Integrated synthetic strategy for higher catechin oligomers

Article abstract of DOI:10.1002/anie.201007473

Chain linked: An integrated strategy for the rapid assembly of catechin oligomers has been developed, thus enabling access to higher oligomers ranging from 6- to 24-mers. The method exploits an orthogonal, block-type coupling strategy (see scheme). Copyright

Full text of DOI:10.1002/anie.201007473

Communications
DOI: 10.1002/anie.201007473
Oligocatechins
Integrated Synthetic Strategy for Higher Catechin Oligomers**
Ken Ohmori, Tomohiro Shono, Yuki Hatakoshi, Takahisa Yano, and Keisuke Suzuki*
Considerable attention has recently been focused on procya-
nidins (oligomeric catechins) for their potential bioactivi-
ties.[1] The pioneering work of Tꢀckmantel, Kozikowski, et al.
Scheme 1. Conventional upward-extension approach.
on the cocoa-derived epicatechin linear oligomers showed
that there was a tendency for the higher degree of oligomer-
ization to lead to enhanced bioactivities.[2]
In principle, such oligomers are synthetically accessible
through repeated Friedel–Crafts reactions of catechin units
by exploiting the inherent C8 nucleophilicity and the C4
electrophilicity.[3] An essential issue, however, is the ubiquity
of these dual polar reactivities in the starting materials as well
as the products. Scheme 1 illustrates the issue in the conven-
tional formation of the key interflavan bonds executed in an
upward direction. A key disadvantage is that one needs to use
larger substrates in excess to avoid the multiple coupling.
Recently, we developed a promising approach that relies
on two key innovations (Scheme 2):[4] 1) The capping of the
top of the chain with Br suppresses the C8 nucleophilicity,
thus enabling the downward extension of catechin units.
Significantly, it provided a clear division of roles between the
nucleophilic and electrophilic reaction partners, thus enabling
their equimolar coupling, even for combining oligomers.
Scheme 2. Downward-extension approach by orthogonal and block
coupling.
2) Orthogonal activation,[5] that is, application of alternate
“hard!soft!hard…” activation modes, which enabled rapid
block oligomerization.
Herein we demonstrate competence of the strategy by
selective synthesis of higher oligomers up to the tetracosamer.
The key for the orthogonal activation sequence was the
choice in the C4 leaving groups (Scheme 3). We selected the
ethoxyethoxy (OEE) group for hard activation[6] and the 2,6-
xylylthio (SXy) group for soft activation. The latter choice
was made to suppress the side reaction (see A in Scheme 3)
when phenylthio was employed for the activation.[7] Of
additional note here is the choice of the acetyl protection
for the C3 hydroxy groups, thus securing the a stereochem-
istry at the coupling stages, and use of [D7]benzyl (Bn*)
groups, for simplifying 1H NMR analysis.
[*] Dr. K. Ohmori, Dr. T. Shono, Y. Hatakoshi, T. Yano, Prof. Dr. K. Suzuki
Department of Chemistry, Tokyo Institute of Technology
SORST-JST Agency, 2-12-1, O-okayama
Megro-ku, Tokyo, 152-8551 (Japan)
Fax: (+81)3-5734-2788
E-mail: ksuzuki@chem.titech.ac.jp
[**] This work was supported by Grant-in-Aid for Scientific Research
(No. 21350050) from MEXT (Japan), the Global COE program
(Chemistry), and Shorai Foundation for Science and Technology.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201007473.
4862
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 4862 –4867

coupling with H-(CA)1-H (3, 1.2 equiv relative to 1 or 2) upon
BF3·OEt2 or NIS activation, thus giving Br-(CA)3-H (6) in
high yield with good stereoselectivity (a/b = 95:5; runs 1, 3).
