Effect of Caffeine on the ATR/Chk1 Pathway in the Epidermis of UVB-Irradiated Mice

In earlier studies, we showed an inhibitory effect of p.o. administration of green or black tea on the formation of UVB-induced keratoacanthomas and squamous cell carcinomas in SKH-1 hairless mice (1, 2). The regular teas were more effective than the decaffeinated teas, adding back caffeine to the decaffeinated teas restored their inhibitory activities, and administration of caffeine alone strongly inhibited tumor formation (3, 4). These results indicated that caffeine is a major component of tea responsible for its inhibitory effect on UVB-induced tumor formation in mice. Mechanistic studies indicated that p.o. administration of green tea or black tea to tumor-bearing mice enhanced apoptosis in skin tumors and inhibited their growth (5, 6). P.o. administration of green tea or caffeine for 2 weeks before a single irradiation with UVB enhanced UVB-induced increases in epidermal wild-type p53 protein and apoptosis in the epidermis of SKH-1 mice (7). In addition, topical application of caffeine immediately after a single irradiation with UVB also enhanced UVB-induced apoptosis, but this treatment had only a small stimulatory effect on UVB-induced increases in wild-type p53 in the epidermis (8). In this study, caffeine was given immediately after UVB to prevent a possible sunscreen effect. P.o. or topical administration of caffeine had a selective proapoptotic effect on UVB-treated skin and did not affect apoptosis in the epidermis of normal non–UVB-treated skin (7, 8). In additional studies, we found that topical application of caffeine to p53^−/− or bax^−/− mice enhanced UVB-induced apoptosis by a p53- and bax-independent mechanism (9).

Cell culture studies indicate that DNA damage activates a wild-type p53-dependent G₁ checkpoint and a p53-independent G₂ checkpoint. Activation of these checkpoints inhibits cell division and allows time for DNA repair before mitosis, thereby preventing the propagation of DNA-damaged cells. Cell culture studies have also shown that treatment of DNA-damaged cells with caffeine overrides the G₂ checkpoint, thereby preventing cells from arresting in G₂, and this effect of caffeine results in attempted replication of the DNA-damaged cells that results in cell death (10–13).

Cell culture studies indicate that cyclin B1 accumulates during the S and G₂ phases of the cell cycle, and these changes are associated with an increase in cyclin B1 mRNA levels (14) and a decrease in protein degradation (15–18). Treatment of the cells with a DNA-damaging agent, such as hydroxyurea or UV, decreases the level of cyclin B1 as well as mRNA for cyclin B1, and these treatments prevent entry into mitosis by both an ataxia-telangiectasia mutated (ATM)/ATM- and Rad3-related (ATR)–dependent and an ATM/ATR-independent mechanism (15).

Because most (but not all) cell culture studies suggest the importance of the ATR/Chk1/cyclin B1 signal transduction pathway for the effects of caffeine in overcoming the DNA damage–induced G₂ cell cycle arrest (see Discussion), we hypothesize that a key component of the proapoptotic effect of caffeine in the epidermis of UVB-treated mice is inhibition of the UVB→ATR→Chk1→cdc25c→ cdc2→cdc2/cyclin B1 pathway that normally results in a decrease in cdc2/cyclin B1 activity (increased phospho-cdc2/cyclin B1 and decreased dephospho-cdc2/cyclin B1) and an arrest before mitosis. In the presence of caffeine, this arrest should be abrogated (cdc2/cyclin B1 level should be increased despite UVB) and cells should proceed into mitosis prematurely leading to p53-independent chromatin condensation and cell death (13, 19), probably by mitotic catastrophe followed by apoptosis, as suggested by Brown and Attardi (20) and by others (see Fig. 1; refs. 21–23). The results described here provide support for this hypothesis by indicating that administration of caffeine stimulates UVB-induced apoptosis, inhibits UVB-induced phosphorylation of Chk1 on Ser³⁴⁵, and curtails the UVB-induced decrease in mitotic cells with cyclin B1 in the epidermis that occurs shortly after UVB irradiation.

