Model studies towards diazonamide A: Synthesis of the heterocyclic core

Full text of DOI:10.1002/1521-3773(20001002)39:19<3473::AID-ANIE3473>3.0.CO;2-E

COMMUNICATIONS
diffractometer. Crystallographic data (excluding structure factors) for
the structures reported in this paper have been deposited with the
Cambridge Crystallographic Data Centre as supplementary publica-
tion no. CCDC-145227 (I), CCDC-145228 (II), and CCDC-145229
(III). Copies of the data can be obtained free of charge on application
to CCDC, 12Union Road, Cambridge CB12EZ, UK (fax:
(‡44)1223-336-033; e-mail: deposit@ccdc.cam.ac.uk).
H
H
N
N
H2N
N
N
Cl
A
O
B
O
O
Cl
O
E
G
C
F
O
NH
Á
[12] G.Ferey, J. Solid. State Chem. 2000, 152, 37.
D
[13] Crystal data for II: [Cd(C2O4)] ´ 2RbCl ´ H2O, Mr ˆ 460.3, crystal
dimensions 0.06 Â 0.14 Â 0.18mm, orthorhombic, space group, Pbcm,
a ˆ 6.018(2), b ˆ 10.861(2), c ˆ 15.503(4) Š, Vˆ 1013.4(5) Š3, Z ˆ 4,
1calcd ˆ 3.004 gcmÀ3, m(MoKa) ˆ 12.21 mmÀ1, l ˆ 0.71073 Š. The struc-
ture was solved by direct methods (SHELXTL-PLUS), 3866 reflec-
tions, 758 independent reflections. Full-matrix least-squares on jF 2 j
OH
OH
1: diazonamide A
ring closure by alkene
bond generation
R
O
oxazole formation
N
N
N
29
30
26
O
(SHELXs-93, G. M. Sheldrick, Gottingen, 1993) to R1 ˆ 0.08 and
O
À3 [11b) ]
24
wR2 ˆ 0.17. Residual density, min./max.: À2.72/2.03 eŠ
.
10
11
18
[14] Crystal data for III: K2Cd(C2O4)3 ´ 2 KBr ´ 2 H2O, M ˆ 841.12, crystal
dimensions, 0.12 Â 0.12 Â 0.24 mm, orthorhombic, space group, Pbca,
a ˆ 11.898(6), b ˆ 10.963(1), c ˆ 15.203(6) Š, Vˆ 1983.2(2) Š3, Z ˆ 4,
1calcd ˆ 2.817 gcmÀ3, m(MoKa) ˆ 7.077 mmÀ1, l ˆ 0.71073 Š. The struc-
ture was solved by direct methods (SHELXTL-PLUS), 7521 reflec-
tions, 1428 independent reflections. Full-matrix least-squares on jF 2 j
NH
16
N
O
Me
O
RO
MeO
2
3
by HWE
olefination
by olefin
metathesis
O
(SHELXs-93, G. M. Sheldrick, Gottingen, 1993) to R1 ˆ 0.03 and
P(OMe)2
N
À3 [11b) ]
wR2 ˆ 0.05. Residual density, min./max.: À0.593/0.369 eŠ
.
N
O
O
[15] T. A. Kodenkandath, J. N. Lalena, W. I. Zhou, E. E. Carpenter, C.
Sangregorio, A. U. Falster, J. R. Simmons, C. J. OꢁConnor, J. B. Wiley,
J. Am. Chem. Soc. 1999, 121, 10743.
O
N
N
Me
Me
O
O
[16] Q. Huang, M. Ulutagay, P. A. Michener, S.-J. Hwu, J. Am. Chem. Soc.
1999, 121, 10323.
MeO
MeO
biaryl coupling
biaryl coupling
5
4
Scheme 1. Structure of diazonamide A (1) and retrosynthetic analysis of
model system 2.
Model Studies towards Diazonamide A:
Synthesis of the Heterocyclic Core**
complex and strained halogenated heterocyclic core trapped
as a single atropisomer harboring a quaternary center at the
epicenter.[1] Beyond the formidable synthetic challenge posed
by such a daunting molecular framework, diazonamide A
possesses potent in vitro cytotoxicity against human colon
carcinoma and B-16 murine melanoma cell lines with IC50
values in the nanomolar range. These biological actions,
however, are exerted through an unknown mode of action.
Unfortunately, more extensive bioassays have been hampered
by an inability to harvest additional material from the original
source. As such, diazonamide A represents one of the most
enticing natural products isolated in recent years and a serious
challenge to synthetic chemists. Although several groups have
reported progress on particular structural subunits of diazon-
amide A,[2] no one has yet successfully prepared the fully
unsaturated 12-membered polycycle. Herein we report the
first synthesis of this heteroaromatic macrocyclic core by a
concise and novel strategy which proceeds to provide
atropisomerically pure product.
