[CANCER RESEARCH 44, 5195-5199,
Photoprotective Role of Epidermal Melanin Granules against Ultraviolet Damage and DMA Repair in Guinea Pig Skin1
Takatoshi Ishikawa,2 Ken-ichi Kodama, Jiro Matsumoto, and Shozo Takayama
Department of Experimental Pathology, Cancer Institute (Japanese Foundation tor Cancer Research) Kami-lkebukuro, Toshima-ku, Tokyo 170 [T. I., K. K., S. T.], and Department of Biology, Keio University, Hiyoshi, Kohoku-ku, Yokohama 223 [J. M.], Japan
ABSTRACT We previously developed a quantitative autoradiographic tech nique with special forceps for measuring unscheduled DNA synthesis (UDS) in mouse skin after treatment with ultraviolet light in vivo. By this method, we investigated the relationship between the protective role of melanin and UV-induced DNA repair in black-and-white guinea pigs. Flat areas containing a sharp border between pigmented and unpigmented skin were selected. The skin of the selected areas was shaved and irradi ated with short-wave UV (254 nm) or UV-AB (270 to 440 nm, emission peak at 312 nm) at various doses. Immediately after irradiation, the skin was clamped off with forceps, and an isotonic aqueous solution of [mef/7y/-3H]thymidine was injected s.c. into the clamped off portion. UDS was clearly demonstrated as silver grains in this portion of the skin after irradiation with 254 nm UV or UV-AB. Errors due to individual differences were avoided by comparing the intensities of UDS in basal cells from pigmented skin and unpigmented skin of the same animals. Unexpectedly, in groups of animals treated with 254 nm UV or UV-AB, no difference in UDS in pigmented and unpigmented skin was seen at any UV dose. These results suggested that epidermal melanin granules do not significantly protect DNA of basal cells against 254 nm UV or UV-AB irradiation. Results of a study on the effect of the wavelength of irradiation on the UDS response of albino guinea pigs are also reported. INTRODUCTION Clinical observations and epidemiolÃ³gica! studies indicate that sunlight is the principal cause of carcinomas of human skin (4, 14). Human skin has various defensive mechanisms against actinic damage (12, 15, 17). In particular, melanin is considered to have a screening effect against harmful UV radiation. Recently, the role of DNA repair in cutaneous carcinogenesis has received much attention (4). Therefore, it seemed important to investigate the relation between the protective role of melanin and UVinduced DNA repair in the skin. However, few methods are available for detecting DNA repair in vivo. Direct measurement of pyrimidine dimers in irradiated skin has been used in studies of UV-induced DNA repair in the skin (2, 3, 16). This method is specific, but it does not give information on the location of DNA repair within the skin. Another method is autoradiographic mea surement of DNA repair in the skin in vivo. It has long been thought to be difficult to demonstrate UDS3 in vivo. In fact,
'This work was supported by Grants-in-Aid for Cancer Research from the
Epstein ef al. demonstrated UV-induced UDS in mouse (5) and human (6) skin by injecting [metfjy/-3H]dThd s.c., but they could not demonstrate dose-dependent responses. We developed an autoradiographic technique with special forceps for measuring UDS in mouse skin after treatment with various chemical carcin ogens (8) or UV irradiation (11) in vivo. This method allowed the first quantitative autoradiographic measurements of DNA repair in individual cells of the skin of intact animals. We used this method in studies on the possible protective role of melanin in guinea pig skin against UV irradiation.
