Protection of Normal Human Reconstructed Epidermis from UV By

ORIGINAL ARTICLE Protection of normal human reconstructed epidermis from UV by catalase overexpression HR Rezvani1, M Cario-Andre´1, C Pain1, C Ged1, H deVerneuil1 and A Taı¨ eb1,2 1Inserm E 0217, University Victor Segalen Bordeaux 2, Bordeaux, France and 2CHU Bordeaux, Department of Dermatology, Hoˆpital St Andre´, Bordeaux, France

Reactive oxygen species (ROS) generated by ultraviolet (UV) irradiation are counterbalanced by endogenous antioxidant systems. To test the hypothesis of a novel photoprotective approach, we irradiated epidermis reconstructed with normal human keratinocytes overexpressing sustainably lentivirus-mediated catalase (CAT), copper/zinc superoxide dismutase (CuZnSOD) or manganese superoxide dismutase (MnSOD) enzymes. We found that following UVB irradiation there was a marked decrease in sunburn cell formation, caspase-3 activation and p53 accumulation in human reconstructed epidermis overexpressing CAT. Moreover, UVA-induced hypertrophy and DNA oxidation (8-oxodeoxyguanosine) were decreased by CAT overexpression. These effects were not achieved by overexpression of CuZnSOD or MnSOD. In conclusion, vector-mediated CAT overexpression could be a promising photoprotective tool against deleterious effects of UV irradiation such skin cancer especially in monogenic/polygenic photosensitive disorders characterized by ROS accumulation. Cancer Gene Therapy (2007) 14, 174–186. doi:10.1038/sj.cgt.7701000; published online 20 October 2006 Keywords: gene transfer; superoxide dismutase; photoprotection; lentiviral vector

Introduction damage under the control of the tumor suppressor protein p53.6,7 However, several studies on different cells have The solar ultraviolet (UV) radiation reaching the earth’s shown that UVBalso acts as an inducer of death 8,9 10,11 surface is the most important risk factor for the receptors and ROS, both of which participate in 12 development of skin cancer. Non-melanoma skin cancer UVB-induced apoptosis. consists of basal cell carcinoma (BCC), the most prevalent Thus, UVA and UVBcan induce ROS such as the À form of skin cancer, and squamous cell carcinoma superoxide anion radical (O2K) and hydrogen peroxide 1,2 (SCC). The rapid rising incidence of both BCC and (H2O2) which could be further converted to the highly SCC over the past decades, suggests that atmospheric reactive hydroxyl radical (KOH) via the Fenton 2 þ - 3 þ K À ozone depletion may further increase these rate. Thus, (Fe þ H2O2 Fe þ OH þ OH) and Haber–Weiss À - K À 13 there is a vital requirement to develop a strategy for (O2K þ H2O2 O2 þ OH þ OH) reactions. In order to prevention and therapy of skin cancer. neutralize these ROS, living cells have acquired various Solar UV radiation includes UVC (200–280 nm), UVB defense systems including non-enzymatic (a-tocopherol, À (280–320 nm) and UVA (320–400 nm) light. Whereas vitamin C) and enzymatic antioxidants. In particular, O2K UVC is absorbed by the atmospheric layer, UVA and is converted to less reactive H2O2 and O2 by superoxide UVBboth function as initiator and promoter in photo- dismutase (SOD), an enzyme present in three forms in carcinogenesis.2 UVA, which is the major component of humans, namely a cytosolic copper/zinc superoxide solar UV radiation, is poorly absorbed by DNA and dismutase (CuZnSOD), a mitochondrial manganese affects various cell components such as activator protein- superoxide dismutase (MnSOD) and an extracellular 3 4 1 (AP-1) and heme oxygenase, mainly through reactive one (EC-SOD). H2O2 is converted to H2O and O2 either oxygen species (ROS) production.4,5 UVBis initially by catalase (CAT) located in peroxisomes or glutathione 13 absorbed by DNA and causes direct DNA damage. It was peroxidase located in cytoplasm and mitochondria. originally thought that sunburn cell (SBC) formation was Heme oxygenase decreases iron availability and thereby 14 only the consequence of UVB-induced DNA photo- restricts KOH formation. However, an imbalance between ROS-producing levels and the antioxidant defense is in favor of the former upon UV irradiation Correspondence: Professor A Taı¨ eb, Inserm E 0217, Universite´ and leads to cytotoxicity by proteins, lipids and DNA Victor Segalen Bordeaux 2, 146 rue Le´ o Saignat, 33076 Bordeaux 15–18 cedex, France. oxidation both in vitro and in vivo. It has been shown E-mail: [email protected] that UVA and UVBirradiation of murine skin results in Received 1 May 2006; revised 13 June 2006; accepted 19 August impairment of enzymatic and non-enzymatic antioxi- 5 2006; published online 20 October 2006 dants. In addition, several studies have demonstrated Protective effect of catalase against UV HR Rezvani et al 175 that an increase in antioxidant defense systems can green fluorescent protein (EGFP), CAT, CuZnSOD and reduce the deleterious effects of UV-induced ROS. MnSOD was driven by the human phosphoglycerate The protective effects of non-enzymatic antioxidant or kinase (PGK) promoter. Lentiviral particles were pro- radical scavenger addition on UVB-,18–20 UVA-20 and duced by transient transfection of 293T cells as described UVA þ B-21 induced damage have been reported. We previously.26 To determine the titer of each viral vector, have recently demonstrated that overexpression of CAT serial dilutions were used for transduction of 293T cells. reduces UVB-induced apoptosis in human keratino- Following transduction, EGFP expression was measured cytes.22 Furthermore, a severe disturbance in antioxidant directly by cytofluorimetric analysis. Typical titers after balance with diminished antioxidant enzymes was found concentration for lentiviral vectors were between 2 Â 108 in non-melanoma tumors.23 Taken together, these data and 2 Â 109/ml. Enzyme-linked immunosorbent assays of suggest that increase in antioxidant defense may protect p24 were used to determine the concentration of viral p24 keratinocytes from UV-induced oxidative stress. Mean- protein in the different viral supernatants. By comparing while, controversial results from various isolated cells the p24 of different supernatants and specific titers of point out that in vivo data are necessary to understand the EGFP, the titration of different viral supernatants was complex of oxidative stress and antioxidant defense in determined. response to UV irradiation. Thus, we decided to investigate the effects of UVA and UVBirradiation on Transduction of keratinocytes reconstructed epidermis (RE) made with human kerati- A total of 5 Â 105 cells were plated in T25 flasks and nocytes transduced by CuZnSOD, MnSOD or CAT, incubated for 24 h in complete medium. Before infection, because this system is more similar to human skin than medium was removed and cells were infected with viral keratinocyte monolayers. supernatants for 24 h at 371C in the presence of 8 mg/ml protamine sulfate. After 5 days, the cells were trypsinized and used to prepare RE. Before reconstruction, the percentage of EGFP-positive cells was analyzed by Materials and methods cytofluorimetry. Source of keratinocytes and epidermal reconstruction Keratinocytes were isolated from normal human skin in Irradiation procedure 22 patients undergoing plastic surgery as already described. UVBand UVA irradiation was delivered with a Biotronic Briefly, fresh skin fragments were immediately cut into device (Vilber Lourmat, Marne la Valle´ e, France) 5 Â 5 mm pieces and treated with trypsin-ethylenediamine- equipped with a dosimeter as already described.21,22 tetraacetic acid (EDTA) for 3 h at 371C or overnight at Briefly, the UVB lamp emitted a continuous spectrum 41C in order to separate the epidermis from the dermis. between 280 and 380 nm with a major peak at 312 nm. 5 Keratinocytes were seeded at a concentration of 10 cells The UVA lamp delivered UV in the range of 312–400 nm 2 per cm in MCDB153 medium, which included hydro- with a major peak at 365 nm. After rinsing in phosphate- cortisone (0.5 mg/ml), epidermal growth factor (10 ng/ml), buffered saline (PBS), RE were irradiated at a dose of 180 insulin (5 mg/ml) and bovine pituitary extract. This or 220 mJ cmÀ2 UVBand at a dose of 10 J cm À2 UVA. medium was changed three times a week. When the These doses were chosen after preliminary tests in order to culture reached 70–80% confluence, the cells were induce significant levels of apoptosis (20 and 40%) in the detached with trypsin-EDTA and then resuspended in basal layer of normal RE. Following irradiation, samples MCDB153 medium or were used to reconstruct epidermis. were rinsed in PBS and incubated in fresh medium.

