A new and convenient one-step synthesis of the natural 3-deoxyanthocyanidins apigeninidin and luteolinidin chlorides from 2,4,6-triacetoxybenzaldehyde

Article abstract of DOI:10.1055/s-2003-40982

The total synthesis of apigeninidin (1), luteolinidin (2) and 5,7-dihydroxyflavylium (5) chlorides is performed through a one step reaction from an acetylated derivative of a commercial reagent. Condensation reaction between 2,4,6-triacetoxybenzaldehyde and an acetophenone derivative in anhyd methanolic hydrogen chloride provides the 3-deoxyanthocyanidins in high yields.

Full text of DOI:10.1055/s-2003-40982

1878
SHORT PAPER
A New and Convenient One-Step Synthesis of the Natural 3-Deoxyantho-
cyanidins Apigeninidin and Luteolinidin Chlorides from 2,4,6-Triacetoxy-
benzaldehyde
A
pigeninidin an
h
d
L
uteolinid
i
in
C
hlo
e
rides from
r2,4,6-Tr
r
iacetoxy
y
benzaldehyde Mas*1
Laboratoire de Pharmacognosie, Université Victor Ségalen Bordeaux 2, 146 rue Léo Saignat, 33 076 Bordeaux Cedex, France
E-mail: mthierry@ucmail.ucm.es
Received 11 April 2003; revised 19 May 2003
Abstract: The total synthesis of apigeninidin (1), luteolinidin (2)
and 5,7-dihydroxyflavylium (5) chlorides is performed through a
one step reaction from an acetylated derivative of a commercial re-
agent. Condensation reaction between 2,4,6-triacetoxybenzalde-
hyde and an acetophenone derivative in anhyd methanolic hydrogen
chloride provides the 3-deoxyanthocyanidins in high yields.
Key words: total synthesis, natural products, heterocycles, polycy-
cles, phenols
Figure 1 3-Deoxyanthocyanidins; derivatives of the flavylium (2-
phenylbenzopyrylium) cation
Apigeninidin (1) and luteolinidin (2) chlorides (Figure 1)
phloroglucinaldehyde.10 The products were however gen-
erally obtained as crude mixtures after three or four steps
and in 30% yields in the best cases. More recently, in the
eighties, Sweeny and Iacobucci carried out a method in
which these natural derivatives were synthesized through
an oxidative decarboxylation of the corresponding 4-car-
boxy-2-flavenes obtained from Bülow’s intermediates.11
This synthesis proceeds in four steps and in overall yields
of around 40% from phloroglucinol dimethyl ether. An-
other method consists in the chemical conversion of fla-
vonoid precursors: Sweeny and Iacobucci thus obtained
apigeninidin chloride in two steps from 4 ,5,7-triacetoxy-
flavan in 21% yield.12
are the most common of the natural 3-deoxyanthocyani-
dins, a group of yellow–orange flavylium pigments.2
These 5,7-dihydroxyflavylium salts, possessing at least
one additional hydroxyl group in the 4 position, are con-
tained in food plants such as corn and sorghum.3 Apige-
ninidin has been earlier investigated as a food colorant,4
and hair dye preparations containing this substance have
more recently been patented,5 which underlines the possi-
ble industrial use of these colored derivatives.
In previous studies,6–8 we have shown that natural 3-O-
glucosylated flavylium salts (the monoglucosylated an-
thocyanins from Vitis vinifera grapes) possess DNA tri-
plex stabilizing properties. Such properties are useful in
the antigene strategy, a potential therapeutic approach
consisting in the regulation of gene expression by nucleic
acids, since one of its limitations pertains to the weak sta-
bility of the triplex formed between a target duplex and a
synthetic oligonucleotide.6,7,9 Concerning the stabilizing
properties of the monoglucosylated flavylium salts previ-
ously studied, we have demonstrated that the sugar moiety
at their 3 position could preclude efficient interaction with
the triplex and so limit the activity.6 Thus, we decided to
study non-glucosylated derivatives such as 1 and 2,6,7 that
we have chosen to obtain by total synthesis, since their ex-
traction is greatly time-consuming and leads generally to
only small amounts of substances. In the present work, we
report a new and convenient synthesis of these cationic
derivatives.
