Cross-Validation of Murine UV Signal Transduction Pathways in Human Skin[Dagger]

ABSTRACT Acute UVB irradiation of mouse skin results in activation of phospatidyinositol-3 (PI-3) kinase and mitogen- activated protein kinase (MAPK) pathways leading to altered protein phosphorylation and downstream transcription of genes. We determined whether activation of these pathways also occurs in human skin exposed to 4x minimal erythemic dose of UVB in 23 volunteers. Biopsies were taken prior to, at 30 min, 1 and 24 h post-UVB. In agreement with mouse studies, the earliest UV-induced changes in epidermis were seen in phospho-CREB (two- and five-fold at 30 min and 1 h) and in phospho-MAPKAPK-2 (three-fold at both 30 min and 1 h). At 1 h, phospho-c-JUN and phospho-p38 were increased five- and two-fold, respectively. Moreover, phospho-c-JUN and phospho-p38 were further increased at 24 h (12- and six-fold, respectively). Phospho- GSK-3beta was similarly increased at all time points. Increases in phospho-p53 (12-fold), COX-2 (four-fold), c-FOS (14-fold) and apoptosis were not seen until 24 h. Our data suggest that UVB acts through MAPK p38 and PI-3 kinase with phosphorylation of MAPKAPK-2, CREB, c-JUN, p38, GSK-3beta and p53 leading to marked increases in c- FOS, COX-2 and apoptosis. Validation of murine models in human skin will aid in development of effective skin cancer chemoprevention and prevention strategies.

INTRODUCTION

Over 1 million new skin cancers are diagnosed yearly in the United States accounting for approximately 40% of all new cancer cases. The incidence of skin cancer is increasing because of many factors including aging of the population and larger amounts of UV radiation reaching the earth’s surface because of depletion of the ozone layer (1). Approximately 80% of skin cancer diagnoses in the United States are basal cell carcinomas (BCC), 16% are squamous cell carcinomas (SCC) and 4% are melanomas. SCC involves the malignant transformation and proliferation of squamous cells, which are the most abundant cell type in the epidermis. SCC can become aggressive and metastasize, with approximately 1200 deaths reported in 1998 (a number equivalent to the yearly mortality attributed to Hodgkin’s lymphoma) (2-4). Furthermore, a high percentage of patients with SCC develop a second primary skin cancer within 5 years (5-7). Nonmelanoma skin cancers (NMSC) occur primarily on sun-exposed areas of the body and have been strongly associated with chronic sun exposure (8). Furthermore, the 3-year cumulative risk of recurrence in both BCC and SCC is at least 10 times the rate in a comparable general population (9). NMSCs are associated with relatively low mortality, but the morbidity of these cancers is high, treatment is often disfiguring and the resulting cost to society is substantial. In Medicare recipients over the age of 65 alone, approved physician charges for treatment of NMSC total at least $285 million per year (10). Efforts at chemoprevention of NMSC, especially in the southwestern region of the country where incidence rates are particularly high, must therefore be considered a public health priority.

UV irradiation can act as a complete carcinogen (11) and the UVB portion of the UV spectrum accounts for many of the direct harmful effects induced by sunlight. UV has both acute and chronic effects on mouse and human skin (12). Although not all of these effects have likely been described, it has been shown in both in vitro and mouse carcinogenesis models that acute UVB irradiation leads to the activation of critical molecular targets and multiple signal transduction pathways that result in the induction of expression of specific genes that lead to skin cancer (reviewed in Refs. [13- 17]). These effects have been shown primarily in cultured cells and mouse skin (13,18-21). UVB activates cell surface growth factor and cytokine receptors (13,22,23). This activation leads to stimulation of mitogen-activated protein kinase (MAPK) signal transduction pathways (18,23,24). MAPK are made up of three family members that include extracellular-signal-related protein kinases (ERKs), c-JUN N- terminal kinases/stress-activated protein kinases (JNKs/SAPs) and p38 kinases (25). Activation of these signaling cascades can result in a number of cellular responses that include apoptosis, proliferation, inflammation, differentiation and development (17). Physiologically relevant doses of UVB have been shown to activate p38, JNK and to a lesser extent ERK in vitro (26), while high doses of UVB in human skin have been shown to induce apoptosis through a p53-dependent pathway that likely removes damaged cells with the potential of reducing the risk of skin carcinogenesis (22,27). We and others have demonstrated that the p38 MAPK pathway and the transcription factor, activator protein-1 (AP-1) play a functional role in UVB-induced mouse skin carcinogenesis (12,23,28,29). UVB- mediated AP-1 activation occurs in part through the transcriptional activation and expression of the c-fos gene (30,31). Another UVB signaling pathway leading to c-fos transcriptional activation includes the phospatidyinositol-3 (PI-3) kinase-Akt signaling pathway (31-35). UVB also regulates the expression of another important target gene, cyclooxygenase 2 (COX-2) in keratinocytes (36- 38) and mouse skin (35) and this was found to be transcriptionally regulated through altered phosphorylation of the cyclic AMP response element binding protein (CREB) and its binding to a cyclic AMP response element (CRE) in the promoter region of the COX-2 gene (37,38).

Fisher et al. (23) demonstrated that exposure of human skin to low doses of UVB activates the EGF receptor, GTP-binding regulator protein p21RAS and stimulates the MAPKs (ERKs, JNKs and p38). Both JNK and p38 phosphorylation lead to the activation of the transcription factors c-JUN and ATF thereby increasing c-JUN expression. This group also reported that c-FOS was not induced by UVB, but was expressed constitutively, leading to increased levels of AP-1 (23).

The p53 tumor suppressor gene is a highly regulated protein which acts to stop cell cycle progression, influence DNA repair processes and to promote apoptosis when skin is exposed to UV light (39). Normally p53 is maintained at low levels but when cells are stressed ubiquitination is suppressed, resulting in p53 stability and nuclear accumulation (40). p53 activation is the result of a diverse set of posttranslational modifications that influence and determine which p53 target genes are induced. One of the most common of these modifications is phosphorylation of serines and/or threonines (39). Posttranslational modifications are both cell/tissue type specific and stimuli specific (39). Phosphorylation of p53 increases the sequence-specific DNA binding to p53 target genes (41). UVB has been shown to increase the accumulation and/or activation of p53 by phosphorylation at serine 15 and at serine 392 (42,43).

