The Consequences of UV-Induced Immunosuppression for Human Health

Authors


Corresponding author email: m.norval@ed.ac.uk (Mary Norval)

Abstract

Exposure to UV radiation can cause suppression of specific immune responses. The pathways leading to the down-regulation are complex, starting from the absorption of UV photons by chromophores in the skin and ending with local and systemic changes in immune mediators, the generation of T and B regulatory cells and inhibition of effector and memory T cell activation. The consequences for human health are thought to be both beneficial and adverse. The former are illustrated by protection against polymorphic light eruption, and possible protection against T cell-mediated autoimmune diseases and asthma. The latter are illustrated by skin cancer, cutaneous lupus erythematosus and infectious diseases including vaccination. Many outstanding questions remain in this rapidly developing and controversial area, not least what advice to give the general public regarding their sun exposure. While considerable advances have been made in the development of strategies that preserve the health benefits of sunlight exposure and decrease its detrimental effects, further research is required before optimal levels of protection are achieved.

Introduction

More than 30 years ago, Margaret Kripke and her colleagues demonstrated that exposure of mice to UV radiation (UVR) led to suppression of cell-mediated immunity to an antigen encountered within a short time after the irradiation (1). The down-regulation was not general but was specific to the antigen tested, and was shown subsequently to be mediated by T cells (2). While the first experiments used tumor cells as antigen, later studies have shown the same outcome using a wide variety of antigens including alloantigens, contact sensitizers and microorganisms. In addition to UVR exposure affecting the generation of primary immune responses, it can also suppress already established (memory) immune responses in humans (3). Tracking of T cells in murine experiments has conclusively shown that UVR suppresses the activation of T cells in skin-draining lymph nodes during a primary immune response, as well as their differentiation into skin-homing memory T cells that produce γ-interferon (4). For ethical and logistical reasons, most work has been done in rodent models but there are sufficient results to date involving human subjects to be certain that they respond similarly. In evolutionary terms, the positive advantages of the immunosuppression might involve the inhibition of a chronic inflammatory response to the neoantigens induced by the UVR (5). Furthermore, UVR exposure promotes cell cycle arrest, allowing time for DNA repair; if this is successful, the cell cycle proceeds. Factors involved in the arrest process also contribute to the production of a range of immunosuppressive mediators. Thus, UV-induced immunosuppression may be a consequence of processes that help to maintain the genomic integrity (6). It should be noted here that the skin of rodents and most human subjects adapts to chronic UVR by tanning and epidermal thickening, thus giving a measure of protection against sunburn. In contrast, repeated exposures do not lead to the development of protection against almost all of the immune effects of UVR that have been assessed in both mice and humans (7).

The pathways leading from UV exposure of the skin to reduced T cell activity are complex and are outlined in the first section below. This part is followed by sections in which the potential benefits to human health of UV-induced immunosuppression are considered, and then the potential adverse consequences assessed. A final section deals with the difficult issue of the best advice to give the public concerning personal sun exposure behavior to ensure optimal immunological and other health outcomes.

Pathways leading to UV-induced immunosuppression

While it is not disputed that UVR exposure can suppress immunity, it is clear that the mode of action is complex with the possibility of different pathways being involved. Several of the variables in experimental and natural systems relating to the UVR source and to the antigen are noted in Table 1. In local UV-induced immunosuppression, the antigen is applied to the irradiated site. Below and in Fig. 1, an outline of the steps leading to local suppression of contact hypersensitivity (CHS) can be found. In systemic UV-induced immunosuppression, the antigen is applied to a distant unirradiated site and here the steps are not clear, although no change in the function or the numbers of antigen presenting cells in the lymph nodes draining the site of antigen application have been detected (8,9). Thus, it is likely that other cell types or the release of immune mediators from the irradiated site, such as prostaglandins, are involved. Recently dendritic progenitor cells prepared from the bone marrow of irradiated mice were shown to induce tolerance, suggesting that they could have a key role (10). It is also of interest to note that UV irradiation of the eye alone can lead to systemic immunosuppression, the mechanism thought to be via a nitric oxide-dependent hypothalamo-pituitary promelanocortin pathway (11).

Table 1.   Factors that could determine which pathway leads to suppression of cell-mediated immunity following UV radiation.
UV sourceAntigen
SpectrumType (e.g. sensitizer, replicating microorganism, complex protein)
Dose (suberythemal/erythemal)Dose
IntensityRoute of administration
Frequency of exposureHost species and strain
Area of body irradiatedApplication to irradiated site (local) or distant unirradiated site (systemic)
Length of time between exposure and antigen applicationNew antigen (suppression of induction phase) or previously experienced antigen (suppression of elicitation phase)
Figure 1.

 Outline of the steps in the cascade from UV radiation to immunosuppression.

