Has melanin a photoprotective role ?

Paolo U. Giacomoni

Laboratoires de l'Oréal

188, Rue Paul Hochart

94550, Chevilly-Larue, Franc

It is currently believed that melanogenesis is a natural, protective response to solar irradiation in the course of which melanin is synthesized in the melanocytes and transferred to the keratinocytes. If no one questions the fact that the process is a natural one, the question can be asked about the protecties against sunlight (if any) of melanin inside the keratinocytes.

What makes us think that melanin is protective?

Nordlund and co-workers (1) reports that the idea that melanin is protective has an history, the origin of which has been pinpointed down to Benjamin Franklin. As a matter of fact. Franklin observed that "dark coloured cloths laid upon fresh snow on a bright sunny day melted the underlying ice cristals more rapidly than light colored cloth" in agrement with the observation that, all the rest being equal, black objects exposed to sunlight reach higher temperatures than white ones. "It was known that black-skinned individuals normally inhabited tropical regions and that white individuals came from the north. It seemed perverse to Franklin that nature would endow individuals living in the tropics with a type of skin which was inappropriate for the environment. Everard Home resolved this enigma ... He exposed both his hands to the sun. One was covered with a black cloth, and the other was left bare. He measured the temperature of his skin. He found that the black cloth did indeed elevate the temperature of his skin a few degrees. However, the exposed hand became sunburned. He concluded that pigment did indeed cause a slight warming of the skin but protected it from the nonthermal, i.e., scorching, effects of the sunlight. Thus was born the idea that melanin was a sunscreen which prevented sunburn, a concept which persists to modern age and is only now undergoing reconsideration".

An obvious comment to Home's experiment is that it would have been better to cover both hands with cloths of identical manufacture, the only difference being the colour. An obvious comment to Home's conclusion is that the experiment proves protective properties for topically applied dark coloured molecules, and that nothing can be said about the role of endogenous pigmented granules.

Apparently this is the historical reason for believing that melanin is a natural against sunlight, possibly conforted by the observation that melanin production is a painless consequence of exposure to sunlight.

Upon learning more about the physiology of black skin, it was realized that black-skinned individuals exposed to sunlight experience sunburn as well. Someone also realized that prehistorical men used not to live in the sun-exposed savannah but in the sun-protected rainforest and scientists started wondering about the selective advantage of being black in such an environment. It was pointed out that in a forest environment, some sort of camouflage would be essential and black skin, reflecting only 16% of visible light, could be very effective for this purpose, much more effective than white skin which reflects 45% of visible light. Another selective advantage comes from the fact that radiant energy from sunlight can be absorbed by melanin and converted to heat. As a matter of fact, pigmentation can thus contribute to the maintainance of body temperature and to the conservation of metabolic heat, and be very important for prehumans, who slept without the benefit of fire or clothing and were not always successful in hunting (2).

Of course these discussions did not answer the question about photoprotective properties of melanin inside the keratinocytes and a number of experiments were designed in order to contribute to the advancement of knowledge.

Once it is made clear that the biological properties of melanin depend on its chemical structure (there is not one melanin, there are many different melanins, grouped in the categories of eumelanins and phaemelanins), one of the first questions which can be asked, when the role of melanin is questioned, concerns cell survival after UV irradiation.

Brian Johnson and coworkers observed that sunburn cells contained granules which appeared to be similar to those, known to be melanin, in basal layer cells. They also observed that in biopsies from normal human volounteers, the fraction of sunburn cells in lightly pigmented skin increased linearly with the dose up to nearly 90 per thousand malpighian cells for 8 Minimal Erythemal Doses (MED), while in vitiligo skin, in which no melanocytes are present, the fraction of sunburn cells remained nearly constant (~ 5 per thousand) with doses up to 16 MED. (It has to be noted that Brian Johnson used to work in Dundee, so that if the volounteers were autochtonous there are chances for their epidermis to contain phaeomelanin).

Taking advantage of the fact that macrophages can phagocytose melanin from the environment, Brian Johnson and coworkers exposed to UV from FS 20 fluorescent tubes, macrophages from mouse peritoneum which had been incubated with squid ink melanin.

They observed that macrophages incubated for 24 hours with melanin were slightly more sensitive to UV than macrophages which did not take up melanin (3).

Another question which can be asked concerns the formation of UV-induced DNA damage in cells containing or not containing melanin. Schothorst and coworkers (4) undertook to expose cultured human keratinocytes and melanocytes to monochromatic radiation in the UV range and measured the amount of Endonuclease Sensitive Sites (ESS) versus the dose at different wavelengths. Melanocytes were grown in a medium containing isobuthyl-methyl-xanthine, so it is reasonable to believe that they were pigmented, even though the authors did not present the reader with figures relative to the amount of melanin per cell. The outcome of this experiment is particularly interesting: no difference can be pointed out in the dose- and wavelength- dependence of ESS formation in keratinocytes or in melanocytes in the UV-C and UV-B regions, except for a small difference when 297 nm radiation is utilized, in this case melanocyte DNA is slightly more damaged than is keratinocyte DNA. An analogous experiment was performed by De Leeuw and coworkers, who measured the residual clone-forming ability of cultured human melanocytes and keratinocytes after monochromatic UV irradiation. They found that melanocytes are slightly less sensitive than keratinocytes to UVB and more resistant to UVA than keratinocytes (5).

