Date: 31st January 1998
The following Position Paper was produced by members of the ASCC Technical Committee in accord with its brief: to critically evaluate advances in skin biology as these may have an impact upon the industry and the community it serves. In this case, it is claimed by members of the academic community that a phenomenon known as photoimmunosuppression (the suppression of immune function following exposure sunlight) is a very important , if not necessary, event for the development of skin cancer following chronic sun exposure. From this claim it has followed that sunscreens should therefore protect from photoimmunosuppression if they are to contribute to skin cancer prevention.
Examination of the immediate sunscreen issue leads inevitably into the broad field of cancer immunology which underpins it. This field is daunting with respect to its extent, complexity and a formidable array of experimental and clinical studies, no one of which can be definitive. The approach adopted here has been in accord with the scientific principle that the best science emerges from attempts to disprove theory rather than to prove it. The question before us is whether there is sufficient evidence to provide a basis for accepting the claim that sunscreens should protect from immunosuppression, and its implicit and very important corollary, that if a sunscreen does not so protect then its use may constitute a hazard.
Following examination of the evidence we find that it is far from compelling in favour of the claim made by many immunologists. The following paper presents a number of questions arising from the available data which need to be answered to establish the claim on a basis of fact rather than assumption.
It is not the intention of this paper to represent a judgement upon this issue except to declare a “state of debate” which must be pursued toward a firmer basis for the acceptance or rejection of the claim that sunscreens should protect from photoimmunosuppression.
Exposure of the skin to sunlight results in a very complex cascade of effects. These effects are entirely dependent upon the type of skin, the quality and quantity of sunlight, and the general health status of the body with which the skin as an organ is integrated. For the skin, as for every other biological process there are thresholds beyond which necessary and beneficial stimuli become adverse, provoking equally complex arrays of repair and recovery processes. Sunburn is a familiar signal which tells us that cell damage has occurred, inflammatory mediators are responding, and repair is in progress. It is important that these repair processes are switched off when repair is complete. If this downregulation of repair processes is impaired then these processes themselves will inflict progressive damage, independently of the original stimulus.
The immune system is primarily engaged in recognising and eliminating foreign agents such as bacteria, viruses, fungi, and parasitic infections. The foreignness of an agent is called its antigenicity for which particular antigens can be characterised. The immune system must be very finely regulated if it is to maintain proper recognition and very accurately distinguish between self, and non-self characteristics. To ensure that self is not mistaken for non-self, leading to auto-immunity the immune system will sometimes be downregulated in circumstances such as severe burning.
Much of our understanding of skin immunity is derived from responses to skin grafting where the host animal will recognise cell surface antigens (as foreign) on the donated graft tissue and summon inflammatory mediators to reject the graft. This model of classical allograft rejection provided the basis for the hypothesis which developed into an assumption, that a cancer cell is foreign and should be recognised by the immune system. This assumption was developed further to include a notion of immunosurveillance which cancer cells must escape if they are to remain viable and propagate.
At least 20 years of experimental and therapeutic experience have failed to confirm the reality of immunosurveillance while an increasing body of evidence finds against it. Prominent amid this contrary evidence is the only marginal successes of immunotherapeutic strategies based upon it, and that no cell surface tumour-specific antigens have been isolated*. However, tumour cell nuclear antigens have been isolated. Tumour growth will always constitute a balance between populations of replicating cells and dying cells, and the antigenic status of the cell contents released into the tumour environment as a result of cell death can be expected to invoke immunological responses, but not against the living cell population which remains immunologically invisible. It is to the observation of immune cells interacting with tumours that proponents of immunosurveillance may point, but it becomes increasingly necessary to integrate the greatly increased knowledge supplied by molecular biology, as well as the function of endocrines, paracrines, autocrines and intracrines, which together provide mechanisms other than those of transplantation immunology to account for the regression or inhibition of tumour growth.
* This paper does not ignore the advances and promise of immunological engineering by which tumours cells may be rendered therapeutically immunogenic. This engineering is based upon such phenomena as mutated intracellular proteins and chromosomal translocations which result in the expression of antigens (40 - 42,). The present paper attempts to address the immunological facts regarding spontaneously arising cancers in humans and the conspicuous absence of any direct evidence of effective immunological activity directed against them (where classical antigens such as viruses are not involved).