Since we envisioned from the outset that the heterocyclic
core would pose the greatest synthetic challenge for a total
synthesis of diazonamide A, we focused our efforts primarily
on the preparation of model system 3 (Scheme 1) as a means
to address key synthetic issues such as potential atropisomer-
ism along the C16 ± C18 and C24 ± C26 biaryl linkages. As shown
in Scheme 1, our synthetic rationale was based on two fairly
obvious but strategic bond disconnections. Because numerous
methods are currently available to generate macrocycles by
alkene bond formation, we felt that disconnection of the C29 ±
C30 alkene would be prudent and could potentially arise in the
K. C. Nicolaou,* Scott A. Snyder, Klaus B. Simonsen,
and Alexandros E. Koumbis
Diazonamide A (1, Scheme 1), a secondary metabolite
isolated from the colonial ascidian Diazona chinensis, exem-
plifies an unprecedented molecular architecture encompass-
ing a cyclic polypeptide backbone as well as an admirably
[*] Prof. Dr. K. C. Nicolaou, S. A. Snyder, Dr. K. B. Simonsen,
Dr. A. E. Koumbis
Department of Chemistry
and The Skaggs Institute for Chemical Biology
The Scripps Research Institute
10550 North Torrey Pines Road, La Jolla, CA 92037 (USA)
and
Department of Chemistry and Biochemistry
University of California San Diego
9500 Gilman Drive, La Jolla, CA 92093 (USA)
Fax : (‡1)858-784-2469
E-mail: kcn@scripps.edu
[**] We thank Dr. D. H. Huang, Dr. G. Suizdak, and Dr. R. Chadha for
NMR spectroscopic, mass spectrometric, and X-ray crystallographic
assistance, respectively. We also thank Professor R. H. Grubbs for a
generous gift of catalyst 30. Financial support for this work was
provided by The Skaggs Institute for Chemical Biology, the National
Institutes of Health (USA),
a predoctoral fellowship from the
National Science Foundation (S.A.S.), postdoctoral fellowships from
the Alfred Benzons Foundation and the Danish Natural Science
Research Council (K.B.S.), and grants from Abbott, Amgen, Array-
Biopharma, Boehringer± Ingelheim, GlaxoWellcome, Hoffmann ±
LaRoche, DuPont, Merck, Novartis, Pfizer, and Schering Plough.
Angew. Chem. Int. Ed. 2000, 39, No. 19
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
1433-7851/00/3919-3473 $ 17.50+.50/0
3473

COMMUNICATIONS
forward direction either by metathesis of the olefinic groups
in 4 or by an intramolecular Horner± Wadsworth ± Emmons
(HWE) reaction between the phosphonate and aldehyde
residues in 5; the resultant C29 ± C30 alkene could then serve as
a scaffold upon which to construct the A-ring oxazole, thereby
completing the ABCDEF polycycle 2. From either 4 or 5, our
second key disconnection rupturing the C16 ± C18 biaryl link-
age revealed, by a retro Suzuki or Stille type coupling, two
fragments of roughly equal molecular complexity, thereby
affording a highly convergent retrosynthetic blueprint.
Due to the power of the olefin metathesis reaction in
macrocycle formation, particularly with regard to complex
natural product total synthesis,[3] we initially sought to test this
method as a means to generate the C29 ± C30 alkene. Toward
this end, our synthetic endeavors commenced with prepara-
tion of the BCD indole ± oxazole fragment 14 as shown in
Scheme 2. Starting from 4-bromoindole (6),[4] condensation
with dimethylamino-2-nitroethylene[5] followed by LiAlH4-
mediated reduction readily afforded 4-bromotryptamine (8)
in 87% overall yield. Subsequent coupling of 8 with TBDPS-
protected glycolic acid derivative 9[6] in the presence of EDC
and HOBt smoothly provided intermediate 10 (for abbrevia-
tions of reagents and protecting groups, see legends in
schemes). Although the conversion of 10 to 11 upon treatment
with DDQ is known,[7] we found that far superior yields of 11
could be obtained by reverting to a two-step procedure in
which initial exposure of 10 to DDQ in THF/H2O (9/1) at 08C
resulted in formation of the desired benzylic hydroxy group
which was then readily oxidized by using IBX (90% overall
yield).[8] Notably, protection of the indole nitrogen was
unnecessary during this synthetic sequence and exposure to
IBX did not afford any N-oxidized products. The synthesis of
14 was then completed through the following five-step
protocol: 1) formation of the oxazole ring using a modified
variant of the Gabriel ± Robinson cyclodehydration reac-
tion;[9] 2) N-methyl protection of the indole nucleus employ-
ing phase-transfer conditions (88% over two steps);
3) TBAF-mediated cleavage of the terminal TBDPS group
(92%); 4) oxidation of the resultant hydroxyl group with IBX
(90%); and 5) generation of the desired alkene upon Wittig
Br
Br
NO2
N
a.
NO2
N
H
N
H
6
7
ˆ
reaction of 13 with Ph3P CH2 (89%).
b. LiAlH4
Br
Our next task was to synthesize the requisite EFG fragment
26, as delineated in Scheme 3. Although numerous methods
to access such benzofuranone fragments in racemic form are
known, we sought to develop a novel route to this type of
system with the hope of generating the C10 quaternary center
of diazonamide A stereoselectively. As such, our approach
began with aldehyde 15.[10] After initial treatment of 15 with
PhMgBr, the hydroxy group of the resultant bisbenzylic
compound was smoothly oxidized by utilizing IBX, and the
Br
H
N
NH2
OTBDPS
O
c. 9, EDC, HOBt
N
H
N
H
10
8
HO
OTBDPS
TBDPSO
d. DDQ
e. IBX
9
O
Br
O
Br
O
H
N
N
ˆ
newly generated ketone was treated with Ph3P CH2 to afford
OTBDPS
olefin 16 in 80% overall yield over these three steps.
