C-D-glucopyranosyl derivatives of tocopherols - Synthesis and evaluation as amphiphilic antioxidants

FULL PAPER

DOI: 10.1002/ejoc.200700885

C-D

-Glucopyranosyl Derivatives of Tocopherols – Synthesis and Evaluation as

Amphiphilic Antioxidants

Li He,^[a,b]Stéphanie Galland,^[c]Claire Dufour,^[c]Guo-Rong Chen,*^[b]Olivier Dangles,*^[c]

Bernard Fenet,^[d]and Jean-Pierre Praly*^[a,e,f,g]

Keywords: C-Glycosides / Tocopherol / Chromanol / Antioxidant / Lipid peroxidation

Treatment of dimethylhydroquinone dimethyl ethers (ortho

and meta isomers) with glycopyranose pentaacetates (

gluco, -galacto) in the presence of SnCl₄and F₃CCO₂Ag

selectively afforded the corresponding C-β- -glycosyl deriv-

atives by aromatic electrophilic substitution. Oxidation of the

dimethoxybenzene moiety with ceric ammonium nitrate de-

phytol led to acetyl-protected C-β-

D-glycosyl chromanols or

D-

C-β- -glycosyl tocopherols, the sugar residues of which were

D

deacetylated under base catalysis conditions. These new

molecules were evaluated as antioxidants in terms of their

ability to inhibit the peroxidation of linoleic acid in SDS mi-

celles. The position of the C-glucosyl moiety on the phenolic

nucleus emerges as the critical structural determinant of their

activity.

D

livered C-β-

D-glycosyl-dimethylbenzoquinones, which were

reduced with Na₂S₂O₄to the corresponding C-β-

D-glycosyl-

dimethylhydroquinones. ZnCl₂-catalyzed cyclization either

with methylbut-2-en-1-ol (prenyl alcohol) or with all-racemic

Germany, 2008)

noted by the lettering α, β, γ or δ. The four tocopherols

have asymmetric carbons at positions 2, 4Ј and 8Ј

(Scheme 1), while in the case of the four tocotrienols only

one chiral centre – at C2 – exists. Such naturally occurring

lipophilic molecules, practically insoluble in aqueous media,

have long been known as antioxidants, primarily as a func-

tion of the reducing properties of the phenolic nucleus.

Among them, α-tocopherol is the most abundant, and also

the most biologically active. Recent progress has shed more

light on the mechanisms of its biological activity.^[1]Natural

α-tocopherol, with three methyl groups on the ring and

three (R)-configured chiral centres, is recognized well by α-

tocopherol transfer protein (α-TTP) the binding pocket of

which can accommodate α-tocopherol with a U-shaped

phytyl chain. Vitamin E was originally claimed to be the

fertility vitamin, after it was demonstrated that the embryos

of α-TTP–/– female mice died at mid-gestation, while an α-

tocopherol-supplemented diet allowed full-term pregnancy.

Nowadays, the roles of vitamin E in humans, animals and

plants are all better understood, and they appear quite di-

verse and complex in view of the fact that these dietary

Introduction

Vitamin E constitutes a family of eight related molecules,

in which a chromanol ring bears either a saturated C₁₆H₃₃

phytyl chain (tocopherols) or an unsaturated isoprenoid

chain (tocotrienols). The chromanol ring displays one to

three methyl groups with different substitution patterns, de-

[a] Université Lyon 1,

69622 Lyon, France

[b] Laboratory of Advanced Materials, Institute of Fine Chemicals,

East China University of Science and Technology (ECUST)

130 Meilong Road, P. O. Box 257, 200237 Shanghai, P. R.

China,

Fax: +86-21-64252758

E-mail: mrs_guorongchen@ecust.edu.cn

[c] INRA, Université d’Avignon, UMR A 408 Sécurité et Qualité

des Produits d’Origine Végétale, Domaine St Paul, Site Agro-

parc

84914 Avignon, France

Fax: +33-4-90144441

E-mail: Olivier.Dangles@univ-avignon.fr

[d] Université Claude Bernard Lyon 1, Centre Commun de RMN,

CPE-Lyon, bâtiment 308

43 Boulevard du 11 Novembre 1918, 69622 Villeurbanne,

France

[e] Institut de Chimie et Biochimie Moléculaires et Supramolécul- molecules have both antioxidant and pro-oxidant properties

aires (ICBMS), Laboratoire de Chimie Organique 2, CPE-

and may be involved in the prevention of atherogenesis and

Lyon, bâtiment 308

cancers.^[2,3]A lot of experimental or theoretical work has

43 boulevard du 11 novembre 1918, 69622 Villeurbanne, France

Fax: +33-4-72448349

thus recently been devoted to investigating the activity of

phenolic antioxidants and the possible influence of co-anti-

oxidants.^[4,5]While the beneficial effects of supplementation

with vitamin E have fuelled many debates and stimulated

intensive academic and industrial investigations, how the

protective antioxidants act in vivo has raised many ques-

E-mail: jean-pierre.praly@univ-lyon1.fr

[f] CNRS, UMR5246

69622 Villeurbanne, France

[g] CPE Lyon

69616 Villeurbanne, France

Supporting information for this article is available on the

WWW under http://ww w .eurjoc.org or from the author.

Eur. J. Org. Chem. 2008, 1869–1883

1869

G.-R. Chen, O. Dangles, J.-P. Praly et al.

FULL PAPER

tions, in particular with regard to the exact location and

It is worth mentioning ongoing efforts being made

orientation of antioxidant molecules at the molecular level. towards the synthesis and evaluation of novel compounds

For instance, it is presumed that the regeneration of α-toc- designed as inhibitors of lipid peroxidation with protective

opherol by vitamin C (ascorbate) is made possible by the effects against myocardial ischemia-reperfusion damage.^[18]

location of the tocopheroxyl radical adjacent to the inter- From a series of molecules containing lipoic acid and trolox

face, so as to abstract a hydrogen atom or an electron from connected through spacers^[19]and examined for their anti-

ascorbate.^[6]

oxidant activity and their protective effects against reper-

fusion arrhythmias, it was found that the 2- and the 5-sub-

stituted chromanols exhibited the better cardioprotective

activity.^[20]Other 5-substituted analogues were also synthe-

sized by connecting the chromanol ring to methylsulfo-

nylaminophenyl residues through tertiary amine moieties (a

combination of pharmacophores identified for the most

active class III antiarrhythmics), so as to obtain new com-

pounds that were evaluated with regard to their antiar-

rhythmic and antioxidant activity.^[21]In an extension of this

approach, novel hybrids combining the pharmacophoric re-

dox moieties of vitamin E and motifs responsible for the

antiarrhythmic properties of the class I antiarrhythmics

procainamide and lidocaine were prepared and evaluated.

The tests suggested that the efficacy of the new compounds

in preventing reperfusion arrhythmias could be attributed

Scheme 1. Racemic α-tocopherol and glucosylated derivatives.

In this context, the chemical synthesis of antioxidants, to their combined effects involving inhibition of free-radi-

possibly with improved properties, is an active field. The cal-mediated damage coupled to antiarrhythmic proper-

easily available tocopheryl bromide, for example, has re- ties.^[22]

cently been converted into the corresponding phosphonium

Because ascorbic acid and tocopherols are hydrophilic

salt, further elaborated into various furotocopheryl deriva- and lipophilic molecules, respectively, modified amphiphilic

tives or coupled onto modified polystyrene to provide a new antioxidants have been synthesized with the goal of im-

vitamin-E-loaded resin.^[7]In other approaches, the 5a- proving their accessibility towards various media.^[23]While

azido-tocopheryl derivative, also prepared from tocopheryl novel ascorbic acid derivatives (6-O-acyl-2-O-α--glucopyr-

bromide,^[8]and its 5a-nitro analogue^[9]have provided access anosyl--ascorbic acids) have been evaluated as lipophilic

to tocopheryl-1,2,3-triazoles or -isoxazolines through cyclo- topical prodrugs of ascorbic acid (with transdermal activity

addition reactions. Interestingly, the tocopheryl-1,2,3-tri- in a human living skin equivalent model),^[24]other studies,

azoles might undergo base-induced cleavage of the toco- following an opposite strategy, have involved the modifica-

pheryl moiety, with potential applications in drug-delivery tion of lipophilic chromanols and tocopherols for the prep-

systems. Condensation of trimethylhydroquinone with alde- aration of water-soluble molecules with antioxidant activi-

hydes provided 3-oxachromanol-type antioxidants,^[10]which ties in aqueous media [e.g., the commercially available tro-

were comparatively studied with α-tocopherol and penta- lox (COOH-substituted chromanol) or its reduced deriva-

methylchromanol, as well as with ubichromanol, ubichro- tive (CH₂OH-substituted chromanol)]. After glycosylation,

menol and twin-chromanol. The last of these was found to the latter compound afforded a water-soluble O-alkyl glyco-

be a promising new candidate as an artificial antioxidant in side^[25]showing radical scavenging properties comparable to

biological systems.^[11]all-rac-α-Selenotocopherol has been those of tocopherol, trolox and ascorbic acid, or even

synthesized and found to be a slightly less potent antioxi- higher under specific conditions.^[26]Oligosaccharides of toc-

dant than α-tocopherol, in keeping with the value of the opherols (O-aryl glycoside type; see Scheme 1) represent an-

OH bond dissociation energy, which was found to be ap- other class of water-soluble derivatives, the biological appli-

proximately 1 kcalmol^–1higher than that of α-tocoph- cability of which, however, is dependent on deglycosidation

erol.^[12]Moreover, phenolic compounds with one or two ni- in vivo as the free phenoxyl group is essential for the redox

trogen atoms in the aromatic ring (e.g., 3-pyridinols,^[13]6- properties of tocopherols. Their synthesis was achieved in

amino-3-pyridinols,^[13]5-pyrimidinols^[14]and analogues) are good yield by glycosidation between oligosaccharide per-

extremely effective chain-breaking antioxidants in homo- acetates and tocopherol with BF₃·etherate as catalyst.^[27]

geneous organic solutions, a tetrahydro-1,8-naphthyridin-3- Hence, all-rac-α-tocopheryl 2,3,4,6-tetra-O-acetyl-β--gluc-

ol being almost 30 times more effective than α-tocopherol opyranoside was obtained in 66% yield and was then dea-

at inhibiting lipid peroxidation.^[13,15]Chemical modifica- cetylated (97% yield). From benzyl-protected α--glucopyr-

tions of the alkyl chain have also been considered, as shown anose trichloroacetimidate as glucosyl donor, the two-step

by the synthesis of tocopherol fatty alcohols as microglial synthesis was achieved in 67% overall yield.^[28]All the dea-

activation modulators,^[16]or the preparation of new chro- cetylated tocopheryl oligosaccharides prepared were soluble

man/catechol hybrids, evaluated against oxidative stress-in- in DMSO or pyridine, while their solubilities in CH₂Cl₂,

duced cellular damage and as neuroprotectants.^[17]

MeOH or H₂O depended on the size of the sugar resi-

1870

www.eurjoc.org

Eur. J. Org. Chem. 2008, 1869–1883

Amphiphilic Tocopherol-Derived Antioxidants

due.^[27a]Glycosidation of α-tocopherol and analogues by tures should convey interesting chemical and biological

cultured plant cells has also been investigated.^[29]

properties, in view also of the diverse bioactivities of qui-

This account reports on synthetic routes to C-(β--glyco- noid compounds.^[37]In particular, we found that C-glucopy-

pyranosyl)chromanols and C-(β--glycopyranosyl)toco- ranosyl-hydro- and -benzoquinones were modest inhibitors

pherols (-gluco, -galacto series) with acetylated or deacet- of glycogen phosphorylase, as shown by enzymatic and

ylated sugar moieties. These new C-glucosylated analogues crystallographic studies.^[34b]This study provided a basis for

of vitamin E are then evaluated for their ability to inhibit further elaboration of a new scaffold for obtaining more

the peroxidation of linoleic acid within SDS micelles.

effective inhibitors of the enzyme, while nitration of 2-

(tetra-O-acetyl-β--glycopyranosyl)-1,4-dimethoxybenzenes

led, after reduction and acylation of the resulting amino

groups, to a series of 5-N-amido derivatives.^[38]After pre-

liminary attempts at cyclization between C-β--galactopyr-

anosyl-hydroquinone and all-racemic phytol had afforded

low yields of regioisomers related to C--galactopyranosyl

tocopherol (see Supporting Information), we considered C-

glycopyranosyl-dimethylhydroquinones (o-, m- and p iso-

mers) as appropriate building blocks for preparing C-glyco-

syl-chromanols and C-glycosyl-tocopherols upon annel-

ation with 3-methylbut-2-en-1-ol (prenyl alcohol) or all-ra-

cemic phytol. In line with our previous syntheses,^[34]electro-

philic substitution reactions between dimethyl-substituted

1,4-dimethoxybenzenes (o, m, or p isomers, 2–4; Scheme 2)

and appropriate glycosyl donors were considered a key step

Results and Discussion

1. Synthetic Results

C-Glycosyl arenes (also frequently termed aryl-C-glyco-

sides; see Scheme 1)^[30]and C-glycosyl flavonoids^[31]are

continuously attracting attention, because many of them

are bioactive natural compounds. This has stimulated inter-

est with regard to their biosynthesis and their chemical syn-

thesis.^[32,33]Our recent synthesis of C-glycopyranosyl-hy-

dro- and -benzoquinones^[34]via the known C-glycopyrano-

syl-1,4-dimethoxybenzenes^[35]highlighted a poorly known

route^[36]to small molecules that have not received much at-

tention, although their stabilities and polyfunctional struc-

Scheme 2. Synthesis of C-β--glucosyl-chromanols and C-β--glucosyl-tocopherols.

Eur. J. Org. Chem. 2008, 1869–1883

www.eurjoc.org

1871

G.-R. Chen, O. Dangles, J.-P. Praly et al.

