Photodynamic degradation of vitamin E induced by psoralens

Biochimica et Biopt~ys~aAcra. I [ 16 ( 1992t ~ 1 - 2 % 291 © 1992 EIsevlct Sckmc¢ Publish¢fs B.V, All t'ighLsresev.'ed 0304-4165/92/$05.fl0 BBAGEN ! ~ 7 4 Photodynamic degradation of vitamin E induced by psoralens Claudio Costantini ~, M a r c o d ' l s c h i a ~, A l e s s a n d r a N a p o l i t a n o and Giuseppe Prota ~ ~, G i o v a n n a Misuraca " Deparhnent of Organic and Biol~,gical('konistm; UniversiO"of Naple~ tltab'~ and i, Department of Pharmacolog% Urdl erMlyof Naples. ~'abges ¢lla[t'~ t R ~ ¢ ¢ ~ Iqovew,bef 281h I~}1) Key words: Vitamin E: Psoralen: Photooxidalion: Singlet oxygen Psoralcns and other furocoamarins currcndy used in PUVA photochemotherapy are shown to have, to a variable cxtcm, the ability 1o hasten the rate of ultraviolet-induced photooxidation of ~r-tocopherol (a-T) in ethanol or ethanol-phosphate buffer (pH 6.8L T i c sensitizing effect varies significantly" with the substrate concentration and the natuse of the furocoumarin usod. and is dependent on the p r e s e ~ of oxygen. Scavengers of singlet osygen, e.g., gJdium azide, markedly inhibit the psoralen-sensitizcd photooxidation of a-T. whereas superosid¢ dismatase exerts an opposite, accelerating effect on the reaction rate. Catalase has no signiFa:ant influence on the kinetics of a - T decay, Analysis of the products formed by p~ralcn-scnsitizcd photooxidalion of a - T in ethanol-phosphate buffer showed the presence of tz-tocopherolquinone, its 2,3-epoxide and two related compounds containing the 7-oxaspiro[4~]dcc-I-cnc-3,6~lione ring system. The nature of thc~ products, coupled with the results of the Idnetic experiments..suggest that psoralens induce a type It, oxygen-dependem photodegradation of a - T primarily mediated by singlct o~ygen. Introduction T h e photochemical and photobiologlcal properties o f psoralens and related furocoumarios are the focus o f keen and ever increasing interest to chemists, pharmacologists and dermatologists. This interest stems primarily from the ability o f these compounds to act as potent skil~ photosensitizers when applied in conjunction with UVoA irradiation ( P U V A ) , eliciting a marked hyperpigmentary response and interfering with abnormal cell proliferation in psoriasis and other severe diseases characterize0 By epidermal hyperplasia [1-7]. Yet, the precise molecular mechanism o f action of psoralens is still controversial and is the subject of a lively debate. Traditional theories center mainly on the ability of these molecules to undergo oxygen-independent, type I Abbreviations: a-T. a-tocopherok PS. psoralen: &MOP. 8-methox~W~raten: 5-MOP. 5-methoxypsoralem TMP. 4.5'.8-ttimethylpsoralem AN, angelicin; SOD. supcr~xxid~ dlsmutase: DABCO. l.,l~liazabicyclooctahe: PUVA~psoralen + uhra~qolet A, irradiation: HPLC, high-p~rformance liquid chromatography;, TLC. thin-la~er chromatography. Correspondence: G. Prota. Department of Organic and t3iological Chemistry. Unh-ersi~" of Naples. Via Mezzocannone 16. 1-80134 Naples, Italy. photocycloaddldon reactions with pyrimidiue bases of D N A , leading to the formation of monofunctional and bifunctional adducts which are responsible for interstrand cross-linking, and, hence, mutagenicity and cell death [8,9]. In more recent years, however, it has been recognized that pseralens can undergo efficient intersystem crossing from the excited ~w, ~r* state to the 3~., ~.. state [10]. T h e availability o f the triplet state acceunts for the observed ability of pseralens to act as efficient ~ n s i t i z e r s fur the generation of singlct oxygen (1At) and o t h e r electronically activated forms of oxygen, most notably