Exposure of mice to UV and the preparation of serial sections of skin. Female SKH-1 hairless mice, 6 to 8 wk of age, were fed a Purina Laboratory Chow 5001 diet and irradiated with UV lamps (FS72T12-UVB-HO; National Biological Corp.) that emit UVB (280–320 nm; 75–80% of total energy) and UVA (320–375 nm; 20–25% of total energy) as described in our earlier studies (1). Skin samples (20 mm long; 5 mm wide), which also contained associated dermis, were taken from the middle of the back and placed in 10% phosphate-buffered formalin at 4°C for 18 to 24 h followed by washing and storing in 75% ethanol until the samples were processed for histology as described earlier (7).

All histologic and immunohistochemical determinations were performed with 400-fold magnification and scored blind by two investigators (Y-P.L. and Y-R.L.), who evaluated coded samples randomly. Good agreement was obtained between the two investigators, and the mean value obtained from the examination of multiple fields by each investigator was determined for each mouse and used in the calculation of mean ± SE for the mice in each group. Each microscope field (40-fold magnification) was approximately equivalent to a 0.5-mm length of epidermis.

Measurement of apoptotic and mitotic cells in the epidermis. Identification of apoptotic sunburn cells was based morphologically on cell shrinkage and nuclear condensation attributable to fragmentation of the cells (24, 25). Earlier studies showed that sunburn cells are indeed apoptotic cells (26). Apoptotic sunburn cells were identified in the epidermis by their intensely eosinophilic cytoplasm and small, dense nuclei, which were observed in H&E-stained histologic sections of the skin using light microscopy. The percentage of apoptotic sunburn cells in the epidermis (basal plus suprabasal layers) was calculated from the number of these cells per 100 cells counted from the entire 20-mm length of epidermis for each skin section.

Cells undergoing mitosis were determined as described earlier (6). Mitotic cells were determined by observing (a) chromosome condensation together with breakdown of the nuclear envelope, (b) alignment of the chromosomes on the spindle equator, (c) separation of sister chromatids, and (d) movement to their respective spindle poles. Two separate nonadjacent skin sections from each mouse were analyzed, and an average value for the mitotic index and percentage of apoptotic cells was calculated.

Phospho-Chk1 and cyclin B1 immunostaining in the epidermis. The antibody used for the immunohistochemical detection of phospho-Chk1 (Ser³⁴⁵) was from Santa Cruz Biotechnology (cat. no. sc-17922), and the antibody used for the immunohistochemical measurement of cyclin B1 was from Abcam (cat. no. ab72).

Skin sections were stained by the Biotin-Streptavidin Amplified System (alkaline phosphatase–conjugated streptavidin) using StrAviGen Super Sensitive Universal immunostaining kit purchased from Biogenex, with some modifications. Paraffin sections were first treated with 0.01 mol/L sodium citrate buffer (pH 6.0) in a microwave oven at high temperature for 10 min for phospho-Chk1 or cyclin B1 staining. The sections were then incubated with a protein block (normal goat serum) for 10 min at room temperature. The sections were incubated with phospho-Chk1 antibody (1:100 dilution) or cyclin B1 antibody (1:500 dilution) for 1 h at room temperature. The samples were then incubated with a biotinylated anti-rabbit secondary antibody for 5 min at 37°C followed by incubation with conjugated streptavidin solution for 5 min at 37°C. Color development was achieved by incubation with New Fuchsin Substrate Pack (containing 0.6 mg/mL levamisole solution; Biogenex) for 20 min at room temperature. The slides were then counterstained with hematoxylin and dehydrated, and coverslips were added for permanent mounting.

A positive reaction was shown as a brown precipitate in the cells. The percentage of phospho-Chk1 or cyclin B1–positive cells in the epidermis (combined basal and suprabasal layers) was calculated from the number of phospho-Chk1–stained (nuclear staining) or cyclin B1–stained (staining in nucleus and cytoplasm) positive cells per 100 cells counted from the entire 20-mm length of epidermis for each skin section.

