Inhibition of mTOR Suppresses UVB-Induced Keratinocyte Proliferation and Survival

Nonmelanoma skin cancer (NMSC) is the most frequent malignancy worldwide, with more than 1 million cases diagnosed each year in the United States alone (1). Squamous cell carcinoma (SCC) and basal cell carcinoma are the most common forms of NMSC and account for more than 40% of newly diagnosed cancers (1). The most important risk factor for NMSC is UV radiation exposure, which is the leading cause of skin cell damage, photoaging, and malignant transformation (2). UV radiation from sunlight is composed of UVB (280–315 nm) and UVA (315–400 nm). UVB wavelengths are the most energetic and account for the majority of the biologically damaging effects from sun exposure. UVB is a complete carcinogen and has been shown to act as a tumor initiator by inducing mutations in critical target genes and a tumor promoter by activating signal transduction pathways mediated by kinases and transcription factors that induce cell-cycle progression and proliferation (3).

mTOR, an essential serine/theronine kinase conserved in all eukaryotes, is at the center of an increasingly intricate signaling network that responds to nutrients, growth factors, and cellular stress (4). Hyperactivation of mTOR signaling through mutational activation of upstream signals, such as Ras or phosphoinositide 3-kinase (PI3K), occurs in many cancers and contributes to increased proliferation and reduced sensitivity to apoptotic stimuli (4, 5). mTOR exists in at least two functionally and compositionally distinct protein signaling complexes: the rapamycin-sensitive mTOR complex 1 (mTORC1) and the rapamycin-resistant mTOR complex 2 (mTORC2). The mTORC1 complex, consisting of mTOR, Raptor, and mLST8, is downstream of both the Raf/MEK/ERK and PI3K/Akt pathways and controls protein synthesis and ribosome biogenesis by direct phosphorylation of eukaryotic initiation factor 4E binding proteins (4EBPs) and p70 S6 kinase 1 (S6K), respectively (6). The mTORC2 complex also contains mTOR and mLST8, but Raptor is replaced by Rictor and mSin1. While the function of mTORC2 is less well understood, it has been shown to phosphorylate AKT at Ser473, thus promoting cell survival and proliferation (7). Phosphorylation of AKT by mTORC2 can also serve to activate mTORC1, emphasizing the interdependence of the two mTORCs (8).

Several studies indicate that mTOR signaling may play a critical role in NMSC development. Immunohistochemical analysis of human epidermal tumors showed that mTOR itself as well as its downstream effectors 4EBP1, S6K, and AKT^Ser473 are phosphorylated at much higher levels in SCC and precancerous actinic keratosis than normal skin (9). More recently, reverse phase protein microarray analysis of SCC, actinic keratosis, and normal skin revealed aberrantly activated mTOR pathways in the precancerous and transformed tissues (10). Data compiled from more than 30,000 kidney transplant recipients found the use of mTOR inhibitors as maintenance immunosuppressive therapy produced a remarkable reduction in NMSC incidence compared with the calcineurin inhibitor cyclosporine A (CsA; ref. 11). A dramatic decrease in the incidence of skin malignancies was also observed in transplant patients who were converted to mTOR inhibitors after 3 months of treatment with CsA (the Sirolimus Renal Conversion Trial study; ref. 12). These striking results have led to several ongoing prospective randomized trials evaluating the use of mTOR inhibitors in renal transplant patients, and could significantly impact prevention of NMSC in the general population as well.