Preparation of RE Histological and immunohistochemical studies Normal skin samples were cut into 1 Â 1 cm pieces and Reconstructs were fixed in 4% formaldehyde, embedded then kept at 371C for 15 days. Then, dermis was separated in paraffin, cut in 5 mm sections and stained with from epidermis and conserved at À801C. Normal human hematoxylin–eosin to assess the general morphology of keratinocytes were seeded at 4 Â 105 cells per cm2 on the the epidermis and the amount of SBCs. We counted 100 dead dermis. Twenty-four hours later, the dermis was cells each time on 10 different areas for quantification of immersed in culture medium for 72 h to allow cell SBC. The sections of epidermis reconstructed with proliferation. Then, it was placed at the air/liquid keratinocytes transduced by EGFP (RE/EGFP) were interface for 7 days to allow differentiation.24 counterstained with hematoxylin and analyzed under fluorescence microscopy to quantify transduction. The Construction and production of lentiviral vectors monoclonal antibody against CAT (Rockland, Tebu-Bio, The pRRL.SIN.PPT.PGK.MCS.WPRE and the Le Perray en Yvelines, France) was also used for this pRRL.SIN.PPT.PGK.EGFP.WPRE vectors, hereafter purpose. DNA oxidation was detected 1 h after UVA termed TPW and TPEW, were a generous gift from D irradiation with a monoclonal anti-8-oxodeoxyguanosine Trono (Universite´ de Gene` ve, Switzerland).25 The vectors (anti-8-oxodG) antibody (RD systems, Lille, France) (TPCATW, TPCuZnSODW and TPMnSODW, see according to the manufacturer’s recommendations. Figure 1a) were constructed by inserting CAT, CuZnSOD Accumulation of p53 protein was detected 24 h after or MnSOD cDNAs into the multiple cloning site irradiation using anti-human p53 antibody (clone DO-7 (MCS) of the TPW vector. Expression of enhanced Dako, Trappes, France). We used the EnVision þ HRP

Cancer Gene Therapy Protective effect of catalase against UV HR Rezvani et al 176 RRE a TPW CMV R U5 cPPT CTS PGK WPR E R U5 SD SA TRIPLEX MCS U3 TPEW CMV R U5 RRE cPPT CTS PGK EGFP WPRE R U5 SD SA TRIPLEX U3 RRE TPCATW CMV R U5 cPPT CTS PGK Catalase WPRE R U5 SD SA TRIPLEX U3 RRE TPMnSODW CMV R U5 cPPT CTS PGK MnSOD WPRE R U5 SD SA TRIPLEX U3 RRE TPCuZnSODW CMV R U5 cPPT CTS PGK CuZnSOD WPRE R U5 SD SA TRIPLEX U3

bc1000 12345678910111213

800 10 600 400 6 200 4 0 100 101 102 103 104 3 EGFP 2.5 1000 0 1 5 10 20 1.5 800 600 400 0.5 SSC-Height200 SSC-Height 0.25 0 100 101 102 103 104 EGFP Figure 1 Schematic drawing of the vectors used and analysis of their transduction efficiency. (a) Vectors carry an internal cassette for the EGFP, CAT, CuZnSOD or MnSOD driven by the promoter of human PGK gene. DU3, R and U5 are the LTR (long terminal repeat) regions, with a deletion including the enhancer and the promoter from U3. CMV is the cytomegalovirus promoter, SD is the major splice donor site, SA is the splice acceptor site, RRE is the rev-response element, cPPT is the nuclear import sequence and WPRE is the regulatory element of woodchuck hepatitis virus. (b) Fluorescence-activated cell sorting analysis demonstrates the percentages of keratinocytes expressing EGFP, 5 days after transduction by the lentiviral vector TPEW (approximately 90%). (c) Southern blot analysis was performed to test the integrity of different proviruses. Xho1 and Kpn1 digested DNA was gel-fractionated, blotted and hybridized with 32P-labeled WPRE probe. Lanes 1 and 13: molecular weight markers; lanes 2–7: DNA from cells transduced by EGFP (multiplicity of infection (MOI) ¼ 30), EGFP (MOI ¼ 6), CuZnSOD, MnSOD, CAT and TPW, respectively; lanes 8–12: 0, 11.24, 56.2, 112.4 and 224.8 pg of TPEW plasmid were mixed with 10 mg of normal DNA corresponding to 0, 1, 5, 10 and 20 copies per cell of the transgene, respectively.

Kit (Dako, Trappes, France), which enhances staining was measured by Amplex Red Catalase Assay Kit sensitivity. (Molecular Probes, Invitrogen, Cergy-Pontoise, France). Standard curves for the enzymatic activities of SOD and Determination of CAT, MnSOD and CuZnSOD CAT were plotted using purified enzyme preparations. activities Enzyme-specific activities were expressed as U/mg protein Frozen reconstructs were thawed. The epidermis were in epidermis homogenate. One unit of CAT activity was separated from dermis, homogenized with a Potter defined as 1 mmol H2O2 consumed per minute. One unit of homogenizer in 0.1 M Tris-Cl (pH 7.5) and then sonicated SOD activity was defined as the amount of enzyme that for two 30-s bursts. After 10 min of centrifugation at inhibits 50% of WST-1 formazan formation per minute. 12 000 g, aliquots of the obtained supernatant were stored at À801C. The total protein content of each epidermis was Quantification of proteins by Western blot measured by the BCA kit Reagent (Pierce, Bezons, Equal amounts of total protein were resolved by sodium France). SOD was assayed by using the SOD assay kit- dodecyl sulfate-polyacrylamide gel electrophoresis (12% WST (Dojindo Molecular Technologies, Gaithersburg, gel) and electrophoretically transferred to a polyvinyl- À2 MD). To determine MnSOD activity, KCN (2 mM) was idene difluoride (PVD) membrane at 1.2 V cm for 1 h added to the mixture to inhibit CuZnSOD activity. CAT 20 min. After blocking the membranes with 5% non-fat