Induction of UDS in the Skin of Albino Guinea Pigs. The technique used was essentially the same as that used for mice (8,11) and thus is only briefly described here (Chart 1). Seventy-five noninbred Hartley albino guinea pigs (2 to 3 months old, female), weighing 200 to 250 g, were obtained from Saitama Experimental Animal Farm (Sugito-shi, Japan). Animals were anesthetized by Â¡.p.njection of sodium pentobari bital (15 mg/kg; Abbott Laboratories, Chicago, IL). Their flank skin was shaved with electric clippers, and the remaining hair was carefully re moved with a razor. Flat areas on the flank (about 3x3 cm) were selected. The skin was exposed to short-wave UV from a 15-watt germicidal lamp (Toshiba GL 15 UV lamp) with predominant emission at 254 nm (dose rate, 1.2 J/sq m/sec) at a distance of 50 cm, or to UV-AB from three 20-watt sunlamp fluorescent tubes (Toshiba FL 20 S. E. sunlamp) with emission at 270 to 440 nm and a peak at 312 nm (dose rate, 2.2 J/sq m/sec) at a distance of 20 cm. The dose rate from the short-wave UV lamp was measured with a Black-ray Model J-225 UV meter (Ultraviolet Products, Inc., San Gabriel, CA); the dose rate of UVAB irradiation was measured with a UVR-365 UV radiometer (Tokyo Kogaku Kikai Co., Tokyo, Japan). In this paper, we refer to short-wave UV and UV-AB irradiation as 254 nm UV and sunlamp UV irradiation, respectively. Five animals each were exposed to 254 nm UV irradiation at doses of 72, 144, 216, 360, 500, 720, and 1400 J/sq m. Similarly, 5 animals each were exposed to sunlamp UV irradiation at doses of 330, 660,1300, 2600, 3900, and 8000 J/sq m. Then, the irradiated region of the skin was clamped off with tongue forceps (ring shaped, 20 mm internal diameter), avoiding stretching the skin as much as possible. Immediately after this region was clamped off, isotonic Ringer solution (0.5 ml) containing [rneffcy/-3H]dThd (New England Nuclear, Boston, MA; specific activity, 75 Ci/mmol; 100 fiCi/ml) was injected s.c. into the clamped-off region through a fine needle (0.125 gauge). After this treat ment, the animals were kept at 35Â°for 60 min in an incubator, and then the forceps were removed. The animals were killed 3 hr after removal of the forceps, because during this 3-hr period excess [mef/)y/-3H]dThd was washed out in the blood stream. The skin was excised and fixed in 10% neutral formaldehyde solution. Fixed skin was cut transversely into thin strips (5 x 20 mm), embedded in paraffin, and cut into 4- to 5-iÂ¿mthick sections. Sections were treated with 5% trichloroacetic acid for 45 min at 4Â° with 3 changes of the solution to remove the acid-soluble fraction. Sections were dip-covered with NR-M2 emulsion (Konishiroku Photo Co., Tokyo, Japan) and exposed for 4 weeks at 4Â°.After devel opment, the sections were stained lightly with hematoxylin and eosin.
Ministry of Education, Science, and Culture and the Ministry of Health and Welfare of Japan. 2 To whom requests for reprints should be addressed. 'The abbreviations used are: UDS, unscheduled DNA synthesis; UVAB, UV-A (400 to 315 nm) plus UV-B (315 to 280 nm); dThd, thymidine. Received March 5,1984; accepted August 7,1984.
T. Ishikawa et al.
1. UV Irradiation 254 nm UV or Sunlamp UV 15
2. Clamping Off with Tongue Forceps and [methyl-HldThd Injection
3. Treatment 4. Preparation
at 35Â°for 60 Minutes of Histological Sections
5. Autoradiography 6. Grain Counts on 200 Cells Chart 1. Detection of UDS in guinea pig skin in vivo after UV irradiation. Since the range of variability in the grain numbers was small throughout a sample, grains were counted consecutively (in one direction) on 200 basal cells in the basal layer starting from a randomly selected point. Ten albino guinea pigs were used to determine the minimal erythemal dose. The skin of both flanks was shaved, and various regions were exposed to 8 different doses of sunlamp UV irradiation. Changes were judged by eye 24 hr after exposure. Studies on Photoprotection by Melanin in Pigmented Guinea Pigs. For study of the possible photoprotective effect of melanin, 34 blackand-white guinea pigs were used in place of albino guinea pigs. Noninbred pigmented guinea pigs of both sexes, weighing 250 to 400 g, were obtained from the same source, and only animals that had sharply demarcated black-and-white regions in the middle of the back or flank (Fig. 1) were used. The hair was shaved. Flat areas of approximately 3 x 3 cm containing a sharp border between pigmented and unpigmented skin were selected and marked with a pen (Fig. 1). Four animals each were exposed to 3 doses of 254 nm UV irradiation. In addition, 3 to 4 animals each were exposed to 5 doses of sunlamp UV irradiation. The dose ranges used were based on the dose-response curves obtained in experiments on albino guinea pigs. Immediately after irradiation, areas of skin containing equal amounts of pigmented and unpigmented skin were clamped off with forceps. [Mefr7y/-3H]dThd injection and autoradiographic procedures were as described above. However, before being coated with emulsion, sections were treated for 3 hr at room temperature with 0.25% potassium permanganate to bleach the melanin. Unbleached parallel sections were stained with hematoxylin and eosin and used to identify pigmented areas of the skin. Grains were counted on 200 randomly selected basal cells in the basal layer of pigmented skin and unpigmented skin of each animal.