Many studies have illustrated these responses to UVB irradiation in mouse skin. However, work showing the translation of these findings in human skin is needed to validate these models in a clinical setting. Therefore, the objective of this study was to evaluate activation of MAP kinase, PI-3 kinase, p53 and JNK pathways in human skin following an acute dose of UVB. Similarities in response between mouse models and human skin support the use of murine models as a tool in the development of chemopreventive agents and strategies.

MATERIALS AND METHODS

Study population. Subjects were recruited through personal contact, distribution of flyers and media advertising (i.e. television, radio, newspapers). Healthy males and females aged 18 or older with fair skin and Fitzpatrick skin Types II (burns easily, tans poorly) or III (burns moderately, tans gradually) were recruited for the study. Exclusion criteria included immunosuppression, serious concurrent illness, presence of dyplastic nevi or invasive cancer (including any type of skin cancer or cancer treatment within the past 5 years) and baseline serum chemistry values outside of normal limits. In addition, those using photosensitizing agents or topical medications on the test area during the past 30 days were ineligible. Individuals taking mega doses of vitamins were not eligible (i.e. more than five times the RDA, more than five capsules of multivitamins, 400 IU of vitamin E, 200 [mu]g of selenium and 1 g of vitamin C). Additional exclusion criteria included individuals with a history of sun exposure to the buttocks within 30 days of randomization and participants must have agreed to avoid sun exposure during the study period. Finally, individuals with a known allergy to lidocaine were ineligible. The study was approved by the University of Arizona Institutional Review Board. Informed consent was obtained from all study participants.

Minimal erythemic dose. The minimal erythemic dose (MED) of UVB was determined for each individual using a Multiport UV Solar Simulator Model 600 (Solar Light Co., Philadelphia, PA). MED was defined as the smallest dose of energy necessary to produce confluent erythema with four distinct borders at 22-24 h postexposure. MED was determined on a buttock area previously unexposed to sunlight. Each test area was subdivided into six subsites (each 1 cm^sup 2^) corresponding to the liquid light guide pattern on the solar simulator. The solar simulator was calibrated prior to each use and a series of six increasing UVB radiation exposures (expressed as J cm^sup -2^) were administered to each subsite area. Following exposure, the test sites were covered until evaluations were completed. The spectrum used was UVB plus UVA (290- 390 nm). Although the emission spectrum contained UVA, the exposure time used to achieve a dose of 4x the MED of UVB was only 4 min and it would require significantly longer exposure times to obtain a biologically significant dose of UVA that would likely be significant, although we cannot rule out some contribution from UVA.

Administration of 4 MED. After determination of the MED for each individual, the contralateral buttock was exposed to four times that MED. A 4 mm skin punch biopsy sample was collected from one buttock at baseline prior to UVB exposure and additional 4 mm punch biopsies were removed at 30 min, 1 and 24 h post-UVB irradiation. Biopsy sites were then sutured and subjects returned to the clinic for suture removal at approximately 1 week.

Immunohistochemistry. Biopsies were immediately fixed in 10% neutral buffered formalin for 24 h then transferred to 70% ethanol prior to routine processing and paraffin embedding. Three-micron tissue sections were deparaffinized and rehydrated. All tissue sections were subjected to antigen retrieval using a citrate buffer in a Decloaking Chamber Pro (Biocare, Walnut Creek, CA) for 30 s. Immunohistochemical staining was performed using a streptavidinbiotin peroxidase system with a DAB chromagen and a hematoxylin counterstain (Ventana Medical Systems, Tucson, AZ) on an automated VMS 320 immunostamer (Ventana Medical Systems). Antibodies included those to phospho-MAPKAPK-2 (Thr 334) (rabbit polyclonal; Cell Signaling Technology, Beverly, MA) at 1:100, phospho-p38 (Thr180/Tyr182) (Biosource, Camarillo, CA) at 1:50, phospho-CREB (Ser 133) (rabbit monoclonal; Cell Signaling Technology) at 1:50, phospho-p53 (Ser15) (rabbit polyclonal; Cell Signaling Technology) at 1:50, phospho-c-JUN (Ser63) (rabbit monoclonal; Cell Signaling Technology) at 1:25, COX-2 (mouse monoclonal; Cayman Chemical, Ann Arbor, MI) at 1:100, Phospho-GSK-3beta (rabbit monoclonal; Cell Signaling Technology) at 1:100 and c-FOS (rabbit polyclonal; ABCAM, Cambridge, MA) at 1:80.

Tissue sections were measured using ImagePro Plus (Media Cybernetics, Silver Springs, MD) or Acuity(TM) (The BioAnalytics Group, Jamesburg, NJ) software systems and a Leica DMR microscope (Westzlar, Germany), and a Sony 3CCD color video camera (Japan). The percent positive nuclear or cytoplasmic area per 40x field was determined for each biopsy. Phospho-GSK-3beta was assessed on a scale of 0-3 + where 0 was negative, 0.5 was between negative and lightly positive, 1+ was lightly positive, 2+ was moderately positive and 3+ was intensely positive.

Apoptosis. The number of apoptotic cells was assessed based on morphology (i.e. condensed and/or pyknotic nuclei, eosinophilic cytoplasm, formation of apoptotic bodies) on H&E per 100 basal keratinocytes.

Statistical analysis. All analyses were conducted using Stata version 10 (StataCorp, College Station, TX). The percent positive area was calculated for each marker, with the exception of apoptosis and phospho-GSK-3beta. The distributions for the percent positive area were examined graphically and as a result of the nonnormal distributions nonparametric analyses were used. Data were available from biopsies that were taken prior to UVB exposure, at 30 min, 1 h post UVB exposure and at 24 h after UVB exposure. For each marker, expression levels at these time points were compared with expression levels in skin from a biopsy that had not been exposed to UVB radiation (baseline). Three comparisons were made for each marker- the difference between the percent positive area of the marker at 30 min versus unexposed skin; the difference between the percent positive area of the marker at 1 h versus unexposed skin; and the difference between the percent positive area of the marker at 24 h versus unexposed skin. The Sign-Rank test was used to test whether the median difference in expression level was statistically significant from 0.

