Protective Effects of Catalase Overexpression on UVB-induced Apoptosis in Normal Human Keratinocytes*

UV-induced apoptosis in keratinocytes is a highly complex process in which various molecular pathways are involved. These include the extrinsic pathway via triggering of death receptors and the intrinsic pathway via DNA damage and reactive oxygen species (ROS) formation. In this study we investigated the effect of catalase and CuZn-superoxide dismutase (SOD) overexpression on apoptosis induced by UVB exposure at room temperature or 4 °C on normal human keratinocytes. Irradiation at low temperature reduced UV-induced apoptosis by 40% in normal keratinocytes independently of any change in p53 and with a decrease in caspase-8 activation. Catalase overexpression decreased apoptosis by 40% with a reduction of caspase-9 activation accompanied by a decrease in p53. Keeping cells at low temperature and catalase overexpression had additive effects. CuZn-SOD overexpression had no significant effect on UVB-induced apoptosis. UVB induced an increase in ROS levels at two distinct stages: immediately following irradiation and around 3 h after irradiation. Catalase overexpression inhibited only the late increase in ROS levels. We conclude that catalase overexpression has a protective role against UVB irradiation by preventing DNA damage mediated by the late ROS increase.

One of the major effects of ultraviolet irradiation, especially in the UVB wavelength range, is the induction of apoptosis, which is characterized by morphological changes such as chromatin condensation, fragmentation of the nucleus, contraction of cytoplasmic volume, and emission of apoptotic bodies containing intact organelles.
There are two apoptotic pathways by which caspase activation is triggered. The extrinsic pathway is triggered by the trimerization of cell membrane death receptors followed by the formation of the deathinducing signaling complex (DISC) 2 through recruitment of the adapter molecule FADD (Fas-associated death domain) and procaspase-8. Activation of this initiator caspase by an autocatalytic cleavage in DISC promotes the subsequent procaspase activation cascade (1)(2)(3). The intrinsic pathway is triggered by the release of cytochrome c from mitochondria into the cytoplasm following death signals. Procaspase-9 activation occurs in the apoptosome, which is formed by association of cytosolic cytochrome c, Apaf-1, dATP, and procaspase-9 (4). Activation of these initiator caspases (caspase-8 and -9) via either pathway activates a group of effector caspases such as caspase-3, which cleave a variety of cellular death substrates (such as protein kinase C␦), leading finally to apoptosis (3,(5)(6).
UVB-induced apoptosis is a highly complex process involving the extrinsic and intrinsic pathways, but it is unclear how these pathways are interrelated. Indeed, it has been demonstrated that UVB is able to activate directly cell surface receptors such as CD95/Fas by inducing receptor clustering, without the need for specific ligands (7)(8)(9)(10). Reduction of UVB-induced apoptosis through inhibition of receptor clustering by keeping cells at low temperature (4 -10°C) during UVB irradiation or prevention of DISC formation by expressing a dominant negative FADD supports this notion (11,12). Correlation of apoptosis levels with DNA damage severity and reduction of apoptosis by enhancement of DNA repair enzymes (13,14) substantiates the idea that DNA damage induced by UVB irradiation activates the apoptotic pathway (6,15). While UVB is known to be an inducer of ROS formation (16,17) leading to apoptosis (18), the source of this ROS production is not precisely known. It has been shown that UVB can induce the production of superoxide anion radical ( (19). Several studies have demonstrated that an increase in antioxidant defense systems can reduce the deleterious effects of UV-induced ROS. The protective effects of non-enzymatic antioxidants or radical scavengers on UVB-induced damage have been reported (20 -22). Furthermore, a reduction in UVB-induced apoptosis in SV40-transformed human keratinocytes overexpressing CuZn-SOD has been documented (23). Taken together, the evidence to date suggests that DNA damage, death receptor activation, and ROS formation all contribute to UVBinduced apoptosis in different ways (24). Despite rapid progress in understanding apoptosis, the role of ROS production in cell death and its relationships with other apoptotic events are still unclear. In the present study using UVB irradiation of normal human keratinocytes sustainably overexpressing the lentivirus-mediated catalase and/or CuZn-SOD enzymes at room temperature or at 4°C, we further investigated the role of UVB-induced ROS in keratinocyte apoptosis. A, vectors carry an internal cassette for the EGFP, catalase, CuZn-SOD, or Mn-SOD driven by the promoter of human phosphoglycerate kinase gene (PGK). ⌬U3, R, and U5 are the 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, FACS analysis demonstrates the percentages of keratinocytes expressing EGFP, 5 days after transduction by the lentiviral vector TPEW (ϳ90%). C, immunoblot analysis was performed to determine level of catalase and CuZn-SOD proteins in normal non-transduced keratinocytes (Ker/NT ), keratinocytes transduced by catalase (Ker/CAT ), or CuZn-SOD (Ker/CuZn-SOD) or co-transduced by catalase and CuZn-SOD (Ker/CuZnSODϩCAT). ␤-Actin was used to confirm equivalent protein loading. D, catalase and CuZn-SOD activities were measured in cell lysates 5 days after transduction. Data are mean Ϯ S.D. of five independent experiments.

