Human skin responses to UV radiation: pigment ... - The FASEB Journal

The FASEB Journal • FJ Express Full-Length Article

Human skin responses to UV radiation: pigment in the upper epidermis protects against DNA damage in the lower epidermis and facilitates apoptosis Yuji Yamaguchi,* Kaoruko Takahashi,* Barbara Z. Zmudzka,† Andrija Kornhauser,‡ Sharon A. Miller,† Taketsugu Tadokoro,* Werner Berens,* Janusz Z. Beer,† and Vincent J. Hearing*,1 *Laboratory of Cell Biology, National Cancer Institute, National Institutes of Health, Bethesda, Maryland, USA; †Center for Devices and Radiological Health, Food and Drug Administration, Rockville, Maryland, USA; and ‡Center for Food Safety and Applied Nutrition, U. S. Food and Drug Administration, College Park, Maryland, USA Melanin plays an important role in protecting the skin against UV radiation, and melanomas and basal/squamous cell carcinomas occur more frequently in individuals with fair/light skin. We previously reported that levels of melanin correlate inversely with amounts of DNA damage induced by UV in normal human skin of different racial/ethnic groups. We have now separately examined DNA damage in the upper and lower epidermal layers in various types of skin before and after exposure to UV and have measured subsequent apoptosis and phosphorylation of p53. The results show that two major mechanisms underlie the increased photocarcinogenesis in fair/light skin. First, UV-induced DNA damage in the lower epidermis (including keratinocyte stem cells and melanocytes) is more effectively prevented in darker skin, suggesting that the pigmented epidermis is an efficient UV filter. Second, UV-induced apoptosis is significantly greater in darker skin, which suggests that UV-damaged cells may be removed more efficiently in pigmented epidermis. The combination of decreased DNA damage and more efficient removal of UV-damaged cells may play a critical role in the decreased photocarcinogenesis seen in individuals with darker skin.—Yamaguchi, Y., Takahashi, K., Zmudzka, B. Z., Kornhauser, A., Miller, S. A., Tadokoro, T., Berens, W., Beer, J. Z., Hearing, V. J. Human skin responses to UV radiation: pigment in the upper epidermis protects against DNA damage in the lower epidermis and facilitates apoptosis. FASEB J. 20, E630 –E639 (2006)

ABSTRACT

Key Words: pigmentation 䡠 melanocyte 䡠 melanosome 䡠 photoprotection Adverse effects in the skin following exposure to UV radiation are widely recognized (1–7) and can result in photocarcinogenesis, particularly in fair/light skin. Whether the photoprotection afforded the skin by melanin is due solely to its role as a UV filter, or whether other properties of melanin or skin are also involved, is an important question (8 –10). UnderstandE630

ing the role of racial/ethnic origin in determining individual UV sensitivity and DNA damage should help to elucidate the mechanisms of photocarcinogenesis since the rates of basal/squamous cell carcinomas and melanoma in the United States are 50 and 13 times higher in Whites than in Black or African-Americans, respectively (11–14). Indeed, the incidence of melanoma worldwide is increasing steadily. Pigmentation of the skin is determined by the types and amounts of melanin that melanocytes produce and can vary greatly among individuals of various racial/ethnic origins (15, 16). Melanocytes not only synthesize melanin within membrane-bound organelles (melanosomes) but also distribute those pigment granules to neighboring keratinocytes, which then carry them to the surface of the skin where they are lost by desquamation. Thus, interactions between melanocytes and keratinocytes are critically important to determining skin color, and we conducted studies to characterize the responses of melanocytes and keratinocytes in various types of skin to UV exposure. We previously reported the inverse relationship between melanin content and DNA damage induced by UV exposure in situ in normal human skin of different racial/ethnic groups (17). In that study, the minimal erythema dose (MED) was established for each subject who was then exposed to a single 1 MED exposure of UVA/UVB. Analysis of skin biopsied before, and 7 min, 1 d and 1 wk later, showed great variation among individuals in the amounts of DNA damage incurred and the rates of its removal. The skin of subjects from all groups suffered significant DNA damage, measured as cyclobutane pyrimidine dimers (CPD) and (6 – 4)-photoproducts (6,4PP), and increasing contents of constitutive melanin correlated inversely with amounts of DNA damage. Since supranuclear melanin caps often cover the 1

Correspondence: National Institutes of Health, Laboratory of Cell Biology, Bldg. 37, Rm. 2132, Bethesda, MD 20892-4254, USA. E-mail: [email protected] doi: 10.1096/fj.06-5725fje

0892-6638/06/0020-0630 U.S. government work not protected by U.S. copyright

nuclei of keratinocytes exposed to UV (18), the capacity of melanin to prevent DNA damage is highly significant. The distribution of melanin in the upper layers of the skin may play a significant role in determining its photoprotective value to underlying cells. UV-induced DNA damage to the lower epidermal layer of the skin may be more crucial to photocarcinogenesis than is damage to the upper layer since the lower epidermis contains not only melanocytes but also keratinocyte stem cells that are highly proliferative and are not lost via desquamation and thus could eventually give rise to various types of skin cancers. UV-induced apoptosis is closely associated with the appearance of sunburn cells (19, 20), but it is unclear whether they represent cell death caused by sudden and irreversible DNA damage or whether they result from cumulative DNA damage that was not repaired. The p53 tumor suppressor protein plays important roles in the inhibition of photocarcinogenesis, in part by enhancing the nuclear excision repair of UV-induced DNA damage (21) and by inducing apoptosis (1, 2, 22–25). Nuclear accumulation of high levels of p53 is observed in human skin in response to UV (21), suggesting the overall function of p53 in photoprotection. Phosphorylation of p53 at Ser-46 regulates the transcriptional activation of p53-dependent apoptosisinducing genes [including p53AIP1 (26)], whereas phosphorylation of p53 at Ser-15 and at Ser-20 regulates the transcriptional activation of G1-arrest genes (including p21Waf1) and of DNA repair genes (including p53R2) in response to UV-induced DNA damage (26 –28). There are significant differences in the skin among racial/ethnic groups with respect to the amounts and the distribution of melanin, although the expression levels of melanocyte-specific markers are remarkably similar as is the density of melanocytes in the different types of skin (29). In this study, we investigated DNA damage in the upper and lower layers of the epidermis in different racial/ethnic groups before and after exposure to 1 MED of UV, focusing on the DNA damage that occurs in melanocytes in those types of skin. We also investigated whether apoptotic cells are induced in epidermis exposed to UV, and we used reconstituted 3-dimensional human skin equivalents as a model to investigate UV-induced apoptosis in different types of skin. We also characterized the nuclear accumulation of p53 and its phosphorylation at Ser-46 following UV exposure of various types of skin. These results provide important insights into the effects of UV on skin of different racial/ethnic groups and help clarify why rates of photocarcinogenesis are significantly lower in dark skin than in fair skin. MATERIALS AND METHODS Study subjects and characteristics This study involved 92 volunteer subjects who represented the 6 different racial/ethnic groups defined by the United States

