Studies toward the synthesis of azadirachtin, part 2: Construction of fully functionalized ABCD ring frameworks and unusual intramolecular reactions induced by close-proximity effects

Article abstract of DOI:10.1002/anie.200500217

The remarkable antifeedant agent and insecticide, azadirachtin (1), has, to date, escaped a total synthesis. The preparation of an advanced precursor has revealed some of the inner intricacies of this uniquely crowded and highly functionalized molecular a

Full text of DOI:10.1002/anie.200500217

Angewandte
Chemie
Natural Products Synthesis (2)
Studies toward the Synthesis of Azadirachtin,
Part 2: Construction of Fully Functionalized
ABCD Ring Frameworks and Unusual
Intramolecular Reactions Induced by Close-
Proximity Effects**
K. C. Nicolaou,* Pradip K. Sasmal,
Theocharis V. Koftis, Antonella Converso,
Eriketi Loizidou, Florian Kaiser, A. J. Roecker,
Constantinos C. Dellios, Xiao-Wen Sun, and
Goran Petrovic
In the preceding Communication in this issue[1] we described
the total synthesis of a potential decalin precursor to
azadirachtin (1),[2] its coupling to a suitable norbornene
fragment 4 (Scheme 1), and the elaboration of the product to
an advanced intermediate along the path to this synthetic
target. Herein we report the total synthesis and semisynthesis
from azadirachtin of a more advanced decalin system 3, its
coupling to the same norbornene fragment 4, and the
elaboration of the resulting product to an advanced inter-
mediate for the total synthesis of azadirachtin (Scheme 1).
This report also includes a number of unusual reactions
induced by proximity effects and special steric factors high-
lighting the unique characteristics of the azadirachtin scaffold.
Having successfully synthesized tricyclic decalin system 5,
(Scheme 1) and explored its chemistry toward azadirachtin
(1) as described in the preceding Communication,[1] we turned
our attention to the more advanced tetracyclic decalin
precursor 3, which bears the tetrahydrofuran ring system of
1 within its structure. Our plan to synthesize the targeted
intermediate 3 required key building block 6 (Scheme 2) as
[*] Prof. Dr. K. C. Nicolaou, Dr. P. K. Sasmal, Dr. T. V. Koftis,
Dr. A. Converso, Dr. F. Kaiser, A. J. Roecker, Dr. C. C. Dellios,
Dr. X.-W. Sun, Dr. G. Petrovic
Department of Chemistry and
The Skaggs Institute for Chemical Biology
The Scripps Research Institute
10550 North Torrey Pines Road, La Jolla, CA 92037 (USA)
Fax: (+1)858-784-2469
E-mail: kcn@scripps.edu
and
Department of Chemistry and Biochemistry
University of California, San Diego
9500 Gilman Drive, La Jolla, CA 92093 (USA)
E. Loizidou
Department of Chemistry
San Diego State University
5500 Campanile Drive, San Diego, CA 92182 (USA)
[**] We thank Dr. D. H. Huang, Dr. G. Siuzdak, and Dr. R. Chadha for
NMR spectroscopic, mass spectrometric, and X-ray crystallographic
assistance, respectively. Financial support for this work was
provided by grants from the National Institutes of Health (USA) and
the Skaggs Institute for Chemical Biology, a postdoctoral fellowship
from the Alexander von Humboldt Foundation (to F.K.), and a
predoctoral fellowship from the Division of Organic Chemistry of
the American Chemical Society sponsored by Novartis (to A.J.R.).
Angew. Chem. Int. Ed. 2005, 44, 3447 –3452
DOI: 10.1002/anie.200500217
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3447

Communications
Scheme 1. Structure of azadirachtin (1) and possible precursors 3 and
5.
the starting point, a compound that had been encoun-
tered as an enantiopure substance in our previous
expeditions.[1] This diol 6 was first protected as a
cyclohexanone ketal, the two benzyl protecting
groups were cleaved, the resulting primary hydroxy
group was selectively silylated, and finally the remain-
ing secondary alcohol was oxidized with DMP to
afford ketone 7 in 58% overall yield. Compound 7
was then converted into enone 8 by the standard,
three-step protocol in 95% overall yield and thence
into mesylate 10 via hydroxy compound 9 by desily-
lation followed by mesylation of the primary alcohol.
