Synthesis, crystal structure, and growth inhibition of human hepatoma cell (HepG2) of polyphenolic compounds based on gallates

Article abstract of DOI:10.1139/V07-107

Seven compounds (1-7) based on gallate were synthesized and characterized by elemental analysis, ¹H NMR, and MS spectra. 2-(3,5-Dibenzyloxy-4- methoxy)phenyl-2-propanol (6) was a new compound. Methyl 3,5-dihydroxy-4- methoxybenzoate (3), methyl 3,5-dibenzyloxy-4-methoxybenzoate (4), 3,5-dibenzyloxy-4-methoxybenzyl alcohol (5), and compound 6 were structurally determined by single-crystal X-ray diffraction for the first time. Crystallography data for 3: space group P2₁2₁2 ₁; a = 4.0750(8) A, b = 7.5880(15) A, c = 29.802(6) A; V = 921.5(3) A³ Z = 4. 4: space group P-1; a = 10.068(2) A b = 10.499(2) A, c = 11.388(2) A; α = 76.84(3)°, β= 66.79(3)°, γ = 64.10(3)°; V = 993.0(3) A³, Z = 2. 5: space group P-1; a = 8.1410(16) A, b = 8.7590(18) A, c = 12.879(3) A; α = 91.66(3)°, β= 94.69(3)°, γ = 91.73(3)°; V = 914.4(3) A³; Z = 2. 6: space group P2₁/c; a = 5.8100(12) A, b = 15.778(3) A, c = 23.237(5) A; β= 96.09(3)°; V = 2118.1(7) A³; Z = 4. All of the seven compounds were evaluated for the inhibition of growth of human hepatoma (HepG2) cells. Comparison with the positive-control 5-fluorouracil (IC₅₀ 51.6 μmol/L), 5 showed stronger cytotoxic activity with an IC₅₀ around 15.3 μmol/L, while IC₅₀ value of 3 was 90.3 μmol/L. The effect of slight structural variations in this series of compounds was found to cause a marked change in their activity against HepG2 cells.

Full text of DOI:10.1139/V07-107

951
Synthesis, crystal structure, and growth inhibition
of human hepatoma cell (HepG2) of polyphenolic
compounds based on gallates
Zhu-Ping Xiao, Rui-Qin Fang, Lei Shi, Hui Ding, Chen Xu, and Hai-Liang Zhu
1
Abstract: Seven compounds (1–7) based on gallate were synthesized and characterized by elemental analysis, H
NMR, and MS spectra. 2-(3,5-Dibenzyloxy-4-methoxy)phenyl-2-propanol (6) was a new compound. Methyl 3,5-
dihydroxy-4-methoxybenzoate (3), methyl 3,5-dibenzyloxy-4-methoxybenzoate (4), 3,5-dibenzyloxy-4-methoxybenzyl al-
cohol (5), and compound 6 were structurally determined by single-crystal X-ray diffraction for the first time. Crystallo-
graphic data for 3: space group P212121; a = 4.0750(8) Å, b = 7.5880(15) Å, c = 29.802(6) Å; V = 921.5(3) Å3; Z =
4. 4: space group P-1; a = 10.068(2) Å, b = 10.499(2) Å, c = 11.388(2) Å; α = 76.84(3)°, β = 66.79(3)°, γ =
64.10(3)°; V = 993.0(3) Å3, Z = 2. 5: space group P-1; a = 8.1410(16) Å, b = 8.7590(18) Å, c = 12.879(3) Å; α =
91.66(3)°, β = 94.69(3)°, γ = 91.73(3)°; V = 914.4(3) Å3; Z = 2. 6: space group P21/c; a = 5.8100(12) Å, b =
15.778(3) Å, c = 23.237(5) Å; β = 96.09(3)°; V = 2118.1(7) Å3; Z = 4. All of the seven compounds were evaluated for
the inhibition of growth of human hepatoma (HepG2) cells. Comparison with the positive-control 5-fluorouracil
(IC50 51.6 µmol/L), 5 showed stronger cytotoxic activity with an IC50 around 15.3 µmol/L, while IC50 value of 3 was
90.3 µmol/L. The effect of slight structural variations in this series of compounds was found to cause a marked change
in their activity against HepG2 cells.
Key words: gallates, benzyl alcohol, human hepatoma cells, crystal structure, anti-cancer activities.
