Silibinin up-regulates DNA-protein kinase-dependent p53 activation to enhance UVB-induced apoptosis in mouse epithelial JB6 cells.

In the present study, we employed a well established JB6 mouse epithelial cell model to define the molecular mechanism of efficacy of a naturally occurring flavonoid silibinin against ultraviolet B (UVB)-induced skin tumorigenesis. UVB exposure of cells caused a moderate phosphorylation of ERK1/2 and Akt and a stronger phosphorylation of p53 at Ser(15), which was enhanced markedly by silibinin pretreatment. Kinase activity of ERK1/2 for Elk-1 and Akt for glycogen synthase kinase-3beta was also potently enhanced by silibinin pretreatment. Furthermore, silibinin increased the UVB-induced level of cleaved caspase 3 as well as apoptotic cells. Based on these observations, next we investigated the role of upstream kinases, ATM/ATR and DNA-PK, which act as sensors for UVB-induced DNA damage and transduce signals leading to DNA repair or apoptosis. Whereas UVB strongly activated ATM as observed by Ser(1981) phosphorylation, it was not affected by silibinin pretreatment. However, pretreatment of cells with the DNA-protein kinase (PK) inhibitor LY294002 strongly reversed silibinin-enhanced Akt-Ser(473) and p53-Ser(15) as well as ERK1/2 phosphorylation together with a dose-dependent decrease in cleaved caspase 3 and apoptosis (p < 0.05). In addition, silibinin pretreatment strongly enhanced H2A.X-Ser(139) phosphorylation and DNA-PK-associated kinase activity as well as the physical interaction of p53 with DNA-PK; pretreatment of cells with LY294002 but not caffeine abolished the silibinin-caused increase in both DNA-PK activation and p53-Ser(15) phosphorylations. Together, these findings suggest that silibinin preferentially activates the DNA-PK-p53 pathway for apoptosis in response to UVB-induced DNA damage, and that this could be a predominant mechanism of silibinin efficacy against UVB-induced skin cancer.

In the present study, we employed a well established JB6 mouse epithelial cell model to define the molecular mechanism of efficacy of a naturally occurring flavonoid silibinin against ultraviolet B (UVB)-induced skin tumorigenesis. UVB exposure of cells caused a moderate phosphorylation of ERK1/2 and Akt and a stronger phosphorylation of p53 at Ser 15 , which was enhanced markedly by silibinin pretreatment. Kinase activity of ERK1/2 for Elk-1 and Akt for glycogen synthase kinase-3␤ was also potently enhanced by silibinin pretreatment. Furthermore, silibinin increased the UVBinduced level of cleaved caspase 3 as well as apoptotic cells. Based on these observations, next we investigated the role of upstream kinases, ATM/ATR and DNA-PK, which act as sensors for UVB-induced DNA damage and transduce signals leading to DNA repair or apoptosis. Whereas UVB strongly activated ATM as observed by Ser 1981 phosphorylation, it was not affected by silibinin pretreatment. However, pretreatment of cells with the DNA-protein kinase (PK) inhibitor LY294002 strongly reversed silibinin-enhanced Akt-Ser 473 and p53-Ser 15 as well as ERK1/2 phosphorylation together with a dose-dependent decrease in cleaved caspase 3 and apoptosis (p < 0.05). In addition, silibinin pretreatment strongly enhanced H2A.X-Ser 139 phosphorylation and DNA-PKassociated kinase activity as well as the physical interaction of p53 with DNA-PK; pretreatment of cells with LY294002 but not caffeine abolished the silibinin-caused increase in both DNA-PK activation and p53-Ser 15 phosphorylations. Together, these findings suggest that silibinin preferentially activates the DNA-PK-p53 pathway for apoptosis in response to UVB-induced DNA damage, and that this could be a predominant mechanism of silibinin efficacy against UVB-induced skin cancer.
Ultraviolet B (UVB) 1 radiation plays a major role in the development of non-melanoma skin cancer, which is the most common human malignancy in the United States (1). In recent years, chemoprevention is considered to be a promising strategy for the control of this malignancy, where several studies have shown that various phytochemicals from both nutritive and non-nutritive sources protect against UVB-induced skin cancer (2)(3)(4)(5)(6). In this regard, silymarin (a polyphenolic flavonoid antioxidant from milk thistle) and silibinin (the bioactive component in silymarin) have also shown strong potential in preventing UV-induced skin damages and photocarcinogenesis (2,(5)(6)(7). We have demonstrated the chemopreventive efficacy of silymarin/silibinin in the SKH-1 mouse skin model, where these agents showed strong protection against UVBinduced tumor initiation, tumor promotion, and complete carcinogenesis (2). We also found that silymarin protects against chemical carcinogenesis in SENCAR mouse skin by modulating mitogen-activated protein kinases (MAPKs) and inducing apoptosis (3). In a more recent study, we observed a strong protective effect of silibinin against UVB-induced skin carcinogenesis in SKH-1 mice where topical applications before or after UVB or its dietary feeding were effective in inhibiting tumor multiplicity as well as tumor volume (7). Also, we have reported the dual effects of silibinin in protecting from or enhancing apoptosis in response to UVB-induced moderate/excessive damage in spontaneously immortalized human keratinocyte HaCaT cells (5).
Exposure of skin to UVB results in genotoxic stress which is a major cause for skin cancer initiation (1). In recent years, important roles of ataxia telangiectasia mutated and ataxia telangiectasia and rad3-related (ATM/ATR) and DNA-dependent protein kinase (DNA-PK) were identified in DNA damage recognition and subsequent initiation of downstream protein kinase cascades (8). Both ATM and ATR are activated by DNA damage although ATR has been shown to be specifically associated with damage induced by UV (9). DNA-PK has recently been identified as a crucial molecule activating apoptotic machinery in response to DNA damage, particularly DNA double strand breaks (10). DNA-PK is a nuclear serine-threonine kinase composed of a catalytic subunit and a Ku70/80 subunit, which is activated during DNA damage induced by both ionizing and UV radiations (10), and is regarded as a DNA damage sensor (10). Furthermore, DNA-PK complexed with p53 has been shown to act as a sensor of abnormal DNA structures (11), and p53 has been shown as an effector for DNA-PK-mediated signaling where DNA-PK selectively regulates a p53-dependent apoptotic response (12). Recent data suggest that these damage sensor molecules such as ATM/ATR and DNA-PK play a pivotal role in protecting from UV-caused tumor initiation and promotion (9,10).