superoxidc anions, via both energy-transfer ;2nd charge-transfer processes [I 1,12]. As a result o f these studies, attention has gradually turned to the role of oxygen-dependent, type II photoreactioos in psoralen-induccd skin photosensitization [13,14]. Following the realization that singlet oxygen and supcroxide can survive long enough to diffuse within a relatively long radius from the site of generation, up to 60 nm in the case of singlet oxygen [15], active oxygen species have been increasingly implicated in a variety o f biological effects evoked by pseraleas, including erythema, edema and cytotoxicity, as well as carcinogenicity. T h e actual cell constituents targeted by these species, however, have remained up till now largely unknown. 292 Recently, as a part of a broad range research programme on the pigmentogenic effect of psoralens, we found that 8-methoxypsoralen and other widely used pharmacologically active furocoumarins can remark* ably accelerate the aerial oxidation of glutathione promoted by Pyrex-filtered UV light [16]. Wc report now evidence that psoralens can also sensitize the in vitro photonxidation of a-tocopherul (ot-TL the most active component of vitamin E [17]. Materials and Methods Materials (_+~t-Tocopherol (a-T), psoralen (PS), 8-methoxypsoralen (8-MOP). 5-mctho~,psoralen (5-MOPL 4,5'.8-trimethylpsoralen (TMP), sodium azide and diazabieyclooctanc (DABCO) were [Yore Fluka. Angelicin (AN) was kindly provided by Professor F. Dall'Acqua (University of Padua). Superuxidc dismutase (SOD. 3000 U / m g ) from bovioc-erythrocytes and catala~ (2500 U / m g ) from bovine liver were from Sigma. All other chemicals were of the highest quality available and were used without further purification. HPLC analysis Analytical HPLC was performed on a Waters model 6000A instrument, using a 4 x 9_50 mm RP 18 Lichrochart column. Detection was carried out with a Waters model 440 UV detector. ,z-T was analyzed using methanol as eluant (flow rate 1.5 ml/min, A = 280 urn). Quantitative determination of concentration was carried out by electronical integration of peak ~reas and comparison with external calibration curves. Photonxidation products of ot-T were analyzed using methanol/water, 9: I (v/v), as the eluant. Psoralen-sensitized photooxidation of a-T In a typical experiment, a solution of a-T made up to the desired concentration in ethanol or ethanol/0.02 M phosphate buffer (pH 6.8)~ 7 : 3 (v/v) (2.5 ml), was irradiated in a I × 1 cm euvette thermostated at 25°C with a 500 W Helios Italquartz high-pressure mercury_ lamp equipped with a Pyrex water jacket at a distance of 10 cm. The system produced UV irradiance of 7.0 mW/cm 2 at a I0 cm distance, ms determined with an EG and G photometer-radiometer model 550-1. Stock solutions of furocoumarins (1 m g / m l in ethanol) were prepared fresh before use. When necessary, aliquots of the stock solutions were added to the reaction mixture up to the desired concentration. Addition of sodium azide or DABCO for inhibition experiments in ethanol containing phosphate buffer significantly altered the pH of the medium, and adjustment to the initial value was required before irradiation. Stock solutions of SOD (1 mg/ml) and catalase (I mg/ml) were prepared freshly in deionized water, and 100 V-I aliquots were added when tlecessary. Anaerobic experiments were ~rried out in rubber capped ctwettes, previously flu.