The OsO4–NMO-induced dihydroxylation of enone
10 proceeded stereoselectively from the less-hindered
a face and was accompanied by internal SN2-type
mesylate displacement, affording the expected tetra-
hydrofuran derivative in 95% yield over two steps.
Desilylation of the latter intermediate with TBAF
furnished dihydroxyketone 11 in 96% yield. This
substance was then converted into the exocyclic olefin
13 via its diacetate 12 (95%) by a Wittig reaction with
Scheme 2. Total synthesis of hydroxylactone 3. Reagents and conditions: a) 1,1-
dimethoxycyclohexane (4.0 equiv), PPTS (8.0 equiv), benzene, 808C, 4 h, 74%;
b) Pd/C (10%; 40 wt%), NaHCO3 (10 equiv), H2 (1 atm), EtOAc, 258C, 24 h,
91%; c) TBDPSCl (1.5 equiv), Et3N (3.0 equiv), DMAP (0.1 equiv), CH2Cl2, 08C,
2 h, 94%; d) DMP (1.5 equiv), NaHCO3 (2.0 equiv), CH2Cl2, 0!258C, 2 h,
91%; e) KHMDS (1.5 equiv), TESCl (1.5 equiv), THF, ꢀ788C, 30 min; f) PhSeCl
(1.1 equiv), CH2Cl2, ꢀ788C, 30 min; g) H2O2 (3.0 equiv), THF, 0!258C, 1 h,
95% over three steps; h) TBAF (1.5 equiv), THF, 0!258C, 2 h, 94%; i) MsCl
(2.0 equiv), Et3N (4.0 equiv), DMAP (0.1 equiv), CH2Cl2, 08C, 2 h; j) OsO4
(0.2 equiv), NMO (2.0 equiv), tBuOH/THF/H2O (5:5:1), 258C, 24 h, 95% over
two steps; k) TBAF (1.5 equiv), THF, 258C, 3 h, 96%; l) Ac2O (8.0 equiv), Et3N
=
Ph3P CH2 (93%) followed by deacetylation with
K2CO3 in MeOH (92%). This three-step process was
adopted after encountering difficulties with the
attempted, but failed, direct olefination of 11. The
primary alcohol of diol 13 was selectively oxidized to
=
(12 equiv), DMAP (0.2 equiv), CH2Cl2, 0!258C, 2 h, 95%; m) 1. Ph3P CH2
(8.0 equiv), Et2O, 0!258C, 6 h, 93%; 2. K2CO3 (6.0 equiv), MeOH, 258C, 5 h,
92%; n) TEMPO (0.8 equiv), polystyrene-supported bromite resin (6.0 equiv),
CH2Cl2, 08C, 2 h, 75%; o) Ac2O (5.0 equiv), Et3N (10 equiv), DMAP (0.2 equiv),
CH2Cl2, 0!258C, 1 h, 92%; p) NaClO2 (4.0 equiv), NaH2PO4 (5.0 equiv), 2-
methyl-2-butene (75 equiv), THF/tBuOH/H2O (2:4:1), 258C, 1 h; q) HCl
(0.5m), EtOH/Et2O (1:1), 0!258C, 3 h, 75% over two steps; r) SEMCl
(6.0 equiv), iPr2NEt (12 equiv), nBu4NI (4.0 equiv), CHCl3, 618C, 8 h, 89%;
s) K2CO3 (10 equiv), MeOH, 258C, 2 h, 99%. PPTS=pyridinium p-toluene sulfo-
nate, TBDPS=tert-butyldiphenylsilyl, DMAP=4-(dimethylamino)pyridine,
DMP=Dess–Martin periodinane, HMDS=hexamethyldisilazide, TES=triethyl-
silyl, TBAF=tetra-n-butylammonium fluoride, Ms=mesyl, NMO=N-methyl-
morpholine N-oxide, TEMPO=2,2,6,6-tetramethyl-1-piperidinyloxyl radical,
SEM=2-(trimethylsilyl)ethoxymethyl.