Of course, cultured melanocytes are not melanocytes in the epidermis, moreover their melanin is distributed in melanosomes within the dendrites and only occasionally is interposed between the cell's nucleus and the source of UV light. Therefore it seemed necessary to measure the protection against radiation of cells having ingested different amounts of melanin, making sure that these cells could not be suspected of digesting melanin as it could have been the case in the experiment with macrophages.

Cell biology offers tools and methods for tackling this kind of problems. Ideally one should grow two samples of keratinocytes in the presence of homologous melanocytes, stimulate the first sample with UV light in order to include melanin synthesis and transfer of the pigment from the melanocytes to the keratinocytes, and treat the second sample according to a mock-irradiation protocol. After the transfer, which could be monitored by observation under the microscope, keratinocytes and melanocytes should be separated, the keratinocytes seeded, exposed to UV and checked for some physiological parameters (growth, DNA damage, loss of cytoplasmic enzymes, cell morphology and so forth).

If melanin is photoprotective one expects sample one to be in a better shape after UV xposure than sample two.

Of course such an experiments is extremely difficult to be carried out and some simplified protocols have been designed.

Hill and Hill induced B16 CL 4 mouse melanoma cells in culture to phagocytose melanin particles dispersed in the growth medium and subjected them to ionizing radiation.

The result of the non-irradiated control was that after the incubation in the presence of melanin, the alkaline elution of labelled DNA reveals conspicous nicking, the amount of which is dependent on the concentration of melanin to which cells were exposed. When cells preincubated with melanin are exposed to ionizing radiation, the results indicate that the nicking of DNA provoked by the two agents are additive (6).

In another experiment, Hill and coworkers undertook to measure the survival of three Cloudman S 91 mouse melanoma cell lines after exposure to 137 Cs radiation (7). The three cell lines contain different amount of melanin (respectively 1.2, 1.8 and 3.6 pg/cell) and, all the rest being equal, can be assumed to give responses to insults, which are dependent on the content of melanin. For low irradiation doses (below 5 Grays) there is a direct correlation between survival and melanin content. At 5 Grays, for instance, the surviving fractions for the three cell lines are 0.02, 0.09 and 0.3, respectively.

Of course this result gives informations about the physico-chemical properties of irradiated melanin, but the phagocytosis of melanin particles is not equatable to the melanosomal transfer from cell to cell and it is not sure that, within a melanoma cell, melanin forms a cap around the nucleus as it forms in keratinocytes.

Because of the difficulty to learn in cultured cells about the role of melanin in human epidermis, Young and co-workers designed a clever experiment with human volounteers. In order to have cells containing more or less melanin, all the rest being equal, they exposed the volounteers to a series of suberythemal UV irradiations from a solar simulator, either in the presence of a conventional sunscreen, in order to maintain an "amelanotic" status, or in the presence of the same sunscreen added with trace amounts of 5 methoxypsoralen, in order to obtain an artificially generated "highly pigmented" status, or without xenobiotics in order to obtain a naturally "melanin enriched" status. One week after the end of the series of the suberythemal irradiations, the volounteers where exposed to an erythemal dose of UV and checked for several parameters, such as melanin content and stratum corneum thickness (taken as two possible natural sunscreens) and the extent of Unscheduled DNA Synthesis (UDS or DNA repair), taken as an indicator of the extent of DNA damage, which is a major target of sunlight (8). The results seem to indicate that acquired pigmentation affords better protection against DNA damage, at least in phototypes III, IV and V. Yet the authors conclude that "Photoprotection is often explained by induction of melaninization and/or stratum corneum thickening. As such induction was independent of skin type and similar for the three types of treatment, there is no overall correlation between either or both these parameters with UDS levels, which indicates that photoprotection is more complex than previously thought".

A Symposium on "Melanin: Its Role in Human Photoprotection" was held in March 1994 in Washington, D.C. and the discussions pointed out that the consensus about the role of melanin is far from being reached. The major clinical observation, pointed out by Helen Hill as well as by Albert Kligman, John Pawelek and James Nordlund, that skin cancer rates correlate inversely with skin pigmentation, is the only major evidence in favour of a protective effect of melanin. Yet this protective effect does not necessarily imply that it is exerted via the sunscreen properties of melanin itself.

One could for instance imagine that the vigourous activity of the melanocyte as a modulator of inflammation, manifested as dark or tanned skin, protects the individuals against skin cancers (Nordlund).