Photocarcinogensis is characterized by the inhibition of apoptosis and the enhancement of cell proliferation. Although studies have shown that UVB induces phosphorylation of 4EBP1, S6K, and AKT (13–15), the specific roles of mTORC1- and mTORC2-dependent pathways in UVB-induced proliferation and apoptosis have not been defined. In the present study, we examined the effect of UVB radiation on mTOR signaling in cell culture and in vivo using both the specific mTORC1 inhibitor rapamycin and a genetic model of conditional mTOR deletion in the epidermis, which results in inhibition of both mTORC1 and mTORC2. Our findings indicate that UVB stimulates both mTORC1 and mTORC2 activities. While rapamycin treatment effectively blocked the hyperproliferation response that occurs with UVB exposure, keratinocytes were sensitized to UVB-induced apoptosis only when mTORC2 activity was downregulated. Results obtained using rictor-null cells confirmed that intact mTORC2 is necessary to activate prosurvival signaling. These studies thus provide new insights into the molecular mechanisms of UVB-induced damage leading to skin aging and skin cancer.

HaCaT keratinocytes (obtained from The German Cancer Research Center), were maintained in Dulbecco's modified Eagle's medium (DMEM, Invitrogen) supplemented with 10% heat-inactivated FBS (Atlanta Biologicals) and penicillin/streptomycin (100 μg/mL, Invitrogen), and cultured for less than 20 passages. Control (Rictor^Ex3cond/w) and rictor-null (Rictor^{Ex3del/Ex3del}) mouse embryo fibroblasts (MEF), abbreviated Rictor^+/+ and Rictor^−/−, were a generous gift from Dr. Mark Magnuson, Vanderbilt University (Nashville, TN; ref. 16) and maintained in DMEM supplemented with 10% FBS and penicillin/streptomycin. No further characterization of cell lines was conducted. Primary mouse keratinocytes were isolated from 1- to 3-day-old pups as described previously (17). Briefly, full thickness skin was floated overnight at 4°C in Ca⁺²-free 0.25% Trypsin without EDTA (Cellgro) to separate the epidermis from the dermis. Epidermal sheets were minced and the cell suspension was strained (100-μm nylon filter, Fisher) and plated in keratinocyte growth medium [calcium-free minimum essential medium (MEM) Eagle with Earle's balanced salt solution (BSS), Gln and nonessential amino acids (Lonza), 8% chelexed FBS (Gibco), and penicillin/streptomycin]. Cells were maintained in 7% CO₂ at 37°C, and the medium was changed every other day. Rapamycin (Developmental Therapeutics Program, National Cancer Institute, Bethesda, MD) or vehicle dimethyl sulfoxide (DMSO) was added to medium 1 hour before UVB treatment. Primary keratinocyte cultures from transgenic animals (K5-CreER^T2;mTOR^fl/fl and mTOR^fl/fl) were supplemented with 5 nmol/L of 4-hydroxytamoxifen (4OHT, Sigma) for 4 days to induce recombination.

All experiments involving mice were carried out in compliance with the Guide for the Care and Use of Laboratory Animals and protocols were approved by the Animal Care and Use Committee of the Pennsylvania State University College of Medicine (Hershey, PA).

For mouse experiments, the dorsal surface of FVB/N mice (6–8 weeks) was shaved with electrical clippers and mice were allowed to rest for 24 to 48 hours before all experiments. For rapamycin in vivo studies, mice were treated topically with 100 nmol rapamycin [in 100 μL DMSO:Acetone (1:9), D:A] or vehicle 1 hour before UVB exposure. We used an inducible Cre-LoxP mouse model to ablate mTOR in the epidermis. Floxed mTOR mice (mTOR^fl/fl) contain LoxP sites flanking exons 49 and 50 of the mTOR gene (18); recombination results in a frameshift mutation and loss of the essential kinase domain. K5-CreER^T2 mice (19) express a tamoxifen-activated Cre recombinase fused to a modified estrogen receptor in the basal layer of the epidermis under the control of the keratin 5 promoter. K5-CreER^T2 mice were bred with mTOR^fl/fl mice to ultimately generate mice hemizygous for the K5-Cre-ER^T2 transgene and homozygous for the mTOR floxed allele (K5-CreER^T2;mTOR^fl/fl). K5-CreER^T2;mTOR^fl/fl, and mTOR^fl/fl controls were treated topically with 1 mg 4OHT (in 100 μL D:A) or vehicle daily for 5 consecutive days. All animals were backcrossed for at least 9 generations onto the FVB/N background. Primer sequences used for PCR are as follows: mTOR 1 (wild-type, LoxP, ΔLoxP): GTC CAC CAA CTC GGG CCT CAT T, mTOR 2 (wild-type, LoxP): GCA TGG CGA GGA CAT GTC A, and mTOR 3 (ΔLoxP): CCA CGC ATG GCC CAC TGT CTT T.

