54
C. Marquissolo et al. / Bioorganic Chemistry 37 (2009) 52–56
expressing phenotype multiple drugs resistance) were kindly pro-
vided by Frederick Cancer Research & Development Center – Na-
tional Cancer Institute – Frederick, MA, USA. Stock cultures were
grown in 5 mL of RPMI 1640 (GIBCO BRL, Life Technologies) supple-
oxide (ent-1) and ent-isogoniothalamin oxide (ent-2) was formed
{(ent-1: ½a 2D5
ꢂ
¼ ꢁ121:0 (c 0.7, CHCl3); ent-2: ½a D25
¼ ꢁ32:0 (c 0.9,
ꢂ
CHCl3)}, respectively). It is to be noticed that the absolute value
of the specific optical rotation for ent-2 again differs from that de-
scribed in the literature for isogoniothalamin oxide (2) but is in full
agreement with the one observed in this work.
mented with 5% of fetal bovine serum. Gentamicine (50
added to the experimental cultures. Cells in 96-well plates (100
cells/well) were exposed to various concentrations of epoxides 1,
ent-1, 2 and ent-2 DMSO (0.25, 2.5, 25 and 250 g/mL) at 37 °C, 5%
lg/mL) was
lL
Markó and coworkers have shown that (R)-(+)-goniothalamin
oxide (1) was prepared in 19:1 diastereoisomeric ratio and 98%
yield when purified and water free mCPBA (4.0 equiv.) was em-
ployed in CH2Cl2 at 0 °C [14]. Similar results were recently de-
scribed by Bose and coworkers [15]. However, both Markó and
Bose also reported that the use of commercially available 70%
mCPBA in CH2Cl2 at reflux provided a 3:2 mixture of (R)-goniothal-
amin oxide (1) and (R)-isogoniothalamin oxide (2), in 69% yield. In
order to secure both enantiomeric forms of goniothalamin oxide
(1) and isogoniothalamin oxide (2) for biological evaluation, we
have initially employed commercially available 70% w/w mCPBA
(1.2 equiv.) in CH2Cl2 at rt which in our hands provided a 3:2 mix-
ture of the two diastereoisomers, in 64% overall yield. The diaste-
reoisomeric ratio could be determined by inspection of the 1H-
NMR spectrum of the crude mixture and it was confirmed after iso-
lation of goniothalamin oxide (1) and isogoniothalamin oxide (2)
by chromatographic separation on silica gel.
We have investigated the use of Jacobsen’s catalyst in order to
introduce chiral discrimination by the reagent [21] aiming to pre-
pare the enantiomers of isogoniothalamin oxide as the major epox-
ides. However, when (R,R)-Jacobsen’s catalyst was employed in the
epoxidation of (R)-4 only a slight increase in the diastereoisomeric
ratio in favor of goniothalamin oxide (1) was observed (d.r. = 2.3:1,
52% yield). The use of (S,S)-Jacobsen’s catalyst afforded a 6:1 molar
ratio of goniothalamin oxide (1) and isogoniothalamin oxide (2)
which were obtained in a combined 49% yield. Bose and coworkers
[15] have also employed (S,S)-Jacobsen’s catalyst in the epoxidation
of (R)-goniothalamin (4) but under different experimental condi-
tions compared to those described here: either hexafluoroacetone
or acetone were employed in the presence of OxoneÒ in tetrabutyl-
ammonium sulfate buffered solution and acetonitrile as solvent to
afford (R)-goniothalamin oxide as the major product (98:2 and
95:5 ratio, respectively) in very good yields (90% and 80% yield).
Considering the difference in the experimental conditions em-
ployed by us and those described by Bose and coworkers [15], it
is not easy to rationalize the differences observed in both yield
and diastereoisomeric ratios. In any event, these results indicate
that the use of Jacobsen’s catalysts in the epoxidation of goniothal-
amin is not an efficient route to isogoniothalamin epoxide (2) and
are in accordance with those described by Katsuki and coworkers
for the epoxidation of trans double bonds by optically active (salen)
manganese (III) complexes derived from (S,S)-1,2-diaminocyclo-
hexane where preferencial epoxidation of the Si face of the trans
styrenic double bond (regarding the ipso carbon) was observed al-
beit in low e.e. [22]. In our case, it seems that chiral discrimination
of the diastereotopic faces of the styrenic double bond is enhanced
by the intrinsic Si face preference of the substrate (matched pair).