~dl with o~gen-frec nitrogen for at least 30 rain. Determinatiou of the 13 valae of a-T for tire 8-MOPs o ~ i z e d photooxidatiou 3 ml solutions of ot-T at various concentrations (0.18-4.50 raM) in ethanol containing 0.48 mM 8-MOP were irradiated with 4.5 J/era-" of Pyrex-filtered ultraviolet light at 25°C. Substrate consumption was followed by HPLC and did not exceed 20~L A leastsquares fit of the recipn3eals of the initial rates vs. the reciprocals of the initial concenlrat[oas of a-T ga,'e a slope of 63 +_2 min and an intercept ot" 1.9 _+0.2- 104 rain M - l from which a /3 value of 3.3+0.5- l0 3 M was calculated. Product isolation A stirred solution of a-T (520 me, 1.2l - i0 -3 tool) and 8-MOP (150 me, 5.6- 10-a rood in 1.2 I ethanolphosphate buffer (0.02 M, pH 6.8), 7:3 (v/v), was irradiated in a Pyrex immersion apparatus using a 5fl0 W Helios lralquartz high-pressure mercury lamp equipped with a Pyrex water jacket. After 3 h irradiation, the solution was concentrated to about 300 ml in a rotary evaporator under reduced pressure, and then extracted with ethyl acetate. The organic layer was dried over NazSO. and evaporated to dryness to give about 600 mg of a brownish oil. This was dissolved in light petroleum and fractionated by flash chromatography [18] on a silica-gel column (402.5 cm). After washing with I litr¢ light petroleum/ ethyl ether, 95:5 (v/v), to remove faster moving products (40 me) and the unreactod a-T (108 rag), a major fraction (220 rag) was eluted with 350 ml of light petroleum/ethyl ether, 4 : 6 (v/v). This fraction was ehromatographed on preparative silica gel plates (ethyl aeetate/cyclohexane, ~ : 75 (v/v)) to afford four main distinct ultraviolet-detectable bands at Rf = 0.65, 0.54, 11.48 and 0.42. The first band (88 rag) proved to consist of two products eluting on HPLC after 36.1 and 49 rain. These were isolated by preparativu HPLC (elaant : methanol/I M sodium acetate (pH 4.25), 93 : 7 (v/v); flow rate 6 od/min) as colorless oils. aud were identified as a-tocopherolquinone (I) (22 rag) and its epoxidc (2) (61 mg). The hands at R F = 0.54 and 0.48 (24 and 16 rag, respectively), elutiog on HPLC after 18.6 and 19A rain, afforded two diastereoisomers of 1,2,8-trimethyl-4met hylenc-8-p hytyl-7-oxaspiru[4.5]dcc- 1-ene-3,6-dione O) as coloudess oils, h,o~(cyelohexane)=238 am; ~'m~(CCi4) = 1715, 1708 cm n. The band at Rf=0.42 (12 mgk elutiog after 18.0 min on HPLC, afforded one diastercoisomer of 4-by- 293 a - T t,/o) 40 eo ~1~ (10 ~,n / M4) J / 6o 40 20 O J/era 2 Fig- I- Effect o f furoroumarins (0.48 mM~ on the aerobic ~ t[on o f a - T t0.89 raM) in e t h e n ~ , o , no additbce; o, + A N ; z~, + 8 - M O P ; A, + P S ; £]. ~- 5 - M O P ; I I , + T M P . Each value is [[g¢ mean for three e ~ l ~ m e n t s I SE is [lldJcated by t he VeIlical hat). droxy- 1,2,4,8-tetramethyl-8-phytyl-7-oxaspiro[4.5]dec-1 ene-3,6-dione (4) as colonrless oil, a~(cyciobexane) = 232 nm; ~,=~(CCI,) = 3555, 3418, 1718 (broad) cm-L The structural characterization of the products followed from spectral analysis, =H- and t3C-NMIL mass spectrometry, ultrav/olet and comparison of their chromatographic properties with those of authentic samples obtained by methylene blue-sensitized phot~-xgdation of a-T [19,20]. Resalts The psoralen derivatives used in this study include the parent psoralen (PS), 4,5",8-trimetbylpsoralen (TMP), angelicin (AN), 8 - m e t h o ~ r a l e n (8-MOP) and 5-methoxypsoralen (5-MOP). Fig. ! shows the effect of the various furocoumarins (0.48 raM) on the photooxidation of 0.89 mM a-T in ethanol with Pyrex-filtered ultraviolet light (wavelengths > 320 nm). in the presence of the sensit/zers, a significant acceleration of the rate of decomposition of a-T was observed, to an extent that depends on the nature of the psoralen added. TMP and 5-MOP proved by far the most effective in inducing a-T decay, fob oc~ ps 5-Mop B-MOP Cl.t3 AN TNP 1/c0{103M-' I Fig. 2. Double-reciprocal plot o f the initial rate . f a -T ph,lnoxidalioll vs. the initial concentration o f a-T. with 0.48 mM 8-MOP. Irradiation was carried out under the usual conditions with 4.5 J / c m / of ultraviolet light at 25°C. lowed by PS, 8-MOP and AN in that order, in control experiments, no reaction between psoralens and a-T was observed in the dark. Under the same experimental conditions, a-T was completely stable in the dark. Further experiments on the effect of psoralens on a-T photooxidatinn were carried out using 8-MOP. which is widely used in PUVA therapy. The photosensitizing effect of 8-MOP was found to be dependent on the presence of oxygen: under rigorously anaerobic conditions, only 3.0 _+0.5% consumption of oL-T with 13.5 J / c m 2 irradiation was observed, both in the presence and in the absence of the sensitizer. Under oxygen atmosphere, on the other hand, a-T consumption in the 8-MOP sensitized reaction was increased of 22 + 3% with respect to the same reaction conducted in air. A partial photedestruction of the sensitizer was observed under these latter conditions. At a f'~ed concentration of 8-MOP (0.48 mM), the rate of a-T photooxidation was found to vary significantly with the initial concc;;ttatioa of the suhstrate within the range 0.18-4.50 mM. This is apparent from Fig. 2, which shows a double-reciprocal plot of the initial photooxidation rate of of a-T vs. the initial concentration of ~-T in ethanol. No significant dependence of the rate of a-T (0.89 raM) photoosldation on 8-MOP concentration was observed. After 2h irradiation, recovery of a-T was 52 _+ 3% with 0.15 mM 8-MOP and 49 _+2% with 0.75 mM 8-MOP, variations falling within the experimental error. To investigate the effect of active oxygen scavengers on the 8-MOP-sensitized photooxidation of a-T, in subsequent experiments ethanol containing 30% phosphate buffer (0.02 M, pH 6.8) was used as the solvent, to permit solubilization of the additives at the ncccs- 294 TABLE I a.... Effect of active ox3'ge~z .scal engets ~m the pIzotooMdation of a - T (O.ll9 raM) sensitized by 8-MOP (0.4a raM) m etltamd/phosplzate buffer [002 M. p H 68), 7:3 h / v L A\ Irradiation ~-as earned out under the usual conditions with 5a J/era: of ultraviolet light at 25°C. Additive .[ a-T consumption I%) 33 _+5 52_+5 36_+4 77 + 7 58 _+b 97 + 3 94 + l Control 8-MOP 8-MOP+ NaN~ (1O raM) 8-MOP + DABCO ( 10 raM} 8-MOP +catalase t 100 U/mB 8-MOP + SOD ( 1l0 U/ml) 8-MOP + SOD + catalase = 0 10 20 ~0 aO 50 TIm;~t~in] b sary concentrations. T h e results are reported in T a b l e I. At a concentration of 10 m M sodium azide, a most efficient q u e n c h e r o f singlet oxygen [21], completely suppressed the ~ n s i t i z i n g effect of 8 - M O P . Q u i t e unexpectedly, diazabicyclooctane ( D A B C O ) , a n o t h e r widely used singlet oxygen q u e n c h e r [22], at 10 m M concentration significantly increased the rate o f the 8-MOP-sensitized photodecomposition of a - T . However, it reduced a - T consumption from 51 + 2 % (control) to 24 +_ 3 % when the sensitized photooxidation was carried out in absolute ethanol as the solveuL Superoxide dismutase ( S O D , 1 i 0 U / m l ) caused a dramatic increase in the reaction rate, while catalase at a 100 U / m l concentration had no effect. Control experiments showed that the activities o f S O D and catalase are virtually unchanged in e t h a n o l / p h o s p h a t e buffer, o 2o 30 ao 5o ;ime (mln) Fig. 3. Ts'plcal HPLC e l u t k m prof'd~ of the prodms feinted by. pivatooxldation of a-T t0.89 raM) in ethatml/plmsphate buffet {0.02 M. pH 6:8L 7:3 (v/vL in tat the presence of and in (b) the absence oF 8-MOP (0.48 mMk Irradiation was tamed out under the usual conditions with 50 J / c m 