EXPERIMENTAL PROCEDURES
Source of Keratinocytes and Irradiation Procedure-Keratinocytes were isolated from normal human skin in patients undergoing plastic surgery. Briefly, fresh skin fragments were immediately cut into 5 ϫ 5-mm pieces and treated with trypsin-EDTA for 3 h at 37°C or overnight at 4°C to separate epidermis from dermis. Keratinocytes were seeded at a concentration of 10 5 cells per cm 2 in MCDB153 medium, which included hydrocortisone (0.5 g/ml), epidermal growth factor (10 ng/ml), insulin (5 g/ml), and bovine pituitary extract (70 g/ml) (all from Sigma). The medium was changed three times a week. When the cultures reached 70 -80% confluence, the cells were detached with 10% trypsin-EDTA and then resuspended in MCDB153 medium.
UVB irradiation was delivered with a Biotronic device (Vilber Lourmat, Marne la Vallée, France) equipped with a dosimeter in which the UVB lamp emitted a continuous spectrum between 280 and 380 nm with a major peak at 312 nm, as already described (25). For irradiation at 4°C, the cells grown in monolayers were kept at 4°C for 30 min, rinsed in cold PBS, and irradiated in a cold tank. For irradiation at room temperature, cells were rinsed in PBS at 37°C and irradiated at room temperature. Control cells were subjected to the identical procedure but without exposure to UV.
Viability Testing-Keratinocyte survival rate following irradiation was measured by the trypan blue exclusion test. Cells were grown to 70% confluence and irradiated as described above. At different times after irradiation, cells were detached with 10% trypsin-EDTA and then resuspended in PBS. The resuspended cells were diluted 1:1 in trypan blue solution. Cell viability was expressed as the percentage trypan blue excluded cells versus total cells.
Construction and Production of Lentiviral Vectors-The pRRL.SIN. PPT.PGK.MCS.WPRE and the pRRL.SIN.PPT.PGK.EGFP.WPRE vectors, hereafter termed TPW and TPEW, were a generous gift from D. Trono (Université de Genève, Genève, Switzerland) (26). The different vectors (TPCATW and TPCuZnSODW, see Fig. 1A) were constructed by inserting catalase (CAT) and CuZn-SOD cDNAs into the multiple cloning site of the TPW vector. Expression of EGFP, CAT, and CuZn-SOD was driven by the human phosphoglycerate kinase promoter. Lentiviral particles were produced by transient transfection of 293T cells as described previously (27). The following plasmids were used: a packaging plasmid (pCMV⌬8.91), a vesicular somatitis virus glycoprotein (VSV-G) envelope plasmid pMD.G, and a transducing vector (TPW, TPEW, TPCATW, TPCuZnSODW). Using a calcium phosphate transfection technique, 10 g of pCMV⌬8.91, 4 g of pMD.G, and 10 g of transducing vector were used per 10-cm diameter Petri dish. After 48 h, the viral supernatant was harvested, filtered through a 0.22-m filter, centrifuged at 35,000 ϫ g for 4 h, and frozen in aliquots.
To determine the titer of each viral supernatant, serial dilutions were used for transduction of 293T cells. Following transduction, EGFP expression was measured directly by cytofluorimetric analysis (FACS Calibur, BD Biosciences). Typical titers after concentration for lentiviral vectors were between 2 ϫ 10 8 and 2 ϫ 10 9 infectious particles/ml. Enzyme-linked immunosorbent assays of p24 were used to determine the concentration of viral p24 protein in the different viral supernatants. Titration was done by comparing the p24 concentration of the different supernatants and the specific titers of EGFP.