Résumé : On a réalisé la synthèse de sept composés apparentés au gallate (1–7) dont le 2-(3,5-dibenzoyloxy-4-
méthoxy)phénylpropan-2-ol (6) est un nouveau produit, et on a les a caractérisés par leur analyse élémentaire, leur
spectre de RMN et leur spectre de masse. Faisant appel à la diffraction des rayons X par un cristal unique, on a déter-
miné pour la première fois les structures du 3,5-dihydroxy-4-méthoxybenzoate de méthyle (3), du 3,5-dibenzyloxy-4-
méthoxybenzoate de méthyle (4), de l’alcool 3,5-dibenzyloxy-4-méthoxybenzylique (5) et du composé 6. Les données
cristallographiques sont les suivantes: pour le composé 3, groupe d’espace P212121 avec a = 4,0705(8) Å, b =
7,5880(15) Å et c = 29,802(6) Å, V = 921,5(3) Å3 et Z = 4; pour le composé 4, groupe d’espace P-1 avec a =
10,068(2) Å, b = 10,499(2) Å et c = 11,388(2) Å, α = 76,84(3)°, β = 66,79(3)° et γ = 64,10(3)°, V = 993,0(3) Å3 et
Z = 2; pour le composé 5, groupe d’espace P-1 avec a = 8,1410(16) Å, b = 8,7590(18) Å et c = 12,879(3) Å, α =
91,66(3)°, β = 94,69(3)° et γ = 91,73(3)°, V = 914,4(3) Å3 et Z = 2; pour le composé 6, groupe d’espace P21/c avec
a = 5, 8100(12) Å, b = 15,778(3) Å et c = 23,237(5) Å, β = 96,09(3)°, V = 2118,1(7) Å3 et Z = 4. On a évalué les ca-
pacités inhibitrices des sept composés vis-à-vis de la croissance de cellules d’hépatome humain (HepG2). Par compa-
raison avec le produit de contrôle positif, 5-fluorouracil (IC50 = 51,6 µmol/L), le composé 5 présente une activité
cytotoxique plus importante (IC50 = 15,3 µmol/L), alors que pour le composé 3, la valeur de IC50 = 90,3 µmol/L. On a
observé que de faibles changements de structures dans cette série de composés provoque dans changements importants
dans leur activité contre les cellules HepG2.
Mots-clés : gallates, alcool benzylique, cellules d’hépatome humain, structure cristalline, activités anticancérigènes.
[Traduit par la Rédaction] Xiao et al. 957
Introduction
Gallic acid and its derivatives are widely present in the
plant kingdom and represent a large family of secondary
plant polyphenolic metabolites. Toxicological studies
Received 16 November 2006. Accepted 20 August 2007.
Published on the NRC Research Press Web site at
canjchem.nrc.ca on 3 October 2007.
showed that the no-observed-adverse-effect-level (NOAEL)
of gallic acid is at least 120 mg kg–1 day–1 for F334 rats, and
Z.-P. Xiao,1 R.-Q. Fang, L. Shi, H. Ding, C. Xu, and
H.-L. Zhua.2 Institute of Functional Biomolecules, State Key
Laboratory of Pharmaceutical Biotechnology, Nanjing
University, Nanjing 210093, P.R. China
the level of gallates was reported to be as high as
1000 mg kg–1 for mice (1, 2).
Gallic acid and its alkyl esters derivatives are added to
foods to prevent oxygen-induced lipid peroxidation (3). The
alkyl gallates can inhibit 7,12-dimethylbenz[a]anthracene-
initiated mammary tumorigenesis (4, 5) and decrease
benzo[a]pyrene metabolite formation (6) in vivo, and they
1Present address: Department of Chemistry, Shangrao Normal
College, Shangrao 334000, P.R. China.
2Corresponding author (e-mail: zhuhl@nju.edu.cn).
Can. J. Chem. 85: 951–957 (2007)
doi:10.1139/V07-107
© 2007 NRC Canada

952
Can. J. Chem. Vol. 85, 2007
protect against hydroperoxide-induced cytotoxicity (7) and
mutagenicity (8) in vitro. The ability of the alkyl gallates to
induce apoptosis in several tumor-cell lines has recently at-
tracted much interest (9, 10). At the cellular level, they have
been shown to inhibit cell adhesion (11) and induce
apoptosis of tumor cells (12). Toshihiko Osawa found that
alkyl gallates suppress the 3-methylcholanthrene-mediated
cytochrome P450 1A induction in human hepatoma (HepG2)
cells. The suppressions were dependent on the length of the
alkyl chain of the gallates (13). The cytotoxic activity of
phenolic derivatives depends on their rate of incorporation
into cells, directly related to their lipophilicity as much as to
their antioxidant activity.
The gallates investigated in the literature for their anti-
tumor activity in a human cancer-cell line varied mainly in
the number of phenolic substituents and (or) in the length of
the alkyl side chain. In the present paper, some structurally
distinct derivatives based on gallate were synthesized and
evaluated for the inhibition of growth of liver (HepG2) can-
cer cells. Compounds 4, 5, and 6 showed potent anticancer
activities with IC50 values of 23.1, 15.3, and 20.2, respec-
tively, and crystal structures of 3–6 were described to deter-
mine the effect of the structural characteristics on their
antiproliferative activity.
portional to the number of viable cells (14). To better ex-
plain the inhibitory results, the inhibition of cell population
growth was calculated based on the following formula:
Inhibition percentage = (cell no. in control –
cell no. in treatment) / (cell no. in control –
original cell no. before the compounds were
added) × 100%.
