Total synthesis and structural elucidation of azaspiracid-1. Construction of key building blocks for originally proposed structure

Article abstract of DOI:10.1021/ja0547477

Syntheses of the three key building blocks (65, 98, and 100) required for the total synthesis of the proposed structure of azaspiracid-1 (1a) are described. Key steps include a TMSOTf-induced ring-closing cascade to form the ABC rings of tetracycle 65, a neodymium-catalyzed internal aminal formation for the construction of intermediate 98, and a Nozaki-Hiyama-Kishi coupling to assemble the required carbon chain of fragment 100. The synthesized fragments, obtained stereoselectively in both their enantiomeric forms, were expected to allow for the construction of all four stereoisomers proposed as possible structures of azaspiracid-1 (1a-d), thus allowing the determination of both the relative and absolute stereochemistry of the natural product.

Full text of DOI:10.1021/ja0547477

Published on Web 01/28/2006
Total Synthesis and Structural Elucidation of Azaspiracid-1.
Construction of Key Building Blocks for Originally Proposed
Structure
K. C. Nicolaou,* Petri M. Pihko, Federico Bernal, Michael O. Frederick,
Wenyuan Qian, Noriaki Uesaka, Nicole Diedrichs, Ju¨rgen Hinrichs,
Theocharis V. Koftis, Eriketi Loizidou, Goran Petrovic, Manuela Rodriquez,
David Sarlah, and Ning Zou
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 July 15, 2005; E-mail: kcn@scripps.edu
Abstract: Syntheses of the three key building blocks (65, 98, and 100) required for the total synthesis of
the proposed structure of azaspiracid-1 (1a) are described. Key steps include a TMSOTf-induced ring-
closing cascade to form the ABC rings of tetracycle 65, a neodymium-catalyzed internal aminal formation
for the construction of intermediate 98, and a Nozaki-Hiyama-Kishi coupling to assemble the required
carbon chain of fragment 100. The synthesized fragments, obtained stereoselectively in both their
enantiomeric forms, were expected to allow for the construction of all four stereoisomers proposed as
possible structures of azaspiracid-1 (1a-d), thus allowing the determination of both the relative and absolute
stereochemistry of the natural product.
Introduction
azaspiracid-1; nonetheless, they permitted extensive mass
spectrometric and NMR (1H, COSY, TOCSY, ROESY, HSQC,
Among recently isolated marine biotoxins, none has created
as much concern with regard to seafood poisoning and human
health, or research interest with regard to its structure and
chemical synthesis, as azaspiracid-1. Its story began in The
Netherlands when, in 1995, an incident of human illness with
diarrhetic shellfish poisoning (DSP)-like symptoms was reported
and eventually traced to the consumption of mussels originating
from Killary Harbour, Ireland.1 A subsequent investigation of
these mussels, belonging to the Mytilus edulis family, by
Yasumoto, Satake, and co-workers led to the isolation and
structural assignment of azaspiracid-1 as one of the four possible
structures (1a-d) shown in Figure 1.2 Its discovery and
recognized health hazard led to the declaration of a new toxic
syndrome, named azaspiracid poisoning (AZP), with a molecular
structure and pathological effects sufficiently different from
previously known DSP phenomena to warrant the new name.
Known symptoms in humans include nausea, vomiting and
severe diarrhea. Furthermore, azaspiracid-1 was shown to cause
lung, liver, spleen, and lymphocyte damage as well as lung
tumor formation in mice.3
and HMBC) spectroscopic studies, leading to the proposal of
structure 1a or one of its stereoisomeric forms [epi-FGHI-1a
(1b, the FGHI diastereomer of 1a), ent-1a (1c, the enantiomer
of 1a), and ent-(epi-FGHI-1a) (1d, the enantiomer of 1b)].
Containing no stereocenters, the two-carbon bridge separating
the ABCDE and FGHI domains of the molecule (C26-C27) was
apparently a sufficient barrier to prevent stereochemical cor-
relation across the entire backbone of the structure. Nevertheless,
these structural assignments revealed a unique and unprec-
edented molecular architecture that included in its C47H71NO12
formula no less than 9 rings and 20 stereogenic centers. Among
its most prominent structural motifs are a trioxadispiroacetal
system fused onto a tetrahydrofuran ring, the entire framework
comprising the molecule’s ABCD domain, and an azaspiro ring
system fused onto a 2,9-dioxabicyclo[3.3.1]nonane system,
completing the structure of the tetracyclic FGHI framework.
Given the secondary amino group of the azaspiro system and
the C1 carboxyl moiety that it also carries, azaspiracid-1 is an
amino acid, a feature not to be overlooked, for it may play a
role in the overall conformation and properties of the molecule.
The assigned structure also boasted three sites of unsaturation
in the form of carbon-carbon double bonds (C4-C5, C8-C9,
and C26-C44), a hemiketal moiety (at C-21), and several
stereocenters prone to epimerization (C-10, -13, -14, -21, -36,
and -37) under acidic conditions. All in all, these structural
features presented an appealing, but formidable, synthetic
The heroic isolation and structural elucidation investigations
by the Yasumoto-Satake team yielded only small amounts of
(1) McMahon, T.; Silke, J. Harmful Algae News 1996, 14, 2.
(2) Satake, M.; Ofuji, K.; Naoki, H.; James, K. J.; Furey, A.; McMahon, T.;
Silke, J.; Yasumoto, T. J. Am. Chem. Soc. 1998, 120, 9967.
(3) Ito, E.; Satake, M.; Ofuji, K.; Higashi, M.; Harigaya, K.; McMahon, T.;
Yasumoto, T. Toxicon 2002, 40, 193.
9
2244
J. AM. CHEM. SOC. 2006, 128, 2244-2257
10.1021/ja0547477 CCC: $33.50 © 2006 American Chemical Society