It has been discovered that less than a sunburning dose of UV causes a transitory local immunosuppression. Based on the assumption of immunosurveillance the conclusion has been readily drawn that this is a negative response necessary for the development of skin cancer. In the light of the foregoing, far more compelling evidence can be cited to suggest that this very readily induced immunological response is a positive one which anticipates sunburn, assists in the downregulation of inflammation, protects from potential autoimmunity, and may be irrelevant to skin tumour cell growth.
Thus, and quite contrary to the dominant way of thinking, a sunscreen which does not prevent UV induced immunosuppression may be a better sunscreen than one which does.
Brief History and Elaboration
Most immunologists believe that the immune system of the skin has a central role in the prevention of skin cancer. It is believed that certain cells with immunological functions act as resident or circulating guardians which together detect and destroy foreign transformed (ie carcinogenic or potentially carcinogenic) cells. There are various pieces of evidence to which they point.
The starting point is that the immune system is understood to act in this way in the detection and prevention of disease, and it is central to all that they have been taught.
It is well known that people who have had kidney transplants and who use immunosuppressive drugs are far more susceptible to skin cancer. On the face of it this is strong evidence to support the role of the immune system in prevention of skin cancer.
Seminal work on the effect of sunlight on the skin`s immune system was that of Margaret Kripke and coworkers and Daynes and co-workers. They showed that if cancer cells were implanted into the skin of syngeneic mice they would be rejected, but if the skin had been irradiated with UV-radiation prior to implantation of the tumour cells, the tumours were not rejected and could continue to grow (1-3). They subsequently showed that this effect was mediated by T-suppressor lymphocytes (4,5). If T-lymphocytes from uv-irradiated mice were injected into mice that had not been irradiated they did not reject implanted tumour cells. It is also known that UV irradiation affects the ability of Langerhans cells to present antigen (6). These cells are antigen-presenting cells (ie they are cells which recognise foreign matter and present specific information to lymph tissue which then undergoes differentiation of lymphocytes to react to the foreign material). It is also known that the keratinocytes (the cells which form the majority of the cells in the epidermis and which generate keratin intracellularly) also have an important role to play in the skin's immune system.
It is laborious to implant tumour cells into mice, and unethical to implant them into humans. A simpler, more rapid technique is to measure the response of skin to allergens. Skin which has been immunosuppressed (eg by UV- irradiation) shows reduced reaction to contact allergens (7). Thus it is possible to determine the immunological status of skin by observing the response to contact allergens. If the skin does not react to an allergen at the site at which UV irradiation took place, it is said to be locally immunosuppressed (6,8). If it does not react at a site remote from the site of UV-irradiation it is said to be systemically immunosuppressed.
These techniques have been widely used to determine whether sunscreens protect against uv-induced immunosuppression - with very mixed outcomes. Some workers have reported that sunscreens do not protect whereas others report that they do protect (9-12). Where lack of protection has been reported it has usually been inferred that although sunscreens may protect against sunburn, since they do not protect against immunosuppression they do not therefore protect against skin cancer.
Later workers (such as Yoshikawa (13)) have reported that there are genetic determinants which render some strains of mice and humans more susceptible to immunosuppression and skin cancer.
Superficially these observations together with others provide a plausible support for immunological rejection of tumours. However, upon closer examination of the methodologies, together with recent knowledge including the molecular genetics of tumours, the location of tumour antigens, as well as the phenomenon of apoptosis (genetically programmed cell death/suicide), the case for tumour rejection as an example of allograft immuno-rejection is far from compelling.
There are a number of pieces of evidence which need to be addressed before the conclusion can be drawn that sunscreens which do not protect from immunosuppression do not therefore protect against skin cancer.
1) The experiment, appropriately controlled, which takes a sunscreen which has been shown on humans not to protect from immunosuppression and then applied to the skin of an animal model under carcinogenic solar simulated exposure conditions, has not been done.
2) Reeve and Greenoak (39) showed that hairless mice immunosuppressed with two drugs used in renal transplant recipients, Azothioprine (Az) and Cyclophosphamide (Cy), where Az but not Cy had been shown to be associated with the progression of non melanoma skin cancer to malignancy, only significantly increased the skin cancer incidence following Az treatment (and this only slightly). Cinnamate and o-PABA sunscreens were shown to protect extremely well from skin cancer under these conditions.
