirradiationsareshown.In LanesC, thenonirradiatedcontrolsandthenumbersrepresentthefluencelevelsof UVAandUVBradiationsused.Lane 123. the 123-bp ladder, usedas a molecular weight marker. monitoredon agarosegels. Fig. 2 shows the summaryof the results the phenomenon can be considered as immediate apoptosis. After obtained from these series of experiments. As early as 2 h after UVA irradiation (Fig. 2Aa), the first ladders were visible at a fluence of 500 kJ/m2 in the R6-MCV but not in the R6-Bcl-2 cell population. After 4 h, a characteristic DNA laddering pattern was observed from 250 kJ/m2 of UVA (Fig. 2Ab) in the control strains but not in Bcl-2-expressing cells. In certain experi ments, even the lowest fluence of 100 U/rn2 showed a faint laddenng 24 h, no UVA-induced delayed apoptosis was observed. The expres sion of Bcl-2 prevented the formation of ladders in UVA-irradiated cells, indicating that Bcl-2 exerts an antiapoptotic function during the UVA-induced immediate apoptosis. The results of experimentswith UVB were quite different. The irradiation of both R6-MCV and R6-Bcl-2 with UVB wavelengths did not revealany ladderingup to 24 h (Fig.2, Ba and Bb),indicatingthe patternin the controlR6-MCVcellsafter4 h UVAirradiation(results absence of UVB-induced apoptosis in these cell lines. These results not shown). The pattern of laddering did not change 8, 16, or 24 h differfromthoseof Godaret aL (17) and Godarand Lucas(18), who observedapoptosis(delayed)20 h after UVB irradiationin a murine after irradiation, but the distance that DNA migrated into the gel increased,indicatingthe degradationof nucleosomalmonomers(Fig. lymphoma cell line. The presence of ladders was also monitored with a more sensitive 2, Ac, Ad@and 2Ae) in R6-MCV cells. The maximum intensity of radioactive method (24), which involves the end-labeling of the hy DNA laddering occurred around 4 h after UVA exposure; therefore, 1407 Downloaded from cancerres.aacrjournals.org on January 20, 2022. © 1997 American Association for Cancer Research. BcI.2 OVEREXPRESSIONIN UVA-INDUCEDAPOPTOSIS A.UVA 0 a b.24h a.4h I 00@ I 00@ —O-—Bcl.2 I I 8O@ 160. ,@ 40@ t20. g@ @. I @ I 00 @-- 250 I I • 500 300 —•--MCV —O--Bcl-2 80@ I I 60@ 40@ 20 0I I I 100 250 300 UVA fluences (kJ/m2) 500 UVA fluences (kJIm2) B., UVB . a.4h 0 b.24h 100•100-a —•—MCv I 1: I , @I a) 60@ ‘@ 40@ ‘@ 0@0—U--MCV @ 6040- 20180- 20 J80• —O--Bcl-2 .@ —O—Bd-2 0 0a I• I•12.54 0- .@— @II U C —C- I 2.54 UVB fluences(kJ/m2) UVB fluences (kJ/m2) Fig. 3. Percentage of apoptotic cells as determined by immunostaining (Apoptag kit) 4 and 24 h after irradiation of R6-MCV and R6-Bcl-2 fibroblasts with different fluences of UVA (A) and UVB (B) radiations. MCV, R6-MCV cell line; Bcl-2, R6-Bcl-2 cell line. Between 550 and 2000 cells per fluence and per stain from three independent experiments were counted. Bars, SE. pothetical ladders by the Kienow polymerase. The results (not shown) revealed DNA laddering in an R6-MCV cell line irradiated with UVA fluences of 100 to 500 kJ/m2, but no traces of ladder formation were seen either in the UVA-irradiated R6-Bcl-2 cells or in the UVB irradiated R6-MCV and R6-Bcl-2 cells. Detection of Apoptotic Cells in UVA- and UVB-irradiated R6 Fig. 3Aa shows the percentage of apoptotic cells in R6-MCV and R6-Bcl-2 cells 4 h after irradiation with UVA fluences. These results indicate that the percentage of apoptotic cells increases in a fluence dependentmanner upon UVA irradiationof R6-MCVcells, where after 4 h, the lowest (100 kJ/m2) and highest (500 kJ/m2) fluences yield 12.5 and 96% apoptotic cells, respectively. This is not the case Fibroblast Strains by an Immunostaining Method. The DNAfrag for R6-Bcl-2 cells, where the fraction of apoptotic cells remains below 3%, 4 h after irradiation with