the corresponding hydroxyaldehyde through
a
TEMPO-catalyzed oxidation procedure (75%).[3]
The remaining secondary hydroxy function was
protected as an acetoxy group, and the aldehyde
moiety was further oxidized to the acetate carboxylic
acid. Exposure of the latter to 0.5m ethanolic HCl
gave the dihydroxylactone 14 through cleavage of
both the ketal and the silyl protecting groups and ring
closure (69% overall yield). Lactone 14 crystallized
from CH2Cl2/hexanes as colorless crystals (m.p. 240–
2418C) and X-ray crystallographic analysis[4] confirmed the
ring framework and stereochemical features of this com-
pound (Figure 1). Finally, protection of both free hydroxy
groups within 14 with SEMCl (89% yield) followed by
treatment with K2CO3 in MeOH led to the coveted hydroxy-
lactone 3 in 99% yield.
Having completed the construction of hydroxylactone 3,
we then contemplated obtaining this same key intermediate
3448
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 3447 –3452

Angewandte
Chemie
from the natural product 1 through semisyn-
thesis. The aim of the degradation exercise of
azadirachtin (1) to this key decalin fragment
was not only to enrich our supplies of the target
compound 2 for the purposes of our total
synthesis efforts, but also to discover certain,
unexpected or suspected, intricacies of the
azadirachtin molecule that may be either of
general interest or prove to be of some use in
aiding our sail towards the natural product. In
our search for an expedient route to 3 from 1
we made use of the work of Ley and co-
workers,[5–7] who developed a four-step degra-
dation sequence from 1 to 15,[6] the tetracyclic
compound that served as a beachhead for our
expeditions (Scheme 3). Thus, the two hydroxy
groups within 15 were protected as SEM ethers
by reaction with SEMCl in the presence of
iPr2NEt, and the resulting bis-SEM-protected
ketone was converted into olefin 17 by a
regioselective transformation into the corre-
sponding triflate 16 and then reduction of the
latter compound with Et3SiH in the presence of
[Pd(PPh3)4] (47% overall yield for three steps).
Manual molecular modeling suggested the
a face of the double bond of 17 as the more
accessible site, and this bias was greatly
enhanced by the asymmetric dihydroxylation
of this compound with K2OsO4 and
Scheme 3. Semisynthesis of hydroxylactone fragment 3 from 1. Reagents and conditions:
(DHQD)2PHAL
in
the
presence
of
a) SEMCl (6.0 equiv), iPr2NEt (12 equiv), nBu4NI (4.0 equiv), CHCl3, 608C, 8 h, 89%; b) Comins
reagent (2.0 equiv), KHMDS (0.5m in toluene, 2.0 equiv), THF, ꢀ788C, 30 min, 56%; c) [Pd-
(PPh3)4] (0.2 equiv), Et3SiH (3.0 equiv), LiCl (3.0 equiv), DMF, 558C, 24 h, 95%; d) K2OsO4·H2O
(0.2 equiv), (DHQD)2PHAL (0.5 equiv), K3Fe(CN)6 (3.0 equiv), MeSO2NH2 (1.0 equiv), K2CO3
(3.0 equiv), tBuOH/H2O (3:2), 2 h, 96%; e) Ac2O (5.0 equiv), Et3N (10 equiv), DMAP (1.0 equiv),
CH2Cl2, 0!258C, 8 h, 98%; f) Pd(OH)2/C (1:1 w/w), H2O/MeOH (1:2), pH 7, H2, 258C, 2 h;
g) NaBH4 (10 equiv), MeOH, 0!258C, 4 h; h) Pb(OAc)4 (2.0 equiv), CH2Cl2, 08C, 2 h, 70% over
three steps; i) SOCl2 (20 equiv), py, 0!258C, 8 h, 86%; j) K2CO3 (10 equiv), MeOH, 258C, 2 h,
99%. Comins reagent=2-[N,N-bis(trifluoromethylsulfonyl)amino]-5-chloropyridine, DMF=N,N-
dimethylformamide, (DHQD)2PHAL=1,4-bis(9-O-dihydroquinidinyl)phthalazine, py=pyridine.
