A concise synthesis of fumagillol

Article abstract of DOI:10.1055/s-2001-13359

A concise synthesis of fumagillol was accomplished in twelve steps starting from cyclohexadiene, proceeding by way of ring-opening of a symmetrical protected epoxydiol with a cuprate reagent and subsequent protecting group manipulation to give a key keton

Full text of DOI:10.1055/s-2001-13359

LETTER
661
A Concise Synthesis of Fumagillol
Martin Hutchings,a David Moffatb Nigel S. Simpkinsa*
a The School of Chemistry, The University of Nottingham, University Park, Nottingham, NG7 2RD U.K.
b Celltech-Chiroscience, 216 Bath Road, Slough, SL1 4EN U.K.
E-mail: nigel.simpkins@nottingham.ac.uk
Received 6 February 2001
O
O
Abstract: A concise synthesis of fumagillol was accomplished in
twelve steps starting from cyclohexadiene, proceeding by way of
ring-opening of a symmetrical protected epoxydiol with a cuprate
reagent and subsequent protecting group manipulation to give a key
ketone intermediate. Installation of the required epoxide functions
was achieved by addition of chloromethyllithium to the ketone and
by directed epoxidation of the unsaturated side-chain.
O
H
H
1
3
OMe
O
4
O
OH
4
5
C-1 epoxide
installation
Key words: fumagillol, stereoselective synthesis, epoxides, cuprate
M
OP
7
C-2 ring-opening
O
Fumagillin (1) (Figure 1), a fungal metabolite isolated in
1951 from certain strains of Aspergillus Fumagatus, was
originally noted for its antibiotic and antiparasitic activi-
ty.1 More recently there has been a resurgence of interest
in fumagillin and its relatives, such as ovalicin (2), due to
the discovery that these compounds exhibit potent angio-
genesis inhibitory properties.2 Indeed, the semisynthetic
derivative TNP 470 (3), an unnatural ester of the core cy-
clohexanol 4, known as fumagillol, is in late clinical trials
as an antitumour agent.3
OP
C-3, C-4 protection
6
Scheme 1
We envisaged that the synthesis of fumagillol (4) would
be completed by installation of the required spiro-epoxide
function into ketone 5, followed by acetonide removal and
selective side-chain epoxidation and C-3 alcohol methyla-
tion. The last two steps had been achieved in previous
syntheses4,5b and there was good precedent for direct con-
version of ketones related to 5 into the desired epoxide
function.6 Ketone 5 would be available via ring-opening
of an epoxide 6 (where P represents an appropriate pro-
tecting group) by a suitable organometallic alkenyl 7.
O
O
O
H
O
H
OH
OMe
OMe
O
OR
CO2H
1 R =
3 R =
(
)
2
4
This plan was consciously devised around the opening of
a meso-epoxide taking into account the prospect of an
asymmetric ring-opening by use of emerging methodolo-
gy.7 Synthesis of a suitable starting epoxide was readily
accomplished by oxidation of cyclohexadiene according
to the method of Backvall (Scheme 2).8
O
H
N
Cl
O
O
4 R = H
Figure
1. Pd(OAc)2 5 mol%
LiCl, LiOAc, MnO2
benzoquinone
Although the original landmark synthesis of 1 by Corey
and Snider was achieved over twenty-five years ago,4 the
renewed interest in this family of compounds has sparked
fresh synthetic activity, resulting in several new success-
ful syntheses.5 Herein we describe our own efforts in this
area, which have resulted in a new concise synthesis of
fumagillol (4), which proceeds by ring-opening of a sym-
metrical epoxide intermediate.
OH
AcOH, pentane (93%)
2. K2CO3, MeOH
(43%)
OH
8
9
OPMB
1. mCPBA
EtOAc, Et2O (67%)
O
Our synthetic planning was governed by a requirement for
a flexible route to fumagillol that would employ simple
starting materials, and that might provide a range of side-
chain analogues. The key aspects of the retrosynthetic
analysis are indicated in Scheme 1.
2. NaH, PMBCl (85%)
OPMB
10
Scheme 2
Synlett 2001 No. 5, 661–663 ISSN 0936-5214 © Thieme Stuttgart · New York