In addition to its direct DNA damaging effects causing tumor initiation, UVB also acts as a tumor promoter in the development of non-melanoma skin cancer (13,14). Recent reports show that UVB-induced diverse mitogenic and survival signaling pathways are responsible for its tumor promoting effects (15,16) where these signals converge into activating transcription factors such as AP1 and nuclear factor-B that are crucial in tumor promotion (17)(18)(19). For instance, it has been shown that activation of MAPK ERK1/2 is indispensable for UVBinduced cell transformation and promotion of JB6 cells (20) where UVB-caused p53-Ser 15 phosphorylation is ERK1/2-dependent (21,22). On the other hand, ERK1/2 also induces apoptosis in response to various stress signals, in both p53-dependent and -independent manners (23)(24)(25). Protein kinase B (Akt/PKB) is another kinase that is activated in response to many growth factors and mitogens, including UVB, and controls cellular signaling molecules responsible for preventing cell death (26,27).
Induction of apoptosis is one of the key mechanisms by which various chemopreventive agents provide protection from DNA damage-induced tumor initiation and subsequent clonal expansion of the initiated cells during tumor promotion (28,29). p53 is critical in the UVB-mediated DNA damage response, and depending on the magnitude of DNA damage, p53 aids in either DNA repair or in inducing apoptosis (30). Phosphorylation of p53 in response to UVB radiation leads to its stabilization and activation (21), which is crucial in p53-mediated cell cycle arrest and/or apoptosis (31,32). Taken together, these lines of evidence suggest that agents, which could modulate UVB-induced signaling pathways, especially damage sensors such as ATM/ATR and DNA-PK or an effector molecule such as p53 could be both novel and useful in protecting against photocarcinogenesis. In the present study, we assessed the effect of silibinin on UVB-induced apoptosis and associated molecular mechanisms employing well studied JB6 mouse keratinocytes as a model.

EXPERIMENTAL PROCEDURES
Cell Culture and UVB Irradiation-JB6 cells were maintained in minimal essential medium containing 2 mM L-glutamine, 25 g/ml gentamicin, and 5% serum under standard culture conditions. For all treatments, cells were grown to 80% confluence and then either treated with silibinin (100 M) for 24 h or treated for 2 h with the indicated inhibitors before silibinin treatment. Before UVB irradiation, media was removed from culture plates; cells were washed with phosphatebuffered saline twice and then covered with a thin layer of phosphatebuffered saline followed by UVB irradiation. Control cultures were identically processed but not irradiated. The UVB light source was a bank of four FS24T12-UVB-HO sunlamps equipped with a UVB Spectra 305 Dosimeter (Daavlin Co., Bryan, OH), which emitted about 80% radiation in the range of 280 -340 nm with a peak emission at 314 nm as monitored with a SEL 240 photodetector, 103 filter and 1008 diffuser attached to an IL 1400 Research Radiometer (International Light, Newburyport, MA).
Western Blotting-Cell lysates were prepared in non-denaturing lysis buffer (10 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% Triton X-100, 1 mM EDTA, 1 mM EGTA, 0.3 mM phenylmethylsulfonyl fluoride (PMSF), 0.2 mM sodium orthovanadate, 0.5% Nonidet P-40, 5 units/ml aprotinin), and protein concentration in lysates was determined using Bio-Rad DC protein assay kit (Bio-Rad). For immunoblot analyses, 40 -100 g of protein per sample was denatured in 2ϫ SDS-PAGE sample buffer and subjected to SDS-PAGE on Tris/glycine gels. The separated proteins were transferred onto nitrocellulose membranes followed by blocking with 5% nonfat milk powder (w/v) in Tris-buffered saline (10 mM Tris, 100 mM NaCl, 0.1% Tween 20) for 1 h at room temperature or overnight at 4°C. Membranes were then probed with specific primary antibodies followed by peroxidase-conjugated secondary antibody, and visualized by ECL detection system. Data presented for Western blots are representative of at least two to three independent experiments in each case.
Kinase Activity Assay for ERK1/2 and Akt/PKB-Kinase activity assays for ERK1/2 and Akt were done according to the manufacturer's protocol (Cell Signaling Technology Inc., Beverly, MA) with some modifications. Briefly, cells were lysed (20 mM Tris, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% Triton X-100, 2.5 mM sodium pyrophosphate, 1 mM ␤-glycerophosphate, 1 mM Na 3 VO 4, 1 g/ml leupeptin, and 1 mM PMSF) after the desired treatments and cell extracts were prepared. Two hundred g of protein from each sample was immunoprecipitated using immobilized phospho-specific ERK p44/p42 or Akt antibody at 4°C overnight. The immunocomplexes obtained were washed twice with lysis buffer and once with kinase buffer (25 mM Tris, pH 7.5, 5 mM ␤-glycerophosphate, 2 mM DTT, 0.1 mM Na 3 VO 4, 10 mM MgCl 2 ). The bead pellets were re-suspended in 40 l of kinase buffer with 200 M ATP and 2 g of substrate (Elk-1 for ERK and GSK-3␣/␤ for Akt) and incubated for 30 min at 30°C. The reaction was terminated with 5ϫ SDS sample buffer, and samples were boiled for 5 min, centrifuged, and supernatants were loaded on 12% Tris/glycine gel. Substrate phosphorylation was visualized by immunoblotting using phospho-specific antibody for Elk-1 and GSK-3␣/␤.
Propidium Iodide Staining and Flow Cytometry Analysis-For analysis of cell cycle distribution, JB6 cells were plated in 60-mm dishes, and the next day treated with silibinin (100 M) for 24 h. Cells were then either unexposed or sham-irradiated to 100 mJ/cm 2 UVB. Cells were collected at 24 h after UVB exposure, stained with saponin/ propidium iodide (PI), and analyzed by flow cytometry in the FACS Analysis Core Facility at the University of Colorado Cancer Center, Denver, CO.