FULL PAPER

towards the desired -glycosylated dimethyl-substituted corresponding to the (4ЈRS,8ЈRS) diastereoisomers present

chromanols and tocopherols. Methylation of commercially in (2R)-13 and (2S)-13. Similar observations relating to the

available 2,3-dimethyl-hydroquinone easily afforded the o- choice of solvent were made about α-tocopheryl oligosac-

dimethyl isomer 2. The dimethyl-substituted 1,4-dimethoxy- charides when investigated by NMR, since CD₃OD or [D₆]

benzenes 3^[39]and 4^[40]were prepared in three steps (Fre- DMSO appeared to be suitable, in contrast to D₂O, perhaps

my’s salt oxidation, reduction, methylation) from 2,6-di- due to the existence of supramolecular structures as mi-

methyl- and 2,5-dimethylanisole by adaptation of reported celles.^[27a]Indeed, various sugar-based motifs can be found

methods.

among amphiphilic molecules that produce nanotube archi-

Electrophilic substitution of the dimethyl-substituted 1,4- tectures.^[41]However, well resolved spectra can be interpre-

dimethoxybenzenes 2–4 with β--glucopyranose peracetate ted in terms either of the presence of rotamers in fast equi-

(for the -galacto analogues, see Supporting Information) librium (free rotation) or, for the chromanols 10–13, of the

was accomplished on stirring the reaction mixture contain- existence of a major conformer with possible intramolecu-

ing SnCl₄and silver trifluoroacetate in CH₂Cl₂at ca 28 °C lar H-bonding. In the γ-tocopherol derivatives 12 and 13,

for 4/5 h, as reported.^[34]Although differing significantly, the minimum-energy conformers should, for steric reasons,

the yields recorded for the C-glucosyl compounds 5–7 could have the anomeric hydrogen in a staggered position with

be interpreted on the basis of minimized steric interactions respect to the C4 methylene of chromanol. This arrange-

in the substitution transition states (2, less hindered than 3 ment, which was supported by strong NOESY and

and 4), and on the contributions of the methyl groups in CROESY correlations between the anomeric hydrogen and

stabilizing the cationic Wheland intermediates (best stabi- H4a and H4b, should place the phenoxyl group and the

lized for 3, relative to 2 and 4). The reaction with 4 (higher ring oxygen at a short distance and favour their intramolec-

hindrance, lower reactivity) produced an α/β mixture in ular H-bonding. For compounds 13, the configuration at

modest yield (10%), which was increased to ca. 17% with C2 was tentatively assigned^[34a]from the measured optical

4 in larger excess (4 equiv.). Because of its lack of reactivity rotations, found to be +42.4 [(2R)-13] and +28.5 [(2S)-13].

and stereoselectivity, this last sequence was abandoned.

With β--galactopyranose pentaacetate (see Supporting In-

formation), the yield was 76% for the coupling to 2 [72%

for the desired -galacto analogue of 5, together with 4%

of 1,4-dimethoxy-2,3-dimethyl-5-(3,4,6-tri-O-acetyl-β--ga-

lactopyranosyl)benzene], while 5 (-gluco) was obtained in

67% yield. This again showed the influence of the glycosyl

donor configuration on the outcome of the electrophilic

aromatic substitution, with a higher selectivity in the -gal-

acto series than in the -gluco series.^[34,35]

When condensation of 9a with racemic phytol was first

attempted with ZnCl₂in CHCl₃containing EtOH as stabi-

lizer (0.6%), the expected product 12 was formed in a some-

what lower yield (61% instead of the 66% optimized yield)

because of the formation of 1-ethoxy-4-hydroxy-2,3-di-

methyl-5-(2,3,4,6-tetra-O-acetyl-β--glucopyranosyl)ben-

zene(9b) in 27% yield. This is reminiscent of the formation

of a methoxy derivative upon acid-catalyzed deacetylation

Scheme 3. Structures of compounds 11 and 13 with NMR number-

ing.

of (tetra-O-acetyl-β--glucopyranosyl)benzoquinone in

In the meta-dimethyl series, only the acetylated benzo-

MeOH acidified with AcCl.^[34b]Anhydrous ethanol-free quinone 14 and the hydroquinone 15 gave well-resolved

CHCl₃was therefore used for preparing the glycosyl-chro- spectra in CDCl₃at ambient temperature. The other com-

manols and glycosyl-tocopherols from 9a and 15 upon an- pounds required heating in [D₆]DMSO (6/90 °C; 16 and 18/

nelation with prenyl alcohol or racemic phytol, with anhy- 120 °C, 19/90 °C – spectra not quite clear) or [D₆]DMSO +

drous ZnCl₂as catalyst.

D₂O (17/90 °C). In relation to 5 (see above), restricted rota-

NMR spectroscopy provided a basis for assigning the tion about the glycosidic bond due to steric hindrance could

structures of the prepared compounds, in which the carbon explain the need for heating above the coalescence tempera-

atoms were numbered as shown in Scheme 2 and Scheme 3. ture in order to record well resolved spectra for 6 and 7

In the ortho-dimethyl series derived from 2, all compounds (para-dimethyl). The sterically less demanding structure of

gave well-resolved NMR spectra at room temperature in 14, in relation to 6, might account for the quality of its

CDCl₃(acetylated molecules) or [D₆]acetone (deacetylated spectra in CDCl₃at ambient temperature (free rotation),

structures). For (2R)-13 and (2S)-13, each of which features while for 15, an intramolecularly H-bonded atropoisomer

a hydrophilic sugar residue and a lipophilic chain, use of can be envisaged. While the eight chromanol derivatives ap-

CDCl₃led to badly resolved spectra, as already noted for a pear to be sterically hindered, it is difficult to classify them

O-glucoside,^[27a]but [D₆]acetone was found to be a suitable in terms of hindrance based on a simple analysis. It seems

solvent at ambient temperature, showing smaller signals reasonable to assume that their conformations also depend

1872

www.eurjoc.org

Eur. J. Org. Chem. 2008, 1869–1883

Amphiphilic Tocopherol-Derived Antioxidants

on intramolecular H-bonding between the sugar and chro-

manol moieties. Obviously, H-bonding between the pheno-

lic OH and the pyran oxygen is possible only for 10–13.

Natural tocopherols display a methyl group at C8 and

either a methyl group or hydrogen atom at C5 and C7, be-

ing classified as α (Me at C5 and C7), β (Me at C5, H at

C7), γ (H at C5, Me at C7) or δ (H at C5 and C7). Because

18 and 19 have two methyl substituents adjacent to the phe-

nolic hydroxy group in a nonclassical arrangement that can

be denoted ε, we propose to regard these compounds as

derivatives of ε-tocopherol.

Scheme 4. Mechanism of the peroxidation of linoleic acid (LH) in-

duced by a diazo compound (R–N=N–R) in the presence of an

antioxidant (AH).

2. Evaluation as Antioxidants

Chain-breaking antioxidants such as α-tocopherol ef-

Polyunsaturated fatty lipids are sensitive to oxidation in ficiently inhibit propagation. Consequently, the curves fea-

food during processing, storage or domestic treatments, this turing the accumulation of lipid hydroperoxides (LOOH,

phenomenon being responsible for a loss of sensorial and spectroscopic monitoring at 234 nm through the conjugated

nutritional quality.^[42,43]Polyunsaturated fatty lipids are 1,3-dienyl moiety) display a well-defined lag phase during

also important targets of oxidative stress in living tis- which the peroxidation is very slow before sharply resuming

sues.^[44–49]Their oxidation products, in particular hydroper- once the antioxidant has been consumed. In contrast, pol-

oxides and aldehydes, are potentially toxic and involved in yphenols, although possibly better radical scavengers than

chemical modifications of proteins and nucleic acids that α-tocopherol (both in terms of rate constant and stoichiom-

may initiate events in the development of cancers and ar- etry), inhibit the peroxidation of linoleic acid without a lag

therosclerosis. In addition, lipid peroxidation is particularly phase. This observation was quantitatively interpreted by

pernicious, since it takes place by a radical-chain mecha- assuming that polyphenols essentially act as initiation in-

nism in which a single initiating species can trigger the ac- hibitors.^[50]

cumulation of several equivalents of hydroperoxides upon

From the time dependence of the LOOH concentration,

repetition of the propagation step. The inhibition of lipid it is clear that the tocopherol analogues in which the C-

peroxidation is therefore a major concern in food and glycosyl moiety is meta to the phenolic OH group (16–19)

health research and forms the basis of common antioxidant inhibit the peroxidation of linoleic acid with a lag phase,

tests. In particular, antioxidant tests dealing with biphasic whereas no significant lag phase is apparent with the toc-

systems (micelles, liposomes, emulsions) aimed at modelling opherol analogues in which the C-glycosyl moiety is ortho

of cell membranes or food emulsions afford the opportunity to the phenolic OH group (10–13; Figure 1). Consistently,

to compare antioxidants with similar H-donating or elec- when the ratio of the initial rate of inhibited peroxidation

tron-donating capacities (e.g., the tocopherol analogues (R_p) to the constant rate of uninhibited peroxidation (R⁰_p)

synthesized in this work) but with distinct hydrophilic–lipo- is plotted vs. the total antioxidant concentration, a much

philic balances and thus expected to partition differently sharper decrease is observed with 16–19 (Figure 2). From

between the aqueous and lipid phases.

those curves, IC₅₀values (antioxidant concentration giving

In this work, α-tocopherol and its C-glucosyl analogues R_p/R⁰_p= 1/2) can be extracted (Table 1).

were investigated for their abilities to inhibit the peroxi-

Tocopherol analogues 16–19 are better inhibitors (i.e.,

dation of linoleic acid (LH, H refers to one of the labile bis- giving lower IC₅₀values than 10–13). Rather unexpectedly,

allylic H-atoms) within SDS micelles in a neutral phosphate the relative positions of the C-glucosyl and OH groups on

buffer and under a constant flow (chemical rate R_i) of ini- the aromatic ring emerge as the critical determinant of the

tiating peroxyl radicals (ROO^·) formed by the thermal de- antioxidant activity, whereas the presence or absence of the

composition of the water-soluble diazo compound AAPH phytyl chain plays only a minor role. As a general trend,

[R–N=N–R, R = Me₂C–C(=NH₂⁺)–NH₂]. The mechanism acetylation of the glucosyl moiety slightly enhances the in-

of lipid peroxidation and its inhibition by antioxidants is hibitory capacity.

shown in Scheme 4. The antioxidant (AH) can act by two

For a more quantitative analysis aimed at estimating AE₁

and AE₂, the following strategy was devised.

distinct mechanisms.^[50,51]

– It can compete with LH for the initiating radicals. This – For chain-breaking antioxidants (α-tocopherol, 16–19) at

process is called inhibition of initiation and must take place low concentrations, the A(234 nm) vs. time curves display

in the aqueous phase or at the interface. Its efficiency is well-defined lag phases and are amenable to kinetic analysis

measured by AE₁= k_a1/k₁.

with AE₂, R_iand the antioxidant stoichiometry (n =

– It can compete with LH for the propagating lipid peroxyl number of peroxyl radicals trapped per antioxidant mole-

radicals (LOO^·). This process is called inhibition of propa- cule) as the adjustable parameters [see Equations (1), (2)

gation and must take place in the lipid phase. Its efficiency and (3); C_LH= total lipid concentration, (AH) = nC, C =

is measured by AE₂= k_a2/k₂.

total antioxidant concentration)].^[51]

Eur. J. Org. Chem. 2008, 1869–1883

www.eurjoc.org

1873

G.-R. Chen, O. Dangles, J.-P. Praly et al.

FULL PAPER

Figure 1. Time dependence of the LOOH concentration during uninhibited and inhibited peroxidation of linoleic acid. The concentration

of antioxidant (added at t ≈ 800 s) was about 0.8 µ.

At low antioxidant concentrations, the inhibition of initi-

ation was neglected (AE₁set at 0). With α-tocopherol, n can

be safely set at 2 on the basis of the oxidation products it

forms during the scavenging of the LOO^·radicals.^[52]Thus,

AE₂and R_ican be estimated (Table 2). An average R_ivalue

of 740 ps^–1was then used in all calculations. With 16–19,

the curve-fittings yielded optimized values of AE₂and n.

– The R_p/R⁰_pratio can be expressed as a function of AE₁,

AE₂, R_iand (AH) = nC [Equations (4) and (5)].^[50]

(1)

(2)

Figure 2. Plots of R_p/R⁰_pvs. total antioxidant concentration (n = 3

for each experimental point). The solid lines are the result of the

curve-fitting procedures (see text); ᭡ α-tocopherol, * 10, ᭹ 11, o 18,

᭿ 19.

(3)

(4)

Table 1. IC₅₀values of the antioxidants as inhibitors of the AAPH-

induced peroxidation of linoleic acid.

Antioxidant

α-Toc. 10 11 12 13 16 17 18 19

0.14 0.76 1.40 1.10 1.30 0.32 0.55 0.45 0.35

(5)

IC₅₀[µM]

1874

www.eurjoc.org

Eur. J. Org. Chem. 2008, 1869–1883

Amphiphilic Tocopherol-Derived Antioxidants

Table 2. Inhibition of the AAPH-induced peroxidation of linoleic being set constant at 0.75 (Table 3). As expected, the AE₂

acid. Kinetic analysis of the A(234 nm) vs. time curves for the

chain-breaking antioxidants (r Ͼ 0.999).

values thus estimated are quite low (AE₂Ͻ20).

From Table 2, it can be noted that the chain length of

uninhibited peroxidation (R⁰_p/R_i) lies in the 10–20 range:

that is, on average, each LOO^·can react with 10–20 LH

molecules before it recombines with another LOO^·to ter-

Antioxidant

C/nM R_p⁰/nMs^–1R_i/pMs^–1, n

AE₂

α-Tocopherol^[a]

191

381

10.1

8.7

10.0

11.7

11.5

737 (Ϯ7)

716 (Ϯ4)

728 (Ϯ5)

713 (Ϯ7)

802 (Ϯ3)

3010 (Ϯ100)

4120 (Ϯ140)

3390 (Ϯ100)

3350 (Ϯ150)

4100 (Ϯ120)

minate the chain. Moreover, under our conditions (C_SDS

=

0.1 , C_LH= 2.55 m), each micelle contains approximately

N = 60 SDS molecules^[53]and 1–2 LH molecules. Taking a

CMC value of 7 m for SDS,^[54]the micelle concentration

(C_M) can also be estimated: C_M= (C_SDS– CMC)/N ≈ 1–

2 m. Under such conditions, LOO^·must rapidly diffuse

from one micelle to another so as to propagate the radical

chain efficiently. Similar considerations apply to the antioxi-

dants themselves.