For apoptosis, a ratio of the total number of apoptotic cells over the total number of basal cells was calculated. This was then converted to represent the proportion of apoptotic cells per 100 basal cells. Similar comparisons as described above were also made using the Sign-Rank test.

For phospho-GSK-3beta, the data were measured as an ordinal intensity that was rated as 0, 0.5, 1, 2 or 3+ based on staining intensity. These data were also compared by time of exposure and tested using the Sign-Rank test.

In Table 1 the median and arithmetic mean +- the standard deviation are shown for each marker. Also for all outcomes, excluding phospho-GSK-3beta, the change from baseline to any given time point was measured as a ratio or “fold-change” in the level of expression. Prior to the calculation of each ratio a value of 1 was added to each data point in order to avoid dividing by zero when the baseline expression was absent. Because the fold change was a ratio, the geometric means and 95% confidence intervals were calculated for each marker.

RESULTS

There were 10 males and 13 females in the study with an average age of 57.3 +- 15.3 years for males and 61.2 +- 13.4 years for females. There was not a significant difference in age between males and females. The average (+- SD) MED of UVB for the 23 participants was 25.79 +- 3.13 mJ cm^sup -2^ with a range of 21-31.5 mJ cm^sup – 2^ while the average (+-SD) total dose of UVB (4x MED) was 103.17 +- 12.53 mJ cm^sup -2^ with a range of 84-126 mJ cm^sup -2^. The UVA dose received by subjects (320-390 nm) ranged from 4-6 J cm^sup -2^ with 15 of 20 subjects receiving 5 J cm^sup -2^, four subjects receiving 4 J cm^sup -2^ and one subject receiving 6 J cm^sup -2^.

Figures 1-9 illustrate histologically an example of staining for each marker. Figure 1 shows nuclear phospho-p38 (threonine 180/ tyrosine 182), which was present at low levels at baseline (Fig. 1a). Expression of phospho-p38 was increased at 30 min (Fig. 1b) and 1 h (Fig. 1c), and was markedly increased at 24 h (Fig. 1d). The mean fold change (+- the 95% confidence interval) for the 23 subjects is shown in Fig. 10a with a 1.18 (0.78-1.77), 2.27 (1.56- 3.31) and 5.94 (3.99-8.84) fold change from baseline for phospho- p38 at 30 min, 1 and 24 h, respectively. Table 1 shows the median and mean (+- SD) for the percent positive area with statistically significant differences between baseline and 1 h (P = 0.0013), as well as between baseline and 24 h (P < 0.0001). Additionally, there was a statistically significant linear trend in phospho-p38 over the three time points (P < 0.0001).

Table 1. Median and mean +- standard deviation values of each marker.

Figure 1. Example of phospho-p38 (threonine 180/tyrosine 182) immunohistochemistry at baseline (a), 30 min post-UVB (b), 1 h post- UVB (c) and 24 h post-UVB (d). Brown nuclear staining is positive and blue is negative. The black arrows point to positive nuclei. All images are at a magnification of 400x.

Similarly, Fig. 2 demonstrates an example of phosphomitogen- activated protein kinase-activated protein kinase-2 (phospho- MAPKAPK-2) (threonine 334) at baseline (Fig. 2a), 30 min post-UVB (Fig. 2b), 1 h post-UVB (Fig. 2c) and 24 h post-UVB (Fig. 2d). In this example, there was a dramatic increase in the nuclear expression of phospho-MAPKAPK-2 throughout the epidermis at both 30 min, 1 and 24 h compared to baseline. The mean fold change (+- the 95% confidence interval) for the 23 subjects is shown in Fig. 10a with a 2.74 (1.92-3.90), 2.92 (1.95-4.37, n = 22) and 1.95 (1.58- 2.42) fold change from baseline for phospho-MAP-KAPK-2 at 30 min, 1 and 24 h, respectively. Table 1 shows the median and mean (+-SD) with statistically significant differences at 30 min (P = 0.0001), 1 h (P = 0.0001) and 24 h (P < 0.0001) compared to baseline.

Figure 3 shows phospho-CREB (serine 133) immunohistochemistry at baseline (Fig. 3a), 30 min post-UVB (Fig. 3b), 1 h post-UVB (Fig. 3c) and 24 h post-UVB (Fig. 3d) at a magnification of 400x. There was minimal, but increased, nuclear expression of phospho-CREB compared to baseline with a peak at 1 h. The mean fold change (+- the 95% confidence interval) for the 23 subjects is shown in Fig. 10a with a 2.47 (1.56-3.91), 5.46 (3.87-7.71) and 2.76 (1.86-4.08) fold change from baseline for phospho-CREB at 30 min, 1 and 24 h, respectively. Table 1 shows the median and mean (+- SD) with statistically significant differences at 30 min (P = 0.0001), 1 h (P < 0.0001) and 24 h (P = 0.0003) compared to baseline.

In this example of c-FOS immunohistochemisty there was increased nuclear c-FOS at 24 h (Fig. 4b) throughout the epidermis compared to baseline (Fig. 4a). The mean fold change (+- the 95% confidence interval) for the 23 subjects is shown in Fig. 10a with a 0.75 (0.61- 0.93), 0.98 (0.72-1.34, n = 22) and 12.75 (8.9-18.28) fold change from baseline for c-FOS at 30 min, 1 and 24 h post-UVB, respectively. Table 1 shows the median and mean (+- SD) with statistically significant differences at 30 min (P = 0.0272) and 24 h (P < 0.0001) compared to baseline.

Figure 2. Example of phospho-MAPKAPK-2 (threonine 334) immunohistochemistry at baseline (a), 30 min post-UVB (b), 1 h post- UVB (c) and 24 h post-UVB (d). All images are at a magnification of 400x. Brown nuclear staining is positive and blue is negative. The black arrows point to positive nuclei. In Fig. 5 we show an example of phospho-GSK-3beta (serine 9) at baseline (Fig. 5a), 30 min post- UVB (Fig. 5b), 1 h post-UVB (Fig. 5c) and 24 h post-UVB (Fig. 5d). Phospho-GSK-3beta was present at baseline and was expressed in the cytoplasm in a diffuse pattern throughout the epidermis. There was an increase in the expression of phospho-GSK-3beta throughout the epidermis at 30 min, 1 and 24 h compared to baseline. Due to the ordinal nature of the phospho-GSK-3beta data, the fold change could not be calculated. Table 1 shows the median and mean (+-SD) values for phospho-GSK-3beta expression. The mean (+- SD) values for phospho-GSK-3beta were 0.43 +- 0.27%, 0.96 +- 0.56%, 1.37 +- 0.31% and 1.09 +- 0.57% for baseline, 30 min, 1 and 24 h, respectively. These values were statistically significant at 30 min (P = 0.0001), 1 h (P = 0.0013) and 24 h (P < 0.0001) compared to baseline.