Annexin/PI Staining and Flow Cytometry Analysis-For apoptosis analysis, JB6 cells were plated in 60-mm dishes, and the next day treated with silibinin (100 M) alone, the desired inhibitors alone, or pre-treated with inhibitors for 2 h before silibinin treatment. After 24 h, cells were either unexposed or exposed to 100 mJ/cm 2 UVB. Cells were collected at the indicated times after UVB exposure, stained with annexin V/PI (Molecular Probes) following the manufacturer's protocol, and apoptotic cell population was then analyzed immediately by flow cytometry in the FACS Analysis Core Facility, University of Colorado Cancer Center, as published earlier (6).
Morphological Analysis-Following the desired treatments, pictures for cellular morphology were taken using a Kodak DC290 camera under an inverted microscope at ϫ100 magnification and processed using the Kodak Microscopy Documentation System 290 (Eastman Kodak Co., Rochester, NY).
DNA-PK Pull-down and Kinase Assay-JB6 cells after the desired treatments were harvested by trypsinization, pelleted by centrifugation at 1000 ϫ g at 4°C for 10 min, and then washed once with phosphatebuffered saline containing 0.2 mM PMSF and again with phosphatebuffered saline containing 0.5 mM PMSF. The cell pellets were then re-suspended in 3 ml of lysis buffer (10 mM HEPES, 25 mM KCl, 10 mM NaCl, 1 mM MgCl 2 , 0.1 mM EDTA, pH 7.2, containing 0.1 mM DTT and 0.2 mM PMSF) and centrifuged at 1,000 ϫ g for 5 min. Finally, cells were re-suspended in 100 l of lysis buffer, incubated on ice for 10 min, and flash frozen and stored at Ϫ80°C until further use. The cell extracts were quickly thawed at 37°C and 50 l of the each extract was removed and added with 5.5 l each of 5 M NaCl, 100 mM MgCl 2 , and 5 mM DTT. To dilute the salt, the extracts were then added with 4 volumes of 50 mM HEPES, 0.5 M KCl (pH 7.5). The diluted extracts were centrifuged at 10,000 ϫ g for 3 min and supernatants were collected and used for the assay. Briefly, to the diluted supernatant, a 40-l slurry of 5 mg of DNA-cellulose/sample (pre-equilibrated overnight with wash buffer, 25 mM HEPES, pH 7.5, 50 mM KCl, 0.1 mM EDTA, 0.2 mM EGTA, 20% glycerol, 0.5 mM DTT, 0.1 mM PMSF) was added and incubated with gentle rotation at 4°C for 10 min. Samples were centrifuged and re-extracted with 20 l of DNA-cellulose. The pellets were then re-suspended in 30 l of wash buffer. For the kinase assay, 14 l of the DNA-cellulose bound sample was added with 2 l of 0.1 mM MgCl 2 , 1 mM EDTA, 2 mM EGTA, and 10 mM DTT, 2 l of substrate peptide, 2 l of 1 mM ATP containing [␥-32 P]ATP. The reaction was incubated at 37°C for 10 min and then stopped by the addition of 20 l of 30% acetic acid containing 2.5 mM cold ATP. 15 l of each sample was spotted onto Whatman P81 phosphocellulose paper, washed 4 times with 15% acetic acid, and counted using a liquid scintillation counter.
DNA-PK-p53 Binding Assay-For the DNA-PK-p53 binding assay, we used DNA-cellulose-extracted samples from the DNA-PK pull-down assay. These samples were added with 2ϫ sample buffer, boiled for 5 min, and centrifuged at 14,000 ϫ g for 5 min. The supernatant was then loaded on a 4 (for DNA-PK) or 12% (for p53) Tris/glycine gel and Western blotting was carried out as mentioned above. Membranes were probed with DNA-PK or p53 primary antibody followed by secondary antibodies and ECL detection as mentioned above.
Statistical Analysis-Statistical analysis was performed using SigmaStat 2.03 software (Jandel Scientific, San Rafael, CA) as needed. Data were analyzed using t test as needed and a statistically significant difference was considered to be at p Ͻ 0.05.

Silibinin Enhances UVB-caused Phosphorylation and Kinase
Activity of ERK1/2 and Akt/PKB-First, we assessed the effect of silibinin pre-or post-treatment on UVB-induced MAPK family molecules ERK1/2, p38, and JNK, and Akt/PKB, as these molecules are crucial transducers of UVB-induced signaling cascades. JB6 cells without any serum starvation were treated with Me 2 SO alone or different doses of silibinin (25, 50, and 100 M) in Me 2 SO for 24 h and then irradiated with 100 mJ/cm 2 UVB dose. In some cultures, silibinin was added at similar doses immediately after UVB exposure. In all cases, cells were collected 24 h after UVB exposure and cell lysates were prepared and analyzed for the desired signaling molecules. Exposure of cells to UVB caused a weak phosphorylation of ERK1/2 ( Fig. 1A), p38 (Fig. 1B), JNK1/2 (Fig. 1C), and Akt-Ser 473 (Fig. 1D), which were either not detectable or were at low levels in mock-irradiated controls. Interestingly, UVBcaused phosphorylation of ERK1/2 ( Fig. 1A) and Akt (Fig. 1D) was strongly enhanced by silibinin pre-treatment at 100 M dose without any change in their total protein levels. Pretreatment of cells with lower doses of silibinin or post-treatment at all doses did not show any changes compared with UVB alone (data not shown). In addition, there was no observable change in the phospho-or total levels of p38 and JNK with silibinin treatments (Fig. 1, B and C).