3) There is a severe lack of standardised protocols for conducting these immunosuppression experiments. For example, Roberts (11) has pointed out that many of the sources of UVR that were used contained unacceptably high levels of UVC (wavelengths that are not represented in sunlight). Some workers claim that the action spectrum for UV-induced immunosuppression peaks in the UVB and UVC (15), whereas others claim it is UVA (10).
4) Many of the experiments have used hairless mice as a model for various cancer and UV-induced immunosuppression experiments. However nude mice, which lack a thymus gland - and therefore have no T-lymphocytes to protect them against cancer - do not show increased susceptibility to cancer (17). In addition, tumours in these athymic nude mice can regress following treatment with hormone regulators such as melatonin (44) Such regressions cannot be T-cell mediated.
5) The phenomenon of UV-induced immunosuppression is very widely conserved throughout mammals. It is found in eutherian mammals and marsupials. If UV-induced immunosuppression makes such a wide variety of animals susceptible to skin cancer, why is it so widely conserved? Perhaps it is designed to stop the skin being rejected after a sunburn - but that only happens with any frequency in man. Streilein (25) has suggested that UV-induced immunosuppression may serve to protect against severe immune reactions against parasitic infections - but Kripke (19) argues that UV-induced immunosuppression renders people more subject to disease in the first place - it seems to be a strange evolutionary advantage. Furthermore, if photoimmunosuppession confers a significant predisposition to infection and disease generally, then this impact is not obvious in an Australian population where the skin cancer incidence demonstrates a highly biologically effective sun exposure.
6) The mediating chemical in the skin is thought by some to be urocanic acid (18). This is found in the outermost layers of the stratum corneum. It is highly water-soluble and so is presumably removed from the skin of people who are exposed to daily washing and particularly to those who swim. Such habits are not known to be protective against skin cancer. Most now believe that while urocanic acid may be involved, DNA damage is also necessary for immunosuppression to occur (19, 38).
7) Black people are much less likely to get skin cancer, yet they are just as readily immunosuppressed as white people (20). It is argued that the melanin in the skin protects the DNA of these black skinned people - but sunscreens are also highly protective of the DNA.
(The action spectra for sunburn and DNA damage are very similar, so if a sunscreen protects against sunburn it will be expected to protect the DNA. This has been confirmed in a limited number of cases (21-24).
8) The action spectra (the effectiveness per wavelength to produce the endpoint of interest) for sunburn, non-melanoma skin cancer and DNA damage are very similar. It is therefore reasonable to argue that if you protect against sunburn you should also protect against skin cancer. On the other hand there seems to be no agreement on the wavelengths over which UV-induced immunosuppression operates. If it is UVA, as maintained by Halliday, then it is difficult to understand how UV-induced immunosuppression can play anything but a very secondary role in the initiation and promotion of skin cancer. (Conceivably it might play a role in progression, but sunscreens are not recommended to prevent progression of skin cancers!).
In contrast to Halliday's conclusion, and more recently, Reeve et al. have shown that if all the UVB is eliminated from the experimental light source then UVA is immunologically innocuous. Furthermore, pre or post irradiation with UVA can reverse the immunosuppressive effect of UVB (45). Such contrary findings may be due to light source differences as noted in 3), but this “may be” together with so many others does not constitute a basis for a claim.
9) It has been shown that UVB sunscreens, such as octyl methoxycinnamate and octyl dimethyl PABA, do not protect against UV-induced immunosuppression (at least by some workers). However it has been shown several times, that they are highly protective against skin cancer in hairless mice (26-32). More specifically Bestak (33) showed that octyl methoxycinnamate was only very weakly effective in protecting against UV-induced immunosuppression, but was totally protective against skin cancer when the radiation was continued.
Kerr and Reeve (34) found that octyl methoxycinnamate and octyl dimethyl PABA failed to protect the immune system, which later recovered. Mice which were protected with SPF6 sunscreens (cinnamate and o-PABA) and which received 6 times the dose of UV, developed the same number of tumours (with cinnamate), or significantly fewer tumours (with o-PABA) than the unprotected mice (which had received only 1/6th of the dose). Thus the effect of equal doses of UV-radiation was the same (or less) whether delivered through a sunscreen or in its absence. The interesting point was that although the mice were initially immunosuppressed, they recovered the ability to react to allergens up to the point at which the tumours appeared - at this time the immune system became severely suppressed. That is to say that immunosuppression seemed to be a result of the expression of skin cancer, not the cause of it.
