Enhancement of UVB-Induced Apoptosis by Apigenin in Human Keratinocytes and Organotypic Keratinocyte Cultures

Each year, over 1 million patients present with nonmelanoma skin cancers in United States, making it the most common form of cancer in this country (1). Extensive epidemiologic, clinical, and biological studies have concluded that overexposure to solar UV radiation is responsible for the development and progression of >90% of skin cancers (2–5). UVB radiation acts as a tumor initiator and promoter in the absence of any other agent by causing damage to critical macromolecules such as DNA, proteins, and lipids (ref. 6 and references therein). The cellular damage caused by exposure to low doses of UVB can usually be removed by repair; however, exposure to high doses of UVB overwhelm repair machinery, and apoptosis is initiated to remove damaged cells from the epidermis. Failure to remove the damage either by repair or apoptosis results in the retention of UVB-induced mutations that can lead to aberrant regulation of cell signaling ultimately responsible for tumor formation (7, 8). Education regarding the harmful effects of UVB radiation has been emphasized in recent years; however, primary prevention approaches have had limited success at reducing skin cancer incidence (9). The above findings emphasize the need to develop an agent that can prevent UVB-induced damage and/or the biological effects of UVB exposure by eliminating damaged cells with carcinogenic potential.

Currently there are two major strategies being used for the prevention of UVB-induced skin cancer. The first approach involves use of sun screens and other agents to prevent UVB-induced damage, thus reducing the formation of “initiated” cells. A second chemoprevention strategy aims to eliminate initiated cells with carcinogenic potential. This strategy is attractive clinically because it allows intervention during the longer period of time between development of preneoplastic foci and appearance of the tumor.

4′,5,7-trihydroxyflavone (apigenin) is a nonmutagenic, naturally occurring flavonoid found in a variety of fruits and leafy vegetables (10). Topical application of apigenin to mouse skin reduced the size and frequency of tumors induced by chemical carcinogens (11). Furthermore, Birt and coworkers (12) reported that apigenin applied topically to mouse skin inhibited UVB-induced skin carcinogenesis. Our group and others have shown that apigenin treatment of cells results in a wide variety of antitumorigenic and chemopreventive actions (13–18). More recently, we have focused our investigative efforts on determining the effects of apigenin treatment on keratinocytes exposed to UVB radiation (19, 20), the primary causative agent of skin cancers.

In the present study, we have used multiple methods to investigate the effect of apigenin treatment on UVB-induced apoptosis in three human keratinocyte models: HaCaT keratinocyte cells, primary human keratinocyte cultures derived from neonatal foreskin, and human organotypic keratinocyte cultures. We present results demonstrating that apigenin treatment enhanced UVB-induced apoptosis >2-fold in each of the models tested. We also show that enhancement of UVB-induced apoptosis by apigenin treatment involves both the intrinsic and extrinsic apoptotic pathways. The present work shows that the photochemopreventive properties of apigenin may be explained, in part, by its ability to enhance UVB-induced apoptosis in human keratinocytes. This is the first report demonstrating the effect of apigenin treatment on UVB-induced apoptosis in human keratinocytes.

Reagents and plasmids. DMSO and apigenin were purchased from Sigma Chemical Co. Hexadimethrine bromide (polybrene) was purchased from Fisher Scientific. Epidermal growth factor (EGF) was purchased from Biomedical Technologies, Inc. pBABE vector was purchased from Addgene, and plasmid 1764 was donated by Dr. Robert Weinberg (Massachusetts Institute of Technology, Cambridge, MA). Human Bcl-2 overexpression vector in pBABE was a generous gift from Dr. Vincent Cryns (Northwestern University, Chicago, IL). The Fas-associated death domain (FADDdn)-LZRS expression vector was kindly provided by Dr. Mitch Denning (Loyola University, Maywood, IL).

Cell culture. HaCaT cells, a spontaneously immortalized human keratinocyte cell line, were cultured as described previously (21). HaCaT cells were 85% to 90% confluent at the time of exposure to UVB or “sham” irradiation followed by treatment with 0, 10, or 20 μmol/L apigenin. HaCaT/pBABE-U6 and HaCaT/p53shRNA cells were generated and cultured as described by Boswell et al. (22) and kindly provided by Dr. Sam Lee (Massachusetts General Hospital and Harvard Medical School, Charlestown, MA).

