Total synthesis of the proposed azaspiracid-1 structure, Part 1: Construction of the enantiomerically pure C1-C20, C21-C27, and C28-C40 fragments

Article abstract of DOI:10.1002/anie.200351825

As a build-up to the synthesis of the proposed structure of azaspiracid, both enantiomers of fragments 2, 3, and 4 were synthesized stereoselectively. Piv=pivaloyl, PFP = pentafluorophenyl, Teoc = 2-(trimethylsilyl)ethoxycarbonyl, TES = triethylsilyl.

Full text of DOI:10.1002/anie.200351825

Angewandte
Chemie
Natural Product Synthesis
Total Synthesis of the Proposed Azaspiracid-1
Structure, Part 1: Construction of the
Enantiomerically Pure C1–C20, C21–C27, and
C28–C40 Fragments**
K. C. Nicolaou,* Yiwei Li, Noriaki Uesaka,
Theocharis V. Koftis, Stepan Vyskocil, Taotao Ling,
Mugesh Govindasamy, Wenyuan Qian, Federico Bernal,
and David Y.-K. Chen
An incident of human poisoning in the Netherlands in 1995
with diarrhetic shell poisoning (DSP)-like symptoms was
traced to the consumption of mussels (Mytilus edulis),which
[1]
had been collected in Killary Harbour,Ireland. Following
an intense investigation,Yasumoto and co-workers reported
the isolation of azaspiracid-1,the major biotoxin credited
with these severe symptoms.[2] Both the structure of the
substance and its pathologic effects were sufficiently different
from those associated with the previously known DSP,so that
a new toxic syndrome was declared and named azaspiracid
poisoning (AZP). In addition to its immediate and apparent
harmful effects,azaspiracid-1 has been shown to also cause
lung,liver,spleen,and lymphocyte damage,as well as lung-
tumor formation in mice.[3] It was on the basis of sophisticated
analytical techniques that the Yasumoto group proposed the
structure of azaspiracid-1 as 1. This unprecedented structure
is characterized by a trioxadispiroketal system fused to a
tetrahydrofuran moiety (ABCD ring system),an azaspiro ring
fused to a 2,9-dioxabicyclo[3.3.1]nonane system (FGHI ring
system),a six-membered hemiketal bridge (E ring system),
and a g,d-unsaturated terminal carboxylic acid moiety
(Scheme 1). In total,there are nine rings and 20 stereogenic
centers within this synthetically challenging and novel molec-
ular architecture. Despite the heroic spectroscopic attempts
to elucidate fully the structure of azaspiracid-1,[2] its daunting
Scheme 1. Retrosynthetic analysis of the proposed structure 1 of
azaspiracid-1, leading to key fragments 2 (C1–C20), 3 (C21–C27), and
4 (C28–C40). Relative stereochemistry between ABCDE (C1–C27) and
FGHI (C28–C40) domains and absolute stereochemistry of 1 unknown.
TES=triethylsilyl, Teoc=2-(trimethylsilyl)ethoxycarbonyl.
molecular framework yielded neither its relative stereochem-
istry between domains ABCDE (C1–C27) and FGHI
(C28–C40) nor its absolute stereochemistry. In view of the
health and environmental hazards posed by this molecule,
and to clarify its molecular structure,we initiated a program
directed towards its total synthesis.[4] Herein and in the
following Communication[5] we report the total synthesis of
the proposed structure 1 for azaspiracid-1 and its FGHI
diastereoisomer,which led to the conclusion that this
structure is,in fact,in error.
[*] Prof. Dr. K. C. Nicolaou, Y. Li, Dr. N. Uesaka, Dr. T. V. Koftis,
Dr. S. Vyskocil, Dr. T. Ling, Dr. M. Govindasamy, Dr. W. Qian,
F. Bernal, Dr. D. Y.-K. Chen
As shown in Scheme 1,our proposed synthetic strategy
was based on a twofold disconnection of the polycyclic array
of azaspiracid-1 at the C20–C21 and C27–C28 junctures to
provide the key synthetic precursors 2 (C1–C20 fragment), 3
(C21–C27 fragment),and 4 (C28–C40 fragment). The posi-
tions of the strategic bond scissions were carefully chosen to
ensure optimum convergence and to allow the necessary
flexibility to address the relative stereochemical issue
between the ABCDE and FGHI domains of the molecule.
Department of Chemistry and
The Skaggs Institute for Chemical Biology
The Scripps Research Institute
10550 North Torrey Pines Road, La Jolla, CA 92037 (USA)
Fax: (+1)858-784-2469
E-mail: kcn@scripps.edu
and
Department of Chemistry and Biochemistry
University of California, San Diego
9500 Gilman Drive, La Jolla, CA 92093 (USA)
ꢀ
The C20 C21 bond was envisaged to arise through the
[**] We thank Dr. D. H. Huang and Dr. G. Siuzdak for NMR spectro-
scopic and mass spectrometric assistance, respectively, and Dr. Raj
Chadha for X-ray crystallographic analysis. Financial support for this
work was provided by the National Institutes of Health (USA) and
The Skaggs Institute for Chemical Biology, predoctoral fellowships
from Bristol-Myers Squibb (to F.B.), The Skaggs Institute for
Research (to Y.L. and F.B.), The Scripps Research Institute Society of
Fellows (to F.B.), and Eli Lilly (to Y.L.), and a postdoctoral fellowship
from The Skaggs Institute for Research (to W.Q.).
application of lithiated dithiane coupling technology,[6]
ꢀ
whereas the C27 C28 bridging linkage was reserved for a
palladium-mediated C(sp2)–C(sp3) Stille coupling.[7] We pre-
viously reported the preparation of the ABCD[8] and FGHI[9]
ring systems with the desired skeletal and stereochemical
arrangements,featuring the use of an auxiliary stereocontrol-
ling group at C9 to induce the desired 13R stereochemistry
Angew. Chem. Int. Ed. 2003, 42, 3643 –3648
DOI: 10.1002/anie.200351825
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3643