Chromophores initiate the process leading to immunosuppression by absorbing the UV photons and changing their structure as a result. The recently described action spectrum for UV suppression of memory immunity in humans showed that there are two distinct UV wavebands that suppress immunity. One is centered within the UVB waveband (290–320 nm) at 300 nm, while the other is at 370 nm, the upper end of the UVA waveband (320–400 nm) (12–14). The separation of these wavebands and the different immunosuppressive dose responses indicate that there are likely to be at least two independent chromophores and different pathways that lead to immunosuppression. This becomes even more complex as different times are required for the UVB and UVA chromophores to induce the immune changes, with interactive events between these wavebands eventually occurring (15). For UVB wavelengths, the chromophores are located in the upper layers of the skin in positions enabling this short wavelength part of the UV spectrum, which has poor penetrating power, to reach them. Evidence gathered over a number of years indicates that there are at least three such UVB chromophores and which is the most important may vary from circumstance to circumstance (see Table 1). The main ones are DNA (16), trans-urocanic acid (UCA) (17) and less defined membrane components (18). Photoproducts are formed from DNA following UV irradiation, most commonly cyclobutane pyrimidine dimers, followed by pyrimidine (6-4) pyrimidone. Trans-UCA, formed from histidine as filaggrin breaks down in the stratum corneum, accumulates in this site as the enzyme urocanase, which catabolizes it, is not present in skin. Trans-UCA isomerizes to cis-UCA in a UV dose and wavelength-dependent manner until the photostationary state is reached at about 60%cis-UCA. UVR exposure induces an alteration in cellular redox equilibrium causing free radical formation (oxidative stress) and membrane lipid peroxidation. The chromophore(s) for 370 nm UVA-induced immunosuppression is unknown, but is not riboflavin. Despite absorbing at the correct wavelength for UVA but not UVB-induced immunosuppression, riboflavin protects the human immune response from both of these wavebands (19), possibly because it plays a key role in aerobic metabolism.

It should be noted that some moisturizers and vehicles can affect the pathways leading to immunosuppression. For example, a skin ointment decreased the formation of DNA damage induced by UVR exposure in the skin of hairless mice by 53% (20).

Because of these structural changes in the chromophores upon UV absorption, a range of mediators is produced in the irradiated site from keratinocytes and other cell types. Probably the first is platelet activating factor (PAF) which can bind to receptors on monocytes, mast cells and keratinocytes and activate prostaglandin release (21). This is followed by the release of various cytokines, including interleukin(IL)-4 and IL-10, both of which are immunosuppressive (22). Other molecules are also found locally, such as histamine, prostaglandins, tumor necrosis factor-α (ΤΝF-α), IL-1β, neuropeptides and neurohormones, and these can become systemic. The complement cascade is also involved; UV-activated C3 is important for infiltration of the skin by monocytes/macrophages (CD11b+ cells), which contribute to immunosuppression (23). UVA activation of the alternative complement pathway, with increased levels of properdin and complement factor B, also leads to immunosuppression (24). Whether these complement components are chromophores, or are responding to UV-induced changes in other molecules in the skin, such as oxidized lipids, is unknown, although complement has been associated with immunosuppression in humans (25). Energy crisis in the skin is another important molecular event resulting in immunosuppression in humans. UV irradiation of human keratinocytes causes a blockade in glycolysis, leading to low ATP levels (26). This can be prevented with nicotinamide (vitamin B3), which is an essential cofactor in the electron transport chain. Prevention of this UV-induced energy crisis in humans with oral (27) or topical (28) nicotinamide avoids UVR-induced immunosuppression.

All the above substances have significant effects on the migration and function of several immune cell populations, some of which are involved in antigen presentation. A proportion of Langerhans cells, the major antigen presenting cells of the epidermis, migrates to the lymph nodes draining the site of irradiation (29) (Fig. 2) or, if the UV dose is high, may undergo apoptosis in situ. The dermal dendritic cells also migrate to the draining lymph nodes. Meanwhile the number of mast cell in the dermis increases transiently and then they too migrate to the draining lymph nodes (30). A specialized population of monocytes/macrophages, capable of producing IL-10, move into the skin and then to the draining lymph nodes on stimulation with antigen (31). The end result is thought to be abnormal antigen presentation in the epidermis and dermis and in the draining lymph nodes. In the latter site, there is a lowered production of IL-12 and IL-23, key cytokines that normally promote the activation of a variety of immune cells including T cells and that are capable of reducing UV-induced DNA damage (32). Simultaneously the level of the T helper 1 (Th1) cytokines is reduced, and a special subset of T regulatory cells (Tregs) is stimulated, specific for the antigen encountered shortly after the UV exposure (33). These cells have the phenotype CD4+,CD25+,Foxp3+,CTLA-4+. They are cytotoxic for antigen presenting cells, produce IL-10 on activation and can suppress the activation, cytokine production and proliferation of other types of immunostimulatory T cells. There is also more limited evidence for the involvement of NK-T cells, which secrete IL-4 on activation (34). These cells represent a unique set of lymphocytes that express NK cell markers plus a T cell receptor (35). Recently it has been demonstrated that the Langerhans cells that migrate to the skin-draining lymph nodes following UVR exposure induce the NK-T cells to produce IL-4 (36). Another lymphocyte subset activated in response to UVB, but not UVA, is a unique population of B regulatory cells (UV-Bregs) (8). UV-Bregs inhibit activation of immunity by dendritic cells (37). The end result is significant suppression of immunity with an inhibition of the expansion of effector CD4+ and CD8+ T cells in the skin-draining lymph nodes, and an impaired development of peripheral memory T cells in the skin (4). Once generated, the immunosuppression is long-lasting, leading to tolerance so that, if the same antigen is encountered in the future, the T cell response to it is suppressed.

Figure 2.

 Confocal microscopy after staining of Langerhans cells (CD1a-positive) in epidermal sheets prepared from human skin: (a) unirradiated skin and (b) 24 h after exposure to an erythemal UVB dose.