Furthermore, the orthogonal combination was possible, that
is, Br-(CA)2-OAc (1) was selectively activated by BF3·OEt2,
allowing coupling with H-(CA)1-SXy (4) to give Br-(CA)3-
SXy (7) (a/b = 97:3; run 2), whereas NIS activated Br-(CA)2-
SPh (2) to couple with H-(CA)1-OEE (5) to give Br-(CA)3-
OEE (8) (a/b = 95:5; run 4). It should be noted that all the
trimeric building blocks, 6–8, were diastereomerically pure
after removing the minor isomers by preparative thin layer
chromatography (see the Supporting Information).
Scheme 3. The C4 leaving groups and designation. OEE=2-ethoxy-
Nucleophilic trimers H-(CA)3-H (9) and H-(CA)3-SXy
(10) were prepared by debromination (LiAlH4) of 6 and 7,
respectively. Partial reductive deacetylation occurred, but
these by-products were converted into 9 and 10 with acetic
anhydride.
Scheme 5 illustrates syntheses of the hexamers by the
{3+3} coupling. For activation of Br-(CA)3-SXy (7), AgOTf
proved to be the reagent of choice, enabling smooth coupling
with a nucleophilic trimer H-(CA)3-H (9) to give Br-(CA)6-H
(11) as a diastereomerically pure foam in 74% yield[11] [run 1,
Table 2; MALDI-TOF MS (2,5-dihydroxybenzoic acid
ethoxy.
As arbitrary abbreviations for the linear catechin oligo-
mers, we use the following representations in the text:
structure I is expressed as in Br-(CA)2-OEE, where Br
stands for a C8 bromo group of the top catechin unit, (CA)2
for two catechin units, and OEE for the C4 OEE group in the
terminal catechin unit.
For the block synthesis of higher oligomers, trimers were
chosen as the basic building blocks, and preparation of
suitably functionalized trimers was carried out (Scheme 4).
Table 1 summarizes the coupling of bromo-capped dimers 1[8]
and 2[9] with monomers 3, 4, 5[10] ({2 + 1} coupling) with and
without a terminal activatable group. Importantly, bromo-
capped substrates 1 and 2 nicely allowed the equimolar
Table 2: The {3+3} coupling of the trimers.
Run
X
Y
Activator
AgOTf[a]
T [8C] t [h] Product Yield [%]
1
2
3
SXy
OEE
H
H
ꢀ15
ꢀ35
ꢀ30
14
0.7
1
11
11
15
74
85
89
[b]
BF3·OEt2
[c]
OEE SXy BF3·OEt2
[a] 3.2 equiv in the presence of 4ꢀ M.S. (1 gmmolꢀ1). [b] 4.0 equiv
[c] 7.4 equiv.
(DHB) matrix) m/z 4416.6 ([M + Na]+ calcd for
12
1
2
C
13C3 H63 D16681Br16O4223Na: 4416.5)]. Likewise, Br-
267
(CA)3-OEE (8) was activated by BF3·OEt2 in the presence
of 9 to give 11 in 85% yield (run 2).[11]
To confirm the formation of hexamer 11, an independent
synthesis was carried out with the {2 + 4} combination using
structurally defined building blocks 1 and 14,[12] thereby
sharing full a stereochemistry of the interflavan linkages.
Thus, coupling of Br-(CA)2-OAc (1) with H-(CA)4-H (14) was
promoted by BF3·OEt2 to give a 92% yield of hexamer 11,
which was indistinguishable from the samples of 11 listed in
runs 1 and 2 of Table 2.[13]
Having obtained various oligomers, ranging from dimers
to hexamers, we examined removal of the protecting groups,
including the bromine atom. The major obstacle was that
interflavan bonds were prone to cleavage under reductive
conditions once the phenols were liberated. To get around this
difficulty, we opted to remove the three types of protecting
groups in three separate steps. As the most challenging test
case, we describe here the deprotection of Br-(CA)6-H (11;
Scheme 5). Debromination was achieved upon treatment of
bromide 11 with LiAlH4 (THF, RT, 12 h). Some acetyl groups
remained intact, but were completely removed by exposure to
base (KOH aq./1,4-dioxane/ethanol (1:2:2), 908C, 5 h).