Cancer Gene Therapy Protective effect of catalase against UV HR Rezvani et al 177 dry milk in Tris-buffered saline with 0.1% Tween 20 measured after 5 days of culture (Figure 2a, b and c). In (TBS-T) for 1 h at room temperature in a shaker, the blots the cells transduced by CAT, MnSOD and CuZnSOD or were incubated at 41C with 1:1000 dilution of the co-transduced by CAT and CuZnSOD, activity of the antibody anti-CAT (Rockland, TEBU), anti-MnSOD respective enzymes was highly increased compared to the (Santa Cruz Biotechnology, Tebu-Bio, Le Perray en non-transduced cells. Our results showed that transduc- Yvelines, France) or anti-CuZnSOD (Santa Cruz Bio- tion of cells by the CAT-expressing vector did not affect technology, TEBU). The blots were washed with TBS-T SOD activity or vice versa. Western blot analysis revealed (3 Â 10 min) 16 h later and incubated at room temperature that increased enzyme activities were correlated with an with 1:1000 dilution of anti-immunoglobin horseradish increase in the corresponding proteins (Figure 2d). peroxidase-linked antibody (Vector Laboratories, Bio- valley SA, Marne la Valle´ e, France) for 1 h. The chemi- Epidermis reconstructed with different transduced luminescence ECL Reagent was used to develop the blots keratinocytes reveal efficient expression of transgenes after washing the membrane three times for 10 min with We next examined whether transduced keratinocytes are TBS-T and twice with TBS alone. capable to develop into a fully stratified epidermis with To determine the activation of caspase-3 and the basal, suprabasal, granular and cornified layers. Histol- accumulation of serine-15 phosphorylated p53, epidermis ogy of the normal RE made from all genetically modified were separated from dermis, homogenized with a Potter keratinocytes revealed an architecture similar to the non- homogenizer in lysis buffer (25 mM N-2-hydroxyethylpi- transduced RE (Figure 3a). perazine-N0-2-ethanesulfonic acid, pH 7.4; 150 mM NaCl; To determine whether the genes transferred by lenti- Triton X-100 1% v/v) and then incubated at 41C for vectors were expressed in RE, we assessed expression of 30 min. After 10 min of centrifugation at 12 000 g, aliquots EGFP and CAT by fluorescence microscopy and im- of the supernatant were stored at À801C. Total protein munohistochemistry, respectively. Activities of antioxi- content of each epidermis was measured by the BCA kit dant enzymes were also measured for this purpose. EGFP Reagent (Pierce, Bezons, France). Western blot was fluorescence was absent in epidermis reconstructed with carried out as mentioned above with caspase-3 mono- non-transduced keratinocytes (Figure 3b) and was intense clonal antibody (Cell Signaling Technology, Tebu-Bio, Le in epidermis made with EGFP-transduced keratinocytes Perray en Yvelines, France) or anti-human serine-15 (Figure 3c). By immunohistochemistry with CAT anti- phosphorylated p53 (Cell Signaling Technology). body, we found that cells overexpressing CAT were located in all the layers of RE (Figure 3d and e). Statistics Measurement of antioxidant enzymes revealed that in Student’s t-test was applied for statistical evaluation. P- epidermis reconstructed with cells transduced by CAT values o0.05 were considered significant. Results are (RE/CAT) or co-transduced by CAT and CuZnSOD (RE/ presented as means7s.e.m. of five independent experiments. CuZnSOD þ CAT), CAT activity increased 9.4- and 9.1-fold, respectively. In RE/CuZnSOD and RE/CuZn- SOD þ CAT, CuZnSOD activity increased 7.7- and 8- fold, respectively. In RE/MnSOD, MnSOD activity Results increased 7.4-fold (Figure 4a, b and c). Overexpression Transduction of normal human keratinocytes with of antioxidant enzymes in different types of RE was lentiviral vectors leads to an increase in CAT-, clearly demonstrated by Western blot (Figure 4d). MnSOD- and CuZnSOD-specific activities and proteins Interestingly, CAT activity was more than 10-fold Among various gene transfer techniques for overexpres- higher in normal RE than in keratinocytes grown in sion of antioxidant enzymes in primary human epidermal monolayer culture (see Figures 1b and 3a). In fact, CAT keratinocytes, we used vectors based on the human activity in RE was similar to that found in biopsies from immunodeficiency virus because of the very high effi- human epidermis (353782 U/mg of protein for human ciency and sustained expression of transgenes by this epidermis compared to 3717100 U/mg for RE). This system.27–29 First of all, normal human keratinocytes in result was confirmed by Western blot (Figure 5). monolayer culture were transduced with the different vectors expressing EGFP (TPEW), CAT (TPCATW), Following UVB or UVA irradiation, CAT activity MnSOD (TPMnSODW) and CuZnSOD (TPCuZn- decreases in RE with overexpression of SOD enzymes SODW) (Figure 1a). The percentage of EGFP-positive To study the effects of UVB(180 or 220 mJ cm À2)or cells was determined by cytofluorimetry 5 days post- UVA (10 J cmÀ2) irradiation on antioxidant enzymes in transduction and was between 85 and 95% (Figure 1b), different RE, activity of CAT, CuZnSOD and MnSOD demonstrating a very high efficiency of this lentivector in was measured in different RE (Figure 4a, b and c). keratinocytes. This expression was maintained after 15 Twenty-four hours after UVBirradiation, CAT activity days (data not shown). All vectors transmitted an intact transiently increased 1.7- to 2.2-fold in epidermis recon- provirus, as shown by the appropriate-sized single structed with non-transduced cells (RE/NT) or trans- proviral bands obtained in a genomic Southern blot duced by an empty vector (RE/TPW) and was markedly analysis (Figure 1c). To verify that the genes transferred decreased in the epidermis overexpressing SOD by lentivectors were expressed and properly translated (Figure 4a). By contrast, CuZnSOD and MnSOD into active proteins, corresponding enzyme activities were activities did not show any significant changes in the

Cancer Gene Therapy Protective effect of catalase against UV HR Rezvani et al 178 ab260 30 240 25 220 20 200 15

30 3 20 2

Catalase activity (U/mg prot) 10 1 CuZnSOD activity (U/mg prot) 0 0 Ker / NT Ker / NT Ker / Ker / CAT Ker / Ker / CAT Ker / Ker / TPW Ker / Ker / TPW Ker / +CAT +CAT Ker / MnSOD Ker / Ker / MnSOD Ker / Ke / CuZnSOD Ke / Ker / CuZnSOD Ker / Ker / CuZnSOD Ker / CuZnSOD Ker /

cd 4 3.5 3 2.5 NT CAT CuZnSOD+CAT CuZnSOD MnSOD

0.4 / / / / /

0.3 Ker Ker Ker Ker Ker 0.2 Catalase

MnSOD activity (U/mg prot) 0.1 β-actin 0 MnSOD CuZnSOD Ker / NT Ker / Ker / CAT Ker / Ker / TPW Ker / +CAT Ker / MnSOD Ker / Ker / CuZnSOD Ker / CuZnSOD Ker / Figure 2 Antioxidant enzyme activities and protein levels in non-transduced and transduced keratinocytes. CAT (a), CuZnSOD (b) and MnSOD (c) activities were measured in cell lysates 5 days after transduction. Data are means7s.e.m. of five independent experiments. (d) Immunoblot analysis was performed to determine level of CAT, CuZnSOD and MnSOD proteins in non-transduced keratinocytes (Ker/NT) or keratinocytes transduced by CAT (Ker/CAT), CuZnSOD (Ker/CuZnSOD) or MnSOD (Ker/MnSOD) or co-transduced by CAT and CuZnSOD (Ker/ CuZnSOD þ CAT). b-Actin was used to confirm equivalent protein loading.

different irradiated epidermis compared to the corre- types of RE (Table 1). For the two different doses of UVB sponding non-irradiated (nIr) ones. (180 and 220 mJ cmÀ2), the number of SBC in RE/CAT Twenty-four hours after UVA exposure, there was no and RE/CuZnSOD þ CAT was lower (approximately difference in specific activity for MnSOD and CuZnSOD 40%) than the number found in RE/NT. A slight in all types of RE compared to the corresponding nIr RE. reduction (20%) in total SBC formation was observed By contrast, a marked decrease in CAT activity was found in RE/CuZnSOD irradiated at 180 mJ cmÀ2. The number in RE with overexpression of SOD after UVA exposure of SBC in RE/MnSOD was similar to that in controls (Figure 4a, b and c). (RE/NT and RE/TPW). We then calculated the percentage of SBC within each layer of epidermis (Table 1). SBC Overexpression of CAT decreases UVB-induced appeared in the basal layer and stratum spinosum for apoptosis in RE control epidermis and epidermis overexpressing both One of the major biological features of UVBis the SODs. For epidermis overexpressing CAT, there was a induction of apoptotic cell death of keratinocytes, which marked (75–80%) decrease in SBC in the basal layer, but appear as SBC within the epidermis. To investigate the surprisingly the same number of SBC remained in the effect of antioxidant enzyme overexpression on apoptosis, stratum spinosum. In all types of RE, a very low number formation of SBC following UVB exposure was detected of SBC were present in the stratum granulosum (data not by classical histology (Figure 6a) and quantified in all shown).

