Novel tocopherol compounds XI. Synthesis, bromination and oxidation reactions of 3-(5-tocopheryl)propionic acid

Article abstract of DOI:10.1055/s-1999-2594

3-(5-Tocopheryl)propionic acid (γ-tocopherol-5-propionic acid, 13) has been synthesized by a multi-step sequence involving the hetero-Diels-Alder reaction between O-methyl-C,O-bis-(trimethylsilyl)ketene acetal (10) and the ortho-quinone methide of α-tocopherol (5) as the key step. The title compound is the first 5a-substituted tocopherol that shows an oxidation behavior largely similar to α-tocopherol.

Full text of DOI:10.1055/s-1999-2594

LETTER
291
Novel Tocopherol Compounds XI. Synthesis, Bromination and Oxidation
Reactions of 3-(5-Tocopheryl)propionic Acid
Thomas Rosenau,a * Wolf D. Habicher, b Antje Potthast,a Paul Kosma, a
a Christian-Doppler-Laboratory, University of Agricultural Sciences - Vienna, Muthgasse 18, A - 1190 Vienna, Austria
Fax: +43 (1) 36006 6059, trosenau@edv2.boku.ac.at
b Technische Universität Dresden, Institut für Organische Chemie, Mommsenstr. 13, D - 01062 Dresden, Germany
Received 9 November 1998
5-Tocopherylacetic acid (3), carrying a carboxylic substit-
uent at C-5a, forms a reversible redox system which
makes it different from all vitamin E derivatives known so
far.4 However, in 3-(5-tocopheryl)propionic acid (13)
Abstract: 3-(5-Tocopheryl)propionic acid (g-tocopherol-5-prop-
ionic acid, 13) has been synthesized by a multi-step sequence in-
volving the hetero-Diels-Alder reaction between O-methyl-C,O-
bis-(trimethylsilyl)ketene acetal (10) and the ortho-quinone
methide of a-tocopherol (5) as the key step. The title compound is which can be regarded as “homo”-tocopherylacetic acid,
the first 5a-substituted tocopherol that shows an oxidation behavior
largely similar to a-tocopherol.
the direct influence of the substituent on C-5a is interrupt-
ed by a methylene group. Hence, the oxidation behavior
Key words: a-tocopherol, 5a-substituted tocopherols, ortho-quino-
ne methide, ketene acetal, hetero-Diels-Alder reaction
of this compound should “return to normal”, that means,
resemble again that of a-tocopherol. For these reasons, 3-
(5-tocopheryl)propionic acid (13) was expected to be a re-
warding synthetic target as the first 5a-substituted toco-
pherol with possibly largely unchanged oxidation
behavior as compared to the parent compound vitamin E.
a-Tocopherol (vitamin E, 1) is a biological radical scav-
enger and a potent lipophilic antioxidant.1 Derivatives of
a-tocopherol are currently undergoing a surge of interest
as they show changed properties, such as an altered redox
behavior or increased hydrophilicity as compared to vita-
min E. 5a-Substituted tocopherols 2 differ considerably
from a-tocopherol in the chemical behavior, even though
they possess the two structural features crucial for physio-
logical effectiveness, a free phenolic hydroxyl group and
an intact isoprenoid side chain.2 The altered chemical
properties are caused by the electronic influence of the
substituent at C-5a. These effects are much larger than one
would expect, they are even pronounced enough to deter-
mine the reactivity of the whole tocopherol system. The
special role of the 5a-position in vitamin E chemistry is
often named by the rather vague term “Mills-Nixon ef-
fect”, but it is very far from being comprehensively under-
stood.3
The synthesis of 13 turned out to be unexpectedly diffi-
cult. Neither the alkylation of g-tocopherol (6)5 with b-ha-
lopropionic acid derivatives nor the reaction of a-
tocopheryl-Grignard (7)6 with haloacetic acid derivatives
produced the desired compound in exploitable yields.