Acid-catalyzed removal of the methoxymethyl (MOM)
protecting group was then followed by alkylation of the
resultant phenolic residue with chloroacetonitrile in acetone
to afford 17 in 75% yield. Subsequent mCPBA-mediated
epoxidation of the olefin provided key intermediate 18. As
shown in Scheme 4, we envisaged that upon exposure of 18 to
base, deprotonation of the methylene group adjacent to the
cyanide followed by selective 5-exo-tet opening of the
epoxide ring would lead to the desired quaternary system 21
via intermediate 20.[11] Gratifyingly, treatment of 18 with
KOtBu in DMF at À578C followed by immediate quench-
ing with 10% aqueous HCl at that temperature afforded
21 in 54% yield. The only other product observed in this novel
cyclization step was 3-phenylbenzofuran 22, which resulted
O
f. C2Cl6, Ph3P, Et3N
N
H
N
Me
12
g. MeI, nBu4NHSO4
11
h. TBAF
i. IBX
CHO
N
Br
O
Br
O
N
j. H2C=PPh3
N
N
Me
Me
14
13
Scheme 2. Preparation of indole ± oxazole fragment 14: a) Dimethylami-
no-2-nitroethylene (1.1 equiv), TFA, 258C, 30 min; b) LiAlH4 (1.0m in
THF, 6.0 equiv), THF, 258C !658C, 4 h, 87% over two steps; c) 9
(1.0 equiv), EDC (1.0 equiv), HOBt (1.0 equiv), CH2Cl2, 2 58C, 2h, 79%;
d) DDQ (3.0 equiv), THF/H2O (9/1), 08C, 3 h; e) IBX (3.0 equiv), THF/
DMSO (1/1), 258C, 3 h, 90% over two steps; f) Cl3CCCl3 (2.0 equiv), Ph3P
(2.0 equiv), Et3N (4.0 equiv), 258C, 15 min; g) MeI (4.0 equiv), nBu4NH-
SO4 (1.1 equiv), C6H6, 10% aq. NaOH, 258C, 20 min, 88% over two steps;
h) TBAF (1.0m in THF, 1.5 equiv), THF, 258C, 15 min, 92%; i) IBX
(3.0 equiv), THF/DMSO (1/1), 258C, 3 h, 90%; j) methylene triphenyl-
phosphonium bromide (1.5 equiv), nBuLi (1.6m in hexanes, 1.3 equiv),
THF, 0 !258C, 15 min, 89%. TFA ˆ trifluoroacetic acid, EDC ˆ 3-(3-
dimethylaminopropyl)-1-ethylcarbodiimide, TBDPS ˆ tert-butyldiphenyl-
silyl, HOBt ˆ 1-hydroxy-1H-benzotriazole, DDQ ˆ 2,3-dichloro-5,6-dicya-
no-1,4-benzoquinone, TBAF ˆ tetrabutylammonium fluoride, IBX ˆ 1-hy-
droxy-1,2-benziodoxol-3(1H)-one 1-oxide.
from
a
previously reported cyclofragmentation reac-
tion.[12] With 21 in hand, subsequent TBS protection af-
forded crystalline intermediate 23, which when subjected to
X-ray analysis[13] revealed that the cyclization reaction
proceeded to give a racemic mixture of compounds possessing
only the relative trans stereochemistry as indicated for 21 and
23. As such, stereoselective generation of the epoxide in 18
followed by our cyclization protocol would provide an
enantioselective route to the C10 quaternary center of
diazonamide A.
3474
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
1433-7851/00/3919-3474 $ 17.50+.50/0
Angew. Chem. Int. Ed. 2000, 39, No. 19

COMMUNICATIONS
exposure to O2 and SnCl2.[14] Next, the lactone moiety was
partially reduced with DIBAL-H to the corresponding lactol
with subsequent protection of the hydroxy group as a methyl
ether using NaH and MeI to afford 25 in 79% overall yield
from 23. Although the reduction proceeded stereoselectively,
the methylation step resulted in epimerization of the C11
center. The TBS ether was then easily cleaved upon exposure
to aqueous HF in MeCN, and the resultant primary alcohol
was oxidized to the corresponding aldehyde using Dess ±
Martin periodinane. Finally, the desired terminal alkene
group of 26 was generated upon treatment with Tebbe reagent
(70% yield from 25). Significantly, attempts to synthesize the
OMOM
Br
OMOM
Br
a. PhMgBr
b. IBX
OHC
c. H2C=PPh3
15
16
d. HCl
e. ClCH2CN
O
CN
Br
O
CN
Br
O
f. mCPBA
18
17
g. KOtBu
h. TBSCl, imid.
ˆ
final alkene through Wittig homologation utilizing Ph3P CH2
OTBS
OTBS
i. LiHMDS; O2;
then SnCl2, HCl
led solely to 3-phenylbenzofuran 22 (Scheme 4). We postulate
that a betaine intermediate (unobserved in the reaction with
Tebbe reagent presumably due to the high oxophilicity of
titanium) potentially provides a species capable of undergoing
a cyclofragmentation reaction to 22.
NC
O
O
O
Br
Br
24
23
j. DIBAL-H
k. NaH, MeI
With 14 and 26 in hand, the stage was now set to attempt
biaryl coupling of the two fragments and, subsequently, the
key olefin metathesis-based macrocyclization. Efforts to
prepare the boronic acid of either 14 or 26 under standard
conditions (nBuLi, B(OMe)3) failed in numerous attempts.