Measurement of phospho-Chk1 (Ser³⁴⁵), phospho-Chk2 (Thr⁶⁸), cyclin B1, and phospho-cyclin B1 (Ser¹⁴⁷) in the epidermis by Western blots. Dorsal skin samples were removed and immediately placed in a buffer solution containing 75 mmol/L dibasic sodium phosphate and monobasic potassium phosphate (pH 7.7) at 52°C for 20 s. The samples were then submerged immediately in an ice bath containing the same buffer for 1 min. The epidermis was scraped from the dermis and placed in 1.2 mL of lysis buffer containing 20 mmol/L Tris-HCl, 150 mmol/L NaCl, 1 mmol/L Na₂EDTA, 1 mmol/L EGTA, 1% Triton, 2.5 mmol/L sodium pyrophosphate, 1 mmol/L β-glycerophosphate, 1 mmol/L Na₃VO₄, 1 μg/mL leupeptin, and 1 mmol/L phenylmethylsulfonyl fluoride (Cell Signaling Technology, Inc.). The epidermis was sonicated five times (5 s each time) at 4°C. Samples were centrifuged at 17,800 × g for 20 min at 4°C. Equal amounts of supernatant protein (40 μg) were separated by SDS-PAGE (4% stacking and 4–15% gradient) and electroblotted onto a polyvinylidene difluoride membrane. The blots were blocked in 5% nonfat milk in PBS-Tween 20 for 1 h and incubated with phospho-Chk1 (Ser³⁴⁵) antibody (1:250 dilution), phospho-Chk2 (Thr⁶⁸) antibody (1:150 dilution), and cyclin B1 antibody (1:150 dilution). Blots were washed in PBS-Tween 20 and then incubated with a 1:1,500 dilution of peroxidase-conjugated secondary antibody (Amersham) in PBS-Tween 20 for 1 h at room temperature. Blots were again washed in PBS-Tween 20 and then developed by enhanced chemiluminescence (Amersham).

The antibody used for the measurement of phospho-Chk1 (Ser³⁴⁵) by Western blot was obtained from Cell Signaling Technology (cat. no. 2341), and the antibody used for the measurement of phospho-Chk2 (Thr⁶⁸) was obtained from Rockland Immunochemicals, Inc. (cat. no. 600-401-280). The antibody used for the measurement of cyclin B1 by Western blots was obtained from Santa Cruz Biotechnology (cat. no. sc-752), and the antibody used for the measurement of phospho-cyclin B1 (Ser¹⁴⁷) by Western blots was obtained from Cell Signaling Technology (cat. no. 4131).

Inhibitory effect of administration of caffeine or caffeine sodium benzoate on UVB-induced increase in the level of phospho-Chk1 (Ser³⁴⁵) in the epidermis of SKH-1 mice. In recent studies, we identified caffeine sodium benzoate as a complex of caffeine and sodium benzoate that is more active than caffeine at enhancing UVB-induced apoptosis (27). Our hypothesis predicts that administration of caffeine or caffeine sodium benzoate will inhibit UVB-induced activation of the ATR/Chk1 pathway by inhibiting the phosphorylation of Chk1 at Ser³⁴⁵, thereby leading to increased cyclin B1, premature chromatin condensation, premature mitosis, and cell death in DNA-damaged cells (see Fig. 1).

P.o. administration of caffeine (0.4 mg/mL, 2.1 mmol/L) or caffeine sodium benzoate (2.1 mmol/L) in the drinking water to SKH-1 mice for 1 week before exposure to UVB (30 mJ/cm²) markedly inhibited the UVB-induced increase in the level of phospho-Chk1 (Ser³⁴⁵) at 6 h after UVB, but there was little or no effect on the level of phospho-Chk2 (Thr⁶⁸) as measured by Western blots (Fig. 2). Administration of caffeine or caffeine sodium benzoate had little or no effect on the level of phospho-Chk1 (Ser³⁴⁵) in the absence of UVB (Fig. 2). Densitometry measurements on Western blots from the epidermis in three separate experiments indicated that treatment with UVB plus caffeine or UVB plus caffeine sodium benzoate decreased the level of phospho-Chk1 (Ser³⁴⁵) by an average of 82% and 99%, respectively, compared with animals that received only UVB irradiation at 6 h after UVB (Fig. 2B).