MeSO2NH2 in 96% yield, leading to 1,2-diol
18 as the exclusive product. Landing an acetate
group on the secondary alcohol of 18 proved
trivial, as treatment of this compound with
Ac2O, DMAP and Et3N furnished monoacetate
19 (98% yield) as expected. Debenzylation of
compound 19 required carefully controlled
conditions (10% Pd(OH)2/C, MeOH-H2O,
pH 7, H2), otherwise an unusual reaction took
place, namely the reduction of the benzene ring
to afford the corresponding cyclohexylmethyl ether deriva-
tive 19-H6 (10% Pd(OH)2/C, EtOH, H2, 98% yield, 23/19-H6
ꢁ 2.6:1). The origin of this remarkable mode of reactivity of
benzyl ether derivative 19 remains unknown. Facilitated by
the neighboring hydroxy group, the reduction with NaBH4 of
the carbomethoxy group of the debenzylation product
derived from 19 proceeded smoothly to afford triol 20 in
good yield. The Pb(OAc)4-induced cleavage of the 1,2-diol
system within 20 gave g-lactone 21 in 70% overall yield from
19 for the three steps. Finally, exposure of 21 to SOCl2 in the
presence of pyridine led to the corresponding exocyclic olefin,
which was treated with K2CO3 to give the targeted allylic
alcohol 3 in 85% overall yield.
En route to allylic alcohol 3 from compound 19, a
noteworthy reaction was observed which led to a number of
subsequent developments of equal interest. When the crude
reaction mixture from hydrogenation of 19 was subjected to
the intended reduction of the methyl ester with NaBH4
followed by 1,2-diol cleavage with Pb(OAc)4 on a large
scale, the g-lactone formate 22 was obtained, together with
the desired g-lactone alcohol 21 in approximately equimolar
amounts and good overall yield. As shown in Scheme 4, we
Figure 1. ORTEP drawing of 14 at the 30% probability level.
Angew. Chem. Int. Ed. 2005, 44, 3447 –3452
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3449

Communications
Scheme 4. An unexpected Pb(OAc)4-induced carbonyl migration and
preparation of formate 22. Reagents and conditions: NaBH4
(10 equiv), MeOH, 08C!258C, 30 min; then Pb(OAc)4 (10 equiv),
CH2Cl2, 0!258C, 2 h, 95% over three steps.
hypothesized that this rather unusual outcome might be due
to the propensity of the incipient diol 23 to form pentacyclic
lactone 24 by attack of the C8 tertiary hydroxy group at the
neighboring C11-carbomethoxy moiety with subsequent
expulsion of a molecule of MeOH. Exposure of 24 to
NaBH4 would then result in the C11-OH-assisted reduction
of the newly formed g-lactone ring, yielding bis-hemiketal 25,
which could be oxidatively cleaved in the next step by
Pb(OAc)4 to give the observed formate lactone 22. In support
of this hypothesis we synthesized the postulated intermediate
24 from 23 and converted it exclusively into formate 22 by the
proposed two-step sequence in 95% overall yield.
The fortuitous availability of formate 22 inspired a
potential entry into the a,b-unsaturated g-lactone series of
decalin fragments (e.g. compound 30, Scheme 5) as possible
partners in coupling reactions with the norbornene-type
precursors of azadirachtin (e.g. 4). Thus, 22 was heated in
refluxing toluene in the presence of DBU in the hope that the
a,b-unsaturated lactone 29 would be obtained by elimination
of the formate ester group (see Scheme 5). Instead, however,
compound 27 was observed as the only product (70%). This
C8-inverted tertiary alcohol 27 is presumably the product of
ionization of the formate group to carbocation 26 followed by
its immediate quenching by traces of moisture present in the
reaction medium (22!26!27, Scheme 5). The nonpolar
nature of the solvent employed (toluene) accounts for the
observed inversion of configuration at C8, the tight ion pair
formed between the decalin cationic domain and the leaving
formate anion ensuring total blockade of the already hin-
dered decalin a face. An alternative plausible mechanism,
namely that involving stereoselective conjugate addition of
water to an initially formed a,b-unsaturated g-lactone (i.e. 29,
the compound targeted in the first place), has been ruled out
by the failure of 29 to undergo addition of water upon
exposure to the same reaction conditions. Unexpected as it
was, the C8-epi alcohol 27 offers the only point of entry found
so far, even after considerable experimentation, into the a,b-
Scheme 5. An unexpected, DBU-induced inversion of configuration
and preparation of 8-epi hydroxy compound 27 and allylic acetates 28
and 29. Reagents and conditions: a) DBU (20 equiv), toluene, reflux,
4 h, 70%; b) SOCl2 (20 equiv), py, 0!258C, 8 h, 71%; c) DBU
(20 equiv), toluene, reflux, 10 h; d) K2CO3 (10 equiv), MeOH, 258C,
2 h, 99%. DBU=1,8-diazabicyclo[5.4.0]undec-7-ene.
