E2F1 suppresses skin carcinogenesis via the ARF-p53 pathway.

The E2F1 transcription factor, which is deregulated in most human cancers by mutations in the p16-cyclin D-Rb pathway, has both oncogenic and tumor-suppressive properties. This is dramatically illustrated by the phenotype of an E2F1 transgenic mouse model that spontaneously develops tumors in the skin and other epithelial tissues but is resistant to papilloma formation when subjected to a two-stage carcinogenesis protocol. Here, this E2F1 transgenic model was used to further explore the tumor-suppressive property of E2F1. Transgenic expression of E2F1 was found to inhibit ras-driven skin carcinogenesis at the promotion stage independent of the type of promoting agent used. E2F1 transgenic epidermis displayed increased expression of p19(ARF), p53, and p21(Cip1). Inactivation of either p53 or Arf in E2F1 transgenic mice restored sensitivity to two-stage skin carcinogenesis. While Arf inactivation impaired tumor suppression and p21 induction by E2F1, it did not reduce the level of apoptosis observed in E2F1 transgenic mice. Based on these findings, we propose that E2F1 suppresses ras-driven skin carcinogenesis through a nonapoptotic mechanism involving ARF and p53.

Inactivating E2f1 reverts apoptosis resistance and cancer sensitivity in Trp53-deficient mice.

The E2f1 transcription factor, which regulates genes required for S-phase entry, also induces apoptosis by transcriptional and post-translational mechanisms. As E2f1 is inducible by DNA damage we investigated its importance in vivo in ultraviolet (UV)-induced apoptosis, a protective mechanism that prevents the epidermis from accumulating UV-induced mutations. Contrary to expectation, E2f1-/- mice demonstrated enhanced keratinocyte apoptosis after UVB exposure, whereas apoptosis was suppressed by epidermis-specific overexpression of human E2F1. Apoptosis induced by -radiation was also repressed by E2f1. E2f1-/-;Trp53-/- double knockout mice exhibited the elevated UVB-induced apoptosis of E2f1-/- alone, rather than the profound apoptosis defect seen in Trp53-/- mice, indicating that Trp53 (p53) lies functionally upstream of E2f1. Transfecting E2F1 into E2f1-/-;Trp53-/- primary fibroblasts suppressed UVB-induced apoptosis and this suppression was relieved by Trp53. The double knockout also reverted the abnormal sex ratio and early-onset tumours of Trp53-/- mice. These results imply that E2f1 functions as a suppressor of an apoptosis pathway that is initiated by DNA photoproducts and perhaps genetic abnormalities; p53 relieves this suppression.

Ultraviolet-B irradiation alters the cell cycle machinery in murine epidermis in vivo.

Ultraviolet radiation of mouse skin leads to epidermal hyperplasia, inflammation, and subsequent tumor development. In this study we determined to what extent the cell cycle machinery is altered during epidermal proliferation after ultraviolet B radiation. A minimal erythema dose, 90 mJ per cm2, increased the protein expression of the G1 phase cyclins, cyclin D1 and E, by 12 h. The majority of epidermal cells entered S phase between 18 and 24 h as determined by 5'-bromo-2'-deoxyuridine incorporation, proliferating cell nuclear antigen, and cyclin A immunohistochemistry. An increase in cyclin-dependent kinase 2 (cdk-2) protein expression occurred after 12 h, but no changes in cdk-4 or cdk-6 protein levels were observed. The increase in cyclin D1, E, and A protein expression was associated with an increase in cyclin D1-cdk-4, cyclin E-cdk-2, and cyclin A-cdk-2 complex formation. p53 protein expression was elevated through 48 h, and the cdk inhibitor protein p21(Cip1/WAF1) was elevated 6-fold to 7.5-fold between 12 and 24 h. The elevated p21(Cip1/WAF1) protein contributed to an enhanced association with cdk-2 and cdk-4 at 3-24 h and 6-24 h post-ultraviolet B irradiation, respectively. These data indicate that 90 mJ per cm2 of ultraviolet B irradiation induces a DNA damage response, by increasing p53 and p21(Cip1/WAF1) protein expression, but also induces a rapid and sustained increase in S phase by 18 h.

The effect of vitamin E acetate on ultraviolet-induced mouse skin carcinogenesis.