Cells were cultured for 48 to 72 hours until 70% confluence on 60-mm plates under normal culture conditions. For cell-cycle experiments (low-dose UVB), medium was replaced with 0.1% FBS for 24 hours before UVB treatment. Cells were washed twice with PBS, then in a minimal volume of PBS exposed to UVB (FS20 UVB bulbs, National Biological) emitting UV light between 290 to 320 nm. The irradiation intensity was monitored using a UVB 500C meter (National Biological). After irradiation, PBS was removed and saved medium with drug treatments was added back. For in vivo studies, 3 to 4 mice were used for each treatment group and housed together. Mice were exposed to UVB irradiation from UVB lamps (FS20 UVB bulbs, National Biological) at a dose of 120 mJ/cm². Bulb intensity was measured at the beginning of each experiment.

For in vitro studies, cells were washed twice with cold PBS and harvested by scraping into the radioimmunoprecipitation assay (RIPA) buffer (Santa Cruz Biotechnology), centrifuged, and supernatants collected. In selected experiments examining apoptosis, detached cells were collected via centrifugation of medium and combined with the attached cells in RIPA buffer. For in vivo studies, dorsal skin was treated with a depilatory agent for 3 minutes, washed, excised, and underlying fat and connective tissue were removed. Whole skin samples were flash-frozen and later processed in RIPA buffer by homogenizing for 30 seconds on ice using a Polytron homogenizer, centrifuged at 30,000 × g for 30 minutes at 4°C, and supernatants collected. Epidermal samples were collected by scraping the surface of excised skin with a razor blade and placed into ice-cold RIPA buffer. Protein concentrations were determined using Bio-Rad assay. Equal amounts of protein were subjected to electrophoresis. Western blotting was conducted as described previously (20). Antibodies used include AKT, p-AKT^Ser473, S6K, p-S6K^Thr389, caspase-3, cleaved caspase-3, mTOR, actin (all from Cell Signaling, 1:1,000) and glyceraldehyde-3-phosphate dehydrogenase (GAPDH; Proteintech, 1:2,000).

For cell-cycle analysis, cells were harvested 18 hours after UVB and fixed with ice-cold ethanol overnight. Cells were washed twice with PBS and then incubated with propidium iodide (20 μg/mL) containing RNase A (Sigma). The DNA content was determined using a FACSCalibur cytometer (Beckman Coulter) and analyzed with Modfit LT software (Verify Software). Apoptosis was assessed with flow cytometry using the PE Annexin V Apoptosis Detection Kit I (BD Biosciences) according to the manufacturer's instructions.

The cell viability at 24 hours after UVB exposure was determined colorimetrically by MTS assay (CellTiter 96 Aqueous Proliferation Assay, Promega) according to the manufacturer's instructions and monitored at 490 nm using a model 3550-UV plate reader (Bio-Rad). Each experiment was carried out in triplicate and the experiments were repeated at least 3 times.