It has been suggested that although in these experiments immunosuppression was not necessary for non melanoma skin cancer, it may still be important for malignant melanoma. However, although the action spectrum for malignant melanoma has not been established for mammals, we do know from epidemiolgical studies that the incidence of malignant melanoma relates strongly to severe sunburn events - which strongly suggests that it is predominantly UVB. Also DNA damage is involved in melanoma and the action spectrum for DNA damage is predominantly in the UVB region. Thus it is reasonable to suppose that the action spectrum for malignant melanoma is similar to that for non melanoma skin cancer. Yet according to Halliday it necessary that the sunscreen absorbs the longer UV wavelengths to protect against immunosuppression.
10) Menzies and Greenoak et al (35) found that tumours could be induced by irradiating small areas of skin of hairless mice, even though the mice were not immunosuppressed. This study showed that while systemic immunosuppression may be a result of whole body UV irradiation it was not a necessary condition for the growth of skin cancer. This study is one of the very few where the skin cancer model was not a transplanted tumour (with which transplantation antigens as well as viral contaminants may be associated) but tumours directly induced by simulated sunlight UV.
11) The outward manifestation that the skin's immune system is active is inflammation. If it has been suppressed the inflammatory reaction is suppressed. The observation that sunlight suppresses immune responsiveness has lead some to conclude that sunscreens do not prevent skin cancer. Paradoxically anti-inflammatory drugs, which also suppress inflammation have been shown to protect against skin cancer (36,37).
12) One reason that has been advanced for stating that sunscreens do not protect against skin cancer is that there is a threshold dose of UVR that is required to elicit erythema (MED = minimal erythemal dose), whereas de Gruijl and van der Luen (16) showed that as little as 1/32nd of an MED delivered chronically to hairless mice is sufficient to cause skin cancer - which suggests that there is no threshold for skin cancer. By this argument if UV-induced immunosuppression has a critical role to play in skin cancer it should have no threshold dose - but the indications are that the threshold dose for UV-induced immunosuppression is about 0.5 MED.
13) At a very much more fundamental level of understanding cancer, the widely held, and very long standing belief that the body guards against the growth of cancer cells by continual immunosurveillance is not well supported by the only marginal successes of immunotherapeutic strategies based upon it. Immuno-enhancement in some breast cancer patients has been shown to worsen prognosis, and the most effective chemotherapeutic agents are also strongly immuno suppressive. It remains that no cell surface specific-tumour antigens have been isolated. That a cancer cell should be recognised as foreign is rather an assumption than a well supported fact (also see 4) above.
14) In the light of the above, an emerging school of thought considers cancer as “an aberrant regulatory process whose defects and clinical course are reversible”. This “evolved” model represents a revolutionary distinction from the classical model (parasite) for which direct attack and killing of cancer cells has been the dominant approach to treatment. Since killing cancer cells almost always involves killing normal cells and preventing cure, the evolved model is attempting to address a very much more complex and subtle notion of what cancer is (43). According to this model immunological factors are not given the central role in permitting or eliminating cancer development, but are seen as one among many processes important for carcinogenesis.
15) The evolved model presents for example a different way of interpreting Langerhan cell migration as a result of UV exposure. If the photoimmunosuppressive effect is in fact a protective and positive response which anticipates burning and potential auto-immunity, then the disappearance of the antigen presenting Langerhan cells can be seen to be rationally consistent with such a function.
16) The evolved model represents a response to the (IT driven) accelerated availability of considerably more data relevant to any and every phenomenon and the need to develop multi-disciplinary approaches to the interpretation of research findings. A branch of photobiology examines the effects of light on hormone regulation and in particular the central role the pineal gland plays in the circadium rhythms. High lux visible light, especially at the blue end of the spectrum has been shown to have very significant effects on the immune system via the serotonin/melatonin pathways (46). Mice used in photoimmunological studies are often maintained under low, “red-end”, lighting conditions which would be expected to have the same effect on pineal function as “dark” conditions. Solar simulation is likely to significantly entrain the circadium rhythms in a manner directly affecting immune responsiveness. However, photoimmunologists do not yet control for this variable which may lead to significant misinterpretations of data.
