AMP-activated Protein Kinase Contributes to UV- and H2O2-induced Apoptosis in Human Skin Keratinocytes*

AMP-activated protein kinase or AMPK is an evolutionarily conserved sensor of cellular energy status, activated by a variety of cellular stresses that deplete ATP. However, the possible involvement of AMPK in UV- and H2O2-induced oxidative stresses that lead to skin aging or skin cancer has not been fully studied. We demonstrated for the first time that UV and H2O2 induce AMPK activation (Thr172 phosphorylation) in cultured human skin keratinocytes. UV and H2O2 also phosphorylate LKB1, an upstream signal of AMPK, in an epidermal growth factor receptor-dependent manner. Using compound C, a specific inhibitor of AMPK and AMPK-specific small interfering RNA knockdown as well as AMPK activator, we found that AMPK serves as a positive regulator for p38 and p53 (Ser15) phosphorylation induced by UV radiation and H2O2 treatment. We also observed that AMPK serves as a negative feedback signal against UV-induced mTOR (mammalian target of rapamycin) activation in a TSC2-dependent manner. Inhibiting mTOR and positively regulating p53 and p38 might contribute to the pro-apoptotic effect of AMPK on UV- or H2O2-treated cells. Furthermore, activation of AMPK also phosphorylates acetyl-CoA carboxylase or ACC, the pivotal enzyme of fatty acid synthesis, and PFK2, the key protein of glycolysis in UV-radiated cells. Collectively, we conclude that AMPK contributes to UV- and H2O2-induced apoptosis via multiple mechanisms in human skin keratinocytes and AMPK plays important roles in UV-induced signal transduction ultimately leading to skin photoaging and even skin cancer.

Ultraviolet (UV) spectrum is divided into UVC (200 -280 nm), UVB (280 -320 nm), and UVA (320 -400 nm). UVB and UVA are of environmental significance and social concern, because UVC is filtered through the ozone layer. UV penetrates into the papillary area of the dermis (ϳ0.2 mm) and induces DNA damages of residing keratinocytes and dendritic cells. They are perturbed both phenotypically and functionally undergoing apoptosis upon UV radiation (1,2). Previous studies in human keratinocytes in vitro and in human skin in vivo have demonstrated that UV response comprises trans-activation of cell surface growth factor receptors, such as EGFR, 3 and their attendant downstream signal transduction machinery such as MAPK and phosphatidylinositol 3-kinase/AKT (2)(3)(4)(5)(6). Although MAPK, including JNK and p38, is responsible for UV-induced cell apoptosis and skin aging, other cellular signals such as AKT (also known as protein kinase B) serve as survival signals in skin cells to fight against UV-induced widespread cell death (1)(2)(3)(4). However, the possible involvement of other signals, AMP-activated protein kinase or AMPK, for example, in UV-induced cell apoptosis leading to skin aging or cancer has not been fully studied.
AMPK is a heterotrimeric serine-threonine kinase that senses depletion of intracellular energy and responds by stimulating catabolic pathways that generate ATP (5,6). One mechanism for sensing cellular energy levels involves allosteric activation of AMPK. Under conditions in which cellular energy demands are increased (such as enhanced cellular activities or cellular stresses) or when fuel availability is decreased (because of a reduced rate of glucose uptake), intracellular ATP is reduced and AMP levels rise. AMP then allosterically activates AMPK. In addition to allosteric activation, AMPK activity can be regulated by a mechanism involving covalent modification through the addition of a phosphate group by other molecules such as LKB1 and CaMK or calmodulin-dependent protein kinase (5)(6)(7)(8)(9)(10). A number of stimuli (11,12) have been found that can induce AMPK activation. However, the question how UV radiation, the major cause of skin aging and skin cancer, activates AMPK remains unknown.
