A relay route for the synthesis of azadirachtin

Article abstract of DOI:10.1002/anie.200703028

(Chemical Equation Presented) 22 Years in the making: Azadirachtin (1) was synthesized for the first time by a highly convergent approach, utilizing a Claisen rearrangement and a radical cyclization as key steps. End-game strategies relied on intermediate 2, which could be obtained by synthetic methods as well as by degradation of 1. Bn = benzyl, TBS = tert- butyldimethylsilyl.

Full text of DOI:10.1002/anie.200703028

Angewandte
Chemie
DOI: 10.1002/anie.200703028
Total Synthesis
A Relay Route for the Synthesis of Azadirachtin**
Gemma E. Veitch, Edith Beckmann, Brenda J. Burke, Alistair Boyer, Carles Ayats, and
Steven V. Ley*
The study of a plantꢀs natural defence mechanisms against
predatory insect attack often leads to the discovery of novel
molecules that have important biological effects. Such is the
case with azadirachtin (1), a fascinating natural product first
Azadirachtin (1) represents an exceptionally challenging
synthetic target by virtue of its sixteen contiguous stereogenic
centers and complex pattern of oxygen-containing function-
alities. The conformation and reactivity of azadirachtin are
strongly influenced by the presence of intramolecular hydro-
gen-bonding interactions, evident from X-ray crystallographic
studies.[6,25] In addition, both the hemiacetal at C11 and the
À
C22 C23 enol ether are sensitive functionalities and are
therefore important considerations in the design of any total
synthesis.
À
Potential strategies for the masking of the C22 C23 enol
ether had been investigated previously,[21,22] where it was
found that a C23 methyl acetal was particularly stable in
subsequent transformations. Although this derivative was
accessed as an epimeric mixture, further studies[21] employed
only the b-C23 epimer. We anticipated that this might present
a severe disadvantage in our forward synthesis, as it could be
difficult to control this stereogenic center. Work reported
herein addresses this problem by providing a new relay route
that can utilize both epimeric forms (Scheme 1).
Accordingly, preparation of relay target 4 commenced
with a methoxybromination/reduction protocol, thus convert-
ing the reactive enol ether present in 1 to a 1:1 mixture of its
C23 a and b methyl acetals. Silver oxide mediated alkylation
then permitted selective protection of the C11 and C20
hydroxy groups providing dibenzyl ether 2.[26] Oxidation of
the remaining C7 hydroxy group present in 2, followed by
saponification of the C12-methylcrotonoyl (tigloyl) and C3
acetyl esters afforded intermediate 3. At this stage the C23
diastereomers were separated and the individual epimers
converted to their C3 silyl ethers (a-4 and b-4).[27]
Attention was then turned towards the synthesis of
azadirachtin from a-4 and b-4, in order to establish these
silyl ethers as our new targets for total synthesis (Scheme 2).
TBAF-mediated desilylation of a-4 and b-4 followed by
selective hydrogenolysis of the C20 benzyl ether generated
triols a-5 and b-5 in excellent yield.[28] As the C23 b epimer of
intermediate 5 had been converted to azadirachtin in a first-
generation relay approach,[21] it was desirable to epimerize the
C23 a epimer in order to link our new route to these previous
relay studies. Consequently, the a epimer (a-5) was treated
with Amberlyst15 and methanol in acetonitrile resulting in a
1:1 mixture of diastereomers, from which the desired epimer
(b-5) could be isolated cleanly.
