Studies toward diazonamide A: Initial synthetic forays directed toward the originally proposed structure

Article abstract of DOI:10.1021/ja040090y

A brief introduction into the chemistry of diazonamide A (1) is followed by first-generation sequences to access the originally proposed structure for this unusual marine natural product. These explorations identified a route capable of delivering a model compound possessing the complete heteroaromatic core of the natural product, highlighting in the process several unanticipated synthetic challenges which led both to new methodology as well as an improved synthetic plan that was successfully applied to fully functionalized intermediates.

Full text of DOI:10.1021/ja040090y

Published on Web 07/27/2004
Studies toward Diazonamide A: Initial Synthetic Forays
Directed toward the Originally Proposed Structure
K. C. Nicolaou,* Scott A. Snyder, Xianhai Huang, Klaus B. Simonsen,
Alexandros E. Koumbis, and Antony Bigot
Contribution from the Department of Chemistry and The Skaggs Institute for Chemical Biology,
The Scripps Research Institute, 10550 North Torrey Pines Road, La Jolla, California 92037,
and Department of Chemistry and Biochemistry, UniVersity of California, San Diego,
9500 Gilman DriVe, La Jolla, California 92093
Received March 29, 2004; E-mail: kcn@scripps.edu
Abstract: A brief introduction into the chemistry of diazonamide A (1) is followed by first-generation
sequences to access the originally proposed structure for this unusual marine natural product. These
explorations identified a route capable of delivering a model compound possessing the complete
heteroaromatic core of the natural product, highlighting in the process several unanticipated synthetic
challenges which led both to new methodology as well as an improved synthetic plan that was successfully
applied to fully functionalized intermediates.
Introduction
In 1991, Fenical and Clardy disclosed the structure of
diazonamide A (1, Figure 1), a secondary metabolite isolated
from the colonial ascidian Diazona angulata whose unprec-
edented molecular architecture included a cyclic polypeptide
backbone, a strained halogenated heteroaromatic core trapped
as a single atropisomer, and a lone quaternary center at the
epicenter of its two major macrocyclic subunits.1 Due to the
unique challenges posed by this amazing molecular framework,
in concert with impressive in vitro cytotoxicity (nanomolar
activity against human colon carcinoma and B-16 murine
melanoma cell lines)2 and an inability to harvest additional
material from the original source, this molecule beckoned to
synthetic chemists worldwide. After nearly a decade of prodi-
gious effort from a number of research teams,3-11 including our
own,12 the synthetic gauntlet issued by these motifs was finally
(1) Lindquist, N.; Fenical, W.; Van Duyne, G. D.; Clardy, J. J. Am. Chem.
Soc. 1991, 113, 2303-2304.
Figure 1. The originally proposed structure of diazonamide A (1) and the
revised assignment (2).
(2) For recent explorations into the chemical biology of diazonamide A, see:
(a) Cruz-Monserrate, Z.; Vervoort, H. C.; Bai, R.; Newman, D. J.; Howell,
S. B.; Los, G.; Mullaney, J. T.; Williams, M. D.; Pettit, G. R.; Fenical,
W.; Hamel, E. Mol. Pharmacol. 2003, 63, 1273-1280. (b) Cruz-Monserrate,
Z.; Mullaney, J. T.; Harran, P. G.; Pettit, G. R.; Hamel, E. Eur. J. Biochem.
2003, 270, 3822-3828.
met with the first successful synthesis of the proposed structure
of diazonamide A by the Harran group at the Southwestern
Medical Center in Dallas, Texas.13
(3) (a) Li, J.; Chen, X.; Burgett, A. W. G.; Harran, P. G. Angew. Chem., Int.
Ed. 2001, 40, 2682-2685. (b) Chen, X.; Esser, L.; Harran, P. G. Angew.
Chem., Int. Ed. 2000, 39, 937-940. (c) Jeong, S.; Chen, X.; Harran, P. G.
J. Org. Chem. 1998, 63, 8640-8641.
(9) (a) Lach, F.; Moody, C. J. Tetrahedron Lett. 2000, 41, 6893-6896. (b)
Bagley, M. C.; Hind, S. L.; Moody, C. J. Tetrahedron Lett. 2000, 41, 6897-
6900. (c) Bagley, M. C.; Moody, C. J.; Pepper, A. G. Tetrahedron Lett.
2000, 41, 6901-6904. (d) Moody, C. J.; Doyle, K. J.: Elliott, M. C.;
Mowlem, T. J. J. Chem. Soc., Perkin Trans. 1 1997, 2413-2419. (e)
Moody, C. J.; Doyle, K. J.; Elliott, M. C.; Mowlem, T. J. Pure Appl. Chem.
1994, 66, 2107-2110.
(4) (a) Vedejs, E.; Zajac, M. A. Org. Lett. 2001, 3, 2451-2454. (b) Vedejs,
E.; Wang, J. Org. Lett. 2000, 2, 1031-1032. (c) Vedejs, E.; Barba, D. A.
Org. Lett. 2000, 2, 1033-1035.
(5) (a) Wipf, P.; Methot, J.-L. Org. Lett. 2001, 3, 1261-1264. (b) Wipf, P.;
Yokokawa, F. Tetrahedron Lett. 1998, 39, 2223-2226.
(6) (a) Kreisberg, J. D.; Magnus, P.; McIver, E. G. Tetrahedron Lett. 2001,
42, 627-629. (b) Magnus, P.; McIver, E. G. Tetrahedron Lett. 2000, 41,
831-834. (c) Chan, F.; Magnus, P.; McIver, E. G. Tetrahedron Lett. 2000,
41, 835-838. (d) Magnus, P.; Kreisberg, J. D. Tetrahedron Lett. 1999,
40, 451-454.
(10) (a) Hang, H. C.; Drotleff, E.; Elliott, G. I.; Ritsema, T. A.; Konopelski, J.
P. Synthesis 1999, 398-400. (b) Konopelski, J. P.; Hottenroth, J. M.; Oltra,
H. M.; Veliz, E. A.; Yang, Z. C. Synlett 1996, 609-611.
(11) Boto, A.; Ling, M.; Meek, G.; Pattenden, G. Tetrahedron Lett. 1998, 39,
8167-8170.
(7) Fuerst, D. E.; Stoltz, B. M.; Wood, J. L. Org. Lett. 2000, 2, 3521-3523.
(8) (a) Schley, S.; Radspieler, A.; Christoph, G.; Liebscher, J. Eur. J. Org.
Chem. 2002, 369-374. (b) Radspieler, A.; Liebscher, J. Synthesis 2001,
745-750.
(12) (a) Nicolaou, K. C.; Snyder, S. A.; Simonsen, K. B.; Koumbis, A. E. Angew.
Chem., Int. Ed. 2000, 39, 3473-3478. (b) Nicolaou, K. C.; Huang, X.;
Giuseppone, N.; Bheema Rao, P.; Bella, M.; Reddy, M. V.; Snyder, S. A.
Angew. Chem., Int. Ed. 2001, 40, 4705-4709.
9
10162
J. AM. CHEM. SOC. 2004, 126, 10162-10173
10.1021/ja040090y CCC: $27.50 © 2004 American Chemical Society