Cancer Gene Therapy Protective effect of catalase against UV HR Rezvani et al 179 a associated with decrease in DNA damage, total p53 and serine-15 phosphorylated p53 proteins were assessed (Figure 6c and d). Immunohistochemical analysis on RE 24 h after UVBirradiation demonstrated that accumulation of p53 was significantly reduced in epidermis overexpressing CAT. Western blot analysis revealed that RE / NT-nIr RE / TPW-nIr RE / MnSOD-nIr the level of serine-15 phosphorylated p53 was lower in RE/CAT than RE/NT. These data indicate that UVB- induced DNA damage and consequently p53 induction were reduced by CAT overexpression.

Overexpression of CAT decreases UVA-induced cellular RE / CAT-nIr RE / CuZnSOD+CAT-nIr RE / CuZnSOD-nIr hypertrophy and DNA oxidation in RE Because UVA irradiation constitutes over 90% of solar b UV at ground level, it may be the most important contributor to UV-induced carcinogenesis. Thus, we studied the effect of UVA on different RE. UVA irradiation induced hypertrophy (swelling of cell cytoplasm) and disorganization of basal keratinocytes in control RE and RE/SOD (compare Figures 3a and 7a). RE / NT RE / NT The level of hypertrophy was considerably reduced in RE/ CAT and RE/CuZnSOD þ CAT compared to other RE c (controls, RE/CuZnSOD and RE/MnSOD) with an appearance of normal, nIr epidermis. Furthermore, we observed a few SBC in all types of RE other than RE/ CAT, where these cells were absent. Analysis of caspase-3 activation was not possible because of the low content of apoptotic cells after UVA irradiation (Figure 6b). The RE / EGFP RE / EGFP cellular processes leading to UVA-induced DNA damage are mediated by ROS production leading to oxidative de DNA modifications such as formation of 8-oxodG. Analysis of DNA oxidation by an anti-8-oxodG antibody demonstrated that the number of positive cells in RE/ CAT and RE/CuZnSOD þ CAT was lower than that found in other types of RE (Figure 7b). Quantification of 8-oxodG-positive cells on each layer or the entire RE / NT RE / CAT epidermis demonstrated that the percentage of positive Figure 3 Histological examination and transduction efficiency cells in RE/CAT and RE/CuZnSOD þ CAT was lower analysis in RE. Epidermis reconstructed with non-transduced (approximately 80%) than found in RE/NT (Table 2). keratinocytes (RE/NT) or transduced by TPW vector (RE/TPW), EGFP (RE/EGFP), CAT (RE/CAT), CuZnSOD (RE/CuZnSOD) or MnSOD (RE/MnSOD) were fixed in paraformaldehyde, embedded in paraffin and sectioned in 5 mm. (a) Paraffin sections were stained Discussion with hematoxylin and eosin. RE/NT (b) and RE/EGFP (c) were used to evaluate EGFP expression and are shown under visible (left) and Being the first line of defense against oxidative damage fluorescent lamp (right). RE/NT (d) and RE/CAT (e) were immuno- induced by environmental factors such as UV irradiation, labeled using a CAT-specific antibody, followed by EnVision þ HRP, the skin is equipped by a well-established cellular revealed with DAB (diaminobenzidine) and counterstained with hemalun. CAT-positive cells appear in brown. Scale bar: 100 mm. antioxidant defense system. This system is composed of multiple low molecular weight components (such as a- tocopherol and ascorbic acid) and several enzymes (such as CAT and SOD). Owing to their mutual interactions To assess the global rate of apoptotic cells (SBC) in and their action in different cellular sites, the elucidation different RE, we looked at the activation of caspase-3 by of their roles is difficult. Our present study focused on the specific antibody staining. Immunoblot analysis revealed respective roles of CAT, MnSOD and CuZnSOD in the again that there was a clear decrease in activation of protection of RE against deleterious effect of UVA and caspase-3 in the epidermis overexpressing CAT compared UVBirradiation. to the controls (Figure 6b). This result was specific for Measurement of antioxidant enzyme activities upon CAT and was not found with the overexpression of the transduction with various lentivectors indicated that two forms of SOD. To investigate whether reduction in overexpression of CAT, CuZnSOD or MnSOD does not UVB-induced apoptosis upon CAT overexpression is adversely affect the activities of each other in nIr primary

Cancer Gene Therapy Protective effect of catalase against UV HR Rezvani et al 180 a b 4200 22 nIr 4000 21 nIr UVB180 UVB180 3800 20 UVB220 UVB220 3600 UVA 19 UVA 3400 18 * * * 800 * 4 600 3 400 2 ** * Catalase activity (U/mg prot) 200 * 1 * * CuZnSOD activity (U/mg prot) 0 0 RE / NT RE / NT RE / RE / CAT RE / CAT RE / RE / TPW RE / TPW RE / +CAT +CAT RE / MnSOD RE / MnSOD RE / RE / CuZnSOD RE / CuZnSOD RE / CuZnSOD RE / CuZnSOD RE /

cd3.8 3.6 nIr UVB180 3.4 UVB220 UVA 3.2

0.6 0.4 RE / NT-nIr RE / NT-UVB180 RE / NT-UVA RE / CAT-nIr RE / CAT-UVB180 RE / CAT-UVA RE / CuZnSOD-nIr RE / CuZnSOD-UVB180 RE / RE/CuZnSOD-UVA CuZnSOD+CAT-nIr RE / CuZnSOD+CAT-UVB RE / CuZnSOD+CAT-UVA RE / MnSOD-nIr RE / MnSOD-UVB180 RE / MnSOD-UVA RE / 0.2 Catalase MnSOD activity (U/mg prot) β 0 -actin

MnSOD RE / NT RE / RE / CAT RE / RE / TPW RE / CuZnSOD +CAT RE / MnSOD RE / RE / CuZnSOD RE / CuZnSOD RE / Figure 4 Antioxidant enzyme activities and protein levels in genetically modified RE 24 h after UVB (180 or 220 mJ cmÀ2) or UVA irradiation (10 J cmÀ2). CAT activity (a), CuZnSOD activity (b) and MnSOD activity (c) were measured in nIr and irradiated epidermis reconstructed by different transduced and non-transduced (NT) keratinocytes. The results are expressed as mean7s.e.m. of five independent experiments. The asterisk indicates significant difference (*Po0.05) from control (nIr) samples. (d) Western blot analysis for CAT, CuZnSOD and MnSOD was performed on different epidermis lysates.