IC50 is the extract concentration under which a 50% inhi-
bition of cell population growth occurs.
Single-crystal analysis
The purified products were dissolved in ethyl acetate, then
petroleum ether was added carefully. The crystals suited for
X-ray analysis were obtained after 3 to 4 days by volatiliza-
tion of the solvents.
The crystal data, data collection, and refinement parame-
ters for compounds 3–6 are listed in Table 1.3 The crystal
data were collected with a Bruker SMART APEX CCD
diffractometer, using graphite monochromatized Mo Kα
radiation (λ = 0.71073 Å) (15). All calculations were per-
formed by using the SHELXTL package and refined with
full-matrix least-square procedures on desktop computers
(16). All non-hydrogen atoms were located from a differ-
ence Fourier map and were placed at calculated positions,
refined isotropically with riding model position parameters.
Experimental
Reagents
Chemistry
MTT cell-proliferation assay kits were purchased from
ATCC. Cell-death detection ELISA kits were purchased
from Roche. The HepG2 cell line was purchased from
ATCC.
The compounds previously reported were synthesized ac-
cording to the methods in the references with appropriate
modifications. Separation of the compounds by column
chromatography was carried out with silica gel 60 (200–300
mesh ASTM, E. Merck). The quantity of silica gel used was
50–100 times the weight charged on the column. The eluates
were monitored using TLC. Melting points (uncorrected)
were determined on a XT4 MP apparatus (Taike Corp.,
Beijing, China). ESI mass spectra were obtained on a Mari-
Cell cultures and cell viability assay
The HepG2 cells were cultured in ATCC minimum essen-
tial medium (Eagle) with 2 mmol/L L-glutamine and Earle
BSS adjusted to contain 1.5 g/L sodium bicarbonate,
0.1 mmol/L non-essential amino acids, 1.0 mmol/L sodium
pyruvate (90%), and 10% fetal bovine serum. Cells were in-
cubated under 37 °C with 5% CO2. The medium was
changed 2–3 times per week.
1
ner System 5304 mass spectrometer, and H NMR spectra
were recorded at DPX500 on a Bruker spectrometer. Ele-
mental analyses were performed on a CHN-O-Rapid instru-
ment and were within 0.4% of the theoretical values.
Compounds 1 to 5 were synthesized according to refs. 17
and 18.
After digestion with trypsin-EDTA, uniform amounts
(~2 × 104) of HepG2 cells in growth media were inoculated
into each well of a 96-well flat-bottomed plate. After 24 h of
incubation at 37 °C in 5% CO2, the growth medium was re-
placed with 100 media, containing different concentrations
of compounds 1 to 7, respectively. DMSO was added ini-
tially to the test compounds to help dissolve the sample. Af-
ter 48 h of incubation, cell population growth was
determined using the ATCC MTT cell proliferation assay at
570–655 nm with a Bio-Rad Model 680 Microplate Reader
(Hercules, California). Briefly, a mitochondrial enzyme in
living cells, succinate dehydrogenase, reduced the yellow
tetrazolium salt (MTT) to insoluble purple formazan crys-
tals. Therefore, the amount of formazan produced was pro-
Methyl 3,4,5-triacetoxybenzoate (1)
Melting point 130 to 131 °C. 1H NMR (CDCl3) δ: 2.29 (s,
9H, C(=O)CH3), 3.89 (s, 3H, OCH3), 7.79 (s, 2H, ArH).
ESI-MS: C14H15O8 [M + H]+ 311. Anal. calcd. for
C14H14O8: C 54.18, H 4.56; found: C 54.20, H 4.55.
Methyl 3,5-diacetoxy-4-methoxybenzoate (2)
1
Melting point 83–83.5 °C. H NMR (CDCl3) δ: 2.32 (s,
6H, C(= O)CH3), 3.87 (s, 6H, OCH3). ESI-MS: C13H15O7
[M + H]+ 283. Anal. calcd. for C13H14O7: C 55.35, H 4.98;
found: C 55.32, H 5.00.
3 Supplementary data for this article are available on the journal Web site (canjchem.nrc.ca) or may be purchased from the Depository of Un-
published Data, Document Delivery, CISTI, National Research Council Canada, Ottawa, ON K1A 0R6, Canada. DUD 5214. For more in-
formation on obtaining material refer to cisti-icist.nrc-cnrc.gc.ca/irm/unpub_e.shtml. CCDC 642103–642106 contain the crystallographic
data for 3–6. These data can be obtained, free of charge, via www.ccdc.cam.ac.uk/conts/retrieving.html (Or from the Cambridge Crystallo-
graphic Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax +44 1223 336033; or deposit@ccdc.cam.ac.uk).
