Chem. Res. Toxicol. 1993,6, 542-547
642
Isolation and Identification of Singlet Oxygen Oxidation Products of &Carotene Steven P. Stratton,? William H. Schaefer,t and Daniel C. Liebler'tt Department of Pharmacology and Toxicology, College of Pharmacy, University of Arizona, Tucson, Arizona 85721, and Drug Metabolism and Pharmacokinetics, SmithKline Beecham Pharmaceuticals, P.O. Box 1539, King of Prussia, Pennsylvania 19406-0939 Received February 23,1993
Singlet oxygen is a highly reactive form of oxygen produced by many toxic photosensitizers. @-Carotenequenches singlet oxygen catalytically through a very efficient physical reaction. However, concomitant chemical reactions during photosensitized oxidations consume @-carotene. T o investigate the hypothesis that chemical reactions with singlet oxygen consume P-carotene, we characterized products of the photosensitized oxidation of P-carotene. @-Caroteneand the photosensitizer rose bengal were dissolved in toluene/methanol(85:15 v/v), which was bubbled with 02and illuminated with a quartz-halogen lamp for 30 min a t 5 "C. Reaction products were analyzed by reverse-phase HPLC, UV-vis spectrophotometry, and mass spectrometry. @Carotene oxidation products were identified as @-ionone,P-apo-l4'-carotenal, @-apo-10'-carotenal, @-apo-8'-carotenal,and @-carotene5,8-endoperoxide. Formation of these products was dependent on the presence of the photosensitizer. The products apparently were formed from the action of singlet oxygen rather than by photochemically-initiated @-carotene autoxidation, since suppression of autoxidation by equimolar a-tocopherol did not diminish product formation. @-Caroteneautoxidation initiated by 2,2'-azobis(2,4-dimethylvaleronitrile),which generates peroxyl radicals, yielded a different product distribution than that from photosensitized oxidation. Specific products formed by singlet oxygen oxidation of @-carotenemay serve as markers for singlet oxygen quenching in biological systems.
Introduction
and the formation of @-caroteneoxidation products:
Both singlet oxygen and free radicals produced during photosensitized oxidations can cause cellular damage by reacting with DNA and proteins, or by inducing lipid peroxidation ( 1 ) . @-Carotene(1, Chart I) may protect against photosensitized tissue injury by scavenging free radicals and by quenchingsinglet oxygen (2,3).@-Carotene is employed clinically to prevent photosensitized tissue damage in humans with erythropoietic porphyria (4). Several studies have shown that @-caroteneinhibits free radical-mediated lipid peroxidation in homogeneous solution (5-7), in lipid bilayers (8-lo), and in microsomal membranes ( 11-13). @-Caroteneprobably scavenges peroxy1radicals by forming @-carotene-radicaladducts, which can trap other radicals to form nonradical @-carotene oxidation products (5, 9, 14). @-Caroteneis among the most effective singlet oxygen quenchers known (15, 16). @-Carotenequenches singlet oxygen primarily by a physical reaction, in which excess excited-state energy from singlet oxygen is absorbed by @-caroteneand then dissipated by the polyene chain while oxygen is returned to the triplet ground state (15,17):
k,
singlet oxygen + @-carotene
-
+ 3@-carotene
+ University of
Arizona.
kc
(1)
3@-carotene @-carotene+ heat (2) In contrast to physical quenching, concomitant chemical reactions result in destruction of the polyene chromophore t SmithKline Beecham Pharmaceuticals.
-
@-carotene+ singlet oxygen @-caroteneoxidation products (3) Although it is a minor quenching pathway (151, this chemical quenching reaction results in the loss of @-carotene (18)and thus of antioxidant protection. The specific reactions that contribute to chemical quenching are not understood (2). Although @-carotenephotooxidation has been studied previously, it is not clear from previous reports whether the observed products were formed by reactions of @-carotenewith singlet oxygen or with other oxidants ( 1 S 2 3 ) . Products specifically formed by @-carotene-singlet oxygen reactions could serve as chemical markers for total singlet oxygen quenching. Analysis of specific products thus might be used to monitor singlet oxygen quenching in experimentalmodels of phototoxicity. Here we report studies of the singlet oxygen-dependent photooxidation of @-carotene. The reaction products, which include apocarotenal chain cleavage products and a novel endoperoxide, were shown to be dependent on singlet oxygen rather than photoinitiated autoxidation reactions.