Other attempts to obtain 13 failed similarly. Consequent-
ly, we used a multi-step approach which was admittedly
less direct, but provided 3-(5-tocopheryl)propionic acid in
a satisfying overall yield. The key step of the sequence is
the ZnCl2-catalyzed hetero-Diels-Alder reaction of O-me-
thyl-C,O-bis-(trimethylsilyl)ketene acetal (10) with the
ortho-quinone methide 5.7
Intermediate 5 is readily formed from 5a-bromo-a-toco-
pherol (4) by elimination of hydrogen bromide at temper-
atures above 50°C8 and is conveniently prepared by
adding a solution of 4 into the preheated reaction mixture
(Scheme 1). The ortho-quinone methide normally under-
goes dimerization to the spiro-dimer of a-tocopherol (15),
unless it is trapped in a faster reaction, as in the present ex-
ample.
Scheme 1
Figure 1
Synlett 1999, No. 3, 291–294 ISSN 0936-5214 © Thieme Stuttgart · New York

292
T. Rosenau et al.
LETTER
To obtain the trapping reagent 10, methyl bromoacetate acid derivative, a surprising result since one should expect
(8) was treated with trimethylsilyl chloride and zinc to a-halogenation. However, the mechanism for the forma-
produce methyl trimethylsilylacetate (9).9 Treatment with tion of 3-(5-tocopheryl)-3-bromopropionic acid (14) is
lithium diisopropyl amide and quenching with an excess completely different, namely a two-step oxidation-addi-
of trimethylsilyl chloride finally provided O-methyl-C,O- tion process analogous to the one which had been estab-
bis-(trimethylsilyl)ketene acetal (10) in a good overall lished for the bromination of a-tocopherol.7 The outcome
yield (Scheme 2).10
of the reaction is consequently a first indication that 1 and
13 behave chemically quite similarly.
Apart from the bromination reaction, the following exper-
iments clearly demonstrated that 3-(5-tocopheryl)prop-
ionic acid (13) and vitamin E (1) react alike in oxidative
processes. Oxidation of 13 with Ag2O in n-hexane pro-
duced the spiro-dimer 16 by dimerization of the interme-
diate ortho-quinone methide in 69% isolated yield.15 An
analogous reaction, the formation of spiro-dimer 15, is
well-known for vitamin E.16 Oxidation of 13 with FeCl3 in
a water/methanol mixture afforded the correspondent
para-quinone 18.17 This reaction too, finds its counterpart
in the conversion of vitamin E to para-tocopheryl quinone
(17).18
Scheme 2
The primary product of the hetero-Diels-Alder reaction
between 5 and an excess of 10, the ortho-ester derivative
11,11 was not isolated, but immediately hydrolyzed to me-
thyl 3-(5-tocopheryl)-2-trimethylsilyl-propionate (12).12
Further treatment with tetrabutylammonium fluoride
(TBAF) and acidic hydrolysis finally produced 3-(5-toco-
pheryl)-propionic acid (13) in 72% overall yield13 from
5a-bromo-a-tocopherol (4), (Scheme 3).
Figure 2
In summary, we have prepared 3-(5-tocopheryl)propionic
acid (13), a less lipophilic, base soluble tocopherol deriv-
ative. Contrary to all 5a-substituted tocopherols known so
far, the compound resembles a-tocopherol in its chemical
behavior as demonstrated by bromination and oxidation
reactions typical of vitamin E. Moreover, 13 appears to be
easily applicable as a starting material for further deriva-
tization due to its carboxylic function as site of attach-
ment. This opens the way to novel tocopherol derivatives
that exhibit some modified properties, such as solubility
or lipophilicity, but maintain essentially the same chemi-
cal behavior as vitamin E itself.
Scheme 3
It is recommended not to replace O-methyl-C,O-bis-(tri-
methylsilyl)ketene acetal (10) by another dienophile. O-
methyl-O-(trimethylsilyl)ketene acetal, for instance, gave
very poor yields when employed instead of 10. The addi-
tional step in the synthesis of 13, the removal of the trim-
ethylsilyl group, is more than compensated by the gain in
yield in the addition step.
