Antioxidant Activities of the Antheraxanthin-Related Carotenoids, Antheraxanthin, 9-Cis-Antheraxanthin, and Mutatoxanthins

NOTE Antioxidant Activities of the Antheraxanthin-related Carotenoids, Antheraxanthin, 9-cis-Antheraxanthin, and Mutatoxanthins Saki Shimode, Kana Miyata, Michiko Araki, and Kazutoshi Shindo＊ Department of Food and Nutrition, Japan Women’s University, 2-8-1 Mejirodai, Bunkyo-ku, Tokyo 112-8681, JAPAN

Abstract: In this study, we investigated the antioxidant activities of antheraxanthin-related carotenoids. Antheraxanthin and 9-cis-antheraxanthin were prepared from persimmon and orange fruit, respectively, and converted to other carotenoids under acidic conditions. Resulting carotenoids were purified using preparative silica gel HPLC, and their structures were analyzed in detail by NMR spectra. Both antheraxanthin and 9-cis-antheraxanthin were found to be converted to (8R)- and (8S)-mutatoxanthin at an approximate ratio of 3:2. High antioxidant activities were observed for antheraxanthin, 9-cis-antherxanthin, 1 (8R)-mutatoxanthin, and (8S)-mutatoxanthin, with potent lipid peroxidation inhibitory and moderate O2- quenching activities.

Key words: antheraxanthin, 9-cis-antheraxanthin, mutatoxanthin, antioxidant activity

1 INTRODUCTION more, we examined lipid peroxidation inhibitory（i.e., the Violaxanthin1）and 9-cis-violaxantin2（） Fig. 1）are the most inhibition of rat brain homogenate peroxidation initiated by 2＋ 3＋ 8） abundant epoxycarotenoids biosynthesized in higher the Fe -Fe -O2 complex and progressed by lipid perox- 9） 1 plants, and these compounds primarily accumulate in their ide-derived radicals ）and O2-quenching（i.e., inhibition of acylated forms in various fruits3）. Recently, we reported methylene blue-sensitized linoleic acid photo-oxidation10）） that both violaxanthin and 9-cis-violaxanthin are converted activities of 1–4. to auroxanthins（Fig. 1）by acidic treatment with the antioxidant activities of violaxanthin, 9-cis-violaxanthin, and auroxanthins4）. Aside from violaxanthin and 9-cis-violaxanthins, anther- 2 MATERIALS AND METHODS axanthin（1）5）and 9-cis-antheraxanhin（2）6（） Fig. 1）are 2.1 Spectroscopic Analysis major epoxycarotenoids in many fruits（Fig. 1）. Both 1 and HR-ESI-MS spectra were obtained using a JMS-T100LP 2 have been reported to convert to（8R）-mutatoxanthin（3） mass spectrometer（Jeol, Tokyo, Japan）, with reserpine as and（8S）-mutatoxanthin（4）（ Fig. 1）by acidic treatment7）, an external standard. NMR spectra were measured by an and the detailed NMR structural analyses of 3 and 4 have AVANCE400（Bruker BioSpin, Karlsruhe, Germany）in 7） been also described in the previous report . On the con- CDCl3, using the residual solvent peak as the internal trary, the production quantity ratio of 3 to 4 from 1 and 2, standard（δC 77.0, δH 7.26 ppm）. and the plausible conversion mechanism from 1 and 2 to 3 and 4, have not been reported. Furthermore, the antioxi- 2.2 Preparation of antheraxanthin（1） dant activities of 1–4 have never been reported previously Six hundred grams of persimmon（Diospyros kaki）, primarily due to their unstable natures. purchased at a fruit shop in Tokyo, was cut into small In this report, we confirmed the production quantity blocks（3 cm×3 cm×3 cm, approximately）, suspended in ratios of 3 to 4, from 1 and 2, and proposed plausible saturated aqueous NaHCO3 water（400 mL）, and agitated in mechanisms of the conversion for the first time. Further- a blender for 30 s. One liter of acetone was added to the