It is well established that the key function of AMPK is to regulate the energy balance within the cell. Once activated, AMPK phosphorylates downstream substrates, the overall effect of which is to switch off ATP consuming pathways (e.g. * This work was supported, in whole or in part, by National Institutes of Health Grant P20 RR016457 from the INBRE Program of the National Center for Research Resources. This work was also supported by grants for biomedical research from the Rhode Island Foundation and the Slater Center for Environmental Biotechnology. 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. 1 Both authors contributed equally to this paper. 2 To whom correspondence should be addressed: Providence College, 549 River Ave., Providence, RI 02918-0001. Tel.: 401-865-2507; Fax: 401-865-1438; E-mail: yswan@providence.edu. 3 The abbreviations used are: EGFR, epidermal growth factor receptor; MAPK, mitogen-activated protein kinase; AMPK, AMP-activated protein kinase; mTOR, mammalian target of rapamycin; MEF, mouse embryonic fibroblasts; MTT, 3-(4,5-dimethylthylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; ROS, reactive oxygen species; FACS, fluorescence-activated cell sorter; NAC, N-acetyl-L-cysteine; siRNA, small interfering RNA; PBS, phosphate-buffered saline; CHAPS, 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonic acid; AICAR, 5-aminoimidazole-4-carboxamide-1-␤-4-ribofuranoside; S6K, S6 kinase.
fatty acid synthesis and cholesterol synthesis) and to switch on catabolic pathways that generate ATP (e.g. fatty acid oxidation and glycolysis). Activation of AMPK requires phosphorylation of Thr 172 in the activation loop of ␣ subunit by upstream AMPK kinase (6,(13)(14)(15). AMPK activation also triggers a phosphorylation cascade that regulates the activity of various downstream targets, including transcription factors, enzymes, and other regulatory proteins, such as mTOR pathways (16), p53 (17), and p38 (18). However, the possible role of AMPK in UV-induced signal transduction and skin aging or cancer remains to be elucidated.
In this study, we found for the first time that UV and H 2 O 2 induce AMPK activation and downstream ACC and PFK2 phosphorylation in cultured human skin keratinocytes and reactive oxygen species (ROS)-mediated EGFR trans-activation is involved in LKB1/AMPK activation. Using AMPK inhibitor (Compound C), AMPK␣ siRNA knockdown as well as AMPK activator AICAR, we found that AMPK is actively involved in UV-and H 2 O 2 -induced signal transduction and skin cell damage probably by positively regulating downstream p38, p53 activation, and inhibiting mTOR activation. Our study provides new insights into understanding the cellular and molecular mechanisms involved in UV-induced skin cell damage leading to skin aging and skin cancer.

EXPERIMENTAL PROCEDURES
UV Light Apparatus-As previously reported (19,20), the UV-irradiation apparatus used in this study consisted of four F36T12 EREVHO UV tubes. A Kodacel TA401/407 filter was mounted 4 cm in front of the tubes to remove wavelengths Ͻ290 nm. Irradiation intensity was monitored using an IL443 phototherapy radiometer and a SED240/UV/W photodetector. Before UV irradiation, cells were washed with 1 ml of phosphate-buffered saline (PBS) and changed to fresh 0.5 ml of PBS for each well. Cells were irradiated at the desired intensity without a plastic dish lid. After UV irradiation, cells were returned to incubation in basal medium with treatments for various time points prior to harvest. Mouse skin dendritic cells (XS 106 cell line) were cultured in 10% fetal bovine serum in RPMI 1640 with granulocyte-macrophage colonystimulating factor (Sigma).
Western Blot Analysis-As reported previously (20,22), cultured cells with and without treatments were washed with cold PBS and harvested by scraping into 150 l of RIPA buffer with protease inhibitors. 20 -40 g of proteins were separated by SDS-PAGE and transferred onto polyvinylidene difluoride membrane (Millipore, Bedford, MA). After blocking with 10% milk in Tris-buffered saline, membranes were incubated with specific antibodies in a dilution buffer (2% bovine serum albumin in Tris-buffered saline) overnight at 4°C followed by horseradish peroxidase-conjugated anti-rabbit or anti-mouse IgG at appropriate dilutions and incubated at room temperature for 1 h. Antibody binding was detected using a enhanced chemiluminescence (ECL) detection system from GE Healthcare following the manufacturer's instructions and visualized by fluorography with Hyperfilm.
