´ ´
J. L. Lopez-Perez et al. / Bioorg. Med. Chem. Lett. 14 (2004) 1283–1286
1285
solution of BF3OEt2 was carefully added at ꢀ78 ꢁC in an
inert atmosphere, no reaction was observed, whereas if the
temperature was slowly increased, an intense coloration
appeared, giving place to a mixture of products that could
not be resolved. When other catalysts such as SnCl4, TiCl4,
AlCl3 or ZnCl3 were added at ꢀ78 ꢁC, the reaction did not
take place and it was necessary to increase the temper-
ature up to 0 ꢁC to obtain the desired adducts. In this
case, however, no increase in the rate of the endo/exo
selectivity was observed with respect to the reaction in
thermal conditions. For this reason, all the cycloaddi-
tion reactions described in this work, were finally per-
formed at ꢀ10 ꢁC, in absence of catalyst.
Table 1. Cytotoxicity of lignans esters against four neoplastic cell
lines. (Etoposide and podophyllotoxin as reference compounds. Data
are expressed as IC50 nM values)
Compd
P-388
A-549
HT-29
MEL-28
1a*
2a*
3a
4a
5a
170
12
21
5
170
12
21
4
1700
24
21
4
850
—
21
4
5
4
4
4
6a
7a
190
24
190
48
190
48
190
48
8a
9a
10a
2b
47
94
94
94
184
454
60
184
454
60
184
454
60
184
454
—
3b
55
19
53
19
55
19
53
19
1930
188
9182
The determination of the diastereomeric excess of the
four Diels–Alder reactions was performed by HPLC
analysis. The complete chemistry results will be published
elsewhere. In the four cases, the two endo compounds
were the major products while the two exo compounds
appeared in proportion less than 0.5% of the crude
reaction products. NMR studies of the diastereoisomeric
pairs obtained from the four Diels–Alder reactions,
showed that the major compounds were the endo
adducts, in agreement with theoretical predictions. The
mixture of the endo stereoisomers 4a and 5a was
resolved by flash chromatography. The identification of
every compound of each diastereoisomeric pair was
achieved on the basis of specific rotation studies.13
4b+5b
7b
8b
9b
2c
1930
188
4591
6000
213
1122
2900
213
468
1930
188
9182
6000
213
1122
2900
213
468
1930
188
9182
6000
213
1122
2900
213
468
3c
213
1122
—
213
468
4c+5c
2d
3d
4d+5d
is quite obvious and can be justified by their close simi-
larity or identity in lipophilicity, configuration at C-7,
global conformation and spatial distribution, in spite of
the different configuration at position C-5 of the nor-
bornene moiety, with the additional assumption of free
rotation around the CO–norbornene bond.
The preparation of podophyllotoxin esters 6a–10a was
carried out with good yield, by treatment of podophyllo-
toxin, suspended in methylene dichloride, with the corres-
ponding acid chloride (heptanoic, benzoic, phenylacetic,
trans-cinnamic and norbornylacetic acids) in the presence
of pyridine. Nevertheless, in the case of the preparation
of epipodophyllotoxin esters (6b–10b), due to the axial
disposition of the hydroxyl group, better results were
achieved when using the Hassner14 procedure, in which,
epipodophyllotoxin is treated with the corresponding
free acid (heptanoic, phenylacetic, trans-cinnamic and
norbornylacetic) in presence of 1,3-dicyclohexylcarbodi-
imide (DCC) and 4-dimethylaminopyridine (4-DMAP).
In conclusion, the introduction of linear or aromatic
acyl moieties at the C-7-OH of podolignans, generally
induces either a loss of, or no effect upon, the cytotoxic
activity of the parent hydroxylic derivatives. Never-
theless, when the acyl residue contains a norbornene
fragment, almost a one-fold increase of activity is
observed in the case of trans-lactones. This improve-
ment of activity is more significant in the case of podo-
phyllotoxin norbornene derivatives. The introduction of
one additional methylene unit between the bicylic sys-
tem and the carboxyl group (2-norborneneacetate of
cyclolignan, 10a) that would increase the lipophicity
and the volume virtually occuped by the substituent, led
to a diminished activity. These findings indicate that the
main responsibility for the cytotoxicity change is not
due to the lipophilic factor, but to the precise spatial
arrangement of the bulky moiety, which may also con-
tribute with additional non-polar interactions to
enhance the binding to the target site.
Cytotoxicities of all of the compounds prepared were
tested in vitro following an adapted procedure of the
method described by Bergeron15 against different neo-
plastic cell lines: P-388 (suspension culture of a lymphoid
neoplasm from DBA/2 mouse), A-549 (monolayer culture
of human lung carcinoma), HT-29 (monolayer culture
of human colon carcinoma) and MEL-28 (monolayer
culture of human malign melanoma). The results of
these assays were used to obtain the corresponding
dose–response curves, from which IC50 (nM) values
were calculated.
Other compounds, mainly 40-demethylderivatives will be
prepared in order to clarify the structure–activity rela-
tionship for this family of compounds in relation with
the DNA-topoisomerase II inhibition.
The results obtained are shown in Table 1. As it can be
seen, the IC50 values of the majority of esters were either
similar or better than those of the parent alcohols.
Nevertheless, both podophyllotoxin norbornene-
carboxylates 4a and 5a showed IC50 values about 4 nM,
that represented an improvement in their potency with
respect to the podophyllotoxin 2a. The equipotency
shown by compounds 4a and 5a against neoplastic cells
Acknowledgements
Financial support came from Spanish MEC (AGL2000-
0039-P4-02). This research has been performed under
the auspices of CYTED Program. Subprogram X.