16^[b]

17^[b]

18^[b]

193

387

11.5

13.0

15.4

13.2

2.22 (Ϯ0.03)

1.61 (Ϯ0.01)

2.47 (Ϯ0.05)

1.37 (Ϯ0.01)

1.38 (Ϯ0.01)

1030 (Ϯ30)

1170 (Ϯ30)

400 (Ϯ20)

1180 (Ϯ10)

1650 (Ϯ30)

774

191

381

10.3

11.1

10.1

11.3

10.0

1.65 (Ϯ0.03)

1.67 (Ϯ0.03)

1.15 (Ϯ0.01)

1.02 (Ϯ0.01)

670 (Ϯ30)

410 (Ϯ20)

1140 (Ϯ40)

1080 (Ϯ40)

1060 (Ϯ30)

Indeed, Figure 2 shows that efficient chain-breaking anti-

oxidants at concentrations as low as 1–5 µ (i.e., at least

200 times lower than the micelle concentration) can reduce

the peroxidation rate by a factor of 10. Part of this effi-

ciency might reflect their ability to search micelles for the

propagating lipid peroxyl radicals. Hence, intermicellar dif-

fusion of the antioxidants and lipid peroxyl radicals might

be a crucial factor in the overall process of radical scaveng-

ing.^[55]For instance, the efficiency of α-tocopherol in in-

hibiting the peroxidation of linoleic acid in SDS micelles is

higher than that of trolox, which essentially partitions in

the aqueous phase, but lower than that of trolox methyl

ester, which is mainly located in the micellar phase like α-

tocopherol, but displays the additional advantage of diffus-

ing more rapidly from one micelle to another.^[55]The main

parameters governing the relative abilities of antioxidants

to inhibit lipid peroxidation in micelles are therefore: the

intrinsic reactivity of the antioxidants (H-donating or elec-

tron-donating capacity), their ability to enter the micellar

pseudo-phase with a correct orientation relative to the lipid

peroxyl radicals, and their intermicellar mobility.

762

200

400

800

13.4

12.5

16.1

13.8

15.3

12.9

2.12 (Ϯ0.03)

2.34 (Ϯ0.02)

0.83 (Ϯ0.01)

1.05 (Ϯ0.01)

1.83 (Ϯ0.02)

1.22 (Ϯ0.04)

520 (Ϯ10)

450 (Ϯ10)

650 (Ϯ10)

970 (Ϯ20)

390 (Ϯ10)

810 (Ϯ40)

19^[b]

198

396

10.9

11.1

10.6

10.4

2.37 (Ϯ0.05)

1.89 (Ϯ0.02)

1.85 (Ϯ0.02)

1.78 (Ϯ0.01)

1.83 (Ϯ0.01)

1.57 (Ϯ0.01)

940 (Ϯ50)

890 (Ϯ30)

850 (Ϯ30)

810 (Ϯ10)

840 (Ϯ10)

880 (Ϯ10)

793

1524 10.6

[a] n set at 2. [b] R_iset at 740 ps^–1.

For chain-breaking antioxidants (α-tocopherol, 16–19),

the R_p/R⁰_pvs. C curves were analyzed so as to extract AE₁

and n (Table 3), AE₂being set at a constant average value

deduced from Table 2. The low stoichiometries (in the 0.5–

1.0 range) probably point to the antioxidants being only

partially oxidized during the early stage of the peroxidation

process. For antioxidants believed (from the absence of lag

phase) to act mainly as initiation inhibitors (i.e., 10–13),

AE₁and AE₂were selected as the adjustable parameters, n

With regard to the antioxidants studied in this work, the

following remarks can be made.

– From their structural similarity, it seems reasonable to

assume that the antioxidants should display similar H-do-

nating capacities. In fact, the C-glucosyl group is likely to

deactivate the phenolic nucleus slightly through a conjunc-

tion of steric and electronic effects including the electron-

withdrawing effect of the O atoms and acetyl groups. In 10–

13 (Glc ortho to the phenolic OH), an additional deactivat-

ing effect could arise from the possible formation of an in-

tramolecular H-bond with the pyran O atom of Glc (six-

membered ring), which would be likely to enhance the bond

dissociation energy (BDE) of the phenolic OH.^[56]However,

calculations on a series of simplified structures suggest that

the replacement of a methyl group by a C-glycosyl group,

either acetylated or not and either ortho or meta to the OH

group, induces only minor changes in the BDE value

(Table 4).

Table 3. Inhibition of the AAPH-induced peroxidation of linoleic

acid. Mathematical analysis of the R_p/R⁰_pvs. total antioxidant con-

centration plots (R_iset at 740 pMs^–1).

Antioxidant

R⁰_p/ns^–1AE₁

AE₂

r

n

α-Tocopherol 10.3

8400 (Ϯ6400) 3600^[a]

14100 (Ϯ1300) 10 (Ϯ4)

0.997 0.55 (Ϯ0.11)

0.998 0.75^[b]

10

11

12

13

16

17

18

19

12.1

9.2

5800 (Ϯ500)

18 (Ϯ5)

0.998 0.75^[b]

10.9

10.2

13.0

10.7

13.3

10.0

9700 (Ϯ1200) 10 (Ϯ5)

0.995 0.75^[b]

7200 (Ϯ500)

1270 (Ϯ990)

460 (Ϯ410)

270 (Ϯ370)

1320 (Ϯ740)

7 (Ϯ2)

1090^[a]

870^[a]

630^[a]

870^[a]

0.998 0.75^[b]

0.990 0.75^[b]

– The AE₂values of chain-breaking antioxidants (16–19)

show that these vitamin E analogues react less rapidly than

α-tocopherol with the lipid peroxyl radicals (Table 2). For

the more hydrophilic 17 and 19 (deacetylated Glc), this

0.997 0.65 (Ϯ0.08)

0.993 0.92 (Ϯ0.13)

0.999 0.89 (Ϯ0.11)

[a] Mean value from Table 2. [b] Set constant.

Eur. J. Org. Chem. 2008, 1869–1883

www.eurjoc.org

1875

G.-R. Chen, O. Dangles, J.-P. Praly et al.

FULL PAPER

Table 4. Bond dissociation energies (BDEs) of the phenolic OH antioxidants determines the mechanism of inhibition. Since

groups in compounds 20–22.

10–13 appear to be effective at scavenging the initiating rad-

Compound^[a]

α/°^[b]

BDE/kcalmol^–1

icals, it is likely that their lower overall inhibitory activities

do not reflect lower H-donating activities but rather less

favourable positioning with respect to the lipid peroxyl radi-

cals. A possible positioning of the best antioxidants in each

series (i.e., 10 and 16) in the SDS micelles is proposed in

Scheme 5. Finally, the antioxidant activities displayed by

the different vitamin E analogues are summarized in

Scheme 6.

20

21

45.0^[c]

57.6

43.1^[c]

63.0

–

74.1

73.8

74.6

73.1

73.9

20b (peracetylated Glc)

21b (peracetylated Glc)

22

[a] 2-β--glucopyranosyl-4-methoxy-3,5,6-trimethylphenol (20), 3-

β--glucopyranosyl-4-methoxy-2,5,6-trimethylphenol (21), 4-meth-

oxy-2,3,5,6-tetramethylphenol (22). [b] Dihedral angle about the

glycosidic bond (from O-5(pyran) to C(OR), R = H or Me). [c] H-

bonded conformers, between O-5(pyran) and the phenolic OH, d

= 0.182 nm.

could reflect a less favourable partitioning in the micelles.

More generally, the steric hindrance brought about by the

C-glycosyl moiety, whether acetylated or not, might signifi-

cantly decrease the rate constant for LOO^·scavenging (k_a2).

The additional activity of 16–19 as initiation inhibitors can-

not be precisely assessed, as evidenced by the high standard

deviations in the AE₁parameter (Table 3). We may men-

tion, however, that there is no indication that the C-glycosyl

moiety of 17 and 19, which would be likely to favour their

location in the aqueous phase, increases their potency as

initiation inhibitors in relation to α-tocopherol. However,

since the initiating peroxyl radicals are positively charged

and the micelle surface negatively charged, the interface

could be the privileged site for inhibition of initiation. Over-

all, 16 and 19 emerge as the most efficient antioxidants after

α-tocopherol. Remarkably, 16 (no phytyl chain, acetylated

Glc) and 19 (phytyl chain, deacetylated Glc) can be consid-

ered intermediate antioxidants in terms of hydrophilic–lipo-

philic balance. This character might favour their localiza-

tion at the micelle surface. In the same series (Glc meta to

the phenolic OH), the most hydrophilic antioxidant 17 (no

phytyl chain, deacetylated Glc) and the most lipophilic one

18 (phytyl chain, acetylated Glc) appear to be less efficient.

The observation that 19 is better than 17, due to a higher

AE₁value for the former, is consistent with the inhibition

of initiation taking place predominantly at the interface.

When the C-glycosyl group is acetylated, the presence of

the phytyl chain might be considered unfavourable (16 bet-

ter than 18). It can be proposed that the highly hydrophobic

antioxidant 18 is strongly bound to the micelles and thus

less prone to intermicellar diffusion.

– Although poorer antioxidants than 16–19, compounds

10–13 (Glc ortho to the phenolic OH) can be regarded as

efficient scavengers of the hydrophilic initiating radicals

(Table 3). In the series, 10 (no phytyl chain, acetylated Glc)

emerges as the most active antioxidant. In particular, it re-

acts with ROO^·more rapidly than the corresponding de-

acetylated analogue 11, once more suggesting that inhibi-

tion of initiation (like initiation itself) is an interfacial phe-

nomenon.

Scheme 5. Hypothetical positioning of 10 and 16 in the SDS mi-

celles.

Conclusions

In this work, simple and efficient synthetic routes to β-

C-glycosyl analogues of tocopherols and chromanol are de-

Overall, the quantitative analysis confirms that the ortho scribed. The C-glycosidation of phenolic antioxidants could

vs. meta position of Glc relative to the phenolic OH of the be a means to improve water solubility while preserving the

1876

www.eurjoc.org

Eur. J. Org. Chem. 2008, 1869–1883

Amphiphilic Tocopherol-Derived Antioxidants

Scheme 6. Antioxidant activities of the vitamin E analogues.

otherwise. For column chromatography, Kieselgel 60 (Merck, par-

ticle size 0.063–0.200 mm) was used. Melting points were measured

in open capillary tubes with a Büchi apparatus or on a Kofler hot-

stage and are uncorrected. Optical rotations were determined with

a Perkin–Elmer 241 polarimeter at room temperature. NMR spec-

tra were recorded with Bruker AC 200 (200/50 MHz for ¹H/¹³C),

Bruker DRX 300/Avance 300 (300/75 MHz for ¹H/¹³C) or Bruker

DRX 500 (500/125 MHz for ¹H/¹³C) spectrometers. Chemical shifts

are referenced to Me₄Si (¹H), or to the residual solvent signals

(¹³C). NMR solvents were purchased from Euriso-Top (Saint Au-

phenolic OH, which is critical to the antioxidant activity. It

could also help deliver antioxidants to specific cells through

binding to glucose transporters or lectins. Moreover, C-gly-

cosylated antioxidants would be expected to be metaboli-

cally more stable than their O-glycosylated counterparts.

The antioxidants synthesized in this work inhibit the per-

oxidation of linoleic acid through two distinct mechanisms,

depending on the position of the C-glucosyl moiety

(whether acetylated or not). In the series in which the C-

glucosyl moiety lies ortho to the phenolic OH group, the bin, France). The following abbreviations are used to indicate the

observed multiplicities: s singlet, d doublet, dd doublet of doublet, t

triplet, q quadruplet, m multiplet, br. broad. Assignments of signals

marked with asterisks (*) are uncertain. HRMS (LSIMS) mass

spectra were recorded in the positive mode (unless stated otherwise)

with a Thermo Finnigan Mat 95 XL spectrometer. MS (ESI) mass

spectra were recorded in the positive mode with a Thermo Finni-

gan LCQ spectrometer. Elemental analyses were performed at the

Service Central d’Analyses du CNRS (Vernaison, France).

antioxidants essentially scavenge the hydrophilic peroxyl

radicals derived from the diazo initiator and are only mod-

erately effective (IC₅₀Ͼ0.7 µ). In the series in which the

C-glucosyl moiety lies meta to the phenolic OH group, the

antioxidants also scavenge the lipophilic peroxyl radicals

derived from the lipid and are significantly more effective

(IC₅₀Ͻ0.6 µ). Overall, the best antioxidants in each series

display intermediate hydrophilic–lipophilic balances either

with an acetylated glucosyl moiety or with a phytyl chain.

1,4-Dimethoxy-2,3-dimethylbenzene (2): Commercially available

2,3-dimethylhydroquinone (1.38 g) dissolved in DMSO (8 mL) was

methylated as described^[57]to afford 1,4-dimethoxy-2,3-dimeth-

ylbenzene (1.52 g, 92% yield) as white crystals, m.p. 73–75 °C (PE/

CH₂Cl₂) (ref.^[57]74–75 °C).

Experimental Section

General Methods: Dichloromethane and chloroform were washed

three times with water before drying over CaCl₂and distillation

over CaH₂. Dry methanol was obtained by distillation from magne-

sium methoxide. Acetonitrile was distilled from CaH₂and stored

over molecular sieves (3 Å). Other solvents were of commercial an-

alytical grade quality and were used without further purification.

Organic solutions were dried with anhydrous MgSO₄and concen-

trated in vacuo at 40–50 °C (bath temperature). TLC was per-

formed on DC-Alurolle Kieselgel 60 F₂₅₄(Merck), and the plates

were visualized by gentle heating and/or were inspected under UV

light: depending on the absorbance of the molecule present, the

spots visible under different wavelengths (254, 312 nm) might differ

in colour and/or intensity. When inspected with 312 nm wave-

length, most compounds were visible as violet spots, unless stated

1,4-Dimethoxy-2,6-dimethylbenzene (3): 1,4-Dimethoxy-2,6-di-

methylbenzene was prepared from 2,6-dimethylphenol in three

steps (oxidation with Fremy’s salt, 89%; reduction in ca. 1:1 (v/v)

H₂O/CHCl₃with 4 equiv. Na₂S₂O₄, 88%; methylation with KOH,

CH₃I in DMSO, 84%), as adapted from ref.^[39]The physical data of

the 2,6-dimethylbenzoquinone, -hydroquinone and -hydroquinone

dimethyl ether were in accordance with ref.^[39]

1,4-Dimethoxy-2,5-dimethylbenzene (4): 1,4-Dimethoxy-2,5-di-

methylbenzene was prepared from 2,5-dimethylphenol in three

steps (oxidation with Fremy’s salt, 94%; reduction in ca. 1:1 (v/v)

H₂O/CHCl₃with 4 equiv. Na₂S₂O₄, 93%; methylation with KOH,

CH₃I in DMSO, 91%), as adapted from ref.^[39]The physical data

for the 2,5-dimethylbenzoquinone, -hydroquinone, and -hydroqui-

Eur. J. Org. Chem. 2008, 1869–1883

www.eurjoc.org

1877

G.-R. Chen, O. Dangles, J.-P. Praly et al.