In Vitro Kinase Assay for mTOR Activity-The in vitro kinase assay for mTOR activity was performed as described previously (23). Briefly, HaCaT cells (2 ϫ 10 6 ) were lysed in 200 l of lysis buffer containing 40 mM HEPES, pH 7.5, 120 mM NaCl, 0.3% CHAPS, 1 mM EDTA, 2.5 mM sodium pyrophosphate, 1 mM ␤-glycerophosphate, 1 mM Na 3 VO 4 , and protease inhibitor mixture. Cell lysate (500 g) was incubated with anti-mTOR antibody (Santa Cruz Biotechnology, Santa Cruz, CA) and 30 l of Protein A/G-agarose beads at 4°C for 3 h. Following the incubation, beads were washed 3 times with lysis buffer and 2 times with kinase buffer containing 25 mM HEPES, pH 7.4, 50 mM KCl, 20% glycerol, 10 mM MgCl 2 , 4 mM MnCl 2 , 1 mM dithiothreitol, 1 mM glycerophosphate, and 1 mM Na 3 VO 4 . After the final wash, beads were assayed for kinase activity against purified 4E-BP1 substrate by adding 4E-BP1 substrate (Santa Cruz Biotechnology, Santa Cruz, CA) to the beads for 30 min at 30°C. The reactions were terminated by boiling in the presence of 1ϫ SDS sample buffer. The samples were subjected to SDS-PAGE, and phosphorylation of 4E-BP1 was detected by Western blotting using anti-phospho-4E-BP1 (Ser 65 ) antibody.
AMPK␣ RNA Interference Experiments-As described previously (22), siRNA for AMPK␣1/␣2 (sc-45312) was purchased from Santa Cruz Biotechnology. HaCaT cells were cultured in a complete medium that did not contain antibiotics for 4 days. 50 ϫ 10 s cells were seeded in a 6-well plate 1 day prior to transfection and cultured to 60 -70% confluence the following day. For RNA interference experiments, 6.25 l of Lipofectamine TM LTX together with 2.5 l of PLUS TM Reagent (Invitrogen) was diluted in 90 l of Dulbecco's modified Eagle's medium for 5 min at room temperature. Then, 8 l of AMPK␣ siRNA was mixed with Dulbecco's modified Eagle's medium containing Lipofectamine together with PLUS reagent and incubated for 30 min at room temperature for complex formation. Finally, the complex was added to the wells containing 2 ml of medium with

W I T H D R A W N W I T H D R A W N
the final AMPK␣ siRNA concentration of 100 nM. AMPK␣ protein expression was determined by Western blot.
Cell Viability Assay (MTT Dye Assay)-Cell viability was measured by the 3-(4,5-dimethylthylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) method (20). Briefly, cells were collected and seeded in 96-well plates at a density of 2 ϫ 10 5 cells/cm 2 . Different seeding densities were optimized at the beginning of the experiments (data not shown). After incubation for 24 h, cells were exposed to fresh medium containing reagents at 37°C. After incubation for up to 24 h, 20 l of MTT tetrazolium (Sigma) salt dissolved in Hank's balanced salt solution at a concentration of 5 mg/ml was added to each well and incubated in a CO 2 incubator for 4 h. Finally, the medium was aspirated from each well and 150 l of dimethyl sulfoxide (Sigma) was added to dissolve formazan crystals and the absorbance of each well was obtained using a Dynatech MR5000 plate reader at a test wavelength of 490 nm with a reference wavelength of 630 nm.
Assessment of the Percentage of Apoptotic Cells-To detect apoptotic cells (20), cells were stained with DNA binding dye Hoechst 33342 (Sigma). After the cells were exposed to UV and the test compounds for the allotted time periods, they were fixed with 4% formaldehyde in PBS for 10 min at 4°C, and then washed with PBS. To stain the nuclei, cells were incubated for 20 min with 20 g/ml of Hoechst 33342. After washing with PBS, the cells were observed under a fluorescence microscope (Zeiss Axiophoto 2, Carl Zeiss, Germany). Cells exhibiting condensed chromatin and fragmented nuclei were scored as apoptotic cells. A minimum of 200 cells were scored from each sample.
