SCHEME 2
diol monoactetate 5 appears to enter an oxygenation
“matrix” and is converted into a host of polyols, we have
focused on the downstream oxygenation reactions of 6, a
verified taxol pathway metabolite. However, production
of 6 by bioconversion of 5 gives at best, sub-milligram
quantities of this substance, which is insufficient for the
substrate requirements of our planned bioconversion
experiments. We have thus devoted considerable effort
to devising totally synthetic and semisynthetic methods
to prepare this compound. A further restriction we have
placed on our synthetic planning requires that the
convenient installation of radioisotopic tags be easily
attainable.
Herein, we report a method to semisynthetically
prepare 6 from taxa-4(20),11(12)-dien-2R,5R,10â-triac-
etoxy-14â-(2-methyl)butyrate (7), a component of J apa-
nese Yew heart wood.11 This method appears easily
applicable for the preparation of multi-milligram quanti-
ties of tritium-labeled 6.
Taxa-4(20),11(12)-dien-2R,5R,10â-triacetoxy-14â-(2-me-
thyl)butyrate 7 was prepared by extraction from J apa-
nese yew heart wood. Although the yield of 7 is less than
that of taxusin,12 the main taxoid component in Yew
heart wood, compound 7 does not possess the C-13 allylic
hydroxyl group. We have devoted considerable efforts to
deoxygenate taxusin at C-13, but all attempts to reduc-
tively remove this hydroxyl function were attended by
allylic transposition of the bridgehead olefin to the 12-
(13) position.13 Compound 7, therefore, appeared to be
an excellent precursor for the synthesis of lightly oxygen-
ated C-13-deoxy taxoids.
As shown in Scheme 2, all of the acyl groups of 7 were
removed by treatment with LAH to give taxa-4(20),11-
(12)-dien-2R,5R,10â,14â-tetraol 814 in good yield. Next,
the C-10-hydroxyl group was selectively protected as the
corresponding tert-butyldimethylsilyl ether (TBS) 9a or
as the corresponding diethylisopropylsilyl ether (DEIPS)
9b by treatment with the corresponding silyl chlorides
at low temperature. Deoxygenation via hydride attack
on C-2- and/or C-14- mesyl or tosyl groups was consid-
ered, but appeared to be difficult due to steric hindrance
on the congested and convex R-face at C-14 as well as to
steric hindrance imposed by the C-17 methyl group on
the â-face at C2.
enter a very complex matrix and the elucidation of a
single linear path to taxol has proven extremely chal-
lenging. This appears to be due to the softening of
substrate specificity by several of the remaining cyto-
chrome P-450 enzymes that can accept 5 as a suitable
substrate in vitro. Two major approaches are being
concurrently investigated to identify the genes and
associated intermediates from 5 to taxol. In the first
approach, we have obtained a set of related full-length
cytochrome P-450 clones by the method of differential
display of mRNA-reverse transcription-PCR, followed by
traditional library screening. Clones were selected based
on homology to other plant cytochrome P-450s and used
to individually transform Saccharomyces cerevisiae and
the transformed yeast clones were screened for oxygen-
ation activity with several taxoids as substrates. One
such clone produced taxa-4(20),11(12)-dien-5R-acetoxy-
10â-ol (6) using taxa-4(20),11(12)-dien-5R-acetate (5) as
a substrate.7 This material has been characterized by 1H
NMR and mass spectroscopy and has been shown to
incorporate into taxol in vivo.
In a second approach, microsomal bioconversion of 5
with Taxus sp. microsomes to more polar products,
however, yields several diol monoacetates including
taxa-4(20),11(12)-dien-5R-acetoxy-2R-ol8 and taxa-4(20),-
11(12)-dien-5R-acetoxy-13R-ol along with other, as yet
unidentified oxygenation products.9
Thus, Barton deoxygenation strategies were exam-
ined.15 Xanthate esters were formed from the sterically
more accessible C-2 hydroxyl group of triols 9a and 9b
by treatment with base followed by O-alkylation with
carbon disulfide and methyl iodide. This protocol fur-
nished the requisite substrates 10a and 10b in 55% and
44% yields, respectively (Scheme 3). Resubjecting 10a
and 10b to the xanthate ester acylation conditions yielded
Due to the very low yield of the intermediate metabo-
lites that may be obtained from natural sources, we have
relied heavily on synthetic, tritium-labeled taxadienes
3,10 4,5 and 55 as substrates from which in vivo and in
vitro bioconversion strategies have been utilized to
identify lightly oxenated taxoids downstream of these
substances. Using tritium-labeled 5 as a substrate, we
have identified several diol, triol, and higher polyols from
Taxus sp. suspension cell cultures by GC-MS, and, in a
few instances where more than 20 µg was produced and
purified, 1H NMR techniques were applied to identify the
structures of three diol monoacteates (vida supra). Since
(10) (a) Rubenstein, S. M.; Williams, R. M. J . Org. Chem. 1995, 60,
7215-7223. (b) Rubenstein, S. M.; Vazquez, A.; Williams, R. M. J .
Labelled Cmpd. Radiopharm. 2000, 43, 481-491.
(11) Sugiyama, T.; Oritani, T.; Oritani, Takashi. Biosci. Biotech.
Biochem. 1994, 54, 1923-1924.
(12) Ho, T.-L.; Lee, G.-H.; Peng. S, -M.; Yeh, H.-M.; Chen, F. C.;
Yang, W. L. Acta Crystallogr. 1987, C43, 1378.
(13) Nicolaou, K. C.; Nantermet, P. G.; Ueno, H.; Guy, R. K. J . Chem.
Soc., Chem. Commun. 1994, 295-296.
(14) Keki, C.; Weiming, C.; Weihuua, Z. Qicheng, F.; XiaoTian, L.;
J iyu, G. PCT Int. Appl. 1994, 31. J apanese Patent Appl. No. Wo
9406740.
(7) Schoendorf, A.; Rithner, C. D.; Williams R. M.; Croteau, R. Proc.
Natl. Acad. Sci. U.S.A. 2001, 98, 1501-1506.
(8) Vazquez, A.; Williams, R. M. J . Org. Chem. 2000, 65, 7865-
7869.
(9) (a) Wheeler, A. L.; Long, R. M.; Ketchum, R. E. B.; Rithner, C.
D.; Williams R. M.; Croteau, R. Arch. Biochem. Biophys. 2001, 390,
265-278. (b) J ennewein, S.; Rithner, C. D.; Williams, R. M.; Croteau,
R. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 13595-13600.
(15) Barton, D. H. R.; McCombie, S. W. J . Chem. Soc., Perkin Trans.
1 1975, 1574-1585.
4902 J . Org. Chem., Vol. 67, No. 14, 2002