Because we found a further increase in UVB-induced ERK1/2 and Akt phosphorylation by silibinin pre-treatment, we next assessed whether this is accompanied by an increase in their kinase activity as well as phosphorylation of their downstream substrates Elk-1 and GSK-3␣/␤, respectively. Phospho-ERK1/2 and phospho-Akt were immunoprecipitated from total lysates employing their specific primary antibodies and then in vitro kinase assays were performed using Elk-1 and GSK-3 as substrates, respectively. Kinase activities of ERK1/2 and Akt as observed by the phosphorylation of Elk-1 (Fig. 1E) and GSK-3␣/␤ (Fig. 1F), respectively, were also enhanced strongly by silibinin pre-treatment as compared with UVB alone. Similarly, silibinin pre-treatment also caused a strong increase in the phosphorylation of Elk-1 (Fig. 1G) and GSK-3␣/␤ (Fig. 1H); actin immunoblotting confirmed equal protein loading (Fig. 1I). All together, these results were comparable with an increase in ERK1/2 and Akt phosphorylation observed by Western blotting, and support the notion that silibinin pre-treatment upregulates kinase activities of ERK1/2 and Akt when cells are exposed to UVB.
Silibinin Enhances UVB-mediated p53-Serine 15 Phosphorylation-p53 is known to play a crucial role in UVB-induced cell cycle arrest and/or apoptosis, and serine 15 phosphorylation is essential for stabilization and activation of p53 (21). Accordingly, next we assessed the effect of silibinin pre-treatment on UVB-induced p53 phosphorylation and its total levels by Western immunoblotting. Under identical treatment proto-cols as for ERK1/2 and Akt analyses ( Fig. 1A-I), compared with sham irradiated and silibinin (100 M) alone treated controls, UVB exposure at 100 mJ/cm 2 dose resulted in a strong p53-Ser 15 phosphorylation and accumulation (Fig. 1J); however, pre-treatment of cells with silibinin for 24 h before UVB exposure resulted in a much stronger increase in both p53-Ser 15 phosphorylation and stabilization (Fig. 1J). The effect of silibinin pre-treatment on UVB-induced p53 phosphorylation was specific to Ser 15 , as we did not observe any considerable change in Ser 6 , and Ser 392 phosphorylation sites of p53 (Fig. 1, K and  L). Additionally, phospho-p53-Ser 9 and phospho-p53-Ser 20 were not detectable in our experimental conditions following UVB, silibinin plus UVB, and silibinin alone treatments (data not shown).
Differential Kinetics of Silibinin plus UVB-Versus UVBcaused ERK1/2, Akt, and p53 Phosphorylation and p53 Stabilization-Based on our findings that silibinin pre-treatment for 24 h followed by UVB irradiation at 100 mJ/cm 2 dose and analyzing samples 24 h thereafter causes strong ERK1/2, Akt, and p53 phosphorylation and p53 stabilization, two different studies were next performed seeking a UVB dose-response and a time kinetics of the observed effects following UVB irradiations. Under the identical experimental conditions mentioned FIG. 1. Silibinin enhances UVB-induced ERK1/2 and Akt activation, and p53 stabilization in JB6 cells. JB6 cells were treated with Me 2 SO or 100 M silibinin for 24 h and then irradiated with a 100 mJ/cm 2 dose of UVB. Cells were harvested 24 h later, lysates were prepared and Western blotting was carried out for phosphorylated and total levels of ERK1/2 (A), p38 (B), JNK (C), and Akt (D). Cell lysates were also analyzed for the kinase activity of ERK1/2 (E) and Akt (F), as observed by substrate phosphorylation of Elk-1 and GSK-3 as described under "Experimental Procedures." In similar treatments, Western blot was carried out for p-Elk-1 (G) and p-GSK-3␣/␤ (H), and protein loading was checked by stripping and re-probing the membranes with ␤-actin (I). These cell lysates were also analyzed for Ser 15 -phosphorylated and total levels of p53 (J), phospho-p53-Ser 6 (K), and phospho-p53-Ser 392 (L). IP, immunoprecipitation; IB, immunoblotting.
above, an increase in UVB dose from 100 to 200 and 400 mJ/cm 2 did result in an alteration in ERK1/2 ( Fig. 2A) and Akt ( Fig. 2B) phosphorylation, but interestingly decreased both UVB alone and silibinin plus UVB-caused p53-Ser 15 phosphorylation and stabilization (Fig. 2C). Based on our observations that a UVB dose higher than 100 mJ/cm 2 does not result in an additional increase in the phosphorylation of ERK1/2, Akt, or p53, all further studies were done employing 100 mJ/cm 2 UVB as the dose.
In the time kinetic studies at 3, 6, and 12 h, compared with UVB alone, silibinin pre-treatment started showing a very strong increase in ERK1/2 phosphorylation beginning 3 h after UVB irradiation and remained sustained at 6 and 12 h (Fig.  2D). Additional studies at 24, 48, and 72 h showed similar sustained ERK1/2 phosphorylation in silibinin plus UVB-exposed cells compared with a marginal response in UVB alone samples (data not shown). In the case of Akt phosphorylation, compared with UVB alone, silibinin pre-treatment started showing a further increase at 6 h after exposure (Fig. 2E), and remained sustained up to 72 h (data not shown). In time kinetics studies for p53 phosphorylation and stabilization, silibinin pre-treatment resulted in comparable levels of p53 phosphorylation to UVB alone up to 12 h after UVB exposure (Fig. 2F); however, a clear and sustained increase was evident starting at 24 up to 72 h; interestingly, UVB alone treatment showed no p53-Ser 15 phosphorylation at 72 h (Fig. 2G). Comparable responses were also evident throughout the 3-72-h time kinetics study for total p53 levels (Fig. 2, F and G). Together, these results suggest that the silibinin-caused increase in UVB-induced ERK1/2 and Akt phosphorylation could have a role in p53 phosphorylation and stabilization resulting in the biological responses such as cell cycle arrest and apoptosis. Accordingly, studies were next conducted to answer these questions.
Silibinin Alters UVB-mediated Cell Cycle Progression and Enhances UVB-induced Apoptosis-Induction of cell cycle arrest after DNA damage is known to provide cells more time for DNA damage repair, and is one of the mechanisms by which chemopreventive agents exert their protective effect against UVB (34,35). Accordingly, first we assessed the effect of silibinin on UVB-caused changes in cell cycle progression where cells were pre-treated with silibinin (100 M) for 24 h followed by UVB irradiation at 100 mJ/cm 2 dose, and 24 h later cells were harvested and stained with saponin/PI and analyzed by FACS (Fig. 3A). UVB exposure resulted in a significant (p Ͻ 0.005) S phase arrest as compared with sham-irradiated control (31% in control versus 41% in UVB), whereas silibinin pre-treatment shifted the S phase arrest toward the G 1 phase (47.2 Ϯ 2.4 in UVB alone versus 53.3 Ϯ 0.7 in silibinin plus UVB) as well as the G 2 -M phase (11.6 Ϯ 0.5 in UVB alone versus 17.7 Ϯ 0.4 in silibinin plus UVB, p Ͻ 0.001) arrests as compared with UVB alone (Fig. 3B).
