Exposure of humans [1,2] and rodents [3-7] to ultraviolet

Transcription

1 nalysis of the Protective Effect of Different Sunscreens on Ultraviolet Radiation-Induced Local and Systemic Suppression of Contact Hypersensitivity and Inflammatory Responses in Mice Peter Wolf, Cherrie K. Donaho, and Margaret L. Kripke Department of Immunology, The University of Texas M.D. nderson Cancer Center, Houston, Texas, U.S.. We investigated the capacity of three sunscreen compounds to protect mice from the inflammatory and immunosuppressive effects of ultraviolet radiation (UVR). The sunscreen preparations contained 7.5% -ethylhexyl-p-methoxycinnamate, % octyl-n-dimethyl-p-aminobenzoate, or 6% benzophenone-3 in an oil-in-ater emulsion. Skin selling as used as the measure of their effect on UVR-induced inflammation, and immunosuppression as assessed by contact sensitization ith,4-dinitrofluorobenzene applied to UV -irradiated skin (local suppression) or a distant site (systemic suppression). The sunscreens ere applied to the shaved dorsal skin of C3H mice, hich ere then given a single dose of UVR ranging from to 3 kj/m ithin the UV (- 3 nm) region. ll three sunscreens gave complete protec- tion against local suppression of contact hypersensitivity caused by a dose of kj/m UV. They also protected against both inflammation and systemic immunosuppression caused by UVR; hoever, protection as highly dependent on the UVR dose. Furthermore, the sunscreens ere less effective in protecting against systemic immunosuppression than against inflammation. These results indicate that immunosuppression is less sensitive to the protective effects of the sunscreens than inflammation and that protection against UVR-induced inflammation does not necessarily imply prevention of immunologic alterations. In addition, these studies suggest that UVR-induced immunosuppression and inflammation may involve different mechanisms. ] Invest Dermatol 1:54-59, 1993 Exposure of humans [1,] and rodents [3-7] to ultraviolet radiation (UVR) decreases their contact hyper~e~sitivity (CHS) response to haptens. The ImmunologIC Impairment caused by UVR can be divided into local and systemic effects [3]. Local immunosuppression is defined as the diminished CHS response observed hen haptens are applied through UV-irradiated skin. Exposure to UVR can also result in a diminished CHS response hen haptens are applied at a distant, non - UV -irradiated site; this is referred to as systemic immunosuppression. In experimental animals, the UVR-induced unresponsiveness to haptens is associated ith the presence of hapten-specific T -suppressor cells [7-9]. The mechanisms by hich UVR activates the suppressor rather than the effector arm of the immune response are not completely understood; hoever, an alteration in the activity of epidermal Langerhans cells has been implicated in local immu- Manuscript received September 17, 199; accepted for publication November,199. Reprint requests to: Dr. Margaret L. Kripke, Department of Immunology-17, The University of Texas M.D. nderson Cancer Center, 1515 Holcombe oulevard, Houston, TX 773. bbreviations: P-3: benzophenone-3 CHS: contact hypersensitivity DNF:,4-dinitrofluorobenzene -EHMC: -ethylhexyl-p-methoxycinnamate NIH: National Institutes of Health P: p-aminobenzoic acid : octyl-n-dimethyl-p-aminobenzoate SPF: sun-protection factor UVR: ultraviolet radiation nosuppression [5,1,11]. The production of soluble factors may be involved in both local [11-13] and systemic immunosuppression [6,14-16]. Recently, strong evidence as presented indicating that on the molecular level, DN damage in the form of pyrimidine dimers is the initiating event for local [17] and systemic [17,1] immunosuppressive effects of UVR. UVR-induced immunosuppression plays a crucial role in the development of skin cancers in mice. Exposure of mice to UVR suppresses their ability to reject highly antigenic, UVR-induced tumors l19,]. The mechanisms responsible for this immunosuppressive effect are similar to those involved in suppression of CHS [7-9]. UVR is the major carcinogenic agent in the etiology of human skin cancers, at least for basal and squamous cell carcinomas [1,]. There is increasing evidence no that UVR may also contribute to the development of skin cancers in humans via its effects on the immune system [3-5]. Sunscreens are highly protective against sunburn in humans [6] and protect laboratory animals from chronic UVR-induced skin damage [7,]' carcinogenesis [9], and co-carcinogenesis [3]; hoever, a controversy has arisen hether sunscreens can protect against the immunosuppressive effects of UVR. In a number of studies, different sunscreens did not protect rodents from a variety of UVR-induced effects on the immune system, such as local [31,3] and systemic [33] suppression of CHS and induction of susceptibility to tumor transplantation [34]. Moreover, sunscreens lacked the capacity to protect against suppression of delayed-type hypersensitivity, natural killer-cell activity, and immunoglobulin production, according to Hersey et af [35]. and suppression of mixed lymphocyte reaction, according to Van Prag et af [36], by UVR in human subjects. -Xj93j$6. Copyright 1993 by The Society for Investigative Dermatology, Inc. 54