Tissue sections were collected from 3 to 4 mice for each condition at each timepoint (mock animals were not irradiated and collected at 24 hours). Dorsal skin was treated with a depilatory agent for 3 minutes, washed, excised, and fixed overnight in 10% neutral buffered formalin. Skin was embedded in paraffin and 5 μm sections were cut for immunohistochemistry. The effect of rapamycin and mTOR deletion on UVB-mediated epidermal hyperplasia was studied by histopathologic examination of hematoxylin and eosin (H&E)-stained tissue sections. Epidermal thickness was measured at 5 locations in 4 different sections for each mouse. Epidermal proliferation was assessed in vivo using 5-bromo-2-deoxyuridine (BrdUrd) incorporation. Mice received intraperitoneal injections of BrdUrd (Sigma) at 100 μg/g body weight in 0.9% NaCl 1 hour before killing. Sections were deparaffinized, rehydrated, and stained with an anti-BrdUrd antibody as described previously (21). Epidermal proliferation index (PI) was determined by calculating the percentage of basal cells positive for BrdUrd. A minimum of 1,000 basal cells were counted. Apoptosis was assessed using immunohistochemical staining with cleaved caspase-3 antibody. Interfollicular apoptotic keratinocytes were counted microscopically in at least 5 nonoverlapping low-power fields.

Data are expressed as the mean of at least 3 independent experiments analyzed by 2-sided Student t test. A P value of < 0.05 was considered significant.

In photocarcinogenesis, UVB-stimulated activation of serine/theronine and tyrosine kinase signaling pathways and transcription factors mediate a number of pathologic changes in the skin. The resulting keratinocyte proliferation is responsible for epidermal hyperplasia and tumor promotion (3). We first sought to investigate the role of mTORC1 in UVB-induced cell-cycle progression and proliferation using rapamycin. It was previously shown that subapoptotic doses of UVB (2.5–10 mJ/cm²) activate EGF receptor and AKT and induce G₁–S cell-cycle progression in serum-deprived HaCaT cells (22). To determine whether mTORC1 contributes to UVB-induced cell-cycle progression, HaCaT cells were serum-starved to synchronize them in G₀ and then subjected to low-dose UVB. As shown in Fig. 1A, low-dose UVB activates both mTORC1 and mTORC2 signaling pathways as measured by phosphorylation of S6K and AKT^Ser473, respectively. To determine whether mTORC1 activation is involved in UVB-mediated cell-cycle progression, cells were incubated with 50 nmol/L rapamycin to block mTORC1 activation. Rapamycin suppressed UVB-induced S6K phosphorylation, but did not alter phosphorylation of AKT^Ser473 (Fig. 1B). Low-dose UVB stimulated cell-cycle progression in vehicle-treated cells, as measured by the percentage of cells in S-phase at 18 hours after irradiation (Fig. 1C). The proportion of cells in the S-phase decreased significantly in cells treated with rapamycin (Fig. 1C).

Because cellular proliferation in response to UVB is a key feature of photocarcinogenesis, we next sought to explore the role of mTORC1 in UVB-mediated responses in vivo. Wild-type FVB/N adult mice were treated topically with rapamycin (100 nmol) 1 hour before irradiation with a single doses of UVB (120 mJ/cm²), and whole-skin samples were subjected to immunoblot analysis at 6 hours after UVB exposure to assess mTOR activation. Concentrations of p-S6K and p-AKT^S473 were too low in mock-irradiated samples to discern any noticeable inhibitory effect of rapamycin (Fig. 2A). However, there was a dramatic increase in p-S6K and p-AKT^Ser473 levels in vehicle-treated skin after UVB irradiation. Rapamycin pretreatment prevented UVB-stimulated phosphorylation of S6K, but had no inhibitory effect on p-AKT^Ser473. Epidermal hyperplasia (as measured by epidermal thickness) and proliferation (as measured by BrdUrd incorporation) were examined in skin sections at 24 and 48 hours after UVB exposure. Rapamycin had no effect on epidermal thickness or proliferation in mock-irradiated animals. There was noticeable epidermal thickening in vehicle-treated animals at 24 and 48 hours following UVB exposure, but this effect was significantly reduced in animals treated with rapamycin (Fig. 2B and C). Similar blocking effects were seen with rapamycin when epidermal thickening was stimulated by treatment with the chemical tumor promoter 12-O-tetradecanoylphorbol-13-acetate (TPA, see Supplementary Fig. S1). BrdUrd labeling of actively proliferating cells at 48 hours showed a similar pattern. Vehicle-treated mice showed a significant increase in BrdUrd-positive cells at 48 hours after UVB exposure, and this effect was significantly reduced by rapamycin treatment (Fig. 2D, E). Collectively, these results and those in Fig. 1 suggest that mTORC1 plays an important role in keratinocyte cell-cycle progression and epidermal hyperproliferation following UVB irradiation.