Figure 6 shows an example of COX-2 immunohistochemistry where in comparison with baseline (Fig. 6a), 30 min (Fig. 6b) and 1 h (Fig. 6c) there was increased cytoplasmic COX-2 at 24 h post-UVB (Fig. 6d) with expression primarily limited to the basal layer. The mean fold change (+- the 95% confidence interval) for the 23 subjects is shown in Fig. 10b with a 1.15 (0.72-1.82, n = 14), 1.3 (1.08-1.56) and 3.78 (2.90-4.92) fold change from baseline for COX-2 at 30 min, 1 and 24 h, respectively. Table 1 shows the median and mean (+-SD) with statistically significant differences at 1 h (P = 0.0047) and 24 h (P < 0.0001) compared to baseline.

Figure 7 gives an example of phospho-p53 (serine 15) immunohistochemistry where there was increased nuclear expression at 24 h post-UVB (Fig. 7b) with expression throughout the epidermis in comparison with baseline (Fig. 7a). The mean fold change (+- the 95% confidence interval) for the 23 subjects is shown in Fig. 10b with a 1.03 (1.0-1.06, n = 22), 1.09 (1.05-1.14) and 12.23 (10.58-14.13) fold change from baseline for phospho-p53 at 30 min, 1 and 24 h, respectively. Table 1 shows the median and mean (+- SD) with statistically significant differences 30 min (P = 0.0457), at 1 h (P = 0.0001) and 24 h (P < 0.0001) compared to baseline.

Figure 3. Example of phospho-CREB (serine 133) immunohistochemistry at baseline (a), 30 min post-UVB (b), 1 h post- UVB (c) and 24 h post-UVB (d). Images are at a magnification of 400x. Brown nuclear staining is positive and blue is negative. The black arrows point to positive nuclei.

Figure 4. Example of c-FOS immunohistochemistry at baseline (a) and 24 h post-UVB (b). Brown nuclear staining is positive and blue is negative. The black arrows point to positive nuclei. All images are at a magnification of 400x.

Figure 5. Example of phospho-GSK-3beta (serine 9) immunohistochemistry at baseline (a), 30 min post-UVB (b), 1 h post- UVB (c) and 24 h post-UVB (d). All images are at a magnification of 400x. Brown cytoplasmic staining is positive and blue is negative.

Figure 8 shows phospho-c-JUN (serine 63) immunohistochemistry where there was a modest increase in nuclear phospho-c-JUN expression at 30 min (Fig. 8b), a further increase at 1 h (Fig. 8c) followed by a dramatic increase at 24 h post-UVB (Fig. 8d) expressed throughout the epidermis in contrast to baseline (Fig. 8a) where expression was limited to the granular layer. The mean fold change (+- the 95% confidence interval) for the 23 subjects is shown in Fig. 10b with a 2.86 (2.12-3.86), 4.86 (3.56-6.65) and 12.1 (9.56- 15.32) fold change from baseline for phospho-c-JUN at 30 min, 1 and 24 h, respectively. Table 1 shows the median and mean (+- SD) with statistically significant differences at 30 min (P = 0.0001), 1 h (P < 0.0001) and 24 h (P < 0.0001) compared to baseline. Moreover, there was a statistically significant linear trend in phospho-c-JUN over the three time points (P < 0.0001).

Finally, Fig. 9 illustrates the increase in apoptotic cells at 24 h (Fig. 9b) as demonstrated by the arrows compared to baseline (Fig. 9a). The mean fold change (+- the 95% confidence interval) for the 23 subjects is shown in Fig. 10b with a 1.19 (1.01-1.39), 1.09 (0.88- 1.34) and 23.76 (18.16-31.07) fold change from baseline for apoptotic cells at 30 min, 1 and 24 h, respectively. Table 1 shows the median and mean (+- SD) with statistically significant differences at 24 h (P < 0.0001) compared to baseline.

DISCUSSION

The objective of this study was to cross-validate results observed in mouse skin models for the activation of the MAP kinase, PI-3 kinase, p53 and JNK pathways through the use of phospho- specific antibodies in human skin after an acute dose of 4x MED of UVB irradiation. We acknowledge the possibility that there was some small potential contribution of UVA to one or more of the markers in the study. A dose of 4x MED was chosen based on previous studies by Katiyar et al. (44) showing that at 4x MED the number of cyclobutane dimers was maximal and by Buckman et al. (45) demonstrating COX-2 expression at 24 h post-UVB irradiation. Our group and others had previously studied these pathways, although primarily in the setting of in vitro cell culture and mouse skin models (13,18,21,22,30- 33,36-38,46). We found that, in general, our results were in agreement with the previous findings from mouse skin studies and a more limited number of in vivo human studies (12,23,34,42,45).

Figure 6. Example of COX-2 immunohistochemistry at baseline (a), 30 min post-UVB (b), 1 h post-UVB (c) and 24 h post-UVB (d). All images are at 400x. Brown nuclear staining is positive and blue is negative. The black arrows point to positive cytoplasmic staining. All images are at a magnification of 400x.