The Phoenix-Ampho retroviral packaging cells used to transduce HaCaT cells with retrovirus encoding FADDdn or overexpression of Bcl-2 were obtained from American Type Culture Collection with permission from Dr. Gary P. Nolan (Stanford University Medical Center, Stanford, CA). The packaging cells were cultured as decribed by Sitailo et al (23). J2-3T3 fibroblasts were cultured in DMEM containing 10% heat-inactivated bovine serum (Life Technologies) and gentamicin/amphotericin solution (Cascade Biologics).

Isolation and culture of primary keratinocytes. Primary keratinocytes were isolated from normal human neonatal foreskin obtained from the Pathology Core Facility (Robert H. Lurie Comprehensive Center of Northwestern University) as described previously by Sitailo et al (23). Primary keratinocytes were then washed with M154CF keratinocyte medium containing 0.07 μmol/L CaCl₂, GA solution, and human keratinocyte growth supplements (Cascade Biologics). Experiments on primary keratinocytes were conducted on cells at passage ≤4.

Preparation of collagen gels and human organotypic keratinocyte cultures. Aliquots of 1 × 10⁶ J2-3T3 fibroblasts were resuspended in 10× reconstitution buffer, 10×DMEM (Sigma), and rat tail collagen I (BD Biosciences). The pH was adjusted to 7.4 with NaOH. This mixture was then dispensed into Falcon Cell Culture Inserts (pore size, 3.0 μm) placed in BD Companion Plates (BD Biosciences) and incubated at 37°C for 45 min. The collagen/fibroblast gels were cultured for 24 h submerged in E medium as described by Wu et al. (24). Primary keratinocyte cultures were trypsinized, and 1 × 10⁶ keratinocytes were resuspended in E medium containing 5 ng/mL EGF and seeded on top of each collagen gel. Keratinocytes form a monolayer sheet on top of the collagen/fibroblast matrix while cultured submerged in E medium for 2 d containing 5 ng/mL EGF with daily medium changes. Medium was then aspirated carefully, and the Falcon inserts were placed into a BD BioCoat Deep-well Plate designed for use with 6-well size BD Falcon Cell Culture Inserts and cultured at an “air/liquid interface” for 6 d with daily medium changes. On day 6, the human organotypic keratinocyte cultures were exposed to UVB or sham radiation and then incubated with 0, 25, or 50 μmol/L apigenin for 24 h; cultures were harvested on day 7.

Transduction of keratinocytes. Phoenix cells were transfected with 6 μg of either Bcl-2–overexpressing vector, FADDdn-expressing vector, or control vector (pBABE vector) using Lipofectamine 2000 (Invitrogen) for 6 h as directed by manufacturer. Phoenix cells were then selected for 2 to 3 d in medium containing 1 μg/mL puromycin. The virus was generated, harvested, and used for infection as described previously (23).

UVB irradiation and apigenin treatment. Cells were exposed to UVB and treated with apigenin as described previously by Tong et al. (19).

Western blot analysis. Cells were harvested and lysed in Triton Lysis Buffer as described previously (19). Whole cell lysates were collected and then sonicated thrice for 3 sec followed by centrifugation. Aliquots containing equal amounts of protein from samples were resolved by SDS-PAGE, transferred onto a polyvinylidene difluoride membrane, and blocked for 1 h using 5% nonfat dry milk. The membranes were probed with primary antibodies for poly(ADP)ribose polymerase (PARP), phospho-Ser¹⁵-p53 (Cell Signaling Technology), glyceraldehyde-3-phosphate dehydrogenase (GAPDH) from Chemicon, and p53-FL-393 (Santa Cruz Biotechnology) overnight at 4°C in 5%. After incubation with appropriate horseradish peroxidase–conjugated secondary antibodies, signals were detected with an enhanced chemiluminescence system (Amersham Biosciences).

Apoptotic assays. Cells were treated as indicated, harvested by trypsinization, and labeled using Annexin-V-APC as recommended by the manufacturer (BD Biosciences). Annexin-V–positive cells were measured using a Dako Cytomation CyAn flow cytometer (Flow Cytometry Core Facility). The average of replicate experiments was used to determine statistical significance by paired t test using Prism 3.0 software.