unsaturated g-lactone series of decalin fragments. This was
converted into 29 by exposure to SOCl2 in the presence of
pyridine in a reaction that also produced the exocyclic olefinic
acetate 28 in a combined yield of 71% (28/29 ꢁ 2:1). Allylic
alcohols 3 and 30, obtained by treatment of this mixture with
K2CO3 in MeOH, were conveniently separated by silica gel
chromatography.
Allylic alcohol 3, which was now available through two
different routes, was then subjected to bromoketalization
with norbornene derivative 4 under the conditions described
in the preceding Communication[1] and remarkably afforded a
single bromoketal 31 in 76% yield (Scheme 6).[10] It is
interesting to recall that the bromoketalization of the
previously employed decalin substrate 5, which lacks the
tetrahydrofuran system, led to a mixture of two diastereo-
meric bromoketals in approximately 1:1 ratio.[1] From this
contrast, it is apparent that the distribution of the two isomers
in these reactions is dictated not only by the reaction time and
temperature, but also, and most profoundly, by the nature of
the decalin partner.[8]
It was again found that (Me3Si)3SiH served as an excellent
H-atom donor in the radical-based cyclization of bromoketal
31, yielding heptacyclic compounds 35 and 36 (Table 1) in 32
and 42% yield, respectively.[9] In this case, the secondary
radical 32, initially generated from 31, underwent both the 6-
endo-trig and the 5-exo-trig modes of ring closure,[8] thus
leading to the tertiary and primary radicals 33 and 34. These,
3450
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 3447 –3452

Angewandte
Chemie
Scheme 7. Conversion of heptacyclic compound 36 into advanced
system 2. Reagents and conditions: a) NaBH4 (5.0 equiv), MeOH,
08C, 15 min, 88%; b) BzCl (7.0 equiv), DMAP (5.0 equiv), pyridine,
608C, 5 h; c) 0.5m HCl, MeOH/Et2O (1:1), 0!258C, 6 h, 72% over
two steps; d) Ac2O (8.0 equiv), Et3N (12 equiv), DMAP (0.4 equiv),
CH2Cl2, 0!258C, 6 h, 79%; e) H2SO4 (5.0 equiv), CH2Cl2, 08C, 30 min;
f) PCC (15 equiv), DCE, 658C, 16 h, 60% over two steps; g) K2CO3
(10 equiv), MeOH, 258C, 30 min; h) TESOTf (10 equiv), 2,6-lutidine
(20 equiv), CH2Cl2, ꢀ78!258C, 24 h, 62% over two steps. Bz=ben-
zoyl, PCC=pyridinium chlorochromate, DCE=1,2-dichloroethane,
OTf=trifluoromethanesulfonate.
which was benzoylated and then desilylated under acidic
conditions to furnish dihydroxy benzoate 37 in 63% overall
yield. Exposure of diol 37 to Ac2O, Et3N, and DMAP gave the
corresponding bisacetate (79% yield), which was subse-
quently converted into hemiketal 38. Finally, oxidation of 38
with PCC, cleavage of the acetate group, and TES protection
furnished diketone 2 (Table 1) in 37% yield for the four steps.
