PHOTOCHEMISTRY OF UROCANIC ACID 155 distinct photochemistries (13). Specifically, photoacoustic (ultrasonic) results showed that following absorption near 308 nm, isomerization by t-UA occurs efficiently and reactive long-lived intermediate(s) are not produced. Supporting time-resolved laser studies showed that upon absorption in the vicinity of 300 nm, t-UA isomerizes in less than 100 x 10 -12 seconds (20). In contrast and in the same study, it was determined that isomerization following excitation at 264 nm near the absorption maximum is so inef- ficient because t-UA undergoes rapid intersystem crossing to a long-lived electronically excited triplet state that is approximately 230 kJ/mol above the ground state. This value is in agreement with triplet sensitization studies (21). Cis-urocanic acid was also found to generate a long-lived triplet state following absorption of 264 nm irradiation, but, like t-UA, c-UA also released essentially all of the absorbed energy after excitation near 308 nm without the generation of a long-lived intermediate state. This study deter- mined that the photochemistry seen at 264 nm is the result of an electronic transition peaking near the absorption maximum (270-280 nm), extending past 290 nm and weakening in its absorptivity past 300 nm, where a second transition begins to dominate the absorption in this region and therefore alters the photochemistry exhibited by the molecule. This conclusion is supported by similar isomerization quantum yield values seen between 254-289 nm and 300-313 nm, respectively (19). Cis-urocanic acid's role as an immunomodulator is proposed to be greater than that of trans-urocanic acid (6) however, the Cosmetic Ingredient Review Expert Panel could "not conclude whether UA is safe for use" based upon the data used in their assessment study, and they question whether cis-urocanic acid is the sole culprit or if reactive intermediate(s) should be of concern as well (3). Ultrasonic techniques have shown that both t-UA and c-UA not only generate long-lived triplet state intermediates but also generate singlet oxygen under an 02 or air-saturated solution (13) following absorption at 264 nm. These results caution against urocanic acid's use as a cosmetic ingredient. For example, as Figure 2 shows, the probability of absorption at 264 nm in the UV-C is negligible. Little if any UV-C penetrates through the earth's stratospheric ozone layer, and therefore should have little effect upon the UA photochemistry. However, when we consider that the transition excited at 264 nm also dominates the absorption spectrum and thereby the photochemistry between 264 nm and 289 nm, we conclude that triplet-state and singlet-oxygen chemistry results in this lower UV-B region and there- fore is important when considering the effects UA may have in the skin. In Figure 2, this region where solar UV reaches the earth's surface and t-UA isomerization is inefficient, but where triplet-state and singlet-oxygen production dominates the photochemistry, is marked by the striped region. The dotted region marks the wavelengths where isom- erization to c-UA from t-UA is the dominant photochemical pathway. Evidence that both UA isomers produce a long-lived electronically excited triplet state upon absorption of UV between 264 nm and 289 nm raises concerns about the subse- quent effects of the excitation of UA in the epidermis. Such electronically excited triplet molecules are of concern because of their ability to sensitize triplet reactions with natural chromophores in the skin, which in turn indicate the possibility of antigen production within the cell. Of concern as well is the application of other topical ingredients that will generate a triplet state of either greater or lesser energy than UA's (approximately 230 kJ/mol) and in turn undergo energy transfer to or from UA, respectively. In addition, our data show that both UA isomers undergo energy transfer to produce singlet oxygen following absorption of UV near the absorption maximum. Singlet oxygen is believed to

156 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS contribute to cellular damage (22), and it is possible that this photoreactive pathway contributes to the immunosuppressive characteristics exhibited by urocanic acid. The data summarized herein provides experimental evidence to the Cosmetic Review Expert Panel and to those who use urocanic acid as an ingredient in cosmetic agents that long-lived reactive intermediates do exist following absorption of solar UV by urocanic acid. The effects of these intermediates should be explored further to determine their significance in vivo, where, because both isomers of urocanic acid generate triplet state and singlet oxygen intermediates in the UV-B, both should be of concern when con- sidering urocanic acid as an additive to topical formulations. ACKNOWLEDGMENTS This work is supported by the Institute of General Medicine of NIH and by a graduate fellowship from ARCS. We are grateful to Professor Doug Magde for the use of his Excimer laser, to Professor Roger Tsien for the use of his fluorometer, and to Dr. Bulang Li for technical assistance. REFERENCES (1) A. Zenisk and J. A. Kral, The occurrence of urocanic acid in human sweat, Blochim. Biophys. Acta, 12, 479-480 (1953). (2) A. Zenisk, J. A. Kral, and I. M. Hals, "Sun-screening" effect of urocanic acid, Blochim. Biophys. Acta, 18, 589-591 (1955). (3) Cosmetic Ingredient Review Expert Panel, Cosmetic ingredient review final report on the safety assessment of urocanic acid, J. Am. Coil. ToxicoL, 14, 386-421 (1995). (4) J. H. Anglin, Jr., A. T. Bever, M. A. Everett, and J. M. Lamb, Ultraviolet-light-induced alterations in urocanic acid in vivo, Blochim. Biophys. Acta, 53, 408-409 (1961). (5) E. C. De Fabo and F. P. Noonan, Mechanism of immune suppression by ultraviolet irradiation in vivo. Evidence for the existence of a unique photoreceptor in skin and its role in photoimmunology,J. Exp. Med., 157, 84-98 (1983). (6) M. Norval, N. K. Gibbs, and J. Gilmour, The role of urocanic acid in UV-induced immunosuppres- sion: Recent advances (1992-1994), Photochem. PhotobioL, 62, 209-217 (1995). (7) A. El-ghorr, F. Pierik, and M. Norval, Comparative potency of different UV sources in reducing the density and antigen presenting capacity of murine Langerhans cells, Photochem. PhotobioL, 60, 256-261 (1994). (8) I. Kurimoto and J. W. Streilein, Cis-urocanic acid suppression of contact hypersensitivity is mediated via tumor necrosis factor-alpha, J. lmmunoL, 148, 3072-3078 (1992). (9) M. Norval, T.J. Simpson, E. Bardshiri, and S. E. M. Howie, Urocanic acid analogues and the sup- pression of the delayed hypersensitive response to herpes simplex virus, Photochem. PhotobioL, 49, 633-639 (1989). (10) J. W. Gilmour and M. Norval, The effect of UVB irradiation, cis-urocanic acid, and turnout necrosis factor-alpha on delayed hypersensitivity to herpes simplex virus, PhotodermatoL PhotoimmunoL Photoreed., 9, 250-259 (1993). (11) J. w. Gilmour, J.P. Vestey, S. George, and M. Norval, The effect of phototherapy and urocanic acid on natural killer cell function, J. Invest. DermatoL, 101, 169-174 (1993). (12) R. H. Guymer and T. E. Mandel, Urocanic acid as an immunosuppressant in allotransplantation in mice, Transplantation, 55, 36-43 (1993). (13) K. M. Hanson, B. Li, and J. D. Simon, A spectroscopic study of the epidermal ultraviolet chromophore trans-urocanic acid, JACS, 119, 2715-2721 (1997). (14) F. P. Noonan, private communication. (15) J. H. Anglin and M. A. Everett, Photodimerization of urocanic acid in vitro and in vivo, Biochim. Biophys. Acta, 88, 492-501 (1964).