＊Correspondence to: Kazutoshi Shindo, Department of Food and Nutrition, Japan Women’s University, 2-8-1 Mejirodai, Bunkyo-ku, Tokyo 112-8681, JAPAN E-mail: [email protected] Accepted April 18, 2018 (received for review April 5, 2018) Journal of Oleo Science ISSN 1345-8957 print / ISSN 1347-3352 online http://www.jstage.jst.go.jp/browse/jos/ http://mc.manusriptcentral.com/jjocs

977 S. Shimode, K. Miyata, M. Araki et al.

Fig. 1 The structures of violaxanthin, 9-cis-violaxanthin, antheraxanthin（1）, and 9-cis-antheraxanthin（2）.

agitated persimmon solution, and the solution was stirred flow rate: 3.0 mL/min, detection: photodiode array（PDA）, for 5 min and filtered. The remaining solid on the filter and monitored at 250–700 nm. Under these conditions, paper was then collected in a 2 L beaker. The solid was pure 1, violaxanthin, and 9-cis-violaxanthin were eluted at

added to 750 mL of CH2Cl2/acetone（2:1, v/v）and stirred for 15.0, 18.0, and 23.0 min, respectively. The amount of 2 in 15 min at room temperature to extract the carotenoids persimmon was very little, and we could not identify the （this was repeated twice）. The combined filtrate（1.5 L） corresponding peak. The peak area（at 460 nm）ratio of 1:

was concentrated into a small volume to remove CH2Cl2 violaxanthin: 9-cis-violaxanthin was approximately 0.2:1:2.

and acetone and separated into EtOAc（200 mL）and H2O （200 mL）without pH adjustment. The EtOAc layer was 2.3 Preparation of 9-cis-antheraxanthin（2） concentrated to dryness to give a crude carotenoid extract Seven hundred grams of navel orange（Citrus sinensis）, （0.19 g）containing diacyl 1, diacyl violaxanthin, and diacyl purchased at a fruit shop in Tokyo, was cut into small 9-cis-violaxanthin. blocks（one navel orange was cut into 16 equivalent To prepare 1, the carotenoid extract was saponified by pieces）, and handled the same way as described in the resuspension in 20 mL of KOH solution（5 g KOH/100 mL previous section. In the final preparative HPLC, pure 2 was

90％ EtOH）and 5 mL of CH2Cl2 and stirred for 1 h. The so- eluted at 20.0 min. Since the amount of 1 in navel orange lution was added to a separation funnel containing EtOAc was very little, we could not identify the corresponding

（200 mL）and H2O（200 mL）, and the two layers（EtOAc/ peak. The peak area（at 460 nm）ratio of violaxanthin: 2:

H2O）were shaken well. The EtOAc layer was collected and 9-cis-violaxanthin was approximately 1:3:5.

dried using anhydrous Na2SO4 and concentrated to dryness to give a red oil containing 1 and 2（180 mg）. The red oil 2.4 Acid conversion of antheraxanthin（1）and 9-cis-an- was then subjected to chromatography on a 10×2 cm theraxanthin（2）, and the puri cation of products column of Chromatorex FL60D silica gel（Fuji Silysia Compounds 1 and 2（each 2.0 mg）were dissolved in 10 Chemical Ltd., Aichi, Japan）and eluted with hexane-ace- mL of 0.1 M HCl in 50％ EtOH and stirred for 5 min at tone （3:1, v/v）（ 300 mL）. The fractions containing 1 and 2 room temperature. The solution was then transferred to a

were collected and concentrated to dryness to give a red test tube containing 60 mL EtOAc and 60 mL H2O and powder（31.0 mg）. Finally, the red powder was separated mixed well using a separatory funnel. The EtOAc layer was

using a preparative HPLC（Hitachi L-7000 HPLC system; dried over anhydrous Na2SO4 and concentrated to dryness Hitachi High-Technologies Co., Tokyo, Japan）column（250 to give a mixture of mutatoxanthin［a mixture of（8R）-mu- mm×10 mm i.d.; Cosmosil 5SL; Nakarai Tesque, Inc., tatoxanthin（3）and（8S）-mutatoxanthin（4）］in each experi-