© 2007 NRC Canada

Xiao et al.
Table 1. Crystal data and structure refinement for compounds 3, 4, 5, and 6.
953
Compounds
3
4
5
6
Empirical formula
C9H10O5
C23H22O5
C22H22O4
C24H26O4
Formula weight
Crystal system
Space group
198.17
Orthorhombic
P212121
378.41
Triclinic
P-1
350.40
Triclinic
P-1
378.45
Monoclinic
P21/c
a (Å)
b (Å)
c (Å)
4.0750(8)
7.5880(15)
29.802(6)
90
90
90
921.5(3)
4
1.428
0.118
416
10.068(2)
10.499(2)
11.388(2)
76.84(3)
66.79(3)
64.10(3)
993.0(5)
2
8.1410(16)
8.7590(18)
12.879(3)
91.66(3)
94.69(3)
91.73(3)
914.4(3)
2
5.8100(12)
15.778(3)
23.237(5)
90
96.09(3)
90
2118.1(7)
4
1.187
α (°)
β (°)
γ (°)
Volume (Å3)
Z
Calculated density (Mg/m3)
1.266
0.089
400
1.273
0.087
372
µ (mm–1)
F(000)
0.080
808
Reflections collected
Unique reflections
Rint
2185
1111
0.0472
4126
3894
0.0131
3847
3581
0.0117
4112
3716
0.0431
Data/restraints/parameters
Goodness-of-fit on F2
R1; wR2 [I > 2σ(I)]
1111/0/133
1.069
0.0500, 0.1195
3894/0/256
1.026
0.0580, 0.1412
3581/0/238
1.026
0.0441, 0.1036
3716/19/254
0.975
0.0658, 0.1511
R1; wR2 (all data)
0.0678, 0.1304
0.1009, 0.1674
0.0682, 0.1179
0.1444, 0.1761
Extinction coefficient
0.145(15)
0.345/–0.245
0.049(6)
0.200/–0.209
0.373(13)
0.140/–0.183
0.028(3)
0.188/–0.190
∆ρmax/∆ρmin (eÅ–3)
R1 = Σ*(*Fo* – *Fc*)*/Σ*Fo*. wR2 = [Σw(*Fo* – *Fc*)2/Σw*Fo*2] , w = [σ2(Fo)2 + (0.1(max(0, Fo ) + 2Fc2)/3)2]–1.
½
2
mixture was cooled and poured into an NH4Cl solution. The
ethereal layer was separated, and the aqueous layer was ex-
tracted with ether. The combined ethereal extracts were
washed with solutions of NaHSO3, NaHCO3, and water and
dried over anhyd. Na2SO4. The solvent was removed, and
the residue was purified by flash chromatography with
EtOAc – petroleum ether (1:4) to afford colorless crystals
Methyl 3,5-dihydroxy-4-methoxybenzoate (3)
Melting point 148 to 149 °C. 1H NMR (CDCl3) δ: 3.73 (s,
3H, ArOCH3), 3.77 (s, 3H, OCH3), 6.94 (s, 2H, ArH), 9.51
(s, 2H, OH). ESI-MS: C9H11O5 [M + H]+ 199. Anal. calcd.
for C9H10O5: C 54.52, H 5.10; found: C 54.55, H 5.09.
Methyl 3,5-dibenzyloxy-4-methoxybenzoate (4)
1
Melting point 119 to 120 °C. 1H NMR (CDCl3) δ: 3.88 (s,
3H, ArOCH3), 3.94 (s, 3H, OCH3), 5.16 (s, 4H, ArCH2),
7.33 (t, 2H, J = 7.0 Hz, benzyl H-4′), 7.31 (s, 2H, H-2, 6),
7.39 (t, 4H, J = 7.3 Hz, benzyl H-3′, 5′), 7.46 (d, 4H, benzyl
H-2′, 6′). ESI-MS: C23H23O5 [M + H]+ 379. Anal. calcd. for
C23H22O5: C 73.04, H 5.84; found: C 73.00, H 5.86.
(3.45 g, 87%). Melting point 63 to 64 °C. H NMR (CDCl3)
δ: 1.49 (s, 6H, CH3), 2.05 (s, 1H, OH), 3.90 (s, 3H, OCH3),
5.14 (s, 4H, ArCH2), 6.77 (s, 2H, H-2, 6), 7.31 (t, 2H, J =
7.3 Hz, benzyl H-4′), 7.38 (t, 4H, J = 7.4 Hz, benzyl H-3′,
5′), 7.46 (d, 4H, J = 7.4 Hz, benzyl H-2′, 6′). ESI-MS:
C24H25O3 [M – H2O + H]+ 361. Anal. calcd. for C24H26O4:
C 76.09, H 6.94; found: C 76.17, H 6.92.