UVA. This small fraction of apoptotic mentation observed in many apoptotic cells can be detected by im munostaining with the Apoptag kit. This kit marks the apoptotic cells cells in the R6-Bcl-2 line could reflect a resting rate of apoptosis unresponsive to Bcl-2 expression in a minor fraction of the cell with green fluorescence (Fluorescein), because of their high DNA fragmentation, and marks the nucleus of intact cells with red fluores population (Fig. 3Aa). Fig. 3Ab displays the percentage of apoptotic cells 24 h after UVA irradiation in both cell lines. This percentage cence (propidium iodide), which binds nonspecifically to DNA. The percentage of apoptotic cells compared to nonirradiated controls can decreases dramatically in R6-MCV cells, where at the lowest and highest doses of UVA, only 3 and 7% of apoptotic cells were counted, also be easily estimated. The R6-MCV and R6-Bcl-2 cells were immunostained using the Apoptag kit, after 4 or 24 h irradiation time respectively. In the same experiment, the percentage of apoptotic cells with UVA and UVB fluences. The results are summarized in Fig. 3. remains less than 1% for R6-Bcl-2 cells with all UVA fluences. These 1408 Downloaded from cancerres.aacrjournals.org on January 20, 2022. © 1997 American Association for Cancer Research. @ . . . Bcl-2 OVEREXPRES5ION IN UvA-INDUCED APOPTO5IS A. lOOkJ/m2UVA I B. -.-Mcv lkJ/m2UVB 1I@ —•—MOV -.-@4 I @ 12 ‘E @ S I •• 0 @ H@is Ha)w$ aftSr Iff@ R6—pMCV @ 4b**@*@ 0; R6—pMCV R6-Bcl--2 R6-Bcl-2 HO-I t “ @ “.- BcI-2 , . . GAPDH C 1/2 Ih 2h 3h 4h6h C 1121h2h3h4h6h C 1/2 Ih 2h 3h 4h 6h C 1/2 Ih 2h 3h 4h 6h Fig. 4. Effect of Bcl-2 expression on HO-i induction in R6-MCV and R6-Bcl-2 cell populations after UVA (A) and UVB (B) irradiations. RNA extraction was perfonned at various thne points after irradiations with 100 id/rn2 UVA and 1 kJ/m2 of UVB. Northern blots of the total cellular RNA were first probed for HO-I mRNA and then for bcl-2 and OAPDH mRNA. HO-I, bcl-2, and GAPDH mRNA signals were quantified by Phosphorhuager laser scanning, where the GAPDH signal was used as an internal loading control. The HO-l mRNA signal, normalized against GAPDH for differences in loading between samples, was expressed as a fold-increase above basal level and plotted as a function of time after UVA and IJVB irradiations. results clearly demonstrate that UVA induces only “immediate― ap optosis (within 4 h) in control cells and that Bcl-2 expression protects the cells against this phenomenon. The percentages of apoptotic cells in control cells, 24 h after irradiation with UVA, are in agreement with the DNA ladderingresults (Fig. 2Ae).These low fractionsprobably display the staining of fragmented DNA in a few cells remaining from the UVA-induced immediate apoptosis. Fig. 3B shows the results of Apoptag staining 4 (Fig. 3Ba) and 24 h (Fig. 3Bb) after UVB irradiation of both R6-MCV and R6-Bcl-2 cells. From these results, it is obvious that UVB does not induce either immediate or delayed apoptosis in R6 control and R6 Bcl-2-express ing cell lines. These results are also in agreement with the DNA laddering experiments. Effect of Bcl-2 Expression UVA and UVB Irradiation on HO-i mRNA Accumulation after of R6-Bcl-2 Fibroblasts. It is known that followingUVA-inducedoxidativestress,the HO-i gene is tran scnptionallyactivatedin humanfibroblastcells (21, 25). Becausethe involvement of ROS in apoptosis and the role of Bcl-2 as an antiox idant. For this purpose, R6-MCV and R6-Bcl-2 cells were irradiated with a l00-kJ/m2 fluence of UVA (a fluence corresponding to 90 ± 5% survival; see above), and in a series of Northern blot analyses, the levels of HO-l mRNA accumulation were measured in both strains. We also carried out Northern blot analyses with an equitoxic fluence of 1 kJ/m2 