All these reactions, though leading to the desired com-
pounds, proceed in several steps and in limited yields.
Moreover, mixtures of organic salts are often formed dur-
ing these syntheses, which can greatly complicate the pu-
rification of the product. It should be underlined here that
purification is a crucial and difficult step in the synthesis
of such hydroxylated flavylium derivatives.
We propose here an interesting alternative to the above-
mentioned reactions. Our method leads to pure 3-deoxy-
anthocyanidins in good yields and in one step from a com-
mon acetylated derivative. Moreover, only very small
amounts of salt by-products are formed, which allows eas-
ier and faster purification steps.
We first used a synthetic scheme derived from the most
current one for the preparation of flavylium ions non-hy-
droxylated in the 5 position (5-deoxyflavylium ions). The
reaction consists of an acid-catalyzed condensation be-
tween an acetophenone derivative and an ortho-hydroxy-
benzaldehyde [a stream of anhyd HCl (g) is bubbled into
the solution of these reagents in EtOAc or HCO2H13]. Us-
ing this scheme, mono- or polyhydroxylated 5-deoxyfla-
vylium chlorides can be readily obtained in good yields
The synthesis of apigeninidin and luteolinidin chlorides
were firstly realized by Robinson et al. in the thirties by
condensation between an acetophenone and a protected
Synthesis 2003, No. 12, Print: 02 09 2003. Web: 30 07 2003.
Art Id.1437-210X,E;2003,0,12,1878,1880,ftx,en;Z05003SS.pdf.
DOI: 10.1055/s-2003-40982
© Georg Thieme Verlag Stuttgart · New York

SHORT PAPER
A New and Convenient One-Step Synthesis of 3-Deoxyanthocyanidins
1879
without any protection of the hydroxyl groups of the re- anthocyanidins after filtration and washing with the same
agents. However, when 2,4,6-trihydroxybenzaldehyde (3) solvent. The pure products are then obtained by simple
is used without any protection, the yields of 3-deoxyan- crystallization in MeOH–HCl (12 N; 95:5) for 1, and in
thocyanidin are very low whatever the conditions. Indeed, acidified EtOAc–MeOH (80:20) for 2. These purification
compound 3 mainly gives complex mixtures of salts from steps (easier in comparison with those of other syntheses)
which the desired compound, formed in small amount, is are due both to the low quantity of salt by-products
hardly isolated. For example, using this synthetic scheme, formed in the reaction and to their relatively high solubil-
final yields of compound 1 from 3 range between 5% and ity in methanol. Compound 5 was purified by preparative
13% in different solvent and temperature conditions HPLC on a C-18 reversed phase column. Structure and
(EtOAc or HCO2H; 0 °C, r.t. or 60 °C), even with a large purity of the synthesized derivatives were confirmed by
excess of acetophenone.
1H and 13C NMR analysis, UV-visible absorption, and
mass spectroscopy.
We greatly improved these preliminary results by using
instead of 3, 2,4,6-triacetoxybenzaldehyde (4) (obtained Moreover, in order to check whether the method could be
by a simple O-acetylation of 3 by acetic anhydride in di- performed on a larger scale, we realized a gram-scale syn-
ethyl ether in the presence of anhyd potassium thesis of apigeninidin (1). Following the general proce-
carbonate14), and by replacing the classical solvents used dure, 5 g of the pure 3-deoxyanthocyanidin was obtained
in the flavylium synthesis by methanol. In our optimized without any difficulty from 6 g of 2,4,6-triacetoxybenzal-
procedure, an anhydrous methanolic solution of 4 is added dehyde (4), which opens the way to potential industrial us-
dropwise to an anhyd hydrogen chloride methanolic solu- es.