However, as shown in Scheme 5, the boronate ester of 26 was
OTBS
l. aq. HF
m. Dess–Martin [O]
11
n. Tebbe reagent
MeO
O
MeO
O
Br
Br
26
25
Scheme 3. Preparation of quaternary center fragment 26: a) PhMgBr
(1.0m in THF, 1.3 equiv), THF, À78 !258C, 2h, 100%; b) IBX (2.0 equiv),
THF/DMSO (1/1), 258C, 1 h, 93%; c) methylene triphenylphosphonium
bromide (1.4 equiv), nBuLi (1.6m in hexanes, 1.2equiv), THF, 0 8C, 12h,
86%; d) conc. HCl, EtOH/CH2Cl2 (2/1), 408C, 12h, 100%; e) ClCH 2CN
(3.0 equiv), K2CO3 (2.0 equiv), acetone, 568C, 2h, 75%; f) mCPBA
(3.0 equiv), CH2Cl2, 2 58C, 6 h, 91 %; g) KOtBu (1.0m in THF, 1.5 equiv),
DMF, À578C, 2min; then 10% HCl (excess), À578C, 54%; h) TBSCl
(1.5 equiv), imidazole (2.0 equiv), DMF, 258C, 12h, 99%; i) LiHMDS
(1.0m in THF, 2.0 equiv), THF, À788C, 15 min; O2, 15 min, À788C; 1m
SnCl2 in 10% aq. HCl, À78 !08C, 30 min, 94%; j) DIBAL-H (1.0m in
toluene, 1.5 equiv), toluene, À788C, 1 h, 92%; k) NaH (60% dispersion in
mineral oil, 3.0 equiv), MeI (5.0 equiv), THF, 258C, 24 h, 91%; l) aq. HF,
MeCN, 08C, 1 h, 99%; m) Dess ± Martin periodinane (1.2equiv), CH 2Cl2,
0 !258C, 2h, 96%; n) Tebbe reagent (0.5 m in toluene, 1.5 equiv), THF,
08C, 10 min, 74%. MOM ˆ methoxymethyl, imid. ˆ imidazole, mCPBA ˆ
m-chloroperoxybenzoic acid, TBS ˆ tert-butyldimethylsilyl, LiHMDS ˆ
lithium salt of 1,1,1,3,3,3-hexamethyldisilazane, DIBAL-H ˆ diisobutyl-
aluminum hydride.
a. [Pd(dppf)Cl2]
O
O
O
O
MeO
O
B
B
MeO
B
O
26
O
Br
O
27
b. 14, [Pd(dppf)Cl2]
N
N
O
O
X
16
N
11
Me
O
18
O
N
MeO
MeO
Me
4
3
O
CN
Br
O
N
N
O
CN
Br
iPr
iPr
O
PCy3
Cl
Cl
Cl
KOtBu
-57 °C
N
Ph
(CF3)2MeCO
(CF3)2MeCO
Ru
Ru
Me
Mo
Ph
Cl
Ph
Me
PCy3
PCy3
18
19
28
29
30
cyclization
O
Scheme 5. Biaryl coupling to generate 4 and attempted olefin metatheses
with catalysts 28 ± 30 to provide 3: a) Bis(pinacolato)diboron (1.2equiv),
[Pd(dppf)Cl2] ´ CH2Cl2 (0.15 equiv), KOAc (3.0 equiv), DMSO, 908C, 6 h,
42%; b) 14 (1.0 equiv), [Pd(dppf)Cl2] ´ CH2Cl2 (0.15 equiv), K2CO3
(5.0 equiv), DME, 858C, 2h, 62%. dppf ˆ diphenylphosphanylferrocene,
Cy ˆ cyclohexyl.
-[ CN]
-[H2C=O]
OH
H
-57 °C
NC
NC
O
O
O
Br
Br
Br
22
20
21
generated in modest yield (42%) using the conditions
described by Ishiyama et al.[15] and, gratifyingly, 27 could
readily be coupled with 14 using [Pd(dppf)Cl2] and K2CO3 in
DME at 858C to afford 4 in 62% yield. As expected, H NMR
analysis of 4 indicated a diastereomeric mixture of four
Scheme 4. Novel cyclization reaction to generate 21.
1
The synthesis of 26 from 23 was then completed in six
additional transformations. First, 23 was readily converted to
benzofuranone 24 upon treatment with LiHMDS followed by
compounds due to the mixture of C11 epimers and atropisom-
Angew. Chem. Int. Ed. 2000, 39, No. 19
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
1433-7851/00/3919-3475 $ 17.50+.50/0
3475

COMMUNICATIONS
ers around the newly formed C16 ± C18 biaryl
linkage, confirming a result observed earlier in
a related system.[2c] Unfortunately, although we
were extremely pleased to have prepared 4 in
only 16 linear steps from known starting
materials, our excitement was quickly dashed
when attempted olefin metatheses (toluene,
708C, several days) in the presence of catalysts
28, 29, or 30[16] failed to result in 3 and led only
to recovered starting material or decomposi-
tion. Although we recognized from the outset
that olefin metathesis in this system would
prove challenging, with potential difficulties
arising from the presence of heterocycles con-
taining basic nitrogen atoms and the fact that
styrene derivatives are less favorable substrates
for the reaction,[17] we believe that steric
hindrance close to the double bonds is the
overriding factor which prevents successful
formation of 3.[18]
O
TBDPSO
P(OMe)2
Br
O
Br
O
a. TBAF
N
N
b. I2, Ph3P, imid.