Measurement of Keratinocytes Mitochondrial Membrane Potential-HaCaT cell mitochondrial membrane potential (⌬⌿ m ) was assessed with fluorescent probe  OCTOBER

W I T H D R A W N W I T H D R A W N
JC-1 (20). At 490 nm, cells with depolarized mitochondria contained JC-1 predominantly in monomeric form and fluoresced green. Cells with polarized mitochondria predominantly contain JC-1 in aggregate form, and mitochondria fluoresce redorange. HaCaT cells were incubated with 5 M JC-1 (Invitrogen) for 30 min at 37°C, washed, and fluorescent images were visualized by a Zeiss fluorescence microscope with excitation at 490 nm and emission at 520 nm (monomeric form for depolarized ⌬⌿ m ) and 590 nm (aggregate form for polarized ⌬⌿ m ).
HaCaT cells with polarized mitochondria were seen with distinct mitochondria fluorescing red-orange, and with depolarized mitochondria, cell cytoplasm and mitochondria appeared green. The aquired signal was analyzed with image analysis software (20). A minimum of six fields were selected and average intensity for each region was quantified. The ratio of J-aggregate to JC-1 monomer intensity for each region was calculated. A decrease in this ratio was interpreted as loss of ⌬⌿ m , whereas an increase in the ratio was interpreted as gain in ⌬⌿ m .
ROS Detection-ROS generation was detected by FACS analysis as described previously (24,25). Briefly, cultured human skin keratinocytes (HaCaT cells) were loaded with 1 M fluorescent dye dihydrorhodamine 2 h before UV radiation, which reacts with ROS in cells and results in a change of fluorescence. After being treated with UV with or without reagents for the desired time points, keratinocytes were trypsinized, suspended in ice-cold PBS, and fixed in 70% ethyl alcohol in Ϫ20°C. The changes in fluorescence in drug-treated cells were quantified by FACS analysis. Induction of ROS generation was expressed in arbitrary units.
Statistical Analysis-The values in the figures are expressed as the mean Ϯ S.E. The figures in this study were representative of more than 3 different experiments. Statistical analysis of the data between the control and treated groups was performed by a Student's t test. Values of p Ͻ 0.05 were considered statistically significant.

UV Radiation and H 2 O 2 Induce AMPK Activation in Cultured Human Skin
Keratinocytes-To investigate the role of AMPK in UV signaling, we first tested whether UV or H 2 O 2 induces AMPK activation using cultured human skin keratinocytes (HaCaT cells). The results showed that UV radiation induces AMPK␣ phosphorylation in a dose (Fig. 1, A and B) and time (Fig. 1, C and D)-dependent manner. Similarly, H 2 O 2 induces AMPK␣ activation in a dose-and time-dependent manner ( Fig. 1, E-H). Furthermore, as expected, AMPK activator 5-aminoimidazole-4-carboxamide-1-␤-4-ribofuranoside (AICAR) also induces AMPK activation in a dose- (Fig. 1, I and J) and time (Fig. 1, K and L)-dependent manner. AMPK␣-specific siRNA down-regulates AMPK␣ expression and largely inhibits UV-and H 2 O 2 -induced AMPK activation (Fig. 1, M  and N). UV also induces AMPK activation in cultured human skin fibroblasts (Fig. 1O) and dendritic cells (XS 106 cell line) (Fig. 1P).
ROS-mediated EGFR Activation Is Involved in UV-induced LKB1/AMPK Activation-The data above show that both UV and H 2 O 2 induce AMPK activation (AMPK␣ phosphorylation at Thr 172 ) in HaCaT cells. However, the cellular signals involved in this AMPK activation are not fully studied. Previous studies using human skin keratinocytes and dendritic cells have revealed a important role of EGFR in UV-induced cellular signals (3,26), studies also show a critical role of LKB1 for AMPK activation. As expected, in this study, we observed that both UV and H 2 O 2 induce EGFR activation in HaCaT cells (Fig. 2, A and  B). Interestingly, UV radiation also induces LKB1 phosphorylation in a time-and dose-dependent manner in HaCaT cells (Fig. 2, C and D). EGFR inhibitor PD 153035 and AG 1478 inhibit UV-induced AMPK and LKB1 activation (Fig. 2E). EGFR ligand, EGF, also induces LKB1/AMPK activation, which is blocked by PD 153035 (Fig. 2F). To further confirm the key role of EGFR in UV-induced AMPK activation, EGFR knockout MEFs were used. As shown in Fig. 2G, UV induces LKB1/ AMPK activation in wild type but not in EGFR knock-out MEFs. Furthermore, the induction of AMPK is also inhibited by pretreatment with the antioxidant NAC and pyrrolidine dithiocarbamate (Fig. 2H) in UV-treated HaCaT cells. Antioxidant NAC and EGFR inhibitor PD 153035 also reduce H 2 O 2 -induced LKB1/AMPK activation (Fig. 2I). As expected, UV-and H 2 O 2 -induced ROS production is inhibited by NAC pre-treatment (Fig. 2J). Collectively, our data suggest that ROS-mediated EGFR trans-activation is involved in UV-induced LKB1/ AMPK activation.