tion of the acetophenone in excess at 0 °C. This leads rap-
idly to the corresponding 3-deoxyanthocyanidins 1 and 2
In summary, we have developed a simple, rapid, and con-
venient one step synthesis of the 3-deoxyanthocyanidins
without any deprotection step, and in respectively 84%
apigeninidin and luteolinidin chlorides from a common
and 67% yields after work-up and purification
acetylated derivative. With this new method, these are
(Scheme 1). In order to check whether this method could
rapidly obtained in a pure state and in very good yields
be generalized to other 5,7-dihydroxyflavylium chlorides,
since by-products, only present in small amount, are easi-
ly eliminated. Furthermore, this cheap and efficient syn-
we synthesized the less polar compound 5 (not hydroxy-
lated on the B-ring) using the same procedure. This latter
thesis can be carried out on a large scale, compatible with
compound is also easily and rapidly obtained, in 55%
potential industrial uses.
yield (Scheme 1).
2,4,6-Trihydroxybenzaldehyde, 4-hydroxyacetophenone and ace-
R1
tophenone were of commercial source. 3,4-Dihydroxyacetophe-
O
R2
HO
OH
H
AcO
OAc
none was obtained by reduction of 2-chloro-3 ,4 -
dihydroxyacetophenone with zinc powder in THF–HOAc (4:1).15
NMR spectra were recorded on a Bruker AMX 500 spectrometer.
The UV-VIS spectra were measured on a Hitachi U-200 spectrom-
eter. The MS spectra were recorded on a Finigan Mat T Q700 spec-
trometer.
Ac2O, K2CO3
Et2O, r.t.
acetophenone
O
O
MeOH–HCl
0 °C
OH
OAc
H
4
3
R1
Apigeninidin Chloride (1); General Procedure
Cl
A stream of anhyd gaseous HCl was first passed through an anhyd
methanolic solution (50 mL) of 4-hydroxyacetophenone (840 mg;
6.10 mmol) at 0 °C, until the concentration of HCl rose to 15% by
mass. A solution of 4 (240 mg; 0.86 mmol) in anhyd MeOH (15
mL) was then added dropwise at 0 °C. At the end of the addition,
the reaction mixture was kept for a few minutes at r.t. and evaporat-
ed to dryness. The resulting residue was vigorously stirred in Et2O,
which afforded the crude 3-deoxyanthocyanidin after filtration and
washing with the same solvent. The pure product apigeninidin (1)
was obtained by crystallization in MeOH–HCl (12 N; 95:5).
HO
O
R2
1: R1 = OH, R2 = H; 84%
2: R1 = R2 = OH;
5: R1 = R2 = H;
67%
55%
OH
Scheme 1
In each case, 4 reacts as soon as it is added to the solution
of the acetophenone (reaction followed by thin-layer
chromatography). The first step must correspond to a clas-
sical condensation between this latter compound and 4
(nucleophilic attack of the enol form on the carbonyl of 4).
In a second step, the obtained adduct may undergo rapid
de-acetylation and cyclisation to give the flavylium cation
after loss of water molecules. Cyclization and dehydration
must proceed via a classical acidic catalyzed pathway,
while de-acetylation should proceed via an acidic metha-
nolysis. At the end of the addition of 4, evaporation of the
solvent to dryness followed by vigorous stirring of the in-
soluble residue in diethyl ether affords the crude 3-deoxy-
Yield: 210 mg (84%).
1H NMR [500 MHz, CD3OD–TFA (99:1)]: = 6.64 (d, J = 2.0 Hz,
1 H, H-6), 6.92 (dd, J = 2.0, 0.7 Hz, 1 H, H-8), 7.06 (d, J = 9.0 Hz,
2 H, H-3 and H-5 ), 8.01 (d, J = 8.7 Hz, 1 H, H-3), 8.27 (d, J = 9.0
Hz, 2 H, H-2 , H-6 ), 9.06 (dd, J = 8.7, 0.7 Hz, 1 H, H-4).
13C NMR [125 MHz, CD3OD–TFA (99:1)]: = 96.1 (C-8), 103.2
(C-6), 110.3 (C-3), 113.8 (C-4a), 118.4 (C-3 and C-5 ), 121.3 (C-
1 ), 133.1 (C-2 and C-6 ), 149.5 (C-4), 160.0 (C-8a), 160.6 (C-5),
167.3 (C-4 ), 172.4 (C-7), 172.7 (C-2).