FULL PAPER

none dimethyl ether were in accordance with those reported in

ref.^[40]

to a column (mobile phase: EtOAc/PE, 1:2) to afford compound 7

(128 mg, 17% yield) as a 3:7 α/β anomeric mixture (NMR). R_f=

0.55 (EtOAc/PE, 2:3). α-Anomer: ¹H NMR ([D₆]DMSO,

500 MHz, 120 °C): δ = 6.76 (s, 1 H, Ar-H), 5.70 (d, J_1Ј,2Ј= 2.8 Hz,

1 H, H_1Ј), 5.21 (t, J = 5 Hz, 1 H), 5.13 (t, J = 5.3 Hz, 1 H), 5.10

(1 H, hidden signal revealed by integral), 4.49 (dd, J_5Ј,6aЈ= 8.0,

J_6aЈ,6bЈ= 13.1 Hz, 1 H, H_6aЈ), 4.26 (m, 2 H, H_5Ј, H_6bЈ), 3.75, 3.70

(2 s, 6 H, OMe), 2.32, 2.23 (2 s, 6 H, CH₃-arom.), 2.11, 2.09, 2.00,

1.81 (4 s, 12 H, acetyl) ppm. ¹³C NMR ([D₆]DMSO, 125 MHz,

120 °C): δ = 170.7, 170.0, 169.5, 169.3 (C=O, acetyl), 154.9, 151.0

(C₁, C₄),* 129.2, 128.2, 126.6 (C₂, C₃, C₆),* 115.2 (C₅), 75.1, 72.3,

71.1, 69.3, 68.0 (C_1Ј, C_2Ј, C_3Ј, C_4Ј, C_5Ј),* 62.5 (C_6Ј), 61.2, 57.1

(OCH₃), 21.1, 21.0, 21.0, 20.8 (CH₃, acetyl), 16.8, 13.4 (CH₃,

1,4-Dimethoxy-2,3-dimethyl-5-(2,3,4,6-tetra-O-acetyl-β-D-glucopyr-

anosyl)benzene (5): Compounds 1 (985 mg, 2.5 mmol) and 2

(830 mg, 5.0 mmol, 2 equiv.) were dissolved in anhydrous dichloro-

methane (6 mL). Silver trifluoroacetate (831 mg, 3.75 mmol,

1.5 equiv.) was added, and the mixture was stirred at about 28 °C

under argon, protected from light with aluminium foil. A solution

of SnCl₄in CH₂Cl₂(1 , 7.5 mL, 7.5 mmol, 3 equiv.) was added

dropwise to the mixture with stirring. After 3 h, TLC showed com-

pletion of the reaction, with 1, R_f= 0.36 (EtOAc/PE, 1:2), having

been converted into several products, R_f= 0.43, 0.31, 0.26, 0.24

(EtOAc/PE, 1:2). Saturated aq. NaHCO₃(40 mL) was added. After

stirring for 15 min, the mixture was filtered through a bed of celite,

which was rinsed with CH₂Cl₂(3ϫ25 mL). The filtrate was washed

with a satd. NaCl solution (30 mL) and then with water (30 mL),

then dried with MgSO₄. After filtration and concentration, the resi-

due was applied to a column (mobile phase: EtOAc/PE, 1:2) to

afford compound 5, which was recrystallized from CH₂Cl₂/PE

(825 mg, 66.5% yield). White crystals, m.p. 89.5–91 °C; R_f= 0.43

(EtOAc/PE, 1:2). [α]²_D²= –20.8 (c = 0.76 in CH₂Cl₂). ¹H NMR

(CDCl₃, 200 MHz): δ = 6.71 (s, 1 H, Ar-H), 5.52 (t, J_2Ј,3Ј= 9.1 Hz,

1 H, H_2Ј), 5.38 (t, J_3Ј,4Ј= 9.1 Hz, 1 H, H_3Ј), 5.24 (t, J_4Ј,5Ј= 9.7 Hz,

1 H, H_4Ј), 4.90 (d, J_1Ј,2Ј= 10 Hz, 1 H, H_1Ј), 4.24 (dd, J_5Ј,6aЈ= 5.0,

J_6aЈ,6bЈ= 12.3 Hz, 1 H, H_6aЈ), 4.12 (dd, J_5Ј,6bЈ= 2.3 Hz, 1 H, H_6bЈ),

3.90 (ddd, 1 H, H_5Ј), 3.81 (s, 3 H, OMe), 3.72 (s, 3 H, OMe), 2.21,

2.12, 2.07, 2.04, 2.03, 1.80 (6 s, 18 H, CH₃) ppm. ¹³C NMR

(CDCl₃, 50 MHz): δ = 170.7, 170.4, 169.6, 169.3 (C=O, acetyl),

154.0, 151.2 (C₂, C₅),* 131.0, 128.1, 125.3 (C₁, C₄, C₆),* 106.6 (C₃),

76.2, 74.9, 74.5, 70.8, 68.9 (C_1Ј, C_2Ј, C_3Ј, C_4Ј, C_5Ј),* 62.7 (C_6Ј), 62.0,

55.8 (OCH₃), 20.8, 20.7, 20.7, 20.6 (CH₃, acetyl), 12.8, 12.2 (CH₃,

Ar) ppm. C₂₄H₃₂O₁₁(496.19): C 58.06, H 6.50, O 35.45; found C

57.90, H 6.54, O 34.90.

1

Ar) ppm. β-Anomer: H NMR ([D₆]DMSO, 500 MHz, 120 °C): δ

= 6.79 (s, 1 H, Ar-H), 5.60 (t, J_2Ј,3Ј= 9.6 Hz, 1 H, H_2Ј), 5.32 (t,

J_3Ј,4Ј= 9.3 Hz, 1 H, H_3Ј), 5.10 (t, J_4Ј,5Ј= 9.7 Hz, 1 H, H_4Ј), 5.08

(d, J_1Ј,2Ј= 9.8 Hz, 1 H, H_1Ј), 4.18 (d, J = 3.2 Hz, 2 H, H_6aЈ, H_6bЈ),

4.04 (dt, J = 3.6, J = 3.6, J = 9.7 Hz, 1 H, H_5Ј), 3.75, 3.70 (2 s, 6

H, OMe), 2.26, 2.23 (2 s, 6 H, CH₃-arom.), 2.03, 1.99, 1.95, 1.68

(4 s, 12 H, acetyl) ppm. ¹³C NMR ([D₆]DMSO, 125 MHz, 120 °C):

δ = 170.5, 170.2, 170.0, 169.1 (C=O, acetyl), 154.5, 152.7 (C₁, C₄),*

129.0, 128.8, 125.6 (C₂, C₃, C₆),* 115.8 (C₅), 76.3, 75.6, 74.1, 71.4,

69.9 (C_1Ј, C_2Ј, C_3Ј, C_4Ј, C_5Ј),* 63.1 (C_6Ј), 61.9, 57.1 (OCH₃), 21.0,

21.0, 20.9, 20.6 (CH₃, acetyl), 16.7, 12.4 (CH₃, Ar) ppm. MS (ESI,

positive mode): m/z (%) = 519.1 (100) [M + Na]⁺.

2,3-Dimethyl-5-(2,3,4,6-tetra-O-acetyl-β-D-glucopyranosyl)-1,4-

benzoquinone (8): Compound 5 (1120 mg, 2.26 mmol), dissolved in

acetonitrile (4 mL), was oxidized with ceric ammonium nitrate

(CAN, 3717 mg, 6.78 mmol, 3 equiv.) dissolved in water (5 mL) at

room temperature with stirring. After 25 min, TLC showed that

the reaction was complete, compound 5, R_f= 0.43 (EtOAc/PE,

1:2), having been converted into a single new product with R_f=

0.45 (EtOAc/PE, 1:2). The reaction mixture was extracted with

CH₂Cl₂(3ϫ25 mL), washed with brine (40 mL) and then dried

with MgSO₄. After filtration and concentration, the residue was

applied to a column (mobile phase: EtOAc/PE, 1:2) to afford com-

pound 8 (1028 mg, 97.6% yield). Yellow needles, m.p. 119–120 °C

(CH₂Cl₂/PE); R_f= 0.45 (EtOAc/PE, 1:2, violet under UV 312 nm,

1,4-Dimethoxy-3,5-dimethyl-2-(2,3,4,6-tetra-O-acetyl-β-D-glucopyr-

anosyl)benzene (6): By the preceding procedure for the production

of 5, compounds 1 (985 mg, 2.5 mmol) and 3 (830 mg, 5.0 mmol,

2 equiv.) reacted within 3 h to afford product 6 (916.6 mg, 73.9%

yield) after workup and chromatography (mobile phase: EtOAc/

PE, 1:2). White foam, [α]²_D²= –10.1 (c = 0.86 in CH₂Cl₂); R_f= 0.40

(EtOAc/PE, 1:2). ¹H NMR ([D₆]DMSO, 300 MHz, 90 °C): δ = 6.69

1

dark spot at 254 nm). [α]²_D²= –1.2 (c = 0.8 in CH₂Cl₂). H NMR

(CDCl₃, 200 MHz): δ = 6.80 (d, J_1Ј,3= 0.8 Hz, 1 H, Ar), 5.34 (t,

J_3Ј,4Ј= 9.4 Hz, 1 H, H_3Ј), 5.11 (t, J_4Ј,5Ј= 10.0 Hz, 1 H, H_4Ј), 4.97

(t, J_2Ј,3Ј= 9.2 Hz, 1 H, H_2Ј), 4.66 (dd, J_1Ј,2Ј= 9.7, J_1Ј,3= 0.8 Hz, 1

H, H_1Ј), 4.20 (dd, J_5Ј,6aЈ= 4.7, J_6aЈ,6bЈ= 12.4 Hz, 1 H, H_6aЈ), 4.10

(dd, J_5Ј,6bЈ= 2.3 Hz, 1 H, H_6bЈ), 3.79 (ddd, 1 H, H_5Ј), 2.06, 2.02,

1.99, 1.99, 1.98, 1.87 (6 s, 18 H, CH₃) ppm. ¹³C NMR (CDCl₃,

50 MHz): δ = 186.9, 185.7 (C=O, benzoquinone), 170.6, 170.1,

169.7, 169.5 (C=O, acetyl), 143.4, 141.0, 140.9 (C₂, C₅, C₆),* 133.4

(C₃), 76.2, 73.8, 72.3, 72.2, 68.3 (C_1Ј, C_2Ј, C_3Ј, C_4Ј, C_5Ј), 62.0 (C_6Ј),

20.8, 20.6, 20.6, 20.5 (CH₃, acetyl), 12.4, 12.2 (benzoquinone

(s, 1 H, Ar-H), 5.54 (t, J_2Ј,3Ј= 9.3 Hz, 1 H, H_2Ј), 5.30 (t, J_3Ј,4Ј

=

9.5 Hz, 1 H, H_3Ј), 5.10 (d, J_1Ј,2Ј= 9.3 Hz, 1 H, H_1Ј), 5.05 (t, J_4Ј,5Ј

= 10.0 Hz, 1 H, H_4Ј), 4.16 (dd, J_5Ј,6aЈ= 3.4, J_6aЈ,6bЈ= 12.4 Hz, 1 H,

H_6aЈ), 4.11 (dd, J_5Ј,6bЈ= 4.1 Hz, 1 H, H_6bЈ), 4.00 (m, 1 H, H_5Ј), 3.76

(s, 3 H, OMe), 3.58 (s, 3 H, OMe), 2.33, 2.22, 2.03, 2.01, 1.95, 1.69

(6 s, 18 H, CH₃) ppm. ¹³C NMR ([D₆]DMSO, 75 MHz, 90 °C): δ

= 170.1, 169.7, 169.5, 168.7 (C=O, acetyl), 154.6, 151.3 (C₁, C₄),*

131.8, 131.7, 121.7 (C₂, C₃, C₅),* 112.9 (C₆), 75.4, 74.7, 73.6, 70.6,

69.2 (C_1Ј, C_2Ј, C_3Ј, C_4Ј, C_5Ј),* 62.5 (C_6Ј), 59.8, 56.9 (OCH₃), 20.6,

20.6, 20.5, 20.2 (CH₃, acetyl), 16.4, 12.5 (CH₃, Ar) ppm. C₂₄H₃₂O₁₁

(496.19): calcd. C 58.06, H 6.50, O 35.45; found C 57.76, H 6.46,

O 35.95.

CH ) ppm. IR (KBr): ν = 1760 (C=O, acetyl), 1650 (C=O, benzo-

˜

3

quinone) cm^–1. C₂₂H₂₆O₁₁(466.15): calcd. C 56.65, H 5.62, O

37.73; found C 56.56, H 5.97, O 36.77.

2,3-Dimethyl-5-(2,3,4,6-tetra-O-acetyl-β-D-glucopyranosyl)-1,4-hy-

1,4-Dimethoxy-2,5-dimethyl-3-(2,3,4,6-tetra-O-acetyl-β-D-glucopyr- droquinone (9a): A solution of Na₂S₂O₄(2052 mg, 12 mmol,

anosyl)benzene (7): By the preceding procedure for the production

of 5 or 6, compounds 1 (585 mg, 1.5 mmol) and 4 (996 mg,

6.0 mmol, 4 equiv.; twice as much as usual), dissolved in anhydrous

dichloromethane (7 mL), were stirred at ca 28 °C after addition of

silver trifluoroacetate (495 mg, 2.25 mmol, 1.5 equiv.) and a solu-

tion of SnCl₄in CH₂Cl₂(1 , 4.5 mL, 4.5 mmol, 3 equiv.). After 1

day, TLC showed the almost complete conversion of 1, R_f= 0.43

(EtOAc/PE, 2:3) into two major products, R_f= 0.55, 0.18 (EtOAc/

6 equiv.) in water (7 mL) was added to compound 8 (932 mg,

2 mmol), dissolved in CHCl₃(6 mL). The reaction mixture was

stirred vigorously at room temperature for 12 min, whereupon the

yellow colour of the organic layer disappeared. TLC showed that

compound 8, R_f= 0.51 (CH₂Cl₂/EtOAc, 8:1), had changed into a

more polar compound. The reaction mixture was washed with

water (3ϫ25 mL) and dried with MgSO₄. After filtration and con-

centration, the residue was applied to a column (mobile phase:

PE, 2:3). After workup and concentration, the residue was applied CH₂Cl₂/EtOAc, 8:1) to afford compound 9a (820 mg, 1.752 mmol,

1878

www.eurjoc.org

Eur. J. Org. Chem. 2008, 1869–1883

Amphiphilic Tocopherol-Derived Antioxidants

87.6% yield). White crystals, m.p. 192–193 °C (CH₂Cl₂/PE); R_f=

0.21 (CH₂Cl₂/EtOAc, 8:1). [α]²_D⁵= –37 (c = 0.8 in CHCl₃). ¹H NMR

(CDCl₃, 200 MHz): δ = 6.49 (s, 1 H, OH, Ar), 6.35 (s, 1 H, Ar-H),

5.37–5.23 (m, 3 H, H_2Ј, H_3Ј, H_4Ј), 4.87 (s, 1 H, OH, Ar), 4.52 (br.

d, J_1Ј,2Ј= 8.0 Hz, 1 H, H_1Ј), 4.32 (dd, J_5Ј,6aЈ= 3.8, J_6aЈ,6bЈ= 12.4 Hz,

1 H, H_6aЈ), 4.16 (d, J_6aЈ,6bЈ= 12.4 Hz, 1 H, H_6bЈ), 3.89 (m, 1 H,

H_5Ј), 2.16, 2.13, 2.12, 2.06, 2.01, 1.86 (6 s, 18 H, CH₃) ppm.