Studies toward Diazonamide A
A R T I C L E S
This accomplishment, though, would not constitute the final
chapter in the intriguing story of diazonamide A, as fully
synthetic 1 did not match natural diazonamide A, revealing that
the original structural assignment based on NMR, X-ray
crystallographic, and mass spectral data was in error. Their
suggested revision (2, Figure 1), one which included alteration
of the terminal amino acid, exchange of the heteroatom in ring
F, and the addition of a tenth ring (ring H) to the natural
product’s architecture, launched a series of new synthetic
campaigns worldwide.14 Just six months later, in August of
2002, the first of these research programs reached fruition in
our laboratory, unequivocally proving the correct molecular
connectivities for diazonamide A as 2.15 A few months
thereafter, our group completed a second total synthesis of this
intriguing natural product through an entirely different synthetic
approach,16 one that shared with the first a reliance upon
carefully tailored strategies, cascade reactions, and methodolo-
gies to handle its most difficult motifs.17,18
In this and the following article in this issue,19a as well as in
two articles soon to be published,19b,c we present a complete
chronicle of our five-year campaign to synthesize the original
(1) as well as the revised structure (2) of diazonamide A and
concurrently explore this natural product’s intriguing chemical
biology using the developed sequences. We begin in this article
with our first-generation strategy to address the structural
complexity posed by 1. While this approach would ultimately
prove unsuccessful, its prosecution identified and solved a
number of key synthetic challenges posed by the diazonamide
framework and resulted in several new synthetic methods and
tactics which would prove to have much utility in our later drive
to complete the total synthesis of 2.
axes are trapped as single atropisomers, but also as a clue that
forming these ring systems in the laboratory is likely to be quite
difficult. Consequently, we tailored virtually all of our retrosyn-
thetic decisions simply around the questions of how and in what
order to form these two formidable rings, rather than focus
heavily on any other specific motif within 1.
Scheme 1 provides our proposed solution to the original
diazonamide problem, starting with two relatively minor
modifications: excision of the L-valine side chain appended to
the C-2 amine and removal of the two aryl chlorines. The first
of these simplifications was implemented because the C-37
stereocenter of 1 remained unassigned,1 thus enabling late-stage
incorporation of both the L and D forms of 3 to verify the
stereochemistry of that position, while the second reflected
concerns about long-term stability to diverse reaction con-
ditions.3a,12b With these operations leading to 4, we then un-
locked what we regarded to be the simpler of the two macro-
cycles to construct in the forward sense, the 12-membered AG-
ring system, at the most obvious site: its central amide bond.20
This operation revealed intermediate 5, a new goal structure in
which the oxidation state of the carboxylic acid needed for
macrolactamization has been adjusted to a protected alcohol.
Having elected to pursue these initial retrosynthetic simpli-
fications, we next turned our attention to the task of disas-
sembling the heteroaromatic core of 5. Although a number of
possible points for ring dissection were conceivable, some
possessing inherently far more forward synthetic risk than others,
a particularly flexible and practical approach became evident
when the A-ring oxazole was opened through a Robinson-
Gabriel cyclodehydration transform to reveal a ketoamide in
reduced form as 6. Indeed, if the functionality needed for
oxazole formation could be built from an alkene precursor such
as 7, then perhaps macrocyclization could be effected at
C29-C30 through either ring-closing olefin metathesis21 be-
tween the two terminal alkenes in 8 or, more classically, via an
intramolecular Horner-Wadsworth-Emmons (HWE) reaction22
using 9. In both cases, we expected that these macrocyclization
events would be atropselective as a consequence of the C-10
stereochemistry and π-stacking between the A-ring oxazole and
the E-ring, achievable only if the rings are oriented in the desired
fashion. Equally enticing from a more general strategic stand-
point, these two potential macrocyclization precursors could be
retrosynthetically traced to the same two generic building blocks,
indole-oxazole 10 and EFG building block 11, simply by
severing their C16-C18 biaryl linkage through a biaryl coupling
transform.
Results and Discussion
1. Retrosynthetic Analysis. Although several of the most
forbidding structural elements possessed by the originally
proposed structure of diazonamide A (1) are quite clear in a
two-dimensional representation, no paper-based drawing can
adequately convey its overall rigidity and compactness, espe-
cially as imposed by its two 12-membered macrocyclic rings.
An appreciation for this fact is of critical importance, not only
as an explanation for why the C16-C18 and C24-C26 biaryl
(13) (a) Li, J.; Jeong, S.; Esser, L.; Harran, P. G. Angew. Chem., Int. Ed. 2001,
40, 4765-4770. (b) Li, J.; Burgett, A. W. G.; Esser, L.; Amezcua, C.;
Harran, P. G. Angew. Chem., Int. Ed. 2001, 40, 4770-4773.
(14) (a) Vedejs, E.; Zajac, M. A. Org. Lett. 2004, 6, 237-240. (b) Sawada, T.;
Fuerst, D. E.; Wood, J. L. Tetrahedron Lett. 2003, 44, 4919-4921. (c)
Feldman, K. S.; Eastman, K. J.; Lessene, G. Org. Lett. 2002, 4, 3525-
3528.
(15) Nicolaou, K. C.; Bella, M.; Chen, D. Y.-K.; Huang, X.; Ling, T.; Snyder,
S. A. Angew. Chem., Int. Ed. 2002, 41, 3495-3499.
(16) Nicolaou, K. C.; Bheema Rao, P.; Hao, J.; Reddy, M. V.; Rassias, G.;
Huang, X.; Chen, D. Y.-K.; Snyder, S. A. Angew. Chem., Int. Ed. 2003,
42, 1753-1758.
(17) A third total synthesis of diazonamide A, proceeding in a longest linear
sequence of 19 steps, was recently accomplished: Burgett, A. W. G.; Li,
Q.; Wei, Q.; Harran, P. G. Angew. Chem., Int. Ed. 2003, 42, 4961-4966.
(18) For highlights of synthetic studies towards the diazonamides, see: (a)
Nicolaou, K. C.; Snyder, S. A. Classics in Total Synthesis II; Wiley-VCH:
Weinheim, 2003; Ch. 20, pp 550-588. (b) Wittmann, V. Nachr. Chem.
2002, 50, 477-482. (c) Ritter, T.; Carreira, E. M. Angew. Chem., Int. Ed.
2002, 41, 2489-2495.
(19) (a) Nicolaou, K. C.; Snyder, S. A.; Giuseppone, N.; Huang, X.; Bella, M.;
Reddy, M. V.; Bheema Rao, P.; Koumbis, A. E.; Giannakakou, P.; O’Brate,
A. J. Am. Chem. Soc. 2004, 126, 10174-10182. (b) Nicolaou, K. C.;
Chen, D. Y.-K.; Huang, X.; Ling, T.; Bella, M.; Snyder, S. A. J. Am. Chem.
Soc., in press. (c) Nicolaou, K. C.; Hao, J.; Reddy, M. V.; Bheema Rao,
P.; Rassias, G.; Snyder, S. A.; Huang, X.; Chen, D. Y.-K.; Brenzovich, W.
E.; Giuseppone, N.; Giannakakou, P.; O’Brate, A. J. Am. Chem. Soc., in
press.
As such, the synthetic challenges posed by diazonamide A
(1) have been reduced to the construction of two fragments of
(20) Examples of such challenging macrolactamizations can be found in: Boger,
D. L.; Kim, S. H.; Mori, Y.; Weng, J.-H.; Rogel, O.; Castle, S. L.; McAtee,
J. J. J. Am. Chem. Soc. 2001, 123, 1862-1871.
(21) For selected examples as applied to the total synthesis of complex molecules,
see: (a) Nicolaou, K. C.; Snyder, S. A. Classics in Total Synthesis II; Wiley-
VCH: Weinheim, 2003; Ch. 7, pp 161-210. (b) Love, J. A. In Handbook
of Metathesis; Grubbs, R. H., Ed.; Wiley-VCH: Weinheim, 2003; pp 296-
322.
(22) For representative examples of the power of Wittig and Horner-
Wadsworth-Emmons reactions to induce intramolecular macrocyclizations
of complex substrates, see: (a) Ernest, I.; Gosteli, J.; Greengrass, C. W.;
Holick, W.; Pfaendler, H. R.; Woodward, R. B. J. Am. Chem. Soc. 1978,
100, 8214-8222. (b) Nicolaou, K. C.; Seitz, S. P.; Pavia, M. R.; Petasis,
N. A. J. Org. Chem. 1979, 44, 4011-4013. (c) Nicolaou, K. C.; Daines,
R. A.; Chakraborty, T. K.; Ogawa, Y. J. Am. Chem. Soc. 1988, 110, 4685-
4696. For a recent review on the Wittig and HWE reactions in total
synthesis, see: (d) Nicolaou, K. C.; Harter, M. W.; Gunzner, J. L.; Nadin,
A. Liebigs Ann. 1997, 1283-1301.
9
J. AM. CHEM. SOC. VOL. 126, NO. 32, 2004 10163