Office of Management and Budget (U.S. OMB Classification 0990 – 0208), as described previously (17). To be eligible for this study, subjects had to be at least 20 yr of age, in general good health, have a uniform skin color at the test sites (both sides of the lower back), be willing and able to follow the guidelines of the study, and to give informed consent. Exclusion parameters included a history of sun hypersensitivity or photosensitivity, the use of medication(s) that might affect photosensitivity, recent sunburn or use of tanning beds within 3 wk of entry into the study (or during the study), a history of skin cancer or any neoplastic disease, a history of excessive scarring or skin infections, any skin pigmentary abnormality and/or substance abuse. This study, which was approved by the FDA Research Involving Human Subjects Committee, reports on three major racial/ethnic groups of White, Intermediate (including Asian and Hispanic), and Black or African-American. UV irradiation A bank of fibrous sheath lamps (National Biological Corporation, Twinsburg, OH) was used as the source of UV radiation. The calculated erythemal effective energy (EEE) of the UVB content was 40% and that of the UVA was 60%, as previously reported (29). The lamps were positioned ⬃25 cm from the skin; a Kodacel filter (Eastman Chemical Products, Kingsport, TN) was used to remove the UVC component. The MED of each subject was determined at day –1 to assess the UV sensitivity on the midback, as described previously (17). We also studied 14 subjects of different skin types who were irradiated with a similar dose (180 –200 J/m2) to compare responses to the same UV dose among different types of skin. Skin biopsy Shave biopsies, ⬃4 mM in diameter, were taken from the midback before UV exposure, immediately (⬃7 min) after exposure, and then 1 d and 7 d later, as detailed in (29). Specimens were fixed in 10% formaldehyde and were paraffin-embedded using standard procedures. Paraffin sections were used to quantitate melanin content using FontanaMasson staining [as described previously (17)] and to study DNA damage and apoptosis by indirect immunofluorescence (as detailed below). DNA damage by immunofluorescence DNA damage in the form of CPD and/or 6,4PP was detected in paraffin sections by indirect immunofluorescence using mouse monoclonal antibodies (TDM-2 and 64M-2, respectively) (18). Samples were sectioned at 3 ␮m thickness and were mounted on silane-coated glass slides. They were deparaffinized twice with xylene for 5 min and were then dehydrated with a graduated series of ethanol, followed by antigen retrieval via boiling in antigen unmasking solution (Vector Laboratories, Inc. Burlingame, CA) for 12 min. They were subsequently incubated with 20% goat serum (Vector Laboratories) for 30 min at 37°C, and then with TDM-2 (at 1:40,000 dilution) or with 64M-2 (at 1:1,000 dilution) in 5% goat serum at 4°C overnight. Bound antibodies were visualized with the secondary antibody (Ab), Alexa Fluor 488 goat antimouse IgG (H⫹L) (Molecular Probes, Inc., Eugene, OR) at 37°C for 30 min at 1: 500 dilution with 5% goat serum. Nuclear DNA was counterstained with 1 ␮g/ml propidium iodide (PI; Sigma, St. Louis, MO). Immunohistochemistry was also performed using p53 and phospho-p53 (Ser-46 (1:100 dilution, Cell Signaling Technology, Beverly, MA) rabbit

HUMAN SKIN RESPONSES TO UV: PHOTOPROTECTION BY MELANIN

E631

primary antibodies and Alexa Fluor 488 goat anti-rabbit IgG (H⫹L) (1:500 dilution, Molecular Probes) as a secondary Ab. The green fluorescence produced by Alexa Fluor 488 and the red fluorescence produced by PI was observed and analyzed using a Leica DMR B/D MLD fluorescence microscope (Leica, Wetzlar, Germany), a Dage-MTI 3CCD 3-chip color video camera (Dage-MTI, MI City, IN), and Scion Image software (Scion, Frederick, MD), as described previously (17). This system allows one to eliminate background fluorescence and to quantitate fluorescence intensity from the original images; each value was recorded as integrated density over a given epidermal area. The formation of CPD or 6,4PP in epidermal nuclei was then expressed as the ratio of the intensity of green fluorescence (for DNA damage) over the intensity of red fluorescence (for localization of nuclei). Ten randomly selected areas of each specimen were photographed and quantitated; % sem was usually ⬍⫾5%. In addition to investigating the overall epidermal DNA damage, each epidermal image captured was separated into upper and lower halves by demarcating the area to analyze the ratio of DNA damage into the upper and lower epidermis. Sections from the same subject (taken immediately after UV) served in each experiment as an internal control for DNA damage.

er’s protocol to stain the paraffin-embedded tissue. This kit is based on the terminal deoxynucleotidyl transferase-mediated dUTP-nick end-labeling (TUNEL) assay. This procedure was followed by further staining with phospho-p53 (Ser-46, a rabbit primary Ab, and Alexa Fluor 594 goat anti-rabbit IgG (H⫹L) as a secondary Ab to double-stain the specimens. Specimens were also analyzed by staining for cleaved caspase-3 using the Asp175 Ab (at 1:10 dilution) obtained from Cell Signaling Technology (Beverly, MA).

Melanocyte-specific DNA damage

MelanoDerm (MatTek Corp., Ashland, MA) is a viable reconstituted 3-dimensional human epidermis containing human melanocytes and keratinocytes, as described previously (31). We used MelanoDerm containing keratinocytes derived from a Hispanic donor and melanocytes derived from Black, Asian, or White donors. Cultures were maintained in medium at the air/liquid interface according to the manufacturer’s instructions. Human skin reconstructs in culture were irradiated using a UV source different from that used for the subjects as described above. Melanoderm cultures were irradiated (at 25 J/m2 or 50 J/m2) using two Philips TL 20W/12RS lamps (Philips, Somerset, NJ) that emit 64% of their total energy within the UVB range and 36% within the UVA range. UV levels were determined using a PMA2100 UV radiometer (Solar Light, Philadelphia, PA) and a Kodacel filter (Eastman Chemical Products, Kingsport, TN) was used to remove the UVC component, as described previously (32). The cultures

In addition to the immunohistochemistry described above, a rabbit polyclonal antibody to tyrosinase (␣PEP7h, at 1:1,500 dilution) (30) was coincubated with TDM-2 as primary antibodies for the detection of DNA damage in melanocytes. This was followed by visualization with appropriate secondary antibodies, Alexa Fluor 488 goat antimouse IgG (H⫹L) or Alexa Fluor 594 goat anti-rabbit IgG (H⫹L). Nuclear DNA was counterstained with 1 ␮g/ml 4⬘,6⬘-diam idino-2-phenylidole (Vector Laboratories). Apoptosis, p53 phosphorylation and caspase-3 activation An ApopTag in situ apoptosis detection kit (Serologicals Corp., Norcross, GA) was used according to the manufactur-

Melanin content Paraffin-embedded specimens were sectioned at 3 ␮m thickness and were stained using the Fontana-Masson method. Melanin quantity was analyzed using the Leica microscope from the integrated density in given areas of the epidermis in each section, as described previously (17). We previously reported a significant correlation between eumelanin content measured chemically and melanin content determined by Fontana-Masson staining (17). Reconstructed 3-dimensional human skin

Figure 1. Morphology of Fair, Intermediate, and Dark human skin according to (A) melanin content, stained by Fontana-Masson, and (B) melanocyte distribution, stained by MITF. Panels show unexposed skin, and skin 1 d or 1 wk after a single 1 MED UV exposure. (———) demarks the top of the granular layer of the epidermis and (- - - -) demarks the epidermal: dermal junction.