l
ofCO2 for48 h. ThefinalconcentrationofDMSOdidnotaffectthecell
viability. Then, a 50% of trichloroacetic acid solution was added and
after incubation (30 min at 4 °C), washing and drying, the cell prolif-
eration was determined by spectrophotometric quantification
(540 nm) of cellular protein content using sulforhodamine B assay
[16]. The background absorbencies were subtracted from the appro-
priate control and drug-blank measurements. To assess the effect of
goniothalamin oxide (1) and its isomers (2, ent-1 and ent-2) on cell
growth, three measurements were obtained at time zero (T0) values
for all cells at the beginning of incubation, and control (C) and test (T)
values at the end of incubation without and with the test substance,
respectively. For T value 6T0 (cytostatic effect), the calculation was
100x [(T ꢁ T0)/C ꢁ T0]. While for T < T0 (cytocidal effect), the calcula-
tion was 100x[(T ꢁ T0)/T0]. The GI50 values (test substance concen-
tration eliciting 50% inhibition of cell growth) and TGI (test
substance concentration eliciting 100% inhibition of cell growth)
were determined by non-linear regression analysis using Origin
software, version 7.5. These results presented here refer to a repre-
sentative experiment since all assays were run in triplicate and the
average standard error was always <5%.
3. Results and discussion
3.1. Chemistry
Goniothalamin oxide (1), isogoniothalamin oxide (2) and their
respective enantiomers (ent-1 and ent-2) were obtained from the
epoxidation of goniothalamin enantiomers, prepared in three steps
from trans-cinnamaldehyde, as previously described (Scheme 1)
[17–20].
Epoxidation of enantiomerically enriched (>95% enantiomeric
excess by chiral gas chromatography) (R)-goniothalamin (4) with
commercially available 70% m-chloroperbenzoic acid (mCPBA) un-
der standard conditions (1.2 equiv., CH2Cl2, rt) provided a 3:2 mo-
lar ratio of goniothalamin oxide (1) and isogoniothalamin oxide (2)
in 64% yield, after separation by column chromatography on silica
gel. The trans configuration of epoxides 1 and 2 was assigned by
the inspection of the coupling constant between H7 and H8 (1.8
and 2.4 Hz, respectively). The stereochemistry of each diastereoiso-
mer was established by comparison with literature data [5]. De-
spite the good correlation of the NMR data of the two
diastereoisomers formed in the mCPBA epoxidation of (R)-goni-
thalamin (4) with those described by Sam and coworkers [5], the
optical rotation of isogoniothalamin oxide (2) prepared by us
{[
a
]
D = +31.3 (c 1.3, CHCl3)} was significantly different from that
D = ꢁ106.0 (c 1.5, CHCl3)}. Con-
described in the above reference [
a]
3.2. Biological activities
sidering that the specific optical rotation found by us for goniothal-
amin oxide (1) {½a D25
¼ þ113:0 (c 0.8; CHCl3)} nicely matches the
ꢂ
Since it is well known that different cell lines display different
sensitivities toward a cytotoxic compound, in the present study
we have used cell lines of various histological origin [MCF-7
(breast), NCI-ADR/RES (ovarian expressing the resistance pheno-
type for adryamycin), NCI-H460 (lung, non-small cells), UACC-62
(melanoma), 786-0 (kidney), OVCAR-03 (ovarian), PC-3 (prostate),
and HT-29 (colon)] for the initial evaluation of the cytotoxicity of
goniothalamin oxide (1), isogoniothalamin oxide (2) and its
respective enantiomers. Cell proliferation was determined by spec-
trophotometric assay using sulforhodamine B as protein-binding
one described in the literature by Sam and coworkers
{(½a 2D5
¼ þ107:0 (c 0.7; CHCl3)}, we consider that a revision of the
ꢂ
specific optical rotation for isogoniothalamin oxide (2) may be in
order. The stereoselective epoxidation of (R)-goniothalamin (4) is
consistent with the reaction taking place at the Si face of the styre-
nic double bond to afford the major goniothalamin epoxide (1).
Accordingly, when enantiomerically enriched (>95% enantio-
meric excess) (S)-goniothalamin (4) was employed under the same
epoxidation conditions, a 3:2 molar mixture of ent-goniothalamin