Studies in polyphenol chemistry and bioactivity. 4. Synthesis of trimeric, tetrameric, pentameric, and higher oligomeric epicatechin-derived procyanidins having all-4β,8-interflavan connectivity and their inhibition of cancer cell growth through cell cycle arrest

VOLUME 68, NUMBER 5

MARCH 7, 2003

Stu d ies in P olyp h en ol Ch em istr y a n d Bioa ctivity. 4.¹Syn th esis of

Tr im er ic, Tetr a m er ic, P en ta m er ic, a n d High er Oligom er ic

Ep ica tech in -Der ived P r ocya n id in s Ha vin g All-4â,8-In ter fla va n

Con n ectivity a n d Th eir In h ibition of Ca n cer Cell Gr ow th th r ou gh

Cell Cycle Ar r est¹

Alan P. Kozikowski,*^,†Werner Tu¨ckmantel,^†Gesine Bo¨ttcher,^†,‡and Leo J . Romanczyk, J r.^§

Georgetown University Medical Center, Department of Neurology, Drug Discovery Laboratory and

Lombardi Cancer Center, 3900 Reservoir Road NW, Washington, D.C. 20007, and Mars, Incorporated,

High Street, Hackettstown, New J ersey 07840

kozikowa@georgetown.edu

Received J une 11, 2002

We report an improved synthesis of bis(5,7,3′,4′-tetra-O-benzyl)epicatechin 4â,8-dimer (3) from

5,7,3′,4′-tetra-O-benzylepicatechin (1) and 5,7,3′,4′-tetra-O-benzyl-4-(2-hydroxyethoxy)epicatechin

(2) by replacing the previously employed Lewis acid, titanium tetrachloride, with the clay mineral

Bentonite K-10. Under the same conditions, the benzyl-protected all-4â,8-trimer, -tetramer, and

-pentamer were obtained regioselectively from their lower homologues, albeit in rapidly decreasing

yields. Reaction of 2 with an organoaluminum thiolate generated from 2-mercaptobenzothiazole

and trimethylaluminum followed by acetylation produced 3-O-acetyl-4-[(2-benzothiazolyl)thio]-

5,7,3′,4′-tetra-O-benzylepicatechin (12). Medium-sized protected oligomers with 4â,8-interflavan

linkages are obtained in improved yields by using this compound as the electrophile and silver

tetrafluoroborate as activator and are isolated by reversed-phase HPLC. Their deprotection by ester

saponification followed by hydrogenolysis yielded the free procyanidins, which were characterized

as their peracetates. The synthetic procyanidins are identical by normal-phase HPLC with fractions

isolated from cocoa. The principle of chain extension by two members was demonstrated using a

dimeric electrophile obtained by self-condensation of compound 12. Both the synthetic and natural

pentamer 32 inhibit the growth of several breast cancer cell lines. Using the MDA MB 231 line, it

was established that this outcome is based on the induction of cell cycle arrest in the G₀/G₁phase.

Subsequent cell death is more likely necrotic rather than apoptotic. Control experiments demonstrate

that the polyphenol itself, rather than hydrogen peroxide potentially formed by its autoxidation, is

the causative agent.

In tr od u ction

tetra-O-benzyl-4-(2-hydroxyethoxy)epicatechin (2) (Chart

1, eq 1),¹which leads to the 4â,8-dimer 3 together with

higher oligomers in yields that decrease with their

molecular mass. The investigation of these byproducts

was deferred in favor of the more urgent task to un-

equivocally establish the stereochemical nature of the

interflavan bond in the dimer 3, a problem that has

eluded chemists relying solely on instrumental analytical

Condensed tannins (proanthocyanidins) are wide-

spread in the plant kingdom, form part of the human diet,

and display multiple biological activities that render

them significant to health.^1a,2,3We have undertaken a

program aimed at synthesizing condensed tannins previ-

ously isolated from cocoa⁴and reported on the condensa-

tion of 5,7,3′,4′-tetra-O-benzylepicatechin (1) with 5,7,3′,4′-

(2) Recent reviews and monographs: (a) Polyphenols, Wine and

Health. In Proceedings of the Phytochemical Society of Europe, Bor-

deaux, France, 14-16 April 1999; Che`ze, C., Vercauteren, J ., Verpoorte,

R., Eds; Kluwer Academic Publishers: Dordrecht, 2001. (b) Plant

Polyphenols 2. Chemistry, Biology, Pharmacology, Ecology; Gross, G.

G., Hemingway, R. W., Yoshida, T., Eds; Kluwer Academic/Plenum

Publishers: New York, 2000. (c) Merken, H. M.; Beecher, G. R. J . Agric.

Food Chem. 2000, 48, 577. (d) Harborne, J . B.; Williams, C. A.

Phytochemistry 2000, 55, 481. (e) Ferreira, D.; Li, X.-C. Nat. Prod. Rep.

2000, 193.

^†Georgetown University Medical Center.

^‡Current address: ABX Advanced Biochemical Compounds, Wil-

helm-Roensch-Strasse 9, D-01454 Radeberg, Germany.

^§Mars, Inc.

(1) (a) Part 1: Tu¨ckmantel, W.; Kozikowski, A. P.; Romanczyk, L.

L., J r. J . Am. Chem. Soc. 1999, 121, 12073. See here for a general

introduction of plant polyphenols, especially condensed tannins. (b)

Part 3: Kozikowski, A. P.; Tu¨ckmantel, W.; Hu, Y. J . Org. Chem. 2001,

66, 1287.

Published on Web 10/16/2002

J . Org. Chem. 2003, 68, 1641-1658

1641

Kozikowski et al.

CHART 1

methods and that we recently solved through synthesis

of a specifically protected derivative and its subsequent

degradation.⁵After this issue had been settled, we turned

our attention to the higher epicatechin oligomers. The

isolation and characterization of these compounds, to-

gether with improved methods for their synthesis and a

brief investigation of their anticancer activity, are the

subject of the present paper.

Resu lts

TiCl₄-Med ia ted Ch a in Exten sion . The predominant

larger oligomers isolated from the reported^1acondensa-

tion of compounds 1 and 2 in the presence of TiCl₄have

been identified as the 4â,8:4â,8-trimer 4, the 4â,6:4â,8-

trimer 5, and the all-4â,8-tetramer 6 (Chart 1, eq 1).

Trace amounts of even larger oligomers and possibly

additional regioisomers are present in the crude product

mixture but were not further investigated. The same

products were obtained starting from the epicatechin

dimer 3 rather than the monomer 1. The separation of

the product mixture was initially performed by normal-

phase HPLC on silica gel with ethyl acetate/hexane

mixtures as eluent, but we later found that much sharper

(3) See the following selected recent references. Lowering of plasma

cholesterol levels: (a) Bursill, C.; Roach, P. D.; Bottema, C. D. K.; Pal,

S. J . Agric. Food Chem. 2001, 49, 5639. Inhibition of squalene

epoxidase: (b) Abe, I.; Seki, T.; Umehara, K.; Miyase, T.; Noguchi, H.;

Sakakibara, J .; Ono, T. Biochem. Biophys. Res. Commun. 2000, 268,

767. Inhibition of endothelin-1 synthesis: (c) Corder, R.; Douthwaite,

J . A.; Lees, D. M.; Khan, N. Q.; Viseu dos Santos, A. C.; Wood, E. G.;

Carrier, M. J . Nature 2001, 414, 863. Antiallergic activity: (d)

Fujimura, Y.; Tachibana, H.; Yamada, K. J . Agric. Food Chem. 2001,

49, 2527. Neuroprotection: (e) Levites, Y.; Weinreb, O.; Maor, G.;

Youdim, M. B.; Mandel, S. J . Neurochem. 2001, 78, 1073. (f) Shimada,

Y.; Goto, H.; Shibahara, N.; Sakakibara, I.; Sasaki, H.; Terasawa, K.

Am. J . Chin. Med. 2001, 29, 173. Anti-inflammatory activity: (g) Surh,

Y.; Chun, K.; Han, S. S.; Keum, Y.; Park, K.; Lee, S. S. Mutation Res.

2001, 480-481, 243. Inhibition of metabolic activation of procarcino-

gens: (h) Muto, S.; Fujita, K.; Yamazaki, Y.; Kamataki, T. Mutation

Res. 2001, 479, 197. Induction of apoptosis: (i) Saeki, K.; Hayakawa,

S.; Isemura, M.; Miyase, T. Phytochemistry 2000, 53, 391. (j) Smith,

D. M.; Dou, Q. P. Int. J . Mol. Med. 2001, 7, 645. (k) Hayakawa, S.;

Saeki, K.; Sazuka, M.; Suzuki, Y.; Shoji, Y.; Ohta, T.; Kaji, K.; Yuo,

A.; Isemura, M. Biochem. Biophys. Res. Commun. 2001, 285, 1102.

Reversion of multidrug resistance: (l) Zhu, A.; Wang, X.; Guo, Z. Nucl.

Med. Biol. 2001, 28, 735. Modulation of gene expression: (m) Okabe,

S.; Fujimoto, N.; Sueoka, N.; Suganuma, M.; Fujiki, H. Biol. Pharm.

Bull. 2001, 24, 883. Inhibition of phenol sulfotransferase: (n) Isozaki,

T.; Tamura, H. Biol. Pharm. Bull. 2001, 24, 1076. Modification of

arachidonic acid metabolism: (o) Hong, J .; Smith, T. J .; Ho, C.-T.;

August, D. A.; Yang, C. S. Biochem. Pharmacol. 2001, 62, 1175.

Inhibition of the Ras-MAP kinase signaling pathway: (p) Chung, J .

Y.; Park, J . O.; Phyu, H.; Dong, Z.; Yang, C. S. FASEB J . 2001, 15,

2022. Review of mechanisms of activity: (q) Nijveldt, R. J .; van Nood,

E.; van Hoorn, D. E. C.; Boelens, P. G.; van Norren, K.; Leeuwen, P.

A. M. Am. J . Clin. Nutr. 2001, 74, 418.

(4) (a) Forsyth, W. G. C.; Roberts, J . B. Biochem. J . 1960, 74, 374.

(b) Quesnel, V. C. Phytochemistry 1968, 7, 1583. (c) J alal, M. A. F.;

Collin, H. A. Phytochemistry 1977, 16, 1377. (d) Porter, L. J .; Ma, Z.;

Chan, B. G. Phytochemistry 1991, 30, 1657. (e) Rigaud, J .; Escribano-

Bailon, M. T.; Prieur, C.; Souquet, J .-M.; Cheynier, V. J . Chromatogr.

1993, 654, 255. (f) Romanczyk, L. J ., J r.; Hammerstone, J . F., J r.; Buck,

M. M. US Patent 5,554,645, Sep 10, 1996. (g) Romanczyk, L. J ., J r.;

Hammerstone, J . F., J r.; Buck, M. M.; Post, L. S.; Cipolla, G. G.;

McClelland, C. A.; Mundt, J . A.; Schmitz, H. H. PCT Int. Appl. WO

9736,597, Oct 9, 1997; Chem. Abstr. 1997, 127, 322787. (h) Romanczyk,

L. J ., J r.; Hammerstone, J . F., J r.; Buck, M. M.; Post, L. S.; Cipolla, G.

G.; McClelland, C. A.; Mundt, J . A.; Schmitz, H. H. PCT Int. Appl.

WO 9736,497, Oct. 9, 1997; Chem. Abstr. 1997, 127, 330680. (i)

Romanczyk, L. J ., J r.; Hammerstone, J . F., J r.; Buck, M. M. US Patent

5,712,305, J an 27, 1998. (j) Lee, M. C.; Chang, Y. Y.; Yim, M. H.

Han’guk Nonghwa Hakhoechi 1998, 41, 110-7; Chem. Abstr. 1998,

129, 207072. (k) Sanbongi, C.; Osakabe, N.; Natsume, M.; Takizawa,

T.; Gomi, S.; Osawa, T. J . Agric. Food Chem. 1998, 46, 454. (l)

Hammerstone, J . F.; Lazarus, S. A.; Mitchell, A. E.; Rucker, R.;

Schmitz, H. H. J . Agric. Food Chem. 1999, 47, 490. (m) De Bruyne,

T.; Pieters, L.; Witvrouw, M.; De Clercq, E.; Vanden Berghe, D.;

Vlietinck, A. J . J . Nat. Prod. 1999, 62, 954. (n) Wollgast, J .; Pallaroni,

L.; Agazzi, M.-E.; Anklam, E. J . Chromatogr. A 2001, 926, 211.

(5) Kozikowski, A. P.; Tu¨ckmantel, W.; George, C. J . Org. Chem.

2000, 65, 5371-5381.

1642 J . Org. Chem., Vol. 68, No. 5, 2003

Studies in Polyphenol Chemistry and Bioactivity

peaks are obtained on C₈or C₁₈columns with an

acetonitrile-water gradient. On silica gel, retention

increases with molecular size, presumably because of the

increasing number of alcohol and ether functions. Para-

doxically, the same elution sequence is observed on

reverse phases. It appears that these adsorbents dis-

criminate predominantly by the number of the nonpolar

benzyl groups, which also increases with the degree of

oligomerization. The couple of regioisomers 4 and 5,

however, behaves as expected: whereas compound 4

elutes before 5 on silica gel, the opposite sequence is

observed on reverse phases.

Compounds 4 and 6 are identical with products ob-

tained by alternative protocols and will be discussed

below. Compound 5 was additionally transformed into

its triacetate 7, a type of derivative that was also

routinely employed in the preparation of the all-4â,8

oligomers. Since compound 5 is obtained by way of the

dimer 3, it ought to contain the 4â,8-interflavan linkage

present in that material. The additional interflavan

linkage can then originate from any of the three unsub-

stituted positions of its A rings. With structure 4 already

being assigned to another product, only 5 and the

“branched” trimer (containing simultaneously a 4,6- and

a 4,8-interflavan bond to the “bottom” A ring) remain.

As in the case of other benzylated intermediates, com-

pound 5 was deprotected hydrogenolytically, and the

resulting free polyphenol 8 was acetylated with an excess

of acetic anhydride and pyridine to yield the acetate 9

(Scheme 1). The ¹H NMR spectrum of compound 9

significant nucleophilic reactivity in the present chain

extension reaction, whereas we have never found the 4,6-

dimer in reactions starting from the monomer 1.

The formation of regioisomers and possibly other types

of byproducts constitutes a serious drawback of the TiCl₄-

mediated chain extension reaction, not only in terms of

yield. Even though the all-4â,8-trimer and -tetramer were

isolated in pure form, the same could not automatically

be expected for larger oligomers, for which the number

of possible isomers, and thus contaminants, grows rap-

idly. One potential way of dealing with this problem

would be to carefully purify the chain-extended interme-

diate after each step to ensure that all products are at

least derived from a single isomer of the starting mate-

rial. Upon reaction of the electrophile 2 with 2 equiv of

the trimer 4 (2 equiv of TiCl₄, CH₂Cl₂/THF 9:11, 0 °C, 15

min, then rt, 140 min), we observed, however, not only

the formation of the tetramer, pentamer, and small

amounts of higher oligomers, but also a degradation of

the trimer to the monomer and dimer, which then

returned into the chain-extension reaction, giving rise to

regioisomeric oligomers such as small amounts of the

4â,6:4â,8-trimer 5. While the reaction conditions were not

optimized, we felt that the occurrence of chain degrada-

tion was of such fundamental importance that a search

for a different approach was warranted. Chain degrada-

tion of nonprotected polymeric proanthocyanidins in the

presence of protic acids has previously been observed by

others, and the resulting electrophilic species have to

some extent been utilized in the construction of oligo-

mers.⁷

SCHEME 1

Ch a in Exten sion Med ia ted by Ben ton ite K-10. A

serendipitous discovery led the way. Stirring of the

starting materials 1 (4 equiv) and 2 in CH₂Cl₂with the

commercial acidic clay mineral, Bentonite K-10, resulted

in the almost exclusive formation of the 4â,8-dimer 3 in

a 90% isolated yield together with small amounts of the

4â,8:4â,8-trimer 4; no 4,6-linked material was observed.