662
M. Hutchings et al.
LETTER
Palladium-catalysed bis-acetoxylation of cyclohexadiene of alcohol 13 to the corresponding ketone 5 was reason-
8 was very efficient, the overall yield to diol 9 being com-
promised only by its water solubility, which hampered its
isolation. Nevertheless, this method readily supplied us
with 30g batches of 9.
ably efficient by use of either the Dess–Martin periodi-
nane or typical Swern (DMSO, (COCl)2, Et3N) conditions
(Scheme 4).
Directed epoxidation using mCPBA, according to Huang
et al.,9 gave the required syn epoxide intermediate in good
yield, and subsequent diol protection was straightforward
to give 10. Preliminary work indicated that the para-
methoxybenzyl (PMB) series of compounds was prefera-
ble to either the benzyl series or silicon protected deriva-
tives.
O
H
Swern or
Dess-Martin
13
(73-77%)
O
O
5
O
The key epoxide ring-opening reaction was greatly sim-
plified by the availability of a protocol for the direct gen-
eration of the required alkenyllithium corresponding to
7.10 Thus, metallation of the acetone trisylhydrazone 11
and addition of prenyl bromide, followed by addition of
further base and warming to 0 C, effected a one-pot alky-
lation/Shapiro elimination to give the required alkenyl-
lithium (Scheme 3).
sulfoxonium methylide
see text
OH
14
Scheme 4
However, the use of sulfoxonium methylide to convert ke-
tone 5 into the required epoxide, although precedented on
a very similar intermediate,6 did not proceed as expected.
Instead we observed rapid conversion of 5 into the hy-
droxycyclopropane 14, which was isolated as a mixture of
diastereomers.
Br
1. BuLi,
SO2Ar
NH
N
2. BuLi, TMEDA, 0 °C
11
PMBO
Li
H
This unwanted transformation clearly occurs by -elimi-
nation to give an intermediate enone, which then under-
goes cyclopropanation. We were unable to divert the path
of this process under a range of conditions and so sought
an alternative method for converting 5 into the desired ep-
oxide. Fortunately reaction of ketone 5 with an excess of
chloromethyllithium reagent, prepared as described by
Sadhu and Matteson,12 resulted in the predominant forma-
tion of chlorohydrin 15 (18:1), accompanied by only very
minor amounts of the other diastereomer (Scheme 5).
7 M = Li
10
(2-Th)Li(CN)Cu
room temperature (70%)
OH
OPMB
12
1. NH3, Na (15 equiv.)
tBuOH (5 equiv.) (91%)
HO
H
2. dimethoxypropane
PPTS, acetone (92%)
O
O
Ar = 2,4,6-triisopropylphenyl
13
Scheme 3
Cl
ClCH2I (5 equiv.)
nBuLi, -80 °C (5 equiv.)
HO
This was then mixed with commercial lithium(2-thie-
nyl)cyanocuprate and the so-formed cuprate reagent re-
acted with the epoxide 10. Little or no reaction was
observed at 0 C (the temperature used previously for re-
action with alkyl iodides), but it was found that slow
warming of the reaction mixture to room temperature
overnight gave reproducible results.
5
(64%) (16:1)
O
O
15
two
O
known
steps
1. 2M HCl
THF
4
We believe that this is the first time that this protocol has
been applied to epoxide opening reactions, and the pres-
ence of flanking ether functions in epoxide 10 makes this
a very demanding example.
2. NaOH
EtOH
OH
OH
(70%) overall
16
Alcohol deprotection was found to be relatively ineffi- Scheme 5
cient using DDQ, TFA, CAN or hydrogenolysis, but
pleasingly the use of Birch conditions giving excellent re-
sults.11 Having generated the corresponding triol from 12
it proved straightforward to generate the required hy-
droxyacetonide 13 under standard conditions. Oxidation
Although treatment of this chlorohydrin with base formed
the required epoxide, we found that the acidic conditions
required for subsequent removal of the acetonide were in-
Synlett 2001, No. 5, 661–663 ISSN 0936-5214 © Thieme Stuttgart · New York