In a time course study, p.o. administration of caffeine (0.4 mg/mL in the drinking water) to SKH-1 mice for 2 weeks increased UVB-induced apoptosis at 6 to 10 h after UVB (Fig. 3A) and inhibited the UVB-induced increase in the number of phospho-Chk1 (Ser³⁴⁵)–positive cells at 6 to 16 h after UVB (Fig. 3B).

Effect of administration of caffeine or caffeine sodium benzoate on UVB-induced changes in cyclin B1 and mitotic cells with cyclin B1 in the epidermis of SKH-1 mice. We evaluated the possibility that administration of caffeine enhanced UVB-induced apoptosis through abrogation of the cell cycle blockade at the G₂-M checkpoint by elevating epidermal cyclin B1 prematurely, thereby causing premature and lethal mitosis in the epidermis of UVB-treated animals (see Fig. 1). Irradiation of control mice with UVB (30 mJ/cm²) in the absence of caffeine resulted in a 38% decrease in the number of cyclin B1–positive cells (nuclear staining) at 6 h after UVB, and there was an even greater decrease in the number of mitotic cells with cyclin B1 (Fig. 3C and D). Irradiation with UVB (30 mJ/cm²) decreased the number of epidermal mitotic cells with cyclin B1 by 61% at 2 h after UVB, 62% at 4 h after UVB, 75% at 6 h after UVB, and 75% at 10 h after UVB (Fig. 3D), which allowed time for DNA repair before increased mitosis that occurred at later times after UVB exposure. P.o. administration of caffeine (0.4 mg/mL) to SKH-1 mice for 2 weeks before UVB irradiation blocked the normal UVB-induced decrease in cyclin B1 staining cells and mitotic cells with cyclin B1 at 6 to 10 h after UVB as measured by immunohistochemistry (Fig. 3C and D). An example of a cell undergoing mitosis that was also stained for cyclin B1 is shown in Fig. 4.

Our results show that p.o. administration of caffeine for 2 weeks before UVB irradiation increased the number of cyclin B1–stained cells in UVB-treated mice compared with control mice irradiated with UVB, and the time course for this effect of caffeine administration on cyclin B1 was similar to the time course for the inhibitory effect of caffeine on the UVB-induced increase in phospho-Chk1 (Ser³⁴⁵; Fig. 3B and C) and for the time course for the stimulatory effect of caffeine on UVB-induced apoptosis (Fig. 3A). Enhanced chromatin condensation (attempted mitosis) of DNA-damaged cells before repair and at an inappropriate early stage of the cell cycle would be expected to enhance cell death (Fig. 1). Our results indicate that administration of caffeine overcame the prolonged UVB-induced decrease in the percentage of mitotic cells with cyclin B1 that normally occurred after UVB (Fig. 3D). This effect of caffeine was associated with an increased level of cyclin B1 and enhanced early and inappropriate mitosis in the caffeine-treated mice.

It was of interest that the total number of mitotic cells in the epidermis after UVB or in the epidermis of mice pretreated with caffeine before UVB irradiation closely resembled the number of mitotic cells with cyclin B1 (data not presented). These results indicate that most mitotic epidermal cells in this study also had elevated cyclin B1.