3,5-Dibenzyloxy-4-methoxybenzyl alcohol (5)
Melting point 105 to 106 °C. 1H NMR (CDCl3) δ: 1.76 (s,
1H, OH), 3.90 (s, 3H, OCH3), 4.54 (s, 2H, HOCH2), 5.13 (s,
4H, ArCH2), 6.64 (s, 2H, H-2, 6), 7.32 (t, 2H, J = 7.3 Hz,
benzyl H-4′), 7.38 (t, 4H, J = 7.3 Hz, benzyl H-3′, 5′), 7.45
(d, 4H, J = 7.5 Hz, benzyl H-2′, 6′). ESI-MS: C22H23O4
[M + H]+ 351. Anal. calcd. for C22H22O4: C 75.45, H 6.30;
found: C 75.41, H 6.33.
3,5-Dibenzyloxy-4-methoxyphenol (7)
Boron trifluoride etherate (83 mmol, 10 mL) was added to
a suspension of sodium perborate (15.8 mmol, 2.7 g) in dry
tetrahydrofuran (72 mL) contained in a nitrogen-flushed
flask cooled to 0 °C. The mixture was stirred for 30 min at
room temperature, and then the compound 7 (7.2 mmol,
2.71 g in 20 mL of dry tetrahydrofuran) was added. After
stirring for 13 h, the reaction was quenched with sodium sul-
fite; the product was extracted into ethyl acetate and dried
over MgSO4. Removal of the solvent followed by column
chromatography (EtOAc – petroleum ether, 1:4) produced
compound 7 as colorless crystals (2.25 g, 93%). Melting
2-(3,5-Dibenzyloxy-4-methoxy)phenyl-2-propanol (6)
Methyl iodide (4.26 g, 29.6 mmol) was dissolved in
25 mL of dry ether and the resultant mixture added dropwise
with stirring to magnesium turnings (0.68 g, 28.5 mmol) in
5 mL of dry ether. The mixture was heated to reflux for
0.5 h. A solution of 3,5-dibenzyloxy-4-methoxybenzoate (5;
3.96 g, 10.5 mmol) in 20 mL dry tetrahydrofuran was added
dropwise with stirring. After heating to reflux for 1.5 h, the
1
point 86–88 °C. H NMR (DMSO-d6) δ: 3.64 (s, 3H, OCH3),
5.04 (s, 4H, ArCH2), 6.14 (s, 2H, H-2, 6), 7.33 (t, 2H, J =
6.9 Hz, benzyl H-4′), 7.40 (t, 4H, J = 7.4 Hz, benzyl H-3′, 5′),
7.44 (d, 4H, J = 7.1 Hz, benzyl H-2′, 6′), 9.22 (s, 1H,
© 2007 NRC Canada

954
Can. J. Chem. Vol. 85, 2007
ArOH). ESI-MS: C21H21O4 [M + H]+ 337. Anal. calcd. for
C21H20O4: C 75.02, H 5.97; found: C 74.98, H 5.99.
Scheme 1. (a) Ac2O, Et3N; (b) K2CO3, CH3I; (c) CH3OH/H2O,
K2CO3; (d) DMF, K2CO3, BnCl; (e) THF, LiAlH4; (f ) Et2O,
CH3I, Mg; and (g) THF, NaBO3·4H2O, BF3·Et2O.
CO2Me
CO2Me
Result and discussion
CO2Me
c
a
b
Chemistry
AcO
OAc
AcO
OAc
HO
OH
Gallates are known to display variety of biological func-
tions, including anticancer activity. To investigate the effect
of the ester moieties on their antiproliferative and (or)
cytotoxic activity, we successfully synthesized compounds
5–7 (shown in Scheme 1). Guzmán et al. reported that oxi-
dation of hydroxy and alkoxybenzaldehydes afforded phe-
nols with H2O2 and SeO2 in t-butyl alcohol at room
temperature (19). Thus, we designed a rout to synthesize
compound 7 (shown in Scheme 2). However, compound 8
was obtained in good yield instead of 7 when the route was
carried out. Kabalka et al. reported another method to syn-
thesize phenols. Namely, sodium perborate could convert
benzylic alcohols to phenols with boron trifluoride etherate
in tetrahydrofuran (20). Compound 6 was treated under sim-
ilar conditions to afford 7 in 93% yields. In contrast, com-
pound 7 was not formed by similarly treating 5.
OAc
OMe
OH
1
2
CH2OH
CO2Me
CO2Me
d
e
BnO
OBn
BnO
OBn
HO
OH
OMe
5
OMe
4
OMe
3
f
OH
OH
g
BnO
OBn
BnO
OBn
OMe
OMe
7
6
Structure analysis
Scheme 2. (h) Pyridine/CrO3, THF and (g) SeO2, H2O2, t-butyl
Figures 1–4 represent the crystal structures of compounds
3–6 with thermal ellipsoids at 30% probability, respectively.
