Stereospecific synthesis of the major human metabolite of paclitaxel

Article abstract of DOI:10.1016/S0040-4039(00)00707-3

The stereospecific synthesis of 6-α-hydroxy paclitaxel 10, the major human metabolite of paclitaxel, is described. The 6,7-α-diol 4, obtained from paclitaxel, is converted to the 6,7-β-cyclic sulfate followed by nitrate addition and reduction to afford the title compound. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4039(00)00707-3

Tetrahedron Letters 41 (2000) 4729±4731
Stereospeci®c synthesis of the major human metabolite
of paclitaxel
Mark D. Wittman,* John F. Kadow and Dolatrai M. Vyas
Bristol-Myers Squibb Pharmaceutical Research Institute, Division of Chemistry, PO Box 5100, Wallingford,
CT 06492-7660, USA
Received 8 November 1999; revised 21 April 2000; accepted 24 April 2000
Abstract
The stereospeci®c synthesis of 6-a-hydroxy paclitaxel 10, the major human metabolite of paclitaxel, is
described. The 6,7-a-diol 4, obtained from paclitaxel, is converted to the 6,7-b-cyclic sulfate followed by
nitrate addition and reduction to a€ord the title compound. # 2000 Elsevier Science Ltd. All rights reserved.
Keywords: taxoids; diols; oxidation; nitrates; inversion reactions; metabolites.
The expanding clinical use of TAXOL1 in cancer chemotherapy has fueled an aggressive search
within the medicinal chemistry community for second-generation taxanes with improved safety,
ecacy, and pharmacokinetics. We required a supply of 6-a-hydroxy paclitaxel 10, the major
human metabolite of paclitaxel,1 to validate our in vitro assays for the human metabolism of
paclitaxel and to evaluate analogs for improved metabolic stability.
Initially, we focused on the epimerization of the silylated a-diol 1 (4, TESCl, imidazole,
CH2Cl2) (Scheme 1). Treatment with KHMDS at 0ꢀC gave a mixture of epimers at C-7, which
could not readily be separated. The mixture was then deprotected to give the desired 6-a-hydroxy
paclitaxel 3 in 9% yield with 26% of the deprotected a-diol 2. Due to the low yield of product
and poor recovery of 2, we did not pursue a base-induced epimerization of this substrate. King-
ston2 has reported the epimerization of the 20-OTBS ether derivative of diol 2 using DBU to
obtain the 20-OTBS 6-a-hydroxy paclitaxel in 15% yield along with recovered starting diol.
Alternatively, we found that if the 6-a-hydroxyl group of 4 was protected as a phenoxyacetate
or acetate, a silylative epimerization could be a€ected using TESCl and imidazole at elevated
temperatures to a€ord derivatives 5. This process is likely driven by the fact that the C-7-a
hydroxyl is too hindered to react with silylating agents so that any epimerization is quickly trapped
with silyl chloride. Various attempts to remove the esters (K2CO3, MeOH; KCN, MeOH; Ti(OEt)4,
EtOH) from 5 resulted in removal of both the protecting group and silyl group and re-epimerization
* Corresponding author.
0040-4039/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(00)00707-3