Kyoto, Japan）, with solvent: CH2Cl2: diethyl ether（1:1, v/v）, ment. The auroxanthin mixtures obtained from 1 and 2

978 J. Oleo Sci. 67, (8) 977-981 (2018) Antioxidant Activities of the Antheraxanthin-related Carotenoids

were combined and separated by HPLC（column: Cosmosil ed in CH2Cl2: acetone（2:1, v/v）under alkaline conditions. 5 SL, 250 mm×100 mm i.d., solvent: flow rate: 3.0 mL/min, The extracts were saponified to afford crude antheraxan-

detection: PDA, monitored at CH2Cl2: diethyl ether（3:1, v/ thin（1）and 9-cis-antheraxanthin（2）. Compounds 1 and 2 v）, 250–700 nm）, and the two peaks eluting at 17.5 min were further purified by partitioning between EtOAc and

（3）, and 20.0 min（4）, respectively, were collected. The H2O, silica gel column chromatography, and preparative peak area（at 460 nm）ratio of 3 and 4 was approximately silica gel HPLC to afford pure 1（0.3 mg）and 2（1.4 mg）. 3:2. To obtain acid conversion products from 1 and 2, 1 and 2（each 2.0 mg）were each dissolved in 0.1 M HCl in 50％ 1 2.5 O2-quenching experiment EtOH and stirred for 10 min at room temperature. The so-

Eighty microliters of 25 μM methylene blue, 100 μL of lution was partitioned between EtOAc and H2O, and the 0.24 M linoleic acid, with or without 40 μL of carotenoid EtOAc layers containing the converted products were ana-

（final concentration, 1–100 μM; each dissolved in ethanol）, lyzed by silica gel HPLC using CH2Cl2: diethyl ether（3:1, v/ were added to small glass test tubes（5 mL）. Tubes were v）as the developing solvent. For both HPLC analyses for mixed well and were illuminated at 7,000 lux at 22℃ for 3 the products from 1 and 2, two peaks with similar UV-Vis h in a styrofoam box. Then, 50 μL of the reaction mixture absorbance were observed at 17.5 and 20.0 min, respec-

was removed and diluted to 1.5 mL with ethanol, and OD235 tively, and the peak area ratio of them at 450 nm was ap- was measured to estimate the formation of conjugated proximately 3:2. The two product peaks derived from 1 11） dienes . The OD235 in the absence of carotenoids was and 2 were shown to be identical by co-chromatography of 1 measured as negative control（no O2-quenching activity）, the products from 1 and 2. 1 and the O2-quenching activity of carotenoids was calculat- To isolate the products, we combined the EtOAc layers

ed from OD235 in the presence of carotenoids relative to of 1 and 2, and isolated each peak using preparative silica

this reference value. The activity was indicated as the IC50 gel HPLC and the conditions described above. We then an- value, which represents the concentration at which 50％ alyzed the peaks by HR-ESI-MS and 1D（1H and 13C）and 2D inhibition was observed. （1H-1H DQF COSY, HMQC, HMBC, and NOESY）NMR spec- troscopy. The peaks at 17.5 and 20.0 min were identified to 2.6 Inhibitory experiment of lipid peroxidation in rat brain be（8R）-mutatoxanthin（3）（ 0.5 mg）and（8S）-mutatoxan- homogenate thin,（ 4）（ 0.2 mg）, respectively, by the detailed analyses of Rat brain was homogenized according to the method of the 1D and 2D NMR spectra based on the previously re- Kubo et al.12）, with some modifications. Frozen rat brains ported assigned 1H and 13C NMR data7）. （Wistar, 8-week-old males）, purchased from Funakoshi The present study is the first report showing the direct （Tokyo, Japan）, were defrosted in ice-cold 0.1 M phos- evidence of the production of 3 and 4 by acid treatment of phate buffer at pH 7.4. Then, 0.8 g of the brain was mixed 1 and 2, whereas these conversions has been suggested in with 30 mL of ice-cold phosphate buffer for 2 min in a a previous report7）. This is also the first study reporting the Teflon homogenizer. Two hundred microliters of the ho- production ratio of 3 and 4（3:2）from both 1 and 2. A plau- mogenate, 0.6 mL of 0.1 M phosphate buffer at pH 7.4, 0.1 sible reaction mechanism for the conversion of 1 and 2 to 3 mL of 1 mM sodium ascorbate, and 50 μL of carotenoid so- and 4 is shown in Fig. 2. The cation generated at the lutions dissolved in methanol（final concentration, 0.1–100 epoxide oxygen by protonation conjugates to the olefin μM）were added to small glass test tubes（5 mL）and mixed structure in the intermediate. Then, the 9-cis olefin struc- well. Tubes were incubated at 37℃ for 1 h under reciprocal ture isomerizes to a more stable 9-trans olefin. This finding agitation. Malondialdehyde was produced in the reaction indicated that 1 and 2 convert to 3 and 4 in the stomach mixture according to the concentrations of the lipid perox- after ingestion, and also implies that all the other cis ides; the reaction with thiobarbituric acid was used to isomers of antheraxanthin will be converted to 3 and 4 by