c. (MeO)2P(=O)H,
KHMDS
N
Me
N
Me
31
12
OTBS
OTBS
d. LiBH4
e. MeI, NaH
MeO
O
O
MeO
32
Br
Br
24
f. [Pd(dppf)Cl2], BPD
g. 31, [Pd(dppf)Cl2]
O
O
P(OMe)2
P(OMe)2
TBSO
O
N
O
N
O
h. aq. HF
i. [O]
Despite these discouraging results, we felt
that our overall synthetic plan could be resur-
rected, with little alteration of the previously
developed chemistry, by utilizing an intramo-
lecular variant of the HWE reaction, an
MeO
MeO
N
N
MeO
MeO
Me
Me
34
33
N
j. LiCl,
MeO
DBU
O
À
extremely powerful and mild C C bond-form-
N
Me
ing process which has been widely employed in
organic synthesis.[19] Although we desired to
construct an intermediate such as 5 (see
Scheme 1) possessing the requisite aldehyde
and phosphonate functionalities necessary for
condensation, based on our failure to generate
MeO
35
Scheme 6. Formation of macrocyclic alkene 35: a) TBAF (1.0m in THF, 1.5 equiv), THF,
258C, 15 min, 92%; b) Ph3P (1.7 equiv), imidazole (2.0 equiv), I2 (1.5 equiv), CH2Cl2, 08C,
10 min, 94%; c) (MeO)2P( O)H (2.0 equiv), KHMDS (0.5m in toluene, 1.5 equiv), THF, 08C,
ˆ
10 min, 86%; d) LiBH4 (2.0m in THF, 2.0 equiv), Et2O, 0 8C, 2h, 90%; e) NaH (60%
dispersion in mineral oil, 5.0 equiv), MeI (10.0 equiv), THF, 258C, 12h, 70%; f) bis(pinaco-
lato)diboron (1.2equiv), [Pd(dppf)Cl 2] ´ CH2Cl2 (0.15 equiv), KOAc (3.0 equiv), DMSO, 908C,
4 h, 49%; g) 31 (1.0 equiv), [Pd(dppf)Cl2] ´ CH2Cl2 (0.30 equiv), K2CO3 (5.0 equiv), DME,
858C, 2h, 52%; h) aq. HF, MeCN, 0 8C, 15 min; i) Dess ± Martin periodinane (2.0 equiv),
NaHCO3 (2.0 equiv), CH2Cl2, 0 !258C, 1 h, 85% over two steps; j) DBU (5.0 equiv), LiCl
(5.0 equiv), MeCN, 258C, 6 h, 45%. BPD ˆ bis(pinacolato)diboron, DBUˆ 1,8-diazabicy-
clo[5.4.0]undec-7-ene, KHMDSˆ potassium salt of 1,1,1,3,3,3-hexamethyldisilazane. Selected
NOE interactions shown for 35 confirm the indicated stereochemistry.
ˆ
alkene 26 from 25 using Ph3P CH2 we felt that
the carefully constructed lactol system would
have to be opened prior to the key HWE
reaction to prevent similar formation of a
3-arylbenzofuran product. As such, we initiated
efforts to prepare key intermediate 34, as
delineated in Scheme 6. The indole phospho-
nate 31 was readily generated in 74% overall
yield from the previously synthesized indole ± oxazole 12 by
initial removal of the TBDPS group upon exposure to TBAF,
halogen exchange of the resultant hydroxy group, and mild
formation of the phosphonate residue by SN2displacement of
the primary iodide using the anion generated from dimethyl
phosphite. The other key building block, compound 32, was
synthesized in two steps from 24 in 63% overall yield utilizing
LiBH4 to fully reduce the lactone, followed by methylation of
both the phenolic and the primary hydroxy groups. As before,
the boronate ester of 32 was synthesized in 49% yield using
bis(pinacolato)diboron in the presence of [Pd(dppf)Cl2] and
KOAc, and this intermediate could then be directly coupled to
phosphonate 31 in 52% yield utilizing the Suzuki reaction
protocol employed earlier.[20] Subsequent cleavage of the TBS
ether followed by Dess ± Martin oxidation provided the key
aldehyde phosphonate 34 in 85% yield over the two steps.
Despite the failure of initial attempts to effect macrocycliza-
tion under mildly basic conditions (K2CO3, [18]crown-6),[19b,c]
use of LiCl and DBU[21] smoothly led to the formation of a
new product in just 6 h at ambient temperature in 45% yield.
Although NMR analysis of this HWE product confirmed that
macrocyclization did indeed occur, NOE studies verified that
only a single diastereomer of 34 reacted, providing 35 with the
indicated undesired stereochemistry. As such, this particular
pathway does not represent a viable method to prepare model
system 3 since the previously opened F ring is inaccessible
from 35. However, the power of the HWE reaction to effect
macrocyclization and install such a trans alkene in this highly
strained system was encouraging. As such, we felt that we had
to test whether our earlier reservations about employing
aldehyde phosphonate 5 for an intramolecular HWE reaction
were warranted.
Towards this end, 5 (see Scheme 7) was easily synthesized in
four steps from 25 and 31 utilizing chemistry described earlier
and, most gratifyingly, subsequent treatment of 5 with NaH in
THF at 08C resulted in the formation of macrocycle 3 in 25%
yield with none of the previously feared cyclofragmentation
observed. Although NMR analysis of all intermediates from
3476
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
1433-7851/00/3919-3476 $ 17.50+.50/0
Angew. Chem. Int. Ed. 2000, 39, No. 19

COMMUNICATIONS
Table 1. Selected physical properties for compounds 35 and 3.
underwent the condensation reaction. Despite this unexpect-
ed outcome, the HWE reaction still served as an excellent
means by which to separate the C16 ± C18 atropisomers which
resulted from the Suzuki coupling, providing 3 in atropiso-
merically pure form.