A R T I C L E S
Nicolaou et al.
Scheme 1. Retrosynthetic Analysis of the Originally Proposed Structure of Diazonamide A (1)
relatively equal complexity and size in what is overall a
convergent synthetic blueprint. With the synthesis of compounds
of type 10 expected to pose few difficulties, the only major
concern that remained was how to construct the C-10 quaternary
center of 11 stereoselectively, a problem that published studies
toward 1 (ca. 1999)3c,5b,6d,9d,e,10,11 had failed to solve with
resounding success. Our analysis suggested that if a biaryl
precursor of general structure 12 could be coaxed to form the
indicated anion (using a judiciously chosen substituent X to
adjust the pKa of that site), and then a subsequent 5-exo-tet
cyclization could provide the structure of the desired intermedi-
ate in a single operation.23 If true, then absolute stereocontrol
could result from this key reaction if the epoxide within 12 was
formed stereoselectively, assuming, of course, that the distal
stereocenter was an innocent bystander during the event.
construct any appropriately decorated EFG fragment. With this
in mind, we prepared a variety of biaryl epoxides bearing
different alkoxymethylene units (16, 17, 18, 20, and 21, see
Scheme 2) over the course of a few self-explanatory operations
from aldehyde 1324 and exposed each to LDA in THF and
KOt-Bu in DMF at a variety of temperatures. Table 1 sum-
marizes the key results from these studies.
Under no circumstance tested could we induce epoxides
bearing either a SPh (16) or a TMS-activated methylene (20)
to cyclize. However, if the sulfur atom of 16 was oxidized to
its sulfoxide counterpart (17), then the action of LDA in THF
at -78 °C for 5 min, followed by a quench with 1 M aqueous
HCl at that temperature, delivered the desired 5-exo-tet product
(A) in 56% yield (entry 2). Similar levels of success were
achieved with cyanide 21 using KOt-Bu in DMF at -57 °C
(entry 8). Intriguingly, in both cases, the remainder of the
material balance was accounted for not through decomposition
or recovered starting material, but instead by the formation of
a singular byproduct: 3-phenylbenzofuran 22.
2. Studies Directed toward the EFG Fragment. Armed with
this plan, we immediately sought to explore the feasibility of
its key steps by constructing a series of model compounds that
differed from what we would consider to be the fully function-
alized intermediates of Scheme 1, simply with the replacement
of an unadorned phenyl group for ring G. Thus, the first critical
task was to deduce whether 5-exo-tet cyclization of compounds
of type 12 (see Scheme 1) was actually possible, since the
success or failure of this strategy would dictate our ability to
Scheme 3 provides one possible mechanistic explanation for
this outcome, using compound 21 for illustration purposes.
Simply put, 3-phenylbenzofuran 22 could result if a portion of
alkoxide intermediate 24 underwent a Grob-like fragmentation,25
expelling both formaldehyde and cyanide anion prior to being
quenched with HCl. Such a side-reaction would appear quite
(23) In general, Lewis acid catalysts are required to direct the attack of
nucleophiles toward the more substituted carbon of an epoxide (see:
Mordini, A.; Bindi, S.; Pecchi, S.; Capperucci, A.; Degl’Innocenti, A.;
Reginato, G. J. Org. Chem. 1996, 61, 4466-4468). However, on the basis
of Baldwin’s rules and biaryl activation of the epoxide, we felt that the
greater preference of 5-exo-tet cyclization compared to the competing
6-endo-tet reaction would afford the desired product in this case.
(24) Mizuno, T.; Takeuchi, M.; Shinkai, S. Tetrahedron 1999, 55, 9455-9468.
(25) For the original discovery of this process, see: (a) Grob, C. A.; Baumann,
W. HelV. Chim. Acta 1955, 38, 594-610. For a review of its use in chemical
synthesis, see: (b) Grob, C. A. Angew. Chem., Int. Ed. Engl. 1969, 8, 535-
546.
9
10164 J. AM. CHEM. SOC. VOL. 126, NO. 32, 2004

Studies toward Diazonamide A
A R T I C L E S
Scheme 2. Synthesis of Various Biaryl Epoxides To Evaluate the
Potential for a 5-exo-tet Cyclization To Establish the C-10
Quaternary Center of Diazonamide Aa
Scheme 3. Proposed Mechanism To Account for the Formation of
22 and/or 25 Following the Treatment of Epoxide 21 with Base
a Reagents and conditions: (a) PhMgBr (1.0 M in THF, 1.3 equiv), THF,
-78 f 25 °C, 2 h, 100%; (b) IBX (2.0 equiv), THF/DMSO (1:1), 25 °C,
1 h, 93%, or MnO2 (10 equiv), CH2Cl2, 25 °C, 12 h, 95%; (c) concentrated
HCl, MeOH/CH2Cl2 (1:1), 25 °C, 12 h, 97%; (d) PhSCH2Cl (2.0 equiv),
K2CO3 (1.5 equiv), DMF, 25 °C, 12 h, 93%; (e) trimethylsulfonium iodide
(1.5 equiv), KOtBu (1.0 M in THF, 1.4 equiv), DMSO, 0 °C, 10 min, 96%;
(f) mCPBA (1.1 equiv), NaHCO3 (3.0 equiv), CH2Cl2, -10 °C, 30 min,
86%; (g) mCPBA (2.5 equiv), NaHCO3 (5.0 equiv), CH2Cl2, 25 °C, 1 h,
91%; (h) methylene triphenylphosphonium bromide (1.4 equiv), nBuLi (1.6
M in hexanes, 1.2 equiv), THF, 0 °C, 12 h, 86%; (i) concentrated HCl,
MeOH/CH2Cl2 (1:1), 25 °C, 12 h; (j) ClCH2CN (3.0 equiv) or ClCH2TMS
(3.0 equiv), K2CO3 (2.0 equiv), acetone, 56 °C, 12 h; (k) mCPBA (3.0
equiv), NaHCO3 (5.0 equiv), CH2Cl2, 25 °C, 6 h, 68% over two steps for
20, 71% over two steps for 21. IBX ) o-iodoxybenzoic acid; DMF ) N,N-
dimethylformamide; mCPBA ) meta-chloroperoxybenzoic acid; TMS )
trimethylsilyl.
improved yield of 25 when the temperature was lowered to -78
°C (entry 9, Table 1). It is further supported by a crystal structure
of compound 26, proving the exclusive trans-disposition of the
hydroxymethylene and cyanide groups (see ORTEP drawing,
Scheme 3).
Needless to say, this initial series of results was quite exciting
on many fronts. First, because the 5-exo-tet cyclization reaction
could be driven cleanly to form 3-phenylbenzofuran products,
an important pharmacophore found in a number of natural
products and drug leads,26 we were quite content that we had
uncovered a new approach for their synthesis, one which
ultimately proved amenable to both parallel solution and solid-
phase platforms.27 Second, since the X-ray crystal structure of
compound 26 revealed that we had only formed one group of
diastereomers from the cyclization, we knew that if the epoxide
in compounds such as 17 or 21 could be accessed in enantiopure
form, then this approach would deliver the C-11 quaternary
center of 1 enantioselectively. Finally, because more than one
methylene activating group had succeeded in our 5-exo-tet
studies, we had several options in C-11 functionalization to
achieve the synthesis of a complete EFG fragment.
Pressing forward, we elected to use intermediate 26 (the TBS-
protected variant of 25) to explore the next critical stage of our
general synthetic approach: forming the heteroaromatic core
of diazonamide A through ring-closing olefin metathesis. Our
goal was to convert this intermediate into a suitable partner for
a metal-mediated union of the Suzuki or Stille type, and, as
shown in Scheme 4, that objective was accomplished over
several steps. First, cognizant of the ability of LiHMDS and
oxygen to convert secondary nitriles into ketones,28 we thought
the same reaction conditions might be sufficient to transform
compound 26 into its corresponding lactone. Pleasingly, this
Table 1. Exploration of the 5-exo-tet Cyclization To Establish the
C-10 Quaternary Center of Diazonamide A
conditions
(5 min reaction;
then HCl quench)
yield (%)
5-exo-tet
product (A)
yield (%)
entry
compound
22
1
2
3
4
5
6
7
8
9
16: R ) SPh
17: R ) SOPh
17: R ) SOPh
LDAa, THF, -78 °C
LDA, THF, -78 °C
KOtBu, DMF, -57 °C
recovered s.m.
56
0
0
28
93
91
18: R ) SO2Ph KOtBu, DMF, -57 °C
20: R ) TMS
21: R ) CN
21: R ) CN
21: R ) CN
21: R ) CN
LDA, THF, -78 f 25 °C
LDA, THF, -78 °C
KOtBu, DMF, 0 °C
KOtBu, DMF, -57 °C
KOtBu, THF/DMF (1:1),
-78 °C
recovered s.m.
decomposition
0
45
76
95
52
12
a LDA ) lithium diisopropylamide.
(26) (a) Cagniant, P.; Cagniant, D. In AdVances in Heterocyclic Chemistry;
Katritzky, A. R., Boulton, A. J., Eds.; Academic Press: New York, 1975;
Vol. 18, pp 337-482. (b) Mustafa, A. In Chemistry of Heterocyclic
Compounds; Weissberger, A., Taylor, E. C., Eds.; Wiley: New York, 1974;
Vol. 29, pp 1-514.
reasonable on the basis of first principles, as the process creates
three fragments from one molecule of starting material and the
final product (22) gains enthalpically from aromatic stabilization.
This picture is also consistent with experimental findings such
as the exclusive formation of 22 when the reaction was
conducted at higher temperature (entry 7, Table 1) and the
(27) Nicolaou, K. C.; Snyder, S. A.; Bigot, A.; Pfefferkorn, J. A. Angew. Chem.,
Int. Ed. 2000, 39, 1093-1096.
(28) Freerksen, R. W.; Selikson, S. J.; Wroble, R. R.; Kyler, K. S.; Watt, D. S.
J. Org. Chem. 1983, 48, 4087-4096.
9
J. AM. CHEM. SOC. VOL. 126, NO. 32, 2004 10165