One can also surmise that melanin can play several roles in oxido-reduction reactions triggered by UV radiation within exposed cells, and that the end result will depend on the initial oxidation status of the cell, that is to say, to every experiment a different result (Menter & Willis).

As a matter of fact the photochemical, photophysical and physico-chemical properties of melanins have been discussed by Miles Chedekel who stated that" melanin can contribute to photoprotection by directly scavenging free radicals, especially active oxygen species". On the other hand Menter and Willis wrote:" depending on the reductant, melanin either retards or accelerates ferricyanide reduction. Melanin also acts as an electron conduit in markedly accelerating the tyrosinase catalyzed oxygenation of p-hydroxyanisole...The net result of such melanin mediated processes, if they occur in vivo, could be either beneficial or deleterious to the organism"

The outcome of the meeting was brilliantly expressed by the title of Kligman's abstract: Is melanin photoprotective? Answer: sometimes yes, sometimes no"

Some authors are considering the possibility that the properties of melanin might depend on its chemical and stereochemical properties. Melvin Eisner suggested that "the protective capabilities of melanin may be influenced strongly by the morphology of the melanin granule. The layered structures found in vivo do not seem to make full use of the optical absorptivity of the interior melanin, suggesting perhaps a separate quenching or sequestering role".

On the basis of all these consideration, an observation made in our laboratory might help in designing new experiments. We have indeed observed that in the presence of metal chelators such as ~ 1 millimolar EDTA or ~ 10 millimolar citrate or ~ 100 millimolar histidine, some melanins become water soluble at neutral pH and can be re-precipitated by the addition of millimolar amounts of di-valent cations such as calcium, magnesium, iron, copper and so forth (10,11).

The interesting aspects of the phenomenon is that:

i) divalent cations can also precipitate eumelanin dissolved in sodium hydroxide

ii) the precipitate forms particles the diameter of which can be submicrometric

iii) conditions can be found in which the diameter increases slowly with time.

These findings make it possible to prepare melanins with different physico-chemical properties in oreder to check Eisner's hypothesis. They also bring circumstantial evidence in favour of the model which suggests that some of the protective properties of melanin are linked to its capability to bind iron and other transition elements which might play a role in photofenton phenomena or in metal catalyzed oxidations.


1) Nordlund, J.J., Abdel Malek, Z.A., Boissy, R.E. & Rheins, L.A. (1989) Pigment Cell Biology: an Historical Review. J. Invest. Dermatol. 92, 53S - 60S.
2) Morrison, W. L. (1985) What is the Function of Melanin? Arch. Dermatol. 121, 1160 - 1163.
3) Johnson, B.E., Mandell, G. & Daniels, F., Jr. (1972). Melanin and Cellular Reactions to Ultraviolet Radiation. Nature New Biol. 235, 147 - 149.
4) Schothorst, A.A., Evers, L.M., Noz, K.C., Filon, R. & van Zeeland, A.A. (1991). Pyrimidine Dimer Induction and Repair in Cultured Human Skin Keratinocytes or Melanocytes After Irradiation with Monochromatic Ultraviolet Radiation. J. Invest. Dermatol. 96, 916 - 920.
5) De Leeuw, S.M., Janssen, S., Simons, J.W.I., Lohman, P.H.M., Vermeer, B.J., & Schothorst, A.A. (1994) The UV action spectra for clone-forming ability of cultured human melanocytes and keratinocytes. Photochem. Photobiol. 59, 430 - 436.
6) Hill, H.Z. & Hill, G.J. (1987) Eumelanin Causes DNA Strand Breaks and Kill Cells. Pigment Cell Res. 1, 163 - 170.
7) Hill, H.Z., Cathcart, C.N., Bargellini, J., Trizna, Z., Hill, G.J., Schallreuter, K.U. & Wood, J.M. (1991) Does melanin Affect the Low LET Radiation Response of Cloudman S 91 Mouse Melanoma Cell Lines?  Pigment Cell Res. 4, 80 - 86.
8) Young, A.R., Potten, C.S., Chadwick, C.A., Murphy, G.M., Hawk, J.L.M. & Cohen, A.J. (1991) Photoprotection and 5-MOP Photochemoprotection from UVR-Induced DNA Damage in Humans: The role of Skin Type. J. Invest Dermatol, 97, 942 - 948.
9) Maubru, M., Audousset, M.P., Giacomoni, P. U. & Marrot, L. (1994) Utilisation de sels de Magnesium dans un procédé de teintures des fibres kératiniques mettant en oeuvre le 5-6 dihydroxyindole ou l'un de ses dérivés, procédés et composition les mettant en oeuvre. French patent n° 2700266 published the 13th of July 1994 agreed the 23rd of November 1994.
10) Giacomoni, P.U., Marrot, L., Mellul, M. & Colette, A. (1994) Procédé de préparation d'un pigment mélanique de faible granulométrie et son utilisation en cosmétique. French patent file n° 2704554 published the 4th of November 1994.