In conclusion, although the synthesis of the heterocyclic
core of diazonamide A initially seemed elusive after numer-
ous failed approaches, both as illustrated here as well as in
several attempts whose description will have to await the full
account of this work, the power of the Suzuki and HWE
35: Rf ˆ 0.12(silica gel, ethyl acetate/hexane 1/1); IR (film): nÄmax ˆ 2912,
2851, 1590, 1487, 1451, 1410, 1371, 1226, 1108, 1010, 795, 703 cmÀ1; 1H NMR
(600 MHz, CD3CN): d ˆ 7.98 (d, J ˆ 17.5 Hz, 1H), 7.46 (s, 1H), 7.44 (dd,
J ˆ 8.3, 0.8 Hz, 1H), 7.42(dd, J ˆ 8.3, 1.3 Hz, 1H), 7.36 (t, J ˆ 7.9 Hz, 1H),
7.25 ± 7.23 (m, 2H), 7.15 (dd, J ˆ 8.1, 1.5 Hz, 1H), 6.98 (t, J ˆ 7.7 Hz, 1H),
6.98 (dd, J ˆ 7.4, 0.9 Hz, 1H), 6.85 (s, 1H), 6.82(dd, J ˆ 7.4, 1.7 Hz, 1H),
5.87 (d, J ˆ 17.5 Hz, 1H), 4.26 (AB, J ˆ 9.2Hz, nab ˆ 64.7 Hz, 2H), 3.81 (s,
3H), 3.31 (s, 3H), 2.53 (s, 3H); 13C NMR (150 MHz, CD3CN): d ˆ 162.0,
157.6, 149.3, 148.2, 146.1, 141.1, 139.4, 139.2, 134.0, 130.4, 130.3, 129.0, 128.8,
127.1, 125.5, 125.4, 123.9, 122.3, 122.3, 117.9, 114.5, 110.8, 104.3, 75.3, 60.3,
60.2, 59.5, 33.7; HR-MS (matrix-assisted laser desorption/ionization) for
‡
‡
À
reactions to effect C C bond formation has ultimately
C30H27N2O3 [M‡H ]: calcd: 463.2016, found: 463.2005.
provided a highly convergent solution for the preparation of
model system 3 in just 17 linear steps from known starting
materials. As such, multigram quantities of both 25 and 31 are
available, and analogues of both fragments for future bio-
logical studies can readily be incorporated into the existing
synthetic pathway. Additionally, the chemistry described for
the preparation of 25 has already been employed[22] to
synthesize a tyrosine-derived benzofuranone fragment pos-
sessing the requisite functionality necessary to complete
diazonamide A. Continuing studies aim at the total synthesis
of diazonamide A and further investigations of the chemical
biology of this important new class of antitumor agents.
3: Rf ˆ 0.18 (silica gel, ethyl acetate/hexane 1/1); IR (film): nÄmax ˆ 2925,
1591, 1492, 1446, 1373, 1225, 1190, 1115, 928, 747 cmÀ1; 1H NMR (600 MHz,
CD3CN): d ˆ 7.53 (dd, J ˆ 8.3, 0.8 Hz, 1H), 7.46 (s, 1H), 7.42± 7.38 (m,
5H), 7.33 ± 7.29 (m, 1H), 7.20 (dd, J ˆ 7.0, 0.9 Hz, 1H), 6.97 (dd, J ˆ 8.5,
1.3 Hz, 1H), 6.95 (dd, J ˆ 8.3, 1.3 Hz, 1H), 6.90 (d, J ˆ 12.7 Hz, 1H), 6.86 (s,
1H), 6.81 (t, J ˆ 7.4 Hz, 1H), 6.46 (d, J ˆ 12.7 Hz, 1H), 5.99 (s, 1H), 3.88 (s,
3H), 3.17 (s, 3H); 13C NMR (150 MHz, CD3CN): d ˆ 159.0, 146.0, 143.0,
138.5, 132.5, 131.9, 131.0, 130.9, 130.2, 129.4, 128.7, 127.9, 127.8, 126.4, 126.3,
124.5, 124.0, 123.3, 123.0, 122.0, 121.6, 119.1, 116.7, 110.9, 61.0, 57.9, 33.7;
‡
HR-MS (matrix-assisted laser desorption/ionization) for C29H23N2O3
‡
[M‡H ]: calcd: 447.1703, found: 447.1707.
O
Received: August 2, 2000 [Z15571]
P(OMe)2
N
TBSO
MeO
OTBS
O
a. [Pd(dppf)Cl2], BPD
[1] N. Lindquist, W. Fenical, G. D. Van Duyne, J. Clardy, J. Am. Chem.
Soc. 1991, 113, 2303 ± 2304.
b. 31, [Pd(dppf)Cl2]
N
Me
O
MeO
O
[2] For other approaches towards the synthesis of diazonamide A, see:
a) X. Chen, L. Esser, P. G. Harran, Angew. Chem. 2000, 112, 967 ± 970;
Angew. Chem. Int. Ed. 2000, 39, 937 ± 940; b) E. Vedejs, J. Wang, Org.
Lett. 2000, 2, 1031 ± 1032; c) E. Vedejs, D. A. Barba, Org. Lett. 2000, 2,
1033 ± 1035; d) P. Magnus, E. G. McIver, Tetrahedron Lett. 2000, 41,
831 ± 834; e) F. Chan, P. Magnus, E. G. McIver, Tetrahedron Lett. 2000,
41, 835 ± 838; f) F. Lach, C. J. Moody, Tetrahedron Lett. 2000, 41,
6893 ± 6896; g) M. C. Bagley, S. L. Hind, C. J. Moody, Tetrahedron
Lett. 2000, 41, 6897 ± 6900; h) M. C. Bagley, C. J. Moody, A. G. Pepper,
Tetrahedron Lett. 2000, 41, 6901 ± 6904; i) H. C. Hang, E. Drotleff,
G. I. Elliott, T. A. Ritsema, J. P. Konopelski, Synthesis 1999, 398 ± 400;
j) P. Magnus, J. D. Kreisberg, Tetrahedron Lett. 1999, 40, 451 ± 454;
k) A. Boto, M. Ling, G. Meek, G. Pattenden, Tetrahedron Lett. 1998,
39, 8167 ± 8170; l) P. Wipf, F. Yokokawa, Tetrahedron Lett. 1998, 39,
2223 ± 2226; m) S. Jeong, X. Chen, P. G. Harran, J. Org. Chem. 1998,
63, 8640 ± 8641; n) C. J. Moody, K. J. Doyle, M. C. Elliott, T. J.