In the course of our drive towards azadirachtin (1), we
became keenly, and sometimes painfully, aware of the unique
behavior of this natural product due to the oxygen-rich nature
of its skeleton and the close proximity of its numerous
functional groups. In addition to the above-mentioned
incidents, we consider it of interest to reveal one more
example that speaks to this point. Scheme 8 shows the
outcome of the reaction of diketone 2 with MeLi. It became
apparent that initial loss of the benzoyl protecting group at
C21 unmasked the secondary alcohol function, which was
then in a position to attack the C7 carbonyl group, presum-
ably forming a hemiketal. The existence of the latter
compound was evidently transient under the reaction con-
ditions, as a further ring closure ensued, this time on the C13
carbonyl carbon atom, to give the highly congested cagelike
ketal/hemiketal 40 in 72% overall yield. Lactone 40 did not
constitute the end of the journey in this cascade in the
presence of CeCl3. In this case, further attack by MeLi at the
C11 lactone carbonyl function produced a free primary
alcohol group, which was then able to reach across the decalin
system to attack the C4 carbomethoxy residue to generate a
second lactone ring with the expulsion of a methoxy group,
furnishing, upon further addition of MeLi, the impressively
Scheme 6. Bromoketalization of hydroxylactone 3 with norbornene
derivative 4 and elaboration of the product into 35 and 36. Reagents
and conditions: a) Br2 (1.5 equiv), N,N-dimethylaniline (2.0 equiv),
K2CO3 (10 equiv), 4 (2.0 equiv), CH2Cl2, ꢀ788C, 10 min; then 3,
ꢀ78!08C over 2 h, 08C, 1 h, 76%; b) (Me3Si)3SiH (2.0 equiv), AIBN
(1.0 equiv), toluene (0.007m), 1108C, 30 min, 35 (32%) 36 (42%);
AIBN=2,2’-azobisisobutyronitrile.
in turn, undergo a 1,5-H shift (as shown in Scheme 6), to
afford ketones 35 and 36, respectively, upon oxidative
cleavage of the PMB ether group. This behavior again reveals
the controlling power of the decalin system in deciding the
fate of these chemical species.
From the two intermediates 35 and 36, only the latter is
potentially productive in the campaign towards azadirachtin.
To this end, we needed to demonstrate the facile cleavage of
the temporary bridge that enabled the formation of the
ꢀ
crucial C8 C14 bond in the first place. The heptacyclic ketone
36 (Scheme 7) was reduced with NaBH4 in MeOH, exclu-
sively from the exo side, to afford the corresponding alcohol,
Angew. Chem. Int. Ed. 2005, 44, 3447 –3452
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3451

Communications
Table 1: Selected physical properties for compounds 2, 35, and 36.
2: Rf =0.50 (silica gel, EtOAc/hexane 1:2); [a]3D2 =ꢀ23.0 (c=0.3, CHCl3);
IR (film) n˜max =2955, 2876, 1782, 1763, 1752, 1742, 1717, 1458, 1273,
1
1162, 1113, 1019, 874, 813, 716 cmꢀ1; H NMR (600 MHz, C6D6):
d=7.87 (d, J=7.5 Hz, 2H), 7.51 (t, J=7.5 Hz, 1H), 7.42 (t, J=7.5 Hz,