1) For measuring UV-induced immunosuppression there is no single agreed protocol that has been validated by a double blind inter-laboratory comparison. Nor do many of the other validation checks appear to have been carried out. Often people have used solar simulators that emit UVC (which according to some workers is the most immunologically active part of the UV spectrum ). Furthermore the results reported by different workers present a confused and contradictory picture.
2) There is considerable doubt about the relevance of UV-induced immunosuppression to the initiation and promotion of skin cancer. At the very least, proponents of the idea that UV-induced immunosuppression is relevant have to explain why several workers (whose general writings show that they are firmly committed to the concept of the importance of UV-induced immunosuppression to skin cancer) have found that whereas sunscreens have been shown to protect from skin cancer in animal models, the same sunscreens may or may not protect from UV-induced immunosuppression.
3) At this stage we conclude that there is insufficient evidence to show unequivocally that protection against UV-induced immunosuppression will help prevent skin cancer. In view of the emotive appeal of a claim such as “protects the skin's immune system” to HIV-infected people, or people at risk of such infection, we believe that it would be imprudent (as well as misleading) to permit such claims without a far more convincing case.
4) It might be argued that the where there is both supporting and contrary evidence for a potential benefit of sunscreens which protect from photoimmunosuppression (and a hazard associated with those which do not so protect) then those who wish to invest in pursuing such claims as “this product will protect your immune system” should be free to do so. Those who so argue will support the assumption (and not the fact) that a compromised immune system is necessary for cancer growth. They must also accept the vulnerable assumption that photoimmunosuppression has the potential only of negative effects, despite point 3) above.
If products are marketed claiming protection of the immune system, then it will become a market force driving others to adopt it regardless of its scientific basis. The result of this will be more expensive sunscreens and the real risk of decreased useage. Like the most recent oxybenzone may be carcinogenic story, the cinnamates cause cancer story, the urocanic acid story, the PABA scare, the long and expensive process of correcting the basis for an SPF15+ ceiling, and the “sunscreens don't work” headline, the primary bases for all these issues have been exposed as fundamentally false. However, the bad press stays, and it is the public who have been the losers every time. This history should convince us that the evidence to support claims must be compelling, and sufficient to answer the many contraindicating facts in a scientific way.
There are many basic questions to be answered and many studies to be done to satisfy the demands of the issues addressed by this paper. The ASCC would like to look forward to working with academic scientists, with whom it shares concerns, in laying firm foundations for the understanding of the relevance of immunological processes to skin cancer.
- Kripke ML Adv Cancer Res, 34: 69-106, 1981
- Kripke ML & Fisher MS J Natl Cancer Inst, 57:211-215, 1976.
- Spellman CW, Woodward JG, Daynes RA, Transplantation, 14:112- 119, 1997.
- Fisher MS, Kripke ML, Proc Acad Natl Sci, 74:1688-1692, 1977.
- Daynes RA, Spellman CW, Cell Immunol, 31:182-187, 1977
- Toews GB, Bergstresser PR, Streilein JW, 124:445-453, 1980.
- Noonan FP, De Fabo EC, Kripke ML Photochem Photobiol 34:683-689, 1981
- Haniszk J, Suskind SS, J Invest Dermatol 40:183-191, 1983
- Hersey P, MacDonald M, Burns C, Schibeci S, Matthews H, Wilkinson FJ, J Invest Dermatol, 88:271-276,1987
- Bestak R Barnetson R St C, Nearn MR, Halliday GM, J Invest Dermatol 105:345-351, 1995
- Roberts LK, Beasley DG, J Invest Dermatol 105, 339 -345, 1995
- Wolf P,Cox P, Yarosh DB, Kripke ML, J Invest Dermatol 104, 287-292 1995
- Yoshikawa T, Rae V, Bruins-Slot W, Van den Berg JW, Taylor JR, Streilein JW, J Invest Dermatol 95:530-536, 1990
- Noonan FP, De Fabo EC, Immunol Today, 13: 250-254, 1992
- Elmets CA, Le Vine ML, Bickers DR, Photochem Photobiol 42:397- 1985,
- van Weelden H, deGruijl FR, van der Luen JC in Biological Effects of UV Radiation 137-146 ec Urbach F, Gange RW Praeger Publ. 1986.