E632

Vol. 20

July 2006

The FASEB Journal

YAMAGUCHI ET AL.

were kept at room temperature during the irradiation. They were harvested 2 d later and were fixed with 10% formaldehyde for the paraffin-embedded sectioning and determination of apoptosis, as described above. Statistical analyses The statistical software, JMP 5.1, was used to determine t values and correlation coefficients. A P value of ⬍0.05 is defined as significant using paired or unpaired t tests. The r square of the correlation coefficient is expressed with logarithmic regression, although the values expressed by linear regression were similar. Means ⫾ sd are used in this study.

RESULTS DNA damage in different types of skin Representative images of melanin content and distribution of melanocytes in fair, intermediate and dark skin are shown in Fig. 1. Although the melanin content varies greatly between those skin types, melanocyte density before and up to 1 wk after UV exposure is relatively constant, as previously detailed (29). UV irradiation of human skin results in two major types of DNA lesions; 6,4PP and CPD (33), and in this study we measured the production of those lesions and determined their distribution in skin of varying racial/ ethnic origin. We first determined those parameters for 6,4PP (details for CPD are presented below), and we compared 6,4PP in the lower epidermis with that in the upper epidermis at various times after a single 1 MED UV exposure. Representative images of 6,4PP DNA Table 1.

Figure 2. Representative images of 6,4PP DNA damage in Fair, Intermediate, and Dark human skin, before and 7 min, 1 d, and 7 d after a single 1 MED UV exposure. Green and red fluorescence represent 6,4PP and DNA, respectively. (———) demarks the top of the granular layer of the epidermis, (- - - -) demarks the epidermal:dermal junction, and (䡠䡠䡠䡠䡠) represents the division between the upper and lower epidermal layers.

DNA damage in various types of skin for all subjects analyzed

Skin type

6,4PP content Fair (n⫽4) J/m2 ⫽ 285 ⫾ 42 Intermediate (n⫽5) J/m2 ⫽ 388 ⫾ 55 Dark (n⫽4) J/m2 ⫽ 639 ⫾ 54 CPD content Fair(n⫽7) J/m2 ⫽ 291 ⫾ 43 Intermediate (n⫽5) J/m2 ⫽ 388 ⫾ 55 Dark (n⫽7) J/m2 ⫽ 601 ⫾ 40

Time

Total

Pa

Upper

Pa

Pb

Lower

Pa

7 1 7 7 1 7 7 1 7

min day days min day days min day days

0.85 ⫾ 0.13 0.20 ⫾ 0.13 0.15 ⫾ 0.15 0.55 ⫾ 0.14 0.08 ⫾ 0.03 0.04 ⫾ 0.04 0.22 ⫾ 0.06 0.07 ⫾ 0.03 0.02 ⫾ 0.01

– – – NS NS NS * NS NS

0.96 ⫾ 0.15 0.26 ⫾ 0.16 0.20 ⫾ 0.19 0.65 ⫾ 0.16 0.11 ⫾ 0.03 0.06 ⫾ 0.05 0.32 ⫾ 0.08 0.11 ⫾ 0.04 0.04 ⫾ 0.00


NS NS NS ** * NS ** * NS

0.74 0.15 0.09 0.45 0.05 0.02 0.13 0.03 0.00

⫾ 0.13 ⫾ 0.11 ⫾ 0.11 ⫾ 0.13 ⫾ 0.03 ⫾ 0.04 ⫾ 0.05 ⫾ 0.03 ⫾ 0.02


7 1 7 7 1 7 7 1 7


0.71 ⫾ 0.06 0.52 ⫾ 0.09 0.22 ⫾ 0.04 0.52 ⫾ 0.02 0.26 ⫾ 0.05 0.12 ⫾ 0.02 0.45 ⫾ 0.06 0.22 ⫾ 0.04 0.09 ⫾ 0.04

– – – * * NS ** * *

0.72 ⫾ 0.06 0.56 ⫾ 0.09 0.25 ⫾ 0.04 0.63 ⫾ 0.04 0.35 ⫾ 0.06 0.16 ⫾ 0.03 0.61 ⫾ 0.07 0.34 ⫾ 0.05 0.15 ⫾ 0.04

– – – NS NS NS NS * NS

NS * ** ** ** ** ** ** **

0.69 0.48 0.18 0.42 0.17 0.08 0.29 0.10 0.03

⫾ 0.07 ⫾ 0.10 ⫾ 0.05 ⫾ 0.01 ⫾ 0.04 ⫾ 0.02 ⫾ 0.04 ⫾ 0.04 ⫾ 0.04

– – – ** * NS ** ** *

The mean 1 MED UV dose (and n) is shown in J/m2 for each group. Background staining in unirradiated controls has been subtracted from the data at 7 min, 1 day, and 7 days. Values indicate means of intensity measured by Scion image ⫾ se a P value versus Fair Skin in the same layer at the same time point. b P value versus DNA damage in the upper layer of the same type of skin at the same time point. NS, not significant; *P ⬍ .05; **P ⬍ .01.


E633

damage in fair, intermediate, and dark skin before and at various times after UV exposure are shown in Fig. 2 and data for all subjects in each skin type are summarized in Table 1. The 1 MED dose of UV was on average ⬃3 times higher in dark skin (822⫾247 J/m2, n⫽17) than in fair skin (293⫾94 J/m2, n⫽42, P⬍0.001 vs. dark), and was ⬃30% higher in intermediate skin (522⫾262 J/m2, n⫽10, P⬍0.01 vs. dark) than in fair skin (P⬍0.025 vs. intermediate). Despite that, overall levels of 6,4PP were 3.9-fold lower (P⬍0.05) in dark skin immediately after UV exposure compared with fair skin. DNA damage in the lower epidermis was significantly lower than in the upper epidermis immediately after UV in intermediate (1.4-fold, P⬍0.05) and in dark (2.5-fold, P⬍0.01) skin, but not in fair skin, which suggests that the upper epidermis has a photoprotective effect against UV-induced 6,4PP and that this effect is more remarkable in darker skin. Although the bulk of melanin visible by Fontana-Masson staining appears to be in the basal layer of the epidermis in Fig. 1, a significant amount (⬃30 – 40% of the total) is actually in the upper layers of the epidermis, as detailed previously (29). However, no significant differences were found in 6,4PP levels at days 1 and 7 after UV exposure among the racial/ethnic groups, which demonstrates that the repair of 6,4PP is quick and occurs at similar rates among those various types of skin. Melanin content correlated inversely with the amount of 6,4PP damage immediately after UV in the total epidermis (r2⫽.457, P ⬍ 0.02), in the upper epidermis (r2⫽.490, P⬍0.02) and in the lower epidermis (r2⫽.408, P⬍0.05) (data not shown), again supporting the photoprotective effect of melanin in all epidermal layers. Figure 3A shows representative images of DNA damage as CPD for subjects with fair, intermediate, and dark skin immediately after UV exposure and 1 wk later; data for all subjects are shown in Table 1. CPD damage in the lower epidermis of fair skin did not differ significantly from that in the upper epidermis immediately after UV exposure, or at 1 d or 7 d later. However, CPD damage in the lower epidermis of intermediate skin and more dramatically of dark skin was at least 1.5-fold (P⬍0.01) and at least 2.1-fold (P⬍0.01) lower, respectively, than in the upper epidermis at all times (7 min, 1 and 7 d) after UV exposure. Consistent with the 6,4PP damage, levels of CPD damage in the lower epidermis were higher in fair skin after UV than in dark skin at all time points examined, again despite the fact that the darker skin had received ⬎2fold more UV to reach 1 MED. These results show that UV penetrates deeper in less pigmented skin to generate CPD and that the repair of CPD is impaired in fair skin compared with dark skin after the 1 MED UV exposure. CPD damage in total upper and lower epidermis was again inversely correlated with melanin content (data not shown); the correlation coefficient in the lower epidermis (r2⫽.740, P⬍0.001) and in the total epidermis (r2⫽.598, P⬍0.001) was much higher than in the upper epidermis (r2⫽214, P ⬍ 0.05) immediately after UV. This finding suggests that E634