MS (FAB+, from glycerin suspension): m/z = 255 (M+ =
+
C15H11O4 ).
Synthesis 2003, No. 12, 1878–1880 © Thieme Stuttgart · New York

1880
T. Mas
SHORT PAPER
UV/Vis (H2O, 0.1 N HCl): max ( ) = 273 (20 200), 315 (6 500), 467
nm (31 600 M–1cm–1).
Acknowledgment
I would like to express my sincere thanks to the MENRT (Ministère
de l’Education Nationale, de la Recherche et de la Technologie) for
financial support, and to Professor J. Vercauteren for having welco-
med me in the Laboratoire de Pharmacognosie de Bordeaux when
he was Director of this laboratory.
Larger Scale Synthesis
A stream of anhyd gaseous HCl was first passed through an anhyd
methanolic solution (400 mL) of 4-hydroxyacetophenone (20.7 g;
150 mmol) at 0 °C, until the concentration of HCl rose to 15% in
mass. A solution of 4 (6.00 g; 21.4 mmol) in anhyd MeOH (100 mL)
was then added dropwise at 0 °C. At the end of the addition, the re-
action mixture was kept for a few minutes at r.t. and evaporated to
dryness. The resulting residue was vigorously stirred in Et2O, which
afforded the crude apigeninidin after filtration and washing with the
same solvent. The pure product apigeninidin (1) was obtained by
crystallization in MeOH–HCl (12 N).
References
(1) Present address: Departamento de Química Orgánica I,
Facultad de Ciencias Químicas, Universidad Complutense,
28040 Madrid, Spain, Fax: +34(9)13944103.
(2) (a) Strack, D.; Wray, V. In The Flavonoids, Advances in
Research since 1986; Harborne, J. B., Ed.; Chapman and
Hall: London, 1994, 1–22. (b)Brouillard, R.;Dangles, O. In
The Flavonoids, Advances in Research since 1986;
Harborne, J. B., Ed.; Chapman and Hall: London, 1994,
565–588.
Yield: 5.04 g (81%).
Luteolinidin Chloride (2)
Using the procedure described above, luteolinidin chloride (2) was
obtained from 2,4,6-triacetoxybenzaldehyde (4) (1.00 g).
(3) (a) Mazza, G.; Miniati, E. In Anthocyanins in Fruits,
Vegetables and Grains; CRC Press: Boca Raton, 1993, 225–
248. (b) Timberlake, C. F.; Bridle, P. In Anthocyanins as
Food Colors; Markakis, P., Ed.; Academic Press: New York,
1982, 126–162.
(4) (a) Sweeny, J. G.; Iacobucci, G. A. J. Agric. Food Chem.
1983, 31, 531. (b) Iacobucci, G. A.; Sweeny, J. G. Coca-
Cola Co, Eur. Patent Applic. 24731, 1979; Chem. Abstr.
1981, 95: 5399h.
Yield: 730 mg (67%).
1H NMR [500 MHz, CD3OD–TFA (99:1)]: = 6.63 (d, J = 2.0 Hz,
1 H, H-6), 6.87 (d, J = 2.0 Hz, 1 H, H-8), 6.98 (d, J = 8.5 Hz, 1 H,
H-5 ), 7.65 (d, J = 2.3 Hz, 1 H, H-2 ), 7.78 (dd, J = 8.5, 2.3 Hz, 1
H, H-6 ), 7.94 (d, J = 8.7 Hz, 1 H, H-3), 8.95 (d, J = 8.7 Hz, 1 H,
H-4).
13C NMR [125 MHz, CD3OD–TFA (99/1)]: = 96.1 (C-8), 103.3
(C-6), 110.6 (C-3), 113.5 (C-4a), 115.9 (C-2 ), 117.9 (C-5 ), 121.6
(C-1 ), 125.3 (C-6 ), 148.0 (C-3 ), 149.0 (C-4), 156.1 (C-4 ), 159.6
(C-8a), 160.2 (C-5), 171.9 (C-7), 172.4 (C-2).