Experimental Procedures Chemicals and Instrumentation. @-Carotene,@-ionone,and @-apo-8'-carotenal(Fluka Chemical Corp., Ronkonkoma, NY) were used as received. Retinal and retinol were from Sigma (St. Louis MO). @-Carotene, carotenals, and retinoids were stored at -20 "C under Ar. Rose bengal bis(triethy1ammonium) salt (dye content 90%) was from Aldrich (Milwaukee, WI). 2,2'-
0893-228~/93/2706-0542$04.00/0 0 1993 American Chemical Society
Chem. Res. Toxicol., Vol. 6, No.4, 1993 543
Photooxidation of @-Carotene Chart I
1
0
2
0 3
0 4
5
Azobis(2,4-dimethylvaleronitrile) (AMVN)’ was from Polysciences Inc. (Warrington, PA). d-a-Tocopherol was a gift from Henkel Fine Chemicals (La Grange, IL). Products analyzed by HPLC were detected with a HewlettPackard Model 1040M diode-array detector system coupled to a HP 7994A ChemStation computer. Solvents were delivered with a Spectra-Physics SP8810 isocratic pump. Electron ionization and high-resolution MS was performed with a Finnigan MAT-90 instrument (Finnegan MAT, Bremen, Germany). Samples were introduced by direct probe insertion and ionized with a7O-eV electron beam. GC-MS was performedwith aVarian 3400 gas chromatograph coupled to a Finnigan MAT-90 mass spectrometer. LC/MS and LC-MS/MS were performed with a Sciex API I11 mass spectrometer (Sciex, Toronto, Canada) with Ionspray ionization (a proprietary, pneumatic nebulizer-assisted electrospray ionization interface) operated in the positive-ion mode. Compounds were introduced into the instrument by flow injection in a mixture of methylene chloride/acetonitrile/lOmM ammonium acetate (pH 5) (25752.5 v/v) at a flow rate of 50 pL min-1. The ionspray potential was 5500 V and the interface potential was 650 V for all experiments. The orifice potential was set to 35 or 80 V for analysis of product 5 and at 35 or 50 V for analysis of diol 7. MS/MS experiments were done using argon as a collision gas at a thickness setting of 425 X 10l2atoms cm-2. Photosensitized Oxidationof @-Carotene.fl-Carotene(85.9 pg, 160 nmol) and 25 pg of rose bengal bis(triethy1ammonium)
* Abbreviations:
AMVN, 2,2’-azobie(2,4-dimethylvaleronitrile).
salt were dissolved in 20 mL of toluene/methanol (8515 v/v). This solution was bubbledwith 02in a jacketed borosilicate glass flask and illuminated with a 300-W tungsten-halogen lamp at a distance of 15 cm for 30 min at 5 OC. Solvent was evaporated in uacuo, and the product residue was dissolved in mobile phase for purification by HPLC as described below. All products were handled under reduced light. Analysis of Oxidation Products. @-Caroteneand its oxidation products were separated by reverse-phase HPLC on a 5-pm Spherisorb ODs-2 column (4.6 X 250 mm), eluted with methanolfhexane (8515v/v) at 1.5mL/min. The product eluting at 2.6 min (2) was collected and further purified by normalphase HPLC on a 5-pm Spherisorb CN column (4.6 X 250 mm), eluted with hexane/ethyl acetate (92.57.5 v/v) at 1.5 mL/min. Product 2 was identified as &apo-l4’-carotenak UV-vis (see Table I); MS mlz 310 (M+, 1001, 295 (7), 281 (12), 105 (30); highresolution MS mlz 310.2283 (310.2297 calcd for CzzHaoO). The product eluting at 3.1 min (product 3) was further purified by normal-phase HPLC as described above and identified as fl-apc-lW-carotenak UV-vis (seeTable1); MS mlz 376 (M+,loo), 361 (3), 347 (2), 284 (2); high-resolution MS mlz 376.2762 (376.2766 calcd for C2,HwO). The product eluting at 3.4 min (4) was further purified