We next explored the consequences of deletion of mTOR in the basal layer of the skin, targeting both mTORC1 and mTORC2 signaling, to analyze the contribution of both mTOR-signaling complexes in UVB-mediated responses. Because homozygous deletion of mTOR is lethal in utero (23), we used a transgenic system with 4OHT-inducible deletion of mTOR in the epidermis (K5-CreER^T2;mTOR^fl/fl). We verified that the topical treatment with 4OHT, but not vehicle, leads to CreER^T2-dependent recombination of the mTOR allele (ΔLoxP) by PCR analysis using DNA from the dorsal epidermis harvested from mice 2 weeks after the final 4OHT treatment (Fig. 3A). No recombination was observed in the absence of either CreER^T2 expression or 4OHT induction (data not shown). Western blot analysis confirmed reduction of mTOR protein levels in 4OHT-treated K5-CreER^T2;mTOR^fl/fl animals (Fig. 3B). Downregulation of both mTORC1 and mTORC2 signaling pathways was verified by examining phosphorylation of S6K and AKT^Ser473 in whole skin protein extracts. In the absence of UVB stimulation of mTOR pathways, there was no apparent difference between the vehicle-treated and 4OHT-treated animals (Fig. 3C). However, p-S6K and p-AKT^ser473 levels were dramatically increased at 6 hours after UVB (120 mJ/cm²) in vehicle-treated mice. This effect was significantly attenuated upon treatment of K5-CreER^T2;mTOR^fl/fl animals with 4OHT (Fig. 3C), confirming that 4OHT treatment in K5-CreER^T2;mTOR^fl/fl mice was sufficient to block UVB-induced activation of both mTORC1 and mTORC2. Histologic evaluation of epidermal thickness and proliferation index in mock-irradiated animals revealed no differences between vehicle-treated and 4OHT-treated animals (Fig. 3D–G). UVB irradiation caused a significant increase in epidermal thickness in vehicle-treated K5-CreER^T2;mTOR^fl/fl mice at 24 and 48 hours, whereas mice treated with 4OHT (Fig. 3D and E) showed much less epidermal thickening in response to UVB. K5-CreER^T2;mTOR^fl/fl mice treated with 4OHT also showed a significant reduction in proliferation index at 48 hours after UVB treatment compared with vehicle-treated controls (Fig. 3F and G). All mice used as controls in these experiments responded similarly to each other and to wild-type mice at all timepoints, including mTOR^fl/fl mice treated with 4OHT, mTOR^fl/fl mice treated with vehicle, and K5-CreER^T2;mTOR^fl/fl mice treated with vehicle (see Supplementary Fig. S2).

Because one of the major activities of AKT is to promote cell survival (24), we sought to investigate whether UVB-induced mTOR activation of AKT-dependent pathways plays a role in preventing cell death following high-dose UVB. We investigated mTOR signaling pathways after 50 mJ/cm² UVB exposure by examining the expression of total and phosphorylated S6K and AKT in normal mouse primary keratinocytes by Western blot analysis. Enhanced signaling through both mTORC1 and mTORC2 was indicated by increased levels of p-S6K and p-AKT^Ser473 at 30 minutes after 50 mJ/cm² UVB and remained upregulated at 2 hours (Fig. 4A). Pretreatment of cells with rapamycin for 1 hour completely blocked UVB-induced activation of mTORC1 signaling (p-S6K), but had little, if any, effect on mTORC2 activity (p-AKT^Ser473; Fig. 4B). To determine whether inhibition of mTORC1 could sensitize keratinocytes to apoptosis, cells exposed to various rapamycin concentrations were harvested 24 hours after UVB exposure and assayed for viability. Treatment with a wide range of rapamycin concentrations did not alter cell viability in mock-irradiated cells (Fig. 4C). Exposure to 50 mJ/cm² UVB resulted in an obvious decrease in cell viability (54% ± 3.6%). However, rapamycin did not enhance UVB-mediated cell death in wild-type primary keratinocytes (Fig. 4C), and this same result was seen in HaCaT cells (Supplementary Fig. S3A). In addition, increasing the pretreatment time with rapamycin to 24 hours did not enhance UVB-mediated cell death (Supplementary Fig. S3B and S3C). Western blot analysis of cleaved caspase-3 verified that there was no increase in UVB-mediated apoptosis with rapamycin treatment (Fig. 4D).