Synthesis and Structure of Azaspiracid-1
A R T I C L E S
Figure 1. The four structures originally proposed for azaspiracid-1 (1a-d, Satake et al. 1998).
challenge which became even more pressing due to the extreme
scarcity and hazardous nature of the molecule.
By virtue of its structural complexity, natural scarcity, and
biological importance, azaspiracid-1 prompted many groups to
pursue its total synthesis,4 including ours.5 It would be our team
that reached the target in 2003,6 only to prove, however, that
the assigned structures (1a-d) were wrong. This finding spurred
a collaborative effort with the Satake group to determine the
true structure of the natural product through degradative work
Figure 2. Revised structure of azaspiracid-1 (1).
and chemical synthesis, which would culminate, in 2004, in the
demystification of the structure of azaspiracid-1 and its first
total synthesis.7 In this and accompanying articles8,9 in this
series, we present the details of the total synthesis campaign
that led to the revision of the structure of azaspiracid-1 and the
establishment of its absolute stereochemistry as 1 (Figure 2).
It should be noted that the developed synthetic technology for
the total synthesis of azaspiracid-1 could, in principle, be suit-
ably adopted to deliver its siblings, azaspiracids-2 through -11
(2a-11a, originally assigned structure; 2-11, revised structures)
shown in Figure 3.10 We begin, in this paper, with the retro-
synthetic analysis and construction of the defined key building
blocks required for the total synthesis of the originally proposed
structures of azaspiracid-1 (1a-d).
Results and Discussion
(4) For other studies toward the total synthesis of azaspiracid-1, see: (a) Carter,
R. G.; Weldon, D. J. Org. Lett. 2000, 2, 3913. (b) Carter, R. G.; Graves,
D. E. Tetrahedron Lett. 2001, 42, 6035. (c) Carter, R. G.; Bourland, T. C.;
Graves, D. E. Org. Lett. 2002, 4, 2177. (d) Carter, R. G.; Graves, D. E.;
Gronemeyer, M. A.; Tschumper, G. S. Org. Lett. 2002, 4, 2181. (e) Zhou,
X.-T.; Carter, R. G. Chem. Commun. 2005, 19, 2138. (f) Forsyth, C. J.;
Hao, J.; Aiguade, J. Angew. Chem., Int. Ed. 2001, 40, 3663. (g) Dounay,
A. B.; Forsyth, C. J. Org. Lett. 2001, 3, 975. (h) Hao, J.; Aiguade, J.;
Forsyth, C. J. Tetrahedron Lett. 2001, 42, 817. (i) Hao, J.; Aiguade, J.;
Forsyth, C. J. Tetrahedron Lett. 2001, 42, 821. (j) Geisler, L. K.; Nguyen,
S.; Forsyth, C. J. Org. Lett. 2004, 6, 4159. (k) Sasaki, M.; Iwamuro, Y.;
Nemoto, J.; Oikawa, M. Tetrahedron Lett. 2003, 44, 6199. (l) Ishikawa,
Y.; Nishiyama, S. Tetrahedron Lett. 2004, 45, 351. (m) Ishikawa, Y.;
Nishiyama, S. Heterocycles 2004, 63, 539. (n) Ishikawa, Y.; Nishiyama,
S. Heterocycles 2004, 63, 885.
(5) (a) Nicolaou, K. C.; Pihko, P. M.; Diedrichs, N.; Zou, N.; Bernal, F. Angew.
Chem., Int. Ed. 2001, 40, 1262. (b) Nicolaou, K. C.; Qian, W.; Bernal, F.;
Uesaka, N.; Pihko, P. M.; Hinrichs, J. Angew. Chem., Int. Ed. 2001, 40,
4068.
(6) (a) Nicolaou, K. C.; Li, Y.; Uesaka, N.; Koftis, T. V.; Vyskocil, S.; Ling,
T.; Govindasamy, M.; Qian, W.; Bernal, F.; Chen, D. Y.-K. Angew. Chem.,
Int. Ed. 2003, 42, 3643. (b) Nicolaou, K. C.; Chen, D. Y.-K.; Li, Y.; Qian,
W.; Ling, T.; Vyskocil, S.; Koftis, T. V.; Govindasamy, M.; Uesaka, N.
Angew. Chem., Int. Ed. 2003, 42, 3649.
1. Retrosynthetic Analysis. A brief inspection of the structure
of azaspiracid-1 (e.g., 1a) reveals a double spiroacetal (ABC
ring junction), a hemiketal (E ring), an intramolecular bridged
ketal (FG ring system), and a spiroaminal (HI ring junction) as
interesting structural motifs that may need special attention from
the synthetic point of view due to their stereochemical features
and fragile nature. In addition, the astute observer may recognize
a number of uniquely strategic bonds for retrosynthetic discon-
nection, as outlined in Figure 4. For optimum convergency, our
first retrosynthetic analysis involved disconnections at the
C20-C21 (dithiane technology)11 and the C25-C26 (Nozaki-
Hiyama-Kishi coupling)12 bonds leading, upon suitable func-
tional group manipulation, to aldehyde 13 (or an equivalent
C1-C20 fragment), dithiane aldehyde 12 (or an equivalent
(7) (a) Nicolaou, K. C.; Vyskocil, S.; Koftis, T. V.; Yamada, Y. M. A.; Ling,
T.; Chen, D. Y.-K.; Tang, W.; Petrovic, G.; Frederick, M. O.; Li, Y.; Satake,
M. Angew. Chem., Int. Ed. 2004, 43, 4312. (b) Nicolaou, K. C.; Koftis, T.
V.; Vyskocil, S.; Petrovic, G.; Ling, T.; Yamada, Y. M. A.; Tang, W.;
Frederick, M. O. Angew. Chem., Int. Ed. 2004, 43, 4318.
(8) Nicolaou, K. C.; Chen, D. Y.-K.; Li, Y.; Uesaka, N.; Petrovic, G.; Koftis,
T. V.; Bernal, F.; Frederick, M. O.; Govindasamy, M.; Ling, T.; Pihko, P.
M.; Tang, W.; Vyskocil, S. J. Am. Chem. Soc. 2006, 128, 2258-2267.
(9) Nicolaou, K. C.; Koftis, T. V.; Vyskocil, S.; Petrovic, G.; Tang, W.;
Frederick, M. O.; Chen, D. Y.-K.; Li, Y.; Ling, T.; Yamada, Y. M. A. J.
Am. Chem. Soc. 2006, 128, in press.
(10) (a) Ofuji, K.; Satake, M.; McMahon, T.; Silke, J.; James, K. J.; Naoki, H.;
Oshima, Y.; Yasumoto, T. Nat. Toxins 1999, 7, 99. (b) Ofuji, K.; Satake,
M.; McMahon, T.; James, K. J.; Naoki, H.; Oshima, Y.; Yasumoto, T.
Biosci. Biotechnol. Biochem. 2001, 65, 740. (c) James, K. J.; Diaz Sierra,
M.; Lehane, M.; Brana A.; Furey, A. Toxicon 2003, 41, 277.
(11) Corey, E. J.; Seebach, D. Angew. Chem., Int. Ed. Engl. 1965, 4, 1075.
(12) For recent reviews of Cr(II) mediated reactions, see: (a) Furstner, A. Chem.
ReV. 1999, 99, 991. (b) Wessjohann, L. A.; Scheid, G. Synthesis 1999, 1.
(c) Saccomano, N. A. In ComprehensiVe Organic Synthesis; Trost, B. M.,
Fleming, I., Eds.; Pergammon: Oxford, 1991; Vol. 1, p 173.
9
J. AM. CHEM. SOC. VOL. 128, NO. 7, 2006 2245

A R T I C L E S
Nicolaou et al.
Figure 3. Originally proposed structures for azaspiracids-2 to -11 (2a-11a, left) and revised structures (2-11, right).
Figure 4. First-generation retrosynthetic analysis of the originally proposed structure of azaspiracid-1 (1a).
C21-C25 fragment), and vinyl triflate 14 (or an equivalent
C26-C40 fragment) as potential key building blocks for the
intended construction.
17 as reasonably accessible starting materials for the synthesis
of the ABCD domain of the molecule. Similarly, the tetracyclic
FGHI fragment 14 was traced back to open chain trihydroxy
amino-diketone 21 as a potential progenitor, as shown in Figure
4. The latter intermediate was then disconnected to the potential
starting material 22 (acetal formation) and then further discon-
nected to azido-ketone 23 and aldehyde 24 (aldol reaction). The
strategy that emerged from this retrosynthetic analysis provided
us, in addition to high convergency, the flexibility to construct
both enantiomers of each key building block (12-14) so as to
ensure the definition of the relative stereochemistry of the
Side-chain detachment followed by dismantling of the double
spiroacetal within 13 then led to appropriately protected
monocyclic compounds 15 and 16 as plausible precursors of
the desired polycycle. Equipped with a sulfoxide or a dithiane
group, the latter compounds could then be disconnected through
two carbon-carbon bond-forming reactions, a nucleophilic
addition, and a Grignard addition, yielding fragments 19 (aryl
sulfoxide) and 20 (dithiane), Grignard reagent 18, and aldehyde
9
2246 J. AM. CHEM. SOC. VOL. 128, NO. 7, 2006