Beneficial consequences of UV-induced immunosuppression

It is possible that UV-induced immunosuppression could provide positive advantages. Below, examples are selected, first to illustrate what can go wrong if the normal pattern of UV-induced immunosuppression does not occur, and secondly to illustrate the potential of UV-induced immunosuppression to prevent a disease or to ameliorate the symptoms of a disease.

Polymorphic light eruption

Polymorphic light eruption (PLE) is the commonest of the wide range of photosensitivity disorders, all associated with abnormal cutaneous responses on exposure to UVR and/or visible light. It is classified as an idiopathic (immunological) photodermatosis, and is found in about 5–20% of the population of Europe and North America. It occurs most frequently in the spring or early summer or following the first intense dose of sunlight, such as during a sunshine holiday. Red itchy eruptions develop on body sites exposed to the sun. This response is less likely after repeated small exposures—a process called hardening which can be undertaken clinically using artificial sources of UVR. The most effective waveband for the provocation of PLE is not clear and may vary depending on the UVR dose and its spectrum, the genetic background of the individual and what cutaneous neoantigen is involved. Evidence from limited data in the past indicated that the prevalence of PLE decreased as the latitude decreased, suggesting that the larger seasonal change in ambient solar UVR at higher latitudes could be a significant risk factor. For example, the prevalence of PLE in two Australian cities, Perth (32°S) and Ballarat (37.5°S), was 5.2% and 3.6% respectively compared with 14.8% in London (51.5°N) (38). However Rhodes et al. recently undertook a large-scale European survey covering latitudes from Greece (38°N) to Finland (60°N) and found no such gradient (39). The authors concluded that the disease could be triggered equally in countries with different seasonal patterns of solar UVR.

It is thought that PLE subjects respond to photoinduced neoantigens in the skin via an allergic reaction, a form of delayed type hypersensitivity (40). The exact mechanism resulting in the lack of UV-induced immunosuppression in these patients has not been defined but has been attributed to various factors such as fewer neutrophils (38) and suppressor macrophages (41) migrating into and fewer Langerhans cells (42) migrating away from, the irradiated skin compared with normal subjects. There may also be an enhanced proinflammatory cytokine response (43) or altered neuroendocrine signaling locally in the exposed site (44). Wolf et al. have suggested that a reduced number of Tregs, capable of releasing IL-10, could be circulating in the winter months in individuals with PLE leading to less effective immunosuppression on exposure to the sun in the spring (45). It has also been demonstrated recently that the ability to be tolerized (defined as long-term immune unresponsiveness) by UVR is impaired in PLE (46).

Thus, this disease demonstrates one very real advantage of the immunosuppression that follows UVR exposure in healthy subjects. It would be expected that a failure of UV exposure to suppress immunity in PLE patients might reduce the incidence of skin cancer. In confirmation, a prospective case–control study based in Dublin revealed that the prevalence of PLE was 7.5% in subjects with skin cancer compared with 21.4% in gender and age-matched controls without skin cancer (47). The negative aspect of UV-induced immunosuppression relating to skin cancer is covered in the section “Adverse consequences of UV-induced immunosuppression: skin cancer” below.

Th1-mediated autoimmune diseases

As there is a gradient of increasing prevalence of several Th1-mediated autoimmune diseases with higher latitude, a beneficial protective role for UVR exposure has been suggested (48). Such diseases include multiple sclerosis (MS), type 1 diabetes mellitus and rheumatoid arthritis. Most evidence has been obtained for MS and is described below.

Multiple sclerosis.  MS is an autoimmune disease with destruction of myelin-producing cells and axonal loss in the central nervous system. Its cause is complex, including both genetic and environmental risk factors. The most robust evidence for the latter is the strong latitudinal gradient in occurrence, first noted more than 50 years ago (49), and it is also recognized that increased sun exposure is associated with a decreased risk of MS (50). Recent studies validate these findings; for example, within Europe where the UV Index correlated negatively with the distribution of MS indicating that there was an increased risk in areas with low UV Index (51), within Scotland where latitude (55–60°N) correlated positively with MS patient-linked hospital admissions (52), within France where the regional UVB radiation obtained from the solar radiation database strongly predicted MS prevalence rates (53), within North American States where the UV Index had a strong negative correlation with MS prevalence (54), and from global data gathered from 54 reports where there was a very significant negative correlation between the prevalence of MS and UVR as obtained from NASA satellite data and available sunlight hours (55). A similar latitude variation in incidence of the common precursor of MS, first demyelinating events, has been reported in an Australian incident case-control study in four locations ranging from 27 to 43°S (56). Low ambient UVR or low sun exposure in childhood may present particularly significant risks for MS development in later life (50). In addition an association between the season of birth and the risk has been reported in several studies (57,58) leading to the conclusion that higher ambient UVR in the summer months during pregnancy compared with the winter months confers protection in the early fetus against MS developing in adulthood.