Mowlem, J. Chem. Soc. Perkin Trans. 1 1997, 2413 ± 2419; o) J. P.
Konopelski, J. M. Hottenroth, H. M. Oltra, E. A. VØliz, Z. C. Yang,
Synlett 1996, 609 ± 611; p) C. J. Moody, K. J. Doyle, M. C. Elliott, T. J.
Mowlem, Pure Appl. Chem. 1994, 66, 2107 ± 2110.
[3] Recent examples of macrocyclization resulting from olefin metathesis
include applications to the total synthesis of Sch38516: Z. Xu, C. W.
Johannes, S. S. Salman, A. H. Hoveyda, J. Am. Chem. Soc. 1996, 118,
10926 ± 10927; epothilone A: K. C. Nicolaou, N. Winssinger, J. Pastor,
S. Ninkovic, F. Sarabia, Y. He, D. Vourloumis, Z. Yang, T. Li, P.
Giannakakou, E. Hamel, Nature 1997, 387, 268 ± 272; and manzami-
ne A: S. F. Martin, J. M. Humphrey, A. Ali, M. C. Hillier, J. Am.
Chem. Soc. 1999, 121, 866 ± 867.
[4] P. J. Harrington, L. S. Hegedus, J. Org. Chem. 1984, 49, 2657 ± 2662.
[5] T. Severin, H.-J. Böhme, Chem. Ber. 1968, 101, 2925 ± 2930.
[6] G. A. Roth, E. L. McClymont, Synth. Commun. 1992, 22, 411 ± 420.
[7] See: Ref. [2l] and Y. Oikawa, T. Toshioka, K. Mohri, O. Yonemitsu,
Heterocycles 1979, 12, 1457 ± 1462.
Br
25
36
c. aq. HF
d. Dess–Martin [O]
O
P(OMe)2
N
N
O
O
O
e.
N
N
Me
O
Me
O
MeO
MeO
5
3
Scheme 7. Synthesis of heterocyclic macrocycle 3: a) Bis(pinacolato)di-
boron (1.2equiv), [Pd(dppf)Cl 2] ´ CH2Cl2 (0.15 equiv), KOAc (3.0 equiv),
DMSO, 908C, 4 h, 50%; b) 31 (1.0 equiv), [Pd(dppf)Cl2] ´ CH2Cl2
(0.30 equiv), K2CO3 (3.0 equiv), DME, 858C, 2h, 60%; c) aq. HF, MeCN,
08C, 15 min; d) Dess ± Martin periodinane (2.0 equiv), NaHCO3
(2.0 equiv), CH2Cl2, 2 58C, 8 h, 85% over two steps; e) NaH (2.0 equiv),
THF, 08C, 1 h, 25%. Selected NOE interactions shown for 3 confirm the
indicated stereochemistry.
36 onward indicated a diastereomeric mixture of four com-
pounds (due to a mixture of C11 epimers and atropisomerism
around the C16 ± C18 linkage), similar examination of 3
indicated that only a single diastereomer, which possesses
the stereochemistry shown, resulted from this final trans-
formation. Theoretically, one would expect a 50% yield from
this HWE reaction since the atropisomers cannot interconvert
at the reaction temperatures employed and only half of the
material would have both the aldehyde and phosphonate
groups on the same side of the molecule. Although this
expectation was met, surprisingly only one of the C11 epimers
[8] IBX can readily be prepared from 2-iodobenzoic acid and oxone
following the method of Santagostino et al. see:M. Frigerio, M.
Santagostino, S. Sputore, J. Org. Chem. 1999, 64, 4537 ± 4538.
[9] P. Wipf, C. P. Miller, J. Org. Chem. 1993, 58, 3604 ± 3606.
Angew. Chem. Int. Ed. 2000, 39, No. 19
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
1433-7851/00/3919-3477 $ 17.50+.50/0
3477

COMMUNICATIONS
[10] T. Mizuno, M. Takeuchi, S. Shinkai, Tetrahedron 1999, 55, 9455 ± 9468.
[11] In general, Lewis acid catalysts are required to direct the attack of
nucleophiles towards the more substituted carbon of an epoxide (see:
A. Mordini, S. Bindi, S. Pecchi, A. Capperucci, A. DeglꢁInnocenti, G.
Reginato, J. Org. Chem. 1996, 61, 4466 ± 4468). However, based on
Baldwinꢁs rules, we felt that the greater favorability of 5-exo-tet
cyclization compared to the competing 6-endo-tet reaction would
afford the desired product 21 in this case.
Triplet Di(9-anthryl)carbene Undergoes
Trimerization**
Yasutake Takahashi, Masaaki Tomura,
Ken-ichi Yoshida, Shigeru Murata, and
Hideo Tomioka*
[12] K. C. Nicolaou, S. A. Snyder, A. Bigot, J. A. Pfefferkorn, Angew.
Chem. 2000, 112, 1135 ± 1138; Angew. Chem. Int. Ed. 2000, 39, 1093 ±
1096.