by normal-phase HPLC and identified as 6-apo-8’-carotenal: UV-vis (see Table I); MS mlz 416 (M+,100), 324 (3)) 265 (2), 239 (3); high-resolution MS 416.3039 (416.3079 calcd for C&O). The product eluting at 5.8 min (5) was further purified by normal-phase HPLC and identified as @-carotene 5,8-endoperoxide: UV-vis (see Table I); MS mlz 568 (M+, 38), 550 (12), 534 (2), 476 (41, 404 (91, 205 (ll),165 (100); highresolution MS mlz 568.4263 (568.4280 calcd for C~OHBO~). For GC-MS analysis, photooxidations were done as described above except that the solution was bubbled with 0 2 for 5 min and then capped during the 25-min irradiation period to prevent the loss of volatile products. A 2-pL aliquot of the reaction solution was analyzed by GC-MS with a 30-m DB-5 capillary column (J&W Scientific, Folsom, CA). The column was operated isothermally at 150 OC for 7 min and then temperature programmed to 300 OC at 15 OC/min. Product 6 was identified by GC-MS as @-ionone:MS m/z 192 (M+,4), 177 (loo), 159 (4), 149 (6), 135 (11). Identical results were obtained upon analysis of an authentic standard. NaBH, Reduction of @-Carotene5,bEndoperoxide. Product 5 (approximately 30 pg) was dissolved in 2 mL of methanol/ hexane (8515 v/v) and treated with 15 mg of NaBH4 at 0 OC for 5 min. The solvent then was evaporated under Nz, and the residue was dissolved in 1.5 mL of methanol/hexane (8515 v/v) and analyzed by reverse-phase HPLC as described above. The 5,8dihydroxy-&@-carotene7 eluted at approximately 3.0 min and exhibited a UV-vis spectrum identical to that of 5, which eluted at 5.3 min. The peak corresponding to 7 was recovered after evaporation of the mobile phase under Nz; product yield was approximately 60 % . To improve the overallyield of 7, unchanged 5 was recovered and resubjected to NaBH4 reduction as described above. Effect of a-Tocopherol on Photooxidation and AMVNInduced Oxidation of @-Carotene. To study the effect of a-tocopherolon oxidationof @-carotene,photosensitizedoxidation of @-carotenewas repeated as above with an equimolar amount of a-tocopherol (68.9 pg, 160 nmol). Products were analyzed by HPLC as described above. Peroxyl radical oxidations contained 800 nmol of @-caroteneand 240 nmol of AMVN in 100 mL of toluene/methanol(85:15 v/v) and were done in the dark at 37 OC for 7 h. Some incubations contained a-tocopherol (345 pg, 800 nmol). At hourly intervals, 5-mL samples were taken from each flask and the solvent was evaporated in uacuo. The residue was dissolved in HPLC mobile phase, and oxidation products were analyzed by reverse-phase HPLC.
Rssults Identification of &Carotene Oxidation Products. W e chose t o s t u d y the singlet oxygen oxidation of
544 Chem. Res. Toxicol., Vol. 6, No. 4, 1993
2
536
R
3
0
Stratton et al.
4
6
8
10
[M-0 21
\
12
8
14
Time, min
I
Figure 1. Reverse-phase HPLC of &carotene photooxidation products, showing rose bengal photosensitizer (RB),@-carotene (l),and oxidation product peaks (2-5). Table I. UV-Vis Absorbance Maxima for @-Caroteneand Its Oxidation Products: UV-vis absorbance De& product maxima (nm)b 1 @-carotene 272,450,476 2 ,&apo-l4’-carotenal 384 3 fl-apo-lO‘-carotenal 246,434 4 fl-apo-8’arotenal 264,452 5 &carotene 5,8-endoperoxide 252,314,404,424,450 0 Purified products were analyzed by normal-phase HPLC with diode-arraydetection. b Absorbance maxima were recorded in normalphase HPLC mobile phase (seeExperimentalProcedures). Absolute maxima are listed in boldface.