It is assumed from these consistent results that exposure to UVR confers some protection against MS although the mechanism by which this occurs is not yet clear. There is an increasing number of studies indicating that the main consequence of the UVR could be the production of vitamin D (reviewed in reference 59), but there are also some studies to support the view that UV-induced immunosuppression may play a key role. These concepts may be linked as vitamin D is potently immunosuppressive in humans (60). Immunosuppressive mechanisms, additional to vitamin D, could act by down-regulating the activity of the Th1 cells involved in the demyelination through the generation of specific Tregs or UV-Bregs, or the production of immunosuppressive cytokines. Thus, in a multicenter study in Australia, Lucas et al. found higher levels of past, recent and accumulated leisure-time sun exposure were each associated with a reduced risk of first demyelinating events, as was higher actinic skin damage (reflecting mainly chronic UVA exposure) and higher serum vitamin D status (61). The sun exposure and vitamin D status were revealed to be independent risk factors for the first central nervous system demyelination. Mouse models of MS (experimental autoimmune encephalomyelitis—EAE) showing that UV is protective support the human epidemiological evidence. In one such model, chronic UV irradiation was demonstrated to suppress the clinical symptoms of the disease and this occurred without the apparent involvement of vitamin D (62). To try to resolve the current controversy regarding the relative roles of vitamin D and UV-induced immunosuppression in MS protection, mice which lack the vitamin D receptor gene and/or the 1-α-hydroxylase gene (whose product is required to produce the active form of vitamin D) are being studied to assess whether EAE still develops following chronic UVR exposure (63).

Asthma

Asthma comprises a group of diseases which have divergent triggers and pathways but which all present with wheeze, chest tightness or shortness of breath accompanied by airways obstruction and reversible airflow restriction. The level of severity and frequency of symptoms plus the age of onset are all variable, as are the main inflammatory phenotypes, classified on the basis of their immunopathology as eosinophilic, neutrophilic or paucigranulocytic (64). The pathological changes in the airways in eosinophilic asthma are induced by a chronic inflammatory process where the bronchiolar mucosa is infiltrated with lymphocytes and eosinophils and there is increased mucus production and submucosal edema. Following allergic sensitization, an increased production of the Th2 cytokines occurs, especially IL-4, IL-5, IL-6 and IL-13, which induce and regulate IgE production and eosinophilic airways infiltration (65). There is also an increased level of IFN-γ in patients with severe asthma suggesting enhanced Th1 responses. It is likely that both Th1 and Th2 cells contribute to the development of the symptoms of asthma (66). Recently it was shown that naturally occurring Tregs can effectively suppress Th1-mediated airways inflammation (67).

While there is some published information to indicate that the prevalence of asthma correlates negatively with latitude (the closer to the equator the higher the prevalence) and positively with ambient UVR (48), this view has not been substantiated more recently. For example, Krstic (68) found that a 10° change in latitude from the south to the north regions of the Eastern Seaboard of the USA was significantly associated with a 2% increase in adult asthma prevalence. Hughes et al. (69) analyzed data from an Australian multicenter study based on four regions with latitude 27–43°S. They showed initially that there was a 9% decrease in asthma per increasing degree of latitude, but this correlation disappeared after adjusting for the average daily temperature. Furthermore, there was no association between childhood asthma and past exposure to solar UVR, as assessed by personal cumulative actinic damage and self-reported time in the sun. Three other studies were based on hospital admissions: monthly adult asthma hospitalizations in Taiwan negatively correlated with monthly hours of sunshine (70), and monthly childhood asthma hospitalizations in Singapore (71) and Taiwan (72) also correlated negatively with monthly hours of sunshine. In addition there is some evidence from a range of investigations that asthma may be a consequence of vitamin D deficiency, which is most frequently caused by limited exposure to solar UVR (reviewed in reference 73). Finally there are anecdotal accounts that sunshine holidays by the beach or in the mountains are often beneficial for subjects with asthma. Thus, it might be possible that UVR could ameliorate the allergic respiratory response in at least some forms of asthma by altering immune responses.

As far as we are aware, no experimental studies have examined the effect of UV irradiation on human asthma, and few have investigated mouse models of asthma in this context (reviewed in reference 74). McGlade et al. (75) showed that UV irradiation of the skin of mice down-regulated airways hypersensitivity in a model where the animals were sensitized intraperitoneally with ovalbumin and then an asthma-like response to aerosolized ovalbumin measured. A second model using papain administered intranasally also demonstrated that prior UVR exposure of the mice reduced the airways’ inflammatory response when the animals were subsequently challenged with papain (74). The UVR exposure led to the induction of allergen-specific Tregs. However, further work by the same group (76) revealed that such cells were not involved in down-regulating the airways diseases and naturally occurring Tregs were not involved either. Instead a reduction in effector CD4+ T cells in the lymph nodes draining the trachea and airways was found in the irradiated mice compared with the unirradiated mice, and there was a reduction in the ability of these cells to proliferate, probably due to less efficient sensitization.

Thus, there is some preliminary evidence from epidemiological surveys and experimental models in mice to support the hypothesis that UVR exposure could be beneficial in at least some forms of asthma, but more work is required to substantiate this view.