Chart 2. Dose response to UV of UDS Â¡n pithelial cells of albino guinea pig skin e exposed to 254 nm UV or sunlight UV irradiation. Numbers of grains per nucleus are plotted against the energy dose. Points, averages for number of grains per nucleus Â¡n animals; bars, S.D. A, 254 nm UV irradiation; 8, sunlamp UV irradi 5 ation.
sion). The mean grain counts and standard deviations of the means are given in Chart 2. Only background numbers of grains (0.6 to 1.2 grains/cell) were seen on the nuclei of nonirradiated control animals (Fig. 4). Similar dose-dependent effects were seen with 254 nm UV and sunlamp UV irradiations. With the highest doses of UV irradiation, the numbers of silver grains reached a plateau (Chart 2). The highest average grain numbers in UV-treated groups were about 12 to 20 times that of the control. As shown in Chart 2, individual differences (S.D.) were small. However, unlike in mice (11), the energy levels required to induce UDS responses with 254 nm UV and sunlamp UV irradia tions did not differ markedly. The minimal erythemal dose in albino guinea pigs was 1300 J/sq m. Therefore, it is clear that UDS could be detected after sunlamp UV irradiation at energy levels below the minimal ery themal dose. Studies on Photoprotection by Melanin in Pigmented Guinea Pigs. The epidermis of guinea pigs consists of 3 to 4 layers of living cells with a thin layer of stratum corneum. There is no significant difference in the architecture, such as Â¡n the thickness of the skin or the density of hair follicles of pigmented and unpigmented skin, but melanogenic melanocytes are present in both the dermis and hair follicles of pigmented skin. The melanin granules are concentrated in the perinuclear area of basal cells and are sparsely distributed throughout the rest of the epidermis (Fig. 5). Results on the photoprotective effect of melanin are given in Chart 3. Each dot represents one animal. Photoprotection is expressed as the grain ratio (grain counts on 200 cells in pigCANCER RESEARCH VOL 44
RESULTS Induction of UDS in the Skin of Albino Guinea Pigs. In albino guinea pigs treated with 254 nm UV or sunlamp UV irradiation, UDS was demonstrated clearly as silver grains over the nuclei of epithelial cells. As shown in Figs. 2 and 3, UDS was uniformly distributed, and there was little variability in the numbers of grains on the nuclei of epithelial cells throughout the sections. Grains were counted on 200 randomly selected cells in the basal layer under a microscope with a x100 objective lens (oil immer-
Protective Role of Melanin against UV Damage induced DNA repair in various animal systems. We used this method in studies on UDS induced by UV irradiation in guinea pigs. We were interested to know whether epidermal melanin pro tects DNA of cutaneous cells against UV damage. Pigmented guinea pigs proved to be suitable animals for this study, because they, like humans, have a rich pigmentary melanocyte system in the epidermis (1, 12). Comparative histological studies have shown that the distributions of pigmentation in guinea pig and human skin are similar with large numbers of melanin granules in perinuclear areas of basal cells. Rabbits, mice, and rats were less suitable for this study, because they usually do not have epidermal melanocytes when they have a densely pigmented furry coat (7,12). We confirmed dose-dependent inductions of UDS in albino guinea pigs after 254 nm UV or sunlamp UV irradiation. Then, we investigated the photoprotective effect of melanin using black-and-white guinea pigs. To avoid errors due to individual differences, small variations in thickness of the skin and differ ences in the procedure of thymidine injection, we studied the photoprotective role of melanin in small areas of skin of single animals. In this way, pigmented and unpigmented areas of the skin were compared under as nearly similar conditions as pos sible. For autoradiographic study, it was essential to bleach melanin, since melanin granules seriously interfere with grain counting. Bleaching was done by treating the preparations with potassium permanganate, which we have found does not affect DNA (9). Unexpectedly, our results showed that epidermal melanin granules do not significantly protect DNA of basal cells against UV irradiation. This surprising