¹³C NMR (CDCl₃, 50 MHz): δ = 170.8, 170.5, 169.5, 169.0 (C=O,

acetyl), 146.8, 146.8 (C₁, C₄, Ar), 126.6, 125.1, 117.6 (C₂, C₅and

C₆, Ar), 111.7 (C₃, Ar), 80.0, 76.1, 73.9, 70.7, 68.0 (C_1Ј, C_2Ј, C_3Ј,

C_4Ј, C_5Ј),* 61.7 (C_6Ј), 20.7, 20.7, 20.6, 20.5 (CH₃, acetyl), 12.2, 12.1

(Ar-CH₃) ppm. C₂₂H₂₈O₁₁(468.16): calcd. C 56.41, H 6.02, O

37.57; found C 56.15, H 6.22, O 37.26.

20.6, 20.4 (CH₃, acetyl), 12.2, 11.9 (Me_7a, Me_8b) ppm. C₂₇H₃₆O₁₁

(536.23): calcd. C 60.44, H 6.76, O 32.80; found C 59.98, H 6.74,

O 33.75.

5-(β-D-Glucopyranosyl)-6-hydroxy-2,2,7,8-tetramethylchroman (11):

A solution of sodium methoxide in methanol (4 mL, 0.1 ) was

poured into a flask containing compound 10 (268 mg, 0.5 mmol).

After the mixture had been stirred at room temperature for 2 h,

TLC showed that compound 10 had changed into a more polar

compound, R_f= 0.23 (EtOAc). Then a slight excess of IR 120 was

added (pH paper for monitoring) and stirring of the mixture was

continued for 20 min. After filtration and concentration, the resi-

due was applied to a column (mobile phase: EtOAc) to afford com-

pound 11 (168.4 mg, 91.5% yield). White solid, m.p. 135–137 °C

1-Ethoxy-4-hydroxy-2,3-dimethyl-5-(2,3,4,6-tetra-O-acetyl-β-D

-glu- (EtOAc/Et₂O); R_f= 0.23 (EtOAc). [α]²_D²= +48.2 (c = 0.7 in ace-

copyranosyl)benzene (9b): Compound 9b was formed in 27% yield

as a by-product from 9a (98 mg, 0.209 mmol) and all-racemic phy-

tone). ¹H NMR ([D₆]acetone, 500 MHz): δ = 7.75 (s, 1 H, Ar-OH),

4.74 (d, J_1Ј,2Ј= 9.5 Hz, 1 H, H_1Ј), 4.43, 4.39, 4.09 (3 br. signals, 3 H,

tol (0.07 mL, 0.195 mmol, 0.93 equiv.) on treatment at reflux for sugar OH), 3.90–3.84 (m, 2 H, H_6aЈ, H_6bЈ), 3.76 (t, J_2Ј,3Ј= 9.1 Hz, 1

1 day in CHCl₃(3 mL, commercial grade, stabilized with ca. 0.6%

ethanol) in the presence of anhydrous ZnCl₂(106 mg, 0.778 mmol,

6 equiv.) in which 12 was obtained as a diastereoisomeric mixture

(90 mg, 0.12 mmol, 61.5% yield based on phytol). 9b: White crys-

tals, m.p. 161–162.5 °C (CH₂Cl₂/PE); R_f= 0.24 (PE/EtOAc, 3:1);

H, H_2Ј), 3.69 (t, J_3Ј,4Ј= 9.1, J_4Ј,5Ј= 9.5 Hz, 1 H, H_4Ј), 3.58 (t, J_2Ј,3Ј

= 9.1, J_3Ј,4Ј= 8.8 Hz, 1 H, H_3Ј), 3.50 (dt, J_4Ј,5Ј= 9.5, J_5Ј,6aЈ= J_5Ј,6bЈ

= 3.1 Hz, 1 H, H_5Ј), 2.98 (m, overlapping signals, sugar OH, 2 H,

H_4a), 2.68 (dt, J_gem= 16.4, J = 6.6 Hz, 1 H, H_4b), 2.07, 2.05 (2 s,

6 H, Me_7a, Me_8b), 1.74 (m, 2 H, H_3a,b), 1.28, 1.25 (2 s, 6 H, Me_2a,

Me_2b) ppm. ¹³C NMR ([D₆]acetone, 125 MHz): δ = 147.3 (C₆),

145.0 (C_8a), 125.1 (C₈), 123.1 (C₇), 119.0 (C₅), 117.5 (C_4a), 81.3

(C_5Ј), 79.0 (C_3Ј), 78.4 (C_1Ј), 73.4 (C_2Ј), 72.5 (C₂), 70.2 (C_4Ј), 61.3

(C_6Ј), 33.3 (C₃), 26.8, 26.2 (C_2a, C_2b),* 21.1 (C₄), 11.7, 11.5 (C_7a,

C_8b) ppm. MS (ESI, positive mode): m/z (%): 759.0 (100) [2M +

Na]⁺, 391.1 (15) [M + Na]⁺, 369.0 (10) [M + H]⁺. MS (ESI, nega-

[α]²_D²= –24.9 (c = 1.2 in CHCl₃). H NMR (CDCl₃, 500 MHz): δ

1

= 6.53 (s, 1 H, Ar-OH, exchanged with D₂O), 6.37 (s, 1 H, Ar-H),

5.36–5.28 (m, 3 H, H_2Ј, H_3Јand H_4Ј), 4.54 (d, J_1Ј,2Ј= 9.2 Hz, 1 H,

H_1Ј), 4.31 (dd, J_5Ј,6aЈ= 3.8, J_6aЈ,6bЈ= 12.5 Hz, 1 H, H_6Јa), 4.17 (dd,

J_5Ј,6bЈ= 2.1 Hz, 1 H, H_6Јb), 3.87 (m, 1 H, H_5Ј), 3.86 (m, 2 H, super-

imposed signals, OCH₂CH₃), 2.16, 2.13, 2.12, 2.05, 2.01, 1.84 (6 s,

18 H, 2 CH₃, 4 OAc), 1.38 (t, J = 6.9 Hz, 3 H, OCH₂CH₃) ppm. tive mode): m/z (%) = 770.7 (10) [2M + Cl]^–, 402.8 (100) [M +

¹³C NMR (CDCl₃, 125 MHz): δ = 171.0, 170.8, 169.7, 169.1 (C=O,

acetyl), 150.5 (C₄), 147.6 (C₁), 128.8 (C₅), 127.1 (C₆), 117.1 (C₂),

110.4 (C₃), 81.0 (C_1Ј), 76.5 (C_5Ј), 74.3 (C_3Ј), 70.9 (C_2Ј), 68.3 (C_4Ј),

65.4 (OCH₂CH₃), 62.0 (C_6Ј), 21.1, 21.0, 21.0, 20.9 (CH₃, acetyl),

15.5 (OCH₂CH₃), 12.6, 12.5 (Ar-CH₃) ppm. MS (ESI, +C): m/z

(%) = 519.2 (100) [M + Na]⁺, 514.2 (35) [M + NH₄]⁺.

Cl]^–, 367.0 (85) [M – H]^–.

5-(2,3,4,6-Tetra-O-acetyl-β-D-glucopyranosyl)-γ-tocopherol (12):

Compound 9a (270 mg, 0.577 mmol), dissolved in CHCl₃(5 mL),

was mixed with all-racemic phytol (0.2 mL, 0.556 mmol,

0.96 equiv.) and anhydrous ZnCl₂(453.7 mg, 3.336 mmol, 6 equiv.).

The mixture was stirred under argon at reflux for 24 h, protected

from light. TLC showed that the reaction was almost complete,

with 9a, R_f= 0.05 (EtOAc/PE, 1:3), still being visible under UV

312 nm, while phytol, R_f= 0.52–0.48 (EtOAc/PE, 1:3), was hardly

visible, in contrast to the desired products, visible on TLC as two

spots, R_f= 0.20 and 0.19 (EtOAc/PE, 1:3). The mixture was cooled

to room temperature and filtered. The filtrate was washed with

saturated aq. NaHCO₃(40 mL) and brine (40 mL) and H₂O

(40 mL), and was then dried with MgSO₄. After filtration and con-

centration, the residue was applied to a column (mobile phase: PE/

EtOAc, 3:1) to afford compound 12 as a diastereoisomeric mixture

(272 mg, 0.365 mmol, 65.6% yield, based on phytol), which was

6-Hydroxy-2,2,7,8-tetramethyl-5-(2,3,4,6-tetra-O-acetyl-β-D-gluco-

pyranosyl)chroman (10): Anhydrous ZnCl₂(979 mg, 7.2 mmol,

6 equiv.) was added to compound 9a (560 mg, 1.2 mmol) and 3-

methylbut-2-en-1-ol (prenyl alcohol, 0.123 mL, 1.19 mmol) dis-

solved in dry CHCl₃(6 mL) and the mixture was heated at reflux

under argon for 20 h while being stirred and protected from light.

TLC showed that the reaction was almost complete, with 9a, R_f=

0.08 (EtOAc/PE, 1:2), still being slightly visible, while 3-methylbut-

2-en-1-ol, R_f= 0.56–0.53 (EtOAc/PE, 1:2) had disappeared, with

formation of a more polar product, R_f= 0.23 (EtOAc/PE, 1:2).

After cooling and filtration, the organic phase was washed with

H₂O (40 mL), saturated aq. NaHCO₃(40 mL) and brine (40 mL), recrystallized from CH₂Cl₂/PE. In other experiments using larger

and was then dried with MgSO₄. After filtration and concentration,

the residue was applied to a column (mobile phase: EtOAc/PE, 1:2)

to afford 10 (384 mg, 0.716 mmol, 60% yield), which was recrys-

amounts of phytol (2, 3, or 4 equiv.), the reaction mixture was more

complex, and the desired product 12, obtained in ca. 60% yield,

was contaminated with impurities. Hence, use of minimized

tallized from CH₂Cl₂/PE. White crystals, m.p. 205–206.5 °C amounts of phytol appeared more suitable. Use of CHCl₃stabilized

(CH₂Cl₂/PE); R_f= 0.23 (EtOAc/PE, 1:2). [α]²_D²= –4.4 (c = 0.9 in

CH₂Cl₂). H NMR (CDCl₃, 200 MHz): δ = 6.93 (s, 1 H, Ar-OH), based on phytol), due to the formation of 9b (27% yield) as a by-

with EtOH (ca. 0.6% EtOH) afforded 12 in lower yield (61.5%,

1

5.58 (t, J_2Ј,3Ј= 9.3 Hz, 1 H, H_2Ј), 5.35–5.31 (m, 2 H, H_3Ј, H_4Ј), 4.86

(d, J_1Ј,2Ј= 10.0 Hz, 1 H, H_1Ј), 4.33 (dd, J_5Ј,6aЈ= 3.14, J_6aЈ,6bЈ

product (see above). Compound 12 (2RS,4ЈRS,8ЈRS mixture):

white crystals, m.p. 144–146 °C (CH₂Cl₂/PE); R_f= 0.20 (EtOAc/

=

12.5 Hz, 1 H, H_6aЈ), 4.14 (dd, J_5Ј,6bЈ= 2.0 Hz, 1 H, H_6bЈ), 3.86 (m, PE, 1:3). [α]²_D²= –0.227 (c = 0.88 in CH₂Cl₂). ¹H NMR (CDCl₃,

1 H, H_5Ј), 2.72 (t, J = 6.7 Hz, 2 H, 2 H₄), 2.14, 2.14 2.07, 2.06,

2.01, 1.81 (6 s, 18 H, 4 Ac, Me_7a, Me_8b), 1.75 (2 H, overlapped, 2

H₃), 1.29, 1.23 (2 s, 6 H, Me_2a, Me_2b) ppm. ¹³C NMR (CDCl₃,

50.32 MHz): δ = 170.7, 170.5, 169.3, 168.4 (C=O, acetyl), 146.9,

145.0 (C₆, C_8a, Ar),* 127.6, 124.9, 115.4, 114.9 (C₅, C₇, C₈, C_4a,

200 MHz): δ = 6.93, 6.90 (2 s, exchangeable with D₂O, 1 H, Ar-

OH), 5.58 (t, J_1ЈЈ,2ЈЈ= 9.9, J_2ЈЈ,3ЈЈ= 7.9 Hz, 1 H, H_2ЈЈ), 5.35–5.30

(m, 2 H, H_3ЈЈ, H_4ЈЈ), 4.86 (d, J_1ЈЈ,2ЈЈ= 9.9 Hz, 1 H, H_1ЈЈ), 4.33 (dd,

J_5ЈЈ,6aЈЈ= 2.9, J_{6aЈЈ,6bЈЈ}= 12.4 Hz, 1 H, H_6aЈЈ), 4.15 (d, J_{6aЈЈ,6bЈЈ}

=

12.4 Hz, 1 H, H_6bЈЈ), 3.89 (m, 1 H, H_5ЈЈ), 2.70 (t, J = 6.5 Hz, 2 H,

Ar),* 76.2, 76.2, 74.2, 70.3, 67.7 (C_1Ј, C_2Ј, C_3Ј, C_4Ј, C_5Ј),* 72.4 (C₂), H_4a,b), 2.14, 2.14, 2.08, 2.06, 2.01, 1.82 (6 s, 18 H, CH₃), 1.75 (2

61.4 (C_6Ј), 33.1, 20.8 (C₃, C₄),* 26.9, 26.3 (C_2a, C_2b), 20.7, 20.7, H, overlapped, H_3a,b), 1.60–1.0 (m, 24 H, H_1Ј-12Јand H_2a), 0.89–

Eur. J. Org. Chem. 2008, 1869–1883

www.eurjoc.org

1879

G.-R. Chen, O. Dangles, J.-P. Praly et al.