2 of ultraviolet tight at 2YC. The ¢lmmt was r~ethanol/water 9: l, the flow rate: was i 3 ml/m[n, detcctloo was at 254 urn. =-T,luted at 93 rain under these coeMiti~Ls. products o f a - T , as evidenced by H P L C anab~is (Fig. 3at. Formation o f these products was d e p e n d e n t on the presence o f the psoralen, since in the absence o f 8 - M O P none o f them was apparently g e n e r a t e d in significant yields (Fig. 3b), and was markedly i n h ~ i t e d by addition o f sodium azide. S O D , in spite o f the accelerating effect, did not affect the product pattern. Fractionation o f the photooxidation mixture by preparative T L C and H P L C allowed the isolation o f five products (overall isolated yield about 3 0 % o f the consumed substrate), corresponding to peaks A - E in Fig. 3a. Two of these (peaks E and D ) were identified as a-tocopherolquinone (1) and its epoxide (2) by comparison of their chromatographical and spectral properties with those o f authentic samples [20]. T h e other products proved identical (HPLC, T L C , IH and ;3CN M R , ultraviolet, IR, mass spectrometry) with 1,2,8- 7:3 (v/v). In-subsequent experiments, the nature o f the products formed by 8-MOP-seositized photooxidation o f a - T was investigated. T h e chemical course o f the reaction was found to be markedly dependent on the nature of the solvent system used: when the photooxidation was conducted in absolute ethanol, a very complex and chromatographically ill-defined mixture of products was obtained; however, using e t h a n o l / phosphate buffer. 7 : 3 ( v / v ) , as the solvent system, the reaction afforded a well-defined pattern o f oxidation o T 1o ~ c-~H~ c~oH3~ o ~ o ,H a o a Fig. 4. Structures of the products obtained by 8-MOP-sensalzed photooxidation of a-T. 295 t rimethyl-4-metbylene-8-phytyl-7-oxaspiro[4.5]dce-1ene-3,6-dione O ) (two diastereoisomers, peaks B and C) and 4-hydroxy-l,2,4,8-tetramethyl-8-phytyl-7oxaspiro[4.5]dceol -¢ne-3,6-dione (4)(peak A), two novel oxidation products of a - T recently obtained by dyesensitized photooxidation wRh methylene blu¢ and visible light [19]. The structures of the products isolated from the 8-MOP sensitized photoo~dation of a-T are shown in Fig. 4. The remainder of the consumed a - T could be accounted for b y extensively degraded materials and trace products that could not be isolated and characterized. All the results obtained with 8*MOP were qualitalively confirmed using other more active sensitizing psuralens, e.g., TMP and IS, which, however, proved less stable to irradiation than 8-MOP, and gave, therefore, less reproducible data. Discussion a - T is a mogt efficient phenolic antioxidant present in cell membranes, and plays a central role in the defense mechanisms against peroxidative damage following oxidative stress conditions [2311. It is therefore conceivable that abnormally high fluxes of singlet oxygen and superoxide ions, produced by psoralens upon excitation with UVA light, may ultimately lead to a loss of the a - T engaged in the antioxidant activity. The marked ability of psoralens to accelerate the aerial photooxidation of a-T, reported in this study, would apparently support this view. Chemically, the sensitized reaction is primarily mediated by singlet oxygen, as indicated bY the inhibito~ effect of quenchers like sodium azide and DABCO in ethanol. The opposite, enhancing effect of DABCO on the reaction rate in phosphate containing medium, though surprising, is not unprecedented, as it has been obserced in the previous study on glutathione photooxidation [16] as well as by other authors [24]. It is possible that in aqueous medium this quencher [h not competitive towards singlet oxygen with respect to the substrate [25]. Rather, owing to its basic nature, it may accelerate the photooxidation rate by favouring deprotonation of the phenolic group of aoT by a pH°independent, general-base catalyzed process, thereby enhancing the susceptibility of the substrate to oxidation. The involvement of singlet oxygen in the psoralen sensitized photooxidatinn of a - T would also 1~.