Apoptosis is one of the mechanisms involved in eliminating the cells with severe DNA damage following UVB irradiation thereby preventing their entry into the cell cycle. Because we observed a strong increase in p53-Ser 15 phosphorylation as well as an alteration in UVB-mediated S-phase arrest following silibinin pre-treatment of the cells, which could be crucial for apoptosis induction, we next investigated the effect of silibinin on UVB-induced apoptosis. Cells after identical treatments to those in other studies were harvested and subjected to lysate preparation followed by Western blotting, or processed for annexin V/PI staining and FACS analysis. As shown in Fig. 3B, compared with sham irradiated control or silibinin alonetreated samples showing no reactivity for the cleaved caspase-3, UVB exposure resulted in a strong band for the cleaved caspase-3; however, silibinin pre-treatment very strongly increased the UVB-induced cleaved caspase-3 level (Fig. 3C). Blotting membrane for ␤-actin confirmed equal protein loading (Fig. 3D). In quantitative apoptotic cell death analysis, silibinin alone treatments showed 2.1 Ϯ 0.6% apoptotic cells that were comparable with the sham irradiated control with 1.6 Ϯ 0.05% apoptotic cells (Fig. 3E). UVB irradiation resulted in 11.0 Ϯ 0.3% apoptotic cells that also strongly increased to 40.1 Ϯ 1.6% in silibinin pre-treated and UVBexposed cells. We did not observe the pre-G 1 cell population in UVB or silibinin plus UVB treatment (Fig. 3A); however, analysis of cell morphology data showed a decrease in total cell number with elongated and rounded (floaters) cells in UVB treatment, which was enhanced strongly in silibinin plus UVB treatment as compared with control or only silibinin treatment (Fig. 3F). These results convincingly show the strong apoptosis enhancing effect of silibinin when cells are exposed to UVB, and suggest that this could be one of the major mechanisms associated with silibinin efficacy against UVB damage in JB6 cells.

Enhancement of UVB-induced ERK1/2 Activation by Silibinin Is Not the Major Pathway in Apoptosis Induction by
Silibinin-To assess the role of an increase in ERK1/2 phosphorylation by silibinin pre-treatment followed by UVB irradiation in p53 phosphorylation as well as apoptotic cell death, cells were pre-treated with PD98059 at 50 and 100 M concentrations for 2 h and then treated with silibinin (100 M) for 24 h before UVB exposure (100 mJ/cm 2 ). Western blot analysis of  15 and total p53 (C) by Western blotting as described under "Experimental Procedures." JB6 cells with or without silibinin pretreatment were exposed to a 100 mJ/cm 2 UVB dose and harvested at different times after UVB exposure (D-G) and analyzed for phosphorylated and total levels of ERK1/2 (D), Akt (E), and p53 (F and G) by Western blotting. cell lysates showed that pre-treatment with PD98059 only partially reverses the silibinin-induced increase in ERK1/2 phosphorylation (Fig. 4A), and that only a higher dose of PD98059 results in a moderate reversal of the silibinin-caused increase in p53-Ser 15 phosphorylation (Fig. 4B), but no effect on Akt phosphorylation (data not shown). Similar treatment with PD98059 also did not show a strong reversal in the level of cleaved caspase-3 (Fig. 4C); re-blotting the membrane for ␤-actin confirmed equal protein loading (Fig. 4D). Consistent with these results, in quantitative apoptosis analysis, pre-treatment of cells with 50 or 100 M PD98059 did not result in reversal in silibinin plus UVB-enhanced apoptotic cell death, in fact there was a small increase in apoptotic cell population although statistically insignificant; the quantitative data shown in Fig.  4E is for a 100 M dose of PD98409. Together, these results imply that ERK1/2 phosphorylation and activation by silibinin pre-treatment followed by UVB irradiation occurs only in part through MEK1/2, and that it does not have any considerable role in p53-Ser 15 phosphorylation and apoptosis induction by silibinin. To further ascertain that indeed this is the case, all studies were repeated with another MEK1/2 inhibitor UO126 that showed similar outcomes; increasing the PD98059 concentration to 200 M also did not result in any additional response to the 100 M dose (data not shown).

Silibinin Increases Phospho-Ser 15 and Total p53 and Enhances UVB-induced Apoptosis via the PI 3-Kinase Pathway-
Based on our findings that silibinin pre-treatment results in a strong increase in Akt phosphorylation, we next examined whether the upstream PI 3-kinase pathway plays an important role in a silibinin-caused increase in phospho-Ser 15 and total p53 as well as apoptosis. Cells were pre-treated with LY294002 at 25 and 50 M concentrations for 2 h and then with silibinin (100 M) for 24 h before UVB exposure at 100 mJ/cm 2 dose. After 24 h of UVB exposure, cells were harvested and analyzed for Akt phosphorylation (Fig. 5A), and p53-Ser 15 phosphorylation and stabilization (Fig. 5B). Pre-treatment with LY294002 decreased the silibinin-caused increase in Akt phosphorylation (Fig. 5A) and p53-Ser 15 phosphorylation as well as its total levels (Fig. 5B) in a dose-dependent manner. In addition, a similar trend was observed for cleaved caspase 3 where LY294002 pretreatment dose-dependently decreased silibinin plus the UVB-caused increase in the cleaved caspase-3 level (Fig. 5C). Blotting the same membrane for ␤-actin confirmed equal protein loading (Fig. 5D). In other studies, we also observed that LY294002 pre-treatment of cells considerably abrogates a silibinin plus UVB-caused increase in ERK1/2 phosphorylation (Fig. 5E), suggesting the involvement of the PI 3-kinase pathway independent of a partial ERK1/2 activation via MEK1/2 observed using PD98059 (Fig. 4A). Consistent with all these observations, in the apoptotic cell death assay, LY294002 pre-treatment strongly reversed the silibinin plus UVB-induced increase in apoptotic cells as observed by flow cytometric analysis of annexin V/PI-stained cells (Fig. 5F). Whereas silibinin treatment followed by UVB exposure resulted in 58 Ϯ 6.0% apoptotic cells, pre-treatment with LY294002 decreased apoptotic cell population to 24.8 Ϯ 1.9% (p Ͻ 0.05) (Fig. 5F). Together, these data suggest that activation of the PI 3-kinase family of kinase(s) followed by p53-Ser 15 phosphorylation and stabilization could be the central mediator of apoptotic signaling induced by silibinin following UVB exposure.