Vol. 20

July 2006

Figure 3. A) Representative images of CPD DNA damage in Fair, Intermediate, and Dark human skin immediately and 7 d after UV exposure; green and red fluorescence represent CPD and DNA, respectively. B) Representative images of CPD (green) in melanocytes (stained red for tyrosinase) immediately after UV in Fair, Intermediate, and Dark human skin. (———) demarks the top of the granular layer of the epidermis, (- - - -) demarks the epidermal:dermal junction, and (䡠䡠䡠䡠䡠) represents the division between the upper and lower epidermal layers.

skin containing more melanin incurs less DNA damage in the lower epidermis and that DNA damage in the upper epidermis is similar among racial/ethnic groups immediately after UV. The melanin content correlated inversely with CPD damage even in the upper epidermis at 1 d (r2⫽.393, P⬍0.01) and at 7 d (r2⫽.307, P⬍0.02) after UV, and this was more significant in the lower epidermis at 1 d (r2⫽.552, P⬍0.001) and at 7 d (r2⫽.410, P⬍0.01). The sum of these results demonstrates that: 1) skin containing more melanin suffers significantly less CPD damage not only in the upper epidermis but also in the lower epidermis; 2) the initial DNA damage the subjects suffer correlates inversely with the repair of CPD; and 3) recovery from CPD damage in the epidermis takes significantly longer than seen for 6,4PP. Since DNA damage detected as CPD immediately after 1 MED UV was significant, even deep in the epidermis, we investigated DNA damage in melanocytes measured by costaining for CPD and for tyrosinase (the critical enzyme in melanin synthesis (30). Melanocytes were stained red for tyrosinase and CPD damage was stained green (Fig. 3B). The % melano-

The FASEB Journal

YAMAGUCHI ET AL.

Figure 4. A) % CPD-positive melanocytes in Fair, Intermediate, and Dark human skin 1 d after a single 1 MED UV dose; (B) % CPD-positive melanocytes with detectable levels of CPD sorted by melanin content [(same subjects as in (A)]. C) Graphical representation of CPD DNA damage sorted by melanin content in Fair, Intermediate, and Dark skin, at 7 min, 1 d, and 7 d after 1 MED UV exposure.

cytes with detectable levels of CPD was higher (2.9-fold, P⬍0.05 and 7.5-fold, P⬍0.01, respectively) in fair skin than in intermediate and in dark skin (Fig. 4A), and the constitutive melanin content of the skin correlated inversely with the % melanocytes with DNA damage immediately after UV (Fig. 4B). The inverse correlation of CPD DNA damage with melanin content in the different types of skin at various times after UV exposure is shown in Fig. 4C. Melanin facilitates the induction of apoptosis by UV We used the TUNEL assay to measure apoptotic cells in the skin of different racial/ethnic groups after exposure to 1 MED or to a similar dose of 180 –200 J/m2 UV. Surprisingly, 7-fold more apoptotic cells were observed in dark skin than in fair skin after 1 MED UV (data not shown), although the DNA damage in dark skin was significantly less than in fair and intermediate skin, as

noted above. To rule out that the higher physical UV dose at 1 MED used for dark skin elicited the increase of TUNEL-positive cells, we performed a similar study but using a similar dose (180 –200 J/m2). Representative images of TUNEL-positive cells in fair, intermediate and dark skin at 1 d after UV exposure are shown in Fig. 5A; data for all subjects are summarized in Table 2. TUNEL-positive apoptotic cells were observed at more than 5-fold higher levels in dark skin 1 d after UV than in fair skin. Increased levels of melanin correlated directly with the number of TUNEL-positive cells in the skin at 1 d (r2⫽.350, P⬍0.05) and at 7 d (r2⫽.312, P⬍0.05) after UV exposure (data not shown), demonstrating that skin containing more melanin undergoes significantly more apoptosis in response to a single low UV dose. We hypothesized that it might be the melanin within keratinocytes that is involved in the increased apoptosis after UV exposure since the only apparent difference


E635

Figure 5. A) TUNEL staining in Fair, Intermediate, and Dark human skin 1 d after exposure to 180 –200 J/m2 UV; green and red fluorescence represent TUNEL and DNA, respectively. B) TUNEL staining in reconstituted 3-dimensional human skin equivalents containing melanocytes derived from Black, Asian, or White donors and keratinocytes from the same Hispanic donor. The cultures were UVB-irradiated at 50 J/m2 and were harvested 2 d later; green and red fluorescence represent TUNEL and DNA, respectively. C) p53 phosphorylated at Ser-46 in Fair, Intermediate and Dark human skin 1 d after 180 –200 J/m2 UV; green and red fluorescence represent TUNEL and p53 phosphorylated at Ser-46, respectively, (———) demarks the top of the granular layer of the epidermis, (- - - -) demarks the epidermal:dermal junction, and (䡠䡠䡠䡠䡠) represents the division between the upper and lower epidermal layers.

among skin in various racial/ethnic groups is the amount and distribution of melanin pigment. To examine this, we used reconstructed 3-dimensional human skin equivalents (termed MelanoDerm) containing melanocytes derived from Black, Asian, or White donors and keratinocytes from one Hispanic donor. The differences in melanin distribution are significant and they represent typical skin morphology of those types of skin (31). The MelanoDerm cultures were unirradiated or were UVB-irradiated at 25 or 50 J/m2 and were then harvested 2 d later. TUNEL assays showed that significantly more apoptotic cells were found in Black skin equivalents than in Asian or White skin equivalents at both UV doses (Fig. 5B, data summarized in Table 2). Phosphorylation of p53 and caspase-3 activation after UV exposure The oncogene p53 plays important roles in responses to UV-induced DNA damage and induction of DNA E636

Vol. 20

July 2006

repair and there is an overall nuclear accumulation of p53 in response to UV (21). More than 13 sites of p53 are phosphorylated, one of them a critical site at Ser-46, which is associated with the induction of apoptosis (26). We investigated the effects of UV on the accumulation of p53 in nuclei and its phosphorylation at Ser-46 in skin from the different racial/ethnic groups (Figs. 5C and 6). More p53 accumulated in the nuclei of cells in fair skin than in dark skin at 1 d and at 7 d (Fig. 6, upper row) after UV exposure. However, phosphorylation of p53 at Ser-46 was not seen in fair skin, whereas it was readily seen in dark skin 1 d after UV (Fig. 6, lower row). Overall, the nuclear expression of p53 phosphorylated at Ser-46 colocalized with TUNEL-positive cells, as shown in yellow (Fig. 5C). Increasing melanin content correlated inversely with the number of cells with nuclear accumulation of p53 at 1 d and at 7 d after UV exposure but correlated positively with the number of cells containing phosphorylated p53 at Ser-46 at 1 d and at 7 d after UV exposure (Fig. 7). In an effort to determine whether the TUNEL staining observed truly reflects apoptosis, we stained the specimens with various antibodies detecting mediators of apoptosis such as caspase 3. Staining for cleaved caspase-3 (a specific indicator of apoptosis) revealed highest levels of that 1 d after UV exposure in dark, intermediate, and fair skin, and that levels of cleaved caspase-3 had returned to baseline levels by day 7 (data not shown). The sum of this evidence suggests that the apoptosis occurs through phosphorylated p53-dependent but caspase-3 independent pathways.