~ gupported by the pattern of products isolated, which is --' a-Tocopherol oa • 0~4 , ot433 ~ e~ctptex =-'r +~oa Aaaucts Fig.6. Suggestedmechanismaccountingfor the acceleratingeffectof SOD on the 8-MOP-sensitizedphotoosidation of a-T. identical to that recently obtained by dye-sensitized photonxidation of a - T [19], i.e., under conditions where singlet oxygen is established to be the main oxidizing species. A possible mechanism of formation of the unusual 7-oxaspirol4.5]dec-l-ene-3,6-dione system is depicted in Fig. 5 and involves a ring contracting rearrangement of an epoxyquinone hemiketal or a related species arising by the initial addition of singlet oxygen to a-T. From the double-reciprocal plot shown in Fig. 2, a value was determined of 3.3- 10 -3 M, which is significantly higher than the values of 1.4-10 4 M and 3.2- 10 -4 M reported in the literature for the methylene blue-sensitized photooxidatinn of a-T in methanol [26] and in ethanol [27], respectively. Since the/3 value is the ratio between the decay rate constant of singlet oxygen and its reaction rate constant towards the substrate, and therefore should not depend on the type of sensitizer used [28], it is possible that in the case of the 8-MOP-seusitized reaction there is a significant interaction of the psoralen with singlet oxygen [29], affecting determination of p value. Moreover, though the bulk of the evidence accumulated in our study points to a substantial involvement of singlet oxygen in the psuralen-sensRized photooxidation of a-T, other oxygen species, especially superoxide, may as well participate in the photooxidation reaction. In this connection, the stimulating effect of SOD can be rationalized as follows. Interaction of a-T with singlet oxygen is known to involve as the initial event the formation of an ¢xciplex [30], which, in addition to energetic collapse a n d / o r reversion to reagents, can be engaged with irreversible chemical reactions, affording adducts, and reversible charge transfer processes, yielding superoxide ions and noT phenoxyl radical. Addition of SOD to the reaction medium would shift this latter equilibrium to the right, kinetically favouring the radical-forming route and sidetracking the a-Tsinglet oxygen exciplex from the otherwise prevailing path leading to the photooxygenatiou products (Fig. 6). ~C7% " 1¢~H 3~ 0 ~ C 0 0 16H33 4 Fig.5. Poss~i¢mechanismof formationof the 7-oxaspiro{43]dec-l -ene-3,6~Jion¢ products by singletoxygenationof a-T. 296 A similar accelerating effect o f S O D has been reporied for the autoxidation o f 3-hydroxyanthranilic acid [31 ]. T h e observed ability of psora[ens to sensitize the photooxidation of a - T , coupled with t h e results o f previous studies on glutathione photooxidation [16], may provide a useful lead to understand at the molecular level the mechanism o f skin photosensitization. It is conceivable t h a t a depletion o f biological antioxidants may form the basis o f the t h e r a p e u t i c activity, of psoralens and ultraviolet irradiation; this, however, justifies the existing concern on the use o f sensitizing furocoumarius for purposes o t h e r than clinical. R e l e vant to this discussion is t h e reported protective effect o f vitamin E against lipid peroxidation a n d o t h e r celld a m a g i n g processes occurring during skin irradiation with ultraviolet light [32]. Acknowledgments This work was supported in part by a g r a n t from the M i n i s t e m della Universit~ e delia Rieerca Scientifica e Tecnologica (Rome} and the C N R , Progetti Chimica Fine I I e Qualit~ degli Alimenti. References 1 Song. P.