Acknowledgement
Bromination of 13 with elemental bromine is a very inter-
esting reaction providing 3-(5-tocopheryl)-3-bromoprop-
ionic acid (14) in almost quantitative yield.14 Thus, this re-
action yields a b-bromocarboxylic acid from a propanoic
Financial support by the German National Scholarship Foundation
(Studienstiftung des Deutschen Volkes, fellowship to T.R.) and by
BASF AG, Ludwigshafen, is gratefully acknowledged. The authors
would like to thank Prof. H.-U. Reissig for helpful advice.
Synlett 1999, No. 3, 291–294 ISSN 0936-5214 © Thieme Stuttgart · New York

LETTER
Novel Tocopherol Compounds
293
fication. An analytical sample of 12 was prepared by
chromatography on neutral aluminum oxide.
References and Notes
(1) For reviews see: Halliwell, B.; Gutteridge J. M. C. Free Radi-
cals in Biology and Medicine; Oxford University Press: New
York, 1989. Machlin, L. J. Vitamin E: a Comprehensive Trea-
tise; Marcel Dekker Inc.: New York, 1980. Traber, M. G.;
Cohn, W.; Muller, D. P. R. In Vitamin E in Health and Disea-
ses; Packer, L.; Fuchs, J., Eds.; Marcel Dekker Inc.: New
York, 1993.
(2) Different examples for 5a-substituted tocopherols, their syn-
thesis and chemical behavior are given in part III - IV and IX
of the current series “Novel tocopherol derivatives”, for part
X see: Rosenau, T.; Habicher, W. D. Tetrahedron Lett., 1997,
38, 5959.
(3) For reviews see: Mills, W. H.; Nixon, I. G. J. Chem. Soc.
1930, 2510. Frank, N. L.; Siegel, J.S. In Advances in Theore-
tically Interesting Molecules; Vol. 3; JAI Press Inc., 1995, p.
209. For discussions of effects in the vitamin E system see:
Behan, J. M.; Dean, F. M.; Johnstone, R. A. W. Tetrahedron
1976, 32, 167. Selander, H.; Nilsson, J. L. G. Acta Chem.
Scand. 1971, 25, 1175.
(13) A 2 M solution of tetrabutylammonium fluoride (TBAF) in
DMSO (20 mL, 40 mmol) was added to a solution of the crude
product 12 in 15 ml of DMF. After stirring for 30 min, 50 mL
of n-hexane were added, and the mixture was washed 5 times
with 50 mL of water. The organic layer was evaporated to a
volume of about 3 mL, diluted with 10 mL of ethanol and 20
mL of 5 N HCl, and refluxed for 2h. To obtain an aqueous so-
lution of 13, the mixture was extracted with ether (3 x 20 mL)
and re-extracted into 50 mL of 5 N NaOH. After 5 min of vi-
gorous stirring, the aqueous layer was washed with ether (3 x
10 mL), which was discarded afterwards. The aqueous phase
was acidified by 5 N HCl and extracted with CH2Cl2 (3 x 20
mL). The combined organic layers were washed once with
ice-cold water, dried over MgSO4, and brought to a volume of
approx. 5mL. Petrol ether was added until the mixture became
slightly cloudy. Cooling for several hours afforded 2.11 g
(72%, referred to 1) pure 3-[3,4-dihydro-6-hydroxy-2,7,8-tri-
methyl-2-(4,8,12- trimethyltridecyl)-2H-1-benzopyran-5-
yl]propionic acid (13) as a yellow, waxy solid. 1H NMR
(CDCl3, CD3COOH): δ 1.83 (m, 2H, 3CH2), 2.10 and 2.14 (2s,
6H, 7aCH3, 8bCH3), 2.50 (t, 2H, 5aCH2-CH2), 2.64 (t, 2H,
4CH2), 2.90 (t, 2H, 5aCH2). 13C NMR (300 MHz, CDCl3 /
CD3COOH) δ: 11.8; 12.3 (8bC; 7aC), 20.6 (4C), 21.5 (5aC), 31.6
(3C), 36.4 (5aCH2-CH2), 75.3 (2C), 115.5; 121.8; 122.9; 124.6;
145.5; 147.3 (ArC), 176.2 (COOH). Anal. calcd. for C31H52O4
(488.75): C, 76.18; H, 10.72. Found: C, 76.24; H, 10.88. For
the reason of clarity, the resonances of the isoprenoid side
chain are not listed here and in the following. The convention
of numbering carbon atoms in vitamin E derivatives is explai-
ned in (14b).