overexpression of CuZnSOD increased activity of CAT

E RE KER but did not alter MnSOD activity, whereas overexpres- Catalase sion of MnSOD did not affect activities of CAT and CuZnSOD. However, it is difficult to compare CAT β-actin activities in keratinocytes, JB6 cells and SVHK. We found that CAT activity was lower in keratinocytes Figure 5 Comparison of the level of CAT in keratinocytes, RE and than in RE. This was probably due to a lower level of normal human epidermis (E). Western blot analysis for CAT was performed with lysates of cultured keratinocytes (Ker), RE and CAT in the basal layer (corresponding to keratinocytes in biopsy of human epidermis (E). culture) than in the more differentiated layers present in the RE (see Figure 2c). However, it is important to point out that we found similar levels of CAT in RE compared to normal epidermis (E). cultured keratinocytes or RE. Using mouse epidermal JB6 In the epidermal samples tested 24 h after UVBor UVA cells, Amstad et al.30 reported that overexpression of irradiation, CuZnSOD and MnSOD activities did not CuZnSOD had no effect on CAT activity or vice versa. change as compared to the corresponding nIr epidermis. The results of Takahashi et al.31 in simian virus 40- Several studies have shown that CuZnSOD and MnSOD transformed human keratinocytes (SVHK) showed that activities transiently change following UVBirradiation

Cancer Gene Therapy Protective effect of catalase against UV HR Rezvani et al 181 a

RE / NT-UVB RE / TPW-UVB RE / MnSOD-UVB

RE / CAT-UVB RE / CuZnSOD+CAT-UVB RE / CuZnSOD-UVB

b RE / NT-nIr RE / CAT-nIr RE / CuZnSOD-nIr RE / CuZnSOD+CAT-nIr RE / MnSOD-nIr RE / TPW-nIr RE / NT-UVB RE / CAT-UVB RE / CuZnSOD-UVB RE / CuZnSOD+CAT-UVB RE / MnSOD-UVB RE / TPW-UVB RE / NT-UVA RE / CAT-UVA RE / CuZnSOD-UVA RE / CuZnSOD+CAT-UVA RE / MnSOD-UVA RE / TPW-UVA RE /

procaspase3

cleaved caspase3

β-actin

c d

RE / NT-nIr RE / NT-UVB RE/NT-nIr RE/CAT-nIr RE/NT-UVB RE/CAT-UVB

ser-15-p53

β-actin

RE / CAT-nIr RE / CAT-UVB

Figure 6 Analysis of SBC formation, caspase-3 activation and p53 accumulation following UVB irradiation. (a) Histological examination and analysis of SBC formation in different RE was performed by hematoxylin and eosin staining 24 h after UVB (180 mJ cmÀ2) exposure. Scale bar: 100 mm. (b) Activation of caspase-3 in different RE 24 h after UVB (180 mJ cmÀ2) or UVA irradiation was assessed by immunobloting using an anti-caspase-3 antibody. The proenzyme (30 kDa) and the two active forms (19 and 17 kDa) are shown. (c) P53 accumulation was revealed using anti-p53-specific antibody in non-transduced or CAT-transduced epidermis (RE/NT and RE/CAT). Positive cells appeared in brown. Scale bar: 100 mm. (d) Western blot analysis of serine-15 phosphorylated p53 in RE/NT and RE/CAT was performed using a specific antibody.

and then gradually recover to baseline levels 24 h after finding, we hypothesize that an intense increase in H2O2 irradiation in keratinocyte cell lines22,32 and in murine levels following irradiation in epidermis overexpressing skin.16 Our results indicated that CAT activity was SOD leads to a downregulation of CAT or its inhibition transiently increased in irradiated control epidermis by oxidation. (non-transduced and mock-transduced cells) compared SBC are induced in the epidermis by UVB and are to nIr control ones and was highly decreased in the histologically considered as apoptotic cells. We observed epidermis overexpressing SOD. To explain the latter a reduced formation of SBC in RE overexpressing CAT

Cancer Gene Therapy Protective effect of catalase against UV HR Rezvani et al 182 Table 1 Percentage of SBC in different RE 24 h after UVB irradiation

n UVB (180 mJ cmÀ2) UVB (220 mJ cmÀ2)

Total epidermis Basal and Spinous layer Total epidermis Basal and Spinous layer suprabasal layer suprabasal layer

RE/NT 5 13.471.4 25.271.4 12.7711772.3 38.273.1 14.671.4 RE/TPW 5 13.771.3 26.171.9 11.871.2 17.171.9 38.872.4 14.370.86 RE/CAT 5 7.770.94** 4.270.4** 11.271.5 11.671.3** 8.671.1** 15.671.9 RE/CuZnSOD 5 1171.7* 2071.8 11.672.3 1671.6 36.373.5 2371.4 RE/CuZnSOD+CAT 5 7.870.81** 3.5370.34** 11.971.1 11.571.5** 8.871.7** 14.571.2 Abbreviations: CAT, catalase; CuZnSOD, copper/zinc superoxide dismutase; NT, non-transduced; RE, reconstructed epidermis; SBC, sunburn cell; UVB, ultraviolet B. To quantify the percentage of SBC, 10 Â 100 cells were counted on each layer or the entire epidermis. The results are expressed as mean7s.e.m. The percentage of SBC in non-irradiated RE was less than 0.5%. The statistical significance values were calculated using paired Student’s t-test against RE/NT (*Po0.05 and **Po0.005). n ¼ number of experiments.

RE / NT-UVA RE / TPW-UVA RE / MnSOD-UVA

RE / CAT-UVA RE / CuZnSOD+CAT-UVA RE / CuZnSOD-UVA

RE / NT-nIr RE / NT-UVA RE / TPW-UVA RE / MnSOD-UVA

RE / CAT-nIr RE / CAT-UVA RE / CuZnSOD+CAT-UVA RE / CuZnSOD-UVA

Figure 7 Histological examination and analysis of 8-oxodG formation following UVA irradiation. (a) Histological examination of different RE was performed by hematoxylin and eosin staining 24 h after UVA irradiation. Arrows point to SBC cells. (b) RE were fixed in paraformaldehyde 1 h after UVA irradiation, sectioned in 5 mm and immunostained with an anti-8-oxodG-specific antibody. Positive cells (arrows) appeared in red. Scale bar: 100 mm.

(40%) and CuZnSOD (18%) compared to control RE, strating a protective role in UVB-induced apoptosis in whereas overexpression of MnSOD had no protective SVHK overexpressing CuZnSOD but not MnSOD. We effect. Our previous results implicating protective effects confirmed our results by caspase-3 assay, which is a of CAT overexpression against UVB-induced apoptosis in validated and accurate method to measure the induction normal human keratinocytes are in agreement with these of apoptosis.33–34 The reduced activation of caspase-3 in observations.22 Furthermore, these findings are consistent RE/CAT and RE/CuZnSOD þ CAT confirms the pro- with observations reported by Takahashi et al.31 demon- tective role of CAT against UVB-induced apoptosis.