Synthesis and Structure of Azaspiracid-1
A R T I C L E S
Scheme 1. Construction of Key Intermediate 17a
Figure 5. Anomeric effects providing stabilization for the two isomeric
ABCD domains 25 (two anomeric effects, thermodynamically more stable)
and 26 (one anomeric effect, thermodynamically less stable).
ABCDE and FGHI domains of azaspiracid-1, as well as its
absolute stereochemistry.
2. Construction of the Undesired ABCD Ring System
through a Chiral Sulfoxide. Upon inspection of the ABCD
ring framework and some early modeling studies, it became
clear to us that the reported stereochemical arrangement of the
double spiroacetal moiety (ABC ring junction) represented the
thermodynamically less favored structure (26, 13R) as compared
to its C-13-epimer (25, 13S). Figure 5 depicts these epimeric
compounds with the crucial anomeric effects that provided the
reason for the expected higher thermodynamic stability for the
undesired (25, 13S) structure. Indeed, timely disclosures by the
Carter and Forsyth groups4b,g confirmed this postulate, as these
researchers had synthesized suitable precursors of the ABCD
framework of azaspiracid-1 and demonstrated its preference to
fall, under thermodynamically controlled conditions, exclusively
into the undesired structural motif (25, 13S). It was, therefore,
abundantly evident at the time that a special strategy was
necessary for the casting of the unfavored (26, 13R) ABCD
stereoisomer.
Our first attempt to address this rather thorny problem
involved the use of a chiral sulfoxide moiety, as had previously
been described by Williams.13 Thus, we reasoned that a bulky
group adjacent to the C-13 stereocenter may provide enough
steric bias to override the influence of the anomeric effect,
thereby reversing the outcome of the ring-closing reaction. A
group that would facilitate the construction of the precursor
substrate as well as the required olefinic bond after the
cyclization would be preferred. The p-tolylsulfoxide group
fulfilled these criteria and, therefore, was chosen for incorpora-
tion into the intermediates of the sequence. Schemes 1-3 depict
the results of our investigations employing the chiral sulfoxide
strategy to control the C-13 spiroacetal cyclization.
a Reagents and conditions: (a) (MeO)2P(O)Me (2.2 equiv), n-BuLi (1.6
M in hexanes, 2.2 equiv), THF, -78 °C, 1 h, 84%; (b) 29 (0.67 equiv),
LiCl (1.3 equiv), i-Pr2NEt (1.0 equiv), MeCN, 25 °C, 12 h, 86% based on
29; (c) LiAlH4 (10 equiv), LiI (8.0 equiv), Et2O, -100 °C, 30 min, 98%;
(d) AcOH:H2O (2:1), 40 °C, 5 h, 97%; (e) NIS (5.0 equiv), NaHCO3 (10
equiv), THF, 0 °C, 36 h, 70%; (f) TBDPSCl (1.4 equiv), Et3N (3.0 equiv),
4-DMAP (0.1 equiv), CH2Cl2, 0 °C, 3 h, 90%; (g) TBSOTf (1.6 equiv),
2,6-lutidine (4.0 equiv), CH2Cl2, 0 °C, 30 min, 100%; (h) H2, Raney-Ni
(1:1 w/w), EtOH, 25 °C, 1 h, 99%; (i) H2, 20% Pd(OH)2/C (25% w/w),
EtOH, 25 °C, 3 h, 88%; (j) DMP (2.0 equiv), CH2Cl2, 25 °C, 2 h, 99%.
Abbreviations: THF, tetrahydrofuran; NIS, N-iodosuccinimide; TBDPS,
tert-butyldiphenylsilyl; 4-DMAP, 4-(dimethylamino)pyridine; TBS, tert-
butyldimethylsilyl; Tf, trifluoromethanesulfonate; DMP, Dess-Martin pe-
riodinane.
from dimethylmethylphosphonate and n-BuLi resulted in the
formation of ketophosphonate 28 in 84% yield (Scheme 1).
Reaction of an excess of this ketophosphonate with enantio-
merically pure aldehyde 2915 in the presence of LiCl and
i-Pr2NEt then furnished the R,â-unsaturated ketone 30 in 86%
yield.16 Chelation-controlled reduction of the latter compound
with LiAlH4, orchestrated by LiI in ether at -100 °C,17
pleasingly gave the desired secondary alcohol stereoisomer 31
in 98% yield and g98:2 diastereoselectivity. The acetonide
group was then removed by exposure to AcOH from the latter
compound, furnishing triol 32 in 97% yield. This substrate was
then subjected to iodoetherification with NIS (for abbreviations
of reagents and protecting groups, see the scheme footnotes) in
(14) Saito, S.; Ishikawa, T.; Kuroda, A.; Koga, K.; Moriwake, T. Tetrahedron
1992, 48, 4067.
(15) Schoning, K.-U.; Hayashi, R. K.; Powell, D. R.; Kirschning, A. Tetrahe-
dron: Asymmetry 1999, 10, 817.
(16) Blanchette, M. A.; Choy, W.; Davis, J. T.; Essenfeld, A. P.; Masamune,
S.; Roush, W. R.; Sakai, T. Tetrahedron Lett. 1984, 25, 2183.
(17) Mori, Y.; Kuhara, M.; Takeuchi, A.; Suzuki, M. Tetrahedron Lett. 1988,
29, 5419.
The sequence began with enantiomerically pure acetonide
methyl ester 27,14 whose reaction with the lithio anion derived
(13) Williams, D. R.; Barner, B. A.; Nishitani, K.; Phillips, J. G. J. Am. Chem.
Soc. 1982, 104, 4708.
9
J. AM. CHEM. SOC. VOL. 128, NO. 7, 2006 2247

A R T I C L E S
Nicolaou et al.
Table 1. Conditions for the Cyclization of Sulfoxide 15 to
Scheme 2. Synthesis of Chiral Sulfoxide Cyclization Precursor
15a
Tetracycle 42
entry
conditionsa
yield (%)
1
2
3
4
5
6
TFA (3.0 equiv), THF:H2O (4:1), 0 f 25 °C, 48 h
CSA (1.0 equiv), PhH:MeOH:H2O (10:2:1), 25 °C 24 h
p-TsOH‚H2O (0.1 equiv), MeOH, 25 °C, 36 h
BCl3 (2.5 equiv), CH2Cl2, 25 °C, 3 min
AcOH:H2O (4:1), 25 °C, 18 h
TMSOTf (3.0 equiv), CH2Cl2, -78 f 0 °C, 3 h
25
nr
<10
0
<10
62
a Reactions were carried out on 0.1 mmol scale. Abbreviations: TFA,
trifluoroacetic acid; CSA, camphorsulfonic acid; TMS, trimethylsilyl;
p-TsOH, p-toluenesulfonic acid; nr, no reaction.
Scheme 3. Synthesis of Undesired C6-C20 Isomer 44 of the
ABCD Domaina
a Reagents and conditions: (a) LDA (1.0 equiv), (R)-methyl-p-tolylsul-
foxide (1.0 equiv), THF, -78 °C, 30 min, 83%; (b) 3-butenylmagnesium
bromide 18 (6.0 equiv), THF, -78 f -10 °C, 3 h, 87% (ca. 1:1 mixture
of diastereomers); (c) OsO4 (0.03 equiv), NMO (2.0 equiv), t-BuOH:THF:
H2O (10:2:1), 25 °C, 12 h; then NaIO4 (5.0 equiv), pH 7 buffer, 25 °C, 5
h, 96%; (d) NIS (5.0 equiv), n-Bu4I (2.0 equiv), CH2Cl2, 25 °C, 1 h, 98%;
(e) LDA (1.0 equiv), 19 (1.5 equiv), THF, -78 °C, 15 min, 83%; (f) DMP
(3.0 equiv), CH2Cl2, 25 °C, 2 h, 92%. Abbreviations: LDA, lithium
diisopropylamide; NMO, N-methylmorpholine-N-oxide; Ar, p-tolyl.
THF in the presence of NaHCO3 at 0 °C to afford the
tetrahydrofuran system 33 in 70% yield as a single isomer.18
The preference of this process for a single cyclic ether isomer
is remarkable in that the reaction not only was ring-selective,
but it was also highly stereoselective (g98% de). The following
step was also highly selective, leading from the resulting
dihydroxy iodoether to the mono-TBDPS derivative 34 upon
treatment with a slight excess of TBDPSCl and Et3N in the
presence of 4-DMAP (90% yield). Further silylation of this
compound (34) under TBSOTf-2,6-lutidine conditions afforded
bis-silyl ether 35 in quantitative yield. Reductive removal of
the iodine residue from 35 (H2, Raney-Ni) then gave the desired
tetrahydrofuran system 36 in 99% yield. This compound (36)
was then converted to aldehyde 17 by first hydrogenolyzing its
benzyl ether off [H2, Pd(OH)2] to afford the primary alcohol
(37, 88% yield) and then oxidizing the latter compound with
DMP (99%).19
a Reagents and conditions: (a) TMSOTf (3.0 equiv), CH2Cl2, -78 °C,
1 h; then 0 °C, 10 min, 62%; (b) BzCl (4.0 equiv), py (100 equiv), CH2Cl2,
25 °C, 12 h, 96%; (c) P(OMe)3 (6.0 equiv), toluene, reflux, 12 h, 45%.
Abbreviations: Bz, benzoyl; py, pyridine.
1,2-diol, produced the corresponding lactol, which was then
oxidized to the γ-lactone (40) by treatment with NIS and
n-Bu4NI, in 96% overall yield. Sulfoxide 19 (derived in one
step from (R)-methyl-p-tolylsulfoxide and iodide 38, Scheme
2)20 was converted to its anion with LDA and then reacted with
lactone 40 to afford the coupled product, hydroxy ketone sul-
foxide 41, as a mixture of four diastereomers, in 83% combined
yield. Dess-Martin oxidation (92% yield) converted this
mixture to the diketone sulfoxide 15, setting the stage for the
pending deprotection-cyclization cascade.
Scheme 2 shows the sequence through which aldehyde 17
was advanced to the next stage, namely cyclization precursor
15. Thus, the Grignard reagent 18 generated from 4-bromo-1-
butene was reacted with 17 in THF to afford olefinic secondary
alcohol 39 in 87% yield as a mixture of diastereomers (ca. 1:1).
Dihydroxylation of this compound with catalytic OsO4 in the
presence of NMO, followed by NaIO4 cleavage of the resulting
Indeed, upon brief experimentation (see Table 1), it was found
that exposure of precursor 15 to TMSOTf 21 in CH2Cl2 at -78
(18) (a) Adinolfi, M.; Parrilli, M.; Barone, G.; Laonigro, G.; Mangoni, L.
Tetrahedron Lett. 1976, 16, 3661. (b) Haaima, G.; Weavers, R. T.
Tetrahedron Lett. 1988, 29, 1085.
(20) Allingham, M. T.; Howard-Jones, A.; Murphy, P. J.; Thomas, D. A.;
Caulkett, P. W. R. Tetrahedron Lett. 2003, 44, 8677.
(21) Bou, V.; Vilarrasa, J. Tetrahedron Lett. 1990, 31, 567.
(19) Dess, D. B.; Martin, J. C. J. Org. Soc. 1983, 48, 4155.
9
2248 J. AM. CHEM. SOC. VOL. 128, NO. 7, 2006