(14) The synthesis of 14 followed the procedure described in (7),
13 was used as the starting material instead of 1. 1H NMR
(CDCl3, CD3COOH): δ 1.84 (m, 2H, 3CH2), 2.11 and 2.18 (2s,
6H, 7aCH3, 8bCH3), 2.64 (t, 2H, 4CH2), 3.05 and 3.22 (2dd, 2H,
5aCHBr-CH2COOH), 5.52 (m, 1H, 5aCHBr-CH2). 13C NMR: δ
11.9; 12.4 (8bC; 7aC), 20.4 (4C), 21.7 (5aC), 31.7 (3C), 33.9
(5aC), 43.5 (5aCHBr-CH2) 75.3 (2C), 116.9; 121.5; 123.2;
124.8; 144.9; 147.0 (ArC), 177.2 (COOH). Anal. calcd. for
C31H51O4Br (567.64): C, 65.59; H, 9.06; Br, 14.08. Found: C,
65.45; H, 8.99; Br, 14.12.
(15) Freshly prepared, dry (!) silver oxide (1.160 g, 5.050 mmol)
was added to a solution of 12 (0.489 g, 1.000 mmol) in 50 mL
of dry n-hexane. The mixture was stirred at rt for 10 min, and
the solids were separated. The residue was chromatographed
on neutral aluminum oxide with n-hexane as the eluent, affor-
ding 0.336 g (69%) of 16. 13C NMR (CDCl3) : δ 11.1 (11C),
11.5 (12C), 11.5 (11’C), 14.6 (12’C), 15.2 (10’C), 18.0 (4C), 19.4
(4’C), 23.8 (13C), 25.2 (13’C), 31.0 (3C), 33.8 (3’C), 33.9 (14’C),
34.3 (14C), 74.6 (2C), 74.6 (2’C), 80.5 (5C), 115.4 (5’C), 115.5
(10C), 121.9 (8C), 123.7 (7C), 127.3 (8’C), 142.4 (9’C), 144.0
(9C), 145.2 (7’C), 145.7 (6C), 201.9 (6’C), CH-CH2-COOH (x
2): 21.5, 25.3, 33.4, 34.6, 175.1, 175.7. MALDI-TOF-MS (si-
napic acid, m/z): 974 (MH+). Anal. calcd. for C62H100O8
(973.48): C, 76.50; H, 13.15. Found: C, 76.62; H, 13.23.
(16) a) Synthesis of 15: Schudel, P.; Mayer, H.; Metzger, J.;
Rüegg, R.; Isler, O. Helv. Chim. Acta 1963, 46, 636. Skinner,
W. A.; Alaupovic, P. J. Org. Chem. 1963, 28, 2854. b) Mecha-
nism of formation and structure: Nilsson, J. L. G.; Branstad, J.
O.; Sievertsson, H. Acta Pharm. Suec. 1968, 5, 509. Fales, H.
M. J. Chem. Soc. Perkin Trans. II 1990, 1005.
(4) Rosenau, T.; Habicher, W. D.; Chen, C. L. Heterocycles 1996,
43, 787.
(5) For an overview see: Ames, S. R. In The Vitamins, Vol.5; Se-
brell, W. H., Harris, R. S., Eds.; Academic Press: New York,
1972; p 218. Pratt, D. E. ACS Symp. Ser. 1992, 507, 54.
(6) Rosenau, T.; Habicher, W.D. Synlett 1996, 5, 427.
(7) The trapping of ortho-quinone methides in hetero-Diels-Alder
reactions with vinyl ethers or ketene acetals as electron-rich
dienophiles has already been investigated: Bolon, D. A.
J. Org. Chem. 1970, 35, 3666; Chapman, O. L.; McIntosh,
C. L. Chem. Comm. 1971, 383.
(8) Preparation, reaction mechanism and trapping of the ortho-
quinone methide intermediate 5: Rosenau, T.; Habicher, W.