For compound 6, the hydroxyl O(4) and the two methyl
groups C(23) and C(24) were disordered and were refined in
three different orientations. The cause is the rotation about
C(1)–C(22) bond. The selected bond lengths and bond an-
gles are listed in Table 2.
alcohol.
OH
g
CH2OH
CHO
BnO
OBn
h
OMe
BnO
OBn
7
BnO
OBn
OMe
OMe
In the crystal structures of compounds 4–6, there are three
benzene rings for each molecule. Although they are structur-
ally similar, the dihedral angels between the central benzene
ring and the two peripheral ones in different compounds
have significant differences. In compounds 4–6, C(1), C(2),
C(3), C(4), C(5), and C(6) form the first plane, defined as
plane I; C(8), C(9), C(10), C(11), C(12), and C(13) form the
second plane, defined as plane II; C(16), C(17), C(18),
C(19), C(20), and C(21) form the third plane, defined as
plane III. Plane II and plane III make dihedral angles of
83.064(84) and 62.153(112)°, respectively, with plane I, and
the dihedral angle between plane II and plane III is
88.847(112)° in compound 4. Whereas, plane II and plane
III make dihedral angles of 84.899(61) and 50.761(66)°, re-
spectively, with plane I in compound 5 vs. 36.582(81) and
31.189(98)°, respectively, in compound 6, and the dihedral
angle between plane II and plane III is 53.726(73)° in 5 and
37.767(87)° in 6 (Table 3).
CO2H
5
g
BnO
OBn
OMe
8
Fig. 1. The crystal structure of compound 3 (30% probability
ellipsoids).
In compound 3, the bond length of C(8)—O(4) is
1.202(3) Å, which is comparable with the corresponding
bond length (C(22)—O(4) = 1.197(3) Å) in compound 4.
This is in good agreement with the standard values of C=O
bond. In both compounds 3 and 4, the bond lengths of
C(=O)–O are shorter than that of the other C—O single
bond in the ester moieties, and the bond lengths of
C(=O)–C(phenyl) are slightly shorter than that of normal
C—C single bond (1.530 Å) (Table 2), but they are still
within the normal scope. This indicated that p–π conjugation
exists between O and C=O and the existence of a week π–π
conjugation between C=O group and the benzene ring.
In compound 3, the bond length of C(3)—O(1)
(1.373(3) Å) is longer than that of C(5)—O(3) (1.354(3) Å),
this difference may be caused by the different type of
intermolecular hydrogen bonds. In fact, the intermolecular
hydrogen bonds are as follows: O(3)–H(1)···O(4) has length
of 2.819(4) Å and an angle of 147(4)°, with symmetry code
© 2007 NRC Canada

Xiao et al.
955
Table 2. Selected bond lengths (Å) and angles (°) for compounds 3, 4, 5 and 6
Compound 3
Compound 4
Compound 5
Compound 6
Bond lengths (Å)
C(1)—C(8)
C(3)—O(1)
C(5)—O(3)
C(8)—O(4)
C(8)—O(5)
C(9)—O(5)
1.489(4) C(1)—C(22)
1.373(3) C(3)—O(1)
1.354(3) C(5)—O(3)
1.202(3) C(7)—C(8)
1.328(4) C(15)—C(16)
1.444(4) C(22)—O(4)
C(22)—O(5)
1.503(3)
1.367(3)
1.369(3)
1.500(3)
1.496(4)
1.197(3)
1.322(3)
1.455(3)
C(1)—C(22)
C(3)—O(1)
C(5)—O(3)
C(7)—C(8)
C(15)—C(16)
1.508(2)
1.366(2)
1.3707(19)
1.504(2)
1.495(3)
C(1)—C(22)
C(3)—O(1)
C(5)—O(3)
C(7)—C(8)
C(15)—C(16)
1.525(5)
1.369(4)
1.371(4)
1.504(5)
1.496(6)
C(23)—O(5)
Bond angles (°)
O(4)–C(8)–O(5)
O(4)–C(8)–C(1)
O(5)–C(8)–C(1)
C(4)–O(2)–C(7)
C(8)–O(5)–C(9)
123.8(3) O(3)–C(7)–C(8)
124.2(3) O(1)–C(15)–C(16) 107.7(2)
112.0(2) O(4)–C(22)–O(5)
113.9(3) O(4)–C(22)–C(1)
116.8(2) O(5)–C(22)–C(1)
C(3)–O(1)–C(15)
108.87(18) O(3)–C(7)–C(8)
112.64(14)
O(3)–C(7)–C(8)
O(1)–C(15)–C(16) 111.3(4)
C(4)–O(2)–C(14)
C(5)–O(3)–C(7)
C(22)–O(5)–C(23) 116.7(2)
C(3)–O(1)–C(15)
C(4)–O(2)–C(14)
C(5)–O(3)–C(7)
109.2(3)
O(1)–C(15)–C(16) 109.25(15)
123.6(2)
123.2(2)
113.2(2)
118.22(18)
C(3)–O(1)–C(15)
C(4)–O(2)–C(14)
C(5)–O(3)–C(7)
116.78(13)
113.20(16)
118.23(13)
115.6(2)
116.53(17)
117.8(3)
114.5(3)
117.7(3)
Fig. 2. The crystal structure of compound 4 (30% probability el-
lipsoids).