UVB (fluence corresponding to 87 ±6% survival) using the HO-l probe, although from the literature (26), we did not expect to find a strong transcriptional activation of HO-i gene by UVB irradiation. We also evaluated the level of bcl-2 mRNA accumulation obtained after UVA and UVB radiations to check that the bcl-2 gene is not transcriptionallyactivatedas a result of irradia tion. Fig. 4 show the results obtained from the Northern analyses. In the R6-MCV cell line, the bcl-2 mRNA was neither constitu lively expressed (Fig. 4, bcl-2 cDNA probe, Lanes i) nor induced by UV (UVA and UVB). In Bcl-2-overexpressing cells, the bcl-2 mRNA was constitutively overexpressed as expected (Fig. 4, bcl-2 cDNA probe, Lanes 8), but as for the control cell line, no increase in bcl-2 HO-i gene can be considered as a marker of oxidative stress (25), its induction by UVA was exploited to provide information about the 1409 Downloaded from cancerres.aacrjournals.org on January 20, 2022. © 1997 American Association for Cancer Research. BcI-2 OVEREXPRESSIONIN UVA-INDUCEDAPOPTOSIS mRNA signal was observed after UVA or UVB treatments (Fig. 4, bcl-2 cDNA probe, Lanes 9—i4). The second hybridizations with the HO-I cDNA probe are shown in Fig. 4 with equitoxic fluences of 100 and 1 kJ/m2 for UVA and UVB, respectively. In R6-MCV cell line, the HO-i gene was induced up to 12.5-fold over the control value, 3 h after irradiation with 100 kJ/m2 UVA. For the R6-Bcl-2 cell line, the same fluence of UVA induced HO-l mRNA accumulation to a maximum of 4.9-fold over the maximum peak control value. The HO-l mRNA signal in R6Bcl-2 was 2.5-fold lower than in R6-MCV cells. Higher fluences of 250 and 300 kJ/m2 UVA irradiation did not further modulate the level of HO-l mRNA in the two cell lines (results not shown). In the case of UVB irradiation (lkJ/m2), the HO-i induction was very low and in the same order of magnitude (2-fold the control values) for both strains (Fig. 4B, HO-i cDNA probe). DNA-damaging agent. The oxidative component of UVB is evident only under conditions of compromised antioxidant activity (e.g., glu tathione depletion; Ref. 32). This is not the case for UVA radiation, which exerts biological effects primarily by oxidative pathways (33). The dramatic stimulation of HO-i mRNA accumulation by UVA radiation compared with the low levels of induction (2—3-fold)after UVB treatment of cultured human fibroblasts (26) and R6-MCV fibroblasts (Fig. 4) is entirely consistent with this concept. Further more, UVA is a membrane-damaging agent and induces peroxidation in membrane lipids of cultured human skin keratinocytes and fibro blasts (34, 35). In three strains of munne lymphoma LY cells, both immediate apoptosis and membrane permeability were increased by UVA in a fluence-dependent manner (19). Moreover, the pretreatment of LY cells with vitamin E (a lipophilic free radical scavenger) reduced this effect, suggesting a possible role for free radical-induced membrane damage in immediate apoptosis. Singlet oxygen (‘02)is apparently involved in oxidative damage to membrane lipids in vitro Discussion Our goal in the present study was to investigate whether UV radiation (both UVA and UVB) induces apoptosis in the R6 fibroblast cell line and, if so, to determine whether the expression of Bcl-2 protects the cells against such apoptosis. We found that UVA induces apoptosis in the R6-MCV cell line. Because this apoptotic induction occurred within the first 4 h after UVA irradiation, we referred to the phenomenon as “immediate― apoptosis, in agreement with previous nomenclature (18). In the case of UVB radiation, we observed neither immediate nor delayed apoptosis. The latter result is in contrast to previous observations by Godar and co-workers (17), who showed that