DISCUSSION UV irradiation of human skin results in two major types of DNA lesions, 6,4PP and CPD (33), and the yield of both types of DNA photoproducts is lower in tissues with higher melanin content in vivo (17, 20, 34, 35) and in vitro (36). DNA damage in the upper epidermis immediately after UV exposure was similar among racial/ethnic groups since levels of CPD damage in the different types of skin were similar. The correlation between constitutive melanin content and immediate CPD damage was greater in the lower epidermis (P⬍0.001) than in the upper epidermis (P⬍0.05). Taken together, we conclude that the upper epidermis of dark skin is significantly more photoprotective for the deeper tissue against UV damage than is that of fair skin. The location of TUNEL-positive cells in the middle to upper layers of the epidermis following UV exposure suggests that they are keratinocytes rather than melanocytes. The mechanism(s) underlying the formation of these cells is of course of great interest. It is known that a 351 nM pulse laser causes highly selective injury to melanocytes containing melanosomes (37), suggesting that the UV energy absorbed by melanin may cause selective damage to pigmented structures. In other words, low doses of UV may cause melanin-specific photothermolysis, which could be considered another

The FASEB Journal

YAMAGUCHI ET AL.

Table 2.

TUNEL-Positive cells in skin and skin reconstructs after UV exposure # TUNEL-

Skin type In situ Fair (n⫽7) S2, S9, S18, S23, S31, S77, & S80 Intermediate (n⫽3) S5, S63, & S73 Dark (n⫽3) S65, S78, & S79 Melanoderm Fair (n⫽3) Intermediate (n⫽3) Dark (n⫽3)

UV dose

positive cells

Pa


180–200J/m2 ⬘ ⬘ 180–200J/m2 ⬘ ⬘ 180–200J/m2 ⬘ ⬘

1.12 ⫾ 0.29 0.63 ⫾ 0.25 1.24 ⫾ 0.38 1.28 ⫾ 0.47 3.80 ⫾ 1.49 2.62 ⫾ 1.28 1.03 ⫾ 0.52 5.40 ⫾ 2.18 7.93 ⫾ 3.45

– – – NS * * NS * *

2 days ⬘ ⬘ 2 days ⬘ ⬘ 2 days ⬘ ⬘

0J/m2 25J/m2 50J/m2 0J/m2 25J/m2 50J/m2 0J/m2 25J/m2 50J/m2

0.90 ⫾ 1.10 4.20 ⫾ 1.75 8.30 ⫾ 2.11 1.20 ⫾ 1.48 11.10 ⫾ 3.70 23.90 ⫾ 5.13 1.00 ⫾ 1.15 14.10 ⫾ 5.02 28.10 ⫾ 6.17

– – – NS ** ** NS ** **

Time 7 1 7 7 1 7 7 1 7

S numbers under in situ skin refer to subjects as reported in Table 1 of (17). Background staining in unirradiated controls has been subtracted from the data reported above. Values indicate means of TUNEL-positive cells per field ⫾ sd a P value versus fair skin at the same time point for the in situ results, and at the same UV dose for the melanoderm results. NS, not significant; *P⬍0.05; **P⬍0.01.

Figure 6. Representative images of immunohistochemical staining of p53 and of p53 phosphorylated at Ser-46 in Fair, Intermediate, and Dark human skin, before and 7 min, 1 d, and 7 d after a single 1 MED UV exposure. Green fluorescence represents p53 (top row in each series) and p53Ser46 (bottom row in each series). (———) demarks the top of the granular layer of the epidermis, (- - - -) demarks the epidermal:dermal junction, and (䡠䡠䡠䡠䡠) represents the division between the upper and lower epidermal layers.

form of photoprotection since it would enhance the removal of cells after UV exposure, many of them with significant DNA damage. It was recently shown that eumelanin or pheomelanin could elicit apoptosis in the skin of mice exposed to UV (38). In one study, a White and a Black or African-American (only 1 subject each) showed similar numbers of sunburn cells in response to 4 MED UV, but melanin pigment was more evident in the sunburn cells than in the surrounding cells (39). Additionally, in vitro studies have shown that macrophages containing melanin (termed melanophages) undergo cell death after UV exposure (40) and that melanocytes in White skin prevent sunburn cell formation (41). The presence of melanin in cells facilitates the apoptotic effect of UV, but whether that results from the generation of heat from the absorbed UV energy or whether other properties of melanins are involved will require further study. For example, apoptotic cells are normally removed by macrophages (42), so the increased numbers of apoptotic cells might also reflect the decreased function of melanophages (or keratinocytes which also have phagocytosis activity) laden with melanin that were preferentially killed by the UV. Differences in melanosome phenotype and distribution may also play a role in the different propensities to apoptosis among racial/ethnic groups. The fact that TUNEL-positive cells are similar at 7 d than at 1 d also suggests differences from the kinetics of sunburn cells resulting from UV-induced DNA damage, which typically maximize within 1–2 d (1). We hypothesize that the kinetics and mechanism of apoptosis induced by UV effects on melanin (photothermolysis?)


E637

Figure 7. Graphical representation of p53 and p53Ser46 staining in Fair, Intermediate, and Dark skin at 1 and 7 d after UV irradiation (left panels) and sorted by Melanin Content (right panels). Bars represent means ⫾ sem

differ significantly from apoptosis induced by DNA damage via pathways involving caspase 3 activation. In summary, we propose two major mechanisms that underlie the dramatic differences in photocarcinogenesis of light and dark skin. First, UV-induced DNA damage in the lower epidermis (including keratinocyte stem cells and melanocytes) is not effectively prevented because of low melanin content in the upper (and lower) epidermis of White skin. DNA damage in the upper epidermis is quite similar among all types of skin, which indicates that the epidermal pigmentation is an efficient UV filter. Prolonged DNA repair required by the initial severe concentration of UV damage in fair/ light skin may result in heritable mutations in critical genes involved in photocarcinogenesis. Second, UVinduced apoptosis was absent in White skin after low UV doses but was significant in Black or AfricanAmerican skin, facilitating the effective removal of UV-damaged cells in dark skin. Virtually all epidermal cells had significant DNA damage in fair skin but only ⬃1% of them became apoptotic. In contrast, less than 50% of epidermal cells in dark skin had significant DNA damage, and very few of those cells were in the basal cell layer, yet ⬃5% of those cells were apoptotic. The combination of relatively low DNA damage and efficient removal of UV-damaged cells no doubt contributes to the decreased incidence of skin cancer in darker skin. E638

Vol. 20

July 2006

The authors thank Dr. Seiji Takeuchi for helpful discussions. This research was supported in part by the Intramural Research Program of the NIH, National Cancer Institute, Center for Cancer Research. The mention of commercial products, their sources, or their use in connection with material reported herein is not to be construed as either an actual or implied endorsement of such products by the Department of Health and Human Services.