As illustrated in Fig. 11a, we found that an acute dose of 4x MED of UVB to human skin resulted in the activation of the MAP kinase pathway with early phosphorylation of p38 at threonine 180/tyrosine 182 at 1 h and a further increase at 24 h. Previous studies in mouse epidermis (35,47) found that phosphorylation of p38 occurred within 15-30 min post-UVB. In addition, Bachelor et al. (48) found that phosphorylated p38 returned to basal levels by 24 h using a UVB dose of 600 mJ cm^sup -2^, while Kim et at. (47), using a UVB dose of 360 mJ cm^sup -2^, reported maximal phosphorylation of p38 at 24 h. In human skin Fisher et al. (23) reported that p38 MAPK was activated after an acute dose of 2x MED of UVB within 1-8 h with a return to baseline levels by 24 h using western blot analyses. Similarly, Pfundt et al. (12) reported increased phosphorylation of p38 within 2 h with a decrease at approximately 16-24 h using immunohistochemistry. In addition, Pfundt et al. (12) found that activation of p38 was dose related, with little activation seen at UVB doses below 1x MED. Our results showing a further increase in phospho-p38 at 24 h post-UVB are in agreement with the studies in mouse epidermis by Bachelor et al. (48) and Kim et al. (47). The most likely reason for the discrepancy between our study and the studies of Fisher et al. (23) and Pfundt et al. (12) may be the difference in the dose of UVB (i.e. 4x MED in our study compared with 2x MED). Higher doses of UVB (i.e. 4x MED) may have led to an increase in gene expression that was not evident at the lower dose of 2x MED. In addition, all three studies used different light sources, which may have contributed to the differences between studies. Moreover, we cannot rule out the potential effect of the small UVA dose that subjects received.

Figure 11a also shows that phosphorylation of p38 leads to activation of MAPKAPK-2 at threonine 334 and 222 (35,47), so we next studied phosphorylation of MAPKAPK-2 at threonine 334. In mouse epidermis Bachelor et al. (35) found an increase in phosphorylation of MAPKAPK-2 at threonine 222 between 15 min and 12 h with a decrease in levels by 24 h. Kim et at. (47) found activation of MAPKAPK-2 at both threonine 222 and 334 within 30 min but similar to that seen with p38 they reported that levels were still higher at 24 h compared to baseline. Our findings in human skin, using an antibody to threonine 334, were that phosphorylation of MAPKAPK-2 was maximal between 30 min and 1 h but remained elevated at 24 h compared to baseline. To our knowledge, this is the first report of increased phosphorylation of MAPKAPK-2 in human skin after UVB irradiation.

Figure 7. Example of phospho-p53 (serine 15) immunohistochemistry at baseline (a) and 24 h post-UVB (b). Brown nuclear staining is positive and blue is negative. The black arrows point to positive nuclei. All images are at a magnification of 400x.

Figure 8. Example of phospho-c-JUN (serine 63) immunohistochemistry at baseline (a), 30 min post-UVB (b), 1 h post- UVB (c) and 24 h post-UVB (d). Brown nuclear staining is positive and blue is negative. The black arrows point to positive nuclei. All images are at a magnification of 400x.

Figure 9. Example of apoptotic cells on hematoxylin and eosin- stained tissue sections at baseline (a) and 24 h post-UVB (b). Images are at a magnification of 400x, with an inset a magnification of 1000x in Fig. 1b (c). The black arrows show morphologically apoptotic cells.

Activation of MAPKAPK-2 can result downstream in phosphorylation of CREB at serine 133 (illustrated in Fig. 11a) as shown in mouse epidermis (35) with an increase in phosphorylation of CREB at serine 133 that occurred within 15 min post-UVB. In agreement with the study in mouse skin, in the current study we found activation and phosphorylation of CREB at serine 133 with an increase at 30 min and a peak at 1 h post-UVB in human epidermis. Again, to our knowledge, this is the first report of increased phosphorylation of CREB at serine 133 in human epidermis after an acute dose of UVB.

Phosphorylation of CREB at serine 133 has been shown to lead to increased transcription of c-fos (as illustrated in Fig. 11a), activation of AP-1, and ultimately to cellular events like cell proliferation and tumor promotion in cultured cells and mouse skin (30,31,33,35). Our findings in the current study indicated that exposure of human skin to 4x MED of UVB resulted in a maximal increase in expression of c-FOS at 24 h post-UVB. This is in agreement with the study in mouse epidermis by Bachelor et al. (48) where c-FOS was increased maximally at 24 h post-UVB. In contrast, in human skin using 2x MED of UVB Fisher et al. (23) found that c- FOS was not induced by UVB, but was expressed constitutively in human epidermis. Fisher et al. (23) used a dose of 2x MED while in our study we used a dose of 4x MED and this could account, at least in part, for the differences in the biopsies at 24 h postirradiation. Higher doses of UVB (i.e. 4x MED) may have led to an increase in gene expression that was not evident at the lower dose of 2x MED. In addition, Fisher et al. (23) do not describe the age of the population nor do they state whether the skin was from exposed or non-sun-exposed areas of the body and both of these variables could have an effect on the expression of c-FOS in un- irradiated skin. Another contributing factor may have been the differences in light sources used in the studies. Lastly, variations in immunohistochemistry methods may have contributed to differences between the study by Fisher and our study. Moreover, we cannot rule out the potential effect of the small UVA dose that subjects received.

Figure 10. (a) Graph showing the geometric mean +- the 95% confidence intervals of the fold difference from baseline for phospho-p38 (threonine 180/Tyrosine 182), phospho-MAPKAPK-2 (threonine 334), phospho-CREB (serine 133) and c-FOS at 30 min (black bars), 1 h (red bars) and 24 h (green bars), (b) Graph showing the geometric mean +- the 95% confidence intervals of the fold difference from baseline for COX-2, phospho-p53 (serine 15) and phospho-c-JUN (serine 63), and apoptosis at 30 min (black bars), 1 h (red bars) and 24 h (green bars).

UVB also mediates the activation of the PI-3 kinase pathway through phosphorylation of AKT at threonine 308 and serine 473, as illustrated in Fig. 11a. This has been previously shown in mouse epidermis (35) and in human skin (23,34). Phosphorylation of AKT inhibits GSK-30 by phosphorylating GSK-3beta at serine 9 (Fig. 11a). This in turn inhibits the phosphorylation of CREB at serine 129 which leads to increased transcription of COX-2 (29,33,35) (Fig. 11a). Phosphorylation of GSK-3beta at serine 9 has been previously demonstrated in cultured cells (32,38) as well as in mouse epidermis (35) after exposure to UVB. To our knowledge, this is the first study to show phosphorylation of GSK-3beta at serine 9 in human skin after an acute dose of UVB. In agreement with results from mouse epidermis (35), we found that phosphorylation of GSK-3beta occurred within 30 min after 4x MED of UVB.