0.86 (m, 12 H, Me_{13Ј,4Јa,8Јa,12Јa}) ppm. ¹³C NMR (CDCl₃, 50 MHz): C_5Ј, C_6Ј, C_7Ј, C_9Ј, C_10Ј, C_11Ј), 33.0, 32.9, 28.2, 23.6, 22.6, 22.5, 19.7,

FULL PAPER

δ = 170.7, 170.5, 169.3, 168.4 (C=O, acetyl), 146.8, 144.9 (C₆, C_8a),

127.6, 124.9, 115.6, 114.9 (C₅, C₇, C₈, C_4a), 76.2, 76.2, 74.2, 70.3,

19.6 (C_4Ј, C_8Ј, C_12Ј, C_13Ј, C_4Јa, C_8Јa, C_12Јa, C_2a), 11.8, 11.6 (Me_7a,

Me_8b) ppm. MS (ESI, +C): m/z (%) = 1179.3 (65) [2M + Na]⁺,

67.8 (C_1ЈЈ, C_2ЈЈ, C_3ЈЈ, C_4ЈЈ, C_5ЈЈ), 74.4 (C₂), 61.4 (C_6ЈЈ), 39.4, 39.4, 578.3 (100) [M]⁺. MS (ESI, –C): m/z (%) = 1190.9 (65) [2M + Cl]

37.4, 37.4, 37.3, 31.7, 24.8, 24.5, 21.0, 20.6, 20.6 (methylene: C₃, ^–, 613.2 (100) [M + Cl]^–, 577.3 (65) [M – H]^–.

C₄, C_1Ј, C_2Ј, C_3Ј, C_5Ј, C_6Ј, C_7Ј, C_9Ј, C_10Ј, C_11Ј), 32.8, 28.0, 23.5, 22.8,

3,5-Dimethyl-2-(2,3,4,6-tetra-O-acetyl-β-D-glucopyranosyl)-1,4-

22.6, 20.4, 19.8, 19.7 (C_4Ј, C_8Ј, C_12Ј, C_13Ј, C_4Јa, C_8Јa, C_12Јa, C_2a),

20.7, 20.7, 20.6, 20.6 (CH₃, acetyl), 12.2, 11.9 (C_7a, C_8b) ppm.

C₄₂H₆₆O₁₁(746.46): calcd. C 67.53, H 8.91, O 23.56; found C

67.56, H 8.83, O 22.67.

benzoquinone (14): As described for the synthesis of 8, compound

6 (306 mg, 0.617 mmol), dissolved in acetonitrile (2 mL), was

treated with CAN (1.01 g, 1.85 mmol, 3 equiv.) dissolved in water

(2 mL) with stirring for 25 min at room temperature. TLC showed

that the reaction was over, with 6, R_f= 0.40 (EtOAc/PE, 1:2), hav-

ing been converted into a single new product, R_f= 0.43 (EtOAc/

PE, 1:2). After workup, purification by column chromatography

(mobile phase: EtOAc/PE, 1:2) afforded compound 14 (268 mg,

0.575 mmol, 93.2 % yield). Yellow crystals, m.p. 149–150.5 °C

(CH₂Cl₂/EP); R_f= 0.43 (EtOAc/PE, 1:2, violet under UV 312 nm,

dark spot under 254 nm). [α]²_D²= –63 (c = 0.7 in CH₂Cl₂). ¹H NMR

(500 MHz, CDCl₃): δ = 6.56 (d, J_6,5-Me= 1.4 Hz, 1 H, Ar-H), 5.43

5-(β-D-Glucopyranosyl)-γ-tocopherol [(2R)-13 and (2S)-13]: A solu-

tion of sodium methoxide in methanol (5 mL, 0.1 ) was poured

into a flask containing compound 12 (405 mg, 0.54 mmol). After

the mixture had been stirred at room temperature for 2 h, TLC

showed the conversion of 12 into polar products visible as two

spots on the plates. A slight excess of IR 120 H⁺was added (pH

paper for monitoring), and stirring of the mixture was continued

for 20 min. After filtration and concentration, the residue was ap-

plied to a column (mobile phase: EtOAc/CH₂Cl₂2:1, then 4:1, then

8:1, at last pure EtOAc) to afford three fractions: compound (2S)-

13 (70.4 mg), a mixture of compounds (2S)-13 and (2R)-13

(125.5 mg, the proportion of (2S)-13 and (2R)-13 being about 2:3

according to the ¹H NMR spectrum), and compound (2R)-13

(100.6 mg), with combined total yield of 95%. The ratio of (2S)-13

and (2R)-13 can be estimated as ca. 2:3.

(t, J_1Ј,2’= 9.9, J_2Ј,3’= 9.1 Hz, 1 H, H_2Ј), 5.31 (t, J_2Ј,3’= 9.1, J_3Ј,4’

=

9.7 Hz, 1 H, H_3Ј), 5.20 (t, J_3Ј,4’= J_4Ј,5’= 9.7 Hz, 1 H, H_4Ј), 4.96 (d,

J_1Ј,2’= 9.9 Hz, 1 H, H_1Ј), 4.20 (dd, J_5Ј,6a’= 2.6, J_6aЈ,6b’= 12.5 Hz,

1 H, H_6aЈ), 4.17 (dd, J_5Ј,6b’= 4.0, J_6aЈ,6b’= 12.5 Hz, 1 H, H_6bЈ), 3.76

(dq, 1 H, H_5Ј), 2.27, 2.08, 2.06, 2.01, 1.87 (5 s, 15 H, CH₃at posi-

tion 3, and acetyl), 2.04 (d, 3 H, J = 1.5 Hz, Me at position 5) ppm.

¹³C NMR (CDCl₃, 50 MHz): δ = 187.7, 185.3 (C=O, benzoqui-

none), 170.6, 170.1, 169.6, 169.5 (C=O, acetyl), 145.9, 145.8, 136.8

(C₂, C₃, C₅),* 133.0 (C₆), 76.3, 74.2, 71.9, 70.2, 68.3 (C_1Ј, C_2Ј, C_3Ј,

C_4Ј, C_5Ј),* 61.9 (C_6Ј), 20.7, 20.6, 20.6, 20.4 (CH₃, acetyl), 16.0, 12.7

Compound (2S)-13: Pale yellow amorphous solid, [α]²_D²= +28.5 (c

= 1 in acetone); R_f= 0.50 (EtOAc). ¹H NMR ([D₆]acetone,

500 MHz): δ = 7.76 (s, 1 H, Ar-OH), 4.73 (d, J_1ЈЈ,2ЈЈ= 9.5 Hz, 1 H,

H_1ЈЈ), 4.72, 4.47, 4.26, 3.13 (4 br. signals, 4 H, sugar OH), 3.90–

3.83 (m, 2 H, H_6aЈЈ, H_6bЈЈ), 3.77 (t, J_1ЈЈ,2ЈЈ= 9.5, J_2ЈЈ,3ЈЈ= 8.8 Hz, 1

H, H_2ЈЈ), 3.71 (t, J_3ЈЈ,4ЈЈ= 8.8, J_4ЈЈ,5ЈЈ= 9.2 Hz, 1 H, H_4ЈЈ), 3.60 (t,

(benzoquinone CH ) ppm. IR (KBr): ν = 1750 (C=O, acetyl), 1650

˜

3

(C=O, benzoquinone) cm^–1. C₂₂H₂₆O₁₁(466.15): calcd. C 56.65, H

5.62, O 37.73; found C 56.57, H 5.61, O 37.50.

J_2ЈЈ,3ЈЈ= J_3ЈЈ,4ЈЈ= 8.8 Hz, 1 H, H_3ЈЈ), 3.49 (br. d, J_4ЈЈ,5ЈЈ= 9.2 Hz, 1 3,5-Dimethyl-2-(2,3,4,6-tetra-O-acetyl-β-

H, H_5ЈЈ), 3.00 (m, 1 H, H_4a), 2.65 (m, 1 H, H_4b), 2.07, 2.07 (2 s, 6 droquinone (15): As described for the synthesis of 9a, compound 14

H, Me_7a, Me_8b), 1.75 (m, 2 H, H_3a,b), 1.59–1.13 (m, 21 H, H_1Ј-12Ј), (268 mg, 0.575 mmol), dissolved in CHCl₃(4 mL), was reduced

D-glucopyranosyl)-1,4-hy-

1.22 (s, 3 H, H_2a), 0.91–0.89 (m, 12 H, H_13Ј, H_4Јa, H_8Јa, H_12Јa) ppm. with Na₂S₂O₄(600 mg, 3.45 mmol, 6 equiv.) in water (4 mL). The

¹³C NMR ([D₆]acetone, 125 MHz): δ = 147.2 (C₆), 144.8 (C_8a),

mixture was stirred vigorously at room temperature for 12 min.

125.1 (C₈), 123.1 (C₇), 119.0 (C₅), 117.7 (C_4a), 81.3 (C_5ЈЈ), 78.8

TLC showed that compound 14, R_f= 0.55 (CH₂Cl₂/EtOAc, 8:1),

(C_3ЈЈ), 78.3 (C_1ЈЈ), 74.5 (C₂), 73.3 (C_2ЈЈ), 70.0 (C_4ЈЈ), 61.2 (C_6ЈЈ), 31.9,^ahad changed into a more polar compound, R_f= 0.17 (CH₂Cl₂/

31.8^a(C₃), 20.7 (C₄), 40.3,^a40.1,^a39.6, 38.0,^a37.9,^a37.9,^a37.9,^aEtOAc, 8:1). After workup and concentration, the residue was ap-

37.7,^a37.7, 37.6, 25.1, 24.7, 21.4,^a21.3,^a20.7 (C_1Ј, C_2Ј, C_3Ј, C_5Ј,

C_6Ј, C_7Ј, C_9Ј, C_10Ј, C_11Ј), 33.1, 33.0,^a32.9,^[a]28.2, 23.6, 22.6, 22.5,

plied to a column (mobile phase: CH₂Cl₂/EtOAc, 8:1) to afford

compound 15 (256 mg, 0.547 mmol, 95.1% yield). White crystals,

19.7,^a19.7,^a19.6, 19.5^a(C_4Ј, C_8Ј, C_12Ј, C_13Ј, C_4Јa, C_8Јa, C_12Јa, C_2a), m.p. 198.5–200 °C (CH₂Cl₂/PE); R_f= 0.17 (CH₂Cl₂/EtOAc, 8:1).

1

11.8, 11.6 (C_7a, C_8b) ppm. At 125 MHz, the (4ЈRS,8ЈRS) stereoiso-

[α]²_D⁵= –23.9 (c = 0.8 in CHCl₃). H NMR (CDCl₃, 500 MHz): δ

mers present gave additional small signals, labelled with the super- = 6.95 (s, exchangeable, 1 H, OH), 6.58 (s, 1 H, H₆), 5.50 (t, J_1Ј,2’

=

script ^a. MS (ESI +C): m/z (%) = 1179.3 (100) [2M + Na]⁺, 578.3

(30) [M]⁺. MS (ESI –C): m/z (%) = 190.8 (40) [2M + Cl]^–, 613.2

(100) [M + Cl]^–, 577.2 (20) [M – H]^–.

J_2Ј,3’= 9.5 Hz, 1 H, H_2Ј), 5.37–5.30 (2 overlapping triplets, 2 H,

H_3’and H_4Ј), 4.91 (d, J_1Ј,2’= 9.9 Hz, 1 H, H_1Ј), 4.32 (dd, J_5Ј,6a’

=

2.3, J_6aЈ,6b’= 12.5 Hz, 1 H, H_6aЈ), 4.25 (s, exchangeable, 1 H, OH),

4.15 (dd, J_5Ј,6b’= 1.4, J_6aЈ,6b’= 12.5 Hz, 1 H, H_6bЈ), 3.88 (m, 1 H,

H_5Ј), 2.21, 2.17, 2.12, 2.06, 2.01, 1.81 (6 s, 18 H, CH₃) ppm. ¹³C

NMR (CDCl₃, 50 MHz): δ = 170.7, 170.4, 169.4, 168.6 (C=O, ace-

tyl), 149.4, 145.5 (C₁, C₄), 125.3, 125.3, 122.3 (C₂, C₃, C₅, Ar),

117.6 (C₆), 77.0, 76.1, 73.9, 70.4, 67.8 (C_1Ј, C_2Ј, C_3Ј, C_4Ј, C_5Ј), 61.4

(C_6Ј), 20.7, 20.7, 20.6, 20.3 (CH₃, acetyl), 16.3, 12.3 (Ar-CH₃) ppm.

C₂₂H₂₈O₁₁(468.16): calcd. C 56.41, H 6.02, O 37.57; found C

56.03, H 6.21, O 37.51.

Compound (2R)-13: Pale yellow amorphous solid, [α]²_D²= +42.4 (c

= 1 in acetone); R_f= 0.43 (EtOAc). ¹H NMR ([D₆]acetone,

500 MHz): δ = 7.79 (s, 1 H, Ar-OH), 4.73 (d, J_1ЈЈ,2ЈЈ= 9.5 Hz, 1 H,

H_1ЈЈ), 4.62, 4.47, 4.19, 3.01 (4 s, exchanged with D₂O, 4 H, sugar

OH), 3.89–3.83 (m, 2 H, H_6aЈЈ, H_6bЈЈ), 3.75 (t, J_1ЈЈ,2ЈЈ= 9.5, J_2ЈЈ,3ЈЈ

= 8.8 Hz, 1 H, H_2ЈЈ), 3.70 (t, J_3ЈЈ,4ЈЈ= 8.8, J_4ЈЈ,5ЈЈ= 9.8 Hz, 1 H,

H_4ЈЈ), 3.58 (t, J_2ЈЈ,3ЈЈ= J_3ЈЈ,4ЈЈ= 8.8 Hz, 1 H, H_3ЈЈ), 3.49 (br. d, J =

9.5 Hz, 1 H, H_5ЈЈ), 2.96 (m, 1 H, H_4a), 2.68 (dt, J = 6.3, J = 6.6,

J_gem= 16.4 Hz, 1 H, H_4b), 2.07, 2.07 (s, 6 H, Me_7a, Me_8b), 1.74 (t,

J = 6.6 Hz, 2 H, H_3a,b), 1.62–1.11 (m, 21 H, H_1Ј-12Ј), 1.22 (s, 3 H,

6-Hydroxy-2,2,5,7-tetramethyl-8-(2,3,4,6-tetra-O-acetyl-β-D-gluco-

pyranosyl)chroman (16): By the procedure optimized for the ortho-

H_2a), 0.90–0.88 (m, 12 H, H_13Ј, H_4Јa, H_8Јa, H_12Јa) ppm. ¹³C NMR dimethyl series, compound 15 (500 mg, 1.068 mmol), dissolved in

([D₆]acetone, 50 MHz): δ = 148.2 (C₆), 145.9 (C_8a), 126.2 (C₈), anhydrous CHCl₃(6 mL), was mixed with prenyl alcohol (0.11 mL,

124.2 (C₇), 120.2 (C₅), 118.8 (C_4a), 82.5 (C_5ЈЈ), 80.0 (C_3ЈЈ), 79.5

(C_1ЈЈ), 75.6 (C₂), 74.5 (C_2ЈЈ), 71.3 (C_4ЈЈ), 62.4 (C_6ЈЈ), 33.1 (C₃), 21.9

1.06 mmol) and anhydrous ZnCl₂(878 mg, 6.41 mmol, 6 equiv.).