To investigate the possible role of mTORC2 in suppressing apoptosis after UVB exposure, we used primary keratinocytes isolated from K5-CreER^T2;mTOR^fl/fl mice. PCR analysis verified recombination of the mTOR allele (ΔLoxP) in cells cultured with 4OHT, but not vehicle (Fig. 5A). There was no obvious difference in mTORC1 and mTORC2 activities in 4OHT-treated keratinocytes not exposed to UVB (0 minutes), as measured by p-S6K and p-AKT^Ser473 (Fig. 5B). However, when keratinocytes were exposed to UVB (50 mJ/cm²) to activate both mTOR complexes, a dramatic reduction in phosphorylation of both S6K and AKT^ser473 was observed in 4OHT-treated cells compared with vehicle, confirming downregulation of both mTORC1 and mTORC2 signaling (Fig. 5B). Unlike inhibition of mTORC1 with rapamycin (Fig. 4C), deletion of mTOR enhanced UVB-induced cell death (Fig. 5C). Though there was no difference in cell viability in mock-irradiated cells, K5-CreER^T2;mTOR^fl/fl keratinocyte cultures treated with 4OHT contained significantly fewer viable cells 24 hours after UVB exposure compared with both vehicle-treated K5-CreER^T2;mTOR^fl/fl and 4OHT-treated mTOR^fl/fl cells. To determine whether the significant decrease in cell viability was due to apoptosis, immunoblot analysis of cleaved caspase-3 was examined (Fig. 5D). Cell extracts harvested at 9 hours after UVB showed significantly higher levels of cleaved caspase in mTOR-ablated keratinocytes than in vehicle-treated K5-CreER^T2;mTOR^fl/fl cells.

We further investigated the effects of mTOR deficiency on UVB-meditated apoptosis in vivo using our K5-CreER^T2;mTOR^fl/fl mice. UVB irradiation increased the number of epidermal cleaved caspase-3–positive cells 24 and 48 hours after irradiation in K5-CreER^T2;mTOR^fl/fl mice treated with vehicle. The number of apoptotic cells was significantly increased in the epidermis of mTOR-deficient mice (4OHT treated) compared with vehicle controls at 24 hours (Fig. 5E and F).

Our results show that K5-CreER^T2;mTOR^fl/fl primary keratinocytes treated with 4OHT to induce mTOR deletion have enhanced sensitivity to UVB-induced apoptosis (Fig. 5), but wild-type primary keratinocytes treated with the mTORC1 inhibitor do not (Fig. 4). These data suggest that mTORC2, but not mTORC1, influences cell survival signaling following UVB exposure. To further elucidate the specific role of mTORC2 in UVB-induced apoptosis, we used rictor-null MEFs (Rictor^−/−). UVB exposure (50 mJ/cm²) increased phosphorylation of S6K and AKT^Ser473 in wild-type MEFs (Rictor^+/+; Fig. 6A). UVB induced p-S6K in a similar manner in Rictor^−/− cells, but p-AKT^Ser473 was completely absent, illustrating the loss of mTORC2 signaling (Fig. 6A). Significantly, fewer rictor-null cells were viable compared with wild-type cells at 24 hours after UVB exposure (Fig. 6B). To determine whether this represents an increased sensitivity to apoptosis in the absence of mTORC2 signaling, cells were analyzed using Annexin V by flow cytometry. There was a significant increase in the percentage of apoptotic Rictor^−/− cells following UVB compared with wild-type cells (Fig. 6C). The marked increase in apoptosis was verified by Western blot analysis of caspase-3. The results show that cleaved caspase-3 begins to accumulate in Rictor^−/− cells by 6 hours and continues to increase up to 12 hours after UVB exposure (Fig. 6D). In contrast, wild-type MEFs show considerably less caspase-3 cleavage over the same time course. Taken together with the results presented in Figs. 4 and 5, these data are consistent with the idea that downregulation of mTORC2 signaling sensitizes cells to UVB-induced apoptosis.