2 VOL. 1, NO.3 MRCH 1993 SUNSCREENS ND IMMUNOSUPPRESSION 55 The aim of the present study as to clarify hether three commonly used sunscreen compounds, namely, -ethylhexyl-p-methoxycinnamate (-EHMC), octyl-n-dimethyl-p-aminobenzoate (- P) (both UV absorbers), and benzophenone-3 (P-3) (UV + UV absorber), protect against immunosuppression induced by UVR and to compare immunoprotective capacity ith the capacity of the sunscreens to protect against UVR-induced inflammation. For this purpose, the photoprotective effects of the different sunscreen compounds ere studied in a ell-characterized model of UVR-induced local and systemic suppression of CHS using the hapten,4-dinitrofluorobenzene (DNF) and C3H mice. The capacity of the sunscreens to protect against system.ic suppression of CHS as evaluated In UVR dose-response studies across a broad range of UVR doses. MTERILS ND METHODS nimals Specific-pathogen-free female C3H/HeN(MTV-) mice ere supplied by the National Cancer Institute-Frederick Cancer Research Facility nimal Production rea (Frederick, MD) or Charles River reeding Laboratories (Wilmington, M). The mice ere housed in a pathogen-free barrier facility accredited by the merican ssociation for ccreditation of Laboratory nimal Care, in accordance ith current US Department of griculture, Department of Health and Human Services, and National Institutes of Health (NIH) regulations and standards. ll animal procedures ere approved by the Institutional nimal Care and Use Committee. The mice ere given free access to NIH formula 31 mouse food and sterilized ater. mbient lighting as controlled to provide a regular 1-h light and 1-h dark cycle; -16-eek-old, agematched mice ere used in the experiments. Sunscreens The sunscreen preparations, kindly provided by Estee Lauder Co. (Melville, NY), contained 7.5% -EHMC, % (both UV absorbers), or 6% P-3 (UV + UV absorber) in an oil-in-ater emulsion. ccording to the manufacturers, the -EHMC- and -containing sunscreens blocked 75% ofuv (-3 nm) radiation, the P-3-containing sunscreen blocked 45%, and the vehicle blocked 16%, as measured in vitro by back-transmittance hen the sunscreen preparations ere applied at mg/ cm Their sun-protection factor (SPF) as rated by the manufacturers as approximately 4-6 for both UV sunscreens and beteen and 4 for the P-3 - containing preparation. Determination of Transmission Spectra for Sunscreen Preparations The transmission spectra of the sunscreen preparations ere measured in our laboratory by spreading the preparations in a thin layer over a quartz cuvette and measuring transmission using an LK iochrom Ultrospec Model 45 spectrophotometer (LK iochrom, Cambridge, UK). UV Irradiation UVR as provided by a bank of six FS4 sunlamps (National iology Corp., Tinsburg, OH), hich have a peak emission at 313 nm and deliver 65% of their total energy ithin the UV (-3 nm)-avelength range; their UV irradiance as 5 W /m at a -cm distance, as determined by an IL 7 radiometer equipped ith an SEE 4 detector fitted ith an SES filter and a quartz diffuser (International Light, Inc., Neburyport, M). One day after the dorsal hair of the mice had been removed ith electric clippers, they ere exposed to the radiation source hile housed five per cage in individual compartments. ecause of shielding by the cage lids, the final irradiance received by the animals as approximately 3 W /m The sunscreen preparations ere liberally applied (approximately J.Ll/mouse) and rubbed on the shaved dorsal surface and tail of the mice min before exposure to UVR. Measurement of Inflammatory Response The inflammatory response as determined by measuring the double skin-fold thickness of the dorsal skin ith a spring-loaded micrometer (Mitutoyo, Tokyo, Japan) prior to and 4 and 4 h after UVR exposure. Skin selling as determined by subtracting the average skin thickness before UV irradiation from that after UV irradiation. Prior to this study e found that the application of the sunscreens or the vehicle alone resulted in a small but significant skin selling. In four