The surprisingly high reactivity differential under these

conditions between monomer 1 and dimer 3 that allows

most of the dimer to survive without entering into further

chain extension means, of course, that application of the

same protocol to the extension of the dimer to the trimer

could be expected to proceed less efficiently. Indeed,

reaction of the dimer 3 (3 equiv) with the electrophile 2

yielded only 40% of the 4â,8:4â,8-trimer 4 together with

13% of the all-4â,8-tetramer 6. The poorer ratio of singly

to doubly chain-extended product certainly in part re-

flects the lower excess of nucleophilic component used

here (a consequence of its limited availability) in com-

parison with the same reaction involving reactants 1 and

2, but probably also a lower reactivity differential be-

tween trimer and dimer than between monomer and

dimer. Still, the reactivity of the trimer and tetramer is

so much reduced that their chain extension by this

protocol, while yielding small amounts of the desired

tetramer 6 and pentamer, is impractical. From an

operational point of view, it is also worthwhile to note

that the cleanness of the reaction permitted for the first

time to achieve at least a partial separation of the

demonstrated the presence of a major component exhibit-

ing sharp signals and of one or more minor components

exhibiting broad signals. The major component proved

identical to the peracetate prepared from natural epi-

catechin 4â,6:4â,8-trimer.⁶While it is commonly observed

that NMR spectra of proanthocyanidins and their deriva-

tives are complicated by the presence of several rotamers

that interconvert slowly on the NMR time scale, ref 6

explicitly states that the compound consists of a single

rotamer. Regrettably, therefore, we must conclude that

our preparation contains unknown contaminants; a result

the more surprising since all three compounds 5, 7, and

9 appear homogeneous by HPLC. It is interesting to note

that the position 6 of the top ring of the dimer 3 displays

(7) (a) Fletcher, A.; Porter, L. J .; Haslam, E.; Gupta, R. K. J . Chem.

Soc., Perkin Trans. 1 1977, 1628. (b) Botha, J . J .; Ferreira, D.; Roux,

D. G. J . Chem. Soc., Perkin Trans. 1 1981, 1235. (c) Foo, L. Y.; Porter,

L. J . J . Chem. Soc., Perkin Trans. 1 1983, 1535.

(6) Ho¨r, M.; Heinrich, M.; Rimpler, H. Phytochemistry 1996, 42, 109.

J . Org. Chem, Vol. 68, No. 5, 2003 1643

Kozikowski et al.

SCHEME 2

monomer from the dimer and even of the dimer from the

trimer by column chromatography, thus significantly

reducing the amount of material that needs to be put

through HPLC purification. Together with a simplified

isolation procedure for the electrophilic building block 2,⁸

it is now possible using this protocol to obtain several

grams of the intermediates 3 and 4 with a reasonable

amount of effort.

Ch a in Exten sion Em p loyin g th e 4-[(2-Ben zoth ia -

zolyl)th io] Der iva tive. Since oligomers larger than the

tetramer remained poorly or not accessible, we remained

interested in alternative chain extension protocols. In this

context, the activation of 4-(benzylthio)catechin and

-epicatechin by dimethyl(methylthio)sulfonium tetrafluo-

roborate or, preferably, by silver tetrafluoroborate came

to our attention (eq 2).⁹The reaction has considerable

(10a ,b ) which were isolated in 67 and 2.5% yield,

respectively, by fractional crystallization and column

chromatography (Scheme 2). Only the major isomer was

used for subsequent reactions. To our delight, addition

of AgBF₄to a solution of 1 and 10a in THF resulted in

the formation of the 4â,8-dimer 3 (56%) and the (4â,8)₂-

trimer 4 (14%) together with recovered monomer 1 after

normal-phase HPLC separation. The recovered monomer

proved impure; further separation by reversed-phase

HPLC yielded 5% of a material to which we assign the

similarity with the activation of another type of reactive

sulfides, namely R-alkoxyalkyl sulfides (especially thiogly-

cosides), by various classes of electrophiles.¹⁰The par-

ticular virtue of this protocol resides in the exclusive

formation of 4,8-interflavan linkages even for nonpro-

tected substrates, which would otherwise yield mixtures

of 4,6- and 4,8-linked products. On the other hand, the

noxious nature of benzyl mercaptan involved in the

preparation of the starting material was considered a

serious drawback by us. We therefore decided from the

beginning to replace this reagent with a nonvolatile,

odorless heterocyclic thiol and found that 2-mercapto-

benzothiazole is suitable for this purpose, although

during one particular transformation a side reaction

caused by the nucleophilicity of the nitrogen atom was

encountered (see below). Furthermore, we intended to

perform this chemistry on benzyl-protected intermediates

as we had done before, both because of the easier

handling and better stability of these compounds and

because of the poor accessibility of nonprotected, 4-sub-

stituted epicatechins. The 4-(2-hydroxyethoxy) derivative

2 was selected as the starting material. Its reaction with

an organoaluminum thiolate prepared in situ from

2-mercaptobenzothiazole and trimethylaluminum¹¹fur-

nished the 4-thioether as a mixture of two stereoisomers

1

structure of a 3-O-4-dimer 11 on the basis of its H NMR

spectrum, which displayed two AB quartets for two pairs

of A-ring protons together with a single OH proton and

a single 4-methylene group. The formation of similar

dimers mediated by AgBF₄has been reported for non-

protected flavan-3,4-diols containing a 7,8-dioxygenated

A ring,¹²which is less nucleophilic than the present 6,8-

dioxygenated one, and related compounds have also been

isolated from natural sources.¹³

To avoid the undesired intervention of the 3-hydroxyl

group in the chain elongation process, this group was

protected in both the electrophilic and the nucleophilic

reaction partner (in this case, the dimer 3) by acetylation

(Scheme 3). The yields were near-quantitative. When a

solution of AgBF₄in THF was added to a THF solution

of the resulting acetates 12 and 13 (molar ratio 1:3.6) at

0 °C, the expected trimer 16 and tetramer 17 were

formed, but only in 34% and 4% yield, respectively. Chain

elongation proceeds so slowly in this case that adventi-

tious water successfully competes with the flavanoid

(11) The replacement of an allylic methoxyl group by phenylthio with

a reagent derived from thiophenol and diisobutylaluminum chloride

has been reported: Dzhemilev, U. M.; Ibragimov, A. G.; Morozov, A.

B.; Tolstikov, G. A. Izv. Akad. Nauk SSSR, Ser. Khim. 1988, 2645;

Chem. Abstr. 1989, 110, 153414c.

(12) Bennie, L.; Malan, E.; Coetzee, J .; Ferreira, D. Phytochemistry

2000, 53, 785.

(8) See the Supporting Information

(9) Steynberg, P. J .; Nel, R. J . J .; van Rensburg, H.; Bezuidenhoudt,

B. C. B.; Ferreira, D. Tetrahedron 1998, 54, 8153.

(10) (a) Toshima, K.; Tatsuta, K. Chem. Rev. 1993, 93, 1503. (b)

Garegg, P. J . Adv. Carbohydr. Chem. Biochem. 1997, 52, 179. (c)

Marcune, B. F.; Karady, S.; Dolling, U.-H.; Novak, T. J . J . Org. Chem.

1999, 64, 2446.

(13) Bennie, L.; Coetzee, J .; Malan, E.; Ferreira, D. Phytochemistry

2001, 57, 1023 and references quoted therein.

1644 J . Org. Chem., Vol. 68, No. 5, 2003

Studies in Polyphenol Chemistry and Bioactivity

SCHEME 3

SCHEME 4

21 were isolated in a combined yield of 91%. The reaction

is exceptionally clean; a chromatogram of the reaction

mixture is shown in Figure 1. No 4,6-linked products

were found. Similar results were obtained in the reac-

tions of 12 with the trimer 16 and with the tetramer 17.

From the latter reaction, oligomers up to the undecamer

24 could be isolated by reversed-phase HPLC if ethyl

acetate was admixed to the acetonitrile in the final step

of the gradient. This nonpolar solvent permitted recovery

of the highly retained larger oligomers but also eluted

significant amounts of aliphatic impurities which sub-

sequently had to be removed by additional HPLC steps,

thus reducing the total product recovery.¹⁵

All of the benzyl ether-acetates up to the nonamer

(compounds 16-22) were deacetylated in near-quantita-

tive yield with 40% aqueous tetra-n-butylammonium

hydroxide in THF (Scheme 4). This base was chosen

because of its good solubility in the relatively nonpolar

reaction medium that is required by the lipophilicity of

the starting materials. The ¹H NMR spectra of the

resulting benzyl ethers 4, 6, and 25-29 display signals

of two major rotamers together with trace amounts of

additional rotamers that increase as the oligomer chain

grows. We are confident that these minor components

are rotamers rather than regioisomers because similar

signals are absent from the spectra of the precursor

acetates. Samples of the benzyl ethers prepared in CDCl₃

exhibit well-resolved, characteristic signals for the OH

protons in the δ 1.8-1.1 region.

nucleophile, for the major product of the reaction (44%

yield) was the 4-hydroxy monomer 14. In addition, 4%

of the 4-hydroxy dimer 15 was also isolated, pointing to

self-condensation of the thioether 12 followed by either

chain elongation (to yield 17) or hydrolysis. An attempt

to improve yields by stirring the reactants with powdered

molecular sieves prior to the addition of AgBF₄was

ineffective. Since the major source of moisture in this

reaction is probably the AgBF₄, a very hygroscopic

reagent, it appeared logical to dry it and the dimer with

molecular sieves first and to add compound 12 subse-

quently. From this experiment, however, both organic

reactants were recovered unchanged! It thus appears that

the silver salt is adsorbed on the molecular sieves and

in this form is no longer capable of reacting with the

thioether 12. This result remains puzzling. Inclusion of

molecular sieves in the reaction mixture is a common

practice in glycoside synthesis, and it has actually been

reported that AgNO₃adsorbed on MS 4 Å is an activator

of glycosyl halides.¹⁴Fortunately, simple vacuum-drying

of the AgBF₄immediately before the reaction was sub-

sequently found to reduce the hydrolysis of 12 to an

acceptable level, at least for gram-scale runs. Under these

conditions, and with a 12/13 molar ratio of 1:2.5, a series

of oligomers spanning from the trimer 16 to the octamer

(15) In all of these reactions, the yields of oligomers decrease with

the degree of oligomerization, as should be expected. However, this

decrease is not uniform. Especially the dimer 20 reproducibly furnishes

a percentage yield of the tetramer 22 which comes fairly close to that

of the trimer 21. A weaker effect of the same kind is found for the

tetramer as the starting material, but not for the trimer. There does

not appear to be a simple explanation for these discrepancies.

(14) Garegg, P. J .; Ossowski, P. Acta Chem. Scand. B 1983, 37, 249.

J . Org. Chem, Vol. 68, No. 5, 2003 1645

Kozikowski et al.

F IGURE 1. AgBF₄-mediated reaction between compounds 12 and 13: chromatogram of the reaction mixture after filtration

over silica gel. Conditions: Hewlett-Packard RP-8, 200 × 4.6 mm, flow rate 1.0 mL/min, UV detection at 280 nm; 0-30 min, 80

to 100% CH₃CN in H₂O, then CH₃CN.

Dep r otection a n d Ch a r a cter iza tion . The benzyl

ethers 4, 6, and 25-28 (trimer through octamer) were

deprotected by hydrogenolysis over Pearlman’s catalyst

to yield the free polyphenols 30-35. We found it advis-

able to perform this reaction in bicarbonate-washed

glassware, as partial fragmentation to lower oligomers

was occasionally observed without this precaution, quite

probably as a consequence of an acidic reaction of the

glass surface of the reaction flask.¹⁶To obtain a readily

soluble polyphenol, it was furthermore found necessary,

as similarly reported by others,¹⁷to dilute the filtered

solution of the crude product with water (in place of

dioxane used in ref 17), to evaporate only partially so as

to remove most of the organic solvents, and to lyophilize

the residual solution. If the crude polyphenol solutions

are directly evaporated to dryness, partially insoluble

materials result, indicating that some decomposition has

occurred. Combustion analyses show that the lyophilized

products contain 1.3-2 equiv of water per epicatechin

moiety.

The synthetic trimer 30, tetramer 31, and pentamer

32 were compared against the natural products purified

(16) An alternative explanation, namely the actual hydrogenolysis

of interflavan bonds, appears unlikely since several authors have

reported that this reaction, when performed intentionally, requires

elevated temperatures and hydrogen pressures: (a) Nisi, D.; Panizzi,

L. Gazz. Chim. Ital. 1966, 96, 803. (b) J acques, D.; Haslam, E.; Bedford,

G. R.; Greatbanks, D. J . Chem. Soc., Perkin Trans. 1 1974, 2663. (c)

Foo, L. Y. Phytochemistry 1982, 21, 1741. It is, to the contrary, quite

conceivable that these reported reactions proceed via initial solvolytic

cleavage of the interflavan bond followed by hydrogenolytic removal

of the resulting 4-substituent. Acidic glass surfaces may also be

involved in reported cases of partial fragmentation of benzylated

dimeric procyanidins: (d) J acques, D.; Haslam, E.; Bedford, G. R.;

Greatbanks, D. J . Chem. Soc., Perkin Trans.

1 1974, 2663. (e)

Arnaudinaud, V.; Nay, B.; Nuhrich, A.; Deffieux, G.; Me´rillon, J .-M.;

Monti, J .-P.; Vercauteren, J . Tetrahedron Lett. 2001, 42, 1279.

(17) Yoneda, S.; Kawamoto, H.; Nakatsubo, F. J . Chem. Soc., Perkin

Trans. 1 1997, 1025.

1646 J . Org. Chem., Vol. 68, No. 5, 2003

Studies in Polyphenol Chemistry and Bioactivity

F IGURE 2. Comparison HPLC traces of purified natural (blue) and synthetic (red) procyanidin oligomers: A, 30; B, 31; C, 32.

For conditions, see the Supporting Information.

from Theobroma cacao⁴by normal-phase^4e,lHPLC analy-

sis (Figure 2). Purities ranging from 94 to 96% were

observed for the synthetic procyanidins, which were

2-4% higher than those obtained for the natural pro-

cyanidins. The t_R’s of the synthetic procyanidins matched

those observed for the natural products, thus confirming

epicatechin 4â,8 regio- and stereochemistry in cocoa. All

of the natural procyanidins purified from cocoa showed

impurity peaks preceding and following the main peak.

Scanning these regions by HPLC/MS revealed no change

in the [M]⁺or [M + Na]⁺ions, indicating that these minor

impurities are isomers of the major oligomer. These

minor impurities may contribute to in vitro and in vivo

activities reported in the literature and potentially

confound structure-activity relationships based on natu-

ral material only. As a precaution, both natural and

synthetic samples of procyanidins were therefore used

in the biological assays.

Since free polyphenols are inherently poorly amenable

to purification because of their oxygen and acid sensitiv-

ity (acid being required as a solvent additive to reduce

peak tailing during HPLC), and their NMR spectra are

anyway uncharacteristic because of severe line broaden-

ing, we preferred to characterize these compounds as

their peracetates 36-41. In an attempt to avoid acid-

induced interflavan bond cleavage during this reaction,

we increased the amount of pyridine in the Ac₂O/pyridine

reagent from one (the usual amount) to two volumes

relative to Ac₂O, but as a result, impurities emerged that

eluted immediately before the peracetates and could not

1

be removed on a preparative scale. The H NMR spectra

of the peracetates exhibit sharp signals for two rotamers

J . Org. Chem, Vol. 68, No. 5, 2003 1647

Kozikowski et al.

(in a 2:1 ratio for the trimer 14 and in a 3:2 ratio for all

higher homologues) and are, up to the heptamer or

octamer, quite suitable for compound identification.

