4590
J.-H. Liu et al. / Tetrahedron Letters 50 (2009) 4587–4591
probably due to the similar polarity of compound 26 and its diaste-
reoisomer, which just differed in the configuration at 1-and 7-po-
sition. However, the product was produced almost quantitatively
in this reaction.
When HCl was used to catalyze the intramolecular ketal forma-
tion reaction in MeOH, which was a successful procedure in the
former synthetic route, the exchange of benzyloxyl group to meth-
oxyl group at 3-position was observed, leading to the side product
28. When the solvent was changed to ether or dioxane, the epimer-
ization at 3-position occurred. We proposed the mechanism of the
side reaction as follows (Scheme 5): the reactant was easily elimi-
accomplished under the condition of BF3ÁEt2O in Et2O at 0 °C,
affording the alcohol 27 as the major product in yield of 85%. The
configurations of 27 and its 3-epimer epi-27 were determined with
NOESY (Scheme 6). The existence of NOE effect between Ha and Hb
revealed the S configuration of 3-position in compound 27, while
such NOE effect was not observed in the diastereoisomer epi-27.
Considering the natural product cyclodidemniserinol trisulfate
possessed both antitumor and HIV-1 integrase inhibitory activity,
we chose tumor cell line (MDA-MB435) and HIV-1 integrase to
evaluate the bioactivity of compound 27 and its 3-epimer epi-27.
Neither compound showed inhibition against the HIV-1 integrase,
indicating the important effect of the 22-memberd additional ring
and the sulfate groups on the HIV-1 integrase inhibition. However,
the compound 27 exhibited an IC50 value of 7.8 lM to inhibit the
growth of the MDA-MB 435 cell, whereas the epi-27 displayed
no antiproliferative effect against the tumor cell line at the concen-
tration of 20 lM, suggesting the importance of the 3-chirality on
the antitumor activity. This interesting result provided useful clue
for further structural elaboration and medicinal chemistry study of
3,5,7-trisubstituted 6,8-dioxabicyclo [3.2.1] octane-based focused
library.
In conclusion, by employing either a convergent synthesis or a
linear synthesis approach, we accomplished the construction of
core structure 3,5,7-trisubstituted-6,8-dioxabicyclo [3.2.1] octane.
In the context of further total synthesis, the two strategies were
evaluated. In general, convergent synthesis strategy was made up
of 13 reactions, with the longest route consisting of eight steps.
The final trisubstituted bicycle structure was obtained in 0.9%
overall yield with unsolved chirality at 3-position. However, the
linear synthesis strategy encompassed 19 steps according to the
longest route, with 5.3% overall yield and an unambiguous chirality
at 3-postion.
Obviously, the lower yield of strategy one will hamper its appli-
cation in total synthesis. Although the synthesis route was
relatively long, strategy two was preferred for utilization in the to-
tal synthesis, because of its higher yield and the installed unambig-
uous chiral center at 3-postion. Furthermore, the preliminary
biological assay provided interesting results of the 3,5,7-trisubsti-
tuted-6,8-dioxadicyclo [3.2.1] octane structure and its 3-eipmer
in inhibiting the growth of the tumor cells. Our developed syn-
thetic methodology will be beneficial for the further medicinal
chemistry study and structural elaboration.
nated to yield the a,b-unsaturated system with the catalysis of HCl.
When MeOH was used as solvent, the methoxyl group attacked the
unsaturated system via Michael-addition reaction. As a result, ben-
zyloxy-methoxyl exchanged by-product generated. When ether or
dioxane was used as solvent, there was no other nucleophile spe-
cies present in the reaction system except for the benzyloxyl
group, which was just eliminated from the reactant. So the benzyl-
oxyl group attached again via 1,4-addition non-stereoselectively
and caused epimerization. In order to prove the proposed mecha-
nism, a drop of MeOH was added to the reaction system of catalytic
amount of HCl in Et2O, TLC revealed that there appeared benzyl-
oxy-methoxyl exchanged by-product.
Since the proton acid HCl was the causing factor for the side
reaction, we decided to try non-proton Lewis acid in the non-pro-
ton solvent such as Et2O, to secure the desired dioxabicycle struc-
ture formation. As expected, the final step was successfully
BnO
HO
O
O
OBn
OBn
OBn
HO
HO
26
OH
H+
BnO
O
HO
OH
OH
OBn
O
O
BnO
H+
27
O
OH
HO
HO
OH
Et2O or
dioxane
MeOH
(MeO-)
(BnO-)
Acknowledgement
O
O
OH OBn
OH OMe
OBn
OBn
This work was financially supported by the National Natural
Science Foundation of China (No. 20372068 and 30672528).
HO
OH
OH
Supplementary data
OH
OBn
OH
OBn
O
O
O
O
Synthetic procedure, NMR data and HRMS data of all new
compounds; and 1H and 13CNMR spectra of important interme-
diates, including the gCOSY and NOESY spectrum of 27 and
epi-27, are provided. Supplementary data associated with this
article can be found, in the online version, at doi:10.1016/
MeO
BnO
28
Scheme 5. Proposed mechanism of side reaction in intramolecular ketal formation
step.
References and notes
O
O
Ha
BnO
Ha H
1. Mitchell, S. S.; Rhodes, D.; Bushman, F. D.; Faulkner, D. J. Org. Lett. 2000, 2,
1605–1607.
2. Gonzalez, N.; Rodriguez, J.; Jimenez, C. J. Org. Chem. 1999, 64, 5705–5707.
3. Haustedt, L. O.; Mang, C.; Siems, K.; Schiewe, H. Curr. Opin. Drug Discov. Dev.
2006, 9, 445–462.
4. Wood, D. L.; Browne, L. E.; Ewing, B.; Lindahl, K.; Bedard, W. D.; Tilden, P. E.;
Mori, K.; Pitman, G. B.; Hughes, P. R. Science 1976, 192, 896–898.
5. Ellory, J. C.; Tucker, E. M. Nature 1969, 222, 477–478.
OBn
OH
OBn
OH
BnO
O
O
Hb
b
NOE Effect
No NOE Effect
27
epi-27
Scheme 6. The NOE analysis of compound 27 and epi-27.