experiments (n = 5 in each experimental group), the application of - P, -EHMC, P-3, and the vehicle alone gave a mean skin selling (±SEM) (X 1- mm) of7.5 ±.7, 5.6 ± 1.5,5.4 ±., and.1 ± 1.1 at 4 h, and 5. ± 1.1, 3. ± 1.5, 3.9 ± 1., and 3.4 ±. at 4 h, respectively. These values ere subtracted from the skin selling of sunscreen-treated, UV -irradiated groups to determine net skin selling. The protection from inflammation by a sunscreen as defined as complete protection hen there as no significant skin selling in a sunscreen-treated, UV -irradiated group 4 and/or 4 h after UVR exposure hen compared ith sunscreen-treated, unirradiated mice; the protection as defined as partial protection hen there as significant skin selling, but the selling as significantly different from that of the UV-irradiated group ithout sunscreen. Local and Systemic Suppression of CHS Groups of mice, either untreated or treated ith a sunscreen or the verucle, ere exposed to the UVR source. During irradiation, the ears of the mice ere protected from UVR by opaque tape, hich as removed after exposure. Three days after exposure to UVR, mice ere sensitized ith 5 J.Ll of.3% DNF (Sigma Chemical Co., St. Louis, MO) in acetone on shaved, irradiated dorsal skin (assay for local suppression of CHS) or on shaved, unirradiated ventral skin (assay for systemic suppression of CHS). Five days later, the CHS response as elicited by applying 5 J.Ll of.% DNF in acetone to the ventral and dorsal surfaces of both ears. Ear thickness as measured ith the micrometer before and 4 h after challenge. The percent suppression of CHS as determined by the folloing formula [1 - ( - )/(C )] X 1, here the letters represent the average ear selling in () sensitized and irradiated mice, () unsensitized and unirradiated mice, and (C) sensitized and unirradiated mice. The percent protection by a sunscreen as determined by the folloing formula [1 - (+ SS/-SS)] X 1, here + SS represents the percent suppression of CHS in sunscreen-treated, UV -irradiated mice, and - SS is the percent suppression in UV -irradiated mice ithout sunscreen. The protection by a sunscreen from suppression of CHS as defined as complete hen there as no significant difference in the CHS response of the sunscreen-treated and UV -irradiated group from the positive control group (sensitized and unirradiated group); the protection as defined as partial hen the CHS response of the sunscreen-treated, UV -irradiated group as intermediate to the responses of the sunscreen-untreated, irradiated group and the positive control group. Statistics The statistical significance of the differences in skin or ear selling among groups as evaluated using the Student totailed t test. Each group contained at least five mice. RESULTS Sunscreen Transmission Spectra The transmission spectra of the to UV sunscreens, -EHMC and o-p, shoed very similar absorption values from 6 to 4 nm and distinct absorption peaks near 3 nm (Fig l). The P-3 - containing preparation absorbed approximately half the amount of UV radiation as the UV sunscreens; hoever, its absorption capacity extended across the UV range. Overlap spectra beteen the emission spectrum of the FS4 sunlamp, as measured by an Optronic Model 7 4 spectroradiometer (Optronic Laboratories, Inc., Orlando, FL), and the transmission spectra of the sunscreen preparations ere calculated by multiplication of the relative irradiance values of the FS4 sunlamp spectrum by the relative transmission values of the sunscreen preparations at the different avelengths (Fig I). ssessment of Minimal Inflammatory Dose Prior to this study, e found that measurement of skin selling as a more sensitive indicator of the minimal inflammatory response in C3H mice than visual evaluation of erythema, so e used skin selling as our measure. Groups of mice ere UV irradi.ated at 4% dose increments (Fig ). The minimal inflammatory dose, taken as the ex-

3 56 WOLF ET L THE JOURNL OF INVESTIGTIVE DERMTOLOGY.36 Z 1.. iii <Jl :iii.6 <Jl Z a: t-.4 > t=..j a: EHMC P-3 Yehlcle 36 4 '".5 ~.7..., ~ (f) 1.4 C al > ::> SKIN SWELLING (1 mm) f!l 4 h rzi 4 h Figure _ UVR dose - response for induction of inflammation. Groups of mice ere UV irradiated at 4% dose increments. The minimal inflammatory dose as 1 kj/m UV radiation, as determined by dorsal skin selling at 4 and 4 h after UVR exposure; n = 5; data are mean ± 5EM;.p <.5 versus unirradiated mice Z ~.6 a:!!: >.4 t=..j W a: EHMC + P-3 4 ere. not treated ith sunscreen (Fig 3). The application of the o-p - containing sunscreen completely protected against suppression of the CHS response at kj/m and partially protected at kj/m. No protection as