Transduction of Keratinocytes-5 ϫ 10 5 cells were plated in T25 flasks and incubated for 24 h in complete medium. Prior to infection, medium was removed and cells were infected with viral supernatants for 24 h at 37°C in the presence of 8 g/ml protamine sulfate. After 5 days, the percentage of EGFP-positive cells was analyzed by cytofluorimetry.
Determination of Catalase and SOD Activities-The cells were sonicated in 0.1 M Tris-Cl (pH 7.5) for two 30-s bursts. After 10 min of centrifugation at 12,000 ϫ g, aliquots of the obtained supernatant were stored at Ϫ80°C. Total protein concentration was measured by the BCA kit reagent (Pierce, Bezons, France). SOD was assayed by using the SOD Assay Kit-WST (Dojindo Molecular Technologies, Gaithersburg, MD). Catalase was measured by Amplex Red Catalase Assay Kit (Molecular Probes, Invitrogen, Cergy-Pontoise, France). Standard curves for the enzymatic activities of SOD and catalase were drawn by using purified enzyme preparations. Enzyme-specific activities were expressed as units/mg of protein. One unit of catalase activity was defined as 1 mol of H 2 O 2 consumed per min. One unit of SOD activity was defined as the amount of enzyme that inhibits 50% of the WST-1 formazan per minute.
Apoptosis, Necrosis, and DNA Fragmentation Analysis-For apoptosis and necrosis measurements, keratinocytes were incubated for 16 h after irradiation with 5 mol/liter FITC-VAD-FMK (Promega), which binds to activated caspases, for 30 min in the dark at 37°C. To assess plasma membrane permeability, 2.5 g/ml propidium iodide (PI, Sigma) was added, and cells were immediately analyzed by flow cytometry.
For DNA fragmentation analysis, cells were resuspended in hypotonic staining buffer (50 g/ml PI, 100 mg/ml RNase, 3.8 mM sodium citrate, and 0.3% Triton X-100) in the dark at 4°C for 30 min and then analyzed by flow cytometry. Cells with DNA content less than the G 1 amount of untreated cells were considered apoptotic.
Quantification of Proteins by Western Blotting-To determine activation of caspase-3, -8, and -9, and accumulation of total p53 and serine 15-phosphorylated p53, cells were detached with trypsin-EDTA and resuspended in lysis buffer (25 mM HEPES (pH 7.4), 150 mM NaCl, Triton X-100, 1% v/v protease inhibitor mixture; Sigma) at 4°C for 30 min. After 10 min of centrifugation at 12,000 ϫ g, aliquots of the supernatant were stored at Ϫ80°C. Protein concentration was measured by the BCA reagent kit (Pierce). Equal amounts of total protein were resolved by SDS-polyacrylamide gel electrophoresis (12% gel) and electrophoretically transferred to polyvinylidene difluoride membrane at 1.2 V/cm 2 for 1 h 20 min. Membranes were blocked with 5% nonfat dry milk in TBS with 0.1% Tween 20 (TBS-T) for 1 h at room temperature on a shaker, then were incubated at 4°C overnight with 1:1000 dilution of the anti-caspase-3, -8, and -9 monoclonal antibodies (Cell Signaling Technology, Tebu-Bio, Le Perray en Yvelines, France) or anti-human p53 and serine 15-phosphorylated p53 (Cell Signaling Technology) or anti-catalase (Rockland, Tebu-Bio) or anti-CuZn-SOD (Santa Cruz Biotechnology, Tebu-Bio). The blots were washed with TBS-T (3 ϫ 10 min) and incubated at room temperature with 1:10,000 dilution of anti-immunoglobin horseradish peroxidase-linked antibody (Vector Laboratories, Biovalley S.A, Marne la Vallée, France) for 1 h. The chemiluminescence ECL reagent was used to develop the blots after washing the membrane three times for 10 min with TBS-T and twice with TBS alone.