isolated from the Indian neem tree Azadiracta indica
(A. Juss), in 1968.[1] Azadirachtin exhibits potent antifeedant
and growth-disruptant properties against a broad spectrum of
insect species yet displays very low mammalian toxicity
(LD50(rat) > 5 gkgÀ1) and appears to cause little disruption to
beneficial species such as pollinating bees and ladybirds.[2]
The structure of this complex natural product was finally
elucidated following many years of intense research[1,3–11]
culminating in the publication of three back-to-back full
papers in 1987.[7–9] Likewise, understanding of the precise
mode of action of azadirachtin continues to evolve. Detailed
studies have revealed an elaborate set of interactive pathways,
the full discussion of which is beyond the scope of this short
communication.[12]
In preparation for the total synthesis and to define the
structure–activity relationships of azadirachtin,[13–17] we have
been engaged in an extensive investigation of its reactivity
pattern,[18] in particular its propensity for rearrangement
under acidic[19] basic[13] and photolytic[20] conditions. Based on
our cumulated experience in this regard,[21,22] we have devised
a new relay route that has proved instrumental for the
synthesis of azadirachtin.[23,24]
[*] G. E. Veitch, Dr. E. Beckmann, Dr. B. J. Burke, A. Boyer, Dr. C. Ayats,
Prof. Dr. S. V. Ley
Department of Chemistry
University of Cambridge
Lensfield Road, Cambridge, CB2 1EW (UK)
Fax: (+44)1223-336-442
E-mail: svl1000@cam.ac.uk
Homepage: http://leygroup.ch.cam.ac.uk
Selective acetylation at the C3 position of triol b-5 was
then achieved under standard conditions to provide acetate 6
in good yield (Scheme 3). However, esterification of the
hindered C1hydroxy group proved more problematic. Awide
variety of reagents were screened for this pivotal trans-
formation, most of which proved unsuccessful. Our previously
[**] We gratefully acknowledge the many co-workers who have worked
on this project and the EPSRC (G.E.V. and A.B.), Alexander von
Humboldt Foundation (E.B.), and Novartis (S.V. L.) for generous
funding.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2007, 46, 7633 –7635
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
7633

Communications
Scheme 1. Degradation of azadiractin (1) to relay target 4. Reagents
and conditions: a) NBS, MeOH, RT, quant.; b) Bu3SnH, AIBN, toluene,
reflux, 91%; c) BnBr, Ag2O, DMF, RT, 60%; d) DMP, CH2Cl2, RT, 99%
e) Et3N/MeOH/H2O (1:5:1), 658C, 99%; f) TBSOTf, iPr2NEt, CH2Cl2,
À788C, 90%. AIBN=azobisisobutyronitrile, Bn=benzyl, DMP=Dess–
Martin periodinane, MS=molecular sieves, NBS=N-bromosuccini-
mide, TBSOTf=tert-butyldimethylsilyltrifluoromethanesulfonate.
Scheme 3. Return of 5 to azadirachtin (1). Reagents and conditions:
a) Ac2O, Et3N, DMAP, CH2Cl2, RT, 74%; b)CH3CH-
(CH3)(CO)O(CO)C6H2Cl3, Cs2CO3, toluene, reflux, 6 d, 50% (80%
based on recovered starting material); c) NaBH4, CeCl3·7H2O, MeOH,
08C, 49%; d) Pd/C, H2, MeOH, RT, 81%; e) PhSH, PPTS, CH2ClCH2Cl,
808C, 70%; f) DMDO, CH2Cl2, À788C–RT, then toluene, reflux, 67%.
DMAP=4-dimethylaminopyridine, DMDO=dimethyldioxirane,
PPTS=pyridinium toluene-p-sulfonic acid.
described relay route utilized the highly reactive tigloyl
fluoride to effect formation of 7, albeit in poor yield (19%).[21]
Following extensive optimization studies, we are now able to
obtain tiglate 7 in 50% yield by treatment of 6 with a tiglic
acid/Yamaguchi mixed anhydride[29] and Cs2CO3 in refluxing
toluene. Pleasingly, unreacted starting material can be
recovered from this reaction and further converted to the
desired product.