given by the sunscreen at 3 kj/m. Comparison of the photoprotective capacity for inflammation versus immunosuppression at the different UVR doses shos that o-p as more protective against the inflammatory response than against systemic immunosuppression; similar results ere seen ith -EHMC. n experiment ith the P-3-containing sunscreen is shon in Figure 4. The P-3 preparation as not as protective against inflammation as the to UV sunscreens; hoever, it completely protected against skin selling at and 4 kj/m and partially at kj/m (Fig 4). The P-3 preparation completely protected against suppression of CHS at kj/m and partially at 4 and kj/m (Fig 4). summary of all experiments performed is presented in Table I. The protective effects of the sunscreen preparations against inflam- WVELENGTH (nm) Figure L () UVR transmission spectra. The spectra ere determined by measuring transmission of the sunscreen preparations using a spectrophotometer. () Overlap beteen the emission spectrum of the F54 sunlamp and the transmission spectra of the sunscreen preparations from. posure required to elicit minimal significant skin selling, as 1 kj/m. Effect of Sunscreens on UVR-Induced Systemic Suppression of Contact Sensitization and Inflammation To assess the photoprotective capacity of the sunscreen preparations ith respect to the inflammatory response and systemic suppression of CHS, mice ere either treated or untreated ith a sunscreen or the vehicle; they then ere exposed to a UV dose ranging from to 3 kj/m ( min of exposure). To rule out the possibility that the application of a sunscreen or the vehicle per se might influence contact sensitization, experiments ere initially performed ith all the necessary control groups. The application of sunscre,en or vehicle alone had no significant effect on the CHS response in unirradiated mice or on the background ear-selling reaction of unsensitized mice (data not shon). typical experiment ith the -containing sunscreen at three different UVR doses is shon in Figure 3. significant inflammatory response, as measured by skin selling at 4 h, as observed in all groups exposed to UVR but not treated ith sunscreen (Fig 3). The topical application of the O-P - containing sunscreen completely protected against skin selling at and kj/m and partially protected at 3 kj/m. Significant suppression of the CHS response as found in all UVR-exposed groups that Uli 3 3 Sens UY o- P SKIN SWELLING (1 - mm) () 7 (9) () ER SWELLING (1 mm) Figure 3_ The effect of on UVR-induced inflammation and systemic suppression of CH5. Groups of mice ere exposed to different doses of UVR (kj/m), either unprotected or protected by the sunscreen (55); n = 5; data are mean ± SEM. () Inflammatory response to UVR, as measured by dorsal skin selling 4 h after UVR exposure;.p <.5 and +p <.5 versus -untreated groups; () CHS response to DNF, expressed as ear selling 4 h after challenge. 5ens, sensitized;.p <.1 versus positive control group; +p <.5 versus positive control group and -untreated, kj/m UV-irradiated group; values ill pari/'eses represent percent protection by from suppression of CH5.

4 VOL. 1, NO.3 MRCH 1993 SUNSCREENS ND IMMUNOSUPPRESSION 57 Seas UIl 4 4 VY SS P -3 P-3 P -3 SS P-3 P -3 P SKIN SWELLING (1. mm) ER SWELLING (1. mm) Figure 4. The effect of P-3 on UVR-in~uced inflammation a~d systemic suppression of CHS. Groups of ITIlce ere exposed to different doses of UVR (kj/m), either unprotected or protected by the sunscreen (SS); n = 5-1; data are mean ± SEM. () Inflammatory response to UVR, as measured by dorsal skin selling 4 h after UVR exposure;.p <.5 and +p <.5 versus P-3-untreated groups; () CHS response to DNF, expressed as ear selling 4 h after challenge; Sens, sensitized;.p <.1 versus positive control group; +p <.1 versus positive control group and P-3 -untreated, UV-irradiated groups; values in parentheses represent percent protection by P-3 from suppression of CHS. mation and systemic immunosuppression are shon for each sunscreen preparation at the different UVR doses. The UV ~unscreens containing -EHMC or a-p completely protected mice agamst the inflammatory response up to kj/m UV, and the P-3- containing sunscreen completely protected up to 4 kj/m ; hoe~er, the vehicle did not protect at any dose (Table I). ecause the mml- Table I. Protective Effect of Sunscreens gainst UVR Induced Inflammation and Systemic Immunosuppression (kj/m) UV Treatment" Dose IRb CHS' () +,+ +,- 4 +,+ +,+ +,+, + +,±,±,-,- 16 ± 3 ±,±, -EHMC (7) +,+ +,- 4 +,+ +,- +,+,± +,+,±,± 16 ±,±, 3 ± P-3 (6) +,+ +,± 4 +,+ ±,± ±,± ±,-,- 16 Vehicle (4) 4 16 Numbers in parentheses are number of independent experiments