Especially the acetate region serves as a useful “finger-

print”. As the oligomer chain grows, the chemical shift

differences between analogous protons of epicatechin

units in the inner positions of the chain become eventu-

ally insufficient at 300 MHz, resulting in the growth of

uncharacteristic signal clusters without the appearance

of well-separated new signals. We believe that these

spectra can be useful for future reference, and we have

included suitable sections thereof in the Supporting

Information. ¹³C NMR spectra have been acquired for

oligomers up to the hexamer, beyond which insufficient

amounts of material were available, and have also been

included in the Supporting Information. Regrettably,

some of the literature work is limited to partial thiolytic

degradation as a tool of structural assignment. While this

technique is sound¹⁸as long as the structure of the

smaller fragments can be established with certainty, it

fails to provide reference data for future workers who

reisolate or synthesize the compound in question. Only

of two epicatechin units rather than just one. It was

therefore of interest to search for a synthetic approach

to a protected 4â,8-dimer bearing a (2-benzothiazolyl)-

thio substituent in the 4-position of its bottom epicatechin

unit. The direct oxidation of a methyl/acetyl-protected

dimeric procyanidin with DDQ occurs in the doubly

activated 4-position of its top flavanol moiety.²⁰A small

amount of the 4-hydroxylated dimer 15 was available as

a byproduct, but extension of the organoaluminum thi-

olate protocol to the free triol 42 obtained by acetate

saponification yielded but traces of the thioether 43

(Scheme 5). As an alternative, we envisaged a cross-

coupling between the 4-(2-hydroxyethoxy)monomer 2 and

the 4-[(2-benzothiazolyl)thio]monomer 10a , either under

acid catalysis for activation of the reactant 2 or under

the action of AgBF₄for activation of 10a . These attempts

largely met with failure. However, we found that com-

pound 10a self-condenses under the action of AgBF₄or,

surprisingly, Bentonite K-10 to yield fairly complex

mixtures from which small amounts of the 4-[(2-ben-

zothiazolyl)thio]-substituted dimer 43, trimer 44, and

presumably the tetramer 45 were isolated together with

the rearranged monomer 46 and rearranged dimer 47.

The migration of the flavanoid moiety from sulfur to

nitrogen was confirmed for compound 46 by the observa-

tion of a ¹³C NMR signal at δ 190.3 assignable to the

thiocarbonyl carbon atom. The related silver ion-induced

migration of a reactive alkyl group from sulfur to nitrogen

has been documented for the case of 2-[(methoxymethyl)-

thio]pyridine.^10cWe suspected that the complexity of the

above reaction mixtures was in part due to the formation

of 4-O-3-linked oligomers similar to compound 11, and

applied the same remedy used in that case, namely

acetylation. AgBF₄-induced self-condensation of com-

pound 12 resulted in low yields of the 4-[(2-benzothiaz-

olyl)thio]-substituted oligomers 48 (dimer, 14% yield)

through 50 (tetramer) (eq 3). Together with these prod-

ucts, and in considerable quantities because of the small

reaction scale, the 4-hydroxylated oligomers 14, 15, 51

1

for the trimer 14^6,19fand the tetramer 15⁶have H NMR

spectra been published, and these are in good agreement

with ours. In the case of the tetramer 12, partial

thiolysis⁶has established 4â,8-regio- and -stereochemis-

try for the “upper” interflavan linkage whereas the

“lower” portion of the molecule is identical with the

trimer 11. This compound, in turn, has also been sub-

jected to partial thiolysis, and both interflavan linkages

have been identified as 4â,8.⁶Since, therefore, in the

course of the present chain extension process the first

three interflavan linkages formed are exclusively of the

4â,8-type, we conclude that the same must be true for

the additional interflavan linkages present in the larger

oligomers. Compounds 30-33 are natural products wide-

spread in various groups of vascular plants.¹⁹The

reduced frequency with which the larger oligomers have

been reported in the literature probably reflects, at least

in part, their poor extraction and the difficulty of their

purification, rather than their uncommon occurrence.

(19) Isolation from (a) cocoa (30-33): see ref 4. (b) Pinus taeda

(loblolly pine) phloem (30): Hemingway, R. W.; Foo, L. Y.; Porter, L.

J . J . Chem. Soc., Perkin Trans. 1 1982, 1209. (c) Cinchona succirubra

bark (30): Nonaka, G.; Kawahara, O.; Nishioka, I. Chem. Pharm. Bull.

1982, 30, 4277. (d), (e) Crataegus (hawthorn) fruit (30, 31): Kolodziej,

H.; Ferreira, D.; Roux, D. G. J . Chem. Soc., Perkin Trans. 1 1984, 343.

Rohr, G. E.; Meier, B.; Sticher, O. J . Chromatogr. A 1999, 835, 59. (f)

Nelia meyeri leaves (30): Kolodziej, H. Phytochemistry 1984, 23, 1745.

(g) Kandelia candel bark (30): Hsu, F.-L.; Nonaka, G.; Nishioka, I.

Chem. Pharm. Bull. 1985, 33, 3142. (h) Dioscorea cirrhosa (yam) tubers

(30, 31): Hsu, F.-L.; Nonaka, G.; Nishioka, I. Chem. Pharm. Bull. 1985,

33, 3293. (i) Cinnamomum cassia bark (30-33): Morimoto, S.; Nonaka,

G.; Nishioka, I. Chem. Pharm. Bull. 1986, 34, 633. (j) Rhaphiolepis

umbellata bark (30-32): Ezaki-Furuichi, E.; Nonaka, G.; Nishioka,

I.; Hayashi, K. Agric. Biol. Chem. 1986, 50, 2061. (k), (l) Pseudotsuga

menziesii (Douglas fir) bark (30-32): Foo, L. Y.; Karchesy, J . J .

Phytochemistry 1989, 28, 1743. Foo, L. Y.; Karchesy, J . J . Phytochem-

istry 1991, 30, 667. (m), (n) Grape seeds (30, 31): Ricardo da Silva, J .

M.; Rigaud, J .; Cheynier, V.; Cheminat, A.; Moutounet, M. Phytochem-

istry 1991, 30, 1259. Santos-Buelga, C.; Francia-Aricha, E. M.; Es-

cribano-Bailo´n, M. T. Food Chem. 1995, 53, 197. (o), (p), (q) Apples

(30): Pe´rez-Ilzarbe, F. J .; Mart´ınez, V.; Herna´ndez, T.; Estrella, I. J .

Liquid Chromatogr. 1992, 15, 637. Kanda, T.; Shoji, T.; Ohnishi-

Kameyama, M.; Nagata, T. J . Chromatogr. A 1999, 855, 181. Gomis,

D. B.; Palomino, N. F.; Alonso, J . J . M. Anal. Chim. Acta 2001, 426,

111. (r) Guazuma ulmifolia bark (30, 31): ref 6. (s) almond fruit flesh

(30, 31): de Pascual-Teresa, S.; Gutie´rrez-Fernandez, Y.; Rivas-

Gonzalo, J . C.; Santos-Buelga, C. Phytochem. Anal. 1998, 9, 21. (t)

Hypericum perforatum herb (30): Ploss, O.; Petereit, F.; Nahrstedt,

A. Pharmazie 2001, 56, 509.

Ch a in Exten sion by Tw o Mem ber s. The HPLC

separation of oligomer mixtures would be much improved

if a given oligomer could be homologated in increments

(18) Side reactions occurring during exhaustive thiolysis (McGraw,

G. W.; Steynberg, J . P.; Hemingway, R. W. Tetrahedron Lett. 1992,

34, 987) have led to the questioning of all structures established by

thiolytic methods (Balas, L.; Vercauteren, J . Magn. Reson. Chem. 1994,

32, 386). The latter authors go far beyond the conclusions of the former,

who solely reject the quantitative interpretation of exhaustive thiolysis

experiments. The “sequencing” of condensed tannins takes place under

the milder conditions of partial thiolysis. There is at present no

indication that the reported side reaction occurs to any significant

extent under these conditions. Furthermore, even if this reaction takes

place and the sequence of events postulated by McGraw et al. is entirely

reversible, and therefore re-addition of the substituted phloroglucinol

fragment could initiate a return of the reactants to either a 4,6- or

4,8-linked 4-(benzylthio)oligomer, steric factors would favor the former

over the latter. The finding of 4,8-linkage in the case of compounds 11

and 12 (ref 6) leaves thus no doubt that it is genuine. Furthermore, as

no 4,6-linked fragments are formed from these compounds, the finding

of 4,6-linkage in other cases should be considered genuine as well.

Finally, if re-addition of the substituted phloroglucinol fragment did

actually occur, so would the addition of any nonhindered phenolic

hydroxyl group of other components present in the reaction mixture,

resulting in

a complex mixture of products rather than just the

isomerized 4-(benzylthio)oligomer. It is, of course, assumed that all

major fragments, rather than just a few arbitrarily selected components

of complex mixtures, have been identified in these experiments.

(20) Steenkamp, J . A.; Mouton, C. H. L.; Ferreira, D. Tetrahedron

1991, 47, 6705.

1648 J . Org. Chem., Vol. 68, No. 5, 2003

Studies in Polyphenol Chemistry and Bioactivity

SCHEME 5

(trimer), and 52 (tetramer) were also formed. Reaction

of compound 48 with the tetramer 17 in the presence of

AgBF₄finally resulted in the formation of the expected

hexamer 19 in 12% yield together with the byproducts

21 (octamer), 15, and 52. We have thus demonstrated

the principle that chain elongation can be performed in

increments of two flavanol units. While the present yields

are impractical, they should improve considerably upon

scale-up because of the diminished impact of adventitious

moisture. The procedure can be of potential use for the

chain-elongation of the larger protected epicatechin oli-

gomers as these compounds far exceed the building block

12 in value.

nutrition, medicine, and health due to their wide range

of potentially significant biological activities.³There is a

growing body of evidence suggesting that these com-

pounds act as potent antioxidants in vitro, ex vivo, and

in vivo and may thus alter the pathophysiology of

imbalances or perturbations of free radical and/or oxi-

datively driven processes in many diseases or directly

interfere with many cellular processes.²¹Our interest is

focused in the cancer area where initial observations^4f-h

showed procyanidin-rich fractions from cocoa to elicit in

vitro growth inhibition in several human cancer cell lines.

The development of improved analytical and preparative

HPLC methods^4e,lhas allowed us to obtain quantities of

purified oligomers which were screened against certain

human breast cancer cell lines together with our syn-

thetic products (up to the pentamer). Cytotoxic effects

were clearly associated with high molecular weight

oligomers (pentamer and higher) at the 100 µg/mL dose

level (Figure 3). This effect was also observed in MCF-7,

MDA MB 231, and MDA 435 human breast cancer cell

lines treated in a similar fashion (data not shown).

Curiously, no activity was observed for the dimer, trimer,

and tetramer, which others have reported to show

activity on small cell lung and colorectal cancer cell

lines.²²Cytotoxicity thus increases sharply with the

degree of oligomerization.

Several studies have demonstrated an association

between cell cycle regulation and cancer.²³The cell cycle

machinery regulates cell proliferation, and dysregulated

(21) Nijveldt, R. J .; van Nood, E.; van Hoorn, D. E. C.; Boelens, P.

G.; van Norren, K.; van Leeuwen, P. A. M. Am. J . Clin. Nutr. 2001,

74, 418.

(22) (a) Kolodziej, H.; Haberland, C.; Woerdenbag, H. J .; Konings,

A. W. T. Phytotherapy Res. 1995, 9, 410. (b) Gali, H. U.; Perchellet, E.

M.; Gao, X. M.; Karchesy, J . J .; Perchellet, J . P. Planta Med. 1994, 60,

235.

(23) Hajduch, M.; Havlieek, L.; Vesely, J .; Novotny, R.; Mihal, V.;

Strand, M. Adv. Exp. Med. Biol. 1999, 457, 341.

An tica n cer Activity. The proanthocyanidins have

attracted a great deal of recent attention in the fields of

J . Org. Chem, Vol. 68, No. 5, 2003 1649

Kozikowski et al.

F IGURE 3. Cytotoxicity of procyanidin fractions purified from cocoa on human breast cancer cell line SKBR-3 at 100 and 25

µg/mL.

TABLE 1. Cell Cycle An a lysis of MDA MB 231 Hu m a n

Br ea st Ca n cer Cells Tr ea ted w ith Oligom er ic

P r ocya n id in s P u r ified fr om Cocoa

TABLE 2. Com p a r ison Cell Cycle An a lysis of MDA MB

231 Hu m a n Br ea st Ca n cer Cells Tr ea ted w ith Na tu r a l

ver su s Syn th etic Oligom er ic P r ocya n id in s

% G₀/G₁

% S

% G₂/M

% G₀/G₁

% S

% G₂/M

control

vehicle

dimer (200 µg/mL; 24 h)

control

vehicle

trimer (200 µg/mL; 24 h)

control

vehicle

tetramer (200 µg/mL; 24 h)

control

36.69

38.26

38.13

42.28

43.60

43.22

40.33

43.71

51.46

38.33

37.84

66.03

88.31

23.39

22.43

35.61

34.10

35.98

36.25

34.42

28.25

21.05

21.39

17.23

6.07

39.92

39.30

39.45

22.12

22.30

20.80

23.42

21.87

20.30

40.61

40.77

16.67

5.62

control

vehicle

natural trimer

28.65

27.19

28.46

49.28

49.61

48.49

22.06

23.2

23.05

(200 µg/mL; 24 h)

synthetic trimer (30)

(200 µg/mL; 24 h)

natural tetramer

(200 µg/mL; 24 h)

synthetic tetramer (31)

(200 µg/mL; 24 h)

natural pentamer

(200 µg/mL; 24 h)

synthetic pentamer (32)

(200 µg/mL; 24 h)

26.98

36.82

43.49

45.99

64.15

49.57

43.37

39.39

38.25

23.36

23.45

19.02

17.03

15.76

12.49

vehicle

pentamer (200 µg/mL; 24 h)

pentamer (200 µg/mL; 48 h)

cellular proliferation is a hallmark of cancer.²⁴Many

synthetic cell cycle inhibitors such as flavopiridol, olo-

moucine, roscovitine, and puvalanol B are viewed as a

new generation of anticancer drugs, and some are being

investigated in clinical trials.²⁵Cell cycle analysis of

procyanidin-treated MDA MB 231 cells clearly showed

a G₀/G₁arrest by the pentamer, no effect by the dimer

or trimer, and only a slight effect by the tetramer (Table

1). The increase in G₀/G₁was accompanied with a

decrease of cell numbers in the S phase and in G₂/M. The

pentamer-caused G₀/G₁arrest was reversible in cells

treated up to 8 h and irreversible after a 24 h treatment.

No difference in activity was observed between natural

and synthetic trimer 30; an approximately 15% increase

in G₀/G₁arrest was seen for synthetic tetramer 31

compared to natural material, and an approximately 30%

increase for synthetic vs natural pentamer 32 (Table 2).

These differences can be attributed to the higher purities

(24) (a) Paggi, M. G.; Baldi, A.; Bonetto, F. Giordano, A. J . Cell

Biochem. 1996, 62, 418. (b) McDonald, E. R., III; El Deiry, W. S. Int.

J . Oncol. 2000, 16, 871. (c) Hanahan, D.; Weinberg, R. A. Cell 2000,

100, 57.

(25) (a) Buolamwini, J . K. Curr. Pharm. Des. 2000, 6, 379. (b)

Stadler, W. M.; Volgelzang, N. J .; Amato, R.; Sosman, J .; Taber, D.;

Liebowitz, D.; Vokes, E. E. J . Clin. Oncol. 2000, 18, 371. (c) Senderow-

icz, A. M.; Headlee, D.; Stinson, S. F.; Lush, R. M.; Kalil, N.; Villalba,

L.; Hill, K.; Steinberg, S. M.; Figg, W. D.; Tompkins, A.; Arbuck, S.

G.; Sausville, E. A. J . Clin. Oncol. 1998, 16, 2986.

1650 J . Org. Chem., Vol. 68, No. 5, 2003

Studies in Polyphenol Chemistry and Bioactivity

F IGURE 4. Flow cytometry of procyanidin treated MDA MD 231 human breast cancer cells using Annexin V-FITC and propidium

iodide (control versus 24 h treatment with 200 µg/mL of oligomer): A, 30; B, 31; C, 32. Lower left quadrant: viable cells. Lower

right quadrant: early apoptotic events. Upper right quadrant: late apoptotic events. Upper left quadrant: nonviable cells.

of the synthetic procyanidins (see above). Cell cycle arrest

induced by monomeric catechins, especially epigallocat-

echin gallate (EGCG), has previously been observed by

others.²⁶

Cells treated with both natural and synthetic procyani-

dins showed similar profiles, and increases in cell popu-

lations in the upper right quadrant were observed as the

oligomer size increased. This quadrant represents an-

nexin V positive cells that also take up propidium iodide.