Adverse consequences of UV-induced immunosuppression

Skin cancer

Solar UVR exposure is the major risk factor for all three of the commonest forms of skin cancer—cutaneous malignant melanoma (MM), basal cell carcinoma (BCC) and squamous cell carcinoma (SCC). Examples of each of these are shown in Fig. 3. The UVR exposure causes DNA damage in the skin. There is good evidence that UV-induced immunosuppression is also very important, most clearly shown by the high prevalence of skin cancer in immunosuppressed individuals and from the finding that their tumors arise almost entirely on sun-exposed areas of the body. This picture is particularly apparent for SCCs in organ transplant recipients who are therapeutically immunosuppressed to prevent rejection of the transplant (77). These drugs suppress T cell activity mainly, and so T cells are thought to be the major effector cells in controlling SCCs. Surveys have revealed that the prevalence of SCCs is 65 times higher, of BCCs is 10 times higher and of MMs is 7 times higher in immunosuppressed subjects compared with immunocompetent subjects. UVR exposure can down-regulate the production of the Th1 cytokines, which have been shown to protect against SCCs in experimental mice (78). Many Tregs, induced by UVR exposure, infiltrate SCCs (78) and surround BCCs (79) and are presumably capable of releasing the immunosuppressive cytokine IL-10 locally.

Figure 3.

 Examples of skin cancers: (a) well-differentiated squamous cell carcinoma, diameter about 1 cm, (b) superficial basal cell carcinoma, diameter about 0.8 cm, and (c) superficial spreading melanoma with extensive regression, diameter about 1.3 cm (photographs provided by Dr F. Moloney, Royal Prince Alfred Hospital, Sydney, Australia).

More than 20 years ago, Yoshikawa et al. (80) reported that 92% of individuals with a past history of skin cancer were UV-susceptible (defined as requiring only a low UV dose for local suppression of CHS), while only 35% of individuals without a past history of skin cancer fell into this category. Although the division of people into UV-susceptible and resistant (defined as requiring a higher UV dose for local suppression of CHS) has not been confirmed in more recent studies, it is certainly true for mouse strains and may represent a factor in susceptibility of some humans to skin cancer development. It is known that people with fair skin who sunburn readily and do not tan or tan only with difficulty (phototype I/II) have a higher risk of skin cancer than those with darker skins who burn rarely and tan easily (phototype III/IV). An explanation for this was provided by Kelly et al. (81) who found that, for a particular dose of solar UVR exposure, the subjects in the former group were more susceptible to local suppression of CHS than those in the latter group. This difference was most marked when suberythemal doses of UVR exposure were used. A different approach was taken by Grimbaldeston et al. who noted that a high density of mast cells in MMs and SCCs had been associated with poor prognostic outcomes, and they established a correlation between the density of dermal mast cells and skin cancer: the higher the number, the higher the risk of BCC and MM (82). Mast cells are contributors to immunosuppression by producing a range of immune mediators such as TNF-α and histamine, on activation by UVR exposure, and it has been shown that migration of mast cells from the skin to the draining lymph nodes is an early step in UV-induced immunosuppression (30).

Experiments in mice have confirmed the contribution of UV-induced immunosuppression to skin cancer development, starting with the initial studies carried out in the 1970s by Kripke and colleagues (1,2). More recently it has been shown that photocarcinogenesis in mice can be inhibited if particular steps in the pathway depicted in Fig. 1 are blocked. For example, treating mice with a cis-UCA monoclonal antibody (83), a PAF receptor antagonist (84) or a serotonin receptor antagonist (84) during chronic UVR significantly reduced the number of SCCs that developed. An interesting report, in which a novel transgenic mouse strain was used, has indicated that, following UVR of neonates, macrophages infiltrate the skin and produce IFN-γ (85). This cytokine then has the effect of activating the melanocytes to proliferate and migrate into the epidermis, thus indicating a key role for UV-induced IFN-γ in melanomagenesis. Finally exposing the skin of neonatal mice to UVR results in the generation of Tregs which persist into adulthood. Muller et al. have suggested that these cells could represent a predisposing factor for the subsequent development of MM (86).

If the results from the human and mouse studies are assembled together, the evidence that UV-induced immunosuppression plays a major role in skin cancer is overwhelming.

Cutaneous lupus erythematosus

Lupus erythematosus (LE) is an inflammatory autoimmune disorder with a broad range of cutaneous and systemic symptoms but with photosensitivity being one of the diagnostic criteria. The B cells are hyperactive, producing abundant quantities of autoantibodies to nuclear antigens, cell surface molecules and serum proteins; these can form immune complexes that account for many of the symptoms of the disease. In addition the T cells overproduce proinflammatory cytokines or increase cell-to-cell adhesion leading to apoptosis of target cells. The cutaneous lesions are found mainly on sun-exposed body sites and UVR is thought to be involved in their initiation and progression (87). They can appear as scaly red plaques which persist for months before resolving with or without scarring, or as a more diffuse photosensitivity. They can be provoked by UV wavelengths between 280 and 340 nm (88).

The immune effects of UV irradiation in the skin of cutaneous LE patients are different from those occurring in healthy subjects, although most effort to date has concentrated on determining the systemic immune abnormalities that account for the various manifestations of LE (reviewed in reference 89). However, it is likely that local changes, confined to the skin, may be more important following sun exposure. UV irradiation can induce DNA and cell damage in the epidermis, generating apoptotic keratinocytes. This results in the clustering of autoantigens at their cell surface (90), and there is evidence of delayed or altered clearance of these apoptotic cells in cutaneous LE (91). Recent work has demonstrated that this could be the result of impaired macrophage function due to the presence of autoantibodies with specificity for the class A scavenger receptors expressed on macrophages (92). The accumulation of apoptotic cells leads to the release of proinflammatory cytokines locally, such as TNF-α and the subsequent development of the inflammatory lesions (93). The activities of both Th1 and B cells are increased in the lesions. TNF-α is known to inhibit the function of Tregs. In addition the number of Tregs in the cutaneous LE lesions is reduced compared with lesions from patients with other chronic inflammatory diseases (94), indicating a possible organ-specific abnormality of Tregs in the skin. Furthermore, it has been demonstrated recently that there are twice the number of mast cells in the skin lesions of LE compared with normal skin and more in lesions in sun-exposed sites than in sun-protected sites (95). The accumulation coincided with abundant expression by epidermal keratinocytes of IL-15 and chemokine(C-C motif) ligand 5, both of which are recognized mast cell chemoattractants. What limits the alteration in tolerance to the local irradiated site in the skin of the subjects with cutaneous LE is not known at present.