@-carotenein a relatively simple homogeneous solution model, which facilitates the analysis of potentially labile products and simplifies interpretation of the reaction chemistry. Singlet oxygen was generated by visible light irradiation of rose bengal bis(triethy1ammonium) salt. Direct quenching of the triplet sensitizer is negligible a t the oxygen and @-caroteneconcentrations employed (15). Irradiation of @-caroteneand the photosensitizer in toluene/methanol caused @-caroteneoxidation. Reversephase HPLC analysis ihdicated that several products were formed, all bf which were more polar than @-carotene (Figure 1). The peak labeled RB contained the photosensitizer, which was not retained by the column. Peak 1is unoxidized @-carotene. Peaks 2-5 contain @-carotene oxidation products which were subjected to further analysis. The coelution of minor products with peaks 2-5 in the reverse-phase system necessitated further product purification by normal-phase HPLC. The minor products appearing in the polar region of the chromatogram have not yet been characterized. On the basis of their UV-vis spectra, low- and highresolution mass spectra, and HPLC retention characteristics, products 2-4 were identified as @-apo-l4’-carotenal7 @-apo-10’-carotenal,and @-apo-8‘-carotenal,respectively. Absorbance spectral data for the purified products are shown in Table I. Shifts in UV-vis A-valuesfor products 2-4 are consistent with those for homologousapocarotenals with increasing polyene conjugation (24). Products 2 and 3 show similar mass spectral fragmentation patterns (see Experimental Procedures), including characteristic M 15 (loss of CH3) and M - 29 (loss of CHO) ions. MS of product 4 shows M - 15 to be a very weak fragment, and M - 29 loss was not detected. However, the UV-vis spectrum and MS of product 4 matched those of an authentic @-apo-8’-carotenalstandard. UV-vis analysis of peak 5 showed an absorbance maximum at 424 nm (Table I). Both the HPLC retention
AAA ,
1
1
403
I
too
200
_I.,
.
I
. , ,1
300 m / z
,”
400
.I
500
I
Figure 2. LC-MS/MS collision-induced dissociation product ion spectrum of M + (m/z 569) for product 5. time and the UV-vis spectrum appeared essentially identical to those of @-carotene 5,8-epoxide (mutatochrome; MW 552) (25). However, MS of 5 shows the molecular ion at mlz 568 which corresponds to @-carotene plus 0 2 . High-resolution mass spectrometry supported the identification of 5 as a dioxygen addition product of @-carotene. The striking similarity between absorbance spectra and chromatographic behavior of product 5 and @-carotene5,gepoxide suggests that the two compounds are structurally analogous. Product 5 thus was tentatively identified as @-carotene5,8-endoperoxide. LC/MS and LC-MS/MS Characterizationof @-Carotene 5,8-Endoperoxide5 and Its ReductionProduct, 5,8-Dihydroxy-@,@-carotene. Product 5 was further characterized by LC/MS and LC-MS/MS. The ionspray mass spectrum of 5 obtained a t an orifice potential of 35 V contained an ion at mlz 568, which corresponded to M+. Less intense ions were observed at mlz 569,586,627, and 668, which represented [M + HI+, [M + NH41+, [M + NH4 + CH&N]+, and [M + NH4 + CH&N]+. Adducts of the ion at mlz 568 were not observed. Radical cations typically are not observed in electrospray mass spectrometry. However, ionization by loss of an electron from the electron-rich polyene is not surprising. Ionization by electron transfer during electrospray has been reported previously (26). Collisionally induced dissociation of the M‘+ ion at mlz 568 yielded an intense product ion at mlz 536, which corresponded to elimination of 02 from the endoperoxide (Figure 2). Product 5 was treated with NaBH4, and the reduction product was analyzed by reverse-phase HPLC. The reduction product exhibited increased polarity (i.e., a decreased retention time) compared to endoperoxide 6 (data not shown). Reduction did not alter the polyene of the reduction product was chromophore, as the A, identical to that of the endoperoxide. These observations suggested that the reduction product was the corresponding 5,8-diol 7. The ionspray mass spectrum of the reduction product displayed an intense ion at mlz 553, which corresponded to [M + H - HzOl+ for diol 7. An intact [M + HI+ ion was not observed. The atmospheric pressure chemical ionization spectrum for this product showed the same result with an additional ion at mlz 535, consistent with elimination of two water molecules ([M + H - 2Hz01+. Such facile elimination of H20 from this highly conjugated allylic alcohol is not unexpected. Collision-induced dissociation of the [M + H - HzOl+ ion at
Photooxidation of @-Carotene
545 1500 [M+H-H201+ 553 0
YO Y 0
W Q
1200 900 600 -
300 -
,i5 0
100
200
300
m/z
400
0 -
500
Figure 3. LC-MS/MS collision-induced dissociation product ion spectrum of [M + H - HzOl+ (mlz 553) for product 7.