The next hurdle in our synthesis was the stereoselective
reduction of the C7 ketone present in 7 to provide the
requisite axial alcohol found in the natural product. Owing to
the high degree of steric hindrance about this C7 center, it was
not possible to employ bulky reducing reagents such as L-
Selectride. Matters were complicated further by the presence
Scheme 2. Return of a-4 and b-4 to azadirachtin (1). Reagents and
conditions: a) TBAF, THF, 08C, a: quant., b: quant.; b) Pd/C, 10 bar
H2, MeOH, RT, a: 99%, b: 85%; c) Amberlyst15, 3 Š MS, MeOH,
MeCN, RT, 49%. TBAF=tetra-n-butylammonium fluoride.
7634
www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 7633 –7635

Angewandte
Chemie
[10] S. V. Ley, H. Lovell, D. J. Williams, J. Chem. Soc. Chem.
of the C1tigloyl ester, which was also susceptible to conjugate
reduction. Finally, it was found that Luche[30] conditions
cleanly reduced the C7 ketone, although a 1:1 mixture of
diastereomers was obtained.[21] The unwanted equatorial
alcohol could, after separation, be reoxidized and subse-
quently reduced, to give yields of 8 up to 75% after one
recycle. Cleavage of the benzyl ether of 8 then proceeded
smoothly to furnish C11 lactol 9 in excellent yield, with no
concomitant reduction of the tiglate functionality.
Commun. 1992, 1304 – 1306.
[11] S. V. Ley, K. Doherty, G. Massiot, J.-M. Nuzillard, Tetrahedron
1994, 50, 12267 – 12280.
[12] For a review see A. J. Mordue, A. Blackwell, J. Insect Physiol.
1993, 39, 903 – 924.
[13] S. V. Ley, J. C. Anderson, W. M. Blaney, E. D. Morgan, R. N.
Sheppard, M. S. J. Simmonds, A. M. Z. Slawin, S. C. Smith, D. J.
Williams, A. Wood, Tetrahedron 1991, 47, 9231– 9246.
[14] S. V. Ley, J. C. Anderson, W. M. Blaney, P. S. Jones, Z. Lidert,
E. D. Morgan, N. G. Robinson, D. Santafianos, M. S. J. Sim-
monds, P. L. Toogood, Tetrahedron 1989, 45, 5175 – 5192.
[15] M. S. J. Simmonds, W. M. Blaney, S. V. Ley, J. C. Anderson, P. L.
Toogood, Entomol. Exp. Appl. 1990, 55, 169 – 181.
[16] W. M. Blaney, M. S. J. Simmonds, S. V. Ley, J. C. Anderson, P. L.
Toogood, Entomol. Exp. Appl. 1990, 55, 149 – 160.
[17] S. V. Ley, J. C. Anderson, W. M. Blaney, P. S. Jones, L. Z. , E. D.
Morgan, N. G. Robinson, D. Santafianos, M. S. J. Simmonds,
P. L. Toogood, Tetrahedron 1989, 45, 5175 – 5192.
[18] S. V. Ley, A. A. Denholm, A. Wood, Nat. Prod. Rep. 1993, 10,
109 – 157.
[19] J. N. Bilton, P. S. Jones, S. V. Ley, N. G. Robinson, R. N.
Sheppard, Tetrahedron Lett. 1988, 29, 1849 – 1852.
[20] S. Johnson, E. D. Morgan, I. D. Wilson, M. Spraul, M. Hofmann,
J. Chem. Soc. Perkin Trans. 1 1994, 1499 – 1502.
[21] A. A. Denholm, L. Jennens, S. V. Ley, A. Wood, Tetrahedron
1995, 51, 6591– 6604.