Finally, benzyl groups were removed by hydrogenolysis over
Scheme 4. Trimer building blocks. a) LiAlH4, THF, RT, 5.4 h; b) Ac2O,
DMAP, pyr, RT, 1.5 h (87%, 2 steps); c) LiAlH4, THF, RT, 2 h. d) Ac2O,
DMAP, pyr, RT, 1 h (81%, 2 steps). Bn=benzyl, Bn*=[D7]benzyl,
DMAP= N,N-dimethylaminopyridine, pyr= pyridine, SXy=2,6-
xylylthio, THF=tetrahydrofuran.
Table 1: The {2+1} couplings.
Run
X
Y
Activator T [8C] t [h] Product Yield [%] a/b[c]
[a]
1
2
3
4
OAc
H
BF3·OEt2
ꢀ30 0.7
ꢀ35 1.0
6
7
6
8
93
96
84
62
95:5
97:3
95:5
95:5
[a]
OAc SXy BF3·OEt2
SPh
NIS[b]
SPh OEE NIS[b]
H
ꢀ5
1.4
ꢀ15 1.4
[a] 1.1 equiv. [b] 1.2 equiv of NIS in the presence of 4ꢀ M.S.
(1 gmmolꢀ1). [c] Separable by preparative TLC methods; see the
Supporting Information. M.S.=molecular sieves.
Angew. Chem. Int. Ed. 2011, 50, 4862 –4867
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 4863

Communications
Scheme 5. Hexamer building blocks. a) LiAlH4, THF, RT, 12 h; b) 50 wt% KOH aq., 1,4-dioxane, ethanol, 908C, 5 h (76%, 2 steps); c) H2, cat.
ASCA-2, THF, MeOH, H2O, RT, 3 h (46%); d) Ac2O, pyridine, 258C.
ASCA-2 (5% Pd(OH)2 on carbon;[14] THF/MeOH/H2O
(2:2:1), RT, 3 h) to give free catechin hexamer 12. For
coping with high susceptibility of deprotected compounds to
air oxidation, an effective isolation procedure was devised:
1) careful filtration under an inert atmosphere, 2) precipita-
tion (CH3CN) and membrane filtration under an inert
atmosphere to give almost pure compound 12 as a white
powder, and 3) further purification by the preparative HPLC
methods (Inertsil WP300 Diol) to afford pure hexamer 12 in
46% yield as a white powder [MALDI-TOF MS (DHB
En route to the tetracosamer, these hexamer building
blocks were combined to give two dodecamers, 19 and 20
(Scheme 6). Upon treatment with BF3·OEt2, Br-(CA)6-OEE
(16, MW= 4483) was selectively activated and underwent
attack by H-(CA)6-SXy (17, MW= 4452) to afford an
activatable dodecamer, Br-(CA)12-SXy (19, MW= 8845), in
88% yield. As for the other {6 + 6} coupling, the combination
of I2 and Ag2O proved to be effective to allow clean union of
Br-(CA)6-SXy (15, MW= 4531) and H-(CA)6-H (18, MW=
4316), thus giving terminal dodecamer Br-(CA)12-H (20,
MW= 8708) in 82% yield. Debromination and acetylation of
20 afforded the nucleophilic dodecamer H-(CA)12-H (21,
MW= 8629).
Now the stage was set for addressing the accessibility to
tetracosamer by the {12 + 12} coupling. Pleasingly, upon
exposure of a mixture of Br-(CA)12-SXy (19, MW= 8845)
and H-(CA)12-H (21, MW= 8629) to the combined mixture of
I2 and Ag2O, the desired linear oligomer, Br-(CA)24-H (22,
MW= 17336) was cleanly obtained in 80% yield.[15] Partic-
ularly significant is that combination of two large substrates is
possible by employing essentially equimolar quantities (1:1.2
molar ratio) of the coupling partners to give 22 with a single
molecular-weight dispersion, clearly indicating the efficacy of
the integrated strategy for assembling linear catechin oligo-
mers.