Silibinin Enhances UVB-induced DNA-PK and H2A.X-Because we observed a strong increase in Ser 15 phosphorylation and total p53 levels in silibinin pre-treated UVB-exposed cells, which were abolished by LY294002 pre-treatment, we next analyzed the upstream molecules that could possibly be involved in p53 activation. ATM/ATR and DNA-PK are members of PI 3-kinase family that are early sensors of UVB-induced

FIG. 3. Silibinin alters UVB-mediated cell cycle regulation and enhances UV-induced apoptosis in JB6
cells. JB6 cells were treated with Me 2 SO or 100 M silibinin for 24 h and then either mock-irradiated or irradiated with 100 mJ/cm 2 UVB. Cells were harvested, stained with saponin/PI, and analyzed for cell cycle distribution by flow cytometry as described under "Experimental Procedures." A, representative scans of the fluorescence pattern for PI staining and cell cycle distribution, and B, the quantitative cell cycle distribution data. In similar treatments, cell lysates were prepared 24 h after irradiation and Western blotting was carried out for cleaved caspase-3 (C) and ␤-actin (D); or cells were harvested and processed for flow cytometric analysis of apoptotic cells by annexin V/PI staining (E). F, cellular morphology just before harvesting cells at the end of the treatments is shown at ϫ100 magnification. Quantitative data are presented as mean Ϯ S.E. of triplicate samples. $, p Ͻ 0.05; #, p Ͻ 0.005, and *, p Ͻ 0.001 indicate statistically significant differences between control and UVB alone or UVB alone and silibinin plus UVB-treated cells. Sb, silibinin.
DNA damage, and are also shown to be upstream kinases for p53 activation (9,13). Accordingly, next we assessed whether these kinases are modulated under our experimental conditions and whether such a modulation plays a role in p53 activation and apoptosis. UVB exposure of cells showed a very strong ATM serine 1981 phosphorylation at the earliest time point studied (6 h) after UVB irradiation that was sustained to a high level at 12 h but then diminished by 24 h after exposure; silibinin pre-treatment did not show any effect on this UVBcaused ATM phosphorylation (Fig. 6A). To further assess whether UVB-caused ATM/ATR activation is involved in p53-Ser 15 phosphorylation, we used caffeine, a known inhibitor of the ATM/ATR pathway (36). Pre-treatment of cells with caffeine (1 mM) showed a moderate inhibitory effect on UVBinduced p53-Ser 15 phosphorylation; however, it showed no effect on the silibinin-caused increase in p53-Ser 15 phosphorylation following UVB exposure, and in fact it showed a further increase (Fig. 6B). Conversely, in this particular experiment, like the one shown in Fig. 5C, cells pre-treated with LY294002 resulted in a strong decrease in p53-Ser 15 phosphorylation in silibinin pre-treated and UVB-exposed cells (Fig. 6B). Together, these results suggest that silibinin pre-treatment to UVB irradiation follows an alternate pathway to ATM/ATR, in activating p53 as well as in inducing apoptosis.
DNA-PK is another crucial DNA damage sensor that is activated in response to UV, and together with p53 is involved in inducing the apoptotic pathway (13). Therefore, we next assessed the effect of silibinin pre-treatment on UVB-induced DNA-PK (Fig. 6C) and its substrate H2A.X (Fig. 6D). UVB exposure of cells caused a moderate (at 6 and 12 h) to strong (at 24 h) activation of DNA-PK as observed by an increase in the processed, 180-kDa catalytic subunit of DNA-PK, whereas silibinin pre-treatment showed a very strong activation of DNA-PK at 12 h that remained sustained at 24 h (Fig. 6C). No significant changes were observed in the ϳ350-kDa fragment of DNA-PK in control or any of the UVB-irradiated samples with or without silibinin pre-treatment (data not shown). In other studies, at 6 and 12 h after UVB irradiation, we observed a moderate serine 139 phosphorylation of H2A.X (Fig. 6D), which is shown to be one of the substrates of DNA-PK as well as ATM/ATR (37). In accordance with a strong increase in DNA-PK activation, silibinin pre-treatment also strongly enhanced the UVB-caused H2A.X phosphorylation at both 6 and 12 h after exposure, which unlike UVB alone was sustained moderately even at 24 h (Fig. 6D). This increase in H2A.X phosphorylation by silibinin is possibly not through ATM/ATR activation, as silibinin pre-treatment did not show any further increase in UVB-induced ATM activation (Fig. 6A) and that caffeine pre-treatment had no effect on silibinin plus UVBinduced p53-Ser 15 phosphorylation (Fig. 6B). Overall, these results suggest that UVB activates both ATM/ATR (at early) and DNA-PK (at later) in a time-dependent manner, whereas silibinin pretreatment preferentially activates DNA-PK for p53-Ser 15 and H2A.X phosphorylation.