Better understanding of the signal transduction pathways activated by UVB in keratinocytes is essential for effective prevention of skin cancer. Using rapamycin to inhibit mTORC1, and a Cre/LoxP approach to block both mTORC1 and mTORC2 signaling, the present study shows that the two mTOR complexes play distinct roles in mediating UVB-induced proliferation and prosurvival signaling in the epidermis. Our results fit a model in which mTORC2-dependent pathways maintain the survival of DNA-damaged keratinocytes after exposure to carcinogens, whereas mTORC1-dependent pathways control their proliferation. These studies validate the potential of both mTOR complexes as prospective chemoprevention targets in nonmelanoma skin cancer.

We observed that mTORC1 and mTORC2 downstream pathways were upregulated in response UVB both in vitro and in vivo. Rapamycin treatment or mTOR ablation inhibited UVB activation of S6K, attenuated UVB-induced cell-cycle progression, and blocked the epidermal hyperproliferative response to UVB. Cell proliferation as a result of cell-cycle progression is the key process that leads to clonal expansion of initiated cells in tumor promotion, and there is emerging evidence implicating mTORC1 signaling in epithelial tumor promotion. Rapamycin treatment was found previously to inhibit G₁–S transition by reducing cyclin D1 mRNA and protein stability (25), whereas mice with hyperactivation of mTORC1 in the epidermis display epidermal hyperplasia and spontaneous papilloma formation, which is attenuated by treatment with oral RAD001, a rapamycin derivative (26). Other studies have described that mTORC1 inhibition with rapamycin attenuates tumorigenesis in skin, head and neck, and oral squamous carcinoma models (27–31). More recently, Checkley and colleagues have shown that topical rapamycin is an effective chemoprevention strategy for tumors caused by the classical 2-step skin chemical carcinogenesis model by blocking TPA-induced proliferation (32). Our findings clearly indicate that mTORC1 activation following UVB positively regulates cell-cycle progression and proliferation, confirming and extending previous reports that mTORC1 inhibition blocks tumor promotion.

There was no significant difference in attenuation of UVB-induced hyperproliferation in mice with epidermal mTOR deletion compared with wild-type FVB/N mice receiving topical rapamycin treatment (P = 0.11), suggesting UVB-induced cell-cycle progression and proliferation are not mediated by mTORC2 signaling. This does not, however, preclude a role for mTORC2 in this process. Although we see no evidence of mTORC2 inhibition in the presence of rapamycin in this model, it is difficult to compare pharmacologic inhibition to genetic ablation. The effects of mTOR deletion may be dampened because it is unlikely that recombination and deletion occurred in all keratinocytes within the epidermis. In addition, topical rapamycin has been shown to decrease infiltration of dermal inflammatory cells (32), which may contribute to inhibitory effects on keratinocyte proliferation and confound comparisons to our genetic model of mTOR deletion in the skin.