A R T I C L E S
Nicolaou et al.
Scheme 4. Conversion of Advanced Model Compound 25 into the
Complete EFG System (31)a
Scheme 5. Alternate Racemic Synthesis of Advanced EFG
Fragment 31a
a Reagents and conditions: (a) 32 (1.0 equiv), 33 (1.0 equiv), H2SO4
(70% aq), 0 f 70 °C, 2 h, 42% (95% based on recovered 33); (b) LiHMDS
(1.0 M in THF, 1.5 equiv), HCHO (3.0 equiv), THF, 0 f 25 °C, 1 h; then
TBSOTf (2.0 equiv), Et3N (2.0 equiv), 10 min, 75%.
HF in MeCN, oxidizing the resultant alcohol with Dess-Martin
periodinane and then olefinating with the Tebbe reagent. The
model EFG fragment (31) was then completed in 42% yield
by exchanging the aryl bromide for a boronate ester using
Miyaura’s conditions [Pd(dppf)Cl2‚CH2Cl2, bis(pinacolato)-
diboron, and KOAc] in DME at 90 °C.29
a Reagents and conditions: (a) TBSCl (1.5 equiv), imidazole (2.0 equiv),
DMF, 25 °C, 12 h, 92%; (b) LiHMDS (1.0 M in THF, 2.0 equiv), THF,
-78 °C, 15 min; O2, 15 min, -78 °C; 1 M SnCl2 in 10% aq HCl, -78 f
0 °C, 30 min, 94%; (c) DIBAL-H (1.0 M in toluene, 1.5 equiv), toluene,
-78 °C, 1 h, 92%; (d) Ag2O, MeI (5.0 equiv), CH2Cl2, 25 °C, 2.5 h, 72%;
(e) aq HF (48%, excess), MeCN, 0 °C, 1 h, 99%; (f) Dess-Martin
periodinane (1.2 equiv), CH2Cl2, 0 f 25 °C, 2 h, 96%; (g) Tebbe reagent
(0.5 M in toluene, 1.5 equiv), THF, 0 °C, 10 min, 74%; (h) bis(pinacola-
to)diboron (1.2 equiv), Pd(dppf)Cl2 (0.15 equiv), KOAc, DMSO, 90 °C, 6
h, 42%. TBS ) tert-butyldimethylsilyl; LiHMDS ) lithium bis(trimeth-
ylsilyl)amide; DIBAL-H ) diisobutylaluminum hydride, BDP ) bis(pina-
colato)diboron; dppf ) diphenylphosphinoferrocene.
Before we describe efforts to synthesize this compound’s
coupling partner, there are several observations related to these
final reactions that are worth mentioning. First, the use of any
basic fluoride source (such as TBAF) to remove the silyl group
from 28 led to exclusive formation of 3-phenylbenzofuran 22.
The same outcome also occurred when attempts were made to
olefinate the intermediate aldehyde under Wittig conditions,
presumably due to the existence of a betaine intermediate
unobserved in the reaction with Tebbe reagent based on the
higher oxophilicity of titanium. Taken in conjunction with the
earlier studies in the context of Table 1, these two findings reveal
the impressive fragility of functionalized systems such as 25
under basic conditions and highlight an issue that will reappear
in later steps of the sequence. Second, attempts to improve the
yield of boronation through more “standard” protocols, such as
lithiation followed by a quench with B(OMe)3, failed numerous
attempts, leading in all cases to reduced material (i.e., dehalo-
genation). Third, while the overall sequence to 31 was quite
efficient for its length (15 steps), supplies of this key inter-
mediate could be enriched through the two-step synthesis of
lactone 27 shown in Scheme 5.30 Of course, while this alter-
nate approach was effective for forming racemic 31 in just 8
steps for our model studies, we did not expect that it could
prove useful in applications to access fully functionalized
diazonamide intermediates, both for its failure to offer any
means to control the C-11 stereochemistry and for its overall
harshness (which would likely be incompatible with an elabo-
rated G-ring).
conjecture proved to be true, as we observed the smooth
formation of intermediate 27 in 94% yield following a SnCl2
workup to break apart the initially formed peroxide intermediate.
With this new reactive handle unveiled, partial reduction to the
lactol, as effected with DIBAL-H in toluene at -78 °C, followed
by methylation with Ag2O and MeI in CH2Cl2 at 25 °C, served
to provide a protected variant of the diazonamide F-ring (28)
in 66% yield over two steps. In addition to 28, these conditions
also gave rise to the chromatographically separable C-11 epimer
29 and aldehyde 30, each of which were obtained in 12% yield.
Several other methylation procedures were probed at this stage
in hopes of improving the yield of 28, and while most could
suppress the formation of 30 (such as NaH, MeI, THF, 0 °C),
none provided a better relative ratio of 28:29. As such, we
utilized the Ag2O-based protocol to process material. Although
one might question our concern over commanding the C-11
stereochemistry at this stage in the synthesis, as the correct
disposition of the F-ring lactol in diazonamide A (1) is likely
to be the product of thermodynamic control, it stemmed from
fears that the bulk at C-11 could impact later attempts at
macrocyclization. Specifically, we felt that our chances for
success in these likely difficult ring closures would best be
enhanced by using material in which groups attached to C-11
were pointed as far away as possible from the reactive centers
needed for cyclization (vide infra).
(29) Ishiyama, T.; Murata, M.; Miyaura, N. J. Org. Chem. 1995, 60, 7508-
7510.
(30) The opening operation was patterned after that reported by: Padwa, A.;
Dehm, D.; Oine, T.; Lee, G. A. J. Am. Chem. Soc. 1975, 97, 1837-1845.
The addition of HCHO was achieved using the protocol reported by:
Kobayashi, S.; Hachiya, I. J. Org. Chem. 1994, 59, 3590-3596. For a
related example, see: Bernardelli, P.; Moradei, O. M.; Friedrich, D.; Yang,
J.; Gallou, F.; Dyck, B. P.; Doskotch, R. W.; Lange, T.; Paquette, L. A. J.
Am. Chem. Soc. 2001, 123, 9021-9032.
With ample amounts of intermediate 28 in hand, the instal-
lation of the alkene needed for metathesis was achieved over
three steps in 70% yield by cleaving the TBS ether with aqueous
9
10166 J. AM. CHEM. SOC. VOL. 126, NO. 32, 2004

