Efficient formal synthesis of oseltamivir phosphate (Tamiflu) with inexpensive D-ribose as the starting material

Article abstract of DOI:10.1021/ol9024716

Chemical Equation Presented An efficient formal synthesis of oseltamivir phosphate (Tamiflu) has been achieved in 12 steps with use of the inexpensive and highly abundant D-ribose as the starting material. This concise alternative route does not utilize protecting groups and features the introduction of 3-pentylidene ketal as the latent 3-pentyl ether, the use of a highly efficient RCM reaction to form the Tamiflu skeleton, and selective functional group manipulations.

Full text of DOI:10.1021/ol9024716

ORGANIC
LETTERS
2010
Vol. 12, No. 1
60-63
Efficient Formal Synthesis of
Oseltamivir Phosphate (Tamiflu) with
Inexpensive D-Ribose as the Starting
Material
Hiroshi Osato,† Ian L. Jones,‡ Anqi Chen,*,‡ and Christina L. L. Chai*,‡
Shionogi & Co., Ltd., 1-3, Kuise Terajima 2-chome, Amagasaki, Hyogo 660-0813, Japan,
and Institute of Chemical and Engineering Sciences (ICES), Agency for Science,
Technology and Research (A*STAR), 1 Pesek Road, Jurong Island, Singapore 627833
chen_anqi@ices.a-star.edu.sg; christina_chai@ices.a-star.edu.sg
Received October 27, 2009
ABSTRACT
An efficient formal synthesis of oseltamivir phosphate (Tamiflu) has been achieved in 12 steps with use of the inexpensive and highly abundant
D-ribose as the starting material. This concise alternative route does not utilize protecting groups and features the introduction of 3-pentylidene
ketal as the latent 3-pentyl ether, the use of a highly efficient RCM reaction to form the Tamiflu skeleton, and selective functional group
manipulations.
The recent global pandemic of AH1N1 influenza has resulted
in over 480 000 diagnosed infected cases and more than
6 000 recorded related deaths to date.1 This pandemic
highlights the critical importance of the development and
manufacture of anti-influenza drugs to safeguard public
health in the world. To date, oseltamivir phosphate (Tamiflu,
1) remains the most widely used antiviral drug for the
treatment and prevention of influenzas, with many countries
worldwide stockpiling this drug for the treatment of AH1N1
infections. This, together with the threat of avian and other
seasonal influenzas, has inevitably increased the demand of
Tamiflu, putting pressure on the supply of this drug and its
raw materials. As a result, unabated efforts have been devoted
to the synthesis of 1 and to date, 26 syntheses and/or
modifications have been reported.2 These syntheses, which
have been extensively reviewed recently,3 highlight the
ingenuity of synthetic chemists in devising diverse ap-
proaches toward this important synthetic target.
Many of the reported synthetic routes2f-w focused on the
use of alternative starting materials as there was a potential
supply issue of (-)-shikimic acid used in the current
manufacturing process.4,5 In line with these efforts, our work
has been directed toward the development of an alternative
synthetic route using an inexpensive, reliable, and abundant
source of starting material. A search of the chiral pool
revealed that D-ribose (2), produced on a multithousand ton
scale at the cost of ca. $30/kg,6 can be an attractive alternative
starting material for the synthesis of 1. Herein, we report
our preliminary results on an efficient formal synthesis of
Tamiflu (1) via the advanced aziridine intermediate 3,2b using
D-ribose (2) as the starting material.
† Shionogi & Co., Ltd.
‡ Institute of Chemical and Engineering Sciences.
(1) Statistics obtained from the World Health Organisation (WHO)
website: http://www.who.int/csr/disease/swineflu/updates/en/. The actual
infected cases are far more than what have been reported as many countries
no longer test and report individual cases. Figures were updated as of Nov
1, 2009.
Our strategies would (Figure 1) involve the transformation
of D-ribose (2) to the intermediate aldehyde 4, which would
undergo a known anti-selective Reformatsky-type allylation
10.1021/ol9024716  2010 American Chemical Society
Published on Web 11/25/2009