performed ith each treatment agent. I Protection from inflammatory response (IR), as measured by skin selling., Protection from suppression of (CHS). (+) Complete protection, (±) partial protection, (-) no protection (as defined in Materials alld Methods); each symbol represents the result of an individual treatment group containing five mice from an independent determination. mal inflammatory dose in our mice as 1 kjjm (Fig ), both UV sunscreens gave an anti-inflammatory SPF of at least, and the P-3-containing sunscreen gave an SPF of at least 4 in this study. We noted considerable variability in the amount of immunoprotection observed ith all sunscreen preparations, particularly at the loer doses of UVR; hoever, the sunscreens gave better protection at the loer UVR doses than at higher doses in nearly all experiments. In certain experiments, both UV sunscreens completely protected against systemic suppression of CHS up to kj/m UV; the P-3-containing sunscreen completely protected at kjjm and partially at 4 and kjjm UV (Table I). The minimal UV dose required to elicit systemic suppression of CHS in our mice as kjjm (data not shon); therefore, the maximal achievable immunoprotective SPF as 4 for both UV sunscreens and less than for the P-3 - containing preparation. Effect of Sunscreens on UVR-Induced Local Suppression of Contact Sensitization fter e had evaluated the capacity of the sunscreens to protect mice against systemic suppression of CHS, e determined their effects on UVR-induced local suppression of CHS. To assess the immunoprotective capacity of the sunscreen preparations, mice ere exposed to kj/m UV, a dose previously found in our laboratory to significantly reduce CHS to haptens applied 3 d later to the UV-irradiated site. To test the possibility that the topical application of a sunscreen or the vehicle per se might interfere ith local contact sensitization, groups of unirradiated mice ere treated ith sunscreens or vehicle before sensitization; hoever, the application of neither sunscreen nor vehicle alone significantly altered the CHS response. The topical application of all sunscreens completely protected against the inflammatory response to UVR at this particular dose, hereas the vehicle did not protect (Fig S) significant local suppression of contact sensitization as observed in the mice exposed to UVR but not treated ith a sunscreen (Fig S). The -EHMC-, a-p-, and P-3 - containing sunscreens completely protected against UVR-induced local suppression of CHS, hereas the vehicle did not (Fig S). DISCUSSION In the present study, the sunscreen preparations containing - EHMC or a-p (both UV absorbers) or P-3 (UV + UV absorber) protected against UVR-induced systemic and local immunosuppression, as determined by CHS to the contact allergen DNF. The photoprotective capacity of the sunscreens as UVR dose dependent ith regard to both the skin-selling response and immunosuppression; hoever, our data also demonstrated that their immunoprotective capacity as inferior to their capacity to protect against the inflammatory effects ofuvr. The variability e noted in the amount of immunoprotection observed ith all sunscreen preparations, particularly at the loer doses ofuvr, might have been due to variability in the amount of systemic immunosuppression in different experiments; alternatively, it could be due to uneven distribution of the sunscreens on the skin. y contrast, there as little variability ith regard to protection from inflammation. This observation further supports the lack of correlation beteen protection from inflammation and immunosuppression. The transmission spectra of the sunscreen preparations and their overlap spectra ith the emission spectrum of the FS4 sunlamp implied that the to UV sunscreens should be equally protective. Indeed, e observed approximately equal photoprotective capacity of the U V sunscreens for both inflammatory and immunosuppressive effects of UVR. s expected from its overlap spectrum, the P-3 - containing preparation as less effective in protecting against both UVR effects. T he results of this study differ somehat from a number of earlier studies in hich various sunscreens did not protect against different immunosuppressive effects of UVR [31-36]; hoever, most previous studies [31-36] involved chronic UVR treatment regimens ith at least five exposures and higher cumulative UVR doses than ere applied in our study. Furthermore, UVR dose responses ere not determined in any of these previous studies. If the total