UVB irradiation induces delayed apoptosis (>20 h after irradia tion) in L5178Y-R murine lymphoma cells. The reason for this discrepancy is not clear. One explanation would be the difference of sensitivity that exists between both cell lines regarding the applied UVB fluence. Godar et a!. (17) used 75 JIm2 UVB fluence to induce delayed apoptosis in L5178Y-R cells. This fluence, which corre sponds to 10% survival in the clonogenicity assay of LY cells, is not a lethal fluence for the R6 fibroblasts. On the other hand, it is known that UVB radiation results in the formation of apoptotic keratinocytes called sunburn cells (27). Recent mechanistic studies in this field reveal evidence for the requirement of DNA damage (pynmidine dimers; Ref. 28) and the induction ofp53 gene (29) for UVB-induced sunburn cell formation (or apoptotic cells). Also, it appears that tumor necrosis factor-a, which can be released from keratinocytes and is known to cause apoptosis in particular cells, is partially involved in the formation of UVB-induced sunburn keratinocytes (30). These studies suggest the possibility that UVB-induced delayed apoptosis, in contrast to UVA-induced immediate apoptosis, is cell type specific. In additional experiments, we found that the expression of Bcl-2 inhibits UVA-induced immediate apoptosis even at high fluences (500 kJIm2), where the percentageof viability is less than 2%. Because UVA exerts its biological effects via the induction of an oxidative stress (25, 26, 31), the latter result suggested that Bcl-2, besides its antiapoptotic function, possesses antioxidant properties, as suggested previously (8, 9). To verify this hypothesis, we performed a series of Northern analyses using HO-i gene activation as a marker of oxida tive stress and found that the expression of Bcl-2 dramatically de creases the level of HO-i induction following UVA but not UVB irradiation. These results are in agreement with the known differences between UVA and UVB wavelengths. UVA is weakly absorbed by (36, 37) and in human skin fibroblasts (38) and appears to be the primary effector molecule in UVA-induced transcription of the HO-i gene in human fibroblasts (39). These findings, together with those of the present study in which we have shown that Bcl-2 expression prevents UVA-induced immediate apoptosis and decreases the level of HO-i induction, suggest that Bcl-2 prevents the UVA-mediated apoptosis via suppression of the formation or effects of ROS such as ‘02.Furthermore, the localization of Bcl-2 in the mitochondrial membrane (13, 14) suggests that this may also be the site of the UVA-generated ROS involved in HO-i activation and apoptosis. Acknowledgments We thankPatrickLuscherfor excellent technicalassistanceand Dr. Stefan Ryterfor useful discussions. References I. Kerr, I. F. R., Wyllie, A. H., and Currie, A. R. Apoptosis: a basic biological phenomenon with wide-ranging implications in tissue kinetics. Br. J. Cancer, 26: 239—257, 1972. 2. Wyllie, A. H. Glucocorticoid-induced thymocyte apoptosis is associated with endog enous endonucleasc activation. Nawre (Lond.), 284: 555—556,1980. 3. Oberhammer, F. A., Wilson, J. W., Dive, C., Moms, L D., Hickman, J. A., Wakeling, A. E., Walker, P. R., and Sikorska, M. Apoptotic death in epitheial cells: cleavage of DNA to 300 and/or 50 kb fragments prior to or in the absence of internucleosomal fragmentation. EMBO J., 12: 3679—3684, 1993. 4. Brown, D. 0., Sun, X. M., and Cohen, G. M. Dexamethasone-induced apoptosis involves cleavage of DNA to large fragments prior to internucleosomal fragments tion. J. Biol. Chem., 268: 3037—3039,1993. 5. Arends,M. J., and Wyllie, A. H. Apoptosis:mechanismsand roles in pathology.Int. Rev. Exp. Pathol., 32: 223—254,1991. 6. Reed,J. C. Bcl-2 andthe regulationofprogrammed cell death.J. Cell Biol.. 124: 1—6. 