Histology and immunohistochemistry. Human organotypic keratinocyte cultures were fixed in 10% neutral buffered formalin, processed for histology, and embedded lengthwise in paraffin to ensure a cross-section cut. Each 5-μm section was dewaxed and stained with H&E to evaluate formation of sunburn cells. They stain with a small, condensed pyknotic nucleus and highly eosinophilic (pink) cytoplasm, and are a hallmark of apoptosis in the epidermis as described by Brash et al (25). Quantitative assessment of keratinocytes staining positive for sunburn cell formation from each sample was analyzed by blind counting the number of positive cells per millimeter of tissue. Statistical significance was determined by paired t test.

Immunofluorescence. To observe cytochrome c and Bax, keratinocytes were adhered to glass coverslips and then exposed to 0 or 750 J/m² UVB, and then incubated with 0, 10, or 20 μmol/L apigenin for 16 h. Cells were fixed in 2% paraformaldehyde for 10 min at room temperature and then ice-cold 90% methanol for 5 min. Cells were blocked in 5% goat serum (Life Technologies) in PBS containing 0.02% saponin (Sigma). Cells were then incubated with primary antibody to cytochrome c (BD Biosciences) and Bax (Upstate Cell Signaling) at 1:300 dilution in blocking buffer. The appropriate Alexa Fluor–conjugated secondary antibody (Invitrogen) was used at 1:300 dilution in blocking buffer followed by mounting the cells on slides using Vectashield containing 4′,6-diamidino-2-phenylindole (Vector Laboratories). Cell staining was observed by fluorescence microscopy using a ZEISS axiovert 200 man fluorescence microscope and photographed using an AxioCam HR camera.

UVB induces apoptosis in human keratinocytes. To test whether apigenin treatment of keratinocytes enhanced UVB-induced apoptosis, we first carried out dose-response studies to establish the UVB dose that would produce moderate (but not complete) levels of apoptosis in each of the three model systems tested. These studies were necessary to identify a UVB dose that could then be combined with apigenin treatment to determine whether apigenin enhanced UVB-induced apoptosis in each particular model. The dose of UVB required to induce intermediate levels of apoptosis in HaCaT cells was shown by Western blot analysis to be 300 J/m² UVB. This dose produced modest levels of apoptosis in HaCaT cells 8 hours postirradiation, as indicated by cleavage of PARP, a repair enzyme known to be cleaved during apoptosis (Fig. 1A).

Next, we exposed human primary keratinocyte cultures to various doses of UVB to establish the dose required to induce a moderate level of apoptosis in this system. Interestingly, testing of primary keratinocyte cultures isolated from different donors revealed varying levels of sensitivity to UVB radiation. In the majority of experiments with primary human keratinocyte cultures, we observed PARP cleavage at moderate levels 24 hours after exposure to 750 J/m² UVB (Fig. 1B).

Lastly, we exposed human organotypic keratinocyte cultures to different doses of UVB (0–2,000 J/m²) to determine the dose required to induce moderate levels of apoptosis in this third model system. Our results of Western blot analysis for PARP cleavage (Fig. 1C) show that a slightly higher dose of UVB (1,000 J/m²) was required to induce moderate levels of apoptosis in human organotypic keratinocyte cultures compared with the dose capable of inducing moderate levels of apoptosis in the monolayer primary keratinocyte cultures (750 J/m²; Fig. 1B). The barrier provided by the presence of multiple differentiated epidermal layers in this experimental model may account for the slightly higher dose of UVB necessary to induce moderate levels of apoptosis. It is of note that sensitivity to UVB-induced apoptosis in human organotypic keratinocyte cultures was correlated to the sensitivity of keratinocytes when they were exposed to UVB in monolayer culture (data not shown). The results presented in this section provided us with a specific dose of UVB radiation capable of inducing a moderate level of apoptosis in each experimental model. The dose of UVB identified above for each model system was used in subsequent experiments to determine the ability of apigenin to enhance UVB-induced apoptosis.