The mixture, protected from light with aluminium foil, was stirred

(C₄), 41.4, 40.7, 39.1, 38.8, 38.6, 26.2, 26.2, 25.8, 22.4 (C_1Ј, C_2Ј, C_3Ј, at reflux under argon for 20 h. TLC showed that the reaction was

1880 Eur. J. Org. Chem. 2008, 1869–1883

www.eurjoc.org

Amphiphilic Tocopherol-Derived Antioxidants

almost complete, with 15, R_f= 0.08 (EtOAc/PE, 1:2) still being

visible under UV 312 nm, while 3-methylbut-2-en-1-ol, R_f= 0.56–

0.53 (EtOAc/PE, 1:2) had disappeared, with formation of a new

product, R_f= 0.21 (EtOAc/PE, 1:2). After workup and concentra-

tion, the residue was applied to a column (mobile phase: EtOAc/

PE, 1:2) to afford 16 (320.7 mg, 0.6 mmol, 48.2% yield) as a white

foam. R_f= 0.21 (EtOAc/PE, 1:2). [α]²_D⁵= –1.1 (c = 0.8 in CH₂Cl₂).

¹H NMR ([D₆]DMSO): recording of the spectra at room temp.,

50, 90 °C (300 MHz) and 120 °C (500 MHz) resulted in improved

resolution, although H_1’and H_2’still gave broad signals. ¹H NMR

(dd, J_5ЈЈ,6ЈЈb= 4.5, J_{6ЈЈa,6ЈЈb}= 12 Hz, 1 H, H_6ЈЈb), 3.86 (m, 1 H, H_5ЈЈ),

2.57 (m, 2 H, H_4a, H_4b), 2.27, 2.08, 2.02, 2.00, 1.94 (5 s, 15 H, 2

ArCH₃, 3 OAc), 1.74 (m, 2 H, H_3a, H_3b), 1.71 (2 s, 3 H, acetyl at

C_2ЈЈ), 1.61–1.12 (m, 24 H, H_1Ј-12’and Me_2a), 0.93–0.87 (m, 12 H,

Me_{13Ј,4Јa,8Јa,12Јa}) ppm. ¹³C NMR ([D₆]DMSO, 120 °C, 125 MHz): δ

= 170.4, 170.15, 170.1, 169.9, 169.8, 169.1 (C=O, acetyl), 146.5,

125.3, 125.3, 125.2, 125.1, 120.4, 120.2, 118.5, 118.4 (C₆, C_8a, C₅,

C₇, C₈, C_4a), 76.3, 76.0, 71.6, 70.2, 70.1 (C_1ЈЈ, C_2ЈЈ, C_3ЈЈ, C_4ЈЈ, C_5ЈЈ),

75.7, 75.6 (C₂), 63.5, 63.4 (C_6ЈЈ), 41.5, 39.8, 38.2, 37.8, 37.7, 37.6,

31.9, 24.9, 24.7, 21.5, 21.1 (methylene: C₃, C₄, C_1Ј, C_2Ј, C_3Ј, C_5Ј,

([D₆]DMSO, 120 °C, 500 MHz): δ = 7.00 (s, 1 H, Ar-OH), 5.65 C_6Ј, C_7Ј, C_9Ј, C_10Ј, C_11Ј), 33.1, 28.2, 24.0, 23.7, 23.1, 23.1, 21.0, 21.0,

(brtriplet, 1 H, H_2Ј), 5.24 (t, J_2Ј,3’= 9.1, J_3Ј,4’= 9.5 Hz, 1 H, H_3Ј),

5.11 (brd, 1 H, H_1Ј), 5.07 (t, J_3Ј,4’= 9.5, J_4Ј,5’= 9.8 Hz, 1 H, H_4Ј),

20.7, 20.4, 20.3 (C_4Ј, C_8Ј, C_12Ј, C_13Ј, C_4Јa, C_8Јa, C_12Јa, C_2a), 20.4,

20.3, 20.3, 20.3 (CH₃, acetyl), 13.3, 12.6 (C_7a, C_8b) ppm. MS (ESI,

positive mode): m/z (%): 1515.6 (55) [2M + Na]⁺, 769.4 (100) [M

4.14 (d, J = 3.5 Hz, 2 H, H_6aЈ, H_6bЈ), 3.91 (dt, J_4Ј,5’= 9.8, J_5Ј,6a’

=

J_5Ј,6b’= 3.8 Hz, 1 H, H_5Ј), 2.58 (t, J = 6.9 Hz, 2 H, H_4ab), 2.27, + Na]⁺, 764.1 (40) [M + NH₄]⁺, 747.1 (45) [M + H]⁺. HRMS (FAB,

2.08 (2 s, 6 H, Me_5a, Me_7a), 2.03, 2.00, 1.95, 1.71 (4 s, 12 H, 4 Ac), nitrobenzyl alcohol) calcd. for C₄₂H₆₇O₁₁[M + H]⁺: 747.4683,

1.77 (t, J = 6.9, J = 6.6 Hz, 2 H, H_3ab), 1.33, 1.29 (2 s, 6 H, Me_2a,

Me_2b) ppm. ¹³C NMR ([D₆]DMSO, 120 °C, 125 MHz): δ = 170.5,

170.2, 169.9, 169.2 (C=O, acetyl), 146.8, 146.5 (C₆, C_8a, Ar),*

125.3, 125.1, 120.4, 118.3 (C₅, C₇, C₈, C_4a, Ar),* 76.0, 76.0, 74.5,

71.7, 70.0 (C_1Ј, C_2Ј, C_3Ј, C_4Ј, C_5Ј),* 73.5 (C₂), 63.4 (C_6Ј), 33.6, 21.4

(C₃, C₄),* 27.6, 27.0 (C_2a, C_2b), 21.4, 21.0, 21.0, 20.8 (CH₃, acetyl),

13.3, 12.6 (Me_7a, Me_8b) ppm. C₂₇H₃₆O₁₁(536.23): calcd. C 60.44,

H 6.76, O 32.80; found C 60.92, H 7.05, O 31.23.

found: 747.46790.

8-(β- -Glucopyranosyl)-ε-tocopherol (19): A solution of sodium

D

methoxide in methanol (3 mL, 0.1 ) was poured into a flask con-

taining compound 18 (212 mg, 0.284 mmol). After the mixture had

been stirred at room temperature for 2 h, TLC showed that com-

pound 18, R_f= 0.26 (EtOAc/PE, 1:3) had changed into a more

polar compound, R_f= 0.60 (EtOAc). After neutralization (Am-

berlyst resin IR 120 H⁺), filtration and concentration, the residue

was applied to a column (mobile phase: EtOAc) to afford com-

pound 19 (146 mg, 89% yield) as a brown-red amorphous solid, R_f

= 0.60 (EtOAc). [α]²_D⁵= +18.8 (c = 0.74 in acetone); ¹H NMR ([D₆]-

DMSO, 90 °C, 500 MHz): diastereoisomeric mixture giving a com-

plex and uninformative spectrum for the sugar part, while the chain

and chroman protons were visible. MS (ESI, positive mode): m/z

(%): 1179.4 (100) [2M + Na]⁺, 601.4 (45) [M + Na]⁺; (ESI, negative

mode): m/z (%): 1191.2 (60) [2M + Cl]^–, 623.1 (100) [M +

HCOO]^–, 577.3 (45) [M – H]^–.

8-(β-

A solution of sodium methoxide in methanol (4 mL, 0.1 ) was

poured into flask containing compound 16 (302 mg,

D-Glucopyranosyl)-6-hydroxy-2,2,5,7-tetramethylchroman (17):

a

0.563 mmol), which was stirred at room temperature for 2 h, after

which a more polar compound had been formed. A slight excess

of Amberlyst resin IR 120 H⁺was added, and stirring of the mix-

ture was continued for 20 min. After filtration and concentration,

the residue was applied to a column (mobile phase: EtOAc) to af-

ford 17 (192.6 mg, 93% yield) as a white amorphous solid; R_f=

1

0.19 (EtOAc). [α]²_D⁵= +18.3 (c = 0.6 in CH₃OH). H NMR ([D₆]-

Evaluation of Antioxidant Properties

DMSO + D₂O, 90 °C, 500 MHz): δ = 4.68 (br. signal, 1 H, H_1Ј),

3.86 (br. signal, 1 H, H_2Ј), 3.71 (dd, J_5Ј,6Јa= 2.3, J_6Јa,6Јb= 11.7 Hz,

1 H, H_6Јa), 3.50 (dd, J_5Ј,6Јb= 5.1, J_6Јa,6Јb= 11.7 Hz, 1 H, H_6Јb),

3.30–3.20 (m, 3 H, H_3Ј, H_4Ј, H_5Ј), 2.56 (t, J = 6.6, J = 6.3 Hz, 2 H,

H_4a,b), 2.22, 2.06 (2 s, 6 H, Me_5a, Me_7a), 1.74 (t, J = 6.6, J = 6.0 Hz,

2 H, H_3a,b), 1.25, 1.21 (2 s, 6 H, Me_2a, Me_2b) ppm. ¹³C NMR ([D₆]-

DMSO + D₂O, 90°, 125 MHz): δ = 147.0, 146.4 (C₆, C_8a), 124.9,

123.9, 123.8, 118.2 (C_4a, C₅, C₇, C₈), 81.6, 80.0, 72.9, 71.8, 71.8

(C_1Ј, C_2Ј, C_3Ј, C_4Ј, C_5Ј), 73.2 (C₂), 62.9 (C_6Ј), 33.6 (C₃), 27.7, 26.7

(C_2a, C_2b), 21.5 (C₄), 13.8, 12.6 (C_5a, C_7a) ppm. MS (ESI, +C): m/z

(%) = = 759.0 (100) [2M + Na]⁺, 391.1 (27) [M + Na]⁺.

Chemicals: (Ϯ) α-Tocopherol, linoleic acid, AAPH [2,2Ј-azo-bis(2-

methylpropionamidine) dihydrochloride] and SDS (sodium dode-

cylsulfate) were purchased from Sigma–Aldrich (L’Isle d’Abeau,

France). All reagents were of the highest purity available (95–99%)

and were used without purification. All solvents used were analyti-

cal grade. The phosphate buffer (pH7.4, 50 m NaH₂PO₄) was pre-

pared with Millipore Q-Plus water and eluted on a chelating resin

(chelex 100, 0.4 milliequiv.mL^–1, Bio-Rad) to remove contaminat-

ing metal traces.

UV/Vis Spectroscopy: UV/Vis spectra were recorded on a Hewlett–

Packard 8453 diode array spectrometer fitted with a magnetically

stirred cell (optical pathlength: 1 cm). The temperature in the cell

was maintained by means of a water thermostatted bath.

8-(2,3,4,6-Tetra-O-acetyl-β-

D

-glucopyranosyl)-ε-tocopherol

(18):

Compound 15 (300 mg, 0.64 mmol) dissolved in CHCl₃(5 mL) was

mixed with racemic phytol (0.23 mL, 0.64 mmol, 1 equiv.) and an-

hydrous ZnCl₂(520 mg, 3.84 mmol, 6 equiv.). The mixture was

stirred at reflux under argon for 24 h with protection against light

(aluminium foil). TLC showed that the reaction was almost com-

plete, with compound 15, R_f= 0.06 (EtOAc/PE, 1:3) still being

visible under UV light (λ = 312 nm), while phytol, R_f= 0.52–0.48

(EtOAc/PE, 1:3) was hardly visible, in contrast to the desired prod-

uct, R_f= 0.26 (EtOAc/PE, 1:3). Workup and purification as de-

scribed above for 12 yielded compound 18 as a mixture of dia-

stereoisomers (300.4 mg, 0.40 mmol, 62.7% yield). Brown-red

Inhibition of Linoleic Acid Peroxidation: A freshly prepared solution

of linoleic acid (2.55 m, 2 mL) in a phosphate buffer (pH 7.4)

containing SDS (0.1 ) was placed in the spectrometer cell at 37 °C.

At time zero, a freshly prepared solution of AAPH (80 m, 25 µL)

in the same buffer was added, followed ca. 800 s later by an antioxi-

dant solution in MeOH (25 µL). The experiments were repeated

with different antioxidant concentrations (1 m and lower). The

initial level of hydroperoxides (molar absorption coefficient at

234 nm was 26100 ^–1cm^–1) was below 2% in all experiments. The

syrup, R_f= 0.26 (EtOAc/PE, 1:3). [α]²_D⁵= +4.66 (c = 0.88 in

1

CH₂Cl₂). H NMR ([D₆]DMSO, 120 °C, 500 MHz): δ = 7.26 (s, 1 uninhibited and inhibited peroxidation rates were calculated from

H, Ar-OH, more visible when the spectra was recorded at 90 °C),

5.64 (br. signal, 1 H, H_2ЈЈ), 5.234 and 5.227 (2 t, J = 9.5, J = 9.1 Hz, dant addition with fixed time intervals. Each experimental point

1 H, H_3ЈЈ), 5.11 (br. signal, 1 H, H_1ЈЈ), 5.10–5.04 (m, 1 H, complex was the mean of two to three measurements. Errors were lower

H_4ЈЈ), 4.16 (dd, J_5ЈЈ,6ЈЈa= 3.2, J_{6ЈЈa,6ЈЈb}= 12 Hz, 1 H, H_6ЈЈa), 4.11 than 10%.

the slope of the A(234 nm) vs. time lines before and after antioxi-

Eur. J. Org. Chem. 2008, 1869–1883

www.eurjoc.org

1881

G.-R. Chen, O. Dangles, J.-P. Praly et al.

FULL PAPER

2004, 69, 9215–9223; c) H.-Y. Kim, D. A. Pratt, J. R. Seal, M.