Cancer Gene Therapy Protective effect of catalase against UV HR Rezvani et al 183 Table 2 Percentage of 8-oxodG-positive cells in RE and XP-RE 1 h after UVA irradiation

n UVA

Total epidermis Basal and suprabasal layer Spinous layer

RE/NT 5 4671.3 4071.6 50.772 RE/TPW 5 4371.1 4071.9 4572.1 RE/CAT 5 870.86* 970.5* 7.270.75* RE/CuZnSOD 5 3871.5 3771.6 39.671.9 RE/CuZnSOD+CAT 5 5.770.67* 3.470.74* 6.870.9* RE/MnSOD 5 44.371.8 42.471.6 48.271.98 Abbreviations: CAT, catalase; CuZnSOD, copper/zinc superoxide dismutase; 8-oxodG, 8-oxodeoxyguanosine; NT, non-transduced; MnSOD, manganese superoxide dismutase; RE, reconstructed epidermis; UVA, ultraviolet A. To quantify the percentage of 8-oxodG-positve cells, 10 Â 100 cells were counted on each layer or the entire epidermis. The results are expressed as mean7s.e.m. The percentage of 8-oxodG-positive cells in non-irradiated RE was less than 0.5%. The statistical significance values were calculated using paired Student’s t-test against RE/NT (*Po0.005). n ¼ number of experiments.

Furthermore, our results showed that overexpression of before irradiation does not reduce the level of UVB- CAT protects 80% of basal and supra-basal keratinocytes induced CPD. As CPD is overall a non-oxidative product from UVB-induced apoptosis. Recently, Zuliani et al.35 of UV upon DNA, it was not unexpected that CAT demonstrated a higher sensitivity of undifferentiated basal overexpression had no effect on it. However, the cells to H2O2-induced cell death compared to differen- significant effect of CAT on reduction of UVB-induced tiated cells. The lower CAT protein levels in basal cells apoptosis without affecting the major type of UVB- compared to more differentiated cells (Figure 2c) could induced DNA photoproduct, that is, CPD formation, is explain this sensitivity to UVB-induced ROS production. surprising. We have recently demonstrated that UVB Apoptosis is the major mechanism preserving the induces ROS production at two distinct chronologic cellular homeostasis and thus preventing skin carcinogen- stages and that the second increase in ROS levels, which esis. To determine whether reduction of UVB-induced persists for some hours, is completely prevented in apoptosis by CAT overexpression allows keratinocytes keratinocytes overexpressing CAT.22 Thus, the protective with DNA damage to survive and undergo subsequently effect of CAT overexpression may be due to the malignant transformation, we compared p53 accumula- mitigation of DNA damage caused by the late increase tion in RE. The tumor suppressor protein p53, which in ROS levels. It is also likely that CAT affects the cellular functions as a latent, short-lived transcription factor,36 redox status, leading to a modulation of gene expression accumulates following various stresses including DNA- through the activity of various transcription factors such damaging agents, oxidative stress and UV irradiation.37–39 as nuclear factor-kB, Fos, Jun, Myb48 and p53.49 p53 transcriptionally activates genes encoding proteins Consistent with this hypothesis, it has been reported that involved in cell cycle arrest, DNA repair and/or UVB-induced ROS are involved in transcriptional activa- apoptosis, and represses transcriptional activation of tion of AP-1,50 increased production of matrix metallo- growth-promoting genes.40–41 Our results showed that proteases51 and also biosynthesis and activation of CAT overexpression reduced the accumulation of p53. transforming growth factor-b.52 Stabilization and functional activation of p53 occurs in We demonstrated that UVA-induced hypertrophy was response to DNA damage via a complex mechanism. One decreased by the overexpression of CAT (but not by the of these mechanisms is protein phosphorylation at N- and overexpression of CuZnSOD or MnSOD). This abroga- C-terminal residues, especially at serine-15 by which tion of UVA-induced hypertrophy may be due to an interaction of p53 with its negative regulator, MDM2, is inhibition of the UV-induced increase in ornithine abolished.37–42 Western blot analysis of serine-15 phos- decarboxylase (ODC) already shown in human53 and in phorylated p53 showed that overexpression of CAT mouse epidermis.54 Both keratinocyte hypertrophy and downregulated this protein in normal RE. The interaction the increase in ODC induced by 12-O-tetradecanoylphor- between UV irradiation and DNA produces a number of bol-13-acetate (TPA) (enhancer of H202 formation) or oxidative43,44 and non-oxidative lesions.1,45,46 Among TPA þ UVA were prevented by the ROS inhibitor 10- various types of DNA damage, cyclobutane pyrimidine acetoxychavicol acetate55 or by a vegetal antioxidant.56 It dimers (CPD) or 6–4 photoproducts are believed to be the is well known that UVA does not interact directly with major type of damage resulting from UVBexposure. DNA, but instead appears to contribute to DNA damage Studies of UVB-induced CPD in our system indicated through ROS, resulting in oxidative DNA modifications, that overexpression of CAT did not affect CPD formation including the formation of 8-oxodG, protein–DNA cross- in normal epidermis (data not shown). In the same way, linking, base loss and strand breaks.1,46 Decrease in UVA- Kulms et al.47 reported that addition of the radical induced DNA oxidation (8-oxodG) in RE overexpressing scavenger pyrrolidene-dithiocarbamate on HeLa cells CAT is consistent with this notion.