methanolic hydrochloric solution, providing the ketal 7
in an initial 54% yield. The reaction was significantly
improved to 89% yield by the addition of trimethyl
orthoformate as a dehydrating reagent.8 The alcohol 7 was
then converted to the iodoribose derivative 8 in 90% yield
by a known method (Ph3P/I2/imidazole).9
Scheme 1. Synthesis of 5-epi-Shikimic Acid Derivative 6
Figure 1. Retrosynthetic analysis of oseltamivir phosphate (1) from
D-ribose (2) via the aziridine intermediate (3).
with ethyl 2-(bromomethyl)acrylate,7 leading to the diene
5. This would then be cyclized by a ring-closing olefin
metathesis (RCM) to provide the 5-epi-shikimic acid deriva-
tive 6, which is to be converted to the aziridine 3 that leads
to 1 via known methodologies.2b
The synthesis started with D-ribose (2) whose syn diol
was converted to its 3-pentylidene ketal (Scheme 1), which
served as the latent 3-pentyl ether moiety present in 1.
This was achieved by heating 2 with 3-pentanone in a
(2) (a) Kim, C. U.; Lew, W.; Williams, M. A.; Liu, H.; Zhang, L.;
Swaminathan, S.; Bischofberger, N.; Chen, M. S.; Mendel, D. B.; Tai, C. Y.;
Laver, W. G.; Stevens, R. C. J. Am. Chem. Soc. 1997, 119, 681. (b) Rohloff,
J. C.; Kent, K. M.; Postich, M. J.; Becker, M. W.; Chapman, H. H.; Kelly,
D. E.; Lew, W.; Louie, M. S.; McGee, L. R.; Prisbe, E. J.; Schultze, L. M.;
Yu, R. H.; Zhang, L. J. Org. Chem. 1998, 63, 4545. (c) Federspiel, M.;
Fisher, R.; Henning, M.; Mair, H. J.; Oberhauser, T.; Rimmler, G.; Albiez,
T.; Bruhin, J.; Estermann, H.; Gandert, C.; Gockel, V.; Gotzo, S.; Hoffmann,
U.; Huber, G.; Janatsch, G.; Lauper, S.; Rockel-Stabler, O.; Trussardi, R.;
Zwahlen, A. G. Org. Process Res. DeV. 1999, 3, 266. (d) Karpf, M.;
Trussardi, R. J. Org. Chem. 2001, 66, 2044. (e) Harrington, P. J.; Brown,
J. D.; Foderaro, T.; Hughes, R. C. Org. Process Res. DeV. 2004, 8, 86. (f)
Cong, X.; Yao, Z.-J. J. Org. Chem. 2006, 71, 5365. (g) Yeung, Y.-Y.; Hong,
S.; Corey, E. J. J. Am. Chem. Soc. 2006, 128, 6310. (h) Fukuta, Y.; Mita,
T.; Fukuda, N.; Kanai, M.; Shibasaki, M. J. Am. Chem. Soc. 2006, 128,
6312. (i) Shie, J.-J.; Fang, J.-M.; Wang, S.-Y.; Tasi, K.-C.; Cheng, Y.-
S. E.; Yang, A.-S.; Hsiao, S.-C.; Su, C.-Y.; Wong, C.-H. J. Am. Chem.
Soc. 2007, 129, 11892. (j) Satoh, N.; Akiba, T.; Yokoshima, S.; Fukuyama,
T. Angew. Chem., Int. Ed. 2007, 46, 5734. (k) Yamatsugu, K.; Kamijo, S.;
Suto, Y.; Kanai, M.; Shibasaki, M. Tetrahedron Lett. 2007, 48, 1403. (l)
Bromfield, K. M.; Graden, H.; Hagberg, D. P.; Olsson, T.; Kann, N. Chem.
Commun. 2007, 3183. (m) Mita, T.; Fukuda, N.; Roca, F. X.; Kanai, M.;
Shibasaki, M. Org. Lett. 2007, 9, 259. (n) Kipassa, N. T.; Okamura, H.;
Kina, K.; Hamada, T.; Iwagawa, T. Org. Lett. 2008, 10, 815. (o) Zutter,
U.; Iding, H.; Spurr, P.; Wirz, B. J. Org. Chem. 2008, 73, 4895. (p)
Matveenko, M.; Willis, A. C.; Banwell, M. G. Tetrahedron Lett. 2008, 49,
7018. (q) Shie, J.-J.; Fang, J.-M.; Wong, C.-H. Angew. Chem., Int. Ed. 2008,
47, 5788. (r) Trost, B. M.; Zhang, T. Angew. Chem., Int. Ed. 2008, 47,
3759. (s) Ishikawa, H.; Suzuki, T.; Hayashi, Y. Angew. Chem., Int. Ed.
2009, 48, 1304. (t) Yamatsugu, K.; Yin, L.; Kamijo, S.; Kimura, Y.; Kanai,
M.; Shibasaki, M. Angew. Chem., Int. Ed. 2009, 48, 1070. (u) Mandai, T.;
Oshitari, T. Synlett 2009, 783. (v) Oshitari, T.; Mandai, T. Synlett 2009,
787. (w) Sullivan, B.; Carrera, I.; Drouin, M.; Hudlicky, T. Angew. Chem.,
Int. Ed. 2009, 48, 4229. (x) Nie, L.-D.; Shi, X.-X. Tetrahedron: Asymmetry
2009, 20, 124. (y) Nie, L.-D.; Shi, X.-X.; Ho, K. H.; Lu, W.-D. J. Org.
Chem. 2009, 74, 3970. (z) Karpf, M.; Trussardi, R. Angew. Chem., Int. Ed.
2009, 48, 5760.
Bernet-Vasella reaction10 of 8 in THF-H2O (2:1) under
reflux provided the intermediate aldehyde 4, which was not
isolated and underwent an anti-selective Reformatsky-type
allylation with ethyl 2-(bromomethyl)acrylate in a one-pot
fashion,7 furnishing the diene 5 in 78% yield after separation
of the minor diastereomer (dr ) 5.2:1). In the presence of
the second-generation Grubbs-Hoveyda catalyst (9), the
diene 5 underwent efficient ring-closing olefin metathesis in
1,2-dichloroethane at reflux, providing 5-epi-shikimic acid
derivative 6 in nearly quantitative yield. The robustness of
catalyst 9 allowed its loading to be as low as 2 mol % and
the use of 1,2-dichloroethane (boiling point 81 °C) as the
solvent greatly facilitated the reaction, enabling the cycliza-
tion of 5 to completion within 2 h.
With the 5-epi-shikimic acid derivative 6 in hand, strate-
gies for the transformation of the 3-pentylidene ketal in 6 to
the 3-pentyl ether moiety for 1 were explored. Our initial
plan was to convert 6 to its mesylate 10, followed by
selective cleavage of the ketal group employing a highly
selective protocol (TiCl4, Et3SiH, -34 °C) reported for the
mesylate 11 in the shikimic acid route (Scheme 2).2c
(8) Look, G. C.; Murphy, M. M.; Campbell, D. A.; Gallop, M. A.
Tetrahedron Lett. 1995, 36, 2937.
(3) (a) Shibasaki, M.; Kanai, M. Eur. J. Org. Chem. 2008, 1839. (b)
Magano, J. Chem. ReV. 2009, 109, 4398.
(9) (a) Verheyden, J. P. K.; Moffatt, J. G. J. Am. Chem. Soc. 1964, 86,
2093. (b) Barrett, A. G. M.; Lebold, S. A. J. Org. Chem. 1990, 55, 3853. (c)
Gallos, J. K.; Goga, E. G.; Koumbis, A. E. J. Chem. Soc., Perkin Trans. 1
1994, 611. (d) Paquette, L. A.; Bailey, S. J. Org. Chem. Soc. 1995, 60, 7849.
(10) (a) Bernet, B.; Vasella, A. HelV. Chim. Acta 1979, 62, 1990. (b)
Nakane, M.; Hutchinson, C. R.; Gollman, H. Tetrahedron Lett. 1980, 21,
1213. (c) Fu¨rstner, A.; Jumbam, D.; Teslic, J.; Weidmann, H. J. Org. Chem.
1991, 56, 2213.
(4) The Roche Group, Tamiflu Factsheet, November, 17, 2006.
(5) (a) Enserink, M. Science 2006, 312, 382. (b) Farina, V.; Brown,
J. D. Angew. Chem., Int. Ed. 2006, 45, 7330.
(6) Wulf, P. D.; Vandamme, E. J. Appl. Microbiol. Biotechnol. 1997,
48, 141.
(7) Hyldtoft, L.; Madsen, R. J. Am. Chem. Soc. 2000, 122, 8444.
Org. Lett., Vol. 12, No. 1, 2010
61