UVB-induced apoptosis is enhanced by apigenin treatment of monolayer keratinocyte cultures. The ability of apigenin treatment to enhance UVB-induced apoptosis was first examined in human HaCaT cells. We exposed HaCaT cells to 0 or 300 J/m² UVB radiation followed by incubation with 0, 10, or 20 μmol/L apigenin for 8 hours. Whole cell lysates were collected and analyzed by Western blot analysis for extent of PARP cleavage. UVB-irradiated keratinocytes showed a greater level of apoptosis as measured by PARP cleavage when treated with 20 μmol/L apigenin (Fig. 2A) compared with UVB exposure alone. Because Western blot analysis for PARP cleavage is difficult to quantitate, we used flow cytometry to determine the percentage of annexin-V–positive staining cells per 10,000 events as a second method to measure apoptosis. Figure 2B shows apoptosis as measured by annexin-V staining increased 1.8- and 2.2-fold in UVB-irradiated HaCaT cells treated with 10 and 20 μmol/L apigenin (P < 0.05), respectively, compared with UVB exposure alone.

To rule out the possibility that this phenomenon is limited to HaCaT cells, we next tested the ability of apigenin to enhance UVB-induced apoptosis in multiple preparations of primary human keratinocyte cultures. Primary keratinocyte cultures demonstrating moderate levels of apoptosis when irradiated with 750 J/m² UVB were used to evaluate the ability of apigenin to enhance UVB-induced apoptosis. Primary human keratinocytes were exposed to 0 or 750 J/m² UVB, then incubated with medium containing 0, 10, or 20 μmol/L apigenin for 24 hours, and harvested for Western blot analysis. Treatment of UVB-irradiated keratinocytes with either 10 or 20 μmol/L apigenin resulted in increased apoptosis as measured by PARP cleavage (Fig. 2C), compared with UVB alone. To generate quantitative values for the extent of enhancement of UVB-induced apoptosis in primary keratinocytes treated with apigenin, we also stained the cells with annexin-V and measured the percentage of apoptotic events by flow cytometry. UVB-irradiated cells incubated with 10 and 20 μmol/L apigenin resulted in a respective 1.9- and 2.7-fold increase in apoptosis (P < 0.05), compared with UVB exposure alone (Fig. 2D). The results presented in this section show that apigenin treatment leads to enhancement of the apoptotic response initiated by UVB exposure in primary human keratinocyte monolayer cultures.

UVB-induced apoptosis is enhanced by apigenin treatment in human organotypic keratinocyte cultures. Numerous studies have been conducted on the effectiveness of chemopreventive compounds in monolayer cell culture. Here, we cultured primary human keratinocytes isolated from neonatal foreskin in an organotypic setting to more accurately predict the effect of apigenin treatment on human skin exposed to UVB. Human organotypic keratinocyte cultures were exposed to 0 or 1,000 J/m² UVB then incubated with 0, 25, or 50 μmol/L apigenin for 24 hours. The epidermal layer was then separated from the dermal fibroblast collagen matrix, and tissue lysates were analyzed by Western blot. The data presented in Fig. 3A show that human organotypic keratinocyte cultures irradiated with 1,000 J/m² UVB followed by treatment with 50 μmol/L apigenin exhibited an increased level of cleaved PARP compared with UVB alone. Treatment of human organotypic keratinocyte cultures with 25 μmol/L apigenin had a negligible effect on UVB-induced apoptosis.

To further evaluate the extent of apoptosis in human organotypic keratinocyte cultures, samples were exposed to 0 or 1,000 J/m² UVB followed by treatment with 0, 25, or 50 μmol/L apigenin. The samples were fixed in 10% neutral buffered formalin, sectioned, and stained using H&E to evaluate architecture of the human organotypic keratinocyte cultures and the presence of apoptotic sunburn cells, which are characterized by a pyknotic nucleus and highly eosinophilic (pink) cytoplasm. Representative sections of human organotypic keratinocyte cultures exposed to 0 or 1,000 J/m² UVB in combination with 0, 25, or 50 μmol/L apigenin are shown in Fig. 3B with sunburn cells indicated by arrows. The number of apoptotic sunburn cells in human organotypic keratinocyte cultures were counted and then averaged as described in Materials and Methods to assess the level of enhancement of UVB-induced apoptosis in organotypic keratinocyte cultures (Fig. 3C). In agreement with the Western blot data (Fig. 3A), quantitative analysis of sunburn cell formation in human organotypic keratinocyte cultures showed that 25 μmol/L apigenin treatment had a negligible effect on UVB-induced apoptosis (1.1-fold change; P > 0.5), but when the human organotypic keratinocyte cultures were treated with 50 μmol/L apigenin after UVB exposure, apoptosis increased 2.4-fold (P < 0.01). These findings are the first to show enhancement of UVB-induced apoptosis by apigenin treatment of organotypic keratinocyte cultures.