AMPK Is Involved in UV-and H 2 O 2 -induced p38 Activation-Because previous studies have suggested that p38 MAPK is a downstream signal of AMPK upon various stimuli (18,27,28), and UV induces p38 activation (24), next we tested the possible role of AMPK in UV-and H 2 O 2 -induced p38 activation. As  OCTOBER Fig. 3, A and B, compound C (AMPKi), an AMPK inhibitor, largely impairs the activation of p38 MAPK upon UV radiation, whereas SB203580, a p38 inhibitor, does not affect AMPK activation. Furthermore, AMPK also positively regulates H 2 O 2 -induced p38 activation, because AMPKi inhibits H 2 O 2 -induced p38 activation, whereas AICAR, an AMPK activator, enhances it (Fig. 3, C and D). AICAR alone also induces p38 activation, which is reversed by compound C (Fig. 3, E and F). To further confirm the role of AMPK in UV-     OCTOBER

W I T H D R A W N W I T H D R A W N
induced p38 activation, AMPK␣ siRNA was used. As demonstrated in Fig. 3G, activation of p38 in response to UV or H 2 O 2 is inhibited in AMPK␣ siRNA-treated HaCaT cells. Collectively, our data suggest that AMPK is involved in UV-and H 2 O 2 -induced p38 activation.
AMPK Positively Regulates UVand H 2 O 2 -induced Phosphorylation of p53-p53 tumor suppressor exerts anti-proliferative effects, including growth arrest, apoptosis, and cell senescence, in response to UV radiation (29). Its activation form (both phosphorylation and acetylation) is regulated by a number of proteins such as ATM (30), Chk1/2 (30), M2M, p300 (31), and SIRT1 (15,32). However, whether AMPK also affects UV-induced p53 activation is unknown. As shown in Fig. 4, A and B, UV-and H 2 O 2 -induced phosphorylation of p53 is inhibited by EGFR inhibitors (PD and AG) and antioxidant NAC, suggesting that the upstream components of AMPK regulate p53 activation. Furthermore, AMPK inhibitor compound C (AMPKi) and p38 inhibitor SB 203580 inhibit UV-induced phosphorylation p53 (Fig. 4,  C and D), suggesting the positive regulatory role of AMPK activation in UV-induced p53 phosphorylation, which is consistent with recent studies demonstrating that AMPK directly phosphorylates p53 at a specific site (33). To further confirm this notion, siRNA-mediated AMPK␣ knockdown was used. As shown in Fig. 4E, phosphorylation of p53 in AMPK␣ knockdown cells is impaired, compared with control siRNA-treated cells. Collectively, our data suggest that AMPK positively regulates UV-and H 2 O 2 -induced phosphorylation of p53.
UV Radiation Induces ACC and PFK2 Phosphorylation in an AMPKdependent Manner-AMPK was originally perceived as a regulator of cellular energy balance that acts in a cell-autonomous matter. AMPK appears to play as important a role in whole body energy balance as it does in cellular energy balance. Once it is activated, it helps to   HaCaT cells with or without AMPK siRNA were treated with UV radiation (25 mJ/cm 2 ) or AICAR (2 mM), and cell viability was detected by MTT assay after 24 h (E), cell apoptosis was detected by Hoechst 33342 assay after 18 h (F). HaCaT cells were pretreated with p38 inhibitor (SB 203580, 10 M) for 1 h, followed by UV radiation (25 mJ/cm 2 ) or AICAR (2 mM), cell viability was detected by MTT assay after 24 h (G). Wild type and p53 knock-out MEFs were treated with UV radiation (25 mJ/cm 2 ) or AICAR (2 mM), cell viability was detected by MTT assay after 24 h (H). *, p Ͻ 0.05 versus untreated group. #, p Ͻ 0.05 versus the same time point of UV or AICAR-treated group. Data are presented as the mean Ϯ S.E. for three independent experiments. For the fluorescence experiment, a minimum of six random fields and 200 cells per group were selected and average fluorescence intensity for each group was quantified. Magnification: C, 1:400. For the apoptosis (Hoechst assay) (F) experiment, a minimum of 10 random fields and 500 cells were counted for the apoptotic death rate.