Fig. 4. The crystal structure of compound 6 (30% probability el-
lipsoids).
x–1, y–1, z, respectively. On the other hand, O(1)–
H(3)···O(1) has length of 2.775(4) Å and an angle of
146(3)°, with symmetry code x – 1/2, –y + 3/2, –z, respec-
tively, (Table 4). One hydrogen bond acceptor is the car-
bonyl O(4) atom and the other, however, is the phenolic
hydroxyl O(1) atom. The corresponding bond lengths of
C=O in compounds 4–5 verified that hypothesis. The two
C=O bond lengths in each compound (4 or 5) are almost
identical (Table 2). Obviously, the reason is that the hydro-
gen bonds mentioned do not exist any longer owing to the
hydrogen atoms in hydroxyl groups being substituted.
The packing diagrams in Figs. 5 and 6 show the stacking
patterns of compounds 3 and 5, respectively. In the crystal
cell of compound 3, intermolecular hydrogen bonds lead to
forming three-dimensional supramolecular network (Fig. 5);
meanwhile, the intramolecular hydrogen bonds were also
observed between the two hydroxyl groups and O(2) atom.
On the other hand, the molecules of compound 5 are linked
by the intermolecular hydrogen bonds to form a one-
Fig. 3. The crystal structure of compound 5 (30% probability el-
lipsoids).
© 2007 NRC Canada

956
Can. J. Chem. Vol. 85, 2007
Table 3. Dihedral angles in compounds 4, 5, and 6.
Plane
4
5
6
C1 to C6 and C8 to C13
C1 to C6 and C16 to C21
C8 to C13 and C16 to C21
83.064(84)°
62.153(112)°
88.847(112)°
84.899(61)°
50.761(66)°
53.726(73)°
36.582(81)°
31.189(98)°
37.767(87)
Table 4. Intra and Intermolecular hydrogen bonds in compounds 3 and 5.
Compound
D···H
d(D···H)
d(H···A)
<DHA
d(D···A)
A
3
O1···H3
O1···H3
O3···H1
0.77(3)
0.77(3)
0.73(4)
2.11(3)
2.35(3)
2.18(4)
146(3)
117(2)
147(4)
2.775(4)
2.779(3)
2.819(4)
O1 [x – 1/2, –y + 3/2, –z]
O2
O4 [x – 1, y – 1, z]
5
O3···H1
O4···H4
O4···H4
0.73(4)
0.87(3)
0.87(3)
2.41(3)
2.17(3)
2.47(3)
116(3)
165(3)
125(2)
2.804(3)
3.016(2)
3.052(2)
O2
O2 [x + 1, y, z]
O3 [x + 1, y, z]
Fig. 5. Packing diagram of compound 3, viewed along the c
axis, showing the intermolecular hydrogen-bonding scheme
(dashed lines).
Fig. 6. View of the packing diagram of compound 5 showing
intermolecular hydrogen bonds. The chains run along the a axis.
Table 5. Inhibition of HepG2 cells by com-
pounds 1–7.
Compounds
IC50 (µmol/L)
5-UF
51.6 0.2
77.8 0.7
69.6 0.4
90.3 0.6
23.1 0.1
15.3 0.2
20.2 0.5
67.8 0.3
dimensional hydrogen-bonded linear wire, and the linear
wires interact by face-to-face aromatic interactions (Fig. 6).
Table 4 shows the hydrogen bonds in compounds 3–5. All
the geometrical data fall into the normal range of the hydro-
gen bonding interactions.
1
2
3
4
5
6
7
Biological assays
A few studies have evaluated the inhibitory effects of phe-
nolic compounds on HepG2 cancer-cell growth (21) and
studied the effects of pure polyphenols (quercetin,
chlorogenic acid, and epicatechin) on HepG2 cell population
growth and apoptosis. They found that quercetin was the
most active with an IC50 of 87 µmol/L. Table 5 represented
the cytotoxic effect of compounds 1–6 against the HepG2
cell lines. Compounds 4, 5, and 6 showed more potent-cell-
growth inhibitory activities than the positive control 5-FU
(IC50 51.6 µmol/L). Compound 5 showed the highest
bioactivity with an IC50 around 15.3 µmol/L, and compound
4 (IC50 23.1 µmol/L) was as potent as 6 (IC50 20.2 µmol/L).