These cells are considered to be in either late apoptosis

or in necrosis. The absence of a distinct cell population

in the lower right quadrant (cells associated with early

apoptotic events) in the case of pentamer-treated cells

suggests a necrotic pathway to cell death possibly due to

a direct interaction with the cell membrane leading to

damage, cell crisis, and eventual death. However, this

conclusion will require additional investigations for

verification. We note that a number of other polyphenols,

especially the important green tea constituent, EGCG,

have been demonstrated to induce apoptosis.^3i-k,26a

A recent report has shown H₂O₂to be artifactually

produced in vitro by several different polyphenolic com-

Cell cycle arrest may result in cell death, which was

indeed observed in the present experiments. The manner

of cell death (apoptosis or necrosis) was investigated by

the annexin V-fluorescein isothiocyanate (FITC) assay

using Trevigen’s TACS Annexin V-FITC kit. Cell cycle

analysis of MDA MB 231 cells treated with natural and

synthetic oligomers 30, 31, and 32 is shown in Figure 4.

(26) (a) Ahmad, N.; Feyes, D. K.; Nieminen, A.-L.; Agarwal, R.;

Mukhtar, H. J . Natl. Cancer Inst. 1997, 89, 1881. (b) Liang, Y. C.;

Lin-Shiau, S.-Y.; Chen, C. F.; Lin, J . K. J . Cell Biochem. 1999, 75, 1.

(c) Ahmad, N.; Cheng, P.; Mukhtar, H. Biochem. Biophys. Res.

Commun. 2000, 275, 328. (d) Gupta, S.; Ahmad, N.; Nieminen, A.-L.;

Mukhtar, H. Toxicol. Appl. Pharmacol. 2000, 164, 82. (e) Tan, X.; Hu,

D.; Li, S.; Han, Y.; Zhang, Y.; Zhou, D. Cancer Lett. 2000, 158, 1.

J . Org. Chem, Vol. 68, No. 5, 2003 1651

Kozikowski et al.

% G₂/M

TABLE 3. Cell Cycle An a lysis of MDA MB 231 Tr ea ted Cells

% G₀/G₁

% S

control

vehicle

100 µM H₂O₂; 24 h

100 µM H₂O₂+ catalase; 24 h

100 µM H₂O₂+ heat inactivated catalase; 24 h

control

vehicle

32 (200 µg/mL); 24 h

32 (200 µg/mL) + catalase; 24 h

32 (200 µg/mL) + heat inactivated catalase; 24 h

32 (200 µg/mL) + 100 µM H₂O₂; 24 h

32 (200 µg/mL) + 100 µM H₂O₂+ catalase; 24 h

32 (200 µg/mL) + 100 µM H₂O₂+ heat inactivated catalase; 24 h

33.14

36.44

20.32

35.20

20.27

29.87

30.28

44.94

41.23

42.89

42.67

48.20

39.47

44.06

41.63

44.92

42.93

45.48

46.21

47.25

38.01

39.65

39.43

18.63

31.12

23.39

22.81

21.94

34.76

21.86

34.25

23.92

22.47

17.05

20.12

17.68

38.71

20.68

37.14

Waters µBondapak C₁₈, 300 × 19 mm; column D, Waters

µBondapak C₁₈, 300 × 30 mm, at 42 mL/min; column E,

Whatman Partisil 10, 500 × 9.4 mm, at 5.0 mL/min; column

F, Whatman Partisil 10, 500 × 22 mm, at 26 mL/min.

Detection by UV absorption at 280 nm. Retention times vary

substantially depending on column history and other subtle

circumstances. They are quoted solely for orientation and

should not be employed for product identification without

comparison to an authentic reference sample.

HP LC An a lysis of P r ocya n id in s. Normal-phase separa-

tions were performed on a 250 × 4.6 mm Phenomenex 5 µm

Prodigy column. The detector was a fluorescence detector

operating at λ_ex) 276 nm and λ_em) 316 nm. The ternary

mobile phase consisted of (A) dichloromethane, (B) methanol

and (C) acetic acid/water (1:1 v/v). Separations were effected

by a series of linear gradients of B into A with a constant 4%

C at a flow rate of 1 mL/min as follows: 0-30 min, 14.0-

28.4% B in A; 30-50 min, 28.4-38.0% B in A; 50-51 min,

38.0-86.0% B in A; 51-56 min, 86.0% B in A isocratic.

HP LC/MS An a lysis of P r ocya n id in s. HPLC/MS analyses

of natural and synthetic procyanidins were performed on an

HPLC system as described above which was interfaced to mass

selective detector equipped with an API-ES ionization cham-

ber. Ionization reagents were added via a tee in the eluent

stream just prior to the mass spectrometer. Conditions for

analysis in the positive ion mode included the introduction of

0.05 M sodium chloride at a flow rate of 0.05 mL/min to assist

ionization, a capillary voltage of 3.5 kV, a fragmentor voltage

of 100 V, a nebulizing pressure of 25 psig, and a drying gas

temperature of 350 °C. Scans were performed over a mass

range of m/z 100-3000 at 1.96 s per cycle.

pounds and to be reponsible for causing a variety of

biological activities.²⁷Our results (Table 3) indicate that

if H₂O₂were present at the levels reported in the

literature, it would produce a shift in the cell cycle to

G₂/M with a decrease in G₀/G₁. The addition of catalase

abrogated these effects, causing a shift in the cell cycle

back to control values. The addition of catalase alone to

pentamer-treated cells produced no conclusive change in

the cell cycle attributable to H₂O₂, i.e., the typical G₀/G₁

arrest caused by the pentamer remained essentially

unchanged. To eliminate the possibility that the pen-

tamer might inhibit catalase activity, H₂O₂was added

to pentamer treated cells in the presence and absence of

catalase. The addition of H₂O₂to pentamer-treated cells

led to an increase in G₀/G₁and G₂/M at the expense of

cells in the S phase. Catalase addition caused a shift back

to the G₀/G₁arrest typical of pentamer-treated cells, and

heat-inactivated catalase had no effect. We conclude that

the G₀/G₁arrest is directly caused by the pentamer and

not by H₂O₂.

Collectively, these results confirm the cytotoxicity to

human breast cancer cell lines by an epicatechin pen-

tamer purified from cocoa and assigned the structure 32.

This procyanidin causes a G₀/G₁arrest in MDA MB 231

cells that is independent of any effects caused by H₂O₂.

An increase in annexin V and propidium iodide positive

cells suggests that pentamer-treated cells quickly enter

into a necrotic phase of cell death. However, additional

efforts are required to elucidate the molecular mecha-

nisms by which 32 exerts its effects in vitro and to

conclusively determine whether apoptosis or necrosis is

the pathway of cell death.

Con d en sa tion of 1 w ith 2 Ca ta lyzed by Ben ton ite K-10.

Bis(5,7,3′,4′-tetr a -O-ben zyl)ep ica tech in 4â,8-Dim er (3). To

a

solution/suspension of 9.26 g (14.2 mmol, 4 equiv) of

compound 1 and 5.0 g of Bentonite K-10 in 115 mL of

anhydrous CH₂Cl₂was added with ice cooling, stirring, and

exclusion of moisture within 2.5 h 2.53 g (3.56 mmol) of

compound 2 in 35 mL of anhydrous CH₂Cl₂. The bath tem-

perature rose to +6 °C at the end of the addition. Stirring in

the bath was continued for 1 h, during which time the

temperature rose to +12 °C. The clay was filtered off with

suction over Celite, and the solids were washed with 2 × 50

mL of CH₂Cl₂. Twenty milliliters of toluene was added, and

the solution was evaporated to a small volume. The residue

was chromatographed on silica gel (60 × 5 cm) with EtOAc/

CHCl₃/hexane 1:14:14. Initially, 5.95 g of unreacted 1 was

eluted, followed by 4.01 g of monomer/dimer mixed fractions

and 1.15 g of pure (98% by HPLC) dimer 3. The last traces of

the dimer together with the trimer were eluted as a mixed

fraction (0.27 g) with a solvent ratio of 1:7:7.

Exp er im en ta l Section

Pearlman’s catalyst (20% Pd(OH)₂/C) contained up to 50%

H₂O. For other chemicals, see ref 1a. ¹H and ¹³C NMR spectra

were acquired at nominal frequencies of 300 and 75 MHz,

respectively, in CDCl₃unless specified otherwise. ¹H NMR

spectra are referenced to internal TMS; ¹³C NMR spectra to

internal TMS if so marked, otherwise to the CDCl₃signal (δ

77.00). Combustion analyses: Micro-Analysis, Inc. (Wilming-

ton, DE). Column chromatography (CC): silica gel 60, particle

size 63-200 µm. TLC: silica gel 60 with fluorescent indicator,

layer thickness 250 µm; visualization by UV light (254 nm) or

with alkaline KMnO₄solution. HPLC: column A, Hewlett-

Packard RP-8, 200 × 4.6 mm, at 1.0 mL/min; column B, Waters

µBondapak C₁₈, 300 × 7.8 mm, at 2.8 mL/min; column C,

The mixed fractions were each dissolved in CH₃CN and

separated by preparative HPLC (column D; 0-30 min, 80 to

100% CH₃CN in H₂O, then CH₃CN); the retention times for

the dimer and trimer were 23.3 and 30.1 min, respectively.

After combination of appropriate fractions, evaporation, and

(27) Long, L. H.; Clement, M. V.; Halliwell, B. Biochem. Biophys.

Res. Commun. 2000, 273, 50.

1652 J . Org. Chem., Vol. 68, No. 5, 2003

Studies in Polyphenol Chemistry and Bioactivity

drying in vacuo, the following yields were obtained: 1, 6.89 g

(74% recovery); 3, 4.26 g (92%); 4, 74 mg (2%). Title com-

pound: ¹³C NMR (CDCl₃, TMS) δ 158.34, 158.07, 157.91,

157.07, 156.83, 156.56, 156.49, 155.89, 155.53, 155.07, 154.44,

152.83, 149.17, 149.01, 148.92, 148.66, 148.60, 148.40, 148.18,

137.40, 137.38, 137.30, 137.28, 137.22, 137.17, 137.01, 136.97,

132.61, 132.43, 131.18, 131.14, 128.6-126.6, 119.96, 119.79,

118.79, 118.65, 115.02, 114.89, 114.35, 114.05, 113.52, 112.93,

112.46, 111.58, 111.17, 104.45, 102.29, 101.76, 94.34, 93.96,

93.33, 93.15, 92.93, 91.52, 78.84, 78.07, 75.63, 72.41, 72.14,

71.48, 71.35, 71.22, 70.81, 70.48, 69.92, 69.86, 69.78, 69.47,

69.05, 66.50, 65.15, 35.90, 35.78, 28.74, 28.61. Other data have

been published.^1a

In another run (3.17 mM in 2), an essentially complete

separation of monomer and dimer and of dimer and trimer

was achieved during column chromatography, with only the

trimer and the very dilute tail of the dimer requiring purifica-

tion by HPLC. The following yields were obtained: 1, 6.20 g

(75% recovery); 3, 3.63 g (88%); 4, 0.15 g (5%).

Con d en sa tion of 3 w ith 2 Ca ta lyzed by Ben ton ite K-10.

To a solution/suspension of 5.60 g (4.31 mmol, 3 equiv) of

compound 3 and 2.04 g of Bentonite K-10 in 45 mL of

anhydrous CH₂Cl₂was added with ice cooling, stirring, and

exclusion of moisture within 110 min 1.02 g (1.44 mmol) of

compound 2 in 15 mL of anhydrous CH₂Cl₂. The bath tem-

perature rose to +6 °C at the end of the addition. Stirring in

the bath was continued for 1 h, during which time the

temperature rose to +12 °C. The clay was filtered off with

suction over Celite, and the solids were washed with 4 × 25

mL of EtOAc. The combined solutions were evaporated.

Attempted separation by column chromatography on silica gel

(56 × 5 cm) with EtOAc/CHCl₃/hexane 1:10:10 failed to

separate the dimer and the trimer; subsequent elution with a

solvent ratio of 1:7:7 gave 0.50 g of a fraction consisting mostly

of tetramer together with residual trimer. The dimer/trimer

fraction was again subjected to column chromatography on

silica gel (55 × 5 cm), this time starting with EtOAc/CHCl₃/

hexane 1:14:14. After elution with 20 L of this mixture, the

solvent ratio was switched to 1:12:12 (5 L), then 1:10:10,

resulting in the recovery of 4.40 g of the dimer. Further elution

with a mixing ratio of 1:8:8 gave 1.04 g of crude trimer (purity

90% by HPLC).

matographed on a short SiO₂column with EtOAc/toluene 1:19

(until the beginning elution of product), then 1:9. The eluate

was evaporated to yield an oil, which soon turned into a light-

yellow solid. This material was dissolved in 30 mL of hot

EtOAc, 90 mL of 1-chlorobutane was added, and the solution

was seeded and set aside for crystallization first at room

temperature, then at -20 °C. The precipitate was isolated by

suction filtration, washed with 2 × 20 mL of cold 1-chlorobu-

tane, and dried in vacuo to afford 3.50 g of the predominant

diastereoisomer. Chromatography of the mother liquor (SiO₂,

EtOAc/CH₂Cl₂/hexane 1:18:11 to 2:18:11) followed by crystal-

lization from EtOAc/1-chlorobutane yielded an additional 0.78

g of the major isomer (together 4.28 g, 67%) and 0.16 g (2.5%)

of the less polar minor isomer. Major diastereoisomer 10a : mp

160-161 °C (from EtOAc/1-chlorobutane); [R]_D+106, [R]₅₄₆

+133 (EtOAc, c 10.6 g L^-1); ¹H NMR (CDCl₃) δ 7.89 (ddd, 1 H,

J ) 8, 1.2, 0.7 Hz), 7.78 (ddd, 1 H, J ) 8, 1.2, 0.7 Hz), 7.47-

7.20 (m, 19 H), 7.17 (d, 1 H, J ) 2 Hz), 7.12-7.00 (m, 4 H),

6.95 (B part of an ABq, 1 H, J ) 8.5 Hz), 6.30, 6.29 (ABq, 2 H,

J ) 2 Hz), 5.46 (d, 1 H, J ) 2 Hz), 5.42 (s, 1 H), 5.17 (s, 2 H),

5.16 (s, 2 H), 5.10, 5.05 (ABq, 2 H, J ) 12 Hz), 5.03 (s, 2 H),

4.40 (ddd, 1 H, J ) 6, 2.5, 1 Hz), 2.00 (d, 1 H, J ) 5.5 Hz); ¹³

C

NMR (CDCl₃) δ 165.00, 160.67, 158.76, 155.95, 153.16, 148.96,

148.88, 137.17, 137.07, 136.53, 136.47, 135.29, 130.76, 128.61,

128.45, 128.37, 128.16, 128.09, 127.75, 127.56, 127.49, 127.46,

127.21, 126.57, 126.06, 124.41, 121.83, 120.97, 119.65, 114.91,

113.58, 98.34, 94.48, 75.13, 71.32, 71.22, 70.78, 70.13, 69.86,

44.43. Anal. Calcd for C₅₀H₄₁NO₆S₂: C, 73.60; H, 5.06; N, 1.72.

Found: C, 73.92; H, 4.75; N, 1.74. Minor diastereoisomer

10b: see the Supporting Information.