quantify the amount by OD532. The percent inhibition was acid treatment. calculated as follows:［1-（T-B）（/ C-B）］× 100（％）where T, We also measured the lipid peroxidation inhibitory and 1 C, and B were the OD532 readings of the treated carotenoid, O2-quenching activities of 1–4 and compared them with the control（peroxidation with no carotenoid）, and the ze- the activities of β-carotene and astaxanthin. The results

ro-time control（no peroxidation without homogenate）, re- （IC50）are shown in Table 1. spectively. As shown in Table 1, 1–4 showed potent lipid peroxida-

tion inhibitory activities（IC50 0.72–1.37 μM）, which were

superior to those of β-carotene（IC50 72 μM）and astaxan-

thin（IC50＞100 μM）and almost equivalent to those of vio-

3 RESULTS AND DISCUSSION laxanthin, 9-cis-violaxanthin, and auroxanthins（IC50 0.43– Diacyl antheraxanthin in persimmon（600 g）and diacyl 2.1 μM）. The potent rat brain peroxidation inhibitory 9-cis-antheraxanthin in navel orange（700 g）were extract- activities of 1–4 may be derived from 5,6- or 5,8-epoxide

979 J. Oleo Sci. 67, (8) 977-981 (2018) S. Shimode, K. Miyata, M. Araki et al.

Fig. 2 Plausible reaction mechanism for the conversion of antheraxanthin（1）and 9-cis-antheraxanthin（2）to mutatoxanthins（3 and 4）.

Table 1 An tioxidant activities of antheraxanthin（1）, 9-cis- antheraxanthin（2）,（ 8R）-mutatoxanthin（3）, and（8S） -mutatoxanthin（4）. Lipid peroxidation 1O -quenching activity Compound inhibitory activity 2 (IC50 μM) (IC50 μM) antheraxanthin (1) 1.37 10.9 9-cis-antheraxanthin (2) 0.72 11.4 (8R)-mutatoxanthin (3) 1.1 35.4 (8S)-mutatoxanthin (4) 1.0 56.5 violaxanthin 0.46a 9.8a 9-cis-violaxanthin 0.43a 18a (8R,8’R)-auroxanthin 2.1a ＞100a (8R,8’S)-auroxanthin 0.79a ＞100a b-carotene 72a 12a astaxanthin ＞100a 1.2a a Cited from the data shown in ref 4. structures which scavenge radical species13）, considering 1–5 are in progress. the structural differences among 1–4, violaxanthin, 9-cis- violaxanthin, auroxanthins, β-carotene, and astaxanthin. 1 Concerning O2-quenching activities, 1 and 2 showed weaker activities（IC50 10.9 and 11.4 μM, respectively）than ASSOCIATED CONTENT astaxanthin（IC50 1.2 μM）, whereas the activities were The authors declare no competing financial interests. equivalent to violaxanthin and 9-cis-violaxanthin（IC50 9.8 1 and 18 μM, respectively）. The weak O2-quenching activities of 1 and 2 compared with astaxanthin may be ex- plained by the decrease in conjugated double bond REFERENCES 14） 1 nmbuer . The further weak O2-quenching activities of 3 1） Kuhn, R.; Winterstein, A. Viola-xanthin, das xantho- and 4（IC50 35.4 and 56.5 μM, respectively）may be ex- phyll des gelben stiefmütterchens（Viola tricolor） plained in the same way. Further biological evaluations of （Über konjugierte Doppelbindungen, XVI.）. Berl.