Dermatomyositis (DM) is another antibody-specific autoimmune disease with photosensitive skin rashes. It may fall into a similar category as cutaneous LE with regard to UVR exposure being harmful as the prevalence of DM has been reported to increase with decreasing latitude in Europe (96) and the incidence correlated with surface UVR in one study (97). Interestingly in Punta Arenas, the southernmost city in Chile, the number of cases of DM was six times higher than average in the 2 years when increased terrestrial UVB occurred during the spring and summer months due to the Antarctic ozone hole (J. Abarca, personal communication). There is preliminary evidence suggesting that UVR exposure is a risk factor for DM development by immune targeting of the expression of the autoantibodies (98).

Infectious diseases and vaccination

UVR exposure is known to affect antigen presentation and to down-regulate Th1 responses, both of which are required for the immune control of intracellular infections in particular. At least 20 models of infectious diseases in rodents have been described in which, in almost all cases, UV irradiation before or after infection causes a significant suppression of T cell activity (reviewed in references 99 and 100). This can result in increased microbial load, increased severity of symptoms and even death, on occasion. The organisms range from ones causing skin infections to ones causing systemic infections and include examples of viruses, bacteria, fungi, protozoa and nematodes. These results lead to the possibility that exposure of humans to UVR could make a higher proportion of infections symptomatic, increase the severity of symptoms, increase the oncogenic potential of a microorganism and/or reduce the resistance to reinfection. However, such outcomes are not generally apparent and there are only a few instances where human infections are recognized to be altered by solar UV irradiation. These are listed in Table 2, together with a summary of how the exposure is thought to affect the immune response to the organism, where this is known. Further details can be found in Norval (101).

Table 2.   Human infections altered by UVR.
OrganismClinical outcomeEffect of UVR on immune response
Herpes simplex virusReactivation of latent virus in gangliaReduced antigen presentation in skin
Recrudescent cold sores on UV-exposed body sites (Fig. 4)Reduced systemic Th1 cytokine responses
Production of IL-4 and IL-10 in irradiated site
Human papillomavirusConversion of skin warts to squamous cell carcinomas on UV-exposed body sitesViral immune evasion mechanisms
Reduced antigen presentation in irradiation site
Cytokine imbalance
Loss of Langerhans cells in irradiated site
Varicella zoster virusReactivation of latent virus in gangliaNot known
Shingles more frequent in summer and on UV-exposed body sites
PolyomavirusMerkel cell carcinoma on UV-exposed body sitesNot known
Mycobacterium lepraeLesions most frequent on UV-exposed body sitesNot known
Leishmania donovaniPost-kala-azar lesions on UV-exposed body sitesNot known
Figure 4.

 Subject with recrudescent herpes simplex virus lesions.

The reason for relatively few human infections being modulated by UVR exposure compared with the animal models might lie in differences between human and rodent species in gene regulation and innate defenses particularly the generation of antimicrobial peptides in the skin, and in the size and site of the infecting dose. In addition there could be inadequate study of human infections as most epidemiological surveys of disease prevalence do not record solar UVR or personal UV exposure, and weather conditions affect the transmission and survival of many microbes, which could confound the results. It is also possible that the UVR needs to affect some property of the microorganism itself, perhaps relating to immune evasion or reactivation, as well as down-regulating immune responses, before a clinical effect is apparent. This is thought to be the case for herpes simplex virus and human papillomaviruses, which are the most studied of the microorganisms affected by UVR exposure. Although the number of human infections altered by UVR exposure is limited, it is clear that it can suppress memory (recall) immune response. This has been shown, for example, for the tuberculin purified protein derivative (Mantoux reaction) (102), lepromin and the standard multitest antigens (103). The last result is perhaps of most interest as the suppression occurred following real-life exposure of the volunteers to sunlight, and was apparent if the antigens were applied to either irradiated or unirradiated body sites (103). Whether such a down-regulation might lead to an increased risk of a symptomatic reinfection is not known.

Perhaps of most interest in the area of human infections and UVR exposure is the question of whether sun exposure can adversely affect the immune response to vaccination. At least four animal models have indicated that UVR exposure has the capacity to alter the efficacy of vaccination such that, on challenge, the resistance/memory response generated by the vaccine is significantly reduced. Few studies have addressed this very important issue in terms of human vaccination and only limited data are available to date with scarcely any results relating to protection against reinfection.

