Original Articles 977
Acetylation of (−)-dihydroguaiaretic acid
Results and Discussion
!
50 mg of (−)-dihydroguaiaretic acid (9) were suspended in Ac O
2
(
1.0 mL) and pyridine (1.0 mL). After stirring at room temperature
In our search for bioactive compounds from plant sources, several
phenolic compounds were isolated from Euphorbia lagascae [12–
14], Euphorbia tuckeyana [15], and Pycnanthus angolensis [17], or
obtained by derivatization reactions [12,16]. These compounds
include the stilbenes piceatannol (1), trans-3,5,3′,4′-tetraacetoxy-
piceatannol (2), and trans-3,5,3′,4′-tetramethoxypiceatannol (3),
the flavonoids naringenin (4) and aromadendrin (5), the couma-
rins scopoletin (6) and esculetin (7), the neolignan dehydrodico-
niferyl diacetate (8), as well as the lignans (−)-dihydroguaiaretic
acid (4,4′-dihydroxy-3,3′-dimethoxylignan, 9) and heliobuph-
for 24 hours, the excess of reagents were eliminated with N . Pu-
2
rification was made by preparative chromatography (CHCl3-
MeOH, 39:1) yielding 52 mg of compound 10 (4,4′-diacetyl-3,3′-
dimethoxylignan).
4
,4′-diacetyl-3,3′-dimethoxylignan (10): EIMS, m/z (rel. int.): 414
+
1
[M] (2), 372 (28), 330 (51), 137 (100). H RMN (400 MHz, CDCl ):
3
δ 6.92 (d, J = 8.0 Hz, H-5 and H-5′), 6.67 (dd, J = 1.6 and 8.0 Hz, H-6
and H-6′), 6.66 (br s, H-2 and H-2′), 3.78 (s, OMe-3 and OMe-3′),
2
.61 (dd, J = 6.8 and 13.6 Hz, H-7a and H-7′a), 2.46 (dd, J = 8.0 and
3.6 Hz, H-7b and H-7′b), 2.33 (s, OCOMe-4 and OCOMe- 4′), 1.81
"
1
thalmin (12, l Fig. 1).
(
m, H-8 and H-8′), and 0.87 (d, J = 6.4 Hz, H-9 and H-9′).
Continuing in this study our search for antineoplasic agents, the
lignan (−)-dihydroguaiaretic acid (9) was derivatized to afford
the new compounds 4,4′-diacetyl-3,3′-dimethoxylignan (10) and
4′-hydroxy-3,3′,4-trimethoxylignan (11). The structure elucida-
tion of these compounds was based on a comparison of their MS
and NMR spectra with those of (−)-dihydroguaiaretic acid (9).
The twelve phenolic compounds listed above were investigated
for their potential antiproliferative activity in several human can-
cer cell lines which were derived from three different tumor en-
tities: gastric (EPG85-257), pancreatic (EPP85-181), and colon
cancer cells (HT-29). Furthermore, in each case two different
multidrug-resistant variants of these cells were also investigated:
cell lines with a classical MDR phenotype (associated with the
overexpression of MDR1/P-gp, EPG85-257RDB, EPP85-181RDB,
and HT-29RDB) and cell lines with an atypical MDR phenotype
(no enhanced expression of MDR1/P-gp, EPG85-257RNOV,
EPP85-181RNOV, and HT-29RNOV). For assessment of cytotoxic-
ity of the twelve tested compounds, the IC50 values of each agent
were determined by proliferation assays in each of the different
cell variants after continuous exposure for 5 days. The etoposide
and cisplatin-specific IC50 values were considered as positive
controls for maintenance of the drug-resistant phenotype. Rela-
tive resistance (RR) values were also determined as the relation
between the IC50 of the resistant cell line and the IC50 of the pa-
rental drug-sensitive cell line. When the sensitivity against a giv-
en compound was less than 10% of the corresponding parental
cell line, the compound was assessed to be highly efficient in this
drug-resistant cell line [18,19].
Methylation of (−)-dihydroguaiaretic acid
5
by-drop to a solution of compound 9 (50 mg in 2 mL of anhydrous
ether). The reaction occurred overnight and then was followed by
TLC. The purification of the compounds was performed by prep-
arative chromatography (CH Cl -acetone, 19:1) to yield 14.5 mg
mL of an ethereal solution of diazomethane was added drop-
2
2
of compound 11 (4′-hydroxy-3,3′,4-trimethoxylignan).