REFERENCES 1.

2. 3.

4.

5.

Brash, D. E., Ziegler, A., Jonason, A. S., Simon, J. A., Kunala, S., Leffell, D. J. (1996) Sunlight and sunburn in human skin cancer: p53, apoptosis, and tumor promotion. J. Invest. Dermatol. Symp. Proc. 1, 136 –142 Wikonkal, N. M., Brash, D. E. (1999) Ultraviolet radiation signature mutations in photocarcinogenesis. J. Investig. Dermatol. Symp. Proc. 4, 6 –10 McKay, B. C., Stubbert, L. J., Fowler, C. C., Smith, J. M., Cardamore, R. A., Spronck, J. C. (2004) Regulation of ultraviolet light-induced gene expression by gene size. Proc. Natl. Acad. Sci. U. S. A. 101, 6582– 6586 Agar, N. S., Halliday, G. M., Barnetson, R. S., Ananthaswamy, H. N., Wheeler, M., Jones, A. M. (2004) The basal layer in human squamous tumors harbors more UVA than UVB fingerprint mutations: a role for UVA in human skin carcinogenesis. Proc. Natl. Acad. Sci. U. S. A. 101, 4954 – 4959 Brenner, D. J., Doll, R., Goodhead, D. T., Hall, E. J., Land, C. E., Little, J. B., Lubin, J. H., Preston, D. L., Preston, R. J., Puskin, J. S., et al. (2003) Cancer risks attributable to low doses of ionizing radiation: assessing what we really know. Proc. Natl. Acad. Sci. U. S. A. 100, 13761–13766

The FASEB Journal

YAMAGUCHI ET AL.

6. 7. 8. 9.

10. 11. 12. 13. 14. 15. 16. 17.

18.

19. 20.

21. 22.

23. 24. 25. 26.

Setlow, R. B. (1999) Spectral regions contributing to melanoma: a personal view. J. Invest. Dermatol. Symp. Proc. 4, 46 – 49 Moan, J., Dahlback, A., Setlow, R. B. (1999) Epidemiological support for an hypothesis for melanoma induction indicating a role for UVA radiation. Photochem. Photobiol. 70, 243–247 Zeise, L., Chedekel, M. R., Fitzpatrick, T. B. (1995) Melanin: Its Role in Human Photoprotection. Valdenmar Publishing Company, Overland Park Wagner, J. K., Parra, E. J., Lin, H., Jovel, C., Shriver, M. D. (2002) Skin responses to ultraviolet radiation: effects of constitutive pigmentation, sex, and ancestry. Pigment Cell Res. 15, 385–390 Ortonne, J. P. (2002) Photoprotective properties of skin melanin. Brit. J. Dermatol. 146, 7–10 Preston, D. S., Stern, R. S. (1992) Nonmelanoma cancers of the skin. New Eng. J. Med. 327, 1649 –1662 Halder, R. M., Bridgeman-Shah, S. (1995) Skin cancer in African Americans. Cancer 75, 667– 673 English, D. R., Armstrong, B. K., Kricker, A., Fleming, C. (1997) Sunlight and cancer. Cancer Causes Control 8, 271–283 Gilchrest, B. A., Eller, M. S., Geller, A. C., Yaar, M. (1999) The pathogenesis of melanoma induced by ultraviolet radiation. New Eng. J. Med. 340, 1341–1348 Szabo, G., Gerald, A. B., Pathak, M. A., Fitzpatrick, T. B. (1969) Racial differences in the fate of melanosomes in human epidermis. Nature 222, 1081 Kaidbey, K. H., Agin, P. P., Sayre, R. M., Kligman, A. M. (1979) Photoprotection by melanin - a comparison of black and Caucasian skin. J. Amer. Acad. Dermatol. 1, 249 –260 Tadokoro, T., Kobayashi, N., Zmudzka, B. Z., Ito, S., Wakamatsu, K., Yamaguchi, Y., Korossy, K. S., Miller, S. A., Beer, J. Z., Hearing, V. J. (2003) UV-induced DNA damage and melanin content in human skin differing in racial/ethnic origin and photosensitivity. FASEB J. 17, 1177–1179 Kobayashi, N., Nakagawa, A., Muramatsu, T., Yamashina, Y., Shirai, T., Hashimoto, M. W., Ishigaki, Y., Ohnishi, T., Mori, T. (1998) Supranuclear melanin caps reduce ultraviolet induced DNA photoproducts in human epidermis. J. Invest. Dermatol. 110, 806 – 810 Murphy, G., Young, A. R., Wulf, H. C., Kulms, D., Schwarz, T. (2001) The molecular determinants of sunburn cell formation. Exp. Dermatol. 10, 155–160 Sheehan, J. M., Cragg, N., Chadwick, C. A., Potten, C. S., Young, A. R. (2002) Repeated ultraviolet exposure affords the same protection against DNA photodamage and erythema in human skin types II and IV but is associated with faster DNA repair in skin type IV. J. Invest. Dermatol. 118, 825– 829 Hall, P. A., McKee, P. H., Menage, H. D., Dover, R., Lane, D. P. (1993) High levels of p53 protein in UV-irradiated normal human skin. Oncogene 8, 203–207 Ziegler, A., Jonason, A. S., Leffell, D. J., Simon, J. A., Sharma, H. W., Kimmelman, J., Remington, J., Jacks, T., Brash, D. E. (1994) Sunburn and p53 in the onset of skin cancer. Nature 372, 773–776 Fridman, J. S., Lowe, S. W. (2003) Control of apoptosis by p53. Oncogene 22, 9030 –9040 Weber, J. D., Zambetti, G. P. (2003) Renewing the debate over the p53 apoptotic response. Cell Death Differ. 10, 409 – 412 Oren, M. (2003) Decision making by p53: life, death and cancer. Cell Death Differ. 10, 431– 442 Oda, K., Arakawa, H., Tanaka, T., Nakanishi, H., Ng, C. C., Taya, Y., Monden, M., Nakamura, Y. (2000) p53AIP1, a potential mediator of p53-dependent apoptosis, and its regulation by Ser-46-phosphorylated p53. Cell 102, 849 – 862

27.

28. 29.

30.

31.

32.

33.

34. 35. 36.

37. 38.

39. 40. 41.

42.