Wijtmans, N. A. Porter, J. Med. Chem. 2005, 48, 6787–6789.

a) D. A. Pratt, G. A. DiLabio, G. Brigati, G. F. Pedulli, L. Val-

gimigli, J. Am. Chem. Soc. 2001, 123, 4625–4626; b) L. Valgimi-

gli, G. Brigati, G. F. Pedulli, G. A. DiLabio, M. Mastragostino,

C. Arbizzani, D. A. Pratt, Chem. Eur. J. 2003, 9, 4997–5010.

a) T.-g. Nam, M. Wijtmans, D. A. Pratt, N. A. Porter, Synthesis

2005, 1397–1404; b) T.-g. Nam, C. L. Rector, H.-y. Kim, A. F.-

P. Sonnen, R. Meyer, W. M. Nau, J. Atkinson, J. Rintoul, D. A.

Pratt, N. A. Porter, J. Am. Chem. Soc. 2007, 129, 10211–10219.

a) T. Muller, L. Grandbarbe, E. Morga, P. Heuschling, B. Luu,

Bioorg. Med. Chem. Lett. 2004, 14, 6023–6026; b) T. Muller,

D. Coowar, M. Hanbali, P. Heuschling, B. Luu, Tetrahedron

2006, 62, 12025–12040.

M. Koufaki, E. Theodorou, D. Galaris, L. Nousis, E. S. Kats-

anou, M. N. Alexis, J. Med. Chem. 2006, 49, 300–306.

a) N. S. Dhalla, A. B. Elmoselhi, T. Hata, N. Makino, Cardiov-

asc. Res. 2000, 47, 446–456; b) D. Altavilla, B. Deodato, G. M.

Campo, M. Arlotta, M. Miano, G. Squadrito, A. Saitta, D.

Cucinotta, S. Ceccarelli, M. Ferlito, M. Tringali, L. Minutoli,

A. P. Caputi, F. Squadrito, Cardiovasc. Res. 2000, 47, 515–528.

M. Koufaki, T. Calogeropoulou, A. Detsi, A. Roditis, A. P.

Kourounakis, P. Papazafiri, K. Tsiakitzis, C. Gaitanaki, I. Beis,

P. N. Kourounakis, J. Med. Chem. 2001, 44, 4300–4303.

M. Koufaki, A. Detsi, E. Theodorou, C. Kiziridi, T. Calogero-

poulou, A. Vassilopoulos, A. P. Kourounakis, E. Rekka, P. N.

Kourounakis, C. Gaitanaki, P. Papazafiri, Bioorg. Med. Chem.

2004, 12, 4835–4841.

M. Koufaki, C. Kiziridi, P. Papazafiri, A. Vassilopoulos, A.

Varró, Z. Nagy, A. Farkas, A. Makriyannis, Bioorg. Med.

Chem. 2006, 14, 6666–6678.

M. Koufaki, T. Calogeropoulou, E. Rekka, M. Chryselis, P.

Papazafiri, C. Gaitanaki, A. Makriyannis, Bioorg. Med. Chem.

2003, 11, 5209–5219.

V. Kren in Glycoscience: Chemistry and Chemical Biology, vol. 3

(Eds.: B. Fraser-Reid, K. Tatsuta, J. Thiem), Springer Verlag,

Berlin, 2001, pp. 2471–2529.

A. Tai, S. Goto, Y. Ishiguro, K. Suzuki, T. Nitoda, I. Yamam-

oto, Bioorg. Med. Chem. Lett. 2004, 14, 623–627.

H. Murase, R. Yamauchi, K. Kato, T. Kunieda, J. Terao, Lipids

1997, 32, 73–78.

H. Murase, J.-H. Moon, R. Yamauchi, K. Kato, T. Kunieda,

T. Yoshikawa, J. Terao, Free Radiat. Biol. Med. 1998, 24, 217–

225.

Data Analysis: The Scientist program (MicroMath, Salt Lake City,

USA) was used for all curve-fitting procedures. Details on the cal-

culations [see Equations (1)–(5)] have already been published.^[50,51]

Standard deviations and correlation coefficients are reported.

[14]

[15]

[16]

Molecular Modelling: All calculations were performed in the gas

phase with use of Hyperchem software (Autodesk, Sausalito,

USA). Firstly, the search for the most stable conformation about

the glycosidic bond was achieved by molecular mechanics (MM+

force field). Then, the most stable conformation was further opti-

mized by semiempirical quantum mechanics calculations (PM3 pa-

rametrization, UHF mode) for each phenol and the corresponding

aryloxyl radical. Intramolecular H-bonding was allowed within the

C-glycosyl moiety and, when possible, between the phenolic OH

and the C-glycosyl moiety (O-5).

[17]

[18]

Supporting Information (see also the footnote on the first page of

this article): a) Description of the ZnCl₂-catalyzed reaction between

2-(2,3,4,6-tetra-O-acetyl-β--galactopyranosyl)-hydroquinone and

all-racemic phytol. b) Experimental details for the coupling reac-

tion between 1,2,3,4,6-penta-O-acetyl-β--galactopyranose and

1,4-dimethoxy-2,3-dimethylbenzene as a route to 5-C-β--galacto-

pyranosyl-7,8-dimethylchroman-6-ols and C-β--galactopyranosyl-

γ-tocopherols. c) Evaluation of the solubilities of the prepared acet-

ylated and deacetylated C-β--glycosylchroman-6-ols and C-β--

glycosyltocopherols (-gluco, -galacto series). d) Miscellaneous

bioactivities.

[19]

[20]

[21]

[22]

[23]

Acknowledgments

On the occasion of the 20th anniversary of the scientific and cul-

tural collaboration between Région Rhône-Alpes (France) and

Shanghai City Council (P. R. China), generous financial support,

and in particular stipends (Mobilité Internationale Rhône-Alpes

studentships to H. L. (2002–2005), is gratefully acknowledged.

Support from the the University Claude Bernard, Lyon 1, for co-

tutored theses, from the Ministry of Foreign Affairs (ARCUS

Chine 2005), and from the National Science Foundation of China

(Grant No. 20576034) is also acknowledged.

[24]

[25]

[26]

[27]

a) M. Lahmann, J. Thiem, Carbohydr. Res. 1997, 299, 23–31;

b) S. Witkowski, P. Walejko, Z. Naturforsch. Teil B 2002, 57,

571–578.

R. K. Uhrig, M. A. Picard, K. Beyreuther, M. Wiessler, Car-

bohydr. Res. 2000, 325, 72–80.

K. Shimoda, Y. Kondo, K. Abe, H. Hamada, H. Hamada, Tet-

rahedron Lett. 2006, 47, 2695–2698.

a) S. Satoh, M. Shimojima, K. Ito, T. Kuribayashi, Annu. Rep.

Sankyo. Res. Lab. 2002, 54, 85–104; b) T. Bililign, B. R. Grif-

fith, J. S. Thorson, Nat. Prod. Rep. 2005, 22, 742–760.

a) M. Jay in The Flavonoids: Advances in research since 1986

(Ed.: J. B. Harborne), Chapman & Hall, London, 1994, pp. 57–

93; b) M. Jay, M.-R. Viricel, J.-F. Gonnet in Flavonoids: Chem-

istry, Biochemistry and Applications (Eds.: Ø. M. Andersen,

K. R. Markham), CRC Press, Taylor & Francis Group, Boca

Raton, 2006, pp. 857–915.

[1] C. Schneider, Mol. Nutr. Food Res. 2005, 49, 7–30.

[2] M. H. Gordon, Nat. Prod. Rep. 1996, 13, 265–273.

[3] P. Fresco, F. Borges, C. Diniz, M. P. M. Marques, Med. Res.

Rev. 2006, 26, 747–766.

[4] J. S. Wright, E. R. Johnson, G. A. DiLabio, J. Am. Chem. Soc.

2001, 123, 1173–1183.

[5] R. Amorati, F. Ferroni, G. F. Pedulli, L. Valgimigli, J. Org.

Chem. 2003, 68, 9654–9658.

[6] M. Afri, B. Ehrenberg, Y. Talmon, J. Schmidt, Y. Cohen, A. A.

Frimer, Chem. Phys. Lipids 2004, 131, 107–121.

[7] C. Adelwöhrer, T. Rosenau, W. H. Binder, P. Kosma, Tetrahe-

dron 2003, 59, 3231–3235.

[8] C. Adelwöhrer, T. Rosenau, E. Kloser, K. Mereiter, T. Netscher,

Eur. J. Org. Chem. 2006, 2081–2086.

[9] T. Rosenau, C. Adelwöhrer, A. Hofinger, K. Mereiter, P.

Kosma, Eur. J. Org. Chem. 2004, 1323–1329.

[28]

[29]

[30]

[31]

[32]

[33]

a) O. R. Martin, Carbohydr. Res. 1987, 171, 211–222; b) P.

Verlhac, C. Leteux, L. Toupet, A. Veyrières, Carbohydr. Res.

1996, 291, 11–20; c) N. Girard, C. Rousseau, O. R. Martin,

Tetrahedron Lett. 2003, 44, 8971–8974.

a) C. Jaramillo, S. Knapp, Synthesis 1994, 1–20; b) K. Briner,

A. Vasella, Helv. Chim. Acta 1990, 73, 1764–1778; c) S. Yam-

ago, M. Hashidume, J.-i. Yoshida, Tetrahedron 2002, 58, 6805–

6813; d) G. Giorgi, F. Ponticelli, L. Salvini, A. Trendafilova,

M. Valoti, F. Pessina, Eur. J. Org. Chem. 2003, 106–115; e) E.

Janes, M. Nieger, P. Saarenketo, K. H. Dötz, Eur. J. Org. Chem.

[10] C. Adelwöhrer, T. Rosenau, L. Gille, P. Kosma, Tetrahedron

2003, 59, 2687–2691.

[11] W. Gregor, G. Grabner, C. Adelwöhrer, T. Rosenau, L. Gille,

J. Org. Chem. 2005, 70, 3472–3483.

[12] D. Shanks, R. Amorati, M. G. Fumo, G. F. Pedulli, L. Valgimi-

gli, L. Engman, J. Org. Chem. 2006, 71, 1033–1038.

[13] a) M. Wijtmans, D. A. Pratt, L. Valgimigli, G. A. DiLabio,

G. F. Pedulli, N. A. Porter, Angew. Chem. Int. Ed. 2003, 42,

4370–4373; b) M. Wijtmans, D. A. Pratt, J. Brinkhorst, R.

Swerza, L. Valgimigli, G. F. Pedulli, N. A. Porter, J. Org. Chem.

1882

www.eurjoc.org

Eur. J. Org. Chem. 2008, 1869–1883

Amphiphilic Tocopherol-Derived Antioxidants

2003, 2276–2285; f) G.-R. Chen, J.-P. Praly, C. R. Chim. 2008, [44] G. Spiteller, Chem. Phys. Lipids 1998, 95, 105–162.

in press.

[45] K. S. Montine, J. F. Quinn, J. Zhang, J. P. Fessel, L. J. Roberts,

J. D. Morrow, T. J. Montine, Chem. Phys. Lipids 2004, 128,

117–124.

[46] D. Pratico, J. Rokach, J. Lawson, G. A. FitzGerald, Chem.

Phys. Lipids 2004, 128, 165–171.

[47] B. Halliwell, Cardiovasc. Res. 2000, 47, 410–418.

[48] M. Yang, K. Schaich, Free Radic. Biol. Med. 1996, 20, 225–

236.

[49] D. A. Butterfield, C. M. Lauderback, Free Radic. Biol. Med.

2002, 32, 1050–1060.

[50] M. Roche, C. Dufour, N. Mora, O. Dangles, Org. Biomol.

Chem. 2005, 3, 423–430.

[51] E. Vulcain, P. Goupy, C. Caris-Veyrat, O. Dangles, Free Radiat.

Res. 2005, 39, 547–563.

[52] D. Lieber, J. Burr, Lipids 1995, 30, 789–793.

[53] A. Maciejewski, J. Kubicki, K. Dobek, J. Phys. Chem. B 2005,

109, 9422–9431.

[34] a) J.-P. Praly, L. He, B. B. Qin, M. Tanoh, G.-R. Chen, Tetrahe-

dron Lett. 2005, 46, 7081–7085; b) L. He, Y. Z. Zhang, M.

Tanoh, G.-R. Chen, J.-P. Praly, E. D. Chrysina, C. Tiraidis, M.

Kosmopoulou, D. D. Leonidas, N. G. Oikonomakos, Eur. J.

Org. Chem. 2007, 596–606.

[35] a) T. Kuribayashi, N. Ohkawa, S. Satoh, Tetrahedron Lett.

1998, 39, 4537–4540; b) T. Kuribayashi, Y. Mizumo, S. Gohya,

S. Satoh, J. Carbohydr. Chem. 1999, 18, 371–382.

[36] L. Kalvoda, Coll. Czech. Chem. Commun. 1973, 38, 1679–1692.

[37] a) R. A. Hudson, L. M. V. Tillekeratne, Biologically Active

Natural Products: Pharmaceuticals, CRC Press, 2000, 109–119;

b) M. J. A. Martinez, P. B. Benito, Studies in Natural Products

Chemistry, vol. 30 (Bioactive Natural Products (Part K)), 2005,

303–366; c) J.-M. Gao, Curr. Org. Chem. 2006, 10, 849–871.

[38] Y. Z. Zhang, K. Aouadi, G.-R. Chen, J.-P. Praly, Synthesis

2007, 3473–3488.

[39] M. Delgado, R. E. Wolf Jr, J. R. Hartman, G. McCafferty, R. [54] S. Panja, P. Chowdhury, S. Chakravorti, Chem. Phys. Lett.

Yagbasan, S. C. Rawle, D. J. Watkin, S. R. Cooper, J. Am.

Chem. Soc. 1992, 114, 8983–8991.

2003, 368, 654–662.

[55] L. Castle, M. J. Perkins, J. Am. Chem. Soc. 1986, 108, 6381–

6382.

[56] R. Amorati, M. G. Fumo, S. Menichetti, V. Mugnaini, G. F.

Pedulli, J. Org. Chem. 2006, 71, 6325–6332.

[57] J. Kang, G. Hilmersson, J. Santamaría, J. Rebek Jr, J. Am.

Chem. Soc. 1998, 120, 3650–3656.

[40] L. I. Smith, K. C. Johnson, J. Am. Chem. Soc. 1937, 59, 673–

679.

[41] T. Shimizu, M. Masuda, H. Minamikawa, Chem. Rev. 2005,

105, 1401–1443.

[42] E. A. Decker, Trends Food Sci. Technol. 1998, 9, 241–248.

[43] J. N. Coupland, D. J. McClements, Trends Food Sci. Technol.

1996, 7, 83–91.

Received: September 18, 2007

Published Online: March 3, 2008

Eur. J. Org. Chem. 2008, 1869–1883

www.eurjoc.org

1883

Article Doi

C-D-glucopyranosyl derivatives of tocopherols - Synthesis and evaluation as amphiphilic antioxidants

DOI: 10.1002/ejoc.200700885

Source and publish data:

Authors:

Article abstract of DOI:10.1002/ejoc.200700885

Full text of DOI:10.1002/ejoc.200700885

Products guided by the article

R&D Labs maybe for 868602-40-8

Relevant to this article

Hot Product