- Hoover, TL., Morrison, WL., Kripke, ML, (1987) Transplantation. 44:693-695
- Gibbs NK, Norval M, Traynor NJ, Crosby JC, Johnson BE, Photochem Photobiol, 58:643-647, 1993
- Kripke ML, Cancer Research, 54:6102-6105, 1994
- Vermeer M, Schmeider CJ, Yoshikawa T, Van den Berg JW, Metzman MS, Taylor, JR, Streilein JW, J Invest Dermatol 97:729-734, 1991.
- Susuki M Photodermatol 4 209-211, 1987
- Freeman SE, Ley RD, Ley KD Photodermatol, 5:243-247, 1988
- Lowe NJ, Breeding J Cosmetics & Toiletries 98:65-68, 1983
- Wolf P, Yarosh DB, Kripke ML, J Invest Dermatol 101:523-527
- Streilein JW, J Invest Dermatol, 100:47S - 53S, 1994
- Reeve VE, Greenoak GE, Gallagher, CH, Canfield PJ, Wilkinson FJ, Aust J Exp Biol Med Sci 63:655-665, 1985
- Forbes PD, Davies RE, Sambucco CP, Urbach F, J Toxicol Cutaneous Ocul Toxicol, 8:209-226
- Kligman LH, Akin F. Kligman AM, J Am Acad Dermatol, 3:30-35 1980
- Reeve VE, Bosnic M, Boehm-Wilcox C Photodermatol Photoimmunol Photomed, 7:222-227 1990
- Greenoak GE, Torkamanzehi A, Nearn MR Cosmet Aerosol Toiletries Aus 7:12-17 1993
- Thompson SC, Jolley D, Marks R N Engl J Med 329:1147-1151 1993
- Naylor MF, Boyd A, Smith DW, Cameron GS, Hubbard D, Nelder KH, Arch Dermatol 131:170-175 1995
- Bestak R, Halliday GM Photochem Photobiol 64:188-193 1996
- Reeve VE., The Relationship Between Sunscreen Protection from Erythema, Photoimmunosuppression and Photocarcinogenesis in the Hairless Mouse. Sunscreen Photobiology: Molecular, Cellular and Physiological Aspects, ed. Gasparro FP., Springer - Verlag and Landes Bioscience 1997, Ch 9, pp 157-176.
- Menzies Greenoak GE, Cancer Res 51;2773-2779, 1991
- Slaga TJ, Carcinogenesis, 5:111-125 1980
- Slaga TJ, J Nat Cancer Inst, 51:1723-1725, 1973
- Hurks H.M.H., Out-Luiting C., Vermeer, Bert-Jan, Class, Frans H. J., Mommaas A.M. In situ Action Spectra Suggest that DNA Damage is involved in Ultraviolet Radiation-induced immunosuppression in Humans. Photochem. Photobiol. 1997, 66(1): 76-81
- Reeve VE., Greenoak GE., Gallagher CH., Canfield PJ., Wilikinson FJ., Effect of Immunosuppressive Agents and Sunscreens on UV Carcinogenesis in the Hairless Mouse. Aust. J. Biol. Med. Sci. 63(Pt 6) 655-665 (1985).
- Scott AM, Cebon J. Clinical promise of tumour immunology. Lancet 1997; 349 (suppl II): 19-22
- Kawai A., Woodruff J., Healey JH., Brennan MF., Antonescu CR., Ladanyi M., SYT-SSX gene fusion as a determinant or morphology and prognosis in synovial sarcoma. The New England J. of Med., 1998, Vol. 338, No 3, 153-160
- Rabbitts TH., The clinical significance of fusion oncogenes in cancer. The New England J. of Med., ., 1998, Vol. 338, No 3, 192-194
- Schipper H., Turley EA., Baum M., A new biological framework for cancer research. Lancet 1996; 348: 1149-51
- Narita T., Kudo H., Effect of melatonin on B16 melanoma growth in athymic mice. Cancer Res. 1985, Sep; 45 (9): 4175 -4177
- Reeve VE., Bosnic M., Boehm-Wilcox C., Nishimura N., Ley RD., Ultraviolet A Radiation (UVA; 320-400 nm) Protects Hairless Mice from Immunosuppression Induced by Ultraviolet B Radiation (UVB; 280-320 nm) or Cis-Urocanic acid. Photochem. Photobiol. in press.
- Maestroni GJ., T-Helper-2 Lymphocytes as Peripheral Target of Melatonin Signaling. Journal of Pineal Research. 1995; 18: 84-89