COX-2 has been shown to be increased after UVB irradiation in both human skin and keratinocytes (45,49), in mouse epidermis (35), in skin reconstructs (29) and in human skin (45) with an increase at 12 h and maximal expression 24-48 h post-UVB. This can then lead to increased PGE^sub 2^ production, cell proliferation and tumor promotion (16). In agreement with these previous studies, we showed increased expression of COX-2 in human epidermis that was maximal at 24 h. Using artificial epidermis and western blot analyses, Mahns et al. (29) found that UVA doses of 1.6 and 2.4 5 J cm^sup -2^ resulted in a two- and three-fold increase, respectively.

Previous in vitro studies, mouse skin studies and studies in human skin have shown that UVB irradiation can lead to phosphorylation of p53 at serine 15 (42,50) through activation of p38, MAPKAPK-2 and ERKs (12,23,51), UVB can also activate JNK (23) which can then phosphorylate p53 at serine 20 (39,51) as illustrated in Fig. 11b. Phosphorylation of p53 results in a stabilization and accumulation of p53 in the nucleus and transcription of genes involved in cell cycle arrest, DNA repair and apoptosis (39). Furthermore, as illustrated in Fig. 11b, activation of JNKs can also lead to phosphorylation of c-JUN (12,23) and activation of c-JUN as a transcription factor for genes involved in cell proliferation, differentiation and apoptosis (12).

In the current study, we also found that an acute dose of 4x MED of UVB resulted in activation of the JNK and p53 pathways with early activation and phosphorylation of c-JUN at serine 63 at 1 h followed by a further increase in c-JUN phosphorylation at 24 h. Fisher et al. (23) also reported that c-JUN was activated within 2 h with a further increase at 16 h after a dose of 2x MED of UVB by western blot analyses. Additionally, we found that UVB irradiation also resulted in phosphorylation of p53 at serine 15 at 24 h. Similarly, Beattie et al. (42), using a dose of 3x MED of UVB in human skin reported increased phosphorylation of p53 at both serine 15 and 392 by immunohistochemistry. Beattie et al. (42) found no significant phosphorylation of p53 at serine 15 and 392 at doses of 20-60 J cm^sup -2^ UVA. Lastly, we found that apoptosis was markedly increased at 24 h post-UVB which is in agreement with the study by Beattie et al. (42).

Here we report the largest clinical study using immunohistochemistry to measure the expression of signal transduction proteins after an acute dose of UVB irradiation in human skin. Because of the number of subjects included in the study, we were able to show statistically significant changes in the expression of these markers between individual subjects even though we limited the study to Fitzpatrick skin Types II and III and the UVB dose used was similar between subjects. The magnitude of these differences is shown in Fig. 10a,b. Other factors besides skin type that may influence the absorption of UV in skin include skin thickness, thickness of the stratum corneum and other biologic differences (44).

In conclusion, we have further validated the body of work from mouse epidermis and a limited number of studies in human skin showing activation of AP-I and COX-2 through the p38 MAPK, PI-3 kinase, JNK and p53 pathways. These UVB signal components are likely targets for development of effective skin cancer chemoprevention strategies.

Figure 11. (a) UVB irradiation activates the MAP kinase pathway through phosphorylation of p38 which in turn phosphorylates MAPKAPK- 2 followed by phosphorylation of CREB at serine 133. Phosphorylation of CREB at serine 133 leads to increased transcription of c-FOS, activation of AP-1 and ultimately to cellular events that include cell proliferation and tumor promotion. UVB also mediates the activation of the PI-3 kinase pathway through phosphorylation of AKT. Activation of AKT results in the inhibition of GSK-3beta which can in turn inhibit the phosphorylation of CREB at serine 129 leading to increased transcription of the COX-2 gene. Increased COX- 2 can lead to increased PGE^sub 2^ production, cell proliferation and tumor promotion. Furthermore, phosphorylation of GSK-3beta at serine 9 and inhibition of phosphorylation of CREB at serine 129 can also lead to increased transcription of c-FOS. (b) UVB irradiation can lead to the phosphorylation of p53 at serine 15 through activation of p38 and ERKs. UVB can also activate JNKs, which can then phosphorylate p53 at serine 20. Phosphorylation of p53 results in a stabilization and accumulation of p53 in the nucleus and transcription of genes involved in cell cycle arrest, DNA repair and apoptosis. Activation of JNKs can also lead to phosphorylation of c- JUN and activation of c-JUN as a transcription factor for genes involved in cell proliferation, differentiation and apoptosis.

Acknowledgements-We would like to thank William Meek for assistance with measurement of biomarkers as well as Dr. Gary Pestano at Ventana Medical Systems, and Dr. Scott Lett and Ned Haubein of the Bioanalytics Group for use of the Acuity software system. This work was supported, in part, by PHS Grants CA27502 and CA23074 from the National Cancer Institute, National Institutes of Health. This publication is solely the responsibility of the authors and does not necessarily represent the official views of the National Cancer Institute.

[dagger] This paper is part of a special issue dedicated to Professor Hasan Mukhtar on the occasion of his 60th birthday.

REFERENCES

1. Johnson, T. M., O. M. Dolan, T. A. Hamilton, M. C. Lu, N. A. Swanson and L. Lowe (1998) Clinical and histologic trends of melanoma. J. Am. Acad. Dermatol. 38, 681-686.

2. Greenlee, R. T., T. Murray, S. Bolden and P. A. Wingo (2000) Cancer statistics, 2000. CA Cancer J. Clin. 50, 7-33.

3. Marks, R. (1995) An overview of skin cancers. Incidence and causation. Cancer 75, 607-612.

4. Moller, R., F. Reymann and K. Hou-Jensen (1979) Metastases in dermatological patients with squamous cell carcinoma. Arch. Dermatol. 115, 703-705.

5. Frankel, D. H., B. H. Hanusa and J. A. Zitelli (1992) New primary nonmelanoma skin cancer in patients with a history of squamous cell carcinoma of the skin. Implications and recommendations for follow-up. J. Am. Acad. Dermatol. 26, 720-726.

6. Karagas, M. R., T. A. Stukel, E. R. Greenberg, J. A. Baron, L. A. Mott and R. S. Stern (1992) Risk of subsequent basal cell carcinoma and squamous cell carcinoma of the skin among patients with prior skin cancer. Skin Cancer Prevention Study Group. JAMA 267, 3305-3310.