Synthesis and Structure of Azaspiracid-1
A R T I C L E S
Scheme 4. Synthesis of the Originally Proposed ABCD Ring System (51) of Azaspiracid-1a
a Reagents and conditions: (a) DMP (2.0 equiv), CH2Cl2, 25 °C, 3 h, 95%; (b) HO(CH2)2OH (7.0 equiv), triethylorthoformate (3.0 equiv), p-TsOH (0.1
equiv), 55 °C, 2 h, 98%; (c) OsO4 (0.03 equiv), NMO (2.0 equiv), t-BuOH:THF:H2O (10:2:1), 25 °C, 14 h; NaIO4 (5.0 equiv), pH 7 buffer, 25 °C, 5 h,
100%; (d) n-BuLi (1.6 M in hexanes, 2.6 equiv), 20 (2.6 equiv), THF, -20 °C, 3 h, 87%; (e) DMP (2.0 equiv), CH2Cl2, 25 °C, 1 h, 88%; (f) TMSOTf (3.0
equiv), CH2Cl2, -78 f -30 °C, 1 h, 85%; (g) PivCl (3.0 equiv), py (10 equiv), 4-DMAP (0.1 equiv), 25 °C, 3 h, 95%; (h) NBS (8.0 equiv), 2,6-lutidine
(16 equiv), MeCN, 25 °C, 2 h, 91%. Abbreviations: Piv, trimethylacetyl; NBS, N-bromosuccinimide.
°C, followed by warming to 0 °C, induced selective cleavage
of the acetonide and TBS groups, events that triggered three
spontaneous ring closures, furnishing the anticipated ABCD
fragment 42 as a mixture of two diastereomeric sulfoxides (62%
combined yield, Scheme 3). This mixture of sulfoxides was
benzoylated (BzCl, py, 96% yield) and then thermolyzed in
refluxing toluene in the presence of P(OMe)3, leading to olefinic
benzoate 44 in 45% yield as a single compound. NMR
spectroscopic analysis, however, revealed the undesired (13S)
stereochemistry for the synthesized product (44), as indicated
by the NOEs (nOes) shown in Scheme 3. Apparently, the
influence of the aryl sulfoxide moiety was not sufficient to
override the directing effect exerted by the double anomeric
effect that drives the formation of the observed 13S diastereomer
Figure 6. Rationale for the expectation that the 9-hydroxy ABCD domain
will fold into the 13R isomer 53 as the thermodynamically favored isomer
(44). Given the almost neutral conditions employed to convert
42 to 44, it is unlikely that epimerization at C-13 occurred after
the cyclization event. Furthermore, exposure of 44 to various
Lewis and protic acids failed to change its stereochemistry,
confirming the thermodynamically most stable nature of this
isomer as compared to its epi-C-13 counterpart.
(stabilized by one anomeric effect and a hydrogen-bonding arrangement)
over the 13S isomer 52 (stabilized by two anomeric effects).
stage for testing this hypothesis, the ABCD ketone system 51
was targeted and synthesized as shown in Scheme 4. Thus, the
mixture of diastereomeric alcohols 39 (see Schemes 1 and 2
for its preparation) was oxidized (DMP, 95% yield), and the
resulting ketone (45) was transformed to its ethylene ketal
(ethylene glycol, triethylorthoformate, p-TsOH, 46, 98% yield).
The terminal olefin of the latter compound (46) was then
converted by a standard dihydroxylation/periodate cleavage
procedure (OsO4-NMO; NaIO4)23 to afford aldehyde 47 in
quantitative yield. Addition of the lithio derivative derived from
dithiane 2024 and n-BuLi to aldehyde 47 (87% yield), followed
by DMP oxidation (88% yield), yielded ketone 16, whose
TMSOTf-induced polycyclization proceeded as expected to
afford tetracycle 49 as the thermodynamically most stable
system (in 85% yield). Pivaloate formation (PivCl, py, 4-DMAP,
95% yield) followed by dithiane removal (NBS, 2,6-lutidine,
91% yield)25 resulted in the formation of the anticipated ketone
51 via dithiane pivaloate ester 50. The 13S stereochemistry of
3. Synthesis of the Desired ABCD Ring System through
an Intramolecular Hydrogen-Bonding Strategy. Having
reached this roadblock in our drive toward the ABCD domain
of azaspiracid-1, we adopted an alternate approach which was
to rely on the stabilizing effect of a neighboring hydroxyl group
exerted through hydrogen bonding, as shown in Figure 6.
Specifically, it was reasoned that a hydroxyl group at C-9
occupying the equatorial position on ring A may be in a position
to hydrogen-bond with both the ring B and ring C oxygens,
thereby lowering the energy level of the desired 13R stereo-
isomer sufficiently to revert the equilibrium away from the
undesired isomer. Beyond this rationale, we were also encour-
aged by previous work in the field in which such hydrogen-
bonding effects decisively manifested themselves.22 To set the
(22) (a) Ishihara, J.; Sugimoto, T.; Murai, A. Synlett 1998, 603. (b) Paterson,
I.; Wallace, D. J.; Gibson, K. R. Tetrahedron Lett. 1997, 38, 8911. (c)
Claffey, M. M.; Hayes, C. J.; Heathcock, C. H. J. Org. Chem. 1999, 64,
8267. (d) McCauley, J. A.; Nagasawa, K.; Lander, P. A.; Mischke, S. G.;
Semones, M. A.; Kishi, Y. J. Am. Chem. Soc. 1998, 120, 7647.
(23) Pappo, R.; Allen, D. S.; Lemieux, R. U.; Johnson, W. S. J. Org. Chem.
1956, 21, 478.
(24) Jenkins, P. R.; Selim, M. M. R. J. Chem. Res. Miniprint 1992, 3, 701.
9
J. AM. CHEM. SOC. VOL. 128, NO. 7, 2006 2249

A R T I C L E S
Nicolaou et al.
Scheme 5. Construction of Originally Proposed ABCD Domain
(57) of Azaspiracid-1 through Trifluoroacetic Acid-Induced
Epimerizationa
Scheme 6. Installation of the Side Chain of Azaspiracid-1 and
Synthesis of the ABCD Fragment 65a
a Reagents and conditions: (a) DIBAL-H (1.0 M in toluene, 2.5 equiv),
toluene, -78 °C, 20 min, 92%; (b) (COCl)2 (5.0 equiv), DMSO (11 equiv),
CH2Cl2, -78 °C, 1 h; -60 °C, 1 h; then Et3N (22 equiv), -60 f -30 °C,
1 h, 92%; (c) vinylmagnesium bromide (1.0 M in THF, 1.6 equiv), Et2O,
0 °C, 30 min, 78%; (d) Ac2O (5.0 equiv), py (10 equiv), 4-DMAP (0.1
equiv), CH2Cl2, 0 °C, 1 h, 94%; (e) LDA (1.5 equiv), TBSCl (1.5 equiv),
HMPA (1.5 equiv), THF, -78 f 25 °C, 72 h, 82%; (f) MeOH (10 equiv),
DCC (1.2 equiv), 4-DMAP (0.1 equiv), CH2Cl2, 0 f 25 °C, 2 h, 86%; (g)
superhydride (1.0 M in THF, 5.0 equiv), THF, -78 f 0 °C, 30 min, 96%;
(h) PivCl (3.0 equiv), py (10 equiv), 4-DMAP (1.0 equiv), CH2Cl2, 0 f
25 °C, 12 h, 95%; (i) Pd(dppe)2 (0.1 equiv), DBU (1.0 equiv), 1,4-dioxane,
25 °C, 18 h, 92%; (j) SmI2 (2.0 equiv), MeOH:THF (1:1), -78 f 25 °C,
15 min, 95%. Abbreviations: DIBAL-H, diisobutylaluminum hydride;
HMPA, hexamethylphosphoramide; DCC, 1,3-dicyclohexylcarbodiimide;
superhydride, lithium triethylborohydride; dppe, 1,2-bis(diphenylphos-
phino)ethane; DBU, 1,8-diazabicyclo[5.4.0]undec-7-ene.
a Reagents and conditions: (a) NaBH4 (1.0 equiv), MeOH, -5 °C, 5
min, 92%; (b) TFA (3.0 equiv), CH2Cl2, 25 °C, 4 h, 56% (40% recovered
52); (c) (COCl)2 (5.0 equiv), DMSO (11 equiv), CH2Cl2, -78 °C, 1 h;
Et3N (22 equiv), CH2Cl2, -78 f 0 °C, 1 h, 80%; (d) KHMDS (0.5 M in
toluene, 4.5 equiv), 55 (5.0 equiv), THF, -78 °C, 45 min, 83%; (e)
[Pd(PPh3)4] (0.2 equiv), n-Bu3SnH (10 equiv), THF, 25 °C, 45 min, 90%.
Abbreviations: DMSO, dimethyl sulfoxide; KHMDS, potassium bis(tri-
methylsilyl)amide.
the last three compounds was evident from NOE studies carried
out with 49 (see Scheme 4).
the next milestone intermediate (i.e., 57) by Swern oxidation26
of 53 to the corresponding ketone (80% yield), followed by
enol triflate formation (KHMDS, Comin’s reagent (55),27 83%
yield) and reductive removal of the triflate group (n-Bu3SnH,
Pd(PPh3)4, 90% yield). The 1H NMR data and NOE studies of
the resulting product revealed the desired 13R stereochemistry
(see NOEs on structure 57, Scheme 5).
Having secured the targeted ABCD fragment 57, it was time
to extend its A ring chain by four carbons as required for the
azaspiracid-1 structure. Three methods were developed for this
task; the first two approaches are shown in Scheme 6. Thus,
pivaloate ester 57 was reduced with DIBAL-H (92% yield), and
the resulting alcohol (58) was oxidized under Swern conditions
[(COCl)2, DMSO, Et3N] to afford aldehyde 59 (92% yield),
Scheme 5 summarizes the events that took ketone 51 to the
next stage along the path toward the desired ABCD domain,
intermediate 57. Thus, reduction of 51 with NaBH4 in MeOH
at 0 °C furnished exclusively, and in 92% yield, the desired 9R
(axial) alcohol 52, ready for the intended equilibration experi-
ments. Indeed, and much to our delight, upon exposure to TFA
in CH2Cl2 at 0 °C, hydroxy spiroacetal 52 equilibrated to a
mixture of C-13 isomers (53:52, ca. 2:1) which were conve-
niently separated by silica gel chromatography. Recycling of
the recovered starting isomer (twice) brought the yield of the
new product (53) to 84%, making the sequence into a practical
process for accessing what was, at this stage, presumed to be
the desired 13R spiroacetal stereoisomer. This assertion was
proven beyond doubt by conversion of this compound (53) to
(25) Williams, D. R.; Jass, P. A.; Allan Tse, H.-L.; Gaston, R. D. J. Am. Chem.
Soc. 1990, 112, 4552.
(26) Omura, K.; Sharma, A. K.; Swern, D. J. Org. Chem. 1976, 41, 957.
(27) Commins, D. L.; Dehghani, A. Tetrahedron Lett. 1992, 33, 6299.
9
2250 J. AM. CHEM. SOC. VOL. 128, NO. 7, 2006