DISCUSSION
As well as the numerous metabolic targets of AMPK for which it is best known, AMPK has many other downstream effectors. Recent studies have revealed that AMPK inhibits cell growth and proliferation, and also positively regulates apoptosis (33). These findings are not surprising, given that cell growth, DNA replication, and mitosis are all major consumers of ATP, and also that the upstream regulator of AMPK, LKB1, is known to be a tumor suppressor. In this study, we discovered that UV and H 2 O 2 induce AMPK activation in cultured skin cells (Fig. 1). ROS-mediated EGFR trans-activation is involved in UV-induced LKB1/AMPK activation (Fig. 2). We also observed that AMPK activation contributes to UV-and H 2 O 2induced skin cell damage (Fig. 7) probably by positively regulating p38 (Fig. 3) and p53 phosphorylation (Fig. 4) and inhibiting the mTOR pathway (Fig. 6). Collectively, we propose cell signaling pathways involved in UV-induced AMPK activation, as depicted in Fig. 8.
Although we found that UV-induced AMPK activation consistently occurs in three different cell types (Fig. 1), our results seem inconsistent with what are mentioned in a recent study (35), which showed that UV radiation reduces AMPK activity (Thr 172 phosphorylation) via down-regulation of LKB1 expression. This may be due to the difference of tested time points between previous work and our study. Instead of time course (15,30,60,120,180,240, and 480 min as we tested in this study) (Figs. 1, C and  D, and 2C), a previous study reported the data collected at only 4-h time points (35). We show here that AMPK phosphorylation begins at 0.5-1 h and lasts for at least 4 -6 h after 25 mJ/cm 2 of UV radiation (Fig. 2C). Also we are unable to see LKB1 expression change even at 4 -6 h after UV radiation (Fig. 2C). Instead, we see clearly LKB1 phosphorylation after UV radiation at earlier time points in HaCaT cells (Fig. 2, C-E), and this LKB1 phosphorylation lasts for about 4 h (Fig.  2, C and D). Thus, LKB1 phosphorylation might serve as a upstream activator of AMPK phosphorylation induced by UV. We propose that this LKB1/AMPK phosphorylation is mediated by UV-induced ROS production, because ROS is known as a strong AMPK activator as shown in a number of studies (36 -38), and more importantly, in our studies, antioxidants such as NAC and pyrrolidine dithiocarbamate inhibit UV-induced AMPK phosphorylation (Fig. 2H). Another explanation is that the dose of UV used in our study is slightly different from a previous study (35) where as high as 37.5 mJ/cm 2 of UV was used, which might cause some nonspecific effects. Interestingly, in our study, at 4 h, UV-induced AMPK phosphorylation is most pronounced at 25 mJ/cm 2 (Fig. 2D). However, even at a very high dose of UV (35 and 45 mJ/cm 2 ), we can still see enhanced AMPK/LKB1 phosphorylation (rather than decrease) (Fig. 2D).
LKB1 phosphorylates and activates the catalytic ␣-subunit of AMPK at its T-loop residue Thr 172 in a cell-free system (7,39). AMPK is minimally activated in cells that lack or have decreased LKB1 expression and LKB1 forms a heterotrimeric complex, with regulatory proteins termed STRAD and MO25, which are required for its activation and cytosolic targeting (40). Thus, LKB1 plays a crucial role in activating AMPK. In this study, we found that UV (Fig. 2, C and F) and H 2 O 2 (Fig. 2G) induce LKB1 phosphorylation. However, because of limitation of experimental materials, we can only presume that LKB1 is upstream for AMPK activation upon UV and ROS stimulation. Further experiments using LKB1 deficiency cells are necessary to confirm our hypothesis. Furthermore, the question whether LKB1 activation plays other important roles in UV-induced skin cell damage remains to be further addressed.