4730
Scheme 1. Reagents and conditions: (a) KHMDS (1.3Â), THF 0ꢀC to rt; 1N HCl, CH3CN 0ꢀC; (b) RCOCl, Et3N,
CH2Cl2; TESCl, imidazole, DMF 65ꢀC; (c) K2CO3, MeOH; or KCN, MeOH; or Ti(OEt)4, EtOH; or H2, Pd±C, MeOH
back to the 20-O-TES-a-diol 4. The C-6-hydroxyl group could be protected with a TES group but
silylative epimerization yielded only recovered starting material. The C-6 hydroxyl could be pro-
tected as the Cbz derivative and C-7 hydroxyl group epimerized and silylated but upon removal
of the Cbz group the starting diol was again obtained. Given the diculty of e€ectively epimer-
izing the C-7-hydroxyl group, we opted for an alternative synthetic route which could process the
gram quantities our studies required and possibly provide access to other C-6-a analogs.3
Our strategy then turned to conversion of the a-diol 4 to the C-6 ketone (Scheme 2). Several
oxidants were e€ective at oxidizing the C-6 hydroxyl (TPAP, NMO; Jones; Bu2SnO, Br2; 4-acetoxy-
TEMPO, pTsOH) but oxidation with catalytic 4-benzoyloxy-TEMPO and bleach4 proved to be
optimal. The crude ketone 6 was a single isomer but upon silica gel chromatography a mixture of
6 and the 7-epimerized ketone 7 was evident. Epimerization of the 7-hydroxyl could be driven
simply by stirring crude 6 in methylene chloride with an equal weight of silica gel. Once the
ketone 7 was obtained, we envisioned a directed reduction using sodium triacetoxyborohydride to
provide the metabolite. In the event, reduction provided only the b-diol 8. Other reducing reagents
provided varying mixtures of diols favoring the b-diol (NaBH4, Dibal). The facial selectivity
obtained in this reduction arises from the pseudo-equatoral orientation of the C-7 hydroxyl
resulting in hydride delivery from the more accessible a-face. In practice, the silica gel was ®ltered
o€ and the crude ketone reduced with sodium triacetoxyborohydride in acetic acid/acetonitrile to
give the 6,7-b-diol 8 in 75% yield from 1 without puri®cation of the intermediates. The high yield of
this three-step process, and the ease with which this process is scaled up, makes this longer route a
preferable one for preparing large amounts of the metabolite. The cyclic sulfate was prepared
according to the method of Sharpless5 and opened with tetrabutylammonium nitrate6 at the
6-position. Hydrolysis of the sulfate a€orded nitrate 9. Hydrogenolysis of the nitrate provided 10,
which was identical to 6-a-hydroxy paclitaxel by 500 MHz NMR.1 Incubation of paclitaxel
with a human liver S9 preparation provided material identical to 10 by HPLC.

4731
Scheme 2. Reagents and conditions: (a) 4-BzO-TEMPO (2%), KBr, Chlorox1; (b) silica gel, CH2Cl2; (c) Na(OAc)3BH,
HOAc, CH3CN (75% overall yield for a±c); (d) SOCl2, Et3N; NaIO4, RuCl3, CCl4/CH3CN/H2O (77%); (e) Bu4NNO3,
PhMe, 100ꢀC (76%); (f) H2, Pd/C, MeOH (85%)
In conclusion, we have presented a stereospeci®c synthesis of 6-a-hydroxy paclitaxel that
quickly inverts 6,7-a-diol 1 to the b-diol 8. The 6-a-hydroxyl can then be introduced via an
opening of the 6,7-b-cyclic sulfate.
Acknowledgements
The authors wish to thank Jay O. Knipe for the human liver S9 incubation of paclitaxel and
LC comparison with 10, and the Analytical Research and Development Department for NMR,
IR, MS and C, H and N analyses.
References
1. Harris, J. W.; Katki, A.; Anderson, L. W.; Chmurny, G. N.; Paukstelis, J. V.; Collins, J. M. J. Med. Chem. 1994,
37, 706±709.
2. Yuan, H.; Kingston, D. G. I. Tetrahedron Lett. 1998, 39, 4967±4970.
3. The use of this chemistry to provide C-6 derivatives will be presented in due course.
4. Anelli, P. L.; Bi, C.; Montanari, F.; Quici, S. J. Org. Chem. 1987, 52, 2559±2562.
5. Gao, Y.; Sharpless, K. B. J. Am Chem. Soc. 1988, 110, 7538±7539.
6. Cainelli, G.; Manescalchi, F.; Martelli, G.; Panunzio, M.; Plessi, L. Tetrahedron Lett. 1985, 26, 3369±3372.