Studies toward Diazonamide A
A R T I C L E S
Scheme 7. Preparation of Indole-oxazole Fragment 53a
Scheme 6. Conversion of Advanced Model Compound 35 into an
Alternate EFG System (38)a
a Reagents and conditions: (a) dimethylamino-2-nitroethylene (1.1 equiv),
TFA, 25 °C, 30 min; (b) LiAlH4 (1.0 M in THF, 6.0 equiv), THF, 25 f 65
°C, 4 h, 87% over two steps; (c) 48 (1.0 equiv), EDC (1.0 equiv), HOBt
(1.0 equiv), CH2Cl2, 25 °C, 2 h, 79%; (d) DDQ (3.0 equiv), THF/H2O (9:
1), 0 °C, 3 h; (e) IBX (3.0 equiv), THF/DMSO (1:1), 25 °C, 3 h, 90% over
two steps; (f) Cl3CCCl3 (2.0 equiv), Ph3P (2.0 equiv), Et3N (4.0 equiv), 25
°C, 15 min; (g) MeI (4.0 equiv), nBu4NHSO4 (1.1 equiv), C6H6, 10% aq
NaOH, 25 °C, 20 min, 88% over two steps; (h) TBAF (1.0 M in THF, 1.5
equiv), THF, 25 °C, 15 min, 92%; (i) IBX (3.0 equiv), THF/DMSO (1:1),
25 °C, 3 h, 90%; (j) methylene triphenylphosphonium bromide (1.5 equiv),
nBuLi (1.6 M in hexanes, 1.3 equiv), THF, 0 °C, 15 min, 89%. TFA )
trifluoroacetic acid; EDC ) 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide
hydrochloride; TBDPS ) tert-butyldiphenylsilyl; HOBt ) 1-hydroxyben-
zotriazole trihydrate; DDQ ) 2,3-dichloro-5,6-dicyano-1,4-benzoquinone;
TBAF ) tetra n-butylammoniumfluoride.
a Reagents and conditions: (a) IBX (1.3 equiv), THF/DMSO (1:1), 25
°C, 1 h, 83%; (b) Tebbe reagent (0.5 M in toluene, 2.2 equiv), THF, -20
f 0 °C, 1 h, 78%; (c) [bis(trifluoroacetoxy)iodo]benzene (1.5 equiv),
MeOH, 25 °C, 3 h, 83%. BTIB ) [bis(trifluoroacetoxy)iodo]benzene.
As a final aside, significant effort was also expended to
advance 5-exo-tet product 35 to the same EFG model system
(31). As revealed in Scheme 6, these forays began by oxidizing
35 to aldehyde 36 using o-iodoxybenzoic acid (IBX).31 This
intermediate was then treated with excess Tebbe reagent (2.2
equiv) in toluene at 0 °C for 1 h in hopes of forming the requisite
alkene needed to probe eventual olefin metathesis. While this
transformation would proceed smoothly, these conditions also
led to the reduction of the sulfoxide to the corresponding sulfide,
presumably through the indicated mechanism proceeding by
way of intermediates 39-42, to provide 37 in 78% yield.
Although unexpected, this outcome was quite fortuitous because
it allowed immediate probes aimed at replacing the phenylsulfide
in 37 with methoxide using bis(trifluoroacetoxy)iodobenezene
in methanol.32 Sadly, any initial excitement was soon dashed
because this exchange only provided material with the stereo-
chemical disposition of 38, as verified by nOe comparisons to
similar compounds obtained from intermediates 28 and 29. As
such, we viewed this product (38), and the overall sequence
from 35, as useless for further studies toward diazonamide A
on the basis of the considerations mentioned above. On the
positive side, however, these explorations did induce us to
explore the sulfoxide deoxygenation process (36 f 37) in more
detail, ultimately leading to a new, general, and mild synthetic
method for the reduction of diverse and heavily functionalized
sulfoxides, selenoxides, and N-oxides using titanocene meth-
ylidenes.33
3. Synthesis of the BCD Indole-Oxazole Fragment. Despite
the problems encountered within the context of Scheme 6, since
we had an efficient sequence available to prepare the desired
EFG fragment, effort was immediately directed at preparing an
appropriate coupling partner, namely BCD indole-oxazole 53.
As shown in Scheme 7, explorations to access this fragment
began from the known 4-bromoindole (44),34 a starting material
that already contained the C- and D-rings of the needed
compound. As such, only an oxazole had to be appended to
(31) IBX is readily and safely prepared from 2-iodobenzoic acid and OXONE
following the method of: Frigerio, M.; Santagostino, M.; Sputore, S. J.
Org. Chem. 1999, 64, 4537-4538.
(32) Moriarty, R. M.; Kosmeder, J. W. In Encyclopedia of Reagents for Organic
Synthesis; Paquette, L. A., Ed.; John Wiley & Sons: Chichester, 1995;
Vol. 6, pp 3982-3984.
(33) Nicolaou, K. C.; Koumbis, A. E.; Snyder, S. A.; Simonsen, K. B. Angew.
Chem., Int. Ed. 2000, 39, 2529-2533. For an earlier report of one of the
reactions disclosed in this communication, the conversion of pyridine
N-oxides to 2-methylpyridines, using a tantalum methylidene, see: Mullins,
S. M.; Bergman, R. G.; Arnold, J. Organometallics 1999, 18, 4465-4467.
(34) Harrington, P. J.; Hegedus, L. S. J. Org. Chem. 1984, 49, 2657-2662.
9
J. AM. CHEM. SOC. VOL. 126, NO. 32, 2004 10167

A R T I C L E S
Nicolaou et al.
Scheme 8. Attempted Ring-closing Metatheses To Create
complete the target. Anticipating that a Robinson-Gabriel
cyclodehydration reaction could be called upon to accomplish
this task, the opening operations sought to install the functional-
ity needed to create the ketoamide precursor required for that
event (i.e., compound 50). Thus, TFA-promoted condensation
with dimethylaminonitroethylene (45),35 followed by LiAlH4-
mediated reduction in refluxing THF, served to convert 44 into
4-bromotryptamine (47) in 87% overall yield. Subsequent
peptide coupling with TBDPS-protected glycolic acid (48),36
mediated by EDC and HOBt, then led to the smooth preparation
of 49 in 79% yield, a compound separated from the desired
ketoamide substrate (50) only by the absence of a carbonyl
group. This deficiency was rectified over the course of two steps
in 80% yield through the initial installation of a hydroxyl group,
as accomplished by the action of DDQ in aqueous THF at 0
°C, followed by oxidation of the newly installed functionality
with IBX. Interestingly, although DDQ in aqueous media is
known to convert materials such as 49 directly into compounds
such as 50,11,37 the second oxidation step could never be induced
with this particular substrate. As revealed by several additional
studies, the bromine substituent was the likely culprit, as
application of the same conditions to the debrominated variant
of 50 led directly to a ketone.
Macrocyclic Alkene 55a
With these operations accomplished, the synthesis of 53 was
then completed in five additional steps: 1) formation of the
oxazole ring using Wipf’s variant of the Robinson-Gabriel
cyclodehydration reaction;38 2) N-methyl protection of the indole
nucleus under phase transfer conditions (88% yield over two
steps);39 3) TBAF-mediated cleavage of the terminal TBDPS
group (92% yield); 4) oxidation of the resultant hydroxyl group
with IBX (90% yield); and 5) generation of the desired alkene
upon Wittig reaction with Ph3PdCH2 (89% yield).
4. Efforts To Accomplish Macrocyclization. With our
having synthesized both boronate ester 31 and indole-oxazole
53, the stage was now set to attempt biaryl coupling of the two
fragments, and subsequently, the key olefin metathesis-based
macrocyclization. Pleasingly, as shown in Scheme 8, the first
of these critical operations proceeded quite smoothly as Suzuki
coupling was readily effected in 62% yield over the course of
6 h using catalytic amounts of Pd(dppf)Cl2‚CH2Cl2 and excess
K2CO3 in DME at 85 °C.40 Our excitement at reaching this
advanced stage, however, was quickly dashed as attempted ring-
closing olefin metatheses in the presence of catalyst 56,41 57,42
or 5843 (toluene, 70 °C, several days or CH2Cl2, 40 °C, several
hours) failed to result in macrocycle 55. Instead, we only
recovered starting material (54) or observed complete polym-
erization/decomposition. Due to the proximity of the more
accessible alkene to a Lewis basic nitrogen atom within the
a Reagents and conditions: (a) 31 (1.0 equiv), 53 (1.0 equiv), Pd(dppf)Cl2
(0.2 equiv), K2CO3 (3.0 equiv), DME, 85 °C, 6 h, 62%. Cy ) cyclohexyl.
oxazole ring, it is possible that their failure stemmed from
nonproductive chelation following catalyst insertion; steric hin-
drance close to the other olefin could also have been a deciding
factor.44
In any case, with this reaction defining a limitation for the
power of ring-closing olefin metathesis in complex molecule
synthesis, we needed to alter our strategy. So, as discussed
earlier, we turned to the prospect of an intramolecular HWE
reaction for macrocyclization, an extremely powerful and mild
C-C bond forming process which has been widely employed
in chemical synthesis, especially on highly functionalized
substrates.22 Although that precedent was encouraging, the
pursuit of this plan did have us concerned, not because we
thought that it would be difficult to access the needed fragments
but because of the proclivity of a number of our benzofuranone
intermediates to undergo cyclofragmentation to 3-arylbenzofuran
products. More specifically, since the aldehyde derived from
compound 28 (see Scheme 4) aromatized under standard Wittig
reaction conditions, we thought that the same outcome would
occur from an HWE macrocyclization reaction, especially since
it could relieve the strain that would be imparted with the
formation of a 12-membered ring.
(35) Severia, T.; Bohme, H.-J. Chem. Ber. 1968, 101, 2925-2930.
(36) Roth, G. A.; McClymont, E. L. Synth. Commun. 1992, 22, 411-420.
(37) Oikawa, Y.; Toshioka, T.; Mohri, K.; Yonemitsu, O. Heterocycles 1979,
12, 1457-1462.
(38) Wipf, P.; Miller, C. P. J. Org. Chem. 1993, 58, 3604-3606.
(39) For other phase-transfer catalysts which can effect the same operation using
milder bases, see: Ottoni, O.; Cruz, R.; Alves, R. Tetrahedron 1998, 54,
13915-13928.
(40) It is important to note that attempts to accomplish C16-C18 biaryl bond
formation through Stille coupling met with failure despite several attempts
(though the requisite Me3Sn group could be attached to both reactive
partners).
(41) Schrock, R. R.; Murdzek, J. S.; Bazan, G. C.; Robbins, J.; DiMare, M.;
O’Regan, M. J. Am. Chem. Soc. 1990, 112, 3875-3886.
(42) Schwab, P.; France, M. B.; Ziller, J. W.; Grubbs, R. H. Angew. Chem.,
Int. Ed. Engl. 1995, 34, 2039-2041.
(43) Scholl, M.; Ding, S.; Lee, C. W.; Grubbs, R. H. Org. Lett. 1999, 1, 953-
956.
(44) For pertinent discussion on the roles of sterics in metathesis reactions, see:
(a) Grubbs, R. H.; Chang, S. Tetrahedron 1998, 54, 4413-4450. (b) Grubbs,
R. H.; Fu, G. C. Acc. Chem. Res. 1995, 28, 446-452. For recent reviews
on metathesis, see: (c) Trnka, T. M.; Grubbs, R. H. Acc. Chem. Res. 2001,
34, 18-29. (d) Fu¨rstner, A. Angew. Chem., Int. Ed. 2000, 39, 3012-3043.
9
10168 J. AM. CHEM. SOC. VOL. 126, NO. 32, 2004