AlCl3, it was found that anhydrous aluminum trichloride in
combination with triethylsilane12 afforded a 6:1 regioisomeric
ratio of products, in favor of the desired diol 13 in 67% yield
(Scheme 3).
Scheme 2. Reductive Ketal Cleavge of Mesylate 10
Scheme 3. Completion of the Synthesis of Azirdine 3
Unfortunately, the reaction afforded a poor selectivity (12a:
12b ) 1:3) for the desired product 12a despite a quantitative
combined yield. A variety of other Lewis acids, including
BF3, AlCl3, and ZnCl2, and conditions were investigated, but
no significant improvement was observed in terms of desired
regioselectivity.
The dramatic difference in the regioselectivity of ketal
cleavage between mesylate 10 and 11 promoted us to look
into the likely causes. Conformation analysis11 revealed that
the syn-mesylate 10 (Figure 2. A) is sterically much more
Having achieved the crucial regioselective cleavage of the
ketal group, selective mesylation of the diol 13 was
examined. Although the formation of the dimesylate is a
potential competing reaction,2x-z we envisioned that it would
be possible to selectively mesylate the least sterically
hindered 5-hydroxy group of 13. To our delight, mesylation
of 13 carried out at -20 °C with 1.05 equiv of methane-
sulfonyl chloride with triethylamine as the base provided the
monomesylate 12a as the sole product in 92% yield.
Treatment of 12a with trifluoromethanesulfonic anhydride
in the presence of pyridine at -10 °C afforded the mesyloxy
triflate 14, which was used in the next step of the synthesis
without purification due to its instability.
To transform 14 to the aziridine 3 requires the selective
displacement of the more reactive but also more hindered
4-triflate in the presence of the 5-mesylate group. This was
achieved by treatment of 14 with sodium azide in acetone-
water (9:1) providing the 4-azidomesylate 15 without notice-
able formation of either the 5-azide or diazide. This high
regioselectivity is attributed to the orientation of the trifloxy
and mesyloxy groups in the preferred conformation2o in
which the 4-trifloxy adopts a favored axial position for the
azide substitution reaction whereas the 5-mesyloxy occupies
a disfavored equatorial position (Scheme 3).
Figure 2. Energy minimized conformations of mesylate 10 (A) and
11 (B).
congested than the anti compound 11 (Figure 2. B) between
the mesyloxy and the ketal group. This steric hindrance
imposed by the syn alignment of mesylate in relation to the
ketal moiety in 10 would have hampered the coordination
of the Lewis acid to the adjacent oxygen atoms of the
mesylate and the ketal, resulting in the poor selectivity.
In light of this conclusion, it was postulated that the
cleavage of the ketal moiety could be better controlled with
alcohol 6, which is sterically less hindered than 10, and the
hydroxy group is likely to react with the Lewis acid,
facilitating its desired coordination to the adjacent ketal
oxygen thus leading to better selectivity. After screening
several Lewis acids including TiCl4, BF3-OEt2, BH3, and
(11) The energy minimized conformations were obtained by using
the MMFF94 force field, Spartan 04 version for Windows; Wavefunction,
Inc.
(12) Crimmins, M. T.; Rafferty, S. W. Tetrahedron Lett. 1996, 37, 5649.
Org. Lett., Vol. 12, No. 1, 2010
62