′-hydroxy-3,3′,4-trimethoxylignan (11): EIMS, m/z (rel. int.): 344
4
+
1
[M] (35), 152 (30), 151 (100), 137 (75), 28 (17). H RMN
(400 MHz, CDCl ): δ 6.82 (d, J = 8.0 Hz, H-5′), 6.78 (d, J = 8.0 Hz,
3
H-5), 6.65 (dd, J = 2.0 and 8.0 Hz, H-6), 6.61 (dd, J = 2.0 and 8.0 Hz,
H-6′), 6.59 (d, J = 2.0 Hz, H-2), 6.56 (d, J = 2.0 Hz, H-2′), 5.49 (s, OH-
4
and 13.6 Hz, H-7a and H-7′a), 2.41 (dd, J = 6.8 and 13.6 Hz, H-7b
and H-7′b), 1.77 (m, H-8 and H-8′), and 0.85 (d, J = 7.2 Hz, H-9
and H-9′).
′), 3.88 (s, OMe-4), 3.83 (s, -OMe-3 and -OMe-3′), 2.56 (dd, J = 6.8
Cell lines, cell culture, and cell proliferation assay
Human cancer cell lines and drug-resistant sublines were grown
in modified Leibovitz L-15 medium (Biowhittaker) as described
previously [15,18,19]. Drug-resistant cell lines were established
from parental cell lines by continuous exposure of the cells to
stepwise increasing concentrations of antineoplastic agents [20].
For maintenance of drug-resistant phenotypes, the medium of
drug-resistant sublines was supplemented with the selection
agent. Resistance to etoposide, cisplatin (both from Gry-Pharma),
and tested compounds was assessed using a proliferation assay
based on sulforhodamine B (SRB) staining. Briefly, 800 cells per
well were seeded in 96-well plates in triplicate. After 24-h attach-
ment, the particular agent was added in a dilution series for 5
days incubation. Cells were fixed by chilled 10% trichloroacetic
acid for 1 h at 4°C and washed five times with tap water before
staining was performed with 0.4% SRB (Sigma) in 1% acetic acid
for 10 min at room temperature. After washing with 1% acetic ac-
id, drying, and resolubilization in 20 mM Tris-HCl (pH 10), ab-
sorbance was measured at 562 nm against the reference wave-
length of 690 nm. Mean IC50 values and standard deviations were
calculated from at least two independent experiments in tripli-
cate for each cell line by using the Prism software (GraphPad Soft-
ware, Inc.). Relative resistance (RR) values were also determined
as:
The IC50 and RR values of drug-resistant cell variants in compari-
son to the drug-sensitive parental cell lines are summarized in
"
l Tables 1, 2 and 3 for the three EPG85-257 gastric, EPP85-181
pancreatic, and HT-29 colon carcinomas, respectively. As it can
be observed, in parental drug-sensitive cell lines, all the tested
compounds (1–12) showed a moderate/weak antiproliferative ef-
fect or were inactive. In contrast, when comparing with the pa-
rental sublines, some of the multidrug-resistant variants showed
increased sensitivities to the studied compounds, particularly the
EPG85-257RDB subline.
When considering the results obtained for the stilbenes (1–3), it
can be observed that in the drug-resistant subline EPG85-
257RDB (associated with the classical MDR phenotype) derived
from gastric carcinoma, piceatannol (1) and its tetramethylated
derivative (3) were found to be as effective as the positive control
etoposide (IC50 = 6.2 µM) and slightly less effective than cisplatin
(IC50 = 4.0 µM), showing IC50 values of 5.6 and 6.3 µM, respective-
RR = (IC50 resistant cell line)/(IC50 parental-drug sensitive cell
line)
"
ly (l Table 1). Furthermore, trans-3,5,3′,4′-tetramethoxypicea-
tannol (3) exhibited a RR value of 0.18 and thus can be consid-
ered very effective against this cell line. On the other hand, when
compared to piceatannol, its acetoxy derivative (2) showed just a
Duarte N et al. Phenolic Compounds As… Planta Med 2010; 76: 975–980