NMSC pathogenesis is characterized by both enhancement of cell proliferation and inhibition of apoptosis. Induction of apoptosis following DNA damage is an essential protective mechanism, ensuring the removal of damaged cells that may harbor oncogenic mutations. However, UVB also activates signaling cascades that promote the survival of these potentially cancerous cells. Previous work in other mouse models has shown that targeting prosurvival pathways induced by UVB increases the sensitivity of DNA-damaged keratinocytes to apoptotic signaling, and is sufficient to inhibit skin carcinogenesis (33). The studies reported here investigate whether activation of mTORC1 and mTORC2 downstream effectors by UVB lead to changes in keratinocyte prosurvival signaling. The effects of mTORC1 inhibition on apoptosis vary greatly depending on the system used. Enhancement of AKT signaling can occur in the presence of rapamycin due to relief of an S6K-dependent negative feedback loop targeting PI3K (34, 35). Loss of this feedback inhibition is thought to be responsible for increased mTORC2/AKT activation and decreased sensitivity to apoptotic stimuli in certain malignancies treated with rapamycin (35). On the other hand, there are a number of studies that report enhancement of apoptosis by rapamycin (36–41). It has been shown that prolonged rapamycin treatment reduces mTORC2 complex assembly and AKT activation in approximately 20% of cancer cell lines (42), which could have a direct affect on apoptosis pathways. It is thus possible that the proapoptotic effects of rapamycin seen in some previous studies are the result of mTORC2 inhibition rather than a direct affect on mTORC1. This rationale is supported by our results. We see no affect of rapamycin on mTORC2-dependent pathways in our system, and rapamycin treatment does not result in enhanced activation of apoptosis in UVB-treated cells. In contrast, 4OHT-induced mTOR deletion resulted in a significant increase in apoptosis following UVB exposure in both keratinocyte culture and mouse epidermis. Furthermore, rictor-null cells were more sensitive to UVB-induced apoptosis than their wild-type counterparts. These results indicate that mTORC2 activation by UVB plays a critical role in mediating pathways that control keratinocyte survival, and reinforce previous observations of aberrant AKT activation in mouse models of NMSC (43, 44). mTORC2 was also identified as a therapeutic target in prostate cancer induced by loss of the tumor suppressor PTEN, using mice with conditional deletion of either mTOR or rictor (45, 46). The role of mTORC2 in these tumor types may be to maintain high levels of p-AKT^Ser473, which results in decreased transcription of a number of FOXO1/3-dependent cell-cycle arrest and apoptotic genes (47, 48).

In summary, this is, to our knowledge, the first study to report that mTORC1- and mTORC2-dependent pathways are both activated by UVB, and play unique roles in controlling proliferation and apoptosis in the skin. These results emphasize the need to further elucidate the roles of mTORC1 and mTORC2 in photocarcinogensis and their links to cell proliferation, apoptosis, and tumor development. Our data provide compelling evidence to support the novel hypothesis that both mTORC1 and mTORC2 act as critical mediators of UVB-activated signal transduction in keratinocytes and suggest that the combined targeting of both mTOR complexes, or alternatively mTORC1 and AKT, may be an effective chemoprevention strategy against photocarcinogensis.

No potential conflicts of interest were disclosed.

Conception and design: T.D. Carr, L.M. Shantz

Development of methodology: T.D. Carr, L.M. Shantz

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): T.D. Carr, J. DiGiovanni, C.J. Lynch, L.M. Shantz

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): T.D. Carr, C.J. Lynch, L.M. Shantz

Writing, review, and/or revision of the manuscript: T.D. Carr, J. DiGiovanni, C.J. Lynch, L.M. Shantz

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): T.D. Carr

Study supervision: L.M. Shantz

The authors thank Patricia Welsh for excellent technical support and Dr. David J. Feith for his helpful suggestions and critical reading of the manuscript.

This work was supported by NIH grants CA133945 (to L.M. Shantz), ES19242 (to L.M. Shantz), DK62880 (to C.J. Lynch), CA37111 (to J. DiGiovanni), and the Pennsylvania Department of Health Tobacco CURE funds (to L.M. Shantz). T.D. Carr is the recipient of an MD/PhD predoctoral fellowship (F30 ES19809).