To complete the synthesis, a final two-step, one-pot
sequential transformation of the azide 15 to 3 was carried
out by reduction of the azide to the corresponding amine
via a Staudinger reaction13 followed by a triethylamine-
mediated cyclization to form the aziridine 3 in 84% yield.
This concluded an efficient synthesis of this advanced
intermediate in nine steps in 28% overall yield from 2. As
the conversion of 3 to 1 has been reported previously in three
steps,2b this work constitutes a concise formal synthesis of
1 in 12 steps (Scheme 3).
In summary, we have accomplished an efficient synthesis
of the key aziridine intermediate 3 for the synthesis of
Tamiflu, 1, using the inexpensive and abundant D-ribose as
the starting material. This concise alternative route does not
utilize protecting groups and features the introduction of
3-pentylidene ketal as the latent 3-pentyl ether present in 1,
a highly efficient RCM reaction to form the cyclohexene
skeleton, and selective functional group manipulations. This
route demonstrates the potential of D-ribose as an alternative,
inexpensive, and renewable raw material for the synthesis
of Tamiflu, 1.
Acknowledgment. The authors are grateful to the Agency
for Science, Technology and Research (A*STAR), Singapore
and Shionogi & Co., Ltd. (H.O.) for the financial support of
this work.
Note Added after ASAP Publication. An error was
corrected in Scheme 1 in the version published ASAP
December 7, 2009.
Supporting Information Available: Experimental pro-
cedures and full spectroscopic data for all compounds. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(13) Gololobov, Y. G.; Kasukhin, L. F. Tetrahedron 1992, 48, 1353,
and references cited therein.
OL9024716
Org. Lett., Vol. 12, No. 1, 2010
63