7. Preston, D. S. and R. S. Stern (1992) Nonmelanoma cancers of the skin. N. Engl. J. Med. 327, 1649-1662.

8. Kwa, R. E., K. Campana and R. L. Moy (1992) Biology of cutaneous squamous cell carcinoma. J. Am. Acad. Dermatol. 26, 1-26.

9. Marcil, I. and R. S. Stern (2000) Risk of developing a subsequent nonmelanoma skin cancer in patients with a history of nonmelanoma skin cancer: A critical review of the literature and metaanalysis. Arch. Dermatol. 136, 1524-1530.

10. Joseph, A, K., T. L. Mark and C. Mueller (2001) The period prevalence and costs of treating nonmelanoma skin cancers in patients over 65 years of age covered by medicare. Dermatol. Surg. 27, 955-959. 11. Strickland, P. T. (1986) Photocarcinogenesis by near-ultraviolet (UVA) radiation in Sencar mice. J. Invest. Dermatol. 87, 272275.

12. Pfundt, R., I. van Vlijmen-Willems, M. Bergers, M. Wingens, W. Cloin and J. Schalkwijk (2001) In situ demonstration of phosphorylated c-jun and p38 MAP kinase in epidermal keratinocytes following ultraviolet B irradiation of human skin. J. Pathol. 193, 248-255.

13. Bender, K., M. Gottlicher, S. Whiteside, H. J. Rahmsdorf and P. Herrlich (1998) Sequential DNA damage-independent and -dependent activation of NF-kappaB by UV. EMBO J. 17, 5170-5181.

14. Tyrrell, R. M. (1996) Activation of mammalian gene expression by the UV component of sunlight-From models to reality. Bioessays 18, 139-148.

15. Tyrrell, R. M. (1996) UV activation of mammalian stress proteins. EXS 77, 255-271.

16. Bowden, G. T. (2004) Prevention of non-melanoma skin cancer by targeting ultraviolet-B-light signalling. Nat. Rev. Cancer 4, 23- 35.

17. Bode, A. M. and Z. Dong (2006) Molecular and cellular targets. Mol. Cardnog. 45, 422-430.

18. Karin, M. and T. Hunter (1995) Transcriptional control by protein phosphorylation: Signal transmission from the cell surface to the nucleus. Curr. Biol. 5, 747-757.

19. Price, M. A., F. H. Cruzalegui and R. Treisman (1996) The p38 and ERK MAP kinase pathways cooperate to activate ternary complex factors and c-fos transcription in response to UV light. EMBO J. 15, 6552-6563.

20. Ramaswamy, N. T., Z. Ronai and J. C. Felling (1998) Rapid activation of JNK1 in UV-B irradiated epidermal keratinocytes. Oncogene 16, 1501-1505.

21. Fisher, G. J., S. C. Datta, H. S. Talwar, Z. Q. Wang, J. Varani, S. Kang and J. J. Voorhees (1996) Molecular basis of sun- induced premature skin ageing and retinoid antagonism. Nature 379, 335-339.

22. Webster, G. A. and N. D. Perkins (1999) Transcriptional cross talk between NF-kappaB and p53. Mol Cell. Biol. 19, 3485-3495.

23. Fisher, G. J., H. S. Talwar, J. Lin, P. Lin, F. McPhillips, Z. Wang, X. Li, Y. Wan, S. Kang and J. J. Voorhees (1998) Retinoic acid inhibits induction of c-Jun protein by ultraviolet radiation that occurs subsequent to activation of mitogen-activated protein kinase pathways in human skin in vivo. J. Clin. Invest. 101, 14321440.

24. Davis, R. J. (1993) The mitogen-activated protein kinase signal transduction pathway. J. Biol. Chem. 268, 14553-14556.

25. Kallunki, T., B. Su, I. Tsigelny, H. K. Sluss, B. Derijard, G. Moore, R. Davis and M. Karin (1994) JNK2 contains a specificity- determining region responsible for efficient c-Jun binding and phosphorylation. Genes Dev. 8, 2996-3007.

26. Rosette, C. and M. Karin (1996) Ultraviolet light and osmotic stress: Activation of the JNK cascade through multiple growth factor and cytokine receptors. Science 274, 1194-1197.

27. Ziegler, A., A. S. Jonason, D. J. Leffell, J. A. Simon, H. W. Sharma, J. Kimmelman, L. Remington, T. Jacks and D. E. Brash (1994) Sunburn and p53 in the onset of skin cancer. Nature 372, 773-776.

28. Cooper, S., J. MacGowan, J. Ranger-Moore, M. R. Young, N. H. Colburn and G. T. Bowden (2003) Expression of dominant negative c- jun inhibits ultraviolet B-induced squamous cell carcinoma number and size in an SKH-1 hairless mouse model. Mol. Cancer Res. 1, 848- 854.

29. Mahns, A., R. Wolber, F. Stab, L. O. Klotz and H. Sies (2004) Contribution of UVB and UVA to UV-dependent stimulation of cyclooxygenase-2 expression in artificial epidermis. Photochem. Photobiol. Sci. 3, 257-262.

30. Chen, W., A. H. Borchers, Z. Dong, M. B. Powell and G. T. Bowden (1998) UVB irradiation-induced activator protein-1 activation correlates with increased c-fos gene expression in a human keratinocyte cell line. J. Biol. Chem. 273, 32176-32181.

31. Gonzales, M. and G. T. Bowden (2002) Ultraviolet B (UVB) induction of the c-fos promoter is mediated by phospho-cAMP response element binding protein (CREB) binding to CRE and c-fos activator protein 1 site (FAP1 cis elements. Gene 293, 169-179.

32. Gonzales, M. and G. T. Bowden (2002) The role of PI 3-kinase in the UVB-induced expression of c-fos. Oncogens 21, 2721-2728.

33. Chen, W. and G. T. Bowden (1999) Activation of p38 MAP kinase and ERK are required for ultraviolet-B induced c-fos gene expression in human keratinocytes. Oncogens 18, 7469-7476.

34. Wan, Y. S., Z. Q. Wang, Y. Shao, J. J. Voorhees and G. J. Fisher (2001) Ultraviolet irradiation activates PI 3-kinase/AKT survival pathway via EGF receptors in human skin in vivo. Int. J. Oncol. 18, 461-466.

