Sunburn and mechanisms of protection 37 repair. Excision repair has already been mentioned and is the best known mechanism in both bacteria and mammalian cells. It is a multi-enzyme process in which a segment of the DNA strand containing a dimer is removed and repair synthesis, based on the tem- plate of the sister strand, fills in the gap. The process is completed when the end of the newly synthesised segment is joined into the parent strand. A second repair mechanism, termed recombinational or post-replication repair (59, 60) involves the filling in of gaps left in daughter strand DNA after replication in which thymine dimers in the template are by-passed by DNA polymerase. A third form, photoreactivation, is a uv-A stimulated, enzymic splitting of the dimers. Although its action is well established in bacteria and most animal cells, higher mammals do not appear to possess the typical photoreactiva- tion enzyme (61). Photoreactivation does not occur in higher mammalian cells in culture but some form of uv-A induced recovery of sunburn damage has been demonstrated, and Sutherland (62) has isolated a photoreactivation type enzyme from human leukocytes. Until the late 1960's and the autoradiographic studies of Fukuyama, Epstein and Epstein (63), Epstein, Fukuyama and Epstein (64) and Epstein, Fukuyama and Epstein (45), few studies of the possible involvement of nucleic acids in the reactions of skin to uvr had been attempted. Tickner (64), although demonstrating a decrease in phospho- lipid content, a finding which drew attention to the uvr effects on cellular membranes, could find no change in DNA or RNA content of mouse skin exposed in vitro to massive doses of uvr. This is perhaps not surprising in view of the now known photochemistry of the nucleic acids. Baden and Pearlman (66) foreshadowed the autoradiographic studies by using chemical extraction methods to demonstrate that synthesis of RblA and protein is inhibited in mammalian epidermis shortly after uv-irradiation in vivo, recovering by 24 h and being increased at 72 h after exposure. A similar study of DNA synthesis (67) shows it to be inhibited shortly after exposure, the degree of inhibition and its duration depending on the exposure dose. In 1970, Pathak and his co-workers reported the first isolation of thymine dimers from uv-irradiated skin, confirming this report later and showing that the effect was restricted to the sunburning wavelengths of uvr (68, 69). Bowden et al. (70) isolated thymine dimers from uv-irradiated mouse epidermis and appear to have demonstrated a form of repair replication to take place, although excision repair in rodent cells is thought to operate only at very low efficiency, if at all. Cooke and Johnson (71) have also isolated thymine dimers from mouse skin exposed to uvr in vivo but could find little evidence of excision repair, the dimers being retained in the skin for at least 24 h. Cooke's extensive studies of DNA involvement in mouse skin reactions to uvr have failed to demonstrate a clear cut relationship between thymine dimer formation and inhibition of DNA syn- thesis or the vascular reactions produced. For example, the dimer yield after exposure to 'sunlamp' fluorescent tube irradiation, peak emission around 315 nm, is very low but severe skin reactions are elicited and inhibition of DNA synthesis appears relatively efficient. The involvement of DNA in human-skin reactions to uvr might be inferred from the findings that repair processes may be deficient in cells from patients with the various forms of Xeroderma Pigmentosum (72, 73, 74). The major feature of this autosomal recessive disease is an abnormal skin sensitivity to sunlight. Increased epidermal damage and accelerated development of the chronic effects of exposure such as premature ageing and cancer of the skin appear to be magnified normal skin responses to uvr, and for these re- actions DNA may well be the primary molecular target. However, the vascular reactions

38 B. E. Johnson to uvr in Xeroderma Pigrnentosum differ from those of normal skin, the abnor- mality is not simply a magnification of the normal response and the inference that uvr erythema is mediated through damage to DNA cannot be drawn. Similarly, the vascular reactions in the photosensitised response of 8-methoxypsoralen, almost certainly result- ing from damage to DNA, differ from the normal uvr erythema (75, 76). At the present time, there would appear to be no involvement of DNA in the uvr induced lesions of photodermatoses such as polymorphic light eruption and actinic reticuloid (77). PHARMACOLOGY Ultraviolet erythema has come to be regarded as a typical inflammatory state (78, 79) although the degree of leucocyte migration is significantly less than in other forms of inflammation. Lewis (80) likened sunburn to the reaction obtained with histamine and suggested that the famous 'H' substance was the mediator involved. Histamine depletion and anti-histamine administration prior to irradiation fail to modify the uv-erythema however (81, 82). Serotonin may have a role in rat skin reaction to uvr (83) but does not appear to be involved in human erythema (84). Increased concentrations of vasodilator substances have been detected in blood and skin of uv-irradiated animals (1) but the kinins observed by Epstein and Winkelmann (85) in the perfusate of irradiated human skin are apparently not involved in the delayed erythema reaction. Greaves and Sonder- gaard (86) reported the presence of a prostaglandin-like substance in dermal perfusates of uv-irradiated human skin, the time course of its appearance approximating that of the erythema. Mathur and Gandhi (87) described an increase in prostaglandin content of rat skin after uv-irradiation and indomethacin, a potent inhibitor of prostaglandin synthesis, appears to prevent the development of uvr erythema (88, 89, 90). Greaves and his co- workers have recently shown that Prostaglandin E2 is the major active constituent of uv-irradiated skin, but the site of its derivation has not been ascertained. If it is produced mainly in the epidermis, either as a result of photochemical changes in cellular fatty acids or by a process initiated by lysosomal enzymes, it may well be the mythical, dif- fusible mediator beloved of this field and so elegantly discussed by Leun (13). CHRONIC EFFECTS The long-term effects of repeated exposures to high intensity sunlight have been well reviewed by Blum (91), Epstein (92) and Urbach, Epstein and Forbes (93) a good sum- mary is presented in the book of Magnus (4). The major change observed in the exposed skin of lightly pigmented peoples is an accelerated ageing. The skin loses its natural elasticity, there is marked epidermal atrophy and increased amounts of mucopolysac- charides are found in the dermis, along with deposits of an elastotic material. Focal, benign abnormalities of keratinisation develop to form small, crusty lesions called solar keratoses and these may be succeeded by basal and squamous cell carcinomas. Evidence for the role of solar uvr in skin carcinogenesis in humans is entriely epidemiologic but convincing. The majority of skin cancer, where no other direct carcinogenic influence is known, is found in the sun-exposed skin of lightly pigmented people and is rare in Negroids. There is an approximate correlation between the incidence of skin cancer and the degree of insolation in similar populations living in different areas and turnours develop mostly in areas of the skin most directly exposed to sunlight. This epidemio- logic evidence is supported by data from numerous studies with experimental animals.