Studies toward Diazonamide A
A R T I C L E S
Scheme 9. Synthesis of Indole-oxazole 59a
Having reached 60, we were now in a position to access the
key aldehyde-phosphonate HWE precursor (61). This objective
was accomplished in 85% yield over the course of two steps
by first cleaving the TBS ether in 60 under acidic conditions
with aqueous HF and then oxidizing the resultant alcohol with
a standard Swern protocol. The stage was now set for the most
important reaction up to this point in the sequence. Most
gratifyingly, exposure of this new intermediate to the action of
LiHMDS in THF at 0 °C for 2 h under high dilution conditions
(1.0 × 10-5 M) did initiate the desired ring-closing reaction,
leading to macrocycle 55 in 25% yield. Only a small amount
of material (<5%) participated in the previously feared cyclo-
fragmentation reaction, with the remainder accounted for by
phosphonate decomposition prior to cyclization. While the yield
of 55 might seem unduly low for this process, it partially reflects
the fact that at 0 °C compound 61 exists as two nonintercon-
vertable atropisomers along the C16-C18 biaryl axis (as verified
a Reagents and conditions: (a) Ph3P (1.7 equiv), imidazole (2.0 equiv),
I2 (1.5 equiv), CH2Cl2, 0 °C, 10 min, 94%; (b) (MeO)2P(dO)H (2.0 equiv),
KHMDS (0.5 M in toluene, 1.5 equiv), THF, 0 °C, 10 min, 86%. KHMDS
) potassium bis(trimethylsilyl)amide; imid ) imidazole.
Scheme 10. Successful Execution of an HWE Approach To
Accomplish the Synthesis of Heterocyclic Macrocycle 55a
1
by H NMR analysis), so the maximum possible yield for the
process under these conditions was 50%. At 25 °C, the two
atropsiomers of 61 exchange on the NMR time scale to reflect
signals corresponding to one compound, but any attempt to
accomplish macrocyclization at this temperature led to near-
exclusive benzofuran formation instead of alkene 55. No other
base explored at 0 °C, whether NaH, KHMDS, or NaHMDS,
provided a better yield of 55, though all worked with varying
levels of success (5-20%). Finally, in line with expectations,
compound 55 was formed as a single atropisomer, and any
attempt to employ material bearing the opposite C-11 stereo-
chemistry of 61 failed to provide any product from HWE
macrocyclization.12a
a Reagents and conditions: (a) BPD (1.2 equiv), Pd(dppf)Cl2 (0.15 equiv),
KOAc (3.0 equiv), DMSO, 90 °C, 4 h, 50%; (b) 59 (1.0 equiv), Pd(dppf)Cl2
(0.30 equiv), K2CO3 (3.0 equiv), DME, 85 °C, 8 h, 67%; (c) aq HF,
MeCN, 0 °C, 45 min; (d) DMSO (10 equiv), (COCl)2 (5.0 equiv), CH2Cl2,
-78 °C, 15 min, then substrate, CH2Cl2, -78 °C, 30 min, then Et3N
(20 equiv), -78 f 25 °C, 30 min, 85% over two steps; (e) LiHMDS (1.0
M in THF, 5.0 equiv), THF, 0 °C, 2 h, 25%. DME ) ethylene glycol
dimethyl ether. Selected nOe’s shown for 55 confirm the indicated
stereochemistry.
5. Completion of Model Studies. Having finally found a
way to construct the final C-C bond of the heterocyclic-based
macrocycle of diazonamide A, we wanted to conduct one more
model study before moving on to fully functionalized intermedi-
ates: identifying a sequence capable of fashioning the final
A-ring ozazole from the C29-C30 olefinic residue. Our initial
efforts along these lines were directed at forming a diketone
moiety from the alkene in 55 (see Scheme 11), expecting that
the resultant carbonyl group near the B-ring oxazole could be
readily converted into an amine over the course of a few steps.
As such, we attempted to dihydroxylate that olefin under
standard conditions (OsO4, acetone/H2O, quinuclidine, 25 °C).
What occurred was something we should have expected. The
reaction proceeded smoothly in solution, but any attempt to
isolate the diol product led to significant amounts of ring scission
through the same cyclofragmentation pathway encountered on
several occasions up to this point. Thus, with this approach shut
down due to a diol yield of less than 5%, we sought diketone
formation through less conventional means, and after consider-
able experimentation, found that conditions developed by
Sharpless using KMnO4 and Cu(OAc)2‚H2O in neat acetic
anhydride rose to the occasion,45 providing direct access to the
needed diketone in 35% yield. Key to obtaining a consistent
yield of this product was the portionwise addition of KMnO4
over a 45 min period, using the monohydrate salt of Cu(OAc)2,
and a total reaction time no greater than 1.5 h. Otherwise, the
only major product observed was the ring-opened dicarboxylic
acid.
Nevertheless, we felt that we should still try the idea since it
would be easy to test. We began by converting indole-oxazole
intermediate 52 (Scheme 9) into an appropriate phosphonate
by exchanging the oxazylic alcohol for an iodine under typical
conditions (I2, Ph3P, imidazole, CH2Cl2, 0 °C) and then
displacing this new leaving group with the anion derived from
dimethyl phosphite (81% overall). Standard Arbuzov conditions
[P(OMe)3, ∆, 12 h] delivered 59 in only mediocre yield (<20%)
in this final reaction, while the use of NaH or HMDS bases
other than KHMDS to generate the needed anion provided
significant amounts of reduced material in addition to 59. With
this fragment suitably outfitted, the previously obtained ben-
zofuranone intermediate, 28, was then transformed into an aryl
boronate using the same conditions as before (see Scheme 10).
This piece was then directly coupled to phosphonate 59 in 67%
yield utilizing the previously described Suzuki reaction protocol;
to the best of our knowledge, this reaction represents the first
example of a readily enolizable phosphonate being successfully
employed in a Suzuki-type coupling reaction. While certainly
not a discovery of great magnitude, had this transformation
failed, the HWE strategy could not have been explored, as any
attempt to install the same phosphonate post-Suzuki reaction
met with failure.
(45) Sharpless, K. B.; Lauer, R. F.; Repic, O.; Teranishi, A. Y.; Williams, D.
R. J. Am. Chem. Soc. 1971, 93, 3303-3304.
9
J. AM. CHEM. SOC. VOL. 126, NO. 32, 2004 10169