It was first noted many years ago that the antibody response to poliovirus was higher if the oral vaccine was administered in temperate compared with tropical zones (104). Similarly the antibody response to the same vaccine was higher if it was given in the winter compared with the summer (105). The same was reported for live influenza virus vaccine (106), hepatitis B surface antigen vaccine (107) and live rubella virus vaccine (108). In the only meta-analysis regarding the efficacy of a vaccine to prevent subsequent disease that also considered location, Colditz et al. showed that the bacille Calmette-Guerin (BCG) vaccine had a latitude gradient in protection so that the ability to prevent subsequent tuberculosis increased with increasing distance from the equator, and thus with decreasing solar UVR exposure (109). Many other factors in addition to solar UV irradiation change with latitude such as clothing, diet, living conditions, temperature and daylight hours and no assessment of T cell responses to the BCG vaccine has been carried out in people vaccinated at different times of the year or in different locations. However, the Colditz et al. study does correlate with the suppression in the Mantoux reaction induced experimentally by exposure of subjects to suberythemal solar-simulated radiation (102).

Only one experimental trial involving human subjects, UVR and vaccination has been undertaken. Here, one group were UV irradiated before vaccination with hepatitis B surface antigen and the subsequent immune responses compared with another group not irradiated before vaccination (110). While the natural killer and CHS responses were suppressed in the irradiated subjects, no differences between the two groups were found in either the hepatitis B-specific T cell or antibody responses. It should be noted that the vaccine contains an alum adjuvant which will promote Th2 cytokine responses predominantly, and is administered at high dose to induce immunity in poor responders; both of these factors may mitigate against a UV-induced effect. However, a subset of the irradiated subjects with a minor IL-1β polymorphism did have suppressed hepatitis B antibody responses (111) and those irradiated individuals with high cutaneous cis-UCA levels had suppressed hepatitis B T cell responses (112). Therefore, individuals may vary in their response to vaccination and UVR exposure depending on genetic and other factors.

It is concluded that the immune response to vaccines or the resistance to reinfection following vaccination may be affected by UVR exposure but the evidence thus far is certainly not definitive. Further work is required to investigate other commonly used vaccines, especially those that promote predominantly Th1 responses such as measles, mumps, rubella, varicella, BCG and Salmonella typhi. It will be important to determine if vaccination in the winter months in temperate countries provides better protection than vaccination in the summer months and if it is advisable to vaccinate someone about to go on or return from a sunshine holiday or who is obviously sunburnt.

Public health advice

Sun exposure

From the above sections summarizing the potential positive and negative aspects of UV-induced immunosuppression, it can be seen that it is important to maintain a balance regarding personal sun exposure. The main difficulty here is to provide useful guidance for the general public, especially as the optimal sun exposure for one individual may be quite different from the optimal for another individual. Personal differences will depend on factors including skin phototype, tendency to freckle, age, obesity, diet, clothing and probably other conditions that are unknown as yet. In addition, while the UVB component of sunlight has more biological activity than UVA for inducing erythema, UVA makes a substantial contribution toward sunlight-induced immunosuppression in humans. However, limited exposure at times of the day and year and in locations where solar UVB is particularly strong needs to be recommended. With regard to skin cancer risk, it is most critical to avoid high levels of sun exposure, particularly during childhood. Such exposure may be more likely on moving to a location where the UV Index is higher than at the usual place of residence, as might be experienced during a sunshine holiday. At the same time, it is important to maintain a sufficient level of vitamin D for promoting its possible multiple health benefits. At present, many populations in the world are thought to have insufficient levels of vitamin D, perhaps due to changes in lifestyles, diet, leisure activities and occupation. Thus, public health campaigns need to reflect both the beneficial and adverse consequences of sun exposure. For example, the SunSmart programme in the UK (http://www.sunsmart.org.uk/advice-and-prevention/) stresses the importance of avoiding sunburn and emphasizes that the amount of sun exposure to ensure production of sufficient vitamin D is less than the amount that causes sunburn. In Australia and New Zealand, the SunSmart UV Alert (http://www.bom.gov.au/announcements/uv/) is used only if the UV Index is 3 or higher when sun protection is recommended, and states that there is nothing healthy about a tan.

Personal protection

The obvious and most effective way to avoid the sun is to stay indoors as most sunlight is blocked. However, while window glass absorbs some UVB, it transmits most UVA (113). As UVA alters immune responses, sunlight filtered through window glass needs to be taken into consideration. Shade is also useful although diffuse and scattered UVB may be a problem and different structures offer different degrees of protection, varying from dense foliage being very good to a beach umbrella being poor. Clothing can be a reliable method for covered areas of the body but not all textiles are effective. Some are now assessed by in vitro transmission to give a UV protection factor but this might not accurately predict how the garment responds to wetness, stretching, washing, ironing and humidity (114). Hats can also be used which are very good provided the brim is wide enough to provide shade for all of the face and neck. Sunglasses are the most practical and efficient way to protect the eyes and should be wrap-around or with side-shields. UV-blocking contact lenses protect the cornea, lens and aqueous humor, but not other parts of the eye and should not be used as a substitute for sunglasses.