12
16
matrix) m/z 1753.15 ([M + Na]+ calcd for
C O
1H74 23Na:
90 36
1753.40)]. For characterization purposes, a small sample of 12
was subjected to acetylation to give the corresponding
peracetate [MALDI-TOF MS (DHB matrix) m/z 3014.9
([M + Na]+ calcd for 12C14913C1 H134
O
66
23Na: 3014.7)].
1
16
To test the scope of the block synthesis, we examined the
synthesis of even larger oligomers. An initial question was the
accessibility of the activatable hexameric building blocks.
Pleasingly, equimolar orthogonal coupling of trimers was
again possible; the treatment of a mixture of Br-(CA)3-OEE
(8) and H-(CA)3-SXy (10) with BF3·OEt2 allowed selective
formation of Br-(CA)6-SXy (15) in 89% yield (run 3, Table 2).
Scheme 6 outlines the higher-oligomer synthesis directed
at tetracosamer 22. Among four hexamer building blocks, two
were prepared from Br-(CA)6-SXy (15), a versatile precursor.
Br-(CA)6-OEE (16) was obtained quantitatively by switching
the leaving group (SXy!OEE) through treatment of 15 with
I2 and Ag2O in the presence of 2-ethoxyethanol. Two
debrominated hexamers, H-(CA)6-SXy (17) and H-(CA)6-H
(18), were prepared by debromination of 15 and 11, respec-
tively, with LiAlH4. Some missing acetyl groups were
recovered by the acetylation as before.
1H NMR analysis rapidly became difficult with the
increased chain length of oligomers containing multiple
oxygen atoms (Figure 1; see also the Supporting Informa-
tion). Additional complications come from the atropisomer-
ism resulting from the hindered rotation around the inter-
flavan bonds, which is slow on the 1H NMR time scale
(500 MHz). However, interesting observations were made:
4864 www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 4862 –4867

Scheme 6. Block synthesis of tetracosamer 22 (unless otherwise noted, the reactions were performed at ambient temperatures). a) 2-ethoxyethanol,
I2, Ag2O, CH2Cl2, 4ꢀ M.S., ꢀ788C, 2 h. (99%, a/b=81:19); b) LiAlH4, 7 h; c) Ac2O, DMAP, pyr, 18.5 h (69%, 2 steps); d) LiAlH4, THF, 3 h. e) Ac2O,
DMAP, pyr, 1 h (90%, 2 steps); f) BF3·OEt2, CH2Cl2, ꢀ78!ꢀ258C, 1 h (88%); g) I2, Ag2O, 4ꢀ M.S., CH2Cl2, ꢀ788C, 1 h (82%); h) LiAlH4, THF,
10.5 h; i) Ac2O, DMAP, pyr, 10.5 h (90%, 2 steps); j) I2, Ag2O, 4ꢀ M.S., CH2Cl2, ꢀ788C, 2 h (80%).
1) oligomers larger than a dodecamer showed much simpler
spectra and 2) the proton signals of the internal units
gradually converge with slight broadening as the number of
catechin units increases. Notably, chemical shifts for the C2
benzylic protons of the internal catechin units appeared
upfield (ca. d = 2.9 ppm) in comparison with a typical benzylic
proton attached to an oxygen atom. The tendency suggests
dominance of higher-order structures that we assume to be
helical. Molecular modeling suggested anisotropic effects of
the proximal benzene rings in the C5 benzyl groups of the
upper catechin units when the local conformation between
two consecutive catechin units adopt a right-handed helix (B’;
Scheme 7).