(5) Darmenton, P.; Philippe, M. L’Oreal S.A. FR 2757383,
1996; Chem. Abstr. 1998, 129, 140458n.
(6) Mas, T.; Susperregui, J.; Berké, B.; Chèze, C.; Moreau, S.;
Nuhrich, A.; Vercauteren, J. Phytochemistry 2000, 53, 679.
(7) Mas, T.; Susperregui, J.; Moreau, S.; Vercauteren, J. Color
of Wine: A New Therapeutic Tool? In Oenologic Actualities
1999; 6th International Oenology Symposium: Bordeaux,
1999.
MS (FAB+, from glycerin suspension): m/z = 271 (M+ =
+
C15H11O5 ).
UV/Vis (H2O, 0.1 N HCl):
600 M–1 cm–1).
( ) = 276.5 (18 400), 481 nm (34
max
(8) Vergé, S.; Mas, T.; Nay, B.; Arnaudinaud, V.; Soulet, S.;
Richard, T.; Castagnino, C.; Delaunay, J. C.; Chèze, C.;
Monti, J. P.; Vercauteren, J. Scientific Registers of the
European Institute 2000, 9.
(9) Thuong, N. T.; Hélène, C. Angew. Chem., Int. Ed. Engl.
1993, 32, 666.
(10) (a) Robinson, G. M.; Robinson, R.; Todd, A. R. J. Chem.
Soc. 1934, 809. (b) Leon, A.; Robinson, R. J. Chem. Soc.
1931, 2732. (c) Pratt, D. D.; Robinson, R. J. Chem. Soc.
1925, 1128.
5,7-Dihydroxyflavylium Chloride (5)
Using the procedure described above, compound 5 was obtained
from 2,4,6-triacetoxybenzaldehyde 4 (500 mg).
Yield: 345 mg (55%).
1H NMR [500 MHz, CD3OD–TFA (99:1)]: = 6.72 (d, J = 1.8 Hz,
1 H, H-6), 7.03 (dd, J = 1.8, 0.7 Hz, 1 H, H-8), 7.70 (dd, J = 7.8, 7.5
Hz, 2 H, H-3 , H-5 ), 7.79 (t, J = 7.5 Hz, 1 H, H-4 ), 8.20 (d, J = 8.4
Hz, 1 H, H-3), 8.35 (d, J = 7.8 Hz, 2 H, H-2 , H-6 ), 9.27 (dd unre-
solved, J = 8.4 Hz, 1 H, H-4).
(11) Sweeny, J. G.; Iacobucci, G. A. Tetrahedron 1981, 37, 1481.
(12) Sweeny, J. G.; Iacobucci, G. A. Tetrahedron 1977, 33, 2927.
(13) (a) Pratt, D. D.; Robinson, R. J. Chem. Soc. 1923, 745.
(b) Perkin, W. H.; Robinson, R.; Turner, M. R. J. Chem. Soc
1908, 93, 1085. (c) Perkin, W. H.; Robinson, R. Proc.
Chem. Soc. 1907, 19, 149.
13C NMR [125 MHz, CD3OD–TFA (99:1)]: = 96.4 (C-8), 104.0
(C-6), 111.2 (C-3), 116.3 (C-4a), 129.7 (C-2 and C-6 ), 130.8 (C-
1 ), 131.1 (C-3 and C-5 ), 136.4 (C-4 ), 150.9 (C-4), 161.0 (C-8a),
161.2 (C-5), 171.9 (C-2), 174.6 (C-7).
MS (FAB+, from glycerin suspension): m/z = 239 (M+ =
+
C15H11O3 ).
(14) Dean, F. M.; Patamapongse, C. J. Chem. Soc., Perkin Trans.
1 1974, 8, 952.
(15) El Hajji, H.; Dangles, O.; Statoua, A.; Brouillard, R.
Polyphénols Actualités 1996, 15, 7.
Synthesis 2003, No. 12, 1878–1880 © Thieme Stuttgart · New York