Measurement of Intracellular ROS-The intracellular production of ROS was assayed using the CM-H 2 DCF-DA probe (Molecular Probes). Briefly, after addition of CM-H 2 DCF-DA (5 M), the cells were incubated for 15 min in the dark at 37°C. CM-H 2 DCF is oxidized by ROS to the highly fluorescent CM-DCF compound. After two washes with PBS, the cells were detached by trypsin-EDTA and immediately analyzed by flow cytometry. Ten-thousand individual data points were collected for each sample. To study ROS production by fluorescence microscopy, keratinocytes grown on glass coverslips in 3.5-cm wells were loaded with CM-H 2 DCF-DA as mentioned above. After two washes in PBS, cells were observed and photographed with a fluorescence microscope.
Statistics-Student's t test was applied for statistical evaluation and a p value of Ͻ0.05 (* on Figs. 3-7) was considered significant. Results are presented as means Ϯ S.D. and CuZn-SOD (TPCuZnSODW) was determined by assessing EGFPpositive cell percentage and by measuring catalase-and CuZn-SODspecific activities and proteins (Fig. 1). The percentage of EGFP-positive cells was determined by cytofluorimetry 5 days post-transduction and was between 85 and 95% (Fig. 1B). The increase in the amount of catalase and/or CuZn-SOD proteins following transduction was assayed by Western blotting (Fig. 1C). Corresponding enzyme activities were measured after 5 days of culture (Fig. 1D). In the cells transduced by catalase or CuZn-SOD, the activity of each enzyme was much greater than in the non-transduced cells: 7.1-fold for catalase and 8-fold for CuZn-SOD. In the cells co-transduced with the catalase-and the CuZn-SOD-expressing vectors, both enzyme activities were increased (7.3fold for catalase and 7.5-fold for CuZn-SOD). Transduction of cells by catalase-expressing vector did not affect SOD activity or vice versa.

Keratinocytes Irradiated at 4°C Are Less Susceptible to UVB-induced Apoptosis Compared with Cells Irradiated at Room Temperature, and Catalase Overexpession Reduces This Susceptibility in Both
Conditions-Initially, the trypan blue exclusion assay was performed to evaluate viability  . Analysis of caspase activation and p53 accumulation. A, Western blot analysis for caspase-9, -8, and -3 activation and for p53 and serine 15-phosphorylated p53 accumulation was carried out 16 h after UVB irradiation. Equal loading protein was confirmed by ␤-actin. The protein bands corresponding to activated caspase-9 (B), caspase-8 (C), caspase-3 (D), and to p53 (E) and serine 15-phosphorylated p53 (F ) were quantified. Data represent the average density from three independent sets of Western blots. The level of activated caspases in non-transduced keratinocytes irradiated at room temperature was normalized to 1 for each experiment. The error bar denotes S.D.
lowing UVB irradiation is shown in Fig. 3B. The percentage of apoptotic and late apoptotic/necrotic cells in keratinocytes irradiated at 4°C was lower than the percentage found in keratinocytes irradiated at room temperature. For keratinocytes overexpressing catalase, there was a marked reduction of apoptotic and late apoptotic/necrotic cells in both conditions compared with the corresponding controls. This result was specific for the catalase enzyme and was not found with CuZn-SOD overexpression. The same results were found by measuring DNA fragmentation using flow cytometry (Fig. 4): the percentage of cells with a DNA content less than the G 1 amount (sub-G 1 DNA) after irradiation at 4°C was lower than the percentage found following irradiation at room temperature. Again, catalase (but not CuZn-SOD) overexpression had a protective effect in both conditions.
Catalase Overexpression Decreases Caspase-9 and Caspase-3 Activation and Also Reduces p53 and Serine 15-phosphorylated p53 Accumulation in UVB-irradiated Keratinocytes-UVB-induced apoptosis is a complex process in which extrinsic and intrinsic pathways are involved. The additive effect of death receptor clustering inhibition (by keeping the cells at 4°C) and of ROS formation prevention (by catalase overexpression) on apoptosis suggests that UVB irradiation acts by two independent mechanisms. We thus investigated caspase-8, -9, and -3 activation 16 h after UVB irradiation at 4°C and room temperature. As shown in Fig. 5, in cells irradiated at 4°C, there was a clear decrease in activation of caspase-8 and -3 as compared with cells irradiated at room temperature. Caspase-9 activation was similar in both conditions. Cells overexpressing catalase, irradiated at 4°C or room temperature, showed a 30% reduction in caspase-9 and -3 and a 9% decrease in caspase-8 activation compared with corresponding non-transduced keratinocytes. These data indicate that keeping cells at low temperature reduces UVB-induced apoptosis by inhibition of caspase-8 activation, whereas catalase overexpression reduces apoptosis mostly by inhibition of caspase-9 activation and slightly by reduction of caspase-8 activation.