2H), 5.57 (ddd, J=10.2, 9.0, 4.2 Hz, 1H), 5.01 (d, J=14.4 Hz, 1H), 4.62
(d, J=10.5 Hz, 1H), 4.44 (br s, 1H), 4.13 (d, J=8.4 Hz, 1H), 4.06 (d,
J=8.4 Hz, 1H), 3.91 (s, 1H), 3.83 (d, J=10.5 Hz, 1H), 3.76 (s, 3H),
3.73 (br s, 1H), 3.11 (br d, J=3.0 Hz, 1H), 3.02 (br s, 1H), 2.99 (d,
J=14.4 Hz, 1H), 2.85 (br d, J=4.8 Hz, 1H), 2.66 (dt, J=13.2, 4.2 Hz,
1H), 2.41 (ddd, J=13.2, 10.2, 4.8 Hz, 1H), 2.02 (br d, J=15.6 Hz, 1H),
1.90 (br d, J=15.6 Hz, 1H), 1.87 (br d, J=10.8 Hz, 1H), 1.78 (br d,
J=10.8 Hz, 1H), 1.37 (s, 3H), 0.97 (t, J=7.8 Hz, 9H), 0.96 (t,
J=8.4 Hz, 9H), 0.66 (q, J=7.8 Hz, 6H), 0.61 ppm (q, J=8.4 Hz, 6H);
13C NMR (150 MHz, C6D6): d=212.7, 202.0, 175.4, 174.0, 166.7, 132.9,
130.0, 129.5, 128.5, 76.1, 74.4, 74.3, 70.8, 69.9, 66.5, 56.5, 55.3, 54.5,
53.3, 52.6, 52.0, 48.0, 45.0, 42.6, 36.6, 36.2, 30.8, 20.8, 7.0, 7.0, 4.7 ppm;
HRMS (ESI TOF): calcd for C42H61O11Si2 [M+H+]: 797.3747; found:
+
797.3748
35: Rf =0.25 (silica gel, EtOAc/hexanes 1:1); [a]3D1 =ꢀ30.0 (c=0.3,
CH2Cl2); IR (film): n˜max =2952, 2926, 1780, 1743, 1724, 1437, 1249, 1163,
1054, 836 cmꢀ1; 1H NMR (500 MHz, C6D6): d=4.75 (d, J=7.0 Hz, 1H),
4.70 (dd, J=9.2, 1.0 Hz, 1H), 4.63 (d, J=6.6 Hz, 1H), 4.56 (d,
J=6.6 Hz, 1H), 4.45 (t, J=2.6 Hz, 1H), 4.32 (d, J=11.0 Hz, 1H), 4.31
(d, J=6.6 Hz, 1H), 4.23 (dd, J=11.0, 1.5 Hz, 1H), 4.17 (d, J=7.7 Hz,
1H), 3.93 (ddd, J=10.6, 9.5, 5.8 Hz, 1H), 3.79 (d, J=8.1 Hz, 1H), 3.69–
3.62 (m, 3H), 3.53–3.44 (m, 3H), 3.41 (s, 3H), 3.39 (dd, J=11.5, 1.5 Hz,
1H), 3.15 (s, 3H), 2.50 (d, J=2.5 Hz, 1H), 2.34 (dd, J=5.1, 2.2 Hz, 1H),
2.23–2.13 (m, 4H), 1.72 (br d, J=11.3 Hz, 1H), 1.53 (dd, J=17.8,
4.5 Hz, 1H), 1.51 (br d, J=12.0 Hz, 1H), 1.42 (dd, J=12.8, 3.8 Hz, 1H),
1.35 (dt, J=16.2, 2.9 Hz, 1H), 1.23 (br d, J=10.6 Hz, 1H), 1.19–1.05 (m,
2H), 0.87 (d, J=7.7 Hz, 1H), 0.85 (d, J=8.1 Hz, 1H), 0.08 (s, 9H),
ꢀ0.04 ppm (s, 9H); 13C NMR (125 MHz, C6D6): d=214.1, 177.6, 173.9,
116.8, 93.5, 93.2, 88.0, 77.4, 73.0, 71.3, 70.0, 69.2, 66.1, 65.6, 59.9, 57.7,
54.9, 51.6, 51.4, 48.5, 43.9, 40.8, 40.1, 38.6, 37.2, 36.6, 29.4, 26.9, 18.1,
18.0, ꢀ1.3, ꢀ1.4 ppm; HRMS (MALDI): calcd for C36H58O12Si2Na:
761.3359 [M+Na+]; found: 761.3350
Scheme 8. Intriguing cascade sequences with azadirachtin-type scaffolds.
Reagents and conditions: a) MeLi (5.0 equiv), Et2O, ꢀ788C, 2 h, 72%; b) K2CO3
(10 equiv), MeOH, 258C, 36 h, 91%; c) MeLi (5.0 equiv), CeCl3 (5.0 equiv), Et2O,
08C, 8 h, 80%.
compact polycycle 41 in 80% overall yield from 2. The stable
ketal/hemiketal 40 was also the exclusive product (91% yield)
of the reaction of benzoate 2 with K2CO3 in MeOH.
The described chemistry represents our most recent and
important advancements toward the total synthesis of azadir-
achtin (1) and at the same time reveals some of the inner
intricacies of this uniquely crowded and highly functionalized
molecular architecture. Further progress towards congeners
of azadirachtin is sure to be frustrated or facilitated by these
intricacies.