Apigenin treatment of UVB-irradiated primary keratinocytes enhanced localization of Bax to the mitochondria, leading to subsequent cytochrome c release. We used indirect immunofluorescence to investigate the effect of apigenin treatment on localization of Bax in primary keratinocytes exposed to 0 or 750 J/m² UVB and treated with 0, 10, or 20 μmol/L apigenin for 16 hours. As seen in Fig. 4A, Bax is uniformly distributed in keratinocytes exposed to sham irradiation followed by treatment with 0, 10, or 20 μmol/L apigenin. Bax staining was more distinct and seemed more punctuate in keratinocytes exposed to UVB compared with sham-irradiated keratinocytes. Furthermore, primary keratinocytes exposed to UVB then incubated with medium containing 10 or 20 μmol/L apigenin showed a marked increase in the appearance of punctate localized staining of Bax compared with primary keratinocytes exposed to UVB then incubated with medium containing DMSO (vehicle control).

To evaluate whether the alterations in Bax localization led to the release of cytochrome c, we exposed primary human keratinocytes to 0 or 750 J/m² UVB radiation; then treated with 0, 10, or 20 μmol/L apigenin; and stained the cells with an anticytochrome c antibody. In keratinocytes exposed to 750 J/m² UVB followed by incubation with 0, 10, or 20 μmol/L apigenin, cytochrome c staining seemed more diffuse in a dose-dependent manner, indicating that cytochrome c was released from the mitochondria (Fig. 4B). No change in cytochrome c staining was observed in the sham-irradiated keratinocytes treated with 0, 10, or 20 μmol/L apigenin. The above findings suggest that apigenin treatment of UVB-irradiated cells enhanced localization of Bax to the mitochondria, which stimulated cytochrome c release and initiation of the intrinsic pathway of apoptosis.

Overexpression of the antiapoptotic protein Bcl-2 reduced the ability of apigenin to enhance UVB-induced apoptosis. To investigate further how apigenin enhances UVB-induced apoptosis, keratinocytes were infected with retrovirus overexpressing the antiapoptotic protein Bcl-2. Overexpression of Bcl-2 significantly reduced the level of apoptosis in each treatment group compared with cells transduced with control vector (Fig. 5A). When Bcl-2–overexpressing keratinocytes were exposed to UVB radiation and treated with medium containing 0, 10, or 20 μmol/L apigenin, we observed a 42%, 52%, and 42% reduction in apoptosis, respectively, compared with vector controls (compare lanes 3 and 4, 7 and 8, and 11 and 12 in Fig. 5A). Our results show that overexpression of the antiapoptotic protein Bcl-2 reduced the ability of apigenin to enhance UVB-induced apoptosis.

Expression of a dominant-negative FADD protected cells from enhancement of UVB-induced apoptosis by apigenin treatment. Transduction of HaCaT keratinocytes with a retroviral construct coding for FADDdn was used to elucidate the role of receptor-mediated apoptosis in this experimental system. The percentage of annexin-V–positive cells was lower in FADDdn keratinocytes than in vector controls for all treatment groups (Fig. 5B). In particular, when FADDdn keratinocytes were exposed to UVB then incubated in medium containing 0, 10, or 20 μmol/L apigenin, apoptosis was reduced by 35%, 42%, and 45% respectively, compared with vector controls (compare lanes 3 and 4, 7 and 8, and 11 and 12 in Fig. 5B). FADDdn keratinocytes were also more resistant to enhancement of UVB-induced apoptosis by apigenin treatment (compare lanes 4 and 12) than the corresponding treatment in cells transduced with vector control (compare lanes 3 and 11). Collectively, our results confirm that enhancement of UVB-induced apoptosis by apigenin treatment involves activation of both the receptor-mediated, extrinsic apoptotic pathway, and the intrinsic pathway.