DNA-PK Is Involved in Silibinin-caused p53-Serine 15 Phosphorylation and Apoptosis-DNA-PK has been shown to initiate apoptotic signaling by preferentially binding to p53 and enhancing its Ser 15 phosphorylation (12). Because the above results suggested that silibinin enhances DNA-PK activation, we next assessed the binding of DNA-PK with p53 as well as its FIG. 4. PD98059 inhibits silibinin-induced UVB-caused activation of ERK1/2, but does not reverse silibinin-enhanced p53 stabilization and apoptosis. JB6 cells were pre-treated with the indicated doses of PD98059 for 2 h and then treated with or without silibinin (100 M) for 24 h. Cells were then irradiated or mock-irradiated with UVB (100 mJ/cm 2 ). Twenty-four hours later, cell lysates were prepared and Western blotting was carried out to look for ERK1/2 (phospho and total) (A), p53 (phospho and total) (B), cleaved caspase-3 (C), and ␤-actin (D); or cells were harvested after similar treatments and processed for flow cytometric analysis of annexin V/PI-stained apoptotic cells (E) as described under "Experimental Procedures." Quantitative data are presented as mean Ϯ S.E. of triplicate samples. kinase activity. DNA-PK pull-down assay showed that silibinin pre-treatment strongly increases its binding with p53 (Fig. 7A,  upper panel). Western immunoblotting the same samples further substantiated the observations shown in Figs. 1 and 2, which compared with UVB alone, silibinin plus UVB treatment increases the total p53 levels (Fig. 7A, lower panel). Furthermore, whereas no considerable change was observed in the total DNA-PK level (Fig. 7B), there was a strong increase in the processed 180-kDa catalytic subunit in silibinin pre-treated and UVB-exposed cells (Fig. 7C). Together, these observations showed a consistent trend in terms of an increase in both DNA-PK activity and binding with p53 in silibinin pre-treated plus UVB-irradiated cells. In further studies, the inhibitors of DNA-PK and ATM/ATR were used to assess their effect on DNA-PK activation where we did not observe any considerable change in the total levels of DNA-PK in different treatments (Fig. 7D). However, the 180-kDa catalytic subunit of DNA-PK was strongly increased by silibinin pre-treatment followed by UVB exposure (Fig. 7E). Furthermore, pre-treatment with LY294002 (50 M) inhibited the formation of the 180-kDa subunit following UVB exposure without or with silibinin pretreatment, but caffeine pre-treatment (1 mM) did not affect UVB or the silibinin plus UVB-caused increase in the DNA-PK catalytic subunit (Fig. 7E). In additional studies, as observed by substrate phosphorylation of a synthetic peptide, kinase activity performed by the DNA-PK pull-down assay further confirmed a stronger increase in DNA-PK activation by silibinin pre-treatment followed by UVB irradiation compared with UVB alone and its inhibition by the DNA-PK inhibitor LY294002 (Fig. 7F). Taken together, these results convincingly suggest that DNA-PK is a key player in the silibinin-mediated increase in p53 activation as well as apoptosis following UVB exposure. DISCUSSION Our earlier in vivo studies have shown the photoprotective efficacy of silibinin where it prevented acute UVB-caused DNA damage, cell proliferation, and apoptosis in mouse skin as well as chronic UVB-caused skin cancer in mice (2, 4 -6). We have also shown recently that silibinin modulates UVB-induced apoptosis in HaCaT and A431 cells, which is dependent on the level of UVB damage and cell type (5,6). As there is now increasing evidence to suggest a diversified role of signaling molecules in apoptotic cell death (38 -40), based on our findings in the present study, we propose a novel role of DNA-PK activation in the silibinin-induced increase in UVB-caused apoptosis in JB6 cells. As induction of apoptosis is considered one of the key mechanisms in protecting clonal expansion of initiated cells (41,42), our results show an increase in UVB-induced apoptosis by silibinin in JB6 cells that is in accord with our findings showing the strong preventive effect of this agent against photocarcinogenesis in mouse skin models.
The MAPK family of serine-threonine kinases is being increasingly recognized as signal transducers of genotoxic stress (43), which upon activation translocate to the nucleus and phosphorylate many transcription factors including p53 (44). Whereas MAPK/ERK1/2 has been shown to be critical for JB6 cell transformation and tumor promotion (18), under our experimental conditions, a further increase in UVB-caused . Six hours later, cells were harvested, DNA-PK was pulled down by DNA-cellulose, as described under "Experimental Procedures," and then the levels of p53 (A), lower panel, total p53 in cell lysate, total DNA-PK (ϳ350 kDa) (B), and processed DNA-PK (ϳ180 kDa) (C) were analyzed by Western blotting using appropriate antibodies. JB6 cells were pre-treated with either 50 M LY294002 or with 1 mM caffeine for 2 h and then with or without silibinin for 24 h. Cells were then irradiated with 100 mJ/cm 2 UVB, harvested 24 h after UVB exposure, and analyzed for total DNA-PK (D) and processed DNA-PK (E) by Western blotting. F, JB6 cells were pre-treated with 50 M LY294002 for 2 h and then with or without silibinin for 24 h. Cells were then irradiated with 100 mJ/cm 2 UVB and harvested 6 h after UVB exposure. Samples were processed for DNA-PK kinase activity as described under "Experimental Procedures." Kinase activity of DNA-PK was expressed as nanomoles of phosphate incorporated/min/mg of protein onto the peptide substrate. The data presented are representative of two independent experiments. Sb, 100 M silibinin; LY, 50 M LY294002. ERK1/2 activation and apoptosis following silibinin pre-treatment suggested its possible pro-apoptotic role. Interestingly, however, MEK1 inhibitor PD98059 even at 100 M concentration showed only a partial reversal in silibinin plus UVBcaused ERK1/2 phosphorylation together with a moderate reversal in p53-Ser 15 phosphorylation and accumulation. Consistent with these observations, PD98059 pre-treatment only moderately reversed caspase-3 cleavage with no effect on the enhancement of silibinin plus UVB-induced apoptosis. These findings clearly suggested that the observed increase in ERK1/2 phosphorylation in silibinin plus UVB-exposed cells is not largely via the MEK1 pathway, and that MEK1-mediated ERK1/2 activation does not play an important role in apoptosis induction, supporting the involvement of other pathway(s) in the observed biological effects.