Okamura, S., Arakawa, H., Tanaka, T., Nakanishi, H., Ng, C. C., Taya, Y., Monden, M., Nakamura, Y. (2001) p53DINP1, a p53-inducible gene, regulates p53-dependent apoptosis. Mol. Cell 8, 85–94 Tanikawa, C., Matsuda, K., Fukuda, S., Nakamura, Y., Arakawa, H. (2003) p53RDL1 regulates p53-dependent apoptosis. Nature Cell Biol. 5, 216 –223 Tadokoro, T., Yamaguchi, Y., Batzer, J., Coelho, S. G., Zmudzka, B. Z., Miller, S. A., Wolber, R., Beer, J. Z., Hearing, V. J. (2005) Mechanisms of skin tanning in different racial/ethnic groups in response to ultraviolet radiation. J. Invest. Dermatol. 124, 1326 – 1332 Yamaguchi, Y., Itami, S., Watabe, H., Yasumoto, K., Abdel-Malek, Z. A., Kubo, T., Rouzaud, F., Tanemura, A., Yoshikawa, K., Hearing, V. J. (2004) Mesenchymal-epithelial interactions in the skin: Increased expression of dickkopf1 by palmoplantar fibroblasts inhibits melanocyte growth and differentiation. J. Cell Biol. 165, 275–285 Yoon, T. J., Lei, T. C., Yamaguchi, Y., Batzer, J., Wolber, R., Hearing, V. J. (2003) Reconstituted 3-dimensional human skin as a novel in vitro model for studies of pigmentation. Anal. Biochem. 318, 260 –269 Virador, V., Muller, J., Wu, X., Abdel-Malek, Z. A., Yu, Z.-X., Ferrans, V. J., Kobayashi, N., Wakamatsu, K., Ito, S., Hammer, J. A., Hearing, V. J. (2002) Influence of ␣-melanocyte stimulating hormone and ultraviolet radiation on the transfer of melanosomes to keratinocytes. FASEB J. 16, 105–107 Young, A. R., Chadwick, C. A., Harrison, G. I., Hawk, J. L., Nikaido, O., Potten, C. S. (1996) The in situ repair kinetics of epidermal thymine dimers and 6 – 4 photoproducts in human skin types I and II. J. Invest. Dermatol. 106, 1307–1313 Bykov, V. J., Marcusson, J. A., Hemminki, K. (2000) Effect of constitutional pigmentation on ultraviolet B-induced DNA damage in fair-skinned people. J. Invest. Dermatol. 114, 40 – 43 Hemminki, K., Xu, G., Kause, L., Koulu, L. M., Zhao, C., Jansen, T. (2002) Demonstration of UV-dimers in human skin DNA in situ 3 weeks after exposure. Carcinogenesis 23, 605– 609 Kadekaro, A. L., Kavanagh, R., Wakamatsu, K., Ito, S., Pipitone, M. A., Abdel-Malek, Z. A. (2003) Cutaneous photobiology. The melanocyte versus the sun: who will win the final round? Pigment Cell Res. 16, 434 – 447 Anderson, R. R., Parrish, J. A. (1983) Selective photothermolysis: precise microsurgery by selective absorption of pulsed radiation. Science 220, 524 –527 Takeuchi, S., Zhang, W., Wakamatsu, K., Ito, S., Hearing, V. J., Kraemer, K. H., Brash, D. E. (2004) Melanin acts as a potent UVB photosensitizer to cause a novel mode of cell death in murine skin. Proc. Natl. Acad. Sci. U. S. A. 101, 15076 –15081 Olson, R. L., Gaylor, J., Everett, M. A. (1974) Ultravioletinduced individual cell keratinization. J. Cutan. Pathol. 1, 120 – 125 Johnson, B. E., Mandell, G., Daniels, G. (1972) Melanin and cellular reactions to ultraviolet radiation. Nature 235, 147–149 Cario-Andre, M., Pain, C., Gall, Y., Ginestar, J., Nikaido, O., Taı¨eb, A. (2000) Studies on epidermis reconstructed with and without melanocytes: melanocytes prevent sunburn cell formation but not appearance of DNA damaged cells in fair-skinned Caucasians. J. Invest. Dermatol. 115, 193–199 Iyoda, T., Nagata, K., Akashi, M., Kobayashi, Y. (2005) Neutrophils accelerate macrophage-mediated digestion of apoptotic cells in vivo as well as in vitro. J. Immunol. 175, 3475–3483 Received for publication January 27, 2006. Accepted for publication March 14, 2006.

E639

The FASEB Journal • FJ Express Summary

Human skin responses to UV radiation: Pigment in the upper epidermis protects against DNA damage in the lower epidermis and facilitates apoptosis Yuji Yamaguchi,* Kaoruko Takahashi,* Barbara Z. Zmudzka,† Andrija Kornhauser,‡ Sharon A. Miller,† Taketsugu Tadokoro,* Werner Berens,* Janusz Z. Beer,† and Vincent J. Hearing*,1 *Laboratory of Cell Biology, National Cancer Institute, National Institutes of Health, Bethesda, Maryland, USA; †Center for Devices and Radiological Health, Food and Drug Administration, Rockville, Maryland, USA; and ‡Center for Food Safety and Applied Nutrition, U. S. Food and Drug Administration, College Park, Maryland, USA To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.06-5725fje SPECIFIC AIMS To better understand the role of melanin in protecting the skin against UV radiation and subsequent photocarcinogenesis. We previously reported that levels of melanin correlated inversely with amounts of DNA damage induced by UV in normal human skin of different racial/ethnic groups. In this study, we further investigated DNA damage in the upper and lower epidermal layers in various types of skin before and after exposure to UV, and measured subsequent apoptosis and phosphorylation of p53.

PRINCIPAL FINDINGS 1. DNA damage in different types of skin UV radiation of human skin results in two major types of DNA lesions: 6,4-photoproducts (6,4PP) and cyclobutane pyrimidine dimers (CPD). In this study, we measured levels of 6,4PP and CPD and determined their distribution in skin of varying racial/ethnic origin following a single exposure to a 1 minimal erythemal dose (MED) of UV. The 1 MED dose of UV was on average ⬃3 times higher in dark skin than in fair skin, and was ⬃30% higher in intermediate skin than in fair skin. Despite that, overall levels of 6,4PP were significantly lower in dark skin immediately after UV exposure compared with fair skin (not shown). Immediately after UV, DNA damage was significantly lower in the lower epidermis than in the upper epidermis in intermediate and in dark skin, but not in fair skin, which suggests that the upper epidermis has a photoprotective effect against UV-induced 6,4PP and that this effect is more remarkable in darker skin. Figure 1A shows representative images of DNA damage as CPD for subjects with fair, intermediate and dark 1486

skin. CPD damage density in the lower epidermis of fair skin did not differ significantly from that in the upper epidermis immediately after UV radiation, or 1 d or 7 d later. However, in intermediate skin and more dramatically in dark skin, CPD damage in the lower epidermis was significantly less than in the upper epidermis at 7 min, 1 d and 7 d after UV exposure. Consistent with the 6,4PP damage, levels of CPD damage in the lower epidermis were higher in fair skin after UV than in dark skin at all time points examined, despite the fact that the darker skin had received up to 3-fold more UV to reach 1 MED. These results show that UV penetrates deeper in less pigmented skin. 6,4PP and CPD damage in total, upper and lower epidermis were inversely correlated with melanin content (data not shown), which indicates that darker skin incurs less DNA damage in the lower epidermis while DNA damage in the upper epidermis is similar among racial/ethnic groups after exposure to UV. Since DNA damage immediately after 1 MED UV was significant, even deep in the epidermis, we investigated CPD damage in melanocytes (Fig. 1B, identified by costaining for tyrosinase). The % melanocytes with detectable levels of CPD was significantly higher in fair skin (86%) than in intermediate (30%) and in dark skin (11%) and the constitutive melanin content of the skin correlated inversely with the % melanocytes with DNA damage immediately after UV (r⫽.588, P⬍0.001). 2. Melanin facilitates the induction of apoptosis by UV We used the TUNEL assay to measure apoptotic cells in the skin of different racial/ethnic groups after expo1