observation raises the question of whether it is possible to extrapolate experimental results at relatively high doses to the lower levels to which humans are exposed; a high dose of UV irradiation may exceed the screening power of melanin. Judging from available information, the most carcinogenic rays are in the acute erythemogenic or sunburn spectrum (4). The doses used in the present experiment were compared with the minimal erythemal dose in guinea pig skin to obtain an estimate of their biological magnitude. UDS was clearly detected after exposure to sunlamp UV irradiation at a dose which was insufficient to induce erythema. Therefore, the doses of UV irradiation used in this work seemed to be within the biological range of actinic radiation. As stated above, much clinical and experimental evidence indicates the protective role'of melanin against UV. For example interesting comparative studies on epidermal specimens from Caucasian and Negro subjects (13) showed that the transmission spectrum differed greatly with the extent of pigmentation. It seems quite possible that the stratum corneum or epidermis as a whole in dark skins serves as a sunscreen against UV. More over, Kaidbey et al. (10) showed that the main site of UV filtration is the stratum corneum in Caucasians but the malpighian layers in Negroes. It is also suggested that the remarkable resistance of negroid epidermis to UV is due not only to its higher melanin content but also to other factors related to the packaging and distribution of melanosomes, i.e., the larger, singly distributed melanosomes in Negroes, as opposed to the smaller, aggre gated, and poorly melanized melanosomes in Caucasians (14). In epidermal cells of pigmented guinea pigs, the heavily melan ized melanosomes are ellipsoidal and are predominantly aggre-
Â«50 a O
800 1000 1200 J/m2
Chart 3. Photoprotection by melanin granules against 254 nrn UV and sunlamp UV irradiations. Relative grain ratios (pigmented skin/unpigmented skin x 100) of 34 guinea pigs are plotted. Values are for individual animals. A. 254 nm UV irradiation; B. sunlamp UV irradiation.
merited skin/grain counts on 200 cells in unpigmented skin x 100) of individual animals. The ratios were similar in groups of animals treated with 3 doses (360, 720, and 1080 J/sq m) of 254 nm UV irradiation (Chart 3); the average ratio for 8 guinea pigs exposed to unsaturated doses (360 and 720 J/sq m) of 254 nm UV was 97.4 Â±3.6 (S.D.). Similarly, the ratios of UDS were similar in groups of animals treated with 5 doses (330, 660, 1300, 2600, and 8000 J/sq m) of sunlamp UV irradiation (Chart 3); the average for 11 animals exposed to unsaturated doses (330,660, and 1300 J/sq m) of sunlamp UV irradiation was 99.9 Â±5.9. These results show that epidermal melanin granules in pigmented guinea pigs do not significantly protect DNA of basal cells against 254 nm UV or sunlamp UV irradiation. DISCUSSION In autoradiographic studies on DNA repair in vivo, Epstein ef al. found that DNA repair synthesis occurs in hairless mice (5) and human dermal and epidermal cells (6) in vivo after UV irradiation. However, they did not observe a dose-response relationship. To demonstrate this, we developed a procedure for clamping off regions of mouse skin (8, 11). By this method, we clearly demonstrated dose-dependent UDS in various types of cells in the skin after treatment with chemical carcinogens or UV irradiation. Since the variability in grain counts was small, even within a sample, this method should be useful for studies on UV-
T. Ishikawa et al. gated (packaged) (15).* Previously, we (10) found that hair serves as an effective sunscreen against UV irradiation; our results showed that hair screened about 90% of the UV energy from mouse skin. There fore, we conclude that melanin is not as important as hair for photoprotection against UV irradiation. Defensive mechanisms of the human skin against UV are of considerable interest, since there are few other hairless animals. Although findings in animals cannot be directly related to human responses, our studies on guinea pigs raise the question of whether the relatively few supranuclear melanin granules seen in Mongoloids and in the suntanned skin of fair-skinned subjects are actually important in photoprotection against UV. REFERENCES