In conclusion, rapid assembly of higher linear catechin
oligomers has become possible through block synthesis based
on an orthogonal strategy. The bromo-capped electrophilic
Scheme 7. Local conformation around interflavan bonds.
Angew. Chem. Int. Ed. 2011, 50, 4862 –4867
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 4865

Communications
Figure 1. Stacked 1H NMR spectra of protected oligomers (500 MHz, CDCl3, 300 K).
units were subjected to equimolar coupling even with larger
nucleophilic oligomeric substrates (up to a dodecamer), thus
providing efficient entries into controlled oligomer formation.
Efficacy of the strategy was demonstrated by the synthesis of
the tetracosamer 22 having a single molecular-weight dis-
persion. Formation and deprotection of higher oligomers are
now in progress in our laboratory.
nopoulos, G.-J. E. Nychas, S. A. Haroutounian, J. Agric. Food
Chem. 2009, 57, 457 – 463.
[2] A. P. Kozikowski, W. Tꢀckmantel, G. Boettcher, L. J. Romanc-
zyk Jr, J. Org. Chem. 2003, 68, 1641 – 1658.
[3] Selected examples: a) R. D. Alharthy, C. J. Hayes, Tetrahedron
Lett. 2010, 51, 1193 – 1195; b) A. Saito, Y. Mizushina, A. Tanaka,
N. Nakajima, Tetrahedron 2009, 65, 7422 – 7428; c) K. Oyama, M.
Kuwano, M. Ito, K. Yoshida, T. Kondo, Tetrahedron Lett. 2008,
49, 3176 – 3180; d) A. Saito, N. Nakajima, N. Matsuura, A.
Tanaka, M. Ubukata, Heterocycles 2004, 62, 479—489 and
related references therein; e) W. Tꢀckmantel, A. P. Kozikowski,
L. J. Romanczyk, Jr., J. Am. Chem. Soc. 1999, 121, 12073 – 12081;
f) P. J. Steynberg, R. J. J. Nel, H. Van Rensburg, B. C. B. Bezui-
denhoudt, D. Ferreira, Tetrahedron 1998, 54, 8153 – 8158; g) S.
Yoneda, H. Kawamoto, F. Nakatsubo, J. Chem. Soc. Perkin
Trans. 1 1997, 1025 – 1030; h) H. Kawamoto, N. Tanaka, F.
Nakatsubo, K. Murakami, Mokuzai Gakkaishi 1993, 39, 820 –
824; i) H. Kawamoto, F. Nakatsubo, K. Murakami, Mokuzai
Gakkaishi 1991, 37, 741 – 747; j) H. Kawamoto, F. Nakatsubo, K.
Murakami, J. Wood Chem. Technol. 1990, 10, 59 – 74; k) L. Y.
Foo, R. W. Hemingway, J. Chem. Soc. Chem. Commun. 1984,
85 – 86; l) R. W. Hemingway, L. Y. Foo, J. Chem. Soc. Chem.
Commun. 1983, 1035 – 1036.
[4] K. Ohmori, N. Ushimaru, K. Suzuki, Proc. Natl. Acad. Sci. USA
2004, 101, 12002 – 12007.
[5] O. Kanie, Y. Ito, T. Ogawa, J. Am. Chem. Soc. 1994, 116, 12073 –
12074.
[6] Monomeric C4 acetate proved to be less reactive as a nucleo-
philic building block for higher oligomer synthesis, due presum-
ably to the decreased electron density at its C8-position by
inductive electron-withdrawal. For the ethoxyethoxy group in
hard activation, see Ref. [3d].
Received: November 29, 2010
Revised: January 20, 2011
Published online: February 25, 2011
Keywords: catechins · flavonoids · oligomers · polyphenols ·
.
synthetic methods
[1] a) A. B. Bohm, Introduction to Flavonoids, Harwood Academic
Publishers, Amsterdam, 1998; b) Plant Polyphenols 2: Chemis-
try, Biology, Pharmacology, Ecology (Eds.: G. G. Gross, R. W.