P.o. administration of caffeine or caffeine sodium benzoate to SKH-1 mice for 1 week before exposure to UVB increased the level of cyclin B1 at 6 h after UVB when compared with the level of cyclin B1 in control mice irradiated with UVB as measured by Western blots at 6 h after UVB (Fig. 2). This treatment with caffeine or caffeine sodium benzoate, however, had little or no effect on the level of phospho-cyclin B1 (Ser¹⁴⁷; Fig. 2). Densitometry measurements on Western blots from the epidermis in three separate experiments as shown in Fig. 2 indicate that treatment with UVB plus caffeine or UVB plus caffeine sodium benzoate increased the level of cyclin B1 by an average of 153% and 201%, respectively, above the level of cyclin B1 in the epidermis of animals that received only UVB irradiation.

In another study, topical application of caffeine or caffeine sodium benzoate to SKH-1 mice immediately after irradiation with UVB increased by approximately 2- to 4-fold the number of cyclin B1–positive cells in the epidermis at 6 h after UVB (compared with UVB alone) as measured by immunohistochemistry (Table 1). There was little or no effect of UVB irradiation alone or with topical applications of caffeine or caffeine sodium benzoate immediately after UVB on the number of phospho-cyclin B1 (Ser¹⁴⁷)–positive cells in the epidermis (data not shown).

Comparison of the effects of p.o. administered caffeine on the time course for UVB-induced changes in apoptosis, phospho-Chk1 (Ser³⁴⁵), cyclin B1, and mitotic cells with cyclin B1 in the epidermis of SKH-1 mice. We observed that caffeine-induced changes in apoptosis, phospho-Chk1 (Ser³⁴⁵), and cyclin B1 (relative to UVB-treated control animals) all started at 6 h after UVB, and the greatest differences between UVB-treated control animals and UVB plus caffeine–treated animals occurred between 6 and 10 h after UVB (Fig. 3). Although the caffeine-induced increase in apoptosis in UVB-irradiated mice was maximal between 6 and 10 h after UVB and then declined (Fig. 3A), caffeine continued to affect UVB-induced changes in phospho-Chk1 (Ser³⁴⁵) and cells with cyclin B1 beyond 10 h (Fig. 3B and C). Decreases in apoptosis before the peak effects of caffeine on phospho-Chk1 and cyclin B1 may occur because of an antiapoptotic response after 10 h (such as an increase in E2F1, activated AKT, or survivin).

Treatment of SKH-1 mice with caffeine enhances UVB-induced apoptosis by both enhancing the level of p53 (7) and by a p53-independent pathway involving premature mitosis (Figs. 3 and 5). p53 is not required for caffeine-induced premature mitosis after UVB irradiation as indicated in p53 knockout mice (Fig. 5). It is important to note that the presence of a peak of cyclin B1–positive cells does not by itself lead to apoptosis, but rather, cyclin B1–positive cells that arise before 16 h after UVB (premature increase in cyclin B1 and premature mitosis) is what leads to apoptosis in caffeine-treated mice. In UVB plus water–treated mice, there is no increase in cyclin B1–positive cells (mitotic cells with cyclin B1) before 16 h (Fig. 3C and D). This indicates a functional “replication checkpoint.” At 16 h, the water-treated mice had keratinocytes that were undergoing mitosis as indicated by cyclin B1 expression; yet, because they had 16 h to recover from UVB damage, there was no corresponding increase in apoptosis (Fig. 3A, apoptosis is markedly declining by 16 h). In contrast, caffeine caused “premature mitosis” (visible at 6 and 10 h after UVB), and it is this early mitosis that then led to increased apoptosis in the caffeine-treated mice. The fact that the caffeine-treated mice went on to have a peak in cyclin B1–positive cells at 16 h but did not lead to apoptosis was because these cells began mitosis later and had had sufficient time to recover from UVB damage.

The data described in Fig. 3 at early time intervals after UVB irradiation provide support for our hypothesis that caffeine abrogates the UVB-induced blockade of the cell cycle at the G₂-M checkpoint by inhibiting the ATR/Chk1 pathway and prematurely elevating cyclin B1 and mitosis, thereby resulting in cell death (shown in Fig. 1).