To investigate whether ROS-induced DNA damage is involved in the induction of apoptosis following UVB irradiation, total p53 and serine 15-phosphorylated p53 proteins were assessed (Fig. 5, A, E, and F). Following UVB irradiation, accumulation of p53 protein was not affected by keeping cells at low temperature during irradiation but was significantly reduced in cells overexpressing catalase. These data indicate that UVB-induced ROS participate in apoptosis by DNA damage formation and consequently p53 induction.

ROS Level Increases in Two Distinct Stages following UVB Irradiation, while Catalase Overexpression Prevents the Second Increase-To exam-
ine the effects of UVB-induced ROS on apoptosis in more detail, a kinetic analysis of intracellular ROS formation was performed. As shown in Fig. 6A, UVB irradiation induced ROS production at two different time intervals. The initial increase in ROS levels took place at the time of irradiation and declined rapidly in a time-dependent manner, reaching the base-line level around 1 h post-UVB. The second increase took place around 3 h after irradiation and persisted for 6 h. Kinetic analysis of ROS production by fluorescence microscopy confirmed this result (Fig. 6B). Interestingly, ROS levels at the time of irradiation were higher in cells irradiated at low temperature than in cells irradiated at room temperature (Fig. 6A) but the second peak was similar. Intracellular ROS levels were then measured in catalase-and CuZn-SOD-transduced keratinocytes (Fig. 7). Overexpression of catalase and CuZn-SOD led to a significant decrease in endogenous ROS level in non-irradiated conditions at both 4°C (data not shown) and room temperature. Following irradiation at 4°C (data not shown) or room temperature, catalase overexpression completely prevented the late increase in ROS levels (after 3 h), but it did not abolish the first peak (at the time of irradiation). In contrast, CuZn-SOD overexpression had no significant effect on the ROS level following irradiation.

DISCUSSION
Recent studies on UV-induced apoptosis have shown that DNA damage, death receptor activation, and ROS formation participate independently in the formation of apoptotic cells (24,28,29). The present study focused on the respective roles of catalase and CuZn-SOD in the protection of keratinocytes against UVB-induced apoptosis.
Keeping human keratinocytes at low temperature decreased UVBinduced apoptosis by reducing caspase-8 activation, independently of any change in p53. It has been shown that exposure of HeLa cells and HaCat cells to UVB at low temperature prevents UV-induced clustering of death receptors and that this inhibition is associated with a reduction in apoptosis (8,12). Treatment of cells with zIETD, an inhibitor of caspase-8, led to the same results (24). Aragane et al. (12) suggested that death receptor aggregation may be independent of p53, since they obtained similar findings with p53-mutated (HaCat) and p53 wild-type (SCL-1) cells.
We observed a reduction in UVB-induced apoptosis in catalase-transduced keratinocytes compared with non-transduced or CuZn-SOD-transduced cells and found that this protective effect of catalase is mainly due to a reduction in caspase-9 activation and partially to a reduction in caspase-8 activation and that it is accompanied by a decrease in p53 accumulation. It has been reported that decreased UVB-induced apoptosis due to radical scavengers (such as pyrrolidene dithiocarbamate or N-acetylcysteine) is accompanied by a reduction in cytochrome c release into the cytoplasm (24), a phenomenon that occurs upstream of caspase-9 activation. To explain why catalase has a protective effect against UVB irradiation but not CuZn-SOD, we hypothesized that CuZn-SOD overexpression lowered superoxide anion levels but led to a deleterious excess in intracellular H 2 O 2 , which cannot be quenched by endogenous catalase. The tumor suppressor protein p53, which functions as a latent, short lived transcription factor, accumulates following various stresses, including oxidative stress, DNAdamaging agent, and UV irradiation (30 -32). Moreover, p53 accumulation in response to ROS exposure facilitates the repair of ROS-induced DNA damage (33). Thus, the reduction in p53 accumulation upon catalase overexpression suggests that the protective effect of catalase against UVB-induced apoptosis takes place through a reduction in ROS-induced DNA damage. Stabilization and functional activation of p53 occurs in response to DNA damage via complex mechanisms. One of these mechanisms is phosphorylation of p53 at N-and C-terminal residues, especially at serine 15, by which interaction of the protein with its negative regulator MDM2 is abolished (30,34). Thus, the reduction in serine 15-phosphorylated p53 accumulation reflects the decreased DNA damage noted in irradiated keratinocytes overexpressing catalase (compared with non-transduced cells). UVBinduced apoptosis is a protective mechanism, since it eliminates cells with damaged DNA. Therefore, inhibition of UVB-induced cell death may enhance the carcinogenic risk, unless it is associated with an increase in removal of DNA damage. According to our results, the prevention of UVBinduced apoptosis by catalase overexpression is accompanied by a reduc- Measurement of intracellular ROS demonstrated that UVB irradiation induced ROS production at two distinct stages. Two phases of ROS production were also reported in a model of apoptosis induced by glutamate in the HT22 cell line (35) and in apoptosis induced by ␥-irradiation in the IM-9 cell line (36).