Sunscreens were designed initially to lessen the risk of sunburn and currently they provide a range of levels of protection up to a sun protection factor (SPF) of 50. SPF is a measure of protection from sunburn, which is caused mostly by UVB, although UVA does contribute. Therefore, to achieve high SPFs, sunscreens need to contain some protection from UVA. Modern sunscreens usually provide this, and a sunscreen rated as broad-spectrum protects against both UVB and UVA. In addition to sunburn, sunscreens offer a degree of defense against other acute effects of solar UVR, such as sunburn cell formation in the epidermis, cutaneous DNA damage (115) and recrudescence of herpes simplex cold cores (116). With regard to chronic effects, the regular use of sunscreens reduces the incidence of actinic keratoses and SCCs (117), attenuates the development of new nevi in children, a recognized risk factor for MM and reduces the risk of cutaneous MM (118). Photoageing, also caused by chronic sun exposure, can be reduced by regular sunscreen use (119).

Sunscreens protect the immune system from UVR. As far as we are aware, Bestak et al. (120) were the first to accurately estimate the immune protection factors (IPFs) of sunscreens. IPFs have a direct correlation with the UVA protection factor (121), emphasizing the effect of UVA on immunity in humans and stressing the importance of using broad-spectrum sunscreens. While some sunscreens have a higher IPF than the SPF, others are relatively low (121,122). It needs to be stressed that the immune system is suppressed by lower doses of UVR than those required to cause sunburn.

Personal protection from sunlight is complex. We need to be protected from both UVB and UVA, but the “best exposure” to maintain the health benefits while avoiding the damaging aspects is unknown. Optimally, preservation of the ability of the skin immune system to destroy developing skin tumors and to control infectious agents is desirable, but without inhibiting the production of vitamin D unduly. Achieving such a balance by use of various protective strategies may not be feasible as there are so many environmental and personal variables to consider, in addition to probable variations in the doses required to produce the range of effects.

Chemoprevention is receiving considerable research attention currently. This functions beyond UV absorbance or avoidance and could be provided as skin creams or taken orally in the diet. UVR exposure suppresses immunity in the skin and internal organs via different mechanisms. While UV-induced cis-UCA and oxidative signaling both suppress skin immunity, neither suppresses immunity in internal organs (123). This suggests that it may be possible to develop chemopreventative strategies that stop UVR from suppressing immunity to skin cancers or cutaneous infections, without reducing the inhibitory effects on immune-mediated diseases at sites other than the skin. The inhibition of reactive oxygen species may be useful in this regard. Another promising strategy is to prevent UVR exposure from causing an energy crisis in the skin (26). This can be done with oral or topical nicotinamide (vitamin B3). The nicotinamide is metabolized to NAD+, an essential coenzyme in ATP production, thus preventing the UV-induced depletion of keratinocyte ATP (26) and allowing the immune system to function normally. Topical nicotinamide protects humans from the immunosuppression induced by UVB and UVA irradiation (28), and both topical (25) and oral (27) nicotinamide from the immunosuppression induced by solar-simulated UVR exposure. Oral lycopene, the antioxidant found in tomato paste, reduces damage to mitochondrial DNA and the erythemal response in humans (124), although it has not yet been tested for effects on the immune system. Other dietary compounds including goji berry juice (125) and isoflavonoids have been shown to protect the immune system in mice (126) and humans (127).

Conclusions

Sunlight exposure has both positive and negative outcomes for human health. The International Commission on Illumination has recently considered whether a recommendation can be made regarding a lower limit of UVR that is beneficial, while minimizing any adverse effects (TC 6-58, http://div6.cie.co.at/?i_ca_id=609&pubid=327). However, due to the lack of information and research into the dose effects of UVR on human health, accurate advice that balances these two aspects cannot be given at present. As outlined in this review, UVR exposure affects the local and systemic immune system in many ways. Both UVB and UVA are immunosuppressive in humans but with various dose responses and probably interactive effects between the wavebands. It is likely that different UVR doses regulate the immune system by dissimilar mechanisms. UV-induced immunosuppression is a critical event that makes humans vulnerable to all types of skin cancer. Particularly in countries with high levels of sun exposure such as Australia, this is the most important health consequence of sunlight exposure. UVR exposure also impairs the immune responses to infections and vaccines, although the relevance of this to human health is unclear at present. There is also substantial evidence that suppression of immunity by UVR exposure is beneficial by reducing the risk of some autoimmune diseases such as MS, photoallergic diseases such as PLE, and asthma. It is likely that different mechanisms of UV-induced immune modulation are involved in each of these diseases, or that different UV doses are effective. Progress, particularly in chemoprevention, may lead to personal protective measures that could enable the deleterious effects of UV irradiation to be avoided while preserving the beneficial effects. Dietary biological modifiers including vitamins and antioxidants show promise in this regard, and are the focus of considerable current research. At present, however, the best advice for the general public is to ensure moderate solar exposure, with sensible sun avoidance practices that include the use of sunscreens at times of high solar UV intensity.

Acknowledgments

Acknowledgements—  The authors declare no disclosures of interest. GMH wishes to acknowledge financial support from the National Health and Medical Research Council of Australia, the New South Wales Cancer Council and Epiderm.

Appendix

Author biographies

inline image

Mary Norval, PhD, DSc, is Professor Emeritus at the University of Edinburgh in Scotland. She has major research interests in the effects of UVR on human health, especially immunological aspects.

inline image

Gary Halliday, PhD, DSc, is a Professor of Dermatology at the University of Sydney in Australia. He has made large contributions toward understanding the role of sunlight in skin carcinogenesis, particularly the suppression of immunity induced by UVR, induction of gene mutations, the cell biology of skin cancer and development of effective photoprotection.

Ancillary

Advertisement