A large number of triplet carbenes have been characterized
by electron paramagnetic resonance (EPR) spectroscopy.[1, 2]
The principal information extracted from EPR spectra of
triplet carbenes in randomly oriented matrices are the zero-
field splitting (ZFS) parameters D and E, where D is a
measure of the number of the unpaired electrons and thus
allows the amount of delocalization in a carbene with
conjugated p systems to be determined, while E measures
the difference in the magnetic dipole interaction and, when
weighted by D, allows one to estimate the bond angle at the
carbene center.
[13] Crystallographic data (excluding structure factors) for structure 23
reported in this paper have been deposited with the Cambridge
Crystallographic Data Centre as supplementary publication no.
CCDC-147867. Copies of the data can be obtained free of charge on
application to CCDC, 12Union Road, Cambridge CB21EZ, UK (fax:
(‡44)1223-336-033; e-mail: deposit@ccdc.cam.ac.uk).
[14] Although the conversion of secondary nitriles into ketones is known
(see: R. W. Freerksen, S. J. Selikson, R. W. Wroble, K. S. Kyler, D. S.
Watt, J. Org. Chem. 1983, 48, 4087 ± 4096), this represents the first
reported application of this methodology for the generation of
lactones.
Among the many triplet carbenes known, triplet di(9-
anthryl)carbene (1) is unique as it shows the smallest D
(0.113 cmÀ1) and E (0.0011 cmÀ1) values ever reported. This
[15] T. Ishiyama, M. Murata, N. Miyaura, J. Org. Chem. 1995, 60, 7508 ±
7510.
[16] M. Scholl, S. Ding, C. W. Lee, R. H. Grubbs, Org. Lett. 1999, 1, 953 ±
956.
[17] For an example of a ring-closing metathesis (RCM) in the presence of
an N atom, see: A. Fürstner, J. Grabowski, C. W. Lehmann, J. Org.
Chem. 1999, 64, 8275 ± 8280. In our case, however, the chelating ability
of the nitrogens in the indole and oxazole rings is likely to be poor and
is not an overriding concern for RCM. For an example of RCM
involving an oxazole, see: K. C. Nicolaou, H. Vallberg, N. P. King, F.
Roschangar, Y. He, D. Vourloumis, C. G. Nicolaou, Chem. Eur. J. 1997,
3, 1957 ± 1970.
•
C
•
C
•
•
[18] For pertinent discussion on the role of sterics in metathesis reactions,
see: a) R. H. Grubbs, S. Chang, Tetrahedron 1998, 54, 4413 ± 4450;
b) R. H. Grubbs, S. J. Miller, G. C. Fu, Acc. Chem. Res. 1995, 28, 446 ±
452.
1
means that the carbene has almost linear and perpendicular
geometry with extensive delocalization of the unpaired
electrons into the anthryl portions of the molecule.[3]
Although these spectroscopic results suggest that the
carbene should exhibit a reactivity that is substantially
different from that observed for other diarylcarbenes, its
chemistry has not been studied extensively. Herein we report
that the carbene undergoes an unusual trimerization reaction,
one that has never been observed before for any other
carbenes.[4]
[19] For representative examples of the power of Wittig and Horner±
Wadsworth ± Emmons reactions to induce intramolecular macrocyc-
lizations of complex substrates, see: a) I. Ernest, J. Gosteli, C. W.
Greengrass, W. Holick, D. E. Jackman, H. R. Pfaendler, R. B. Wood-
ward, J. Am. Chem. Soc. 1978, 100, 8214 ± 8222; b) K. C. Nicolaou, S. P.
Seitz, M. R. Pavia, N. A. Petasis, J. Org. Chem. 1979, 44, 4011 ± 4013;
c) K. C. Nicolaou, R. A. Daines, T. K. Chakraborty, Y. Ogawa, J. Am.
Chem. Soc. 1988, 110, 4685 ± 4696. For a recent review on the Wittig
and HWE reactions in total synthesis, see: K. C. Nicolaou, M. W.
Härter, J. L. Gunzner, A. Nadin, Liebigs Ann. 1997, 1283 ± 1301.
[20] To the best of our knowledge, this represents the first example of a
phosphonate being successfully employed in a Suzuki-type coupling
reaction.
[21] M. A. Blanchette, W. Choy, J. T. Davis, A. P. Essenfeld, S. Masamune,
W. R. Roush, T. Sakai, Tetrahedron Lett. 1984, 25, 2183 ± 2186.
[22] K. C. Nicolaou, A. E. Koumbis, S. A. Snyder, K. B. Simonsen, Angew.
Chem. 2000, 112, 2629 ± 2633; Angew. Chem. Int. Ed. 2000, 39, 2 52 9 ±
2533.
[*] Prof. Dr. H. Tomioka, Prof. Dr. Y. Takahashi, K.-i. Yoshida
Chemistry Department for Materials
Faculty of Engineering, Mie University
Tsu, Mie 514-8507 (Japan)
Fax : (‡81)59-231-9416
E-mail: tomioka@chem.mie-u.ac.jp
Dr. M. Tomura
Research Center for Molecular Materials
Institute for Molecular Science
Okazaki, Aichi 444-8585 (Japan)
Prof. Dr. S. Murata
Department of Basic Science
Graduate School of Arts and Sciences, The University of Tokyo
Meguro, Tokyo 153-8902(Japan)
[**] This work was supported by a Grant-in-Aid for Scientific Research
from the Ministry of Education, Science, Sports, and Culture of Japan.
We thank Prof. Dr. K. Tsujimoto (Japan Advanced Institute of Science
and Technology, Hokuriku) for the FAB-MS measurement.
Supporting information for this article is available on the WWW under
http://www.wiley-vch.de/home/angewandte/ or from the author.
3478
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
1433-7851/00/3919-3478 $ 17.50+.50/0
Angew. Chem. Int. Ed. 2000, 39, No. 19