Despite the benefits of sunscreens, ultraviolet (UV) exposure can still lead to skin cancer. In this study we investigated the effect of topical application of the antioxidant vitamin E acetate (VEA) on the inhibition of UV-induced carcinogenesis. Hairless SKH-1 mice received 5.2 mg of VEA 30 min before (VEA/UV) or after (UV/ VEA) a single minimal erythemic dose of UV light. Vehicle-control animals received acetone 30 min before UV exposure (Ace/UV). After 24 h, cyclobutane dimer repair was twofold and 1.5-fold greater in the UVNEA and VEA/UV groups, respectively. Expression of p53 protein in the UV/VEA group was maximum at 12 h after UV exposure, whereas in the Ace/UV- and VEA/UV-treated mice, maximum p53 immunostaining was statistically higher at 15 h (P = 0.03). DNA synthesis as determined by 5-bromo-2'-deoxyuridine incorporation was twofold higher after 15 h in all groups but was not statistically different among treatment groups. Protein levels of cyclin D1 and p21 were increased in both VEA groups by 6 h. In addition, VEA treatments delayed tumor formation and yield for the first 20 wk, although this difference was lost by 30 wk. The telomerase activity of carcinomas from the UV/VEA-treated mice was statistically lower than that of the Ace/UV-treated mice (P = 0.05). This study showed that although VEA may mitigate some of the initial events associated with UV irradiation such as DNA damage and p53 expression, it has limited potential in preventing UV-induced proliferation and tumor formation.

Epidermal proliferation but not quantity of DNA photodamage is correlated with UV-induced mouse skin carcinogenesis.

The hairless SKH-1 mouse strain has a higher skin tumor incidence, shorter tumor latency, and higher tumor yield in response to ultraviolet (UV) irradiation than the SENCAR strain. In this study we assessed the differences in UV susceptibility of both strains by measuring DNA photodamage and epidermal proliferation after one UV treatment and after 1, 3, 6, and 9 wk of chronic UV irradiation. Induction rates for cyclobutane pyrimidine dimers (CPDs) and pyrimidine (6-4) pyrimidone photoproducts [(6-4) PDs] were significantly greater in the SKH-1 strain than the SENCAR strain, but no strain differences in repair kinetics were detected for CPDs or (6-4) PDs. With chronic UV exposure we observed the following: (i) there was an equal amount of DNA photodamage in both strains; (ii) the number of (6-4) PDs was significantly greater than the CPDs after 6 wk; (iii) there were a significantly greater number of epidermal cells (1.5-fold increase) in the SKH-1 strain; (iv) the number of cycling cells, as measured by 5-bromo-2'-deoxyuridine (BrdU), were located both basally and suprabasally and were significantly greater in the SKH-1 strain; and (v) the number of cells immunoreactive to p53 was equivalent in both strains, but immunoreactive cells were located suprabasally in the SKH-1 strain after 9 wk of UV. These results show that the etiologic role of UV in tumorigenesis is dependent on events other than the amount of DNA photodamage in mouse epidermis.

Comparison of ultraviolet light-induced skin carcinogenesis and ornithine decarboxylase activity in sencar and hairless SKH-1 mice fed a constant level of dietary lipid varying in corn and coconut oil.

To investigate the effect of various levels of corn oil and coconut oil on ultraviolet (UV) light-induced skin tumorigenesis and ornithine decarboxylase (ODC) activity, Sencar and SKH-1 mice were fed one of three 15% (weight) fat semipurified diets containing three ratios of corn oil to coconut oil: 1.0%:14.0%, 7.9%:7.1%, and 15.0%:0.0% in Diets A, B, and C, respectively. Groups of 30 Sencar and SKH-1 mice were fed one of the diets for three weeks before UV irradiation; then both strains were UV irradiated with an initial dose of 90 mJ/cm2. The dose was given three times a week and increased 25% each week. For Sencar mice (irradiated 33 wks for a total dose of 48 J/cm2), tumor incidence reached a maximum of 60%, 60%, and 53% for Diets A, B, and C, respectively, with an overall average of one to two tumors per tumor-bearing animal. For the SKH-1 mice (irradiated 29 wks for a total dose of 18 J/cm2), all diet groups reached 100% incidence by 29 weeks, with approximately 12 tumors per tumor-bearing mouse. No significant effect of dietary corn oil/coconut oil was found for tumor latency, incidence, or yield in either strain. The effect of increasing corn oil on epidermal ODC activity in chronically UV-irradiated Sencar and SKH-1 mice was assessed. Three groups of mice from each strain were fed one of the experimental diets and UV irradiated for six weeks. Sencar mice showed no increase in ODC activity until six weeks of treatment, when the levels of ODC activity in the UV-irradiated mice fed Diet A were significantly higher than those in mice fed Diet B or Diet C: 1.27, 0.55, and 0.52 nmol/mg protein/hr, respectively. In the SKH-1 mice, ODC activity was increased by the first week of UV treatment, and by three weeks of treatment a dietary effect was observed; ODC activity was significantly higher in mice fed Diet C (0.70 nmol/mg protein/hr) than in mice fed Diet A (0.18 nmol/mg protein/hr). Although there was no significant effect of dietary corn oil/coconut oil on UV-induced tumor incidence, the data indicate that chronically UV-irradiated hairless SKH-1 mice are more susceptible to UV-induced skin carcinogenesis than Sencar mice and that this susceptibility is correlated with increased in ODC activity, a parameter of cell proliferation.