LETTER
A Concise Synthesis of Fumagillol
663
(2) (a) Wenert, N.; Stanjek, A.; Kiriakidis, A. H.; Jha, H. C.;
compatible with the epoxide functionality. Therefore we
first treated the chlorohydrin 15 with acid in order to re-
move the acetonide protection, and subsequently effected
epoxide formation with base to give 16.
Mazitschek, R.; Giannis, A. Angew. Chem. Int. Ed. 1999, 38,
3228. (b) Ingber, D.; Fujita, T.; Kishimoto, S.; Sudo, K.;
Kanamaru, T.; Brem, H.; Folkman, J. Nature 1990, 348, 555.
(3) Figg, W. D.; Kruger, E. A. Expert. Opin. Investig. Drugs
2000, 9, 1383.
(4) Corey, E. J.; Snider, B. B. J. Am. Chem. Soc. 1972, 94, 2549.
(5) (a) Kim, D.; Ahn, S. K.; Bae, H.; Choi, W. J.; Kim, H. S.
Tetrahedron Lett. 1997, 38, 4437. (b) Vosburg, D. A.; Weiler,
S.; Sorensen, E. J. Angew. Chem. Int. Ed. 1999, 38, 971. (c)
Taber, D. F.; Christos, T. E.; Rheingold, A. L.; Guzei, I. A. J.
Am. Chem. Soc. 1999, 121, 5589.
(6) Amano, S.; Ogawa, N.; Ohtsuka, M.; Ogawa, S.; Chida, N.
Chem. Commun. 1998, 1263. See also reference 5c.
(7) (a) For a general review of desymmetrisation, see Willis, M.
C. J. Chem. Soc., Perkin Trans. 1 1999, 1765. (b) For a review
dealing with epoxide desymmetrisation, see Hodgson, D. M.;
Gibbs, A. R.; Lee, G. P. Tetrahedron 1996, 52, 14361. (c)
Oguni, N.; Miyagi, Y.; Itoh, K. Tetrahedron Lett. 1998, 39,
9023. (d) Alexakis, A.; Vrancken, E.; Mangeney, P. Synlett
1998, 1165.
(8) Backvall, J. E.; Bystrom, S. E.; Nordberg, R. E. J. Org. Chem.
1984, 49, 4619.
(9) Huang, C.; Cabell, L. A.; Anslyn, E. V. Synth. Commun. 1994,
24, 2757.
(10) (a) Corey, E. J.; Lee, J.; Roberts, B. E. Tetrahedron Lett. 1997,
38, 8915. (b) Corey, E. J.; Dittami, J. P. J. Am. Chem. Soc.
1985, 107, 256. See also reference 5b.
(11) Reist, E. J.; Bartuska, V. J.; Goodman, L.; J. Org. Chem. 1964,
29, 3725.
Epoxydiol 16 is a known precursor for fumagillol by the
recent Sorensen synthesis, and our sample proved identi-
cal to that described earlier. We were also able to repeat
the final two steps required to generate fumagillol, these
being (i) directed epoxidation to install the side-chain ep-
oxide (ii) selective methylation of the more reactive C-3
hydroxyl function. The so-formed fumagillol 4 displayed
spectroscopic characteristics entirely consistent with
those described previously.
In conclusion, we have described a very concise new route
to fumagillol, which should be amenable to the prepara-
tion of large quantities of this product and related materi-
als. An attractive aspect of our approach is the potential
for asymmetric induction in the ring-opening of key ep-
oxide 10. Although established protocols for such an
asymmetric opening by an organometallic alkenyl are
lacking at present,13 this synthesis should act as a further
spur to activity in this area.
Acknowledgement
We are grateful to Celltech-Chiroscience for funding of this work.
We also thank Professor E. J. Sorenson of the Scripps Research In-
stitute for sending us useful spectroscopic data.
(12) Sadhu, K. M.; Matteson, D. S. Tetrahedron Lett. 1986, 27,
795.
(13) (a) Alexakis, A.; Vrancken, E.; Mangeney, P. Synlett 1998,
1165. (b) Oguni, N.; Miyagi, Y.; Itoh, K. Tetrahedron Lett.
1998, 39, 9023. (c) Mizuno, M.; Kanai, M.; Iida, A.; Tomioka,
K. Tetrahedron 1997, 53, 10699.
References and Notes
(1) (a) Hanson, T. E. Antibiot. Chemother. (Washington D. C.)
1951, 1, 54. (b) McCowen, M. C.; Callender, M. E.; Lawliss,
J. F. Science 1951, 113, 202. (c) Katznelson, H.; Jamieson, C.
A. Science 1952, 115, 70. (d) Killough, J. H.; Magill, G. B.;
Smith, R. C. Science 1952, 115, 71.
Article Identifier:
1437-2096,E;2001,0,05,0661,0663,ftx,en;D03401ST.pdf
Synlett 2001, No. 5, 661–663 ISSN 0936-5214 © Thieme Stuttgart · New York