A R T I C L E S
Nicolaou et al.
Scheme 11. Completion of Model Studies Leading to the
Complete Heteroaromatic Skeleton (64) of the Originally Proposed
Structure of Diazonamide A (1)a
Table 2. Model Studies Exploring Chlorination via Electrophilic
Aromatic Substitution; the Role of Indole Protection
entry
compound
product
yield (%)
1
2
3
4
5
6
7
65: R ) Me
67: R ) Bn
69: R ) MOMa
71: R ) TBS
73: R ) Boc
75: R ) Tsb
51: R ) H
66: X ) Cl, Y ) Cl
68: X ) Cl, Y ) Cl
70: X ) Cl, Y ) Cl
72: X ) Cl, Y ) Cl
74: X ) H, Y ) Cl
76: X ) H, Y ) Cl
77: R ) Cl, X ) Cl, Y ) Cl
84
68
72
75
81
79
56
a MOM ) methoxymethyl. b Ts ) p-toluenesulfonyl.
a Reagents and conditions: (a) KMnO4 (6.0 equiv), Ac2O, 0 °C, 2 h,
35%; (b) MeONH2‚HCl (20.0 equiv), EtOH, 25 °C, 12 h, 95%; (c) Pd/C
(10%, 2.0 equiv), H2 (3.0 atm), TFA/MeOH (1:20), 25 °C, 12 h, then AcCl
(3.0 equiv), Et3N (3.0 equiv), CH2Cl2, 25 °C, 30 min, 80%; (d) pTsOH,
benzene, 80 °C, 20 h, 50%. pTsOH ) p-toluenesulfonic acid.
to this point. We also needed to determine how we would
accomplish chlorination of the B- and C-rings. Both of these
questions were answered through the studies summarized in
Table 2. We began by attempting to chlorinate indole-oxazole
51 (entry 8) using NCS as an electrophilic source of chlorine
in THF/CCl4 (1:1) at 70 °C. While successful in furnishing the
desired chlorines, this reaction also led to complete N-chlorina-
tion, a side-reaction that proved difficult to counter or subse-
quently ameliorate through dechlorination. Accordingly, indole
protection was obviously required during the chlorination step,
and as indicated by the exposure of compounds 67, 69, 71, 73,
and 75 to the same conditions, as long as that indole protecting
group was electron-donating or electron-neutral, both the B-
and C-rings could still be smoothly chlorinated (entries 1-4).
On the basis of these results, only three options for indole
protection were available other than engagement by a methyl
group, and in light of the diverse reaction conditions used in
our model studies to reach 64, only two of these (a benzyl group
or a methoxymethyl ether) would likely prove resilient enough
to complete a drive toward 1. Neither, though, was an ideal
candidate, as benzyl protecting groups on indoles are typically
cleaved under substrate-dependent hydrogenation conditions
which often require significant screening to identify (and could
initiate dechlorination), while MOM ethers are deprotected only
through the use of strong Lewis acids.47 However, since a
number of reports48 indicated that benzylic groups on indoles
could be stubbornly resilient to cleavage, we elected to press
forward with MOM protection on our indole intermediates.
6. Application of the Developed Strategy to Fully Func-
tionalized Intermediates. With these extensive model studies
completed, we began our drive to apply the same strategy toward
fully functionalized intermediates, starting with work toward
the EFG fragment using the commercially available L-tyrosine
methyl ester (78) as our new G-ring. As shown in Scheme 12,
its conversion into 79 was accomplished in 90% overall yield
via initial Cbz-protection of the free amine followed by LiBH4
reduction of the ester group and acetonide capture (pTsOH, 2,2-
With oxidation achieved, the more activated and less sterically
hindered carbonyl group was then converted into its corre-
sponding methoxime derivative (62, Scheme 11) in 95% yield
over the course of 12 h using excess MeONH2‚HCl in EtOH,
enabling subsequent hydrogenolysis in acidified methanol to
generate an amine at that site. Acetylation of this new function
using AcCl then furnished acetamide 63 in 80% yield from 62.
Now, only oxazole formation remained to complete this model
study, and fortunately, the needed Robinson-Gabriel dehydra-
tion could be initiated by pTsOH in refluxing benzene, giving
rise to 64 in 50% yield.5b,12b,46 By reaching this point, not only
had we verified all of the opening elements of our synthetic
plan as outlined in Scheme 1 but we also had accomplished the
first synthesis of any material resembling the complete het-
eroaromatic core of diazonamide A (1).
At the same time, we had to concede that the relatively modest
yields observed for HWE closure to generate 55 and subsequent
diketone formation suggested that it would be challenging to
process sufficient material to reach diazonamide A (1) in the
context of a fully elaborated G-ring. Accordingly, we recognized
that the functionalization of the C-11 position in “real” system
compounds would have to be better tailored to prevent benzo-
furan formation. If we could achieve that objective, then it would
be quite reasonable to expect dramatic yield improvements,
especially in the HWE macrocyclization step since we could
likely employ a higher reaction temperature to equilibrate the
C16-C18 biaryl atropisomers and, thereby, drive the potential
yield above 50%.
Apart from this key concern, we also knew that we would
have to address the issue of indole protection in any future
studies with fully functionalized intermediates, a matter of
contention that we had side-stepped by placing the convenient,
but noncleavable, methyl group onto all of our compounds up
(47) Meyers, A. I.; Highsmith, T. K.; Buonara, P. T. J. Org. Chem. 1991, 56,
2960-2964.
(46) Parsons, R. L.; Heathcock, C. H. J. Org. Chem. 1994, 59, 4733-4734. It
is important to note that no other oxazole-forming protocols attempted [such
as Martin’s sulfurane or the previously used conditions from the Wipf group
(ref 38)] succeeded on this substrate. This issue would prove problematic
throughout the course of our diazonamide program.
(48) For some discussion on this topic, see: (a) Kocienski, P. J. Protecting
Groups; Georg Thieme Verlag: Stuttgart, 1994; pp 220-227. (b) Watanabe,
T.; Kobayashi, A.; Nishiura, M.; Takahashi, H.; Usui, T.; Kamiyama, I.;
Mochizuki, N.; Noritake, K.; Yokoyama, Y.; Murakami, Y. Chem. Pharm.
Bull. 1991, 39, 1152-1156.
9
10170 J. AM. CHEM. SOC. VOL. 126, NO. 32, 2004

Studies toward Diazonamide A
A R T I C L E S
Scheme 13. Synthesis of MOM-Protected Indole-oxazole 87a
Scheme 12. Synthesis of Advanced EFG Fragment 86a
a Reagents and conditions: (a) NaH (1.3 equiv), THF, 0 °C, 5 min, then
MOMCl (1.2 equiv), 0 °C, 5 min, 95%; (b) TBAF (1.0 M in THF, 1.5
equiv), THF, 25 °C, 15 min, 92%; (c) Ph3P (1.7 equiv), imidazole (2.0
equiv), I2 (1.5 equiv), CH2Cl2, 0 °C, 10 min, 94%; (d) (MeO)2P(dO)H
(2.0 equiv), NaH (1.5 equiv), THF, 0 °C, 10 min, 86%.
of 80, ready for subsequent metal-halogen exchange with
i-PrMgBr to accomplish the synthesis of 81 through a standard
Grignard reaction with aldehyde 13. From this advanced
intermediate, application of the same general sequence employed
earlier served to advance this compound to lactone 84 in yields
that were commensurate to our model studies. The smooth
nature of these operations requires little commentary beyond
that already presented, except to note that, while 14 distinct
steps were required to reach 84 from 78, most of these operations
proceeded without byproduct formation, such that column
chromatography was typically performed only six times during
the entire sequence. As such, it was quite easy to bring material
to this advanced stage. It is also important to note that we could
not form the epoxide within 82 diastereoselectively irrespective
of epoxidation protocol employed (including attempts with
Jacobsen’s salen catalysts50 and Shi’s L-fructose-derived chiral
dioxirane,51 among a number of others). While clearly a major
concern, we left further study of this problem for later, electing
instead to press forward with our C-10 epimeric material in order
to evaluate the remainder of the strategy.
With 84 in hand, we next reduced its lactone ring fully to
the corresponding diol in 99% yield over the course of 2 h with
LiBH4 in THF at 25 °C. The thought behind this operation being
that if we could engage the new diol in an appropriate protecting
device, such as a cyclic carbonate, then we could solve the issue
of C-11 functionalization and prevent future cyclofragmentations
leading to benzofuran products. However, when that intermedi-
ate was exposed to 1,1′-carbonyldiimidazole in refluxing acetone
in the hope of accomplishing that objective, only the five-
membered ring corresponding to 85 was formed, a result with
some precedent under harsher reaction conditions.52 While
unexpected, this outcome was still acceptable since its func-
tionalization would certainly prevent benzofuran formation.
Accordingly, we completed the synthesis of aryl boronate 86
in 76% yield from 85 using the same conditions that had
succeeded earlier.
a Reagents and conditions: (a) BnOCOCl (1.0 equiv), CHCl3, 25 °C, 3
h, 94%; (b) LiBH4 (2.0 M in THF, 3.0 equiv), 0 f 25 °C, THF, 2 h, 99%;
(c) 2,2-dimethoxypropane (10 equiv), pTsOH (0.1 equiv), acetone, 25 °C,
30 min, 97%; (d) MeI (5.0 equiv), K2CO3 (10 equiv), DMF, 25 °C, 10 h,
92%; (e) I2 (1.1 equiv), Ag(OCOCF3)2 (1.5 equiv), CHCl3, 25 °C, 4 h,
77%; (f) iPrMgBr (1.0 M in THF, 1.5 equiv), THF, -20 °C, 15 min; then
13 (1.0 equiv), THF, -78 °C, 15 min, 69%; (g) IBX (1.2 equiv), DMSO/
THF (1:1), 25 °C, 4 h, 98%; (h) methylene triphenylphosphonium bromide
(2.0 equiv), tBuOK (1.0 M in THF, 1.5 equiv), THF, 0 f 25 °C, 6 h, 86%;
(i) concentrated HCl, MeOH/CH2Cl2 (1:1), 25 °C, 12 h, 98%; (j) ClCH2CN
(1.5 equiv), K2CO3 (1.5 equiv), acetone, 56 °C, 3 h, 95%; (k) 2,2-
dimethoxypropane (10 equiv), pTsOH (0.1 equiv), acetone, 25 °C, 30 min,
96%; (l) mCPBA (77%, 2.0 equiv), NaHCO3 (3.0 equiv), CH2Cl2, 0 °C, 8
h, 84%; (m) tBuOK (1.0 M in THF, 1.5 equiv), THF/DMF (1:1), -78 °C,
5 min, then 10% aq HCl, -78 °C, 5 min, 52%; (n) TBSCl (3.0 equiv),
imidazole (6.0 equiv), DMF, 25 °C, 6 h, 88%; (o) LiHMDS (1.0 M in
THF, 2.0 equiv), THF, -78 °C, 15 min; O2, 15 min, -78 °C; 1.0 M SnCl2
in 10% aq HCl, -78 f 0 °C, 30 min, 73%; (p) LiBH4 (2.0 M in THF, 2.0
equiv), Et2O, 0 °C, 3 h, 88%; (q) 1,1′-carbonyldiimidazole (1.3 equiv), THF,
67 °C, 1 h, 90%; (r) bis(pinacolato)diboron (1.2 equiv), Pd(dppf)Cl2 (0.2
equiv), KOAc (3.0 equiv), DMSO, 85 °C, 6 h, 76%. 2,2-DMP )
2,2-dimethoxypropane; 1,1′-CDI ) 1,1′-carbonyldiimidazole.
With a suitably functionalized EFG segment in hand, its
indole-oxazole coupling partner was then prepared exactly as
before, as shown in Scheme 13, with the only alteration being
MOM protection of the indole nucleus. With these four steps
proceeding in a combined yield of 71% from 51, probes at the
rest of the strategy could now begin. As indicated in Scheme
dimethoxypropane, acetone, 25 °C). Subsequent methylation
(MeI, K2CO3, DMF, 25 °C) and iodination49 using molecular
iodine and Ag(OCOCF3)2 then served to complete the assembly
(50) Jacobsen, E. N.; Zhang, W.; Muci, A. R.; Ecker, J. R.; Deng, L. J. Am.
Chem. Soc. 1991, 113, 7063-7064.
(51) (a) Tu, Y.; Wang, Z.-X.; Shi, Y. J. Am. Chem. Soc. 1996, 118, 9806-
9807. (b) Wang, Z.-X.; Tu, Y.; Frohn, M.; Zhang, J.-R.; Shi, Y. J. Am.
Chem. Soc. 1997, 119, 11224-11235.
(49) For an earlier example of this procedure, see: Carruthers, W.; Coggins,
P.; Weston, J. B. J. Chem. Soc., Perkin Trans. 1 1991, 611-616.
(52) Stafford, J. A.; Valvano, N. L. J. Org. Chem. 1994, 59, 4346-4349.
9
J. AM. CHEM. SOC. VOL. 126, NO. 32, 2004 10171