D. Tetrahedron 1995, 51, 7919.
(9) Fessenden, R. J.; Fessenden, J. S. J. Org. Chem. 1967, 32,
3535.
(10) Matsuda, I.; Murata, S.; Izumi, Y. J. Org. Chem. 1980, 45,
237. Compound 10 was obtained in 72% yield as a mixture of
the E- and Z-isomer. 1H NMR (300 MHz, CDCl3): δ 0.03 and
0.05 (s, 9H, O-Si(CH3)3), 0.20 and 0.25 (s, 9H, CSi(CH3)3),
2.95 and 3.08 (s, 1H, HC=C), 3.48 and 3.52 (s, 3H, O-CH3).
13C NMR (75 MHz, CDCl3): δ -0.4 and -1.0 (O-Si(CH3)3), -
0.15 and -0.4 (C-Si(CH3)3), 53.7 and 54.8 (O-CH3), 63.8 and
72.5 (s, 1H, HC=C), 161.4 and 164.5 (HC=C).
(11) The ketene acetal 10 is normally employed as a Michael donor
with exceptionally high reactivity towards α,β-unsaturated
carbonyl compounds, see also ref. 9. However, in the present
case it reacts as a dienophile in a [4+2]-cycloaddition. Thus,
intermediate 11 is an ortholactone formed by a ZnCl2-cataly-
zed hetero-Diels-Alder reaction, but not the respective
Michael adduct, methyl 3-[2,7,8-trimethyl-6-trimethylsilylo-
xy-2-(4,8,12-trimethyltridecyl)-chro-man-5-yl]-2-(trimethyl-
silyl)propionate. This was confirmed by NMR. For instance,
the 13C spectrum shows a ortholactone signal at 108.4 ppm,
but no indication of an ester carbon atom.
(12) In a 100 mL flask, a mixture of anhydrous zinc chloride
(0.028 g, 0.220 mmol), compound 10 (4.400 g, 20.080 mmol),
and 10 mL of CH3CN was heated to 70°C. A solution of 5a-
bromo-α-tocopherol (3.058 g; 6.000 mmol) in 10 ml of
CH3CN was slowly added (approx. 15 min) under constant
stirring, and the reaction vessel kept for 1 h at 70°C, then 10
min at 85°C, and cooled to rt. Hydrolysis of the primary pro-
duct was carried out by addition of 5 mL of water and, after 5
min, 10 mL of a saturated NH4Cl solution. 100 mL of n-hexa-
ne were added, the mixture was washed twice with water and
dried over Na2SO4. The oily residue remaining after evapora-
tion of the solvent was used for further reactions without puri-
(17) Preparation of 18 was carried out according to the procedure
given for compound 17 in (16). 1H NMR (CDCl3): δ 1.71 (2H,
t, ArCH2CH2, C-3), 1.94 (3H, s, CH3, C-7a), 2.04 (3H, s, CH3,
C-8b), 2.06 (2H, t, ArCH2CH2, C-4), 2.42 (2H, t,5aCH2-CH2),
2.78 (t, 2H, 5aCH2-CH2-), 5.23 (2H, b, OH and COOH). 13
C
Synlett 1999, No. 3, 291–294 ISSN 0936-5214 © Thieme Stuttgart · New York

294
T. Rosenau et al.
LETTER
(18) Synthesis and properties of para-tocopheryl quinone (17):
NMR: δ 18.3 (4C), 25.3 (5aC), 26.7 (2aC), 34.9 (-CH2-COOH),
40.4 (3C), 72.0 (2C), 146.1 (4aC), 141.3, 141.4, 144.4 (5C, 7C,
8C), 186.7, 186.8 (6C, 8aC). Anal. calcd. for C31H52O5
(504.75): C, 73.76; H, 10.38. Found: C, 73.89; H, 10.39.
John, W. Z. Physiol. Chem. 1938, 252, 222. John, W.; Dietzl,
E.; Emte, W. Z. Physiol. Chem. 1939, 257, 173-175.
Synlett 1999, No. 3, 291–294 ISSN 0936-5214 © Thieme Stuttgart · New York