Scheme I Photosensitizer
+
Photosensitizer
*
\
Type II
Free Radicals
D-Carotene Autoxidation Products
’0,
D-Carotene 1 0 2 Oxidation Products
Physical Quenching
mlz 553 formed by ionspray yielded product ions at mlz 335,269, and 243, which resulted from cleavages along the alkyl chain (Figure 3). Shifting of the allylic double bonds must occur during fragmentation, as direct cleavage of double bonds is unlikely. Identification of Other Products. Because products of @-carotenephotooxidation were recovered by evaporation of the reaction mixture in vacuo, volatile products may have been lost. To detect volatile oxidation products, aliquots of the reaction mixture were analyzed directly by GC-MS. Reaction mixtures from photooxidations in a sealed reaction flask contained a volatile product that was identified as @-ionone6 on the basis of its retention time and mass spectrum, both of which were identical to those of an authentic standard. Because retinal (@-apo-15-carotenal,vitamin A aldehyde) has been identified in previous studies of @-carotene autoxidation (27,28), we investigated whether retinal or its reduction product retinol were formed in our photooxidation system. Fractions of the product mixture with reverse-phase HPLC retention times corresponding to standard samples of retinal and retinol were collected and reanalyzed by normal-phase HPLC. Comparison of these normal-phase chromatograms to those of standards indicated that detectable amounts of neither retinal nor retinol were present in the reaction product mixture.
Figure 4. AMVN-initiated formationof 5,6-epoxy-j3,fi-~arotene (0) and 15,15’-epoxy-j3,j3-~arotene ( 0 )in the presence (dashed lines) and absence (solid lines) of 8 pM a-tocopherol.
Singlet Oxygen Dependence of Photooxidation Products. In photooxidation reactions, excited-state photosensitizers can produce free radicals by type I reactions and singlet oxygen by type I1 reactions (17). Under conditions where direct reactions with the excited photosensitizer are negligible, three different types of reactions may occur during @-carotenephotooxidation (Scheme I). @-Carotenemay physically quench singlet oxygen (pathway C) or it may be oxidized either by singlet oxygen (pathway B) or by free radicals (pathway A) .Thus, photosensitized @-caroteneoxidation may involve free radical-mediated @-carotene autoxidation rather than oxidation by singlet oxygen. To determine whether autoxidation accounted for any of the observed products, the photooxidation of @-carotenewas repeated in solutions containing an equimolar amount of a-tocopherol, an efficient inhibitor of free radical-dependent autoxidation (29). Singlet oxygen quenching by a-tocopherol in this experiment would be minimal because the rate constant for this reaction (1.48 X lo8 M-I 8-l) is almost 2 orders of magnitude less than that for singlet oxygen quenching by @-carotene(1.2 X 101oM-ls-l) (17). To test the assumption that a-tocopherol would inhibit peroxyl radical-dependent @-caroteneautoxidation in solution, we first studied the effect of a-tocopherol on AMVN-initiated autoxidation of @-carotene in toluene/methanol solution at 37 “C. Reaction products from oxidations with or without a-tocopherol were analyzed by reverse-phase HPLC. AMVNinitiated oxidation of @-carotenealone in toluene/methanol formed 5,6-epoxy-@,@-carotene and 15,15’-epoxy-@,@-carotene, which had been identified as products of similar oxidations in hexane (14). a-Tocopherol inhibited the formation of @-carotene5,g-epoxideand 8-carotene 15,15’epoxide for up to 3 h (Figure 4). Thus, a-tocopherol inhibited free radical-dependent @-caroteneautoxidation under conditions similar to those employed in photooxidation experiments. In addition, AMVN-initiated @-carotene autoxidation in the absence of a-tocopherol did not yield detectable amounts of the products (2-5) produced by photosensitized oxidation of @-carotene. After confirming that a-tocopherol inhibited peroxyl radical-dependent autoxidation under these conditions, we investigated the effect of equimolar a-tocopherol on photosensitized @-carotene oxidation. @-Carotene was photooxidized with or without equimolar a-tocopherol, and aliquots of the reaction mixture were analyzed by reverse-phase HPLC. Peak areas of products 3-5 (normalized to the photosensitizer peak area, which remained constant throughout the experiment) were compared with
546 Chem. Res. Toxicol., Vol. 6,No. 4, 1993 Table 11. Effect of a-Tocopherol on Formation of &Carotene Photooxidation Products.