Our results indicate that the initial increase in ROS levels detected immediately after irradiation, declined very rapidly (within 1 h). This finding is consistent with observations reported by Wang et al. (30) demonstrating that ROS production in irradiated immortalized keratinocytes decreased in a time-dependent manner and was no longer significantly different from controls at 90 min post-UVB.
ROS can be generated by mitochondria via enzyme complexes during normal processes of oxidative phosphorylation. Numerous cytosolic enzymes including cyclooxygenase, nitric oxide synthase, xanthine oxidase, and the plasma membrane-bound NADPH oxidase can also generate ROS. Although the source of ROS production following irradiation is not definitely known, several studies have suggested that UVB irradiation induces an immediate increase in intracellular ROS via the activation of NADPH oxidase activity (37,38). Our results using catalase and CuZn-SOD overexpression are consistent with this mechanism. In keratinocytes overexpressing catalase, activation of NADPH oxidase causes an increase in superoxide anion (detected using CM-H 2 DCF), which cannot be quenched by catalase. In the case of overexpression of CuZn-SOD, superoxide anion is rapidly converted to H 2 O 2 , which can also be detected using the CM-H 2 DCF probe. Pretreatment of keratinocytes with an inhibitor of NADPH oxidase (diphenylene iodonium) prior to irradiation abolished the initial increase in ROS levels (data not shown), thus confirming this hypothesis. The source of the second increase in ROS levels remains to be determined, but the reduction in apoptosis, accompanied by a decrease in caspase-9 activation, suggests that mitochondria could be the main source of this ROS generation.
The reduction in apoptosis when the late increase in ROS levels is prevented in keratinocytes overexpressing catalase suggests that this second peak of ROS production has a major role in ROS-mediated UVB-induced apoptosis. This finding is consistent with the results of Chen et al. (36) demonstrating that the late ROS increase following ionizing radiation is associated with increased mitochondrial permeability transition pore opening, collapse of mitochondrial inner transmembrane potential, and cytochrome c release. After addition of H 2 O 2 to cultured keratinocytes just before irradiation (data not shown), we found that a mild increase in ROS level at the time of irradiation did not increase UVB-induced apoptosis. This result suggests that an increase in ROS levels at the time of irradiation has no effect on apoptosis. This finding is consistent with the results of Kulms et al. (24) demonstrating that addition of a ROS scavenger (pyrrolidene dithiocarbamate) up to 1 h after UVB exposure has a maximal reducing effect on UVB-induced apoptosis. The role of the first increase in ROS levels remains unclear, but recent data suggest that ROS can also function as a second messenger to induce signal transduction. The mode of action of ROS may involve direct interaction with specific receptors and/or redox activation of members of signaling pathways such as protein kinases, protein phosphatases, and transcription factors (39 -41), which in turn can regulate cell proliferation, cell cycle, and cell death.
In conclusion, this study is the first to show that UVB irradiation induces an increase in ROS levels at two different time intervals and that the second (but not the first) increase in ROS production has a major role in UVB-induced apoptosis. These findings further support the concept of a reinforcement of natural antioxidant photoprotective defenses to prevent severe sunburn and probably skin cancer. Catalase overexpression was one of the two potential enzymatic photoprotectors tested and performs better than SOD in this respect.