3-O-Acetyl-4-[(2-ben zoth ia zolyl)th io]-5,7,3′,4′-tetr a -O-

ben zylep ica tech in (12). To a solution of 3.50 g (4.29 mmol)

of 10a and 53 mg (0.43 mmol) of 4-(dimethylamino)pyridine

in 12 mL of anhydrous pyridine was added all at once 2.0 mL

(21.5 mmol) of acetic anhydride. The reaction mixture was kept

at room temperature in a closed flask for 50 h. Ice and 150

mL of 5% aqueous HCl were added, and the product was

extracted into 100 + 20 mL of CH₂Cl₂. The combined organic

phases were washed with 100 mL of H₂O and 2 × 50 mL of

10% aqueous NaOH; after each washing, the aqueous phase

was back-extracted with 20 mL of CH₂Cl₂. The combined

organic phases were dried over MgSO₄and evaporated, and

the residue was taken up in a small volume of toluene and

filtered over SiO₂with EtOAc/hexane 1:3. Evaporation and

drying in vacuo yielded 3.58 g (97%) of the acetate as a

yellowish foam: [R]_D+91.7, [R]₅₄₆+115 (EtOAc, c 13.2 g L^-1);

¹H NMR (CDCl₃) δ 7.90 (d, 1 H, J ) 8 Hz), 7.77 (d, 1 H, J )

8 Hz), 7.46-7.22 (m, 19 H), 7.11 (d, 1 H, J ) 2 Hz), 7.09-7.00

(m, 3 H), 6.99, 6.91 (ABq, 2 H, J ) 8.5 Hz, A part d with J )

2 Hz), 6.31, 6.30 (ABq, 2 H, J ) 2.5 Hz), 5.63 (dd, 1 H, J )

2.5, 1.2 Hz), 5.55 (s, 1 H), 5.31 (d, 1 H, J ) 2 Hz), 5.17, 5.12

(ABq, 2 H, J ) 12 Hz), 5.14 (s, 2 H), 5.10, 5.05 (ABq, 2 H, J

not readable because of overlap), 5.07, 5.02 (ABq, 2 H, J )

11.5 Hz), 1.84 (s, 3 H); ¹³C NMR (CDCl₃, TMS) δ 169.08,

164.07, 160.69, 158.31, 156.03, 153.22, 148.92, 148.89, 137.18,

137.16, 136.53, 136.31, 135.62, 130.29, 128.67, 128.45, 128.24,

128.19, 127.78, 127.65, 127.43, 127.31, 126.87, 126.10, 124.50,

122.16, 121.01, 119.80, 114.97, 113.51, 98.50, 94.46, 94.30,

74.13, 71.44, 71.23, 70.74, 70.19, 70.13, 42.59, 20.84. Anal.

Calcd for C₅₂H₄₃NO₇S₂: C, 72.79; H, 5.05; N, 1.63. Found: C,

73.01; H, 4.79; N, 1.61.

The crude trimer and the trimer/tetramer mixture were

each dissolved in CH₃CN and separated by preparative HPLC

(column D; 0-30 min, 80 to 100% CH₃CN in H₂O, then CH₃-

CN); the retention times for the dimer, trimer, and tetramer

were 22.5 (22.7), 30.1 (30.8), and 33.9 min, respectively. After

combination of appropriate fractions, evaporation, and drying

in vacuo, the following yields were obtained: 3, 4.43 g (79%

recovery); 4, 1.13 g (40%); 6, 0.24 g (13%).

4-[(2-Ben zot h ia zolyl)t h io]-5,7,3′,4′-t et r a -O-b en zylep i-

ca tech in (10a ,b). Caution! Although 2-mercaptobenzothiazole

is odorless, small quantities of malodorous (but not very

volatile) 2-(benzylthio)benzothiazole are formed in this reac-

tion, which should therefore be conducted in a well-ventilated

fume hood. To a solution of 6.5 g (39 mmol) of 2-mercaptoben-

zothiazole in 40 mL of 1,2-dichloroethane [HPLC grade,

filtered over basic alumina (activity I) immediately before use]

was added dropwise in 10 min under N₂with ice cooling and

stirring 19.5 mL of trimethylaluminum solution (2.0 M in

toluene). The resulting amber solution was stirred at 0 °C for

15 min, then a solution of 5.56 g (7.82 mmol) of compound 2

in 60 mL of 1,2-dichloroethane (pretreated as above) was added

dropwise in 20 min. The orange-colored reaction mixture was

stirred at room temperature for 5 h, then cooled in an ice bath,

and a solution of 22.6 g (80 mmol) of potassium sodium tartrate

tetrahydrate in 90 mL of H₂O and 100 mL of 2.5 M aqueous

NaOH was added dropwise (very cautiously at first; gas

evolution!). Methylene chloride (100 mL) was added, and the

phases were separated. The organic phase was washed with

2 × 100 mL of 2.5 M aqueous NaOH and dried over Na₂SO₄.

After evaporation to a small volume, the residue was chro-

Bis(3-O-a cetyl-5,7,3′,4′-tetr a -O-ben zyl)ep ica tech in 4â,8-

Dim er (13). A solution of 1.5 mL (16 mmol) of acetic anhydride

in 4 mL of anhydrous pyridine was added all at once to 3.69 g

(2.84 mmol) of 3. The mixture was occasionally swirled until

all starting material dissolved and then allowed to stand in a

closed flask at room temperature for 99 h. The reaction was

terminated by addition of 30 mL of EtOAc and 2 mL of MeOH

and allowed to stand at room temperature for 1.5 h. Another

20 mL of EtOAc was added, and then the solution was washed

with 200 mL of 0.5 M aqueous H₃PO₄. The aqueous layer was

back-extracted with 50 mL of EtOAc, and the combined organic

J . Org. Chem, Vol. 68, No. 5, 2003 1653

Kozikowski et al.

phases were dried over MgSO₄. After evaporation, the residue

was taken up in a small volume of toluene and chromato-

graphed on a short SiO₂column with EtOAc/hexane (1:9, then

1:3, then 1:1). Evaporation and drying in vacuo yielded 3.82 g

(97%) of the acetate as a colorless foam: [R]_D+44.9, [R]₅₄₆

5%); 15 (34 mg, 6%); trimer (3.07 g, 70% recovery); tetramer

(1.47 g, 62%); pentamer (221 mg, 15%); hexamer (57 mg, 5%);

heptamer (25.2 mg, 2%); octamer (10.8 mg, 1%). Total mass

balance relative to 12: 96%.

Rea ction of 12 w ith 17. The reaction was conducted

analogously, using 0.34 g (1.75 mmol) of AgBF₄, 4.77 g (1.73

mmol) of 17, and 592 mg (690 µmol) of 12. After filtration over

SiO₂with EtOAc/hexane 1:1, the crude product was subjected

in several portions to a preliminary separation by preparative

HPLC (column D; 0-30 min, 80 to 100% CH₃CN in H₂O; 30-

38 min, CH₃CN; 38-65 min, 10% EtOAc in CH₃CN) to yield

the following products: 14 (t_R11.0 min; 19 mg, 4%); 15 (22.2

min; 47 mg, 10%); tetramer (36.0 min; 3.56 g, 74% recovery);

pentamer (39.4 min; 1.03 g); hexamer (43.6 min; 260 mg);

heptamer (46.0 min; 86 mg); octamer (48.9 min; 41 mg);

nonamer (52.2 min; 22 mg); decamer (56.2 min; 13.5 mg);

undecamer (61.4 min; 8.2 mg). All products from the pentamer

on required additional purification because of peak tailing,

which led to a contamination with lower oligomers that

increased with the degree of oligomerization, and because of

increasing contamination with unidentified aliphatic material

from the nonpolar solvent and/or column. For sample prepara-

tion, a small percentage of THF had to be added to the CH₃-

CN from the heptamer on because of limited solubility in

CH₃CN alone. For the pentamer through nonamer, the ad-

ditional purification was performed on column D (0-30 min,

80 to 100% CH₃CN in H₂O, then CH₃CN), For the decamer

and undecamer, column B was used in combination with the

same gradient. The nonamer, decamer, and undecamer still

contained excessive amounts of aliphatic impurities after this

treatment and were subjected to a third HPLC purification

on column A using the same gradient. The following yields of

pure products (97% or better by HPLC) were obtained:

pentamer, 987 mg (41%); hexamer, 226 mg (16%); heptamer,

68 mg (6.1%); octamer, 26 mg (2.7%); nonamer, 11.5 mg (1.3%);

decamer, 6.5 mg (0.8%); undecamer, 2.5 mg (0.3%). Total mass

balance relative to 12: 82%.

1

+54.6 (EtOAc, c 14.2 g L^-1); H NMR (CDCl₃; two rotamers,

86:14) major rotamer and shared multiplets δ 7.47-6.82 (m,

44 H), 6.75, 6.41 (ABq, 2 H, J ) 8 Hz, B part d with J ) 1.5

Hz), 6.31 (s, 1 H), 6.02 (d, 1 H, J ) 2 Hz), 5.70 (s, 1 H), 5.62

(d, 1 H, J ) 2 Hz), 5.46 (t, 1 H, J ) 1.5 Hz), 5.12 (s, 2 H),

5.10-4.89 (m, 11 H), 4.80 (s, 2 H), 4.77 (s, 1 H), 4.63, 4.46

(ABq, 2 H, J ) 11 Hz), 4.08 (s, 1 H), 2.99, 2.90 (ABq, 2 H, J )

18.5 Hz, A part d with J ) 5 Hz), 1.72 (s, 3 H), 1.68 (s, 3 H);

conspicuous signals of minor rotamer δ 6.21 (br, 1 H), 6.17 (s,

1 H), 6.11 (br, 1 H), 5.44 (br, 1 H), 5.34 (s, 2 H), 4.85 (s, 2 H),

4.61, 4.34 (ABq, 2 H, J ) 12 Hz), 1.43 (s, 3 H), 1.31 (s, 3 H);

¹³C NMR (CDCl₃, TMS; major rotamer only) δ 170.10, 169.10,

158.01, 156.35, 155.96, 155.46, 154.50, 149.17, 148.85, 148.42,

148.24, 137.36-137.23, 136.96, 132.35, 130.74, 128.7-127.1,

126.56, 119.84, 119.31, 114.60, 113.78, 113.59, 112.33, 110.58,

104.43, 101.66, 93.76, 92.89, 91.19, 77.96, 71.46, 71.27, 71.20,

70.51, 69.95, 69.69, 69.57, 69.09, 68.34, 33.20, 26.95, 20.89,

20.80. Anal. Calcd for C₉₀H₇₈O₁₄: C, 78.13; H, 5.68. Found:

C, 78.22; H, 5.43.

Rea ction of 12 w ith 13. An 0.80 g (4.1 mmol) sample of

AgBF₄was dried in the reaction flask at 100 °C in an oil pump

vacuum with exclusion of light for 1.5 h. After cooling, the

vacuum was broken with N₂, and a solution of 5.66 g (4.09

mmol) of 13 in 60 mL of anhydrous THF was added all at once.

The flask was placed in an ice bath under dim light, and a

solution of 1.40 g (1.64 mmol) of 12 in 30 mL of anhydrous

THF was added dropwise in 70 min with stirring. The reaction

mixture turned yellow, and a turbidity eventually appeared.

Stirring at 0 °C was continued for 40 min, during which time

period the reaction mixture turned into a milky, whitish

suspension. Triethylamine (1.1 mL, 8 mmol) was added, the

mixture was evaporated to near dryness, and the residue was

filtered over a short SiO₂column with EtOAc/hexane 1:1. The

eluate was evaporated, and the crude product was analyzed

by HPLC (column A; 0-30 min, 80 to 100% CH₃CN in H₂O,

then CH₃CN). The following peaks were observed (assignment/

area%): t_R5.0 (4-OH-monomer, 0.15), 12.6 (4-OH-dimer, 0.25),

15.6 (dimer, 59.4), 24.8 (trimer, 23.4), 30.3 (tetramer, 12.5),

33.3 (pentamer, 3.2), 35.4 (hexamer, 0.8), 37.3 (heptamer, 0.1),

39.1 min (octamer, 0.02). A partial separation was achieved

by column chromatography on SiO₂(38 × 9 cm). Initial elution

with 25 L of EtOAc/CHCl₃/hexane 1:10:9 did not result in

product recovery (this stage is, however, essential for achieving

separation). Another 25 L of EtOAc/CHCl₃/hexane 1:11:8

eluted 4.01 g of the dimer (71% recovery; pure by HPLC). A

fraction (1.72 g) consisting of trimer, tetramer, and some

pentamer was eluted with 20 L of EtOAc/CHCl₃/hexane 1:12:

7; finally, the column was stripped with EtOAc/CHCl₃/hexane

2:12:7 to give 0.87 g of a fraction consisting mostly of the larger

oligomers. The latter two fractions were taken up in CH₃CN

and separated in several portions by prep. HPLC (column D;

0-30 min, 80-100% CH₃CN in H₂O, then CH₃CN), and

appropriate fractions were pooled and dried in vacuo to obtain

the oligomers as colorless films or foams. The retention times

and yields relative to 12 for the trimer through octamer were

31.9 min (1.46 g, 43%), 36.0 min (755 mg, 33%), 39.6 min (204

mg, 11%), 45.0 min (45 mg, 2.6%), 52.8 min (13.8 mg, 0.9%),

and 64.1 min (5.2 mg, 0.3%), respectively; for 15, 22.2 min (13.4

mg, 1.2%). The 4-OH-monomer 14 was not recovered from the

SiO₂column, probably because of its high polarity. The total

mass balance relative to 12 was 92%.

Tris(3-O-acetyl-5,7,3′,4′-tetra-O-benzyl)epicatechin (4â,8)₂-

Tr im er (16). This product was also obtained in 97% yield from

412 mg (211 µmol) of 4 and 1 mL each of acetic anhydride

and pyridine in a similar manner as described for the acety-

lation of 3. Colorless foam; [R]_D+108, [R]₅₄₆+131 (EtOAc, c

1

19.3 g L^-1); H NMR (CDCl₃; three rotamers, 60:29:11; selec-

tion) δ 6.30 (s), 5.94 (narrow m), 5.87 (s), 5.81 (s), 5.63 (d, 1 H,

J ) 1.5 Hz), 5.53 (s) (these are the 6 most intense signals in

the range of δ 6.6-5.2), 3.11-2.88 (m), 1.72 (s, major + third

rotamer), 1.67 (s, third rotamer), 1.55 (third rotamer), 1.45

(major rotamer), 1.36 (second rotamer), 1.27 (major + second

rotamer), 1.17 (second rotamer); ¹³C NMR (CDCl₃, TMS; only

intense signals below δ 145 and above δ 40) δ 170.18, 169.16,

168.60, 158.05, 156.62, 155.96, 155.81, 155.73, 155.69, 155.62,

155.27, 154.58, 152.65, 149.22, 148.89, 148.56, 148.38, 148.24,

148.13, 34.50, 32.99, 25.76, 20.85, 20.60, 20.31; acetate methyl

groups of the second isomer: δ 20.41, 20.21 (× 2); of the third

isomer: δ 20.93, 20.80, 20.75. Anal. Calcd for C₁₃₅H₁₁₆O₂₁: C,

78.17; H, 5.64. Found: C, 78.23 H, 5.58.

T e t r a k i s (3-O -a c e t y l-5,7,3′,4′-t e t r a -O -b e n z y l)e p i -

ca tech in (4â,8)₃-Tetr a m er (17). This product was also

obtained in 95% yield from 145 mg (55.8 µmol) of 6 and 0.25

mL each of acetic anhydride and pyridine in a similar manner

as described for the acetylation of 3. Colorless foam: [R]_D+122,

[R]₅₄₆+148 (EtOAc, c 11.7 g L^-1); ¹H NMR (CDCl₃; two

rotamers, 74:26; selection) δ 6.50, 6.48, 6.31, 6.22, 6.19, 6.00,

5.97, 5.85, 5.78, 5.76, 5.70, 5.57, 5.56, 5.50, 5.44, 5.43 (multi-

plicities unknown, all signals reported as s), 3.09-2.84 (m), 1.72

(s, major rotamer), 1.38 (s, minor rotamer), 1.31 (s, major

rotamer), 1.24 (s, major rotamer), 1.19 (s, both rotamers), 1.12

(s, minor rotamer), 1.08 (s, minor rotamer); ¹³C NMR (CDCl₃,

TMS; only intense signals below δ 145 and above δ 40) δ

170.15, 169.13, 168.98, 168.70, 157.97, 156.51, 156.03, 155.99,

155.70, 155.62, 155.54, 155.36, 155.16, 154.49, 153.28, 152.84,

149.22, 148.92, 148.56, 148.48, 148.37, 148.21, 147.65, 147.51,

Rea ction of 12 w ith 16. The reaction was conducted

analogously, using 0.41 g (2.1 mmol) of AgBF₄, 4.40 g (2.12

mmol) of 16, and 729 mg (850 µmol) of 12. After filtration over

SiO₂with EtOAc/hexane 1:1, the crude product was taken up

in CH₃CN and separated in several portions by preparative

HPLC as above to yield the following products: 14 (31 mg,

1654 J . Org. Chem., Vol. 68, No. 5, 2003

Studies in Polyphenol Chemistry and Bioactivity

34.59, 32.90, 32.13, 25.41, 20.85, 20.34, 20.26 (× 2). Anal. Calcd

N o n a k i s (3-O -a c e t y l-5,7,3′,4′-t e t r a -O -b e n z y l)e p i -

ca tech in (4â,8)₈-Non a m er (22). Colorless foam: [R]_D+117,

[R]₅₄₆+142 (EtOAc, c 9.2 g L^-1); ¹H NMR (CDCl₃; at least two

rotamers, ratio not determinable because of signal overlap;

selection) δ 6.67, 6.64, 6.58, 6.56, 6.53, 6.47, 6.45, 6.41, 6.38,

6.36, 6.30, 6.24, 6.22, 6.08, 6.05, 5.98, 5.87, 5.85, 5.83, 5.78,

5.66, 5.60, 5.56 (d, J ) 2 Hz), 5.54-5.50, 5.45 (multiplicities

unknown except for the narrow d, all other signals reported

as s), 3.08-2.87 (m), 1.71, 1.38 (each s, major rotamer), 1.20

(s, minor rotamer), 1.15, 1.10 (× 2), 1.07 (each s, major

rotamer), 1.02 (s, minor rotamer), 1.01 (s, major rotamer), 1.00

(minor rotamer), 0.98, 0.95 (each s, major rotamer).