product @-apo-lO’-carotenal3 &apo-8’-carotenal4 &carotene 5,8-endoperoxide 6
relative peak area 1 equiv of no a-tocopherol a-tocopherol 3.57 f 0.03 2.69 f 0.01 2.62 f 0.03 2.09 f 0.01 0.59 f 0.01 0.28 f 0.01
a Photooxidations of &carotene were done as described under Experimental Procedures. Reaction mixtures contained either no a-tocopherol or 1 mol of a-tocopherol/mol of &carotene. Values listed are product peak areas normalized to peak area of the photosensitizer and represent the mean f standard deviation from three experiments.
those for the same products in similar oxidations without a-tocopherol (Table 11). a-Tocopherol did not decrease product yield in these experiments; indeed, a-tocopherol appeared to increase product yield slightly. These observations collectively indicate that products 3-5 were not formed by free radical-dependent autoxidation during @-carotenephotooxidation.
Discussion @-Carotenemay protect plants and mammalian tissues from photosensitized oxidative damage largely by quenching singlet oxygen. However, since photosensitized oxidations may be mediated both by singlet oxygen and by free radicals, multiple antioxidant actions of @-carotene may contribute to its protective effect (2, 3). Specific chemical markers for different types of antioxidant reactions thus could be used to monitor the occurrence of these reactions in studies of @-carotene-dependentphotoprotection. Physical quenching of singlet oxygen by @-caroteneper se (eqs 1and 2) does not oxidize @-carotene. In contrast, chemical quenching forms @-caroteneoxidation products, some of which may be specific markers for singlet oxygen quenching. Here we report that singlet oxygen oxidizes @-caroteneto carbonyl-containing chaincleavage products including @-apo-l4’-carotenal2,@-apo10’-carotenal 3, @-apo-8’-carotenal4, and @-ionone6, as well as a novel oxygen addition product, @-carotene5,8endoperoxide 5. Of the five @-carotenephotooxidation products reported here, three (@-apo-l4’-carotena12,@-apo-10f-carotena13, and @-apo-8’-carotenal4)were previously reported to be thermal degradation products of @-carotene(30).@- Apo14’-carotenal 2, @-ionone6, and several other carbonylcontaining chain-cleavage products were reported as products of @-caroteneautoxidation in carbon tetrachloride or benzene (28). Moreover, @-apo-10’-carotenal3 and @-apo-l4’-carotenal2were identified as products of both “spontaneous” and azobis(isobutyronitri1e)-initiated autoxidation of @-carotene(27). This suggested that the carbonyl products found in our photooxidation experiments may have been formed by free radical-dependent @-caroteneautoxidation rather than by singlet oxygendependent oxidation. Indeed, it seemed possible that carbonyl products may have been formed by reactions of @-carotenewith either singlet oxygen or oxygen radicals. However, two separate lines of evidence indicate that carbonyl products were not formed by autoxidation in our experiments. First, a-tocopherol did not inhibit the formation of apocarotenals 2-4 in the photooxidation system (Table 11), whereas it did inhibit the AMVNdependent autoxidation of @-caroteneunder similar con-
Stratton et al.
ditions (Figure 4). Second, we found that apocarotenals 2-4 are not formed in detectable amounts by AMVN-
initiated autoxidation in the same solvent system used for photooxidations. These observations indicate that the observed photooxidation products were not formed by a peroxyl radical-dependent mechanism and that the oxidations were mediated by singlet oxygen. Further evidence for the role of singlet oxygen as the mediator of @-carotenephotooxidation comes from the formation of 5,8-endoperoxide 5. To our knowledge, this product has not been reported previously as a photooxidation product of @-carotene. However, endoperoxides analogous to 5 have been reported as photooxidation products of retinoids and related compounds (22,31,32, 33). Product 5 may have been detected previously by Seely and Meyer, who identified a similar @-carotenephotooxidation product as 5,8-epoxy-@,@-carotene (mutatochrome) on the basis of its chromatographic properties and UVvis spectrum (21). We found that the HPLC elution and UV-vis spectrum of product 5 were virtually identical to those of authentic mutatochrome. The formation of endoperoxide 5 is typical of 1,4cycloaddition reactions of singlet oxygen with cis dienes (34). However, it is not clear what other types of singlet oxygen-dependent oxidations occurred in our photooxidation experiments. Apocarotenals 2-4 and @-ionone6 may be formed by 1,kaddition of singlet