-S. 0979) Photochem. Photohiol. 29, 1177-1197. 2 Parson. B.J. (IgS0) Phetochem. PhotobioL 32. 813-821. 3 R0dighiero. G.. Dall'Acqua. F. and Pathak. M.A. (1984) in Topics in Photomndicine (Smith. K-C.. ed.I, pp. 319-398. P~enum Press+ New York. 4 Hoaigsmann. H. {1987) in Light in BioloD' and Medicine (Douglas, R.H.. Moan, J., DalrAcqua, F.. e ~ ) . Vol. I. pp. 315-319. Plenum Pres~. Ne~vYork. S Forint, P. (1989) in Psoralens. Past, Present and Future of Pho~ mchemoprotection and Other Biological Activities (Fitzpatrick. T.B. Forint. P., Pathak. M.A. and Urbach, F.. eds.), pp. 63-71. John Libhey EurotexL Paris. 6 PathaL M.A., Dane Carbonate, M. t1989) in Psoralens. Past, Present and Future of Photochemoprotection and Other Biologi+ cal Activities (Fitzpatrick, T.B., FodoL P.. Palhak, M_A. and Urbach, F. eds.L pp. 87-101. John Ll~lxy Eurotext. Paris. 7 DalrAequa. F. 11989) in Psoralens. Past, Present and Future of Photochemoprotcetion and Other Biological Activities (Fitzpatrick. T.B~ Forint. P. Pathak, M.A. and Urbach. F.. nds.L pp. 237-Ls0, John Libbey Eurotext, Paris. 8 Kanne. D~ Straub, K., Ratmpo~. It_ add Heztst J.F.. {19821 Biochemislr./21. 861-87L 9 DalPAcqua. F- Marcmni. S- Vnd01dL D. and P.cdighierOo G. (|974) Z. NaturL 29c. 635-63@ 10 Sa E Me[o, M.T~ A~cfbeck. D~ Bcnscsson, [LV~ Lind. EJ. and Salel. C+ (19"/91 Pholochcm. P ~ . 30. l~45 651. I I Dc Mol. N J_ and B c l j c ~ r g c a ~3n HeL~gom~-cn.G.MJ. 11979) photochem. PhotohmL ~0. 33i-335. 12 Pathak. M.A. and JoshL PC. (1984~Bmchim. B/ot~h~s..ag'ta798. 115-126. 13 Saict. C.. Morcno. G. and Vinzel-~.F. tl982) ~h~lc~hcm. Pi'aRobiol. 3h. 291-296. 14 Can~aro. C. and Palhak. M.A. (19Kg) J. Iw.'csl_ Ik.--raatol_ t~, 267-275. 15 Pooicr. J.P. and Valen~no. D.P. {I979) Pholochen'L Pholcbh~L 30. 581-584. 16 d'lschia, M.. Napolitano. A. and prola. G- (l~91 Bmch/m. 3/ophys. Acta 993, 143-14Z 17 Machlin. L L 11980)Vitamin E: A Co*nprehcn,_b.e Treatise. Marcel Dekker. New YorL 18 Still. W.C~ Kahtx M. a~d ~.litra, A_ 11~8~ J- OrB. Ch¢m. 43, 2923--~23. 19 d'lschia. M.. Costantini. C. and proa.a.G. {1991).L Am. Ch~m. Soc. 113°8353-8.]56. 20 Grams. G.W., Eskins. K. and InglctL G_E_(1972) .L ~ Chem. Soc. 94. 866-86S. 21 FoP(e, C.S. Fugimoto. T-T. and Chang. Y.C. t1972}Tetrahndmn Left. 45M8. Ouannes. C. and Wilson. T_ (1968) J. Am- Chem. S~c_ ~0, 6527-6528. 23 Burton. G.W. and bigo!d. K-U_ {19g6l A¢¢. Chem. I1¢s. 19. 194-201. 24 Auclair, C. i1987L Phot0chem. photobioL 45. 35 38. 25 Fahrcnholtz, S.R.. Doleiden. F.H~ Trozxo[o. A.M. ~md L,x,n o ~ A~A. 11¢)74)Photochum. photobioL 20, 505-509. 26 Grams, G.W., Eskins, K. (1972-)BiochcmislPj IL 606~f~08. 27 YamauchL R.. Malsusbita. S. 11977) Agticol. Bh~l. Chem. 41. 142.5-1430. 28 Young, R.H., Wchrly, I~Land Marlin, g . L 11971)j. Am. Chem. Soc- 93. 5774-5779. 29 Wagserman. H.tt. and Berdhal, D.R. 11982)Photochem. Photobiol. 35. 565-567. 30 Gorman, A.A., Gould, I.R., Hamblctt. I. and Statzden. M.C. (1984) J. Am. Chem. Soc. 106. 695h-6959. 31 lshii, T., twahashi. H.. Sugata, R . Kido. R. and Frido~'/~L I. (lgff0) Arch. Bioch¢m. Bioph~. 27~, 248-2.50. 32 KheUab, N.. Amos', M_C., Briand. G.. Bousqaet, B. FofloL P_ Barey. M. and Combm. A. 119891in PSOrdlens. PasL Present and Future of Photochemoproteclion and Other Bioi~,/cal ACti,.-hles (Fitzpatrick. T.B.. ForloL P.. Pathak. M.A_and UroaclL ~ eds.). pp. 207-212. John Libb~- EurotexL Paris.