D e c a k i s (3-O -a c e t y l-5,7,3′,4′-t e t r a -O -b e n z y l)e p i -

ca tech in (4â,8)₉-Deca m er (23). Colorless foam: [R]_D+114,

[R]₅₄₆+139 (EtOAc, c 3.6 g L^-1); ¹H NMR (CDCl₃; at least two

rotamers, ratio not determinable because of signal overlap;

selection) δ 3.09-2.87 (m), 1.71, 1.38 (each s, major rotamer),

1.20 (s, minor rotamer), 1.15, 1.10 (× 2), 1.07, 1.01 (× 2), 0.98,

0.95 (each s, major rotamer).

for C₁₈₀H₁₅₄O₂₈: C, 78.19; H, 5.61. Found: C, 78.09; H, 5.88.

P e n t a k i s (3-O -a c e t y l-5,7,3′,4′-t e t r a -O -b e n z y l)e p i -

ca tech in (4â,8)₄-P en ta m er (18). Colorless foam: [R]_D+124,

[R]₅₄₆+151 (EtOAc, c 11.9 g L^-1); ¹H NMR (CDCl₃; two

rotamers, 72:28; selection) δ 6.59, 6.56, 6.54, 6.47, 6.44, 6.42,

6.40, 6.33, 6.26, 6.24, 5.96, 5.93, 5.87, 5.83, 5.80, 5.66, 5.57 (d,

J

) 2 Hz), 5.52 (d, 1 H, J ) 1.5 Hz), 5.47, 5.26, 5.21

(multiplicities unknown except for the two narrow d, all other

signals reported as s), 3.09-2.86 (m), 1.72 (s, major rotamer),

1.39 (s, minor rotamer), 1.36 (s, major rotamer), 1.20 (s, minor

rotamer), 1.15 (s, × 2, major rotamer), 1.10 (s, minor rotamer),

1.08 (s, minor rotamer), 1.06 (s, minor rotamer), 1.03 (s, major

rotamer); ¹³C NMR (CDCl₃, TMS; only intense signals below

δ 145 and above δ 40) δ 170.13, 169.15, 169.12, 168.89, 168.67,

157.97, 156.53, 156.20, 155.91 (× 2 or × 3), 155.60, 155.46,

155.30, 155.12, 155.03, 154.41, 153.71, 153.20, 153.02, 149.21,

148.90, 148.74, 148.54, 148.36, 148.15, 148.07, 147.82, 147.47,

147.38, 34.75, 32.91, 31.98, 31.72, 25.33, 20.84, 20.40, 20.27,

20.13, 19.95. Anal. Calcd for C₂₂₅H₁₉₂O₃₅: C, 78.20; H, 5.60.

Found: C, 77.89; H, 5.51.

U n d e c a k is (3-O -a c e t y l-5,7,3′,4′-t e t r a -O -b e n zy l)e p i-

ca tech in (4â,8)₁₀-Un d eca m er (24). Colorless foam: [R]_D

+108, [R]₅₄₆+130 (EtOAc, c 1.1 g L^-1); ¹H NMR (CDCl₃; at

least two rotamers, ratio not determinable because of signal

overlap; selection) δ 3.08-2.87 (m), 1.71, 1.38 (each s, major

rotamer), 1.20 (s, minor rotamer), 1.15, 1.10 (× 2), 1.07, 1.01

(× 3), 0.98, 0.94 (each s, major rotamer).

H e xa k is(3-O-a ce t yl-5,7,3′,4′-t e t r a -O-b e n zyl)e p ica t -

ech in (4â,8)₅-Hexa m er (19). Colorless foam: [R]_D+122, [R]₅₄₆

+147 (EtOAc, c 15.4 g L^-1); ¹H NMR (CDCl₃; at least two

rotamers, ratio not determinable because of signal overlap;

selection) δ 6.66, 6.63, 6.61-6.51, 6.50, 6.47, 6.39, 6.37, 6.31,

6.27, 6.24, 5.98, 5.94, 5.88-5.79, 5.67, 5.61, 5.57 (d, J ) 2 Hz),

5.52 (d, J ) 1.5 Hz), 5.46, 5.21 (multiplicities unknown except

for the two narrow d, all other signals reported as s), 3.09-

2.86 (m), 1.71 (s, major rotamer), 1.60 (s, minor rotamer), 1.38

(s, major rotamer), 1.20 (s, minor rotamer), 1.15, 1.10, 1.07,

1.00 (each s, major rotamer); ¹³C NMR (CDCl₃, TMS; only

intense signals below δ 145 and above δ 40) δ 170.14, 169.10,

168.99, 168.88, 168.63, 157.96, 156.52, 156.21, 155.87, 155.85,

155.63, 155.58, 155.51, 155.29, 155.21, 155.17, 155.11, 154.41,

153.94, 153.63, 153.19, 153.07, 149.20, 148.89, 148.69, 148.51,

148.41, 148.37, 148.14, 148.05, 147.86, 147.66, 147.47, 147.36,

34.71, 32.89, 31.97, 31.79, 31.57, 25.35, 20.83, 20.43, 20.21,

20.13, 20.00, 19.96. Anal. Calcd for C₂₇₀H₂₃₀O₄₂: C, 78.20; H,

5.59. Found: C, 78.33; H, 5.38.

H e p t a k is(3-O-a ce t yl-5,7,3′,4′-t e t r a -O-b e n zyl)e p ica t -

ech in (4â,8)₆-Hep ta m er (20). Colorless foam: [R]_D+120,

[R]₅₄₆+145 (EtOAc, c 10.2 g L^-1); ¹H NMR (CDCl₃; at least

two rotamers, ratio not determinable because of signal overlap;

selection) δ 6.68, 6.65, 6.58, 6.55, 6.53, 6.50, 6.47, 6.44, 6.41,

6.38, 6.31, 6.26, 6.23, 6.10, 5.98, 5.95, 5.88-5.83, 5.82, 5.78,

5.68, 5.62-5.58, 5.56 (d, J ) 2 Hz), 5.54-5.50, 5.47 (multiplici-

ties unknown except for the narrow d, all other signals

reported as s), 3.08-2.87 (m), 1.71, 1.38 (each s, major rotamer),

1.20 (s, minor rotamer), 1.16, 1.10, 1.09 (each s, major

rotamer), 1.08 (s, minor rotamer), 1.04 (s, major rotamer), 1.01,

1.00 (each s, minor rotamer), 0.95 (major rotamer); ¹³C NMR

(CDCl₃, TMS; only intense signals below δ 145 and above δ

40) δ 170.14, 169.12, 169.00, 168.93, 168.66, 157.99, 156.54,

156.45, 155.91, 155.87, 155.7-155.45, 155.35, 155.31, 155.24,

155.13, 154.44, 154.00, 153.92, 153.67, 153.22, 153.09, 149.22,

148.91, 148.69, 148.56, 148.43, 148.16, 148.08, 147.93, 147.8-

147.65, 147.50, 147.40, 34.74, 32.94, 31.99, 31.85, 31.69, 31.60,

29.68, 25.40, 20.83, 20.43, 20.23, 20.16, 20.06, 20.03, 19.92.

Tr is(5,7,3′,4′-tetr a -O-ben zyl)ep ica tech in (4â,8)₂-Tr im er

(4). Preparation by hydrolysis of 16: To a solution of 1.54 g

(742 µmol) of 16 in 30 mL of THF was added all at once 5.8

mL (8.9 mmol) of 40% aqueous tetra-n-butylammonium hy-

droxide. The reaction mixture was allowed to stand at room

temperature in a closed flask for 94 h, then partially evapo-

rated to remove THF. The residue was diluted with 20 mL of

H₂O, the product was extracted into 2 × 20 mL of EtOAc, and

the combined organic phases were washed with 10 mL of brine

and evaporated. Filtration over a short SiO₂column with

EtOAc yielded, after evaporation and drying in vacuo, 1.44 g

(99%) of the product as a colorless foam: [R]_D+72.1, [R]₅₄₆

+86.5 (EtOAc, c 6.9 g L^-1); ¹H NMR (CDCl₃; two major

rotamers, 55:45; selection) δ 6.45 (dd, 1 H, J ) 2, 8.5 Hz), 6.39

(d, 1 H, J ) 8.5 Hz), 6.35 (s, 1 H), 6.27 (d, 1 H, J ) 2 Hz), 6.18

(s, 1 H), 6.14 (dd, 1 H, J ) 2, 8.5 Hz), 6.07 (d, 1 H, J ) 2 Hz),

5.95 (d, 1 H, J ) 2.5 Hz), 5.90 (s, 1 H), 5.85 (s, 1 H), 5.73 (s, 1

H), 5.70 (d, 1 H, J ) 2 Hz), 5.50 (s, 1 H), 3.96 (br d, 1 H, J )

5.5 Hz), 3.64 (br d, 1 H, J ) 4 Hz), 3.07-2.78 (m, 2 H,

containing B part of an ABq(d) at δ 2.83; J _AB) 17.5 Hz, J _d

)

4.5 Hz), 1.81 (d, 1 H, J ) 6 Hz), 1.69 (d, 1 H, J ) 6 Hz), 1.45

(d, 1 H, J ) 7.5 Hz), 1.20 (d, 1 H, J ) 8 Hz), 1.19 (d, 1 H, J )

8 Hz); ¹³C NMR (CDCl₃, TMS; only intense signals below δ

145 and above δ 40) δ 158.29, 158.09, 158.00, 157.40, 156.70,

156.43, 156.33, 156.17, 156.09, 156.04, 155.97, 155.66, 155.44,

155.24, 154.67, 153.10, 152.83, 152.41, 148.18, 148.96, 148.93,

148.88, 148.76, 148.72, 148.66, 148.36, 148.24, 148.17, 148.04,

36.54, 36.15, 35.44, 28.61, 28.55. Anal. Calcd for C₁₂₉H₁₁₀O₁₈

C, 79.53; H, 5.69. Found: C, 79.76; H, 6.03.

:

Tet r a k is(5,7,3′,4′-t et r a -O-b en zyl)ep ica t ech in (4â,8)₃-

Tetr a m er (6). Preparation by hydrolysis of 17: Reaction of

1.59 g (573 µmol) of 17 with 5.6 mL (8.6 mmol) of 40% aqueous

tetra-n-butylammonium hydroxide in 29 mL of THF for 96 h

as described for the trimer yielded 1.45 g (97%) of 6 as a

colorless foam: [R]_D+85.9, [R]₅₄₆+104 (EtOAc, c 12.0 g L^-1);

¹H NMR (CDCl₃; two major rotamers in comparable amounts;

selection) δ 6.43-6.35 (m, 4 H), 6.33 (d, 1 H, J ) 8.5 Hz), 6.27

(d, 1 H, J ) 2 Hz), 6.23 (s, 1 H), 6.14 (dd, 1 H, J ) 1.5, 8.5 Hz),

6.09-6.00 (m, 3 H), 5.92-5.87 (m, 3 H), 5.86 (d, 1 H, J ) 2

Hz), 5.74 (s, 2 H), 5.63 (d, 1 H, J ) 1.5 Hz), 5.53 (s, 1 H), 5.43

(d, 1 H, J ) 8 Hz), 5.39-5.37 (m, 2 H), 5.20 (narrow m, 1 H),

3.65 (d, 1 H, J ) 3.5 Hz), 2.98-2.80 (m, 2 H), 1.79 (d, 1 H, J

) 6 Hz), 1.59 (d, 1 H, J ) 6 Hz), 1.43 (d, 1 H, J ) 7 Hz), 1.37

(d, 1 H, J ) 8 Hz), 1.34 (d, 1 H, J ) 7 Hz), 1.14 (d, 1 H, J )

7.5 Hz), 1.05 (d, 1 H, J ) 8 Hz); ¹³C NMR (CDCl₃, TMS; only

intense signals below δ 145 and above δ 40) δ 158.28, 158.03,

O c t a k i s (3-O -a c e t y l-5,7,3′,4′-t e t r a -O -b e n z y l)e p i -

ca tech in (4â,8)₇-Octa m er (21). Colorless foam: [R]_D+119,

[R]₅₄₆+144 (EtOAc, c 8.5 g L^-1); ¹H NMR (CDCl₃; at least two

rotamers, ratio not determinable because of signal overlap;

selection) δ 6.67, 6.65, 6.60-6.51, 6.48, 6.45, 6.37, 6.31, 6.24,

6.22, 6.09, 6.06, 6.04, 5.98, 5.94, 5.88-5.82, 5.78, 5.66, 5.62-

5.56, 5.54-5.50, 5.46 (multiplicities unknown, all signals

reported as s), 3.08-2.87 (m), 1.71, 1.38 (each s, major rotamer),

1.20 (s, minor rotamer), 1.15, 1.101, 1.098 (each s, major

rotamer), 1.07 (s, minor rotamer), 1.06 (s, major rotamer), 1.02,

1.00 (each s, minor rotamer), 0.98, 0.95 (each s, major rotamer).

Anal. Calcd for C₃₆₀H₃₀₆O₅₆: C, 78.21; H, 5.58. Found: C, 77.79;

H, 5.53.

J . Org. Chem, Vol. 68, No. 5, 2003 1655

Kozikowski et al.

158.01, 157.31, 156.67, 156.56, 156.37, 156.29, 156.15, 156.11,

156.03, 155.96, 155.69, 155.63, 155.35, 155.17, 154.74, 153.22,

153.12, 152.78, 152.75, 149.22, 149.01, 148.93, 148.87, 148.84,

148.82, 148.78, 148.75, 148.53, 148.32, 148.24, 148.09, 148.04,

147.72, 147.63, 36.84, 36.28, 35.61, 35.39, 35.12, 29.64, 28.46.

Anal. Calcd for C₁₇₂H₁₄₆O₂₄: C, 79.55; H, 5.67. Found: C, 79.71;

H, 5.43.

157.34, 156.68, 156.58, 156.40, 156.34, 156.28, 156.16, 156.12,

156.04-155.92, 155.83, 155.72, 155.42, 155.39, 155.16, 154.81,

153.33, 153.28, 153.25, 153.17, 153.12, 153.09, 152.78, 149.23,

149.00-148.82, 148.75, 148.69, 148.53, 148.33, 148.27, 148.05,

148.01, 147.85, 147.80-147.74, 147.70, 147.66, 37.03, 36.31,

35.71, 35.47-35.21, 35.13, 28.46. Anal. Calcd for C₃₀₁

H₂₅₄O₄₂:

C, 79.57; H, 5.64. Found: C, 79.50; H, 5.44.

Octa k is(5,7,3′,4′-tetr a -O-ben zyl)ep ica tech in (4â,8)₇-Oc-

ta m er (28). Reaction of 41.2 mg (26.1 µmol) of 21 with 0.10

mL (0.15 mmol) of 40% aqueous tetra-n-butylammonium

hydroxide in 0.5 mL of THF for 126 h as described for the

trimer yielded 39.4 mg (102%) of the product as a colorless

foam: [R]_D+109, [R]₅₄₆+131 (EtOAc, c 10.8 g L^-1).