1994. 7. Thong, L. 1., Sarafian, T., Kane, D. J., Charles, A. C., Mab, S. P., Edwards, R. H., and Bredesen, D. E. BcI-2 inhibits death of central neural cells induced by multiple agents. Proc. Nail. Aced. Sci. USA, 90: 4533—4537,1993. 8. Hockenbery, D. M., Oltvai, Z. N., Yin, X-M., Milliman, C. L.. and Korsmeyer, S. J. BcI-2 functions in an antioxidant pathway to prevent apoptosis. Cell, 75: 241—251, 1993. 9. Kane,D. J., Sarafian,T. A., Anton,R., Hahn,H., Gralla,E B., Valentine,J. S., Ord, T., and Bredesen, D. E. Bcl-2 inhibition of neural death: decreased generation of reactiveoxygenspecies.Science(WashingtonDC),262: 1274—1277, 1993. 10. Hockenbery, D. M.. Nunez, 0., Milliman, C., Schreiber, R. D., and Korsmeyer, S. J. Bcl-2isan innermitochondrialmembraneproteinthatblocksprogrammedcelldeath. Nature (Land.), 348: 334-336, 1990. 11. Monaghan, P., Robertson. D., Amos, T. A. S., Dyer, M. J. S., Mason, D. Y., and Greaves M. F. Ultrastructural localization ofBcl-2 protein. J. Histochem. Cytochem., 40: 1819—1825,1992. 12. Nguyen, M., Millar, D. G., Yong, V. W., Korsmeyer, S. J., and Shore, G. C. Targeting of BcI-2 to the mitochondrial outer membrane by a COOH-terminal signal-anchor most of biomolecules but is oxidative in nature, generating active sequence. J. Biol. Chem., 268: 25265-25268, 1993. oxygen intermediates via a variety of chromophores (25, 26, 31), 13. Lithgow, T., van Driel, R., Bertram, J. F., and Strasser, A. The protein product of the oncogene bcI-2 is a component of the nuclear envelope, the endoplasmic whereas UVB, although it does have an oxidative component, is reticulum, and the outer mitochondrial membrane. Cell Growth & Differ., 5: strongly absorbed at biologically relevant levels by critical macromol 411—417, 1994. ecules, including DNA. As a consequence, UVB is a very strong 14. Nakai, M., Takeda, A., Cleary, M. L., and Endo, T. The bcl-2 protein is inserted into 1410 Downloaded from cancerres.aacrjournals.org on January 20, 2022. © 1997 American Association for Cancer Research. Bcl-2 OVEREXPRESSION IN UVA-INDUCED APOFI'OSIS the outermembranebutnot intothe innermembraneof ratliver mitochondriain vitro. Biochem. Biophys. Rca. Commun., 196: 233-239, 1993. 15. Shimizu, S., Eguchi, Y., Kosaka,H., Kamiike,W., Matsuda,H., and Tsujimoto,Y. Preventionofhypoxia-inducedcelldeathby Bcl-2andBcl-xL.Nature(Land.),374: 811—813,1995. 16. Jacobson,M. D., andRaff, M. C. Programmedcell deathandBcl-2 protectionin very low oxygen. Nature(Load.), 374: 814—816,1995. 17. Godar,D. B., Miller, S. A., andThomas,D. P. Immediateanddelayedapoptoticcell death mechanisms: 59—66, 1994. UVA versus UVB and UVC radiation. Cell Death & Differ., 1: 18. Godar,D. E., and Lucas, A. D. Spectraldependenceof UV-inducedimmediateand delayed apoptosis:the role of membraneand DNA damage.Photochem.Photobiol., 62: 108—113,1995. 19. Haffiwell,B., and Gutteridge,J. M. C. Role of free radicalsand catalyticmetal ions in human disease: an overview. Methods Enzymol., 186: 1—85,1990. 20. Keyse, S. M., and Tyrrell, R. M. Both near ultraviolet radiation and the oxidizing agent hydrogen peroxide induce a 32 kDa stress protein in normal human skin fibroblasts. J. Biol. Chem., 262: 14821—14825, 1987. 21. Keyse, S. M., and Tyrrell, R. M. Heme oxygenase is the major 32 kDa stress protein induced in human skin fibroblasts by UVA radiation, hydrogen peroxide and sodium arsenite. Proc. NatI. Acad. Sci. USA, 86: 99-103, 1989. 22. Chomczynski, P., and Sacchi, N. Single step method of RNA isolation by acid guanidinium thiocyanate-phenol-cMomfonu extraction. Anal. Biochem., 162: 156- J., Remington,L., Jacks,T., and Brash,D. E. Sunburnand p53 in the onset of skin cancer. Nature(Land.), 372: 773—776, 1994. 