Cancer Gene Therapy Protective effect of catalase against UV HR Rezvani et al 184 Comparing the effects of UVA and UVBon RE in our References system indicated that induction of caspase-3 activation and formation of SBC were markedly lower following 1 de Gruijl FR, van Kranen HJ, Mullenders LH. UV-induced UVA irradiation as compared with UVBirradiation DNA damage, repair, mutations and oncogenic pathways in (Figure 5). Furthermore, accumulation of p53 occurred skin cancer. J Photochem Photobiol B Biol 2001; 63: 19–27. following UVB, but not UVA, exposure (data not 2 Boukamp P. Non-melanoma skin cancer: what drives tumor shown). These findings are consistent with the existence development and progression? Carcinogenesis 2005; 26: of distinct responses to UVA and UVBirradiation in 1657–1667 (Review). 3 Djavaheri-Mergny M, Mergny JL, Bertrand F, Santus R, keratinocytes.57-59 Maziere C, Dubertret L et al. Ultraviolet-A induces In summary, this study establishes that CAT gene activation of AP-1 in cultured human keratinocytes. FEBS transfer can prevent a sizable part of the deleterious Lett 1996; 384: 92–96. effects of both UVA and UVBirradiation on normal skin 4 Tyrrell R. Redox regulation and oxidant activation of heme by reduction of UV-induced DNA damage as well as oxygenase-1. Free Radic Res 1999; 31: 335–340. regulation of gene expression or cascade activation via 5 Fuchs J, Packer L. Ultraviolet irradiation and the skin cellular redox modulation. To confirm the effectiveness of antioxidant system. Photodermatol Photoimmunol Photomed CAT overexpression in reduction of UV-induced apop- 1990; 7: 90–92. tosis as well as the risk of photocarcinogenesis, the study 6 Danno K, Horio T. Sunburn cell: factors involved in its of xenografted reconstructs on tolerant mice60 exposed to formation. Photochem Photobiol 1987; 45: 683–690. chronic irradiation as well as clinical trials using topical 7 Bayerl C, Taak S, Moll I, Jung EG. Characterization of Photoder- delivery systems at the protein level are necessary. sunburn cells after exposure to ultraviolet light. matol Photoimmunol Photomed 1995; 11: 149–154. 8 Schwarz A, Bhardwaj R, Aragane Y, Mahnke K, Riemann Conclusion and perspective H, Metze D et al. Ultraviolet-B-induced apoptosis of Non-melanoma skin cancers are among the most pre- keratinocytes: evidence for partial involvement of tumor valent cancers in humans. Owing to their rising necrosis factor-alpha in the formation in the formation of incidences, it is necessary to develop approaches to sunburn cells. J Invest Dermatol 1995; 104: 922–927. prevent and treat these cancers. One strategy can be the 9 Aragane Y, Kulms D, Metze D, Wilkes G, Po¨ ppelmann B, reinforcement of antioxidant enzymes especially CAT. Luger TA et al. Ultraviolet light induces apoptosis via direct Production of ROS following UVA or UVBirradia- activation of CD95 (Fas/APO-1) independently of its ligand tion,10,11 impairment of the natural antioxidant defense CD95L. J Cell Biol 1998; 140: 171–182. upon UVA or UVBexposure, 5 participation of free 10 Masaki H, Atsumi T, Sakurai H. Detection of hydrogen radicals in all steps of carcinogenesis,61 a severe dis- peroxide and hydroxyl radicals in murine skin fibroblasts turbance of antioxidant balance with diminished anti- under UVBirradiation. Biochem Biophys Res Commun 1995; 17: 474–479. oxidant enzymes in non-melanoma (BCC, SCC) tumors,23,62 11 Peus D, Vasa RA, Beyerle A, Meves A, Krautmacher C, correlation between deficiency in CAT activity with the Pittelkow MR. H O is an important mediator of UVB- appearance and intensity of tumor phenotype expression 2 2 63 induced EGF-receptor phosphorylation in cultured kerati- in human keratinocytes and inhibition of carcinogenesis nocytes. J Invest Dermatol 1999; 112: 751–756. 64 with supplementation of antioxidants support this 12 Murphy G, Young AR, Wulf HC, Kulms D, Schwarz T. The hypothesis. This approach can be accomplished using molecular determinations of sunburn cell formation. Exp topical delivery systems at the protein or gene level. Dermatol 2001; 10: 155–160. This strategy might also be suitable to test in severe 13 Mate´ s JM. Effects of antioxidant enzymes in the molecular photosensitive skin disorders such as xeroderma pigmen- control of reactive oxygen species toxicology. Toxicol 2000; tosum, a human cancer-prone disease with a recessive 153: 83–104. gene defect in nucleotide excision repair in which a 14 Vile GF, Tyrrell RM. Oxidative stress resulting from reduction of CAT activity has been reported.65 ultraviolet A irradiation of human skin fibroblasts leads to a heme oxygenase-dependent increase in ferritin. J Biol Chem 1993; 268: 14678–14681. 15 Punnonen K, Puntala A, Ahtupa M. Effects of ultaviolet A Abbreviations and Birradiation on lipid peroxidation and activity of the antioxidant enzymes in keratinocytes in culture. Photoder- CAT, catalase; CPD, cyclobutane pyrimidine dimer; matol Photoimmunol Photomed 1991; 8: 3–6. CuZnSOD, copper/zinc superoxide dismutase; E, epider- 16 Shindo Y, Witt E, Han D, Tzeng B, Aziz T, Nguyen L et al. mis; MnSOD, manganese superoxide dismutase; NT, Recovery of antioxidants and reduction in lipid hydroper- non-transduced; nIr, non-irradiated; ODC, ornithine oxides in murine epidermis and dermis after acute ultraviolet decarboxylase; 8-oxodG, 8-oxodeoxyguanosine; RE, re- radiation exposure. Photodermatol Photoimmunol Photomed constructed epidermis; SBC, sunburn cell; SVHK, SV40- 1994; 10: 183–191. transformed human keratinocytes. 17 Zhang X, Rosenstein BS, Wang Y, Lebwohl M, Wei H. Identification of possible reactive oxygen species involved in ultraviolet radiation-induced oxidative DNA damage. Free Acknowledgements Radic Biol Med 1997; 23: 980–985. 18 Pelle E, Hung X, Mammone T, Marenus K, Maes D, We thank M Longy (Institut Bergonie´ , France) for Frenkek K. Ultraviolet-B-induced oxidative DNA base providing anti-p53 antibody. damage in primary normal human epidermal keratinocytes