Effects of caffeine on the time course for UVB-induced changes in apoptosis and mitotic cells with cyclin B1 in the epidermis of p53 knockout mice. In an earlier study, we found that topical application of caffeine enhanced UVB-induced apoptosis in p53^−/− C57BL/6J mice (9). In the present study, we used stored paraffin blocks from our earlier study to determine the time course for the effect of caffeine on UVB-induced changes in the percentage of mitotic cells with cyclin B1 (Fig. 5D). The results indicate a dramatic UVB-induced decrease in the percentage of mitotic cells with cyclin B1 in p53^−/− mice between 2 and 16 h after UVB, and this decrease was abrogated starting at 6 h after UVB in mice that were treated topically with caffeine immediately after UVB irradiation (Fig. 5D), indicating a p53-independent effect of caffeine in enhancing the number of mitotic cells with cyclin B1 at early times after UVB irradiation. The time course for the stimulatory effect of caffeine administration on the percentage of mitotic cells with cyclin B1 in p53 knockout mice treated with UVB paralleled the time course for the formation of apoptotic sunburn cells (Fig. 5C and D). The results also indicate that the magnitude of the caffeine effect is greater in p53^−/− mice than in their littermate wild-type mice (Fig. 5A  and B versus C and D).

In earlier studies, we found that p.o. administration of caffeine inhibited UVB-induced carcinogenesis (3) and enhanced UVB-induced apoptosis by p53-dependent and p53-independent mechanisms (7–9). In the present study, we investigated the effect of caffeine administration on UVB-induced activation of the p53-independent ATR/Chk1 pathway that normally inhibits mitosis at the G₂-M checkpoint after exposure to UVB, thereby allowing the cells time to repair their damaged DNA.

The mechanisms and molecular targets for the proapoptotic effect of caffeine after DNA damage have been investigated in cultured cell lines. A recent study indicated that ATR is an important proapoptotic target for caffeine in human osteosarcoma cells (13), and loss of p53 function sensitized these cells to premature chromatin condensation caused by the ATR inhibitor caffeine (13, 19). These investigators found that activation of ATR after DNA damage prevented premature chromatin condensation via Chk1 protein kinase regulation, and caffeine abrogated this effect, which resulted in premature chromatin condensation (13). Other studies indicated that caffeine directly disrupts the ATR/Chk1 checkpoint pathway (28). The ATR protein has higher affinity for DNA in UVB-damaged cells than for undamaged DNA, and damaged DNA stimulates the kinase activity of ATR to a significantly higher level than undamaged DNA (29).

Although the present study and others suggest that caffeine overcomes the G₂ checkpoint block by inhibiting the ATR-dependent phosphorylation of Chk1, other mechanisms may also play a role. A recent study indicated that although caffeine inhibited ATM and ATR kinase activity in vitro, studies done with cultured cells unexpectedly indicated that treatment with 2 to 8 mmol/L caffeine in combination with hydroxyurea stimulated ATR-dependent phosphorylation of Chk1 at Ser³⁴⁵ (30). One possible interpretation of this paradoxical result is that inhibition of ATR by caffeine causes stalled replication forks to convert to more severe forms of damage such as double-strand breaks, paradoxically augmenting damage signaling (31). In a recent study, 5 mmol/L caffeine enhanced apoptosis in mitotic checkpoint-arrested HeLa cells by mechanisms that seem to be independent of the ATR/Chk1/cyclin B1 signal transduction pathway (32). Inhibition of p21-activated PAK1, with an antiapoptotic function, was identified as a possible contributor to caffeine-induced apoptosis (32). It is important to point out that most cell culture studies used millimolar concentrations of caffeine that are 50- to 100-fold higher than what is achievable after p.o. administration of caffeine to mice or humans, although these concentrations are achievable after topical application.