36: Rf =0.36 (silica gel, EtOAc/hexanes 1:1); [a]3D1 =ꢀ24.0 (c=0.3,
CH2Cl2); IR (film): n˜max =2953, 2924, 1766, 1741, 1725, 1438, 1317, 1248,
Received: January 19, 2005
Published online: April 22, 2005
1
1169, 1098, 1032, 836 cmꢀ1; H NMR (500 MHz, C6D6): d=4.76 (d,
J=7.0 Hz, 1H), 4.61 (d, J=7.0 Hz, 1H), 4.59 (d, J=6.6 Hz, 1H), 4.39–
4.37 (m, 1H), 4.37 (d, J=6.6 Hz, 1H), 4.33–4.27 (m, 3H), 4.17 (d,
J=10.3 Hz, 1H), 4.07 (d, J=7.7 Hz, 1H), 3.96 (ddd, J=11.0, 9.5,
5.9 Hz, 1H), 3.71 (s, 1H), 3.64 (ddd, J=11.0, 9.9, 5.9 Hz, 1H), 3.56 (br t,
J=2.4 Hz, 1H), 3.54–3.46 (m, 4H), 3.37 (s, 3H), 3.10 (s, 3H), 3.05 (d,
J=4.4 Hz, 1H), 2.42 (d, J=3.7 Hz, 1H), 2.38 (d, J=4.0 Hz, 1H), 2.31
(dd, J=18.0, 4.8 Hz, 1H), 2.13 (dt, J=16.1, 2.6 Hz, 1H), 1.68–1.61 (m,
1H), 1.48 (dd, J=18.0, 4.8 Hz, 1H), 1.39 (dt, J=16.1, 2.9 Hz, 1H), 1.21
(s, 3H), 1.18–1.05 (m, 3H), 0.88 (d, J=8.4 Hz, 1H), 0.87 (d, J=7.7 Hz,
1H), 0.10 (s, 9H), 0.03 ppm (s, 9H); 13C NMR (125 MHz, C6D6):
d=215.0, 176.3, 174.1, 116.0, 93.6, 92.5, 85.0, 73.9, 72.8, 72.3, 70.5, 69.8,
67.3, 66.2, 65.6, 53.9, 53.8, 51.6, 51.1, 49.1, 47.5, 46.6, 42.0, 39.9, 38.9,
35.2, 27.1, 18.2, 18.1, 16.5, ꢀ1.3, ꢀ1.4 ppm; HRMS (MALDI): calcd for
C36H58O12Si2Na: 761.3359 [M+Na+]; found: 761.3366
Keywords: asymmetric synthesis · degradation ·
.
natural products · radical reactions · total synthesis
[1] K. C. Nicolaou, P. K. Sasmal, A. J. Roecker, X.-W. Sun, S.
Mandal, A. Converso, Angew. Chem. 2005, 117, 3509 – 3513;
Angew. Chem. Int. Ed. 2005, 44, 3443 – 3447, preceding Commu-
nication in this issue.
[2] For the isolation, structural elucidation, and previous synthetic
studies in the area of azadirachtin (1), see literature cited in
reference [1].
[3] G. Sourkouni-Argirusi, A. Kirschning, Org. Lett. 2000, 2, 3781 –
3784.
[4] CCDC-261286 contains the supplementary crystallographic data
for this paper. These data can be obtained free of charge from
the Cambridge Crystallographic Data Centre via www.ccdc.cam.
ac.uk/data_request/cif.
[5] S. V. Ley, P. J. Lovell, S. C. Smith, A. Wood, Tetrahedron Lett.
1991, 32, 6183 – 6186.
[8] C. Anies, L. Billot, J.-Y. Lallemand, A. Pancrazi, Tetrahedron
Lett. 1995, 36, 7247 – 7250.
[9] The structures and stereochemistries of compounds 31, 35, and
36 were unambiguously assigned by spectroscopic analysis (1H,
13C, COSY, ROESY, HMQC, and HMBC).
[6] S. V. Ley, P. J. Lovell, A. M. Z. Slawin, S. C. Smith, D. J. Williams,
A. Wood, Tetrahedron 1993, 49, 1675 – 1700.
[7] A. A. Denholm, L. Jennens, S. V. Ley, A. Wood, Tetrahedron
1995, 51, 6591 – 6604.
[10] Note added in proof: After submission of this manuscript, we
isolated the other bromoketal diastereomer in approximately
10% yield from experiments carried out on a larger scale.
3452
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 3447 –3452