35. Bachelor, M. A., S. J. Cooper, E. T. Sikorski and G. T. Bowden (2005) Inhibition of p38 mitogen-activated protein kinase and phosphatidylinositol 3-kinase decreases UVB-induced activator protein-1 and cyclooxygenase-2 in a SKH-1 hairless mouse model. Mol. Cancer Res. 3, 90-99.

36. Chen, W., Q. Tang, M. S. Gonzales and G. T. Bowden (2001) Role of p38 MAP kinases and ERK in mediating ultraviolet-B induced cyclooxygenase-2 gene expression in human keratinocytes. Oncogens 20, 3921-3926.

37. Tang, Q., W. Chen, M. S. Gonzales, J. Finch, H. Inoue and G. T. Bowden (2001) Role of cyclic AMP responsive element in the UVB induction of cyclooxygenase-2 transcription in human keratinocytes. Oncogens 20, 5164-5172.

38. Tang, Q., M. Gonzales, H. Inoue and G. T. Bowden (2001) Roles of Akt and glycogen synthase kinase 3beta in the ultraviolet B induction of cyclooxygenase-2 transcription in human keratinocytes. Cancer Res. 61, 4329-4332.

39. Bode, A. M. and Z. Dong (2004) Post-translational modification of p53 in tumorigenesis. Nat. Rev. Cancer 4, 793-805.

40. Friedman, P. N., S. E. Kern, B. Vogelstein and C. Prives (1990) Wild-type, but not mutant, human p53 proteins inhibit the replication activities of simian virus 40 large tumor antigen. Proc. Natl Acad. Sci. USA 87, 9275-9279.

41. Hupp, T. R. and D. P. Lane (1994) Allosteric activation of latent p53 tetramere. Curr. Biol. 4, 865-875.

42. Beattie, P. E., L. E. Finlan, N. M. Kernohan, G. Thomson, T. R. Hupp and S. H. Ibbotson (2005) The effect of ultraviolet (UV) A1, UVB and solar-simulated radiation on p53 activation and p21. Br. J. Dermatol. 152, 1001-1008.

43. Kapoor, M., R. Hamm, W. Yan, Y. Taya and G. Lozano (2000) Cooperative phosphorylation at multiple sites is required to activate p53 in response to UV radiation. Oncogene 19, 358-364.

44. Katiyar, S. K., M. S. Matsui and H. Mukhtar (2000) Kinetics of UV light-induced cyclobutane pyrimidine dimers in human skin in vivo: An immunohistochemical analysis of both epidermis and dermis. Photochem. Photobiol. 72, 788-793.

45. Buckman, S. Y., A. Gresham, P. Hale, G. Hruza, J. Anast, J. Masferrer and A. P. Pentland (1998) COX-2 expression is induced by UVB exposure in human skin: Implications for the development of skin cancer. Carcinogenesis 19, 723-729.

46. Chen, W. and G. T. Bowden (2000) Role of p38 mitogen- activated protein kinases in ultraviolet-B irradiation-induced activator protein 1 activation in human keratinocytes. Mol. Carcinog. 28, 196-202.

47. Kim, A. L., J. M. Labasi, Y. Zhu, X. Tang, K. McClure, C. A. Gabel, M. Athar and D. R. Bickers (2005) Role of p38 MAPK in UVB- induced inflammatory responses in the skin of SKH-1 hairless mice. J. Invest. Dermatol. 124, 1318-1325.

48. Bachelor, M. A., A. L. Silvers and G. T. Bowden (2002) The role of p38 in UVA-induced cyclooxygenase-2 expression in the human keratinocyte cell line, HaCaT. Oncogene 21, 7092-7099.

49. An, K. P., M. Athar, X. Tang, S. K. Katiyar, J. Russo, J. Beech, M. Aszterbaum, L. Kopelovich, E. H. Epstein Jr, H. Mukhtar and D. R. Bickers (2002) Cyclooxygenase-2 expression in murine and human nonmelanoma skin cancers: Implications for therapeutic approaches. Photochem. Photobiol. 76, 73-80.

50. She, Q. B., N. Chen and Z. Dong (2000) ERKs and p38 kinase phosphorylate p53 protein at serine 15 in response to UV radiation. J. Biol. Chem. 275, 20444-20449.

51. She, Q. B., W. Y. Ma and Z. Dong (2002) Role of MAP kinases in UVB-induced phosphorylation of p53 at serine 20. Oncogene 21, 1580-1589.

Janine G. Einspahr*1,2, G. Timothy Bowden2,3, David S. Alberts1,2, Naja McKenzie2, Kathylynn Saboda2, James Warneke1,2,4, Stuart Salasche2, James Ranger-Moore2,5, Clara Curiel- Lewandrowski1,2,6, Raymond B. Nagle1,2, Brian J. Nickoloff7, Christine Brooks2, Zigang Dong8 and Steven P. Stratton1,2

1 Department of Medicine, University of Arizona, Tucson, AZ

2 Arizona Cancer Center, University of Arizona, Tucson, AZ

3 Cell Biology and Anatomy, Radiation Oncology, Pharmacology and Toxicology, Molecular and Cell Biology, University of Arizona, Tucson, AZ

4 Department of Surgery, University of Arizona, Tucson, AZ

5 Mel and Enid Zuckerman College of Public Health, University of Arizona, Tucson, AZ

6 Section of Dermatology, University of Arizona, Tucson, AZ

7 Department of Pathology, the Cardinal Bernardin Cancer Center, Loyola University of Chicago Medical Center, Maywood, IL

8 Hormel Institute, University of Minnesota, Austin, MN

Received 15 October 2007, accepted 5 December 2007, DOI: 10.1111/ j.1751-1097.2007.00287.x

* Corresponding author email: jeinspahr@azcc.arizona.edu (Janine Einspahr)

(c) 2007 The Authors. Journal Compilation. The American Society of Photobiology 0031-8655/08

Copyright American Society for Photobiology Mar/Apr 2008

Originally published by Einspahr, Janine G Bowden, G Timothy; Alberts, David S; McKenzie, Naja; Saboda, Kathylynn; Warneke, James; Salasche, Stuart; Ranger- Moore, James; Curiel-Lewandrowski, Clara; Nagle, Raymond B; Nickoloff, Brian J; Brooks, Christine; Dong, Zigang; Stratton, Steven P.

(c) 2008 Photochemistry and Photobiology. Provided by ProQuest Information and Learning. All rights Reserved.