In contrast, other compounds showed intermediate IC50 val-
ues. This suggested that the lipophilic group makes it easier
for the molecule to permeate the cell membrane and interact
with the enzymes in the cells. Roy et al. reported that lauryl
gallate is able to inhibit protein tyrosine kinases (PTKs) in
whole cells and in crude extracts (22). This allows us to
make the following consideration: the compounds based on
gallates in this paper may exert their proapoptotic activity
© 2007 NRC Canada

Xiao et al.
957
6. J. Gnojkowski, V. Krajka-Kuzniak, and W. Baer-Dubowska.
Nutr. Cancer. 39, 117 (2001).
7. T. Nakayama. Cancer Res. 54, 1991s (1994).
8. T. Nakayama, M. Hiramitsu, T. Osawa, and S. Kawakishi.
Mutat. Res. 303, 29 (1993).
through the mechanism mentioned earlier. The study of the
detailed mechanism is in progress.
Conclusion
9. K. Saeki, A. Yuo, M. Isemura, I. I. Abe, T. Seki, and H.
A structural analysis of compounds 3–6 was carried out
by single-crystal X-ray diffraction. The cytotoxic effects of
seven derivatives of gallates tested against the HepG2 cell
lines were evaluated. The effect of slight structural varia-
tions in this kind of systems was found to cause a marked
change in their biological activity. Compound 5 showed the
highest bioactivity with an IC50 around 15.3 µmol/L. The un-
derstanding of the structure–activity relationships ruling
their biological function lies in more specific assays, the tar-
get particular binding sites, and the mechanisms of action.
Further structure–activity relationship study is undergoing
and will be reported in due course.
Noguchi. Biol. Pharm. Bull. 23, 1391 (2000).
10. G. Roy, M. Lombardia, C. Palacios, A. Serrano, C. Cespon, E.
Ortega, P. Eiras, S. Lujan, Y. Revilla, and P. Gonzalez-Porque.
Arch. Biochem. Biophys. 383, 206 (2000).
11. T. Murase, N. Kume, T. Hase, Y. Shibuya, Y. Nishizawa, I.
Tokimitsu, and T. Kita. Arterioscler.Thromb. Vasc. 19, 1412
(1999).
12. K. Brusselmans, E. De Schrijver, W. Heyns, G. Verhoeven, and
J.V. Swinnen. Int. J. Cancer 106, 856 (2003).
13. Q. Feng, T. Kumagai, Y. Nakamura, K. Uchida, and T. Osawa,
Mutat. Res-Gen. Tox. En. 537(1), 101 (2003).
14. T. Mosmann. J. Immunol. Methods, 65, 55 (1983).
15. SMART 5.0 and SAINT 4.0 for Windows NT [computer pro-
grams]. Area detector control and integration software. Bruker
Analytical X-Ray Systems, Inc.: Madison, WI. 1998.
16. G.M. Sheldrick. SHELXTL [computer program]: Structure de-
termination software programs. Bruker Analytical X-Ray Sys-
tem, Inc.: Madison, WI. 1997.
Acknowledgements
The work was financed by grant (Projects 30672516 and
30772627) from National Natural Science Foundation of
China.
17. M.L. Cardona, M.I. Fernandez, M.B. Garcia, and J.R. Pedro.
Tetrahedron, 42, 2725 (1986).
References
18. A.J. Pearson and P.R. Bruhn. J. Org. Chem. 56, 7092 (1991).
19. J.A. Guzm, V. Mendoxa, E. Garc, C.F. Garibay, L.Z. Olivares,
and L.A. Maldonado. Synthetic Commun. 25 (14), 2121
(1995).
1. C.A. van der Heijden, P.J. Janssen, and J.J. Strik. Food Chem.
Toxicol. 24, 1067 (1986).
2. N. Niho, M. Shibutani, T. Tamura, K. Toyoda, C. Uneyama, N.
Takahashi, and M. Hirose. Food Chem. Toxicol. 39, 1063
(2001).
20. G.W. Kabalka, N. K. Reddy, and C. Narayana. Tetrahedron
Lett. 34, 7667 (1993).
3. E.R. Sherwin. Food additives. Dekker, New York. 1990.
pp. 139–193.
21. S. Ramos, M. Alia, L. Bravo, and L. Goya. J. Agr. Food
Chem. 53, 1271 (2005).
4. N. Ito and M. Hirose. Adv. Cancer Res. 53, 247 (1989).
5. M. Hirose, A. Masuda, S. Fukushima, and N. Ito.
Carcinogenesis, 9, 101 (1988).
22. G. Roy, M. Lombard, C. Palacios, A. Serrano, C. Cesp, E.
Ortega, P. Eiras, S. Lujan, Y. Revilla, and P. Gonzalez-Porqu.
Arch. Biochem. Biophys. 38, 206 (2000).
© 2007 NRC Canada