In a recent study, we found that topical applications of caffeine to mice with patches of epidermal cells with one or more UVB-induced p53 mutations enhanced the elimination of these patches of p53-mutant cells (33). Interestingly, caffeine administration had selectivity and enhanced the elimination of cells with a p53 mutation on both alleles to a greater extent than cells with a p53 mutation on only one allele (34). These in vivo results are consistent with prior in vitro data that suggest a greater sensitivity to cell death by caffeine or ATR inhibition in p53-defective cells than in p53 wild-type cells (13).

Our results provide the first demonstration of an in vivo effect of caffeine administration on the ATR/Chk1/cyclin B1 pathway in animals, and this effect parallels the stimulatory effect of caffeine administration on UVB-induced apoptosis. In the present study, caffeine in the drinking water (0.4 mg/mL) given for 2 weeks before UVB treatment inhibited UVB-induced phosphorylation of epidermal Chk1 and abrogated the UVB-induced decrease in mitotic cells with cyclin B1 as well as the decrease in total mitotic cells. This is consistent with a model in which caffeine promotes premature mitosis and apoptosis through blocking ATR function as shown in the model in Fig. 1. Based on this model, we anticipated that caffeine-induced increases in epidermal cyclin B1 may be associated with changes in cyclin B1 phosphorylation. However, this was not observed using an antibody that can detect phosphorylation at Ser¹⁴⁷. It seems that the phosphorylation of cyclin B1 at positions other than Ser¹⁴⁷ plays an important role in the control of the cell cycle as shown by the work of Yang et al. (35). Cyclin B1 sites that are phosphorylated to promote nuclear import include Ser⁹⁴, Ser⁹⁶, Ser¹⁰¹, and Ser¹¹³ (35); however, phosphospecific antibodies have not been developed to investigate this process.

We explored the possibility that the effects of caffeine on UVB-induced carcinogenesis were mediated by inhibition of phosphodiesterase and an increased level of cyclic AMP (cAMP). Topical applications of 125 nmol of cAMP twice a day, 5 days a week for 21 weeks, to UVB-initiated high-risk mice had no effect on the percent of mice with tumors or the number of tumors per mouse but tumor size was increased (data not presented). In additional studies, treatment of mice with topical applications of dibutyryl cAMP or other compounds that increase cAMP had no effect on UVB-induced carcinogenesis (36). These studies indicate that alterations in cAMP levels that could occur after caffeine administration are not important for the inhibitory effect of caffeine on UVB-induced carcinogenesis.

The dose of p.o. caffeine used in the present study (0.4 mg/mL in the drinking water) as well as in earlier studies where caffeine inhibited UVB-induced carcinogenesis gave an average plasma level of ∼16 μmol/L (37), which is similar to that observed in people drinking three to five cups of coffee per day, and some coffee drinkers had much higher plasma levels of caffeine (38). Epidemiology studies suggest that coffee drinkers and tea drinkers have a lower risk of skin cancer (39–41). Whether drinking these caffeine-containing beverages modifies UVB-induced apoptosis and the ATR/Chk1 pathway in humans is unknown. Studies from our laboratory showed that administration of caffeine or caffeine sodium benzoate has sunscreen activity, increases the elimination of DNA-damaged cells by apoptosis, and inhibits UVB-induced carcinogenesis in SKH-1 mice (3, 4, 7, 8, 27, 42). Topical applications of caffeine or caffeine sodium benzoate to UVB-pretreated high-risk mice inhibited carcinogenesis in the absence of further UVB treatment (4, 27, 42), and administration of caffeine also enhanced apoptosis in the tumors (42). The results of these studies suggest that caffeine and caffeine sodium benzoate may be mechanistically novel agents for inhibiting sunlight-induced skin cancer.

In summary, the present study indicates that administration of caffeine to SKH-1 mice inhibits UVB-induced phosphorylation of Chk1 and prematurely increases the number of mitotic cells with cyclin B1 that are likely to go on to apoptosis. These effects are associated with an increase in UVB-induced apoptosis and inhibition of UVB-induced carcinogenesis.

Grant support: NIH grants CA114442, AR049832, and NJCCR05-1976-CCR-EO.

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.