P en t a k is(5,7,3′,4′-t et r a -O-b en zyl)ep ica t ech in (4â,8)₄-

P en ta m er (25). Reaction of 1.81 g (524 µmol) of 18 with 6.9

mL (10.5 mmol) of 40% aqueous tetra-n-butylammonium

hydroxide in 35 mL of THF for 118 h as described for the

trimer yielded 1.45 g (97%) of 25 as a colorless foam. The

analytical sample was further purified by preparative HPLC

(column B; 0-30 min, 80-100% CH₃CN/H₂O, then CH₃CN;

Non a k is(5,7,3′,4′-t et r a -O-b en zyl)ep ica t ech in (4â,8)₈-

Non a m er (29). Reaction of 17.9 mg (2.88 µmol) of 22 with 47

µL (72 µmol) of 40% aqueous tetra-n-butylammonium hydrox-

ide in 0.3 mL of THF for 134 h as described for the trimer

yielded 16.8 mg (100%) of the product as a colorless foam: [R]_D

+110, [R]₅₄₆+133 (EtOAc, c 7.8 g L^-1). Anal. Calcd for

1

t_R33.4 min): [R]_D+100, [R]₅₄₆+121 (EtOAc, c 13 g L^-1); H

NMR (CDCl₃; apparent multiplicities; weak signals of minor

rotamers omitted) δ 7.53 (d, 2 H, J ) 7 Hz), 7.5-6.6, 6.57 (s),

6.55 (s), 6.51 (s), 6.49 (s), 6.45 (s), 6.43 (s), 6.39 (s), 6.36 (s),

6.35 (s), 6.334 (s), 6.325 (s), 6.31 (s), 6.28 (s), 6.27 (s), 6.23 (s),

6.17 (d, J ) 2 Hz), 6.14 (d, J ) 1.5 Hz), 6.08 (s), 6.07 (s), 6.04

(s), 5.95 (s), 5.94 (s), 5.88 (s), 5.87 (d, J ) 2 Hz), 5.75 (s), 5.71

(s), 5.63 (d, J ) 2 Hz), 5.51 (s), 5.48 (s), 5.43 (s), 5.40 (s), 5.27

(s), 5.22 (s), 5.20 (d, J ) 2.5 Hz), 5.17 (d, J ) 2 Hz), 5.15-

3.95, 3.89 (br d, J ) 7 Hz), 3.66 (d, J ) 4.5 Hz), 3.0-2.8 (m),

1.80 (d, J ) 6 Hz), 1.59 (d, J ) 6 Hz), 1.42-1.38, 1.37 (d, J )

7 Hz), 1.32, 1.30 (2 overlapping d), 1.24 (d, J ) 7 Hz), 1.14 (d,

J ) 8 Hz), 1.09 (d, J ) 8 Hz); ¹³C NMR (CDCl₃, TMS; only

intense signals below δ 145 and above δ 40) δ 158.28, 158.03,

158.01, 157.33, 156.67, 156.58, 156.39, 156.34, 156.27, 156.15,

156.12, 155.99, 155.92, 155.71, 155.44, 155.37, 155.15, 154.77,

153.24, 153.10, 153.03, 152.94, 152.78, 152.72, 149.21, 148.97,

148.93, 148.84, 148.75, 148.67, 148.51, 148.29, 148.26, 148.24,

148.03, 147.82, 147.72, 147.62, 37.00, 36.29, 35.69, 35.43,

C

₃₈₇H₃₂₆O₅₄: C, 79.58; H, 5.63. Found: C, 79.79; H, 5.34.

Ep ica tech in (4â,8)₂-Tr im er (30). To a solution of 64.3 mg

(33.0 µmol) of 4 in 5 mL of THF were added 5 mL of MeOH,

0.25 mL of H₂O, and 57 mg of 20% Pd(OH)₂/C. The mixture

was stirred under 1 bar of H₂for 80 min and filtered over

cotton. The filtration residue was washed with 2 × 10 mL of

MeOH. The combined filtrates were evaporated, and the

residue was taken up in 10 mL of HPLC-grade H₂O. The

solution was filtered and lyophilized to yield 32.4 mg (101%)

of 30‚6H₂O as a fluffy, amorphous, off-white solid: [R]_D+70.4,

[R]₅₄₆+84.4 (MeOH, c 2.2 g L^-1); (ref 4c: [R]_D+75.2, acetone,

c 8.7 g L^-1; ref 4d: [R]₅₇₈+90, MeOH, c 2 g L^-1; ref 6: [R]_D

+76.4, acetone, c 8.6 g L^-1; ref 19b: [R]₅₇₈+92, H₂O, c 1.9 g

L^-1; ref 19k: [R]_D+80, MeOH, c 1.6 g L^-1); ¹³C NMR (CD₃OD,

TMS; δ 60-85 region only) δ 79.73, 77.08, 73.47, 72.94, 66.84;

MS (API/ES) m/z 865.0 (calcd for [M - H]^-865.2). Anal. Calcd

for C₄₅H₃₈O₁₈‚6H₂O: C, 55.44; H, 5.17. Found: C, 55.71; H,

5.07.

35.32, 35.26, 35.08, 34.94, 28.43. Anal. Calcd for C₂₁₅H₁₈₂O₃₀

C, 79.56; H, 5.65. Found: C, 79.31; H, 5.68.

:

H exa k is(5,7,3′,4′-t et r a -O-b en zyl)ep ica t ech in (4â,8)₅-

Hexa m er (26). Reaction of 486 mg (117 µmol) of 19 with 1.5

mL (2.3 mmol) of 40% aqueous tetra-n-butylammonium hy-

droxide in 8 mL of THF for 101 h as described for the trimer

yielded 455 mg (100%) of 26 as a colorless glass which was

further purified by preparative HPLC as described for com-

pound 25: t_R35.7 min; [R]_D+102, [R]₅₄₆+123 (EtOAc, c 12.2

g L^-1); ¹H NMR (CDCl₃; apparent multiplicities; weak signals

of minor rotamers omitted) δ 7.52 (d, J ) 7 Hz), 7.46-6.63,

6.61 (s), 6.52 (s), 6.50 (s), 6.44-6.38, 6.36 (s), 6.34 (s), 6.31 (s),

6.23 (s), 6.07 (s), 6.04 (s), 5.96 (s), 5.93 (s), 5.91 (s), 5.90 (s),

5.87 (d, J ) 2 Hz), 5.84 (s), 5.76 (s), 5.75 (s), 5.63 (d, J ) 2

Hz), 5.49 (s), 5.47 (s), 5.44 (s), 5.41 (s), 5.40 (s), 5.37 (s), 5.27

(s), 5.22-3.87, 3.66 (dd, J ) 1.5, 4.5 Hz), 2.96-2.82, 1.80 (d, J

) 6 Hz), 1.59 (d, J ) 6 Hz), 1.40 (d, J ) 5 Hz), 1.34 (d, J ) 7

Hz), 1.32-1.20, 1.13 (d, J ) 8 Hz), 1.11 (d, J ) 8 Hz); ¹³C NMR

(CDCl₃, TMS; only intense signals below δ 145 and above δ

40) δ 158.26, 158.03, 157.99, 157.31, 156.66, 156.56, 156.36,

156.20-155.88, 155.83, 155.70, 155.39, 155.37, 155.13, 154.78,

153.25, 153.22, 153.14, 153.10, 153.02, 152.96, 152.74, 149.20,

148.98, 148.78, 148.73, 148.66, 148.50, 148.25, 148.01, 147.93,

147.81, 147.76, 147.73, 147.63, 37.01, 36.29, 35.69, 35.50-

35.00, 28.42. Anal. Calcd for C₂₅₈H₂₁₈O₃₆: C, 79.57; H, 5.64.

Found: C, 79.59; H, 5.57.

Hep ta k is(5,7,3′,4′-tetr a -O-ben zyl)ep ica tech in (4â,8)₆-

Hep ta m er (27). Reaction of 126 mg (26.1 µmol) of 20 with

0.34 mL (0.52 mmol) of 40% aqueous tetra-n-butylammonium

hydroxide in 1.8 mL of THF for 94 h as described for the trimer

yielded 118 mg (100%) of the product as a colorless foam: [R]_D

+105, [R]₅₄₆+127 (EtOAc, c 8.0 g L^-1); ¹H NMR (CDCl₃;

apparent multiplicities; selection) δ 6.53 (s), 6.51 (s), 6.43-

6.36, 6.33 (s), 6.31 (s), 6.28-6.14, 6.07 (s), 6.05-6.02, 5.95 (s),

5.92(s), 5.90-5.85, 5.76 (s), 5.75 (s), 5.63 (d, J ) 2 Hz), 5.48

(s), 5.40 (s), 5.38 (s), 5.27 (s), 2.95-2.82 (m), 1.81 (d, J ) 6

Hz), 1.62 (d, J ) 6 Hz), 1.45-1.40, 1.36 (d, J ) 7 Hz), 1.33-

1.20, 1.15 (d, J ) 8 Hz); ¹³C NMR (CDCl₃, TMS; only intense

signals below δ 145 and above δ 40) δ 158.29, 158.05, 158.02,

Ep ica tech in (4â,8)₃-Tetr a m er (31). To a solution of 56 mg

(21.6 µmol) of 6 in 4 mL of THF were added 4 mL of MeOH,

0.2 mL of H₂O, and 47 mg of 20% Pd(OH)₂/C. The mixture

was stirred under 1 bar of H₂for 75 min and filtered over

cotton. The filtration residue was washed with 2 × 5 mL of

MeOH. The combined filtrates were diluted with 5 mL of

HPLC-grade H₂O and partially evaporated to remove organic

solvents. After dilution with another 10 mL of HPLC-grade

H₂O, the solution was filtered and lyophilized to yield 24.4 mg

(89%) of 31‚6H₂O as a fluffy, amorphous, off-white solid: [R]_D

+93.3, [R]₅₄₆+114 (MeOH, c 9.3 g L^-1) (ref 4d: [R]₅₇₈+73.2,

MeOH, c 3.7 g L^-1; ref 4j: [R]_D+59.8, acetone, c 12 g L^-1; ref

6: [R]_D+109.5, acetone, c 12.3 g L^-1; ref 19i: [R]_D+89.2,

acetone, c 9 g L^-1; ref 19l: [R]_D+81, MeOH, c 1.1 g L^-1); MS

(API/ES) m/z 1153.3 (calcd for [M-H]^-: 1153.3). Anal. Calcd

for C₆₀H₅₀O₂₄‚6H₂O: C, 56.96; H, 5.10. Found: C, 56.98; H,

4.83.

Ep ica tech in (4â,8)₄-P en ta m er (32). To a solution of 76

mg (23.4 µmol) of 25 in 4 mL THF were added 4 mL of MeOH,

0.2 mL of H₂O, and 60 mg of 20% Pd(OH)₂/C. The mixture

was stirred under 1 bar of H₂for 2 h and filtered over cotton.

The filtration residue was washed with MeOH, and the

combined filtrates were partially evaporated to remove organic

solvents. The residue was diluted with 10 mL of HPLC-grade

H₂O, filtered, and lyophilized to produce 34.8 mg of the title

compound as a fluffy, amorphous, off-white solid: [R]_D+116,

[R]₅₄₆+140 (MeOH, c 8.3 g L^-1) (ref 4d: [R]₅₇₈+96, MeOH, c

1 g L^-1; ref 19i: [R]_D+102.1, acetone, c 10 g L^-1; ref 19l: [R]_D

+102, MeOH, c 1.2 g L^-1); MS (API/ES) m/z 1441.4 (calcd for

[M - H]^-1441.3). Anal. Calcd for C₇₅H₆₂O₃₀‚7.5H₂O: C, 57.07;

H, 4.92. Found: C, 56.99; H, 4.79.

Ep ica tech in (4â,8)₅-Hexa m er (33). To a solution of 92.3

mg (23.7 µmol) of 26 in 8 mL of THF were added 8 mL of

MeOH, 0.4 mL of H₂O, and 169 mg of 20% Pd(OH)₂/C. The

mixture was stirred under 1 bar of H₂for 50 min and filtered

1656 J . Org. Chem., Vol. 68, No. 5, 2003

Studies in Polyphenol Chemistry and Bioactivity

over cotton. The filtration residue was washed with MeOH,

and the combined filtrates were partially evaporated after

addition of 10 mL of HPLC-grade H₂O. The residue was diluted

with another 20 mL of HPLC-grade H₂O, filtered, and lyoph-

ilized to produce 47.4 mg of the title compound as a fluffy,

amorphous, off-white solid: [R]_D+123, [R]₅₄₆+149 (MeOH, c

8.6 g L^-1); MS (API/ES) m/z 1730.3 (calcd for [M - H]^-1729.4).

Anal. Calcd for C₉₀H₇₄O₃₆‚9.2H₂O: C, 56.98; H, 4.91. Found:

C, 56.89; H, 4.61.

Ep ica tech in (4â,8)₆-Hep ta m er (34). To a solution of 87.5

mg (19.3 µmol) of 27 in 8 mL of THF were added 8 mL of

MeOH, 0.4 mL of H₂O, and 111 mg of 20% Pd(OH)₂/C. The

mixture was stirred under 1 bar of H₂for 1 h and filtered over

cotton. The filtration residue was washed with MeOH, and the

combined filtrates were partially evaporated after addition of

10 mL of HPLC-grade H₂O. The residue was diluted with

another 10 mL of HPLC-grade H₂O, filtered, and lyophilized

to produce 39.3 mg of the title compound as a fluffy, amor-

phous, off-white solid: [R]_D+134, [R]₅₄₆+164 (MeOH, c 9.6 g

L^-1). Anal. Calcd for C₁₀₅H₈₆O₄₂‚10H₂O: C, 57.33; H, 4.86.

Found: C, 57.49; H, 4.80

Ep ica tech in (4â,8)₇-Octa m er (35). To a solution of 35.7

mg (6.88 µmol) of 28 in 3 mL of THF were added 3 mL of

MeOH, 0.15 mL of H₂O, and 57 mg of 20% Pd(OH)₂/C. The

mixture was stirred under 1 bar of H₂for 55 min and filtered

over cotton. The filtration residue was washed with MeOH,

and the combined filtrates were partially evaporated after

addition of 10 mL of HPLC-grade H₂O. The residue was diluted

with another 10 mL of HPLC-grade H₂O, filtered, and lyoph-

ilized to produce 17.1 mg of the title compound as a fluffy,

amorphous, off-white solid: [R]_D+148, [R]₅₄₆+180 (MeOH, c

5.2 g L^-1). Anal. Calcd for C₁₂₀H₉₈O₄₈‚10.7H₂O: C, 57.66; H,

4.77. Found: C, 57.68; H, 4.79.

combined organic phases were washed with 30 mL of H₂O and

20 mL of saturated aqueous NaHCO₃solution and dried over

MgSO₄. The evaporation residue was prepurified by filtration

over SiO₂with EtOAc and then subjected to preparative HPLC

(column D; 0-45 min, 40 to 80% CH₃CN in H₂O); t_R30.0 min.

Evaporation and drying in vacuo yielded 40.6 mg (52%) of the

peracetate as a glass: [R]_D+82.2, [R]₅₄₆+101 (EtOAc, c 18.7

g L^-1); ¹H NMR (CDCl₃; two rotamers, ratio 3:2; acetate signals

only) δ 2.38, 2.32, 2.29, 2.28-2.21, 2.18, 2.17, 2.04, 1.991 (×

2), 1.988, 1.96, 1.94, 1.91, 1.88, 1.87 (× 2), 1.86, 1.85, 1.79,

spectrum has been reported: ref 6); ¹³C NMR (CDCl₃, TMS;

only signals below δ 145 and above δ 40) δ 170.12, 169.91,

169.13, 169.01, 168.89, 168.79, 168.69, 168.50, 168.40, 168.36,

168.30, 168.28, 168.16, 168.10, 168.08-168.00, 167.98, 167.94,

167.89, 167.87-167.77, 167.74, 167.71, 167.59, 155.84, 154.89,

154.05, 151.88, 151.85, 151.81, 151.65, 151.56, 149.91, 149.87,

149.83, 148.93, 148.72, 148.59, 148.52, 148.46, 147.99, 147.91,

147.44, 147.35, 147.25, 147.13, 35.70, 35.25, 35.17, 34.25,

32.91, 26.33, 21.18, 21.03, 20.94, 20.76, 20.74, 20.72-20.50,

20.45, 20.42, 20.30, 20.18, 20.16, 20.09, 20.03, 19.97, 19.65,

19.58, 19.50, 19.32.