980 J. Oleo Sci. 67, (8) 977-981 (2018) Antioxidant Activities of the Antheraxanthin-related Carotenoids

Deut. Chem. Ges. 64, 326-332（1931）. lipid peroxidation. Evidence against a requirement for 2） Molár, P.; Szabolcs, J.; Radics, L. Isolation and configu- an iron（II）-iron（III）complex. Biochem. J. 258, 617- ration determination of mono- and di-cis-violaxan- 620（1989）. thins. Magy. Kem. Foly. 93, 122-128（1987）. 9） Buettner, R.G. The pecking order of free radicals and 3） Müller, H. Determination of the carotenoid content in antioxidants: lipid peroxidation, α-tocopherol, and selected vegetables and fruit by HPLC and photodiode ascorbate. Arch. Biochem. Biophys. 300, 535-543 array detection. Z. Lebensm. Unters. For. 204, 88-94 （1993）. （1997）. 10） Kobayashi, M.; Sakamoto, Y. Singlet oxygen quenching 4） Araki, M.; Kaku, N.; Harada, M.; Ando, Y.; Yamaguchi, ability of astaxanthin esters from the green alga Hae- R.; Shindo, K. Production of auroxanthin from violax- matococcus pluvialis. Biotechnol. Lett. 21, 265-269 anthin and 9-cis-violaxanthin by acidic treatment and （1999）. the antioxidant activities of violaxanthin, 9-cis-violax- 11） Hirayama, O.; Nakamura, K.; Hamada, S.; Kobayashi, Y. antin, and auroxanthins. J. Agri. Food Chem. 64, Singlet oxygen quenching ability of naturally occurring 9352-9355（2016）. carotenoids. Lipids 29, 149-150（1994）. 5） Karrer, P.; Oswald, A. Carotenoids from the anthers of 12） Kubo, K.; Yoshitake, I.; Kumada, Y.; Shuto, J.; Naka- Lilium tirinum. A new carotenoid: Antheraxanthin. mizo, N. Radical scavenging action of flunarizine in rat Helvetica Chimica Acta 18, 1303-1305（1935）. brain in vitro. Arch. Int. Pharmacodyn. Ther. 272, 6） Curl, A.L.; Balley, G.F. Polyoxygen carotenoids of Va- 283-295（1984）. lencia orange juice. J. Agric. Food Chem. 2, 685-690 13） deAlmeida, A.A.; de Carvalho, R.B.; Silva, O.A.; de （1954）. Sousa, D.P.; de Freitas, R.M. Potential antioxidant and 7） Molnár, P.; Deli, J.; Matus, Z.; Tóth, G.; Steck, A.; anxiolytic effects of（＋）-limonene epoxide in mice af- Pfander, H. Isolation and characterization of mutato- ter marble-burying test. Pharm. Biochem. Be. 118, xanthin-epimers from red paprika（Capsicum annu- 69-78（2014）. um）. Eur. Food Res. Technol. 211, 396-399（2000）. 14） Shimidzu, N.; Goto, M.; Miki, W. Carotenoids as singlet 8） Aruoma, O.I.; Halliwell, B.; Laughton, J.M.; Quinlan, oxygen quenchers in marine organisms. Fish. Sci. 62, J.G.; Gutteridge, M.C.J. The mechanism of initiation of 134-137（1996）.

981 J. Oleo Sci. 67, (8) 977-981 (2018)