5 5 WOLF ET L THE JOURNL OF INVESTIGTIVE DERMTOLOGY ~ : a-p -EHMC P-3 vehicle Sea s UYR SS a-p - EHMC P-3 vehicle a-p -EHMC P-3 vehic le SKIN SWELLING (1. mm) ER SWELLING (1 ' mm) Figure 5. The effect of sunscreens on UVR-induced inflammation and local suppression of CHS. Groups of mice ere exposed to kj/m of UV, either unprotected or protected by a sunscreen (55); n = 5; data are mean ± SEM. () Inflammatory response to UVR, as measured by skin selling 4 h after UVR exposure; 'p <.5 versus the sunscreen-untreated group. () CH5 response to DNF, expressed as ear selling 4 h after challenge; Sens, sensitized; 'p <.1 versus positive control group; values in parentheses represent percent protection by a sunscreen from suppression of CH5. UVR dose exceeded the protective capacity of the sunscreens, no effect of the sunscreens ould be observed, even though they may have been protective at loer UVR doses. Thus, the high doses of UVR and lack of dose-response information might explain hy immunoprotection as not noted in any of those studies. In a recent study, Reeve et at [37] compared the immunoprotective effects of o-p - and -EHMC - containing sunscreen preparations in albino hairless mice. The -containing preparation did not protect against systemic immunosuppression, as determined by CHS to DNF, nor did it protect against the establishment of a tumor-susceptible state, as shon by the groth of transplanted UVR-induced tumors. In contrast, the -EHMCcontaining preparation protected from both UVR effects; hoever, in their study, and -EHMC ere not suspended in the same vehicle, and P-3 as present in both sunscreen preparations but in unequal concentrations. The -EHMC preparation contained 4.5% P-3, hereas the preparation contained only 3% P-3. s demonstrated in our study, P-3 also exhibited immunoprotective capacity, and therefore its presence in higher concentration could have accounted for the superiority of the -EHMCcontaining preparation. Our finding that sunscreens have immunoprotective capacity is in agreement ith earlier ork by Morison and co-orkers, in hich p-aminobenzoic acid (P) gave nearly complete protection against systemic suppression of CHS in C3H mice and guinea pigs after three UVR exposures [3,39]. In addition, Morison and Kelley found that P protected against UVR-induced tumor susceptibility in C3H mice [4]. This latter result differs from that of a previous study using C3H mice [34] in hich preparations containing P, P esters, and P-3 had no effect on the induction of a tumor-susceptible state by UVR. Differences in methodology (i.e., sunscreen formulations and application, light sources, UVR doses, and number of treatments) could account for the discrepancy beteen those studies [34,4]. It may be important that in our and Morison and co-orkers' studies, the sunscreens ere applied liberally to the mice, hich probably resulted in higher concentrations on the skin than the mg or,u1/cm usually used [31,33,35-37]. Our results suggest that different mechanisms are involved in UVR-induced inflammation and systemic immunosuppression. Moreover, they imply that measurement of the skin-selling response to UVR is not an accurate method to determine protection from immunologic damage by UVR. The reasons for the lack of correlation beteen protection from inflammation and immunosuppression are unknon. It has been suggested that the location of the sunscreens in the skin might be important ith regard to the immunosuppressive effects ofuvr [33]. Willis and Kligman [41] shoed that photoprotection by P against UVR-induced erythema persisted despite repeated stripping of the stratum corneum. Therefore, Fisher et at [33] concluded that certain sunscreens may "pool" beneath the stratum corneum, and, consequently, although sunscreens may inhibit UVR-induced damage to deeper structures of the skin, they might not effectively block the access of UVR to urocanic acid, hich may serve as an immunomodulator [6] in the superficial epidermis. lternatively, the relatively lo immunoprotective capacity of sunscreens may be due to chemical interactions of sunscreens ith cellular molecules critical for immunosuppression. In support of this theory, increased mutagenesis as found in irradiated bacteria in the presence of P [4] and cinnamic acid derivatives related to -EHMC [43]. The investigators concluded that the sunscreens inhibited the repair ofuvr-induced DN damage [4,43]. In addition, it as reported that P can augment the formation of pyrimidine dimers in vitro after UV irradiation [44,45]. In another study [46], P esters alone ere capable of inducing DN damage, and the damage as further augmented after UV irradiation. In contrast, e found that sunscreens protected against UVR-induced formation of pyrimidine dimers in our mice (unpublished data). Thus, a fine balance may exist beteen photosensitizing and photoprotective effects of sunscreens [4,46). It as recently demonstrated that systemic suppression of CHS in UV-irradiated C3H mice is, on the molecular level, mainly mediated by DN damage in the form of pyrimidine dimers [17,1]. Therefore, these to balancing effects may explain the lack of correlation beteen the protective effects of sunscreens for inflammation and immunosuppression: sunscreens might, on the one hand, be capable of inhibiting immunosuppression by prevention of direct UVR-induced damage to DN and, on the other, contribute to immunosuppression by inhibiting repair of pyrimidine dimers or by inducing formation of lesions other than pyrimidine dimers. Finally, UVR-induced inflammation and immunosuppression may be dependent 11 overlapping but not identical pathays. Regardless of the mechanisms involved, our results have important practical implications. In contrast to previous reports, our studies demonstrate that sunscreens do have immunoprotective capabilities. These capabilities are limited, hoever, and immunosuppression can occur in the absence of measurable skin selling. This means that hen subjects expose themselves for prolonged times to UVR, although they may be protected from sunburn by sunscreens, immunosuppressive effects might still occur. We thank Pat Cox and Roger Evans,Jor their teclicaf assistallce, and Dr. Stephe" Ullrich, Jar his help in measuring the Slscreell transmission spectra. This ork as supported by grants C-5457 and C-1667 from the Natiollal Institutes oj Health and the Max Kade Fotdation Inc., Ne York, NY. P. W. is a recipient oj a Max Kade Foundatioll Jelloship. REFERENCES 1. Parish J, Kripke ML, Morison WL (cds.): Photoimmunology. Plenum, Ne York, 193. Halprin KM, Comerford M, Presser SE: Ultraviolet light treatment delays contact sensitization to nitrogen mustard. r J Dermatol 15:71-76, Noonan FP, Dc Fabo EC: Ultraviolet-I3 dose response curves for local and systemic immunosuppression arc identical. Photochem Photobioi 5:1-1, Noonan FP, DeFabo EC, Kripkc ML: Suppression of contact hypersen-

6 VOL. 1, NO. 3 MRCH 1993 SUNSCREENS ND IMMUNOSUPPRESSION 59 sitivity by ultraviolet radiation: an experimental model. Springer Semin Immunopathol 4:93-34, Toes G, ergstresser PR, Streilein )W: Epidermal Langerhans cell density determines hether contact hypersensitivity or unresponsiveness follos skin painting ith DNF.) Immunol 14: ,19 6. DeFabo EC, Noonan FP: Mechanism of immune suppression by ultraviolet radiation ill vivo. 1. Evidence for the existence of a unique photoreceptor in skin and its role in photoimmunology.) Exp Med 157:4-9, Noonan FP, De Fabo EC, Kripke ML: Suppression of contact hypersensitivity by UV radiation and its relationship to UV -induced suppression of tumor immunity. Photochem Photobiol 34:63-69, 191. Elmets C, ergstresser PR, Tigelaar RE, Wood P), Streilein )W: nalysis of the mechanism of unresponsiveness produced by haptens painted on skin exposed to lo dose ultraviolet radiation.) Exp Med 15:71-794, Okamoto H, Kriple ML: Effector and suppressor circuits of the immune response are activated i" vivo by different mechanisms. Proc Nat! cad Sci US 4: , Stingl G, Gazze-Stingl L, berer W, Wolff K: ntigen presentation by murine epidermal Langerhans cells and its alteration by UV light.) ImmunoI17: , Vermeer MH, Streilein )W: Ultraviolet light-induced alteration in epidermal Langerhans cells are mediated in part by tumor necrosis factor-alpha. Photodermatol Photoimmunol Photomed 7:5-65, Yoshikaa T, Streilein )W: Tumor necrosis factor-alpha and ultraviolet light have similar effects on contact hypersensitivity in mice. Reg Immunol 3: ,199/ Kurimoto I, Streilein )W: cis-urocanic acid suppression of contact hypersensitivity induction is mediated via tumor necrosis factor-a.) ImmunoI14:37-37, Robertson, Gahring L, Neton R, Daynes R: I" vivo administration ofll-1 to normal mice decreases their capacity to elicit contact hypersensitivity responses: prostaglandins are involved in this modification of the immune response.) Invest Dermatol :3-37, Scharz T, Urbanska, Gschnait F, Luger T: Inhibition of the induction of contact hypersensitivity by a UV-mediated epidermal cytokine.) Invest Dermatol 7:9-91, Kim T-Y, Kripke ML, Ullrich SE: Immunosuppression by factors released from UV -irradiated epidermal cells: selective effects on the generation of contact and delayed hypersensitivity after exposure to UV or UV radiation. ) Invest Dermatol 94:6-3, pplegate L, Ley R, lcalay ), Kripke ML: Identification of the molecular target for the suppression of contact hyperse