Correspondence: National Institutes of Health, Laboratory of Cell Biology, Bldg. 37, Rm. 2132, Bethesda, MD 20892-4254, USA. E-mail: [email protected] doi: 10.1096/fj.06-5725fje

0892-6638/06/0020-1486 U.S. government work not protected by U.S. copyright

posites and they represent typical skin morphology (including pigmentation) of those types of skin. The MelanoDerm cultures were unirradiated or were UVBirradiated at 25 or 50 J/m2 and were then fixed and embedded 2 d later. TUNEL assays showed that significantly more apoptotic cells were found in Black skin equivalents than in Asian or White skin equivalents at both UV doses (Fig. 2B). 3. Phosphorylation of p53 and caspase-3 activation after UV exposure The oncogene p53 plays important roles in responses to UV-induced DNA damage and induction of DNA repair. There is an overall nuclear accumulation of p53 in response to UV. More than 13 sites of p53 are known to be phosphorylated, one of them being a critical site at Ser-46, which is associated with the induction of apoptosis. Thus, we investigated the effects of UV on the accumulation of p53 in nuclei and its phosphorylation at Ser-46 in skin. More p53 accumulated in the nuclei of cells in fair skin than in dark skin at 1 d and at 7 d after UV exposure. However, phosphorylation of p53 at Ser-46 was not seen in fair skin, whereas it was

Figure 1. A) Representative images of CPD DNA damage in fair, intermediate and dark skin immediately and 7 d after UV exposure; green and red fluorescence represent CPD and DNA, respectively. (———) demarks the top of the granular layer of the epidermis, (- - - -) demarks the epidermal:dermal junction, and (䡠䡠䡠䡠䡠) represents the division between the upper and lower epidermal layers. B) Representative images of CPD (green) in melanocytes (stained red for tyrosinase) immediately after UV in fair, intermediate and dark skin.

sure to UV. Surprisingly, 7-fold more apoptotic cells were observed in dark skin than in fair skin after 1 MED UV (not shown), although the DNA damage in dark skin was significantly less than in fair and intermediate skin, as noted above. To rule out that the higher physical UV dose at 1 MED used for dark skin increased TUNEL-positive cells, we performed a similar study but using an approximately identical dose (180 –200 J/m2 UV). TUNEL-positive apoptotic cells were observed at more than 5-fold higher levels in dark skin 1 d after that same dose of UV than in fair skin (Fig. 2A). Increased levels of melanin correlated directly with the number of TUNEL-positive cells at 1 d and at 7 d after UV exposure, demonstrating that skin containing more melanin undergoes significantly more apoptosis in response to a single low UV dose. We hypothesized that it might be the melanin within keratinocytes that is involved in the increased apoptosis after UV exposure. To examine this, we used reconstructed 3-dimensional human skin equivalents (termed MelanoDerm) containing melanocytes derived from Black, Asian or White donors and keratinocytes from one Hispanic donor. The differences in melanin distribution are significant among these com-

Figure 2. A) TUNEL staining in fair, intermediate and dark skin 1 d after exposure to 180 –200 J/m2 UV; green and red fluroescence represent TUNEL and DNA, respectively. B) TUNEL staining in reconstituted 3-dimensional human skin equivalents containing melanocytes derived from Black, Asian, or White donors and Keratinocytes from the same hispanic donor, 2 d after UVB irradiation; green and red fluroescence represent TUNEL and DNA, respectively. C) p53 phosphorylated at Ser-46 in fair, intermediate and dark skin 1 d after 180 –200 J/m2 UV; green and red fluorescence represent TUNEL and p53 phosphorylated at Ser-46, respectively.


1487

Figure 3. Schematic summarizing the results of this study on fair skin (left) and dark skin (right). UV penetration, DNA damage, melanin distribution and an apoptotic cell in various layers of the skin are shown.

readily seen in dark skin 1 d after UV (Fig. 2C). The nuclear expression of p53 phosphorylated at Ser-46 colocalized with TUNEL-positive cells.

CONCLUSIONS AND SIGNIFICANCE DNA damage (6,4PP and CPD) in skin from different racial/ethnic groups after exposure to 1 MED UV was significantly greater in fair skin than in dark skin and constitutive melanin content correlated inversely with that DNA damage (summarized in Fig. 3). DNA damage in the upper epidermis immediately after UV exposure was similar among racial/ethnic groups but levels of DNA damage in the lower concentration of the epidermis was inversely proportional to the melanin content. Taken together, we conclude that the upper epidermis of dark skin is significantly more photoprotective for the deeper tissue against UV damage than that of fair skin. The localization of TUNEL-positive cells in the middle to upper layers of the epidermis following UV exposure suggests that they are keratinocytes rather than melanocytes, and the mechanism(s) underlying

1488

Vol. 20

July 2006

their formation is of great interest. Pulse lasers cause highly selective injury to cells containing melanosomes suggesting that the UV energy absorbed by melanin in the upper epidermis causes photothermolysis (heat damage) to pigmented cells. We investigated the appearance of TUNEL-positive cells in skin from various racial/ethnic groups after UV and in skin equivalents after UV exposure to test the hypothesis that the melanin content is responsible for the apoptosis. Cells containing melanin in the upper epidermis of dark skin tended to undergo more apoptosis after UV than do those of fair skin. Thus, the presence of melanin facilitates the apoptotic effect of UV on cells but whether that results from photothermolysis or whether other properties of melanins are involved will require further study. Our study also demonstrates that the nuclear accumulation of p53 is less in dark skin than in fair skin, suggesting that the overall activation of p53 following UV-induced DNA damage is greater in fair skin. The sustained activation of p53 may also in part cause the higher incidence of photocarcinogenesis in fair skin. Phosphorylation of p53 at Ser-46, which is associated with the induction of apoptosis, occurred at low levels in fair skin after low doses of UV exposure but was significant in dark skin, suggesting that p53 phosphorylation site is involved in UV-induced apoptosis in epidermis with abundant levels of melanin. In summary, we propose two major mechanisms which underlie the dramatic differences in photocarcinogenesis of light vs. dark skin. First, UV-induced DNA damage in the lower epidermis (which contains keratinocyte stem cells and melanocytes) is not effectively prevented in fair skin because of the low melanin content in the upper (and lower) epidermis. DNA damage in the upper epidermis is quite similar among all types of skin, which indicates that epidermal pigmentation is an efficient UV filter for underlying cells. Second, UV-induced apoptosis was virtually absent in fair skin after low UV doses, but was significant in dark skin, facilitating the effective removal of UV-damaged cells in dark skin. Virtually all epidermal cells had significant DNA damage in fair skin but only ⬃1% of them became apoptotic whereas less than 50% of epidermal cells in dark skin had significant DNA damage, yet ⬃5% of those cells were apoptotic. The combination of relatively low DNA damage and efficient removal of UV-damaged cells contributes to the decreased incidence of skin cancer in darker skin.

The FASEB Journal

YAMAGUCHI ET AL.