Synthesis and Structure of Azaspiracid-1
A R T I C L E S
Scheme 7. Olefin Metathesis-Based Attachment of the C1-C5
Side Chain of Azaspiracid-1 To Afford 65a
Scheme 8. Construction of Acetal Aldehyde 24a
a Reagents and conditions: (a) (+)-(Ipc)2Ballyl (2.0 equiv), Et2O, -100
°C, 4 h, 88%; (b) acrylic acid (2.0 equiv), DCC (2.5 equiv), 4-DMAP (0.1
equiv), CH2Cl2, 25 °C, 18 h, 73%; (c) 80 (0.1 equiv), CH2Cl2, 40 °C, 5 h,
95%; (d) H2O:AcOH (1:2), 55 °C, 2 h; (e) TBDPSCl (1.4 equiv), Et3N
(3.0 equiv), 4-DMAP (0.1 equiv), CH2Cl2, 0 f 25 °C, 4 h, 98% over two
steps; (f) TESOTf (2.0 equiv), 2,6-lutidine (3.0 equiv), CH2Cl2, 0 °C, 88%;
(g) Me2Cu(CN)Li2 (2.0 equiv), Et2O, -78 °C, 1 h, 97% (ca. 10:1); (h)
BnOCH2Li (2.0 equiv), THF, -78 f -40 °C, 1 h, 52% (ca. 1:1); (i)
p-TsOH (0.03 equiv), CH2Cl2, 1 h, 25 °C, 90%; (j) TBAF (1.0 M in THF,
1.5 equiv), THF, 25 °C, 1 h, 93%; (k) DMP (2.0 equiv), NaHCO3 (10 equiv),
CH2Cl2, 25 °C, 3 h, 86%. Abbreviation: TBAF, tetra-n-butylammonium
fluoride.
a Reagents and conditions: (a) TESCl (1.5 equiv), imidazole (3.0 equiv),
4-DMAP (0.1 equiv), CH2Cl2, 0 °C, 10 h, 92%; (b) DIBAL-H (1.0 M in
toluene, 3.0 equiv), CH2Cl2, -78 °C, 1 h, 87%; (c) (COCl)2 (5.0 equiv),
DMSO (10 equiv), CH2Cl2, -78 °C, 1 h; Et3N (15 equiv), -78 f 0 °C, 1
h, 88%; (d) Ph3PCH3+Br- (6.0 equiv), n-BuLi (1.6 M in THF, 5.0 equiv),
THF, -78 f 0 °C, 1 h; then -78 °C; then 69, -78 f 0 °C, 30 min, 81%;
(e) 71 (0.1 equiv), 72 (3.0 equiv), CH2Cl2, 40 °C, 12 h, 65% (30% recovered
70); (f) HF‚py (10 equiv), THF:py (1:1), 0 °C, 2 h, 94%; (g) (COCl)2 (5.0
equiv), DMSO (10 equiv), CH2Cl2, -78 °C, 1 h; Et3N (15 equiv), -78 f
0 °C, 1 h, 95%; (h) KHMDS (0.5 M in toluene, 4.5 equiv), 55 (5.0 equiv),
THF, -78 °C, 1 h, 94%; (i) Pd(PPh3)4 (0.2 equiv), n-Bu3SnH (3.0 equiv),
THF, 25 °C, 40 min, 96%. Abbreviations: TES, triethylsilyl; Mes,
mesitylene; Cy, cyclohexane.
hydroxy pivaloate ester 53 was first converted to TES-protected
ether terminal olefin 70 by a series of four steps [(i) TESCl,
4-DMAP, imidazole, 67; (ii) DIBAL-H, 68; (iii) (COCl)2,
DMSO; Et3N, 69; and (iv) Ph3PCH2+Br-, n-BuLi, 70 (57% over
the four steps)]. Terminal olefin 70 was reacted with 3.0 equiv
of olefin 72 in the presence of the Grubbs second-generation
catalyst (71) to afford the desired C1-C20 fragment 73 cleanly
(65% yield, plus 30% recovered 70) and as a single (E)
geometrical isomer. Resubmitting the recovered starting material
to the reaction conditions two more times led to 90% combined
yield for this conversion (70 f 73). Selective desilylation of
the secondary alcohol within 73 was cleanly effected by
exposure to HF‚py (94% yield); Swern oxidation [(COCl)2,
DMSO; Et3N] of the resulting compound (74) then furnished
ketone 75 in 95% yield. Sequential enolization (KHMDS) of
this ketone 75, followed by reaction with Comin’s reagent (55),
then gave enol triflate 76, whose reductive cleavage [Pd(PPh3)4,
which was reacted with vinylmagnesium bromide, leading to
allylic alcohol 60 (ca. 1:1 mixture of diastereomers, 78%
combined yield). This mixture was acetylated (Ac2O, py,
4-DMAP, 94% yield), and the acetates (61) so obtained were
subjected to an Ireland-Claisen rearrangement (LDA, TBSCl,
HMPA, -78 f 25 °C)28 to afford the elongated carboxylic acid
62 in 82% yield. The latter compound was then converted to
its methyl ester (63, MeOH, DCC, 4-DMAP, 86% yield) and
then, through reduction (superhydride, -78 f 0 °C, 64, 96%
yield) and pivaloate ester formation (PivCl, py, 4-DMAP, 95%
yield), to the targeted pivaloate-TBDPS derivative 65. A second
sequence to convert acetates 61 to methyl ester 63 relied on a
coupling reaction employing methyl phenylsulfonyl acetate,
Pd(dppe)2, and DBU as a means to elongate the chain, affording
intermediate 66 (92% yield), whose extraneous phenylsulfonyl
group was reductively removed by treatment with SmI2 (63,
95% yield) (Scheme 6).29
Finally, a more expedient sequence based on an olefin cross-
metathesis reaction30 was developed for the attachment of the
side chain onto the ABCD domain (see Scheme 7). Thus,
(28) Ireland, R. E.; Mueller, R. H. J. Am. Chem. Soc. 1972, 94, 5897.
(29) Molander, G. A.; Hahn, G. J. Org. Chem. 1986, 51, 1135.
(30) For a review on the use of Grubbs’ catalyst in total synthesis, see: Nicolaou,
K. C.; Bulger, P. G.; Sarlah, D. Angew. Chem., Int. Ed. 2005, 44, 4490.
9
J. AM. CHEM. SOC. VOL. 128, NO. 7, 2006 2251