nsitivity by ultraviolet radiation. ) Exp Med 17: , Kripke ML, Cox P, las LG, Yarosh D: Pyrimidine dimers in DN initiate systemic immunosuppression in UV-irradiated mice. Proc Nat! cad Sci US 9: , Kripke ML, Fisher MS: Immunologic parameters of ultraviolet carcinogenesis. ) Nat! Cancer Inst 57:11-15, Fisher MS, Kripke ML: Suppressor T lymphocytes control the development of primary skin cancers in ultraviolet-irradiated mice. SCience 16: ,19 1. Scotto), Kopf W, Urbach F: Nonmelanoma skin cancer in four areas of the U.S. Cancer 34: , Vitaliano PP, Urbach F: The relative importance of risk factors in nonmelanoma carcinoma. rch Dermatol 116: , Kinlen LF, Sheil GR, Peto), Doll R: Collaborative United Kingdom-ustraliasian study of cancer in patients treated ith immunosuppressive drugs. r) Med II : , oyle), MacKie RM, riggs )D, )unor )R, itchison TC: Cancer, arts, and sunshine in renal transplant patients: a case-control study. Lancet 1:7-75, Yoshikaa T, Rae V, ruins-slot W, Van den erg )W, Taylor JR, Streilein)W: Susceptibility to effects ofuv radiation on induction of contact hypersensitivity as a risk factor for skin cancer. ) Invest Dermatol 95:53-536, Pathak M, Fitzpatrick T, Frenk E: Evaluation of topical agents that prevent sunburn - superiority of para-aminobenzoic acid and its ester in ethyl alcohol. N Engl J Med : , Kligman LH, kin F), Kligman M: Prevention of ultraviolet damage to the dermis of hairless mice by sunscreens. J Invest Dermatol 7:11-19, 19. Young R, Harrison ), W alker SL: Lo SPF sunscreens reduce photoageing in mice. Cos met Toilet 13:49-51, 19 <;1. Kligman LH, kin F), Kligman M: Sunscreens prevent ultraviolet photocarcinogenesis. J m cad DermatoI3:3-35, 19 3.' Synder DS, May M: bility of P to protect mammalian skin from ultraviolet light-induced skin rumors and actinic damage. ) Invest Dermatol 65: , Ho KKL, Halliday GM, arnetson RSC: Sunscreens protect epidermal Langerhans cells and Thy-1 + but not local contact sensitization from the effects of ultraviolet light.) Invest Dermatol 9:7-74, Lynch DH, Gurish MF, Dayncs R: Relationship beteen epidermal Langerhans cell density, TPase activity and the induction of contact hypersensitivity. J Immunol 16:19-197, Fisher MS, Menter )M, Willis I: Ultraviolet radiation-induced suppression of contact hypersensitivity in relation to Padimate and oxybenzone.) Invest Dermatol 9: , Gurish MF, Roberts LK, Krueger GG, Daynes R: The effects of various sunscreen agents on skin damage and the induction of tumor susceptibility in mice subjected to ultraviolet irradiation. ) Invest DermatoI76:46-51, Hersey P, MacDonald M, urns C, Schibeci S, Matthes H, Wilkinson F): nalysis of the effects of a sunscreen agent on the suppression of natural killer cell activity induced in human subjects by radiation from solarium lamps.) Invest Dermatol :71-76, van Praag MCG, Out-Luyting C, Claas FH), Vermeer ), Mommas M: Effect of topical sunscreens on the UV -radiation-induced suppression of the alloactivating capacity in human skin itl vivo.) Invest DermatoI97:69-633, Reeve VE, osnic M, oehm-wilcox C, Ley RD: Differential protection by to sunscreens from UV radiation-induced immunosuppression. J Invest Dermatol 97:64-6, Morison WL: The effect of a sunscreen containing para-aminobenzoic acid on the systemic immunologic alterations induced in mice by exposure to UV radiation. ) Invest Dermatol 3:45-4, Morison WL, Pike R, Kripke ML: Effect of sunlight and its component avebands on contact hypersensitivity in mice and guinea pigs. Photodermatol : 195-4, Morison WL, Kelley SP: Sunlight suppressing rejection of - to 3-nm UV-radiation-induced skin tumors in mice. ) Nat! Cancer Inst 74:55-57, Willis I, Kligman M: minobezoic acid and its esters: the quest for more effective sunscreens. rch Dermatol 1:45-417, Hodges NDM, Moss SH, Davies DJG: The sensitizing effect of a sunscreen agent, p-aminobenzoic acid, on ncar UV -induced damage in a repair deficient strain of Escherichia coli. Photochem Photo bioi 6:493-49, Shimoi K, Nakamura Y, Noro T, Tomita I, Fukushima S, Inoue T, Kada T : Methyl cinnamate derivatives enhance UV-induced mutagenesis due to the inhibition of DN excision repair in Escherichia coli /r. Mutat Res 146:15-, Sutherland JC, Griffin KP: P-aminobenzoic acid can sensitize the formation of pyrimidine dimers in DN: direct chemical evidence. Photochem PhotobioI4: , Sutherland M: P-aminobenzoic acid sunlamp sensitization of pyrimidine dimer formation and transformation in human ce ils. Photochem PhotobioI36:95-97, Long SD, Little ): Sunscreen agents induce DN repair activity in mouse embryo fibrobl asts. J Environ Pathol Toxicol Oncol 5:193-,194