oxygen to form dioxetane intermediates at Clg, C ~ ( and Y , CI, respectively. Decompositionof the dioxetanes would then yield carbonyl fragments, as has been proposed previously (20). Although this appears to be the simplest explanation for chain cleavage, it does not explain the apparent selectivity with which photooxidation cleaves the carotenoid chain. Products of chain cleavage at other sites, most notably retinal (from cleavage of the 15,15’ double bond), were not detected. We also found no evidence for the participation of the “ene” reaction (i.e., formation of allylic hydroperoxides with double bond migration) in @-carotenephotooxidation, as no hydroperoxide-containingproducts were identified. Further work will be necessary to determine which products of singlet oxygen oxidation of @-carotenemay be useful chemical markers for singlet oxygen quenching. Endoperoxide 5 appears to be uniquely formed by singlet oxygen and has been reported only as a product of @-carotene photooxidation. In contrast, the carbonyl products identified here as photooxidation products also apparently are formed by @-caroteneautoxidation in some systems (see above). However, apocarotenals 2-4 were not formed by the AMVN-initiated autoxidation under conditions that were otherwise similar to those used in photooxidation. This suggests that product distributions for @-caroteneby both singlet oxygen and free radicals may depend on the reaction medium. Additional studies of @-carotenephotooxidation in chemically defined, biomimetic systems should provide insight into the photooxidative reactions that consume @-carotenein biological systems. Acknowledgment. We thank Peter F. Baker for performing high-resolution and electron ionization mass spectral analyses and Dr. Thomas McClure for helpful comments. This work was supported in part by USPHS Grant CA56875. S.P.S. was supported by National Institutes of Health Predoctoral Training Grant ES 07091.
Photooxidation of 8-Carotene
References (1) Krinsky, N. I. (1979)Biological roles of singlet oxygen. In Singlet
Oxygen (Wasserman, H. H., and Murray, R. W., Eds.) pp 597-641, Academic Press, New York. (2) Krinsky, N. I. (1989)Antioxidant functions of carotenoids. Free Radical Biol. Med. 7,617-635. (3) Krinsky, N.I. (1979)Carotenoid protection against oxidation. Pure Appl. Chem. 61,649-660. (4) Mathews-Roth, M. M. (1986)Beta-carotene therapy for erythropoietic protoporphyria and other photosensitivity diseases. Biochimie 68,875-W. (5) Burton, G.W., and Ingold, K. U. (1984)beta-Carotene: an unusual type of lipid antioxidant. Science 224,569-573. (6)Terao, J. (1989) Antioxidant activity of beta-carotene-related carotenoids in solution. Lipids 24,659-661. (7) Palozza, P., and Krinsky, N. I. (1991)The inhibition of radicalinitiated peroxidation of microsomal lipids by both a-tocoperol and @-carotene. Free Radical Biol. Med. 11, 407-414. (8) Stocker, R., Yamamoto,Y., McDonagh, A. F., Glazer, A. N., and Ames, B. N. (1987)Bilirubin is an antioxidant of possible physiological importance. Science 236, 1043-1046. (9) Kennedy, T. A., and Liebler, D. C. (1992)Peroxyl radicalscavenging by &carotene in lipid bilayers. Effect of oxygen partial pressure. J.Biol. Chem. 267,4658-4663. (10) Lim, B. P., Nagao, A.,Terao, J.,Tanaka, K., Suzuki,T., andTakama, K. (1992)Antioxidant activity of xanthophylls on peroxyl radicalmediatedphospholipid peroxidation. Biochim.Biophys. Acta 1126, 178-184. (11) Vile, G. F.,and Winterboum, C. C. (1988)Inhibition of adriamycinpromoted microsomal lipid peroxidation by beta-carotene, alphatocopherol and retinol at high and low oxygen partial pressures. FEBS Lett. 238,353-356. (12) Palozza, P., Moualla, S., and Krinsky, N. I. (1992) Effects of ,%carotene and a-tocopherol on radical-initiated peroxidation of microsomes. Free Radical Biol. Med. 13, 127-136. (13) Palozza, P., and Krinsky, N. I. (1992)@-Caroteneand a-tocopherol are synergistic antioxidants. Arch. Biochem. Biophys. 297, 184187. (14) Kennedy, T. A., and Liebler, D. C. (1991)Peroxyl radical oxidation of beta-carotene: formation of beta-carotene epoxides. Chem. Res. Toxicol. 4,290-295. (15) Foote, C. S.,and Denny, R. W. (1968)Chemistry of singlet oxygen. VII. Quenching by beta-carotene. J. Am. Chem. SOC.90, 62336235. (16) Di Maecio, P., Kaiser, S., and Sies, H. (1989)Lycopene as the most efficient biological carotenoid singlet oxygen quencher. Arch. Biochem. Biophys. 274,532-538. (17)Foote, C. S.(1979)Quenching of singlet oxygen. In Singlet Oxygen (Wasserman,H. H., andMurray, R. W., Eds.) pp 139-171,Academic Press, New York.