A R T I C L E S
Nicolaou et al.
Scheme 9. Synthesis of Azido-Ketone 23a
Scheme 10. Failed Attempt To Construct the FGHI Ring
Fragments of Azaspiracid-1 from Azido-Hydroxy Ketone 22 via 94a
a Reagents and conditions: (a) MeO(NHMe)‚HCl (5.0 equiv), AlMe3
(2.0 M in toluene, 5.1 equiv), THF, -15 °C, 2 h, 96%; (b) p-TsCl (1.5
equiv), Et3N (5.0 equiv), CH2Cl2, 25 °C, 18 h, 96%; (c) NaN3 (2.0 equiv),
DMF, 25 °C, 76 h, 96%; (d) MeLi (1.25 M in Et2O, 1.0 equiv), THF, -78
°C, 40 min, 82%.
n-Bu3SnH] led to the coveted segment 65 in 90% overall yield
for the two steps.
a Reagents and conditions: (a) LDA (1.5 equiv), TESCl (3.0 equiv), THF,
-78 °C, 86%; (b) SnCl4 (1.0 M in CH2Cl2, 1.0 equiv), CH2Cl2, -78 °C;
then 93 (1.0 equiv), 1.5 h, 50%; (c) Ph3P (3.0 equiv), THF:H2O (10:1), 25
°C, 21 h.
4. First Attempt To Prepare the FGHI Domain of
Azaspiracid-1. We now turn our attention to the construction
of the FGHI ring system 14 as defined by the retrosynthetic
analysis depicted in Figure 4. Our plan called for the interme-
diacy of the open-chain amino bis-carbonyl trihydroxy system
21 (Figure 4) or its equivalent, whose acid-catalyzed folding
was expected, upon suitable elaboration, to furnish tetracycle
14 with the two spiro centers hopefully in the proper stereo-
chemical arrangement. To test this hypothesis, we targeted azido
acetal 22 (Figure 4) as a potential precursor to the desired
compound (14).
Thus, and as shown in Scheme 8, Brown allylboration31 of
D-isopropylidene glyceraldehyde (77)32 with (+)-(Ipc)2Ballyl
gave allylic alcohol 78 (ca. 96:4 mixture, 88% combined yield),
whose reaction with acrylic acid in the presence of DCC gave
acrylate ester 79 (73% yield). Ring-closing metathesis within
79 was initiated by the Grubbs first-generation catalyst (80),
furnishing the R,â-unsaturated δ-lactone 81 in 95% yield,
whose destination was to become ring F of the final target. At
this stage, it was considered necessary to differentially protect
the two hydroxyl groups of the growing molecule; to that end,
the acetonide group was removed (AcOH-H2O), and the
resulting diol was sequentially and selectively engaged first with
TBDPSCl-Et3N (98% yield over two steps) and then with
TESOTf-2,6-lutidine (88% yield), leading to intermediates 84
via 83 and 82.
Biased by the side chain of the substrate, the 1,4-addition of
Me2Cu(CN)Li2 to R,â-unsaturated lactone 84 (see Scheme 8)
delivered, stereoselectively, methylated lactone 85 in 97% yield
(ca. 10:1 mixture of isomers).33 This compound (85) was then
reacted with BnOCH2Li34 to afford the hemiacetal 86 (ca. 1:1
mixture of isomers, 52% combined yield). Exposure of the latter
compound (86) to p-TsOH in refluxing benzene resulted in its
conversion to intramolecular bridged ketal 87 (90% yield),
whose casting proceeded spontaneously upon the expected
departure of the TES group. Sequential removal of the TBDPS
group from 87 with TBAF (93% yield), followed by oxidation
of the resulting primary alcohol (88) with Dess-Martin perio-
dinane, led to aldehyde 24 (86% yield).
The synthesis of aldol partner 23 commenced from known
lactone 8935 and proceeded as shown in Scheme 9. Thus,
treatment of 89 with MeONHMe‚HCl and AlMe3 afforded
Weinreb amide 90 in 96% yield.36 Subsequent tosylation of the
liberated primary alcohol (p-TsCl, Et3N, 96% yield) within 90,
followed by displacement of the newly formed tosylate (91)
with NaN3, afforded azide 92 in 96% yield. Finally, reaction of
MeLi with 92 resulted in displacement of the amide moiety,
leading to the desired azido-methyl ketone 23 (82% yield).
With both fragments 23 and 24 in hand, their coupling became
the next task (see Scheme 10). To this end, a SnCl4-catalyzed
Mukaiyama aldol reaction37 with TES enol ether 93 (obtained
from methyl ketone 23, LDA, and TESCl, in 86% yield) and
aldehyde 24 was performed, furnishing the coveted azido-
hydroxyketone 22 stereoselectively in 50% yield. Carrying all
the carbon atoms required for the targeted FGHI ring system
95, the latter compound was only two steps away from its final
destination. Staudinger reaction38 with 22 (Ph3P, THF, H2O)
provided the desired intermediate, primary amine 94, whose
desired ring-opening/polycyclization, however, could not be
achieved, as treatment with protic or Lewis acids led to either
decomposition or a large number of products, none of which
resembled the desired tetracycle 95. Faced with this roadblock,
we resorted to a new strategy.
5. Second-Generation Retrosynthetic Analysis of Aza-
spiracid-1 (1a). To circumvent the problems encountered in
our first strategy toward the construction of the targeted aza-
spiracid-1 (1a), and having gathered enough intelligence regard-
ing this challenge, we proceeded to develop a new strategy based
on a second retrosynthetic analysis whose basic disconnections
are shown in Figure 7. Thus, opening ring G of 1a led, upon
suitable modification, to hydroxy enol ether 96 as a potential
precursor to the final target through a short sequence involving
selective iodo-etherification/reductive de-iodination as a key
(35) Collum, D. B.; McDonald, J. H.; Still, W. C. J. Am. Chem. Soc. 1980,
102, 2118.
(36) Nahm, S.; Weinreb, S. M. Tetrahedron Lett. 1981, 22, 3815.
(37) Mukaiyama, T.; Narasaka, K.; Banno, K. Chem. Lett. 1973, 1011.
(38) Staudinger, H.; Meyer, J. HelV. Chim. Acta 1919, 2, 635.
(31) Racherla, U. S.; Brown, H. C. J. Org. Chem. 1991, 56, 401.
(32) Ahrendt, K. A.; Williams, R. M. Org. Lett. 2004, 6, 4539.
(33) Lipshutz, B. H. Synthesis 1987, 325.
(34) Still, W. C. J. Am. Chem. Soc. 1978, 100, 1481.
9
2252 J. AM. CHEM. SOC. VOL. 128, NO. 7, 2006

Synthesis and Structure of Azaspiracid-1
A R T I C L E S
Figure 7. Second-generation retrosynthetic analysis of the originally proposed structure of azaspiracid-1 (1a).
protocol. The next disconnection was performed at the
C27-C28 bond (in contrast to the first retrosynthetic analysis
which involved disconnection of the C25-C26 bond), unraveling
allylic acetate 97 and vinyl stannane 98 as potential precursors.
In the synthetic direction, it was envisioned that a Stille
coupling39 would ensure the assembly of these fragments to form
96. The C1-C27 fragment 97 was then disconnected, employing
a retro-dithiane addition reacting onto the activated ester 99 (note
exchange of the acetate with the more robust pivaloate group),
whose origin was traced to the truncated ABCD domain 70 via
a retro cross-olefin metathesis reaction with terminal olefin 72
[the former intermediate could be envisioned to arise from key
building blocks 17, 18 (see also Figure 4), and 20]. The FHI
segment 98 was stepwise disassembled, first to bicycle 101 by
rupturing ring I and then to δ-lactone 102 by opening ring H,
as shown in Figure 7. Finally, a retro-aldol transform on 102
pointed to aldehyde 103 and azido-methyl ketone 23 as starting
points for the construction of the FGHI domain of the target
molecule.
The construction of the FHI stannane 98 (Figure 7) com-
menced with the (+)-diethyltartrate-derived PMP acetal 10540
and proceeded along the path delineated in Schemes 11-13.
Thus, the diol 105 (Scheme 11) was monoprotected as a TBDPS
ether (n-BuLi, TBDPSCl, 87% yield), and the resulting product
(106) was subjected to a regioselective acetal opening with
BH3‚THF, furnishing the PMB ether 1,2-diol 107 in 76% yield.
Cleavage of the vicinal diol system within 107 with NaIO4 gave
(39) (a) Stille, J. K. Angew. Chem., Int. Ed. Engl. 1986, 25, 508. (b) Del Valle,
L.; Stille, J. K.; Hegedus, L. S. J. Org. Chem. 1990, 55, 3019.
(40) Somfai, P.; Olsson, R. Tetrahedron 1993, 49, 6645.
9
J. AM. CHEM. SOC. VOL. 128, NO. 7, 2006 2253