1

1.77, 1.64, 1.52, 1.44, 1.433, 1.428, 1.35 (The complete H NMR

P er -O-a cetylep ica tech in (4â,8)₄-P en ta m er (38). A mix-

ture of 0.25 mL of pyridine and 0.25 mL of acetic anhydride

was added to 47.4 mg (32.8 µmol) of 32. The reaction mixture

was kept in a closed flask at room temperature for 50 h, and

then 1.5 mL of CH₂Cl₂was added followed by 0.5 mL of MeOH.

After another 1 h at room temperature, 10 mL of 0.5 M

H₃PO₄was added, and the product was extracted into 2 × 10

mL of EtOAc. The combined organic phases were washed with

10 mL of H₂O and 10 mL of saturated aqueous NaHCO₃

solution and dried over MgSO₄. The evaporation residue was

prepurified by filtration over SiO₂with EtOAc and then

subjected to preparative HPLC (column B; 0-30 min, 40-80%

CH₃CN in H₂O, then 80%); t_R30.5 min. Evaporation and

drying in vacuo yielded 37.6 mg (46%) of the peracetate as a

P er -O-a cetylep ica tech in (4â,8)₂-Tr im er (36). A mixture

of 0.25 mL of pyridine and 0.25 mL of acetic anhydride was

added to 30.6 mg (35.3 µmol) of 30. The reaction mixture was

stirred at room temperature for 27 h, 20 mL of saturated

aqueous NaHCO₃solution was added, and the product was

extracted into 3 × 10 mL of EtOAc. The organic layers were

washed with 30 mL of H₂O, 20 mL of 0.5 M H₃PO₄, 10 mL of

H₂O, and 10 mL of saturated aqueous NaHCO₃solution, and

dried over MgSO₄. After filtration over a short SiO₂column

with EtOAc/hexane 85:15, the crude product was purified by

preparative HPLC (column E, EtOAc/hexane 85:15); t_R12.2

min. Evaporation and drying in vacuo yielded 28.6 mg (54%)

of the peracetate as a colorless glass: [R]_D+57.0, [R]₅₄₆+69.8

(EtOAc, c 9.2 g L^-1); [R]_D+55.3, [R]₅₄₆+67.7 (acetone, c 10.4

g L^-1) (ref 19b: [R]₅₇₈+51.6, acetone, c 3.8 g L^-1); ¹H NMR

(CDCl₃; two rotamers, ratio 2:1; acetate signals only) δ 2.38,

2.32 (× 2), 2.30, 2.30-2.25, 2.23, 2.218 (× 2), 2.215 (× 2), 2.164,

2.158, 2.02 (× 2), 1.95 (× 3), 1.89, 1.88 (× 4), 1.82 (× 2), 1.79

1

glass: [R]_D+106, [R]₅₄₆+129 (EtOAc, c 11.3 g L^-1); H NMR

(CDCl₃; two rotamers, ratio 3:2; acetate signals only) δ 2.38,

2.32, 2.31, 2.29-2.21, 2.18, 2.17, 2.04, 2.00, 1.99-1.96, 1.94,

1.92, 1.90, 1.88, 1.87, 1.83, 1.80, 1.79, 1.78, 1.70, 1.53, 1.43,

1.42, 1.41, 1.39, 1.38, 1.33; ¹³C NMR (CDCl₃, TMS; only signals

below δ 145 and above δ 40) δ 170.12, 170.08, 169.93, 169.15,

169.00, 168.96, 168.85, 168.70, 168.60-168.44, 168.40, 168.32,

168.26, 168.18, 168.12, 168.10-167.98, 167.93-167.79, 167.72,

167.66, 167.63, 155.86, 154.89, 154.05, 151.90-151.74, 151.72-

151.62, 149.91, 149.87, 149.82, 148.90, 148.84, 148.76-148.42,

147.99, 147.93, 147.87, 147.48, 147.41, 147.36, 147.27, 147.22,

147.16, 147.02, 146.96, 35.56, 35.41, 35.11, 34.24, 32.91, 26.35,

21.18, 21.04, 20.96, 20.76, 20.70-20.52, 20.48, 20.43, 20.32-

20.02, 19.69, 19.51, 19.34, 19.28. Anal. Calcd for C₁₂₅H₁₁₂O₅₅

C, 60.19; H, 4.53. Found: C, 60.07; H, 4.35.

:

1

(× 2), 1.71, 1.55 (× 2), 1.50, 1.38 (× 2) (The complete H NMR

P er -O-a cetylep ica tech in (4â,8)₅-Hexa m er (39). A mix-

ture of 0.27 mL of pyridine and 0.27 mL of acetic anhydride

was added to 39.5 mg of 33. The reaction mixture was kept in

a closed flask at room temperature for 52 h, and then 2 mL of

EtOAc was added followed by 1 mL of MeOH. After another

80 min at room temperature, 20 mL of 0.5 M H₃PO₄was added,

and the product was extracted into 3 × 10 mL of EtOAc. The

combined organic phases were washed with 20 mL of H₂O and

10 mL of saturated aqueous NaHCO₃solution. The evapora-

tion residue was prepurified by filtration over SiO₂with EtOAc

and then subjected to preparative HPLC (column D; 0-30 min,

40-80% CH₃CN in H₂O, then 80%); t_R32.5 min. Evaporation

and drying in vacuo yielded 34.4 mg (58% over 2 steps) of the

peracetate as a glass: [R]_D+126, [R]₅₄₆+153 (EtOAc, c 5.5 g

L^-1); ¹H NMR (CDCl₃; two rotamers, ratio 3:2; acetate signals

only) δ 2.38, 2.32, 2.30-2.26, 2.25, 2.24, 2.235-2.21, 2.18, 2.17,

2.04, 2.01, 1.99, 1.98-1.97, 1.96, 1.95, 1.93, 1.925, 1.895, 1.89,

1.88, 1.875, 1.86, 1.835, 1.83, 1.785, 1.78, 1.71, 1.435, 1.43, 1.38,

1.37, 1.365; ¹³C NMR (CDCl₃, TMS; only signals below δ 145

and above δ 40) δ 170.12, 170.08, 169.95, 169.14, 169.01,

168.96, 168.90, 168.82, 168.71, 168.59, 168.54, 168.47, 168.39,

spectrum has been reported: refs 6, 19f); ¹³C NMR (CDCl₃,

TMS; only signals below δ 145 and above δ 40) δ 170.10,

170.06, 169.94, 169.07, 169.00, 168.80, 168.77, 168.68, 168.54,

168.40, 168.35, 168.29, 168.14, 168.07, 168.01, 167.99, 167.94,

167.90, 167.87, 167.80, 167.76, 167.68, 167.61, 167.57, 167.43,

155.76, 154.88, 154.06, 151.86, 151.84, 151.73, 149.88, 149.84,

149.78, 148.79, 148.58, 148.49, 148.29, 147.96, 147.58, 147.17,

146.89, 35.61, 35.04, 34.36, 32.96, 26.70, 26.41, 21.16, 21.05,

20.97, 20.76, 20.72-20.56, 20.54, 20.40, 20.26, 20.21, 20.13,

20.11, 20.05, 19.78, 19.72, 19.48. Anal. Calcd for C₇₅H₆₈O₃₃

C, 60.16; H, 4.58. Found: C, 59.80; H, 4.41.

:

P er -O-a cetylep ica tech in (4â,8)₃-Tetr a m er (37). A 102

mg (39.2 µmol) sample of 6 was deprotected as described above.

The crude product solution was evaporated to dryness, and a

mixture of 0.3 mL of pyridine and 0.3 mL of acetic anhydride

was added to the residue. The reaction mixture was kept in a

closed flask at room temperature for 41 h, and then 1 mL of

CH₂Cl₂was added followed by 1 mL of MeOH. After another

80 min at room temperature, 20 mL of 0.5 M H₃PO₄was added,

and the product was extracted into 3 × 20 mL of EtOAc. The

J . Org. Chem, Vol. 68, No. 5, 2003 1657

Kozikowski et al.

168.32, 168.25, 168.20, 168.12, 168.06, 168.01, 167.95, 167.87,

167.82, 167.77-167.67, 167.63, 155.86, 154.90, 154.06, 151.88,

151.82, 151.79, 151.76, 151.66, 149.90, 149.86, 149.82, 148.86,

148.75-148.68, 148.65, 148.60, 148.51, 147.98, 147.93, 147.47,

147.43, 147.28, 147.05, 35.6-35.0, 34.25, 32.91, 26.36, 21.18,

21.05, 20.97, 20.78, 20.74-20.52, 20.48, 20.43, 20.29, 20.24-

20.02, 19.68, 19.57-19.50, 19.35, 19.29. Anal. Calcd for

microculture tetrazolium assay²⁸modified for use with crystal

violet rather than MTT. Briefly, (1-2) × 10³cells were added

per well and allowed to culture in a humidified, 5% CO₂

atmosphere until they reached approximately 50% confluence.

Sterile filtered test compounds were added at various concen-

trations, and the plates were allowed to culture for an

additional 12-36 h. The growth medium was then removed,

and each well was washed twice with 200 µL each of pH 7.4

PBS. After washing, 50 µL of filtered crystal violet solution

(2.5 g/125 mL of methanol + 375 mL of H₂O) was added. At

the end of 5 min, the crystal violet was removed, and the plate

was washed three times with H₂O. Plates were allowed to dry,

and the crystal violet stained cells were resolubilized in 100

µL of 0.1 M sodium citrate in ethanol/H₂O (1:1, v/v). At the

end of 1 h, the plates were scanned at 570 nm (ref 405 nm)

with a microtiter plate reader, and the data were recorded with

the SOFTMAX software program. The average of three read-

ings was taken for each blank, control, vehicle, and test

concentration for statistical data manipulation.

F low Cytom etr y. MBA MB 231 cells were cultured as

described above until they reached approximately 50% con-

fluence. Sterile filtered test compound or catalase adjusted to

100 U/mL or heat-inactivated catalase (solution immersed in

boiling water for 15 min) or H₂O₂was then added, and the

cells were allowed to culture for an additional 24 h. The cells

were then trypsinized and counted, and 1.5 × 10⁶cells were

taken for cell cycle analysis by the Vindelov method.²⁹Analyses

were performed by the Lombardi Cancer Center Flow Cytom-

etry Core Facility at Georgetown University Medical Center.

An n exin V-F ITC. The annexin V-FITC assay was per-

formed on procyanidin-treated MDA MB 231 cells using the

TACS Annexin V-FITC kit according to the manufacturer’s

procedure.

C

₁₅₀H₁₃₄O₆₆: C, 60.20; H, 4.51. Found: C, 60.44; H, 4.20.

P er -O-a cetylep ica tech in (4â,8)₆-Hep ta m er (40). A mix-

ture of 0.25 mL of pyridine and 0.25 mL of acetic anhydride

was added to 34.8 mg of 34. The reaction mixture was kept in

a closed flask at room temperature for 44 h, and then 1 mL of

EtOAc was added followed by 0.5 mL of MeOH. After another

1.5 h at room temperature, 20 mL of 0.5 M H₃PO₄was added,

and the product was extracted into 3 × 10 mL of EtOAc. The

combined organic phases were washed with 20 mL of H₂O and

10 mL of saturated aqueous NaHCO₃solution. The evapora-

tion residue was prepurified by filtration over SiO₂with EtOAc

and then subjected to preparative HPLC (column D; 0-30 min,

40-80% CH₃CN in H₂O, then 80%); t_R29.6 min. Evaporation

and drying in vacuo yielded 24.6 mg (41% over two steps) of

the peracetate as a glass: [R]_D+140, [R]₅₄₆+170 (EtOAc, c

1

11.4 g L^-1); H NMR (CDCl₃; two rotamers, ratio 3:2; acetate

signals only) δ 2.38, 2.32, 2.30-2.22, 2.18, 2.17, 2.04, 2.00,

1.99-1.96, 1.93, 1.92, 1.91-1.87, 1.83, 1.825, 1.785, 1.78, 1.71,

1.435, 1.43, 1.39, 1.38, 1.37, 1.36, 1.34, 1.33.

P er -O-a cetylep ica tech in (4â,8)₇-Octa m er (41). A mix-

ture of 0.1 mL of pyridine and 0.1 mL of acetic anhydride was

added to 14.5 mg of 35. The reaction mixture was kept in a

closed flask at room temperature for 44 h, and then 0.5 mL of

EtOAc was added followed by 0.25 mL of MeOH. After another

70 min at room temperature, the solution was evaporated, and

the evaporation residue was prepurified by filtration with

EtOAc over a short SiO₂column to the top of which a small

amount of p-toluenesulfonic acid had been applied (as a

solution in EtOAc + a minimum volume of MeOH). The crude

product was subjected to preparative HPLC (column D; 0-30

min, 40 to 80% CH₃CN in H₂O, then 80%); t_R33.6 min.

Evaporation and drying in vacuo yielded 7.9 mg (34% over 2

steps) of the peracetate as a glass: [R]_D+150, [R]₅₄₆+184

(EtOAc, c 4.0 g L^-1); ¹H NMR (CDCl₃; two rotamers, ratio 3:2;

acetate signals only) δ 2.38, 2.32, 2.30-2.21, 2.18, 2.17, 2.04,

2.01-1.96, 1.93, 1.92, 1.91-1.87, 1.83, 1.825, 1.79, 1.78, 1.71,

1.53, 1.44-1.42, 1.39, 1.38, 1.37, 1.36, 1.35, 1.33.

Cell Lin es. The human breast cancer cell lines MCF-7,

SKBR-3, MDA 435, and MDA MB 231 were obtained from the

Lombardi Cancer Center Cell Culture Core Facility at Geor-

getown University Medical Center. The MDA MB 231 cell line

is P53 defective, ER negative, and constitutively expresses

K-ras. Cells were cultured in T-75s in IMEM medium supple-

mented with 10% FBS in a humidified 5% CO₂atmosphere at

37 °C.

Ack n ow led gm en t. We are grateful to Mars, Inc.,

for funding this project. We are also grateful to Karen

Creswell and Owen Blair of the Lombardi Cancer

Research Center at Georgetown University Medical

Center for performing the flow cytometry analyses and

to J immy Rae and Michael J ohnson for assistance with

cell cultures. We also are grateful to Sheryl Lazarus of

Mars, Inc., for performing the HPLC-MS analyses and

to Eleanor Eggert and Dave Zuber for preparing the

cover art.

Su p p or tin g In for m a tion Ava ila ble: Preparation/char-

acterization of compounds 2, 5, 9, 10b, 11, 14, 15, 42-44, and

46-52; IR spectra; selections from the ¹H NMR spectra of

compounds 37-41; selections from the ¹³C NMR spectra of

compounds 36-39 (pdf). This material is available free of

charge via the Internet at http://pubs.acs.org.

J O020393F

(28) Carmichael, J .; DeGraff, W. G.; Gazdar, A. F.; Minna, J . D.;

Mitchell, J . B. Cancer Res. 1987, 47, 936.

(29) Vindelov, L. L.; Christensen, I. J .; Nissen, N. I. Cytometry 1983,

3, 323.

Cytotoxicity Assa y. Cytotoxicity assays were performed

on several human breast cancer cell lines treated with test

compounds in a 96-well microtiter plate format using the

1658 J . Org. Chem., Vol. 68, No. 5, 2003

Article Doi

Studies in polyphenol chemistry and bioactivity. 4. Synthesis of trimeric, tetrameric, pentameric, and higher oligomeric epicatechin-derived procyanidins having all-4β,8-interflavan connectivity and their inhibition of cancer cell growth through cell cycle arrest

DOI: 10.1021/jo020393f

Source and publish data:

Authors:

Article abstract of DOI:10.1021/jo020393f

Full text of DOI:10.1021/jo020393f

Products guided by the article

R&D Labs maybe for 99297-47-9

Relevant to this article

Hot Product