It has also been reported that in response to UVB, Akt/PKB is phosphorylated at Ser 473 by upstream kinases such as PI 3-kinase and DNA-PK (44); Ser 473 phosphorylation of Akt is essential for its kinase activity (45,46). Although Akt activation is known as a survival and anti-apoptotic response, in the present study, we found that silibinin enhances UVB-induced Akt activation, which appears to be a stress response consistent with recent reports showing that Akt is one of the substrates of DNA-PK, a genotoxic stress response gene (44). To further support this notion and to identify the pathway(s) regulating Akt activation in JB6 cells under our study conditions, we used LY294002, which in addition to its inhibitory activity toward PI 3-kinase, is also known as a potent and selective inhibitor of DNA-PK (39), which belongs to the PI 3-kinase-like kinase having a PI 3-kinase homology domain (10,39). Pre-treatment of cells with LY294002 completely abolished the silibinin-caused increase in Akt-Ser 473 phosphorylation as well as p53-Ser 15 phosphorylation and accumulation by UVB irradiation. LY294002 pre-treatment of the cells also completely reversed the silibinin plus UVB-caused caspase-3 cleavage together with a strong and significant reversal in the silibinin-caused increase in apoptosis. Furthermore, at the highest dose used in the study (50 M), LY294002 pre-treatment also caused an almost complete reversal in silibinin plus UVB-induced ERK1/2 phosphorylation; this particular observation explains a lack of the strong inhibitory effect of the MEK1 inhibitor PD98059 on ERK1/2 phosphorylation and suggests this to be a major mechanism of ERK1/2 activation by silibinin plus UVB treatment of JB6 cells. In other studies, we also observed that silibinin plus UVB treatment of the cells up-regulates the DNA-PK kinase activity, which was reversed by LY294002 pre-treatment. All together, these results suggested that silibinin possibly sensitizes UVB-exposed JB6 cells to apoptosis by activation of the DNA-PK most likely in a p53-dependent manner.
Phosphorylation of p53 leading to its stabilization and activation is known to mediate an apoptotic response (47,48), and that skin cancer chemopreventive agents such as caffeine, genistein, and EGCG enhance apoptosis in vitro as well as in vivo by up-regulation of p53 (49 -51). Consistent with these studies, we also observed that UVB irradiation alone as well as silibinin pre-treatment followed by UVB exposure leads to strong p53-Ser 15 phosphorylation and accumulation; however, UVB alone exposure resulted in a faster accumulation of p53 that peaked at 12 h after exposure and started declining at 24 h, as compared with its slower accumulation in silibinin pre-treatment conditions that peaked at 24 h after UVB exposure and were sustained until 72 h after exposure. These differences in time kinetics of p53 activation suggest that UVB alone-and silibinin plus UVB-mediated p53 activation occurs possibly via different signaling intermediates or via further sensitization of similar molecules by silibinin treatment prior to UVB irradiation. This possibility is supported by the observation where pre-treatment of the cells with caffeine, an inhibitor of the ATM/ATR pathway, moderately reversed UVB-induced but not silibinin plus UVB-induced p53 phosphorylation; in fact, caffeine pretreatment more strongly enhanced silibinin plus UVB-induced p53-Ser 15 phosphorylation. Conversely, whereas LY294002 also showed a moderate reversal of UVB-induced p53 phosphorylation, almost complete reversal was observed with this inhibitor in the case of silibinin plus UVB-caused p53-Ser 15 phosphorylation. Taken together, these data suggest that while both ATM/ATR and DNA-PK are involved in UVB-caused p53 activation as an early response, silibinin pre-treatment predominantly acts via the DNA-PK activation pathway as a late and sustained response for p53 activation. This suggestion is further supported by the time kinetics studies showing a very strong increase in the 180-kDa subunit of DNA-PK in silibinin pre-treated cells starting at 12 h after UVB exposure, and its association with an increase in p53-Ser 15 phosphorylation and accumulation only from 24 h after UVB exposure again suggesting that DNA-PK could be the key player in enhanced p53 phosphorylation observed with silibinin treatment of the cells prior to UVB exposure.
DNA-PK, consisting of a catalytic subunit and DNA end binding complexes Ku70 and Ku80, is a serine-threonine kinase, whose activity is stimulated by double strand DNA breaks (10). It is present in high levels in the nucleus, and its activation has been shown to be dependent on conformational changes in the DNA (52). Although there are reports showing that the protein level of DNA-PK is not affected by genotoxic agents (53), some studies suggest that DNA-PK is a target for ICE-like protease activity during apoptosis (54). It has also been shown that DNA damage-induced apoptosis is abolished in DNA-PK-deficient cells and that p53-Ser 15 phosphorylation is critical for DNA-PK-mediated apoptosis induction (55). Although DNA-PK is primarily involved in DNA repair, it has also been implicated in cell cycle arrest and induction of apoptosis (10,55). Recent studies using DNA-PK knock-out mice have clearly delineated the role of DNA-PK in apoptosis, rather FIG. 8. Proposed mechanism of apoptosis induction involving the DNA-PK-p53 pathway by silibinin in UVB-exposed tumor promotion-sensitive JB6 cells. UVB activates DNA-PK, ATM/ATR, and ERK1/2 pathways; however, the presence of silibinin preferentially enhances the activation of the DNA-PK-p53 pathway, which is most likely aided by Akt and GSK-3 activation, and G 2 -M arrest for apoptosis in JB6 cells. Broken arrow represents that ATM/ATR is known to activate p53; however, in case of silibinin ϩ UVB, this is not the major mechanism for p53 activation in JB6 cells. than cell cycle regulation (56). In the present study, whereas UVB-caused S-phase arrest was shifted to a moderate G 1 and G 2 -M phase, apoptosis induction by silibinin pre-treatment was the most prominent biological response, suggesting that this could be the primary means by which silibinin eliminates cells with severe damage. Important to re-emphasize here is the ability of a cell to initiate apoptosis in response to severe DNA damage as a critical step for the prevention of cancer progression (57). In summary (Fig. 8), our results suggest that enhancement of UVB-induced apoptosis by silibinin could be one of the mechanisms of its efficacy against photocarcinogenesis and that silibinin treatment prior to UVB exposure causes preferential activation of DNA-PK rather than ATM/ATR leading to p53 phosphorylation and activation, which plays a crucial role in apoptosis induction in JB6 cells.