A R T I C L E S
Nicolaou et al.
Scheme 11. Synthesis of Key Intermediate 102a
Scheme 12. Preparation of Tricycle 119a
a Reagents and conditions: (a) PPTS (0.3 equiv), MeOH, 25 °C, 2 h,
93% (ca. 1:1); (b) H2, 10% Pd/C (25% w/w), EtOAc, 25 °C, 7 h; (c) Teoc-
O-(p-NO2Ph) (3.0 equiv), Et3N (4.0 equiv), EtOAc, 25 °C, 15 h, 80% over
two steps; (d) Nd(OTf)3 (0.1 equiv), MeCN, 25 °C, 15 min, 81%; or
Yb(OTf)3 (0.1 equiv), MeCN, 25 °C, 3 min, 72%. Abbreviation: PPTS,
pyridinium p-toluenesulfonate.
a Reagents and conditions: (a) n-BuLi (1.6 M in THF, 1.1 equiv),
TBDPSCl (1.1 equiv), THF, -78 f 0 °C, 18 h, 87%; (b) BH3‚THF, THF,
65 °C, 4 h, 76%; (c) NaIO4 (4.0 equiv), THF:H2O (3:2), 25 °C, 4 h, 88%;
(d) (+)-(Ipc)2Ballyl (2.0 equiv), Et2O, -100 °C, 2 h, 100% (ca. 96:4); (e)
acrylic acid (3.0 equiv), DCC (3.0 equiv), 4-DMAP (0.1 equiv), CH2Cl2,
40 °C, 18 h, 78%; (f) 80 (0.1 equiv), CH2Cl2, 40 °C, 18 h, 91%; (g)
Me2Cu(CN)Li2 (2.0 equiv), Et2O, -78 °C, 1 h, 95%; (h) TBAF (1.0 M in
THF, 1.5 equiv), THF, 25 °C, 30 min, 87%; (i) DMP (1.4 equiv), py (10
equiv), CH2Cl2, 25 °C, 92%; (j) 23 (1.1 equiv), Cy2BCl (1.3 equiv), i-Pr2NEt
(1.5 equiv), CH2Cl2, 0 °C, 1.5 h; then 114 (1.0 equiv), -78 °C, 3.5 h, 93%
(ca. 38:1); (k) BzCl (3.0 equiv), py, 0 °C, 4 h, 88%; (l) DDQ (1.5 equiv),
CH2Cl2:H2O (10:1), 0 °C, 3 h, 96%.
spectroscopy). This welcome result was then followed by pro-
tecting group exchanges, leading sequentially from alcohol 115
to benzoate 116 (BzCl, py, 88% yield) and then to secondary
alcohol 102 (DDQ, 96% yield). The latter intermediate (102)
contains, besides all required carbon atoms for the target FHI
ring system 98, the appropriately equipped functional groups
for elaboration to the targeted model FGHI system (i.e., 130,
Scheme 14).
aldehyde 108, whose reaction with (+)-(Ipc)2Ballyl, according
to Brown,31 furnished allylic alcohol 109 as a 96:4 mixture with
its stereoisomer (100% combined yield). Esterification of 109
with acrylic acid, as facilitated by DCC, yielded acrylate 110
(78% yield), which served admirably as a substrate for a ring-
closing metathesis initiated by the Grubbs first-generation
catalyst (80) to afford R,â-unsaturated lactone 111 in 91% yield.
A conjugate addition to 111 by Me2Cu(CN)Li2 installed stereo-
selectively (see Scheme 8, structure 84 for a mechanistic ration-
ale) the required methyl group, yielding compound 112 in 95%
yield. Removal of the TBDPS group from 112 through the action
of TBAF then gave alcohol 113 (87% yield), whose DMP
oxidation (92% yield) led to aldehyde 114. The latter com-
pound (114) was then reacted with the boron enolate derived
from methyl ketone 23 (Cy2BCl, i-Pr2NEt) according to the
procedure of Patterson41 to furnish stereoselectively the aldol
product 115 (93% yield, ca. 38:1 ratio of isomers by 1H NMR
Scheme 12 shows the advancement of intermediate 102 to
the desired FHI ring system 119. Thus, exposure of 102 to PPTS
in methanol resulted in the formation of methoxyacetal 117 (ca.
1:1 mixture of stereoisomers, 93% combined yield) with the
second required ring (H) in place. Subsequent reduction of the
azido group within 117 with H2-10% Pd(OH)2, followed by
treatment with Teoc-O-(p-NO2Ph),42 furnished the Teoc-
protected amine 101 (80% yield over two steps) through in situ
trapping of the incipient primary amine (118). Finally, exposure
of this amino-acetal (101) to Nd(OTf)3 in MeCN at 25 °C led
to expulsion of the methoxy group and ring closure, generating
the desired spiroaminal 119 in 81% yield as a single stereoiso-
mer. The stereochemical arrangement within 119 was confirmed
(41) For a review on boron-mediated aldol reactions, see: Cowden, C. J.;
Patterson, I. In Organic Reactions; Paquette, L. A., Ed.; John Wiley &
Sons: New York, 1999; Vol. 1, pp 1-200.
(42) Rosowsky, A.; Wright, J. E. J. Org. Chem. 1983, 48, 1539.
9
2254 J. AM. CHEM. SOC. VOL. 128, NO. 7, 2006

Synthesis and Structure of Azaspiracid-1
A R T I C L E S
Table 2. Optimization of Ring I Cyclization
Scheme 13. Synthesis of Vinyl Stannane 98a
entry
conditionsa
yield (%)b
1
2
p-TsOH (0.1 equiv), MeOH, 25 °C, 24 h
CSA (0.1 equiv), MeOH, 25 °C, 1 h
0
0
3
4
5
6
7
8
9
10
TfOH (0.1 equiv), CH2Cl2, 0 °C, 10 min
BF3‚Et2O (0.1 equiv), CH2Cl2, 0 °C, 10 min
Yb(OTf)3 (0.1 equiv), MeCN, 25 °C, 3 min
Sn(OTf)2 (0.1 equiv), MeCN, 25 °C, 5 min
Ag(OTf)2 (0.1 equiv), MeCN, 25 °C, 30 min
Y(OTf)3 (0.1 equiv), MeCN, 25 °C, 10 min
Nd(OTf)3 (0.1 equiv), MeCN, 25 °C, 15 min
Eu(OTf)3 (0.1 equiv), MeCN, 25 °C, 15 min
0
60
72
0
0
22
81
65
a Reactions were carried out at 0.1-0.2 mmol scale. b Yields refer
to isolated products, the desired isomer (119) being obtained as a single
isomer.
by NOE studies (see Scheme 12). The use of Nd(OTf)3 came
about after a systematic study to optimize this ring closure,
which proved nontrivial. Thus, and as shown in Table 2, initial
attempts to employ protic acids (entries 1-3) failed to produce
any of the desired product (119), leading, instead, to several
unidentified products or decomposition. The first good results
were obtained with the Lewis acid BF3‚Et2O (entry 4) in
CH2Cl2 at 0 °C, which led to ring-closed product 119 in 60%
yield. The Forsyth group demonstrated the use of Yb(OTf)3 as
an effective catalyst in inducing such a ring closure in a similar
system,4f and indeed the same conditions proved quite effective
in converting 101 to 119 (72% yield, entry 5). Further
experimentation with other species (entries 6-10) led to the
discovery that Nd(OTf)3 was an even more effective catalyst
in closing ring I, producing 119 from 101 in 81% yield as
already mentioned above.
At this juncture, it was necessary to invert the stereochemistry
of the C-34 hydroxyl group within the last intermediate (119,
Scheme 13), for it was all along opposite of that embedded
within the natural product. Manual molecular modeling sug-
gested that a simple oxidation-reduction sequence may suffice
to accomplish the required inversion on the basis of steric
hindrance alone (see structure 122 in Scheme 13). To this end,
the benzoate group was cleaved from tricyclic compound 119
by DIBAL-H reduction, an operation which also reduced the
lactone moiety to the corresponding lactol, leading to intermedi-
ate 120. To recover the required dicarbonyl compound (i.e., 122)
for the intended elaboration, it was necessary to employ a two-
step protocol, namely treatment with NIS and n-BuNI, to afford
the intermediate hydroxy lactone 121 (70% yield from 119),
followed by oxidation with DMP (87% yield). Delightfully, the
reduction of ketone 122 with L-selectride, a bulky reagent,
furnished the desired 34S-hydroxy compound (123) in 79%
yield. TES protection of this alcohol (123) (TESOTf, 2,6-
lutidine, 93% yield) then furnished intermediate 124, whose
reaction with KHMDS and Comin’s reagent (55) led to the
expected vinyltriflate 125 (89% yield), from which the targeted
FHI stannane 98 was finally generated in 98% yield by the
action of (SnMe3)2 in the presence of Pd2(dba)3, TFP, and LiCl
(see Scheme 13).
a Reagents and conditions: (a) DIBAL-H (1.0 M in toluene, 4.0 equiv),
toluene, -78 °C, 30 min; (b) NIS (10 equiv), n-Bu4NI (2.0 equiv), CH2Cl2,
25 °C, 1 h, 70% over two steps; (c) DMP (1.4 equiv), py (10 equiv), CH2Cl2,
25 °C, 3 h, 87%; (d) L-selectride (1.0 M in THF, 2.0 equiv), THF, -78 °C,
20 min, 79%; (e) TESOTf (1.5 equiv), 2,6-lutidine (3.0 equiv), CH2Cl2,
-78 °C, 10 min, 93%; (f) KHMDS (1.0 M in toluene, 4.0 equiv), 55 (5.0
equiv), THF, -78 °C, 45 min, 89%; (g) (Me3Sn)2 (10 equiv), TFP (0.5
equiv), LiCl (3.0 equiv), Pd2dba3 (0.1 equiv), THF, 25 °C, 1 h, 98%.
Abbreviations: L-selectride, lithium tri-sec-butylborohydride; TFP, trifu-
rylphosphine; dba, dibenzylidene acetone.
Having obtained both the vinyl triflate 125 and stannane 98
as described above, we were in a position to test their suitability
as partners for the intended Stille coupling reaction and the
subsequent ring G closure. Toward this end, the vinyl triflate
125 proved highly effective in coupling with allyl-tri-n-butyltin
in the presence of Pd2dba3, TFP, and LiCl, furnishing diene
126 in 95% yield (Scheme 14). Removal of the TES group from
the latter substance (126) with the aid of HF‚py afforded the
secondary alcohol 127 in 94% yield. A brief search for a suitable
reagent to close ring G, employing hydroxy diene 127 as a
substrate, revealed both Hg(OAc)2 (75% yield) and NIS (77%
yield) as equally effective (as it turned out, the mercuric acetate-
based method would fail in the real system, leaving the NIS
method as the procedure to be called upon in the end of the
campaign to synthesize the targeted azaspiracid). Iodide 129,
obtained by exposure of 127 to NIS in the presence of NaHCO3
in THF at 0 °C, crystallized nicely, allowing its X-ray
crystallographic analysis (see ORTEP in Scheme 14), which
confirmed its structure as the desired one. Reductive removal
of the extraneous iodide residue from 129 was finally effected
with n-Bu3SnH:toluene (1:2) and Et3B at 0 °C. The large excess
9
J. AM. CHEM. SOC. VOL. 128, NO. 7, 2006 2255