p38 is involved in inflammation, cell growth and differentiation, cell cycle, and cell death (41,42). Recent studies have indicated that AMPK serves as an upstream signal for p38 activa-

W I T H D R A W N W I T H D R A W N
tion by increasing recruitment of p38 MAPK to TAB1. And yet, other studies have yielded different conclusions (27,28,43). Our studies using AMPK inhibitor and AMPK␣ siRNA knockdown have demonstrated that inhibition of AMPK impairs p38 activation in response to UV or H 2 O 2 (Fig. 3, A-D and G). We conclude that AMPK may serve as a positive regulator for p38 activation upon UV radiation, which might contribute to the pro-apoptotic effect of AMPK on UV-treated cells (Fig. 7G).
AMPK activation causes a G 1 /S phase cell cycle arrest in different cell lines, and this is associated with accumulation of tumor suppressor p53 and cyclin-dependent kinase inhibitors p21 and p27, which act downstream of p53 (17, 44 -46). This process involves phosphorylation of p53 at Ser 15 (17, 44 -46), although it has not been shown that there is a direct phosphorylation. In this study, as demonstrated in Fig. 4, AMPK activation serves as a positive regulator for UV-induced p53 phosphorylation (Ser 15 ). However, as seen in Fig. 4, C-E, UV-and H 2 O 2 -induced p53 phosphorylation is reduced (but not largely impaired nor totally blocked) as a result of AMPK inhibition, which suggest that UV-induced p53 phosphorylation is not solely dependent on AMPK activation, and AMPK may be one of many molecules that phosphorylate p53 upon UV radiation. We propose that this positive regulatory effect of AMPK on p53 phosphorylation might contribute to its pro-apoptotic effect (Fig. 7H).
AMPK activation may also inhibit cell growth and survival because of its general effects on biosynthesis, including its ability to inhibit fatty acid synthesis and the mTOR pathway (8,47,48). AKT/PKB phosphorylates TSC2 at sites that are believed to inhibit its Rheb-GAP activity, whereas AMPK is thought to have the opposite effect. Our previous studies have indicated that UV radiation induces mTOR activation in skin keratinocytes in an AKT-dependent manner. 4 By using AMPK site-specific mutation plasmids (8,47,48), our previous work also indicated that AMPK activates TSC2 by phosphorylation at Ser 1345 in response to UV radiation, which ultimately inhibits mTOR activation. 4 This site-specific phosphorylation and activation of TSC2 explains the inhibitory role of AMPK in UV-induced mTOR activation. 4 In this study, using AMPK inhibitor and specific siRNA knockdown as well as AMPK activator AICARand TSC2-deficient cells, we found that AMPK serves as a negative feedback signal against UV-induced mTOR activation (Fig. 6), which may also be involved in the pro-apoptotic effects of AMPK (Fig. 7A).
AMPK stimulates catabolic pathways that generate ATP, including the fatty acid oxidation pathway. Most fatty acids are oxidized in mitochondria, and their entry into the organelle is blocked by malonyl-CoA, which inhibits carnitine palmitoyltransferase-1, an enzyme required for the transport of fatty acids across the inner mitochondrial membrane. Malonyl-CoA is produced by one of the target enzymes for AMPK, acetyl-CoA carboxylase, which exists as two isoforms, ACC1 (␣) and ACC2 (␤) (5). Both are phosphorylated and inactivated by AMPK. In this study, we also discovered that UV and H 2 O 2 induce ACC phosphorylation and inactivation in an AMPK-dependent manner (Fig. 5), which might be responsible for reduced fat accumulation or adipogenesis (49). Furthermore, we also found that PFK2, the key protein of glycolysis is also phosphorylated upon UV radiation in an AMPK-dependent manner (Fig. 5). Based on these observations, together with the critical role of AMPK in energy balancing and metabolism, we propose that the metabolic system might be impaired after UV radiation in human skin cells, which might be another critical mechanism responsible for skin cell damage. The fat accumulation and energy storage in human skin cells might also be impaired (as the possible consequence of AMPK-dependent ACC phosphorylation in keratinocytes). Thus, this study provides novel insights into understanding the cellular mechanisms involved in UV-induced skin aging or skin cancer.