Hemingway, T. Yoshida, S. J. Branham), Kluwer Academic/
Plenum Publishers, New York, 1999; c) T. D. Bruyne, L. Pieters,
M. Witvrouw, E. D. Clercq, D. V. Berghe, A. J. Vlietinck, J. Nat.
Prod. 1999, 62, 954 – 958; d) Y. Hamauzu, H. Yasui, T. Inno, C.
Kume, M. Omanyuda, J. Agric. Food Chem. 2005, 53, 928 – 934;
e) Flavonoids: Chemistry, Biochemistry and Applications, (Eds.:
Ø. M. Andersen, K. R. Markham, CRC Press/Taylor & Francis,
Boca Raton, 2006; f) M. Kusuda, K. Inada, T. Ogawa, T.
Yoshida, S. Shiota, T. Tsuchiya, T. Hatano, Biosci. Biotechnol.
Biochem. 2006, 70, 1423 – 1431; g) Y. Hamauzu, C. Kume, H.
Yasui, T. Fujita, J. Agric. Food Chem. 2007, 55, 1221 – 1226; h) R.
Mayer, G. Stecher, R. Wuerzner, R. C. Silva, T. Sultana, L.
Trojer, I. Feuerstein, C. Krieg, G. Abel, M. Popp, O. Bobleter,
G. K. Bonn, J. Agric. Food Chem. 2008, 56, 6959 – 6966; i) M.
Zhuang, H. Jiang, Y. Suzuki, X. Li, P. Xiao, T. Tanaka, H. Ling,
B. Yang, H. Hiroki, L. Zhang, C. Qin, K. Sugamura, T. Hattori,
Antiviral Res. 2009, 82, 73 – 81; j) M. Anastasiadi, N. G. Choria-
[7] Z. Li, J. C. Gildersleeve, J. Am. Chem. Soc. 2006, 128, 11612 –
11619.
[8] Br-(CA)2-OAc (1) was prepared by combining Br-(CA)1-SPh
and H-(CA)1-OAc promoted by NIS (83% yield, a/b = 95:5).
For details, see the Supporting Information.
4866 www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 4862 –4867

[9] Br-(CA)2-SPh (2) was prepared by combining Br-(CA)1-OAc
and H-(CA)1-SPh in the presence of BF3·OEt2 (95% yield, a/b =
94:6). For details, see the Supporting Information.
[10] For preparation of monomers 3–5; see the Supporting Informa-
tion.
[13] These results imply that a new a linkage could be formed at the
coupling stage. The viability was clearly shown for assembling
the linear structures. In spite of the large molecules with many
functionalities, excellent selectivity/reactivity persisted for regio-
chemistry at C4 and C8.
[11] 1H NMR analysis of crude reaction mixture suggested a mixture
of diastereomers. However, after purification, a single diaste-
reomer was isolated; see the Supporting Information.
[12] H-(CA)4-H (14) was prepared by combining Br-(CA)2-OEE and
H-(CA)2-H with BF3·OEt2 and then debromination (LiAlH4)
and O acetylation (63% yield).
[14] N. E. CHEMCAT Co.
[15] Diffusion coefficients (D) of Br-(CA)6-H (11), Br-(CA)12-H (20),
and Br-(CA)24-H (22) were measured by the diffusion-ordered
NMR experiments (CDCl3, 300 K). The values (D) were
estimated as follows; 11: (3.36 ꢁ 0.19) 10ꢀ10 m2 sꢀ1, 20: (2.43 ꢁ
0.12) 10ꢀ10 m2 sꢀ1, 22: (1.95 ꢁ 0.07) 10ꢀ10 m2 sꢀ1, which were well
correlated to the order of molecular sizes.
Angew. Chem. Int. Ed. 2011, 50, 4862 –4867
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 4867