A R T I C L E S
Nicolaou et al.
Scheme 14. Suzuki Coupling and HWE Condensation To
Generate One of the 12-Membered Macrocycles of the Originally
Proposed Structure of Diazonamide A (1)a
Table 3. Exploration of Conditions To Accomplish the HWE
Macrocyclization Leading to 90
entry
conditions
yield (%)
0-25a
1
2
3
4
5
6
7
8
9
LiHMDS, THF, 0 °C
DBU, CH3CN, 25 °C
decomposition
35
DBU, LiCl, CH3CN, 25 °C
TMG, LiCl, CH3CN, 25 °C
TMG, LiCl, DMSO, 25 °C
TMG, LiCl, hexane, 25 °C
TMG, LiCl, toluene, 25 °C
TMG, LiCl, CH2Cl2, 25 °C
TMG, LiCl, DMF, 25 °C
TMG, LiCl, DMF, 70 °C
decomposition
55
no reaction
decomposition
no reaction
23
10
55-60a
a Range indicates maximum and minimum values obtained for several
runs.
elevated the temperature to 70 °C, the formation of 90 could
be effected in 55-60% yield. Similar yields were also realized
at ambient temperature using DMSO as solvent (entry 5). As
such, simply by altering C-11 functionalization, we had
developed a sequence that could reliably deliver macrocycle
90 in yields sufficiently acceptable to provide enough material
supplies to study the completion of the heterocyclic core of
diazonamide A (1).
Unfortunately, despite our ability to prepare over a hundred
milligrams of 90 with ease, the olefin within 90 proved resilient
to oxidation of any kind. Use of any dihydroxylation protocol
with all published amine accelerants and, in some instances,
stoichiometric OsO4, only led to recovered starting material.55
Efforts to enlist KMnO4 to form a diketone similarly provided
recovered starting material along with small amounts of open
dicarboxylic acid products. Hydroboration with a variety of
hydride sources also failed to convert 90 into anything useful
to explore the sequence further. Thus, for some reason that is
not distinctly clear at present, the mere addition of a few atoms
to ring G provided enough steric bulk to completely shut down
reactive pathways that proceeded in all of our slightly simpler
model systems. We now had to identify a new way to create
the heterocyclic core of 1.
a Reagents and conditions: (a) 86 (1.05 equiv), 87 (1.0 equiv),
Pd(dppf)Cl2 (0.2 equiv), K2CO3 (5.0 equiv), DME, 85 °C, 12 h, 83%; (b)
aq HF (48%, 3.0 equiv), MeCN, 0 °C, 45 min; (c) 2,2-DMP, acetone, 25
°C, 5 min; (d) Dess-Martin periodinane (3.0 equiv), NaHCO3 (10 equiv),
CH2Cl2, 25 °C, 1 h, 70% over three steps; (e) TMG (5.0 equiv), LiCl (5.0
equiv), DMF, 70 °C, 12 h, 55-60%. TMG ) 1,1,3,3-tetramethylguanidine.
Conclusion
14, initial Suzuki coupling proceeded in an excellent yield of
83% over the course of 12 h in refluxing DME as catalyzed by
Pd(dppf)Cl2‚CH2Cl2 in the presence of K2CO3. Aldehyde
formation (89) was then accomplished over three steps in 70%
overall yield, enabling probes of HWE macrocyclization to
commence. At first, these studies were disappointing, as the
use of LiHMDS in THF provided only a modest and highly
variable yield of macrocycle 90 irrespective of temperature (see
Table 3). However, better yields (35%) were obtained if we
switched to Masamune-Roush conditions (LiCl, DBU, MeCN,
25 °C),53 and when we changed the base to the bulkier 1,1,3,3-
tetramethyguanidine (TMG),54 used DMF as solvent, and
As often occurs during efforts directed toward the synthesis
of complex molecules, the presence of a singular functional
group or motif can instigate a number of unanticipated chal-
lenges that can complicate, and sometimes even thwart, a well-
conceived synthetic plan. This article has detailed a number of
such findings, some of which served a useful purpose in leading
to new synthetic methodology (including the facile synthesis
(54) Our use of this base was inspired by the elegant studies of Allen, J. R.;
Harris, C. R.; Danishefsky, S. J. J. Am. Chem. Soc. 2001, 123, 1890-
1897.
(55) For a review summarizing many of the procedures we attempted, see: (a)
Kolb, H. C.; VanNieuwenhze, M.; Sharpless, K. B. Chem. ReV. 1994, 94,
2483-2547. Other conditions probed included those reported in: (b) He,
F.; Bo, Y.; Altom, J. D.; Corey, E. J. J. Am. Chem. Soc. 1999, 121, 6771-
6772. (c) Corey, E. J.; Sarshar, S.; Azimioara, M. D.; Newbold, R. C.;
Noe, M. C. J. Am. Chem. Soc. 1996, 118, 7851-7852.
(53) Blanchette, M. A.; Choy, W.; Davis, J. T.; Essenfeld, A. P.; Masamune,
S.; Roush, W. R.; Sakai, T. Tetrahedron Lett. 1984, 21, 2183-2186.
9
10172 J. AM. CHEM. SOC. VOL. 126, NO. 32, 2004

Studies toward Diazonamide A
A R T I C L E S
of 3-arylbenzofurans, a way to make benzofuranones using
5-exo-tet cyclizations, and a means to deoxygenate sulfoxides),
while others could not be circumvented. Nevertheless, as the
following articles will detail,19 the chemistry described in this
first-generation approach toward the original structure of di-
azonamide A (1) laid a solid foundation which would inform a
new strategy that ultimately enabled the construction of both
the macrocyclic rings within 1. Equally important, this platform
would help to shape both of our later drives to accomplish the
total synthesis of the revised structure of diazonamide A (2).
for this work was provided by The Skaggs Institute for Chemical
Biology, the National Institutes of Health (USA), predoctoral
fellowships from the National Science Foundation, Pfizer, and
Bristol-Myers Squibb (all to S.A.S.), postdoctoral fellowships
from the Alfred Benzons Foundation (K.B.S) and The Skaggs
Institute for Chemical Biology (X.H.), and a grant from
American Biosciences.
Supporting Information Available: Experimental procedures
and compound characterization (PDF, CIF). This material is
available free of charge via the Internet at http://pubs.acs.org.
Acknowledgment. We thank Drs. D. H. Huang, G. Siuzdak,
and R. Chadha for NMR spectroscopic, mass spectrometric, and
X-ray crystallographic assistance, respectively. Financial support
JA040090Y
9
J. AM. CHEM. SOC. VOL. 126, NO. 32, 2004 10173