A R T I C L E S
Nicolaou et al.
Scheme 14. Construction of Model FGHI Compound 130 and
ORTEP Drawing of Compound 129a
Scheme 15. Construction of C21-C27 Fragment 100a
a Reagents and conditions: (a) DIBAL-H (1.0 M in CH2Cl2, 1.1 equiv),
CH2Cl2, -78 °C, 1.5 h; (b) 1,3-propanedithiol (1.1 equiv), BF3‚Et2O
(1.5 equiv), CH2Cl2, 0 °C, 1 h, 99% over two steps; (c) (COCl)2 (1.2
equiv), DMSO (2.4 equiv), CH2Cl2, -78 °C, 30 min; then Et3N (5.0
equiv), -78 f -20 °C, 94%; (d) TMSCl (1.2 equiv), NaI (1.2 equiv),
H2O (0.6 equiv), MeCN, 0 f 25 °C, 1.5 h, 51%; (e) TBSCl (1.2 equiv),
imidazole (2.5 equiv), DMF, 25 °C, 36 h, 96%; (f) NiCl2 (0.02 equiv),
CrCl2 (4.0 equiv), DMF, 0 °C; then 12 (1.0 equiv), 104 (2.5 equiv), 0 f
25 °C, 15 h, 95%; (g) IBX (2.0 equiv), DMSO:THF (4:1), 25 °C, 2 h,
90%; (h) Red-Al (2.5 equiv), toluene, -78 °C, 1 h, 80%; (i) TBAF
(2.2 equiv), THF, 25 °C, 1 h, 99%; (j) t-Bu2Si(OTf)2 (1.6 equiv), 2,6-luti-
dine (4.0 equiv), CH2Cl2, -30 °C, 30 min, 75%. Abbreviations: IBX,
o-iodoxybenzoic acid; Red-Al, sodium bis(2-methoxyethoxy)aluminum
hydride.
a Reagents and conditions: (a) allyltri-n-butyltin (10 equiv), TFP (0.5
equiv), LiCl (3.0 equiv), Pd2dba3 (0.1 equiv), THF, 25 °C, 1 h, 95%; (b)
HF‚py (5.0 equiv), THF:py (1:1), 0 f 25 °C, 2 h, 94%; (c) NIS (10 equiv),
NaHCO3 (30 equiv), THF, 0 °C, 12 h, 77% for 129; (d) Hg(OAc)2 (3.5
equiv), THF:H2O (3:1), -5 °C, 20 min; (e) NaBH4 (10 equiv), 75% from
127 via 128; (f) Et3B (1.0 M in hexanes, 3.0 equiv), n-Bu3SnH:toluene
(1:2), 0 °C, 5 min, 94%.
of n-Bu3SnH was necessary in order to avoid ring closure of
the initially formed carbon-centered radical onto the terminal
olefin leading to a pentacyclic byproduct. The mercurial
intermediate (128) could be converted to the same FGHI ring
system 130 by reduction with NaBH4 in THF:H2O (3:1) at -5
°C (75% yield over two steps).
Conclusion
In this article, we described stereoselective constructions of
all three key building blocks required for the total synthesis of
the originally proposed structure of azaspiracid-1 (1a). The
developed chemistry was applied to the synthesis of both
enantiomers of these intermediates, in the event that they were
needed in reaching the correct structure of the targeted molecule,
since neither its relative (between the ABCDE and FGHI
domains) nor its absolute stereochemistry was known at the time.
Furthermore, within this paper we described synthetic technolo-
gies that could potentially be used to assemble the entire skele-
ton of what we believed was the structure of azaspiracid-1,
including a method for connecting all fragments and casting
ring G. In the following paper8 we describe the application of
these technologies to the construction of the targeted structures
and our findings relating to the true identity of the natural
substance.
The last synthesis to be described herein is the construction
of the C21-C27 fragment corresponding to the E-ring of the
targeted azaspiracid-1, as shown in Scheme 15, and starting with
δ-lactone 89. DIBAL-H reduction of 89, followed by exposure
of the resulting lactol (131) to 1,3-propanedithiol in the presence
of BF3‚OEt2, furnished hydroxydithiane 132 in 99% yield over
two steps. Swern oxidation [(COCl)2, DMSO; Et3N] of the latter
compound (132) then gave aldehyde 12 (94% yield), whose
Nozaki-Hiyama-Kishi coupling with vinyl iodide 104 (ob-
tained in 49% overall yield from propargyl alcohol 133 by
reaction with NaI and TMSCl, followed by silylation of the
resulting allylic alcohol 134 with TBSCl and imidazole, see
Scheme 15) in the presence of CrCl2 and NiCl2 (cat.) yielded
allylic alcohol 135 as a 1:1 mixture of C-25 diastereomers in
95% yield. Oxidation of this mixture with IBX in DMSO
furnished ketone 136 (90% yield), whose reduction with Red-
Al at -78 °C proceeded stereoselectively to afford the desired
allylic alcohol 137 as a single stereoisomer in 80% yield.
Desilylation of 137 (TBAF, 99% yield), followed by treatment
of the resulting 1,3-diol (138) with t-Bu2Si(OTf)2 and 2,6-
lutidine, gave the targeted cyclic silyl ether 100 in 75% yield.
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
for this work was provided by The Skaggs Institute for Chemical
Biology and the National Institutes of Health (USA), predoctoral
fellowships from Bristol-Myers Squibb, The Skaggs Institute
for Research, and The Scripps Research Institute Society of
9
2256 J. AM. CHEM. SOC. VOL. 128, NO. 7, 2006

Synthesis and Structure of Azaspiracid-1
A R T I C L E S
Fellows (all to F.B.), and the National Science Foundation
(M.O.F.), postdoctoral fellowships from the Academy of
Finland, the Ella and Georg Ehrnrooth Foundation, and the
Tauno To¨nning Foundation (all to P.M.P.), The Skaggs Institute
for Research (W.Q.), Bayer AG (N.D. and J.H.), and Array-
Biopharma (N.Z.), and grants from Abbott, Amgen, ArrayBio-
pharma, Boehringer-Ingelheim, Glaxo Smith Kline, Hoffmann-
LaRoche, DuPont, Merck, Novartis, Pfizer, and Schering
Plough.
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.
JA0547477
9
J. AM. CHEM. SOC. VOL. 128, NO. 7, 2006 2257