30. Schwarz, A., Bhardwaj, R., Aragane, Y., Mahnke, K., Riemann, H., Metze, D., Luger, T. A., andSchwarz,T. Ultraviolet-B-inducedapoptosisofkeratinocytes:evidence for partial involvement of tumor necrosis factor-a J. Invest. Dermatol.,104: 922—927, 1995. in the formation of sunburn cells. 31. Tyrrell, R. M. UVA (320-380 nm) radiation as an oxidative stress. In: H. Sies (ed), Oxidative Stress: Oxidants and Antioxidants, pp. 57-83. London: Academic Press, 1991. 32. Tyrrell, R. M., and Pidoux, M. Correlation between endogenous glutathione content and sensitivity of cultured human skin cells to radiation at defined wavelengths in the solar UV range. Photochem. Photobiol.. 47: 405-412. 1988. 33. Tyrrell, R. M. Activation of mammalian gene expression by the UV component of sunlight—from models to reality. BioEssays, 18: 139—148, 1996. 34. Punnonen, K., Jansen, C. T., Puntala, A., and Ahotoupa, M. Effects of in vitro UVA irradiation and PUVA treatment on membrane fatty acids and activities of antioxidant enzymes in human keratinocytes. 3. Invest. Dermatol., 96: 255—259,1991. UVA-induced lipid peroxidation in cultured human fibroblasts. Biochem. Biophys. Acts, 1084:261—268, 1991. 36. Bose,B., Argawal,S., and Chanerjee,S. N. UV-Ainducedlipid peroxidationin 23. Davis, L. 0., Dibner, M. D., and Battey, I. F. Basic Methods in Molecular Biology, p. 143. New York: Elsevier, 1986. liposomal membrane. Radiat. Environ. Biophys., 28: 59—65,1989. in human cells. Nucleic AcidsRes.,20: 5243,1992. 25. Keyse,S. M., Applegate,L A., Tmmvoukis,Y., andTyrrell,R. M. Oxidantstress leads to transcriptional activation ofthe human heme oxygenase gene in cultured skin fibroblasts.Mol.Cell.Biol.,10:4967—4969, 1990. 26. Applegate,L. A., Luscher,P., and Tyrrell,R. M. Inductionof hemeoxygenase:a general response to oxidant stress in cultured mammalian cells. Cancer Rem.,51: 974—978, 1991. 27. Young,A. R. The sunburncell.Photodermatology, 4: 127—134, 1987. Derniatol.,85: 365—367, 1985. 29. Ziegler,A.,Jonason,A. S., Leffel,D. J., Simon,J. A., Sharma,H. M., Kimmelman, 35. Morliere,P., Moysan,A., Santus,R., Huppe,G., Maziere,i-C., and Ertret, L. 159, 1987. 24. Rosi, F. A simple and rapid method for detection ofapoptosis 28. Lay, R. D., and Applegate, L. A. Ultraviolet radiation-induced histopathological changes in the skin of the marsupial Monodeiphis domestica. II. Quantitative studies of the photoreactivation of induced hyperplasia and sunburn cell formation. J. Invest. 37. Bose, B., Argawal, S., Chatterjee, S. N. Membrane lipid peroxidation by UV-A: mechanism and implications. Biotech. Appl. Biochem., 12: 557—561,1990. 38. Vile,0. F.,andTyrrell,R.M.UVAradiation-induced oxidativedamageto lipidsand proteins in vitro and in human skin fibroblasts is dependent on iron and singlet oxygen. Free Radical Biol. Med., 18: 721—730,1995. 39. Basu-Modak, S., and Tyrrell, R. M. Singlet oxygen: a primary effector in the ultravioletA/nearvisible light inductionof the humanheme oxygenase gene. Cancer Res., 53: 4505—4510, 1993. 1411 Downloaded from cancerres.aacrjournals.org on January 20, 2022. © 1997 American Association for Cancer Research. The Overexpression of Bcl-2 Inhibits UVA-mediated Immediate Apoptosis in Rat 6 Fibroblasts: Evidence for the Involvement of Bcl-2 as an Antioxidant Charaeh Pourzand, Gregoire Rossier, Olivier Reelfs, et al. Cancer Res 1997;57:1405-1411. Updated version E-mail alerts Reprints and Subscriptions Permissions Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/57/8/1405 Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at pubs@aacr.org. To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/57/8/1405. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site. Downloaded from cancerres.aacrjournals.org on January 20, 2022. © 1997 American Association for Cancer Research.