Enhancement of UVB-induced apoptosis by apigenin treatment is independent of p53 status in HaCaT keratinocytes. HaCaT cells possess a mutant low-functioning p53 with an increased p53 protein half-life (ref. 22 and references therein). Therefore, we investigated whether the enhancing effect of apigenin treatment on UVB-induced apoptosis in HaCaT cells was p53-dependent by using two stable clones of HaCaT cells in which p53 expression was ablated (Fig. 6A) with shRNA-targeting p53 or control shRNA (kindly provided by Dr. Sam Lee; ref. 22). HaCaT/p53shRNA cells were exposed to UVB or sham radiation, incubated in medium containing 0 or 20 μmol/L apigenin for 8 hours, and then harvested for Western blot analysis. Cell lysates from UVB-irradiated, apigenin-treated, and UVB + apigenin–treated HaCaT/pBABE-U6 cells exhibited elevated levels of phospho-Ser¹⁵ p53 compared with untreated HaCaT/pBABE-U6 cells (Fig. 6A). In both HaCaT/p53shRNA clones (lanes 8 and 12) and in HaCaT/pBABE-U6 cells (lane 4), we observed enhancement of UVB-induced apoptosis by apigenin treatment as evidenced by increased PARP cleavage (Fig. 6A).

To quantitate the effect of UVB, apigenin, and UVB + apigenin treatment on apoptosis, we measured annexin-V+ cells by flow cytometry (Fig. 6B). Apigenin treatment significantly enhanced UVB-induced apoptosis in both clone 15 and clone 22 of HaCaT/p53shRNA cells, as well as in HaCaT/pBABE-U6 cells (lane 8 versus 6, lane 12 versus 10, and lane 4 versus 2, respectively; Fig. 6B). Although it is possible that p53 modulates the enhancement of UVB-induced apoptosis by apigenin treatment, our findings show that the enhancement of UVB-induced apoptosis by apigenin treatment occurs largely via a p53-independent pathway.

More than 15 years ago, Wei and coworkers (11) showed that topical application of apigenin effectively inhibited chemical carcinogenesis in mouse skin. Subsequent studies by Birt and coworkers (12) revealed that apigenin applied topically also inhibited UVB-induced photocarcinogenesis in SKH-1 hairless mice. Further studies by others have investigated which molecular processes are affected by apigenin treatment, thus providing additional evidence that apigenin is an effective chemopreventive agent. For example, Kuo and coworkers (14) have shown the ability of apigenin to inhibit transformation. Experiments by Way et al. and Yin and coworkers (17, 18) showed that apigenin induced growth inhibition, cell cycle arrest, and apoptosis in breast cancer cell lines. Similiarly, Wang and coworkers (26) reported that apigenin induced G₂-M arrest in human colon cancer cell lines. Previous studies in our laboratory have shown that apigenin treatment increased wild-type p53 stability in keratinocytes (15). We also showed that apigenin can induce reversible G₂-M cell cycle arrest (27), and that G₂-M arrest was accompanied by inhibition of the p34(cdc2) cyclin-dependent kinase protein level and activity in a p21(waf1)-independent manner (28). Van Dross et al. (29) showed apigenin treatment can activate extracellular signal-regulated kinase 1/2, p38 mitogen-activated protein kinase pathways, and downstream transcriptional activators in 308 keratinocytes and human colon carcinoma cell line HCT116.

More recently, our group has focused on the effects of apigenin treatment on keratinocytes exposed to UVB radiation (19, 20). These reports have shown that apigenin treatment down-modulates both basal and UVB-induced cyclooxygenase (COX)-2 expression in keratinocytes. A number of other laboratories have shown that selective COX-2 inhibition can reduce tumor formation (30–34), presumably by increasing apoptosis (35, 36) in cells that have carcinogenic potential. A recent study by Akunda and colleagues (34) reported COX-2 deficiency significantly increased UVB-induced epidermal apoptosis, in agreement with previous findings by Tripp et al. (32).

Two principal apoptotic pathways have been characterized in the literature: the death receptor-mediated extrinsic pathway and the stress-mediated intrinsic pathway as reviewed by Fulda et al. (37). UVB radiation induces both intrinsic and extrinsic pathways of apoptosis. Sitailo and coworkers (23) showed the critical involvement of the intrinsic mitochondria-mediated apoptotic pathway in response to UVB radiation in keratinocytes by showing that expression of dominant-negative caspase-9 led to almost complete inhibition of caspase-3, caspase-8, and caspase-9 activation typically induced by UVB. In the present study, incomplete protection from enhancement of UVB-induced apoptosis in keratinocytes overexpressing the antiapoptotic protein Bcl-2 allows for the possibility that enhancement of UVB-induced apoptosis by apigenin treatment involves potentiation of receptor-mediated apoptosis. A recent report by Horinaka et al. (38) has shown that apigenin can sensitize cells to receptor-mediated cell death in various tumor cell lines if used in combination with tumor necrosis factor–related apoptosis-inducing ligand. It is possible that direct activation of death receptors, such as Fas and TNFR1, caused by UVB radiation (39–41) could be involved in the enhancement of UVB-induced apoptosis observed in irradiated keratinocytes treated with apigenin. We show herein that FADDdn expression in keratinocytes significantly reduced apigenin's enhancement of UVB-induced apoptosis compared with keratinocytes transduced with control vector. The reduced ability of apigenin to enhance UVB-induced apoptosis in keratinocytes expressing FADDdn suggests that the extrinsic apoptotic pathway plays a critical role in apigenin's enhancement of UVB-induced apoptosis.