À
At the conclusion of the relay synthesis, the sensitive C22
C23 enol ether was unmasked by first converting methyl
acetal 9 to phenyl sulfide 10 upon treatment with thiophenol
and catalytic PPTS in toluene. Oxidation of 11 with DMDO,
followed by pyrolysis, reinstalled the enol ether, thereby
completing the relay sequence between azadirachtin (1) and
5.[21]
The work reported herein defines key intermediates that,
through a series of selective transformations, can be con-
verted to the natural product azadirachtin (1), a potent insect
antifeedant and growth-disrupting agent. This study sets the
stage for the effective total synthesis of this elusive molecule,
the details of which are disclosed in the following communi-
cation.[23]
[22] S. V. Ley, J. C. Anderson, W. M. Blaney, Z. Lidert, E. D. Morgan,
N. G. Robinson, M. S. J. Simmonds, Tetrahedron Lett. 1988, 29,
5433 – 5436.
Received: July 7, 2007
Published online: July 30, 2007
[23] G. E. Veitch, E. Beckmann, B. J. Burke, A. Boyer, S. L. Maslen,
S. V. Ley, Angew. Chem. 2007, 10.1002/ange.200703027; Angew.
Chem. Int. Ed. 2007, 10.1002/anie.200703027.
[24] The availabilty of azadirachtin has permitted its use in this relay
approach, however, it can only be obtained commercially as
crude neem extract (30% azadirachtin), which requires exten-
sive purification prior to use.
Keywords: azadirachtin · degradation · natural products ·
total synthesis
.
[1] J. H. Butterworth, E. D. Morgan, Chem. Commun. 1968, 23 – 24.
[2] H. Schmutterer, The Neem Tree, VCH, Weinheim, 1995.
[3] P. R. Zanno, I. Miura, K. Nakanishi, D. L. Elder, J. Am. Chem.
Soc. 1975, 97, 1975 – 1977.
[25] V. Kabaleeswaran, S. S. Rajan, T. R. Govindachari, G. Gopa-
lakrishnan, Curr. Sci. 1994, 66, 392 – 394.
[26] The C11,C20 dibenzylated compound was the major product,
but it was accompanied by small quantities of mixed-alkylation
products, the ratios of which were dependent on the reaction
conditions.
[27] C1reactivity during alkylation reactions was in all examples
lower than at any other position, presumbaly owing to steric
constraints imposed by the moleculeꢀs conformation.
[28] Selectivity in the debenzylation was obtained by careful
experimentation and solvent screening.
[29] The tiglic acid/Yamaguchi mixed anhydride was purified prior to
use, for details see M. Ball, S. P. Andrews, F. Wierschem, E.
Cleator, M. Smith, S. V. Ley, Org. Lett. 2007, 9, 663 – 666.
[30] J. L. Luche, J. Am. Chem. Soc. 1978, 100, 2226 – 2227.
[4] J. N. Bilton, H. B. Broughton, S. V. Ley, Z. Lidert, E. D. Morgan,
H. S. Rzepa, R. N. Sheppard, J. Chem. Soc. Chem. Commun.
1985, 968 – 971.
[5] W. Kraus, M. Bokel, A. Klenk, H. Poehnl, Tetrahedron Lett.
1985, 26, 6435 – 6438.
[6] H. B. Broughton, S. V. Ley, A. M. Z. Slawin, D. J. Williams, J.
Chem. Soc. Chem. Commun. 1986, 46 – 47.
[7] C. J. Turner, M. S. Tempesta, R. B. Taylor, M. G. Zagorski, J. S.
Termini, D. R. Schroeder, K. Nakanishi, Tetrahedron 1987, 43,
2789 – 2803.
[8] J. N. Bilton, H. B. Broughton, P. S. Jones, S. V. Ley, Z. Lidert,
E. D. Morgan, H. S. Rzepa, R. N. Sheppard, A. M. Z. Slawin,
D. J. Williams, Tetrahedron 1987, 43, 2805 – 2815.
[9] W. Kraus, M. Bokel, A. Bruhn, R. Cramer, I. Klaiber, A. Klenk,
G. Nagl, H. Poehnl, H. Sadlo, B. Volger, Tetrahedron 1987, 43,
2817 – 2830.
Angew. Chem. Int. Ed. 2007, 46, 7633 –7635
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7635