1. Billingham, R. E., and Medawar. P. B Pigment spread and cell heredity in guinea-pig's skin. Heredity, 2: 29-47,1948. 2. Bowden, G. T., Trosko, J. E., Shapas, B. G., and Boutwell, R. K. Excision of pynmidme dimers from epidermal DNA and nonsemiconservative epidermal DNA synthesis following ultraviolet irradiation of mouse skin. Cancer Res., 35: 3599-3607, 1975. 3. Cooke, A., and Johnson, B. E. Dose response, wavelength dependence, and rate excision of ultraviolet radiation-induced pyrimidine dimers in mouse skin DNA. Biochim. Biophys. Acta, 577: 24-30,1978. 4. Epstein, J. H. Photocarcinogenesis: a review. Nati. Cancer Inst. Monogr. 50: 13-25,1978. 5. Epstein, J. H. Ultraviolet carcinogenesis. In: A. C. Giese (ed.), PhotophyskJtogy, ' T. Ishikawa, J. Matsumoto, and S. Takayama, unpublished data. Vol. 5, pp. 235-273. New York: Academic Press, Inc., 1978. 6. Epstein, J. H., Fukuyama, D., Reed, W. B., and Epstein, W. L. Defect in DNA synthesis in skin of patients with xeroderma pigmentosum demonstrated in vivo. Science (Wash. DC), 768: 1477-1478,1970. 7. Holmes, R. L. Patterns of cutaneous pigmentation: rodents. J. Anat., 87:163168,1953. 8. Ishikawa, T., Kodama, K., Ide, F., and Takayama, S. Demonstration of in vivo DNA repair synthesis in mouse skin exposed to various chemical carcinogens. Cancer Res., 42: 5216-5221,1982. 9. Ishikawa, T., Sakakibara, K., Kurumado, K., Shimada, H., and Yamaguchi, K. Morphologic and microspectrophotometric studies on spontaneous melano mas in Xiphophorus hellen. J. Nati. Cancer Inst., 54:1373-1378,1975. 10. Kaidbey, K. H., Poh Agin, P., Sayre, R. M., and Kligman, A. M. Photoprotection by melaninâ€”a comparison of black and Caucasian skin. J. Am. Acad. Dermatol., 7:249-260, 1979. 11. Kodama, K., Ishikawa, T., and Takayama, S. Dose response, wavelength dependence, and time course of ultraviolet radiation-induced unscheduled DNA synthesis in mouse skin in vivo. Cancer Res., 44: 2150-2154,1984. 12. Montagna, W. The epidermis. In: The Structure and Function of Skin, Chap. 2, pp. 73-87. New York: Academic Press, Inc., 1962. 13. Pathak, M. A., and Fitzpatrick, T. B. The role of natural photoprotective agents in human skin. In: T. B. Fitzpatrick, M. A. Pathak, L. C. Harber, M. Sieji, and A. Kukita (eds.), Sunlight and Man, pp. 725-750. Tokyo: University of Tokyo Press, 1974. 14. Quevedo, W. C., Jr., Fitzpatrick, T. B., Pathak, M. A., and Jombow, K. Light and skin color. In: T. B. Fitzpatrick, M. A. Pathak, L. C. Harber, M. Seiji, and A. Kukita (eds.), Sunlight and Man, pp. 165-194. Tokyo: University of Tokyo Press, 1974. 15. Seiji, M., Toda, K., Okazaki, K., Uzuka, M., Monkawa, F., and Sugiyama, M. Melanocyte-keratinocyte interaction in pigment transfer. Pigm. Cell, 3: 393405, 1976. 16. Sutherland, B. M., Harber, L. C., and Kochever, I. E. Pyrimidine dimer excision and repair in human skin. Cancer Res., 40:3181-3185,1980. 17. Urbach, F., Epstein, J. H., and Forbes, P. D. Ultraviolet carcinogenesis: Experimental, global, and genetic aspects. In: T. B. Fitzpatrick, M. A. Pathak, L. C. Harber, M. Seiji, and A. Kukita (eds.), Sunlight and Man, pp. 259-284. Tokyo: University of Tokyo Press, 1974.
Fig. 1. Black-and-white guinea pigs used for photoprotection experiments. Fig. 2. Autoradiograph showing silver grains on the nuclei of epithelial cells and dermal fibroblastic cells indicative of UDS after exposure to 254 nm uv irradiation (720 J/sq m). Guinea pig skin was treated with [mefhyi-'HJdThd (50 >iCi). Cells with heavily labeled nuclei were in S phase during treatment. Lightly stained with H & E, x 1200. Fig. 3. Autoradiograph showing silver grains on the nuclei of epithelial cells, hair follicle cells, and dermal fibroblastic cells, indicative of UDS after exposure to sunlamp UV irradiation (1300 J/sq m). Guinea pig skin was treated with [merhy/-3H]dThd (50 >iCi). Cells with heavily labeled nuclei were in S phase during treatment. Lightly stained with H &E,x 1200. Fig. 4. Control autoradiograph showing few background grains. Guinea pig skin was treated with [mefriy/-3H]dThd (50 *iCi). Cells with heavily labeled nuclei were in
S phase during treatment. Lightly stained with H & E, x 1200. Fig. 5. Histology of the skin of a pigmented guinea pig. Arrows, melanin granules accumulated in the perinudear area of basal cells. H & E, x 1200.