A number of laboratories have shown that treatment of keratinocytes with natural products such as sanguinarine (42), silibinin (43), curcumin (44), and caffeine (45, 46) can also enhance UVB-induced apoptosis in keratinocytes. Induction of apoptosis in cells repeatedly exposed to UVB exposure is thought to be an important protective mechanism aimed at removing irreversibly damaged and potentially carcinogenic keratinocytes from the epidermis. In accordance with the above studies, we show herein that the natural bioflavonoid apigenin also is capable of enhancing UVB-induced apoptosis. We have identified additional pathways through which apigenin exerts its photochemopreventive effects in human keratinocytes. We observed enhancement of UVB-induced apoptosis by apigenin treatment in HaCaT cells, in primary human keratinocyte cultures, and in human organotypic keratinocyte cultures.

Our laboratory has previously shown that apigenin treatment induced the stabilization and transactivational activity of the tumor suppressor p53 (15). In addition, a number of other laboratories have reported that the chemopreventive properties of apigenin can occur through p53-independent pathways as well. For example, Shukla and coworkers (16) showed that apigenin treatment induced a dose-dependent increase in apoptosis and a dose-dependent decrease in protein expression of cyclin D1, cyclin D2, and cyclin E in DU145 prostate cancer cell lines that possess a mutant form of p53. In addition, apigenin treatment increased apoptosis and expression of WAF1/p21, KIP/p27, INK4a/p16, and INK4c/p18 compared with vehicle-treated controls similarly in both 22Rv1 and PC-3 (p53 null) tumor xenograft mice in a dose-dependent manner, suggesting that apigenin can exert its chemopreventive properties independently of p53 status (47). These findings collectively support the premise that although apigenin treatment mediates many p53-dependent pathways, apigenin can also induce apoptosis and cell cycle arrest in a p53-independent manner. In agreement with these findings, we have shown herein that apigenin treatment enhanced UVB-induced apoptosis in human HaCaT keratinocytes that have had p53 ablated (Fig. 6). Our findings also show that although it is possible that p53 modulates enhancement of UVB-induced apoptosis by apigenin, it seems to occur largely via a p53-independent pathway. Furthermore, apigenin treatment enhanced UVB-induced apoptosis in stable HaCaT/p53shRNA cells, demonstrating that enhancement of UVB-induced apoptosis by apigenin treatment is not a p53-dependent phenomenon.

Taken together, our results show that apigenin enhances UVB-induced apoptosis in human keratinocytes through both the extrinsic and the intrinsic apoptotic pathways. This conclusion is based on the following observations: (a) Apigenin treatment enhanced UVB-induced apoptosis in each model of human keratinocytes as evaluated by Western blot analysis, annexin-V staining by flow cytometry, and/or quantification of sunburn cells; (b) Apigenin treatment of human keratinocytes facilitates changes in Bax localization and cytochrome c release when exposed to UVB radiation; (c) Overexpression of Bcl-2 incompletely protects keratinocytes from enhancement of UVB-induced apoptosis when treated with apigenin; and (d) FADDdn expression partially blocked apigenin enhancement of UVB-induced apoptosis. Although numerous studies have evaluated the effects of apigenin on various cell types, we believe this is the first study to report the ability of apigenin to enhance UVB-induced apoptosis in human keratinocyte cultures.

Grant support: NIH grants RO1 CA072987 and RO1 CA104768 (J.C. Pelling), RO1 AR41836 (K.J. Green), the Zell foundation (J.C. Pelling is a Zell Scholar), and supported in part by NIH grant T32 ES07079 (A.O. Abu-Yousif).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.