Chem. Res. Toxicol., Vol. 6, No. 4, 1993 647 (18) Hasegawa, K., Macmillan, J. D., Maxwell, W. A,, and Chichester, C. 0.(1969)Photosensitized bleaching of &carotene with light at 632.8 nm from a continuous-wave gas laser. Photochem. Photobiol. 9, 165-169. (19)Tsukida, K., Cho, S. C., and Yokota, M. (1969)Photosensitized oxygenation ofall-trans beta-carotene. Chem.Pharm. Bull. 17,17551756. (20) Schenck, G. O.,and Schade, G. (1970)Photosensitized O&ansfer in presence of &carotene under exclusionof singlet oxygen. Chimia 24,13-16. (21) Seely, G. R., and Meyer, T. H. (1971)The photosensitized oxidation of beta-carotene. Photochem. Photobiol. 13,27-32. (22) Dalle, J. P., Mousseron-Canet, M., and Mani, J. C. (1969)Photosensitized oxidation of carotenoid analogs. Bull. SOC.Chim. Fr., 232-239. (23) Isoe, S.,Hyeon, S. B., Katsumura, S., and Sakan, T. (1969)Photooxygenationof carotenoids. I. The formation of dihydroactinidiolide and j3-ionone from &carotene. Tetrahedron Lett., 279-281. (24) Vetter, W., Englert, G., Rigassi, N., and Schwieter, U. (1971) Spectroscopic methods. In Carotenoids (Isler, O., Ed.) pp 192-202, Halsted Press, New York. (25) Tsukida, K., and Zechmeister, L. (1958)The stereoisomerism of beta-carotene epoxides and the simultaneous formation of furanoid oxides. Arch. Biochem. Biophys. 74,408-426. (26) Van Berkel, G.J., McLuckey,S. A.,and G b h , G. L. (1991)Preforming ions in solution via charge transfer complexation for analysis by electrospray ionization mass spectrometry. Anal. Chem. 63,20642068. (27) Handelman, G.J., van Kuijk, F. J. G. M., Chatterjee, A., and K r h k y , N. I. (1991) Characterization of products formed during the autoxidation of 8-carotene. Free Radical Biol. Med. 10,427-437. (28) Mordi, R. C., Walton, J. C., Burton, G. W., Hughes, L., Ingold, K. U., Lindsay, D. A., and Moffatt, D. J. (1993)Oxidative degradation of @-caroteneand j3-apo-8’arotenal. Tetrahedron 49,911-928. (29) Burton, G. W., and Ingold, K. U. (1986)Vitamin E application of the principles of physical organic chemistry to the exploration of its structure and function. Acc. Chem. Res. 19,194-201. (30) Marty, C., and Berset, C. (1988)Degradation products of trans&carotene produceddwing extrusion cooking. J.Food Sci. 53,18801886. (31) Lerner, D. A., Mani, J. C., and Mousseron-Canet, M. (1970) Luminescence and photoreactivity of the vitamin A series. Quantitative studies. Bull. SOC.Chim. Fr., 1968-1974. (32) Olive, J. L., and Mousseron-Canet, M. (1969) Photosensitized oxidation of the vitamin A series. Bull. SOC.Chim. Fr., 3252-3257. (33) Oyler, A. R., Motto, M. G., Naldi, R. E., Facchine, K. L., Hamburg, P. F.,Burinsky, D. J., Dunphy, R., and Cotter, M. L. (1989) Characterization of autoxidation products of retinoic acid. Tetrahedron 45,7679-7694. (34) Gollnick, K., and Kuhn, H. J. (1979)Ene-reactions with singlet oxygen. In Singlet Oxygen (Wasserman, H. H., and Murray, R. W., Eds.) pp 287-427,Academic Press, London.