Cancer Gene Therapy Protective effect of catalase against UV HR Rezvani et al 185 and inhibition by a hydroxyl radical scavenger. J Invest 34 Bedner E, Smolewski P, Amstad P, Darzynkiewicz Z. Dermatol 2003; 121: 177–183. Activation of caspases measured in situ by binding of 19 Maalouf S, El-Sabban M, Drawiche N, Gali-Muhtasib H. fluorochrome-labeled inhibitors of caspases (FLICA): corre- Protective effect of vitamin E on ultraviolet Blight- lation with DNA fragmentation. Exp Cell Res 2000; 259: induced damage in keratinocytes. Mol Carcinogen 2000; 34: 308–313. 121–130. 35 Zuliani T, Denis V, Noblesse E, Schnebert S, Andre´ P, 20 Morley N, Curnow A, Salter L, Campbell S, Gould D. N- Dumas M et al. Hydrogen peroxide-induced cell death in acetyl-L-cysteine prevents DNA damage induced by UVA, normal human keratinocytes is differentiation dependent. UVBand visible radiation in human fibroblasts. J Photo- Free Radic Biol Med 2005; 38: 307–316. chem Photobiol B Biol 2003; 72: 55–60. 36 Harris CC. p53: at the crossroads of molecular carcinogen- 21 Cario-Andre M, Briganti S, Picardo M, Nikaido O, Gall Y, esis and risk assessment. Science 1993; 262: 1980–1981. Ginestar J et al. Epidermal reconstructs: a new tool to study 37 Shieh SY, Ikeda M, Taya Y, Prives C. DNA damage- topical and systemic photoprotective molecules. J Photochem induced phosphorylation of p53 alleviates inhibition by Photobiol B Biol 2002; 68: 79–87. MDM2. Cell 1997; 91: 325–334. 22 Rezvani HR, Mazurier F, Cario-Andre´ M, Pain C, Ged C, 38 Achanta G, Huang P. Role of p53 in sensing oxidative DNA Taı¨ eb A et al. Protective effects of catalase overexpression on damage in response to reactive oxygen species-generating UVB-induced apoptosis in normal human keratinocytes. agents. Cancer Res 2004; 64: 6233–6239. J Biol Chem 281: 17999–18007. 39 Giaccia AJ, Kastan MB. The complexity of p53 modulation: 23 Sander CS, Hamm F, Elsner P, Thiele JJ. Oxidative stress in emerging patterns from divergent signals. Genes Dev 1998; malignant melanoma and non-melanoma skin cancer. Br J 12: 2973–2983. Dermatol 2003; 148: 913–922. 40 Ko LJ, Prives C. p53: puzzle and paradigm. Genes Dev 1996; 24 Prunieras M, Re´ gnier M, Schlotterer M. Nouveau proce´ de´ 10: 1054–1072. de culture des cellules e´ pidermiques humaines sur derme 41 Levine AJ. p53, the cellular gatekeeper for growth and homologue ou he´ te´ rologue: pre´ paration de greffons recom- division. Cell 1997; 88: 323–331. bines. Ann Chir Plast 1979; 24: 357–362. 42 Steegenga WT, vandereb AJ, Jochemsen AG. How phos- 25 Cudre-Mauroux C, Occhiodoro T, Konig S, Salmon P, phorylation regulates the activity of p53. J Mol Biol 1996; Bernheim L, Trono D. Lentivector-mediated transfer of 263: 103–113. Bmi-1 and telomerase in muscle satellite cells yields a 43 Mitchell DL, Jen J, Cleaver JE. Relative induction of Duchenne myoblast cell line with long-term genotypic and cyclobutane dimers and cytosine photohydrates in DNA phenotypic stability. Hum Gene Ther 2003; 14: 1525–1533. irradiated in vitro and in vivo with ultraviolet-C and 26 Ge´ ronimi F, Richard E, Redonnet-Vernhet I, Lamrissi- ultraviolet-Blight. Photochem Photobiol 1991; 54: 741. Garcia I, Lalanne M, Ged C et al. Highly efficient lentiviral 44 Doetsch PW, Zasatawny TH, Martin AM, Dizdaroglu M. gene transfer in CD34+ and CD34+/CD38À/linÀ cells from Monomeric base damage products from adenine, guanine, mobilized peripheral blood after cytokine prestimulation. and thymine induced by exposure of DNA to ultraviolet Stem Cells 2003; 21: 472–480. radiation. Biochemistry 1995; 34: 737. 27 Chen M, Li W, Fan J, Kasahara N, Woodley D. An efficient 45 Cadet J, Anselmino C, Douki T, Voituriez L. Photochem- gene transduction system for studying gene function in istry of nucleic acids in cells. J Photochem Photobiol B Biol primary human dermal fibroblasts and epidermal keratino- 1992; 15: 277. cytes. Clin Exp Dermatol 2003; 28: 193–199. 46 Ravanat JL, Douki T, Cadet J. Direct and indirect effects of 28 Serrano F, Del Rio M, Larcher F, Garcia M, Munoz E, UV radiation on DNA and its components. J Photochem Escamez MJ et al. A comparison of targeting performance of Photobiol B Biol 2001; 63: 88–102 (Review). oncoretroviral versus lentiviral vectors on human keratino- 47 Kulms D, Zeise E, Poppelmann B, Schwarz T. DNA cytes. Hum Gene Ther 2003; 14: 1579–1585. damage, death receptor activation and reactive oxygen 29 Ghazizadeh S, Katz AB, Harrington R, Taichman LB. species contribute to ultraviolet radiation-induced apoptosis Lentivirus-mediated gene transfer to human epidermis. in an essential and independent way. Oncogene 2002; 21: J Investig Dermatol Symp Proc 2004; 9: 269–275. 5844–5851. 30 Amstad P, Peskin A, Shah G, Mirault ME, Moret R, 48 Xanthoudakis S, Curran T. Identification and characteriza- Zbinden I et al. The balance between Cu,Zn-superoxide tion of Ref-1, a nuclear protein that facilitates AP-1 DNA- dismutase and catalase affects the sensitivity of mouse binding activity. EMBO J 1992; 11: 653–665. epidermal cells to oxidative stress. Biochemistry 1991; 30: 49 Jayaraman L, Murthy KG, Zhu C, Curran T, Xanthoudakis 9305–9313. S, Prives C. Identification of redox/repair protein Ref-1 as a 31 Takahashi H, Hashimoto Y, Aoki N, Kinouchi M, Ishida- potent activator of p53. Genes Dev 1997; 11: 558–570. Yamamoto A, Iizuka H. Copper, zinc-superoxide dismutase 50 Garmyn M, Degreef H. Suppression of UVB-induced protects from ultraviolet B-induced apoptosis of SV40- c-fos and c-jun expression in human keratinocytes by transformed human keratinocytes: the protection is N-acetylcysteine. J Photochem Photobiol B Biol 1997; 37: associated with the increased levels of antioxidant enzymes. 125–130. J Dermatol Sci 2000; 23: 12–21. 51 Brenneisen P, Wenk J, Klotz LO, Wlaschek M, Briviba K, 32 Sasaki H, Akamatsu H, Horio T. Effects of a single exposure Krieg T et al. Central role of ferrous/ferric iron in the to UVBradiation on the activities and protein levels of ultraviolet Birradiation-mediated signaling pathway leading copper–zinc and manganese superoxide dismutase in cul- to increased interstitial collagenase (matrix-degrading me- tured human keratinocytes. Photochem Photobiol 1997; 65: talloprotease (MMP)-1) and stromelysin-1 (MMP-3) mRNA 707–713. levels in cultured human dermal fibroblasts. J Biol Chem 33 Hurwitz SA, Spadau DF. Quantitative analysis of UVB- 1998; 273: 5279–5287. induced apoptosis in human epidermis. Exp Dermatol 2000; 52 Wang H, Kochevar IE. Involvement of UVB-induced 9: 185–191. reactive oxygen species in TGF-beta biosynthesis and

Cancer Gene Therapy Protective effect of catalase against UV HR Rezvani et al 186 activation in keratinocytes. Free Radic Biol Med 2005; 38: keratinocytes exhibit specialized UV-responsive signal trans- 890–897. duction pathway. DNA Cell Biol 2003; 22: 665–677. 53 Lowe N, Koo L, Breeding J. Ultraviolet light induces human 59 Bode AM, Dong Z. Mitogen-activated protein kinase epidermal ornithine decarboxylase activity. Br J Dermatol activation in UV-induced signal transduction. Sci STKE 1980; 103: 18. 2003; 2003: RE2 (Review). 54 Berton T, Mitchell DL, Fischer SM, Locniskar MF. 60 Casoli V, Cario-Andre M, Costet P, Pain C, Taieb A. Epidermal proliferation but not the quantity of DNA Comparison of long-term survival of pigmented epidermal photodamage is correlated with UV-induced mouse skin reconstructs cultured in vitro vs xenografted on nude mice. carcinogenesis. J Invest Dermatol 1997; 109: 340–347. Pigment Cell Res 2004; 17: 87–92. 55 Nakamura Y, Murakami A, Ohto Y, Torikai K, Tanaka T, 61 Sander CS, Chang H, Hamm F, Elsner P, Thiele JJ. Role of Ohigashi H. Suppression of tumor promoter-induced oxida- oxidative stress and the antioxidant network in cutaneous tive stress and inflammatory responses in mouse skin by a carcinogenesis. Int J Dermatol 2004; 43: 326–335 (Review). superoxide generation inhibitor l0-acetoxychavicol acetate. 62 Kobayashi T, Matsumoto M, Iizuka H, Suzuki K, Taniguchi Cancer Res 1998; 58: 4832–4839. N. Superoxide dismutase in psoriasis, squamous cell 56 Ishizaki C, Oguro T, Yoshida T, Wen CQ, Sueki H, carcinoma and basal cell epithelioma: an immunohistochem- Iijima M. Enhancing effect of ultraviolet A on ornithine ical study. Br J Dermatol 1991; 124: 555–559. decarboxylase induction and dermatitis evoked by 63 Rabilloud T, Asselineau D, Miquel C, Calvayrac R, Darmon 12-o-tetradecanoylphorbol-13-acetate and its inhibition M, Vuillaume M. Deficiency in catalase activity correlates by curumin in mouse skin. Dermatology 1996; 193: with the appearance of tumor phenotype in human 311–317. keratinocytes. Int J Cancer 1990; 45: 952–956. 57 Schieke SM, Ruwiedel K, Gers-Barlag H, Grether-Beck S, 64 F S, Afaq F, Mukhtar H. Photochemoprevention of skin Krutmann J. Molecular crosstalk of the ultraviolet a and cancer by botanical agents. Photodermatol Photoimmunol ultraviolet Bsignaling responses at the level of mitogen- Photomed 2003; 19: 56–72 (Review). activated protein kinases. J Invest Dermatol 2005; 124: 65 Vuillaume M, Calvayrac R, Best-Belpomme M, Tarroux P, 857–859. Hubert M, Decroix Y et al. Deficiency in the catalase activity 58 Adachi M, Gazel A, Pintucci G, Shuck A, Shifteh S, of xeroderma pigmentosum cell and simian virus 4-trans- Ginsburg D et al. Specificity in stress response: epidermal formed human cell extracts. Cancer Res 1986; 46: 538–544.

Cancer Gene Therapy