Communications
and the employment of a Hg(OAc)2-
mediated ring closure to forge the
G ring of the FGHI ring system.
Herein we describe the adaptation
and refinement of our original
sequence,and its application to the
construction of appropriately function-
alized fragments 2, 3,and 4 needed for
the key carbon–carbon bond-forming
reactions outlined in this retrosynthetic
blueprint. To determine the relative
and absolute stereochemical assign-
ment of the nominal azaspiracid-1 (1)
unambiguously,all three fragments 2,
3,and 4 were prepared in both enan-
tiomeric forms. For conciseness,the
following discussion will be limited to
only one series of enantiomers (arbi-
trarily chosen). The synthesis of 1 and
its FGHI diastereoisomer is sufficient
to reveal both the relative and absolute
stereochemistry of the nominal azas-
piracid-1 (1),even though one more
attempt may be needed to synthesize
it.
The construction of key intermedi-
ate 2 (C1–C20 fragment) began with
the protection of the previously
reported alcohol 5[8] as a pivaloate
ester (with the wrong 13S configura-
tion,see Scheme 2),followed by oxi-
dative removal (NBS,26, -lutidine; for
abbreviations of reagents and protect-
ing groups,see legends in schemes) of
the dithiane group to provide ketone 6
(75% yield over two steps). NaBH4-
mediated reduction of 6 afforded alco-
hol 7 as the major isomer (> 9:1 ratio),
which was isolated in 87% yield,set-
ting the stage for the proposed C9-OH-
induced equilibration to establish the
desired 13R spirocyclic configuration.
As anticipated,upon exposure of alco-
hol 7 to TFA in CH2Cl2 at ambient
temperature,an equilibrium mixture
consisting of the desired tetracycle 8
and starting alcohol 7 was reached (ca.
2:1 ratio),from which 8 was isolated in
pure form by flash column chromatog-
raphy (66% yield). Recycling of 7
increased the supply of 8. The auxiliary
hydroxy group at C9 was then removed
by a three-step sequence involving
Swern oxidation,enol triflate forma-
tion[10] (KHMDS,reagent 10,92%
yield),and [Pd(PPh 3)4]/nBu3SnH-
mediated reductive elimination[11]
Scheme 2. Construction of 2: a) PivCl (3.0 equiv), pyridine (10.0 equiv), DMAP (catalytic),
CH2Cl2, 0!258C, 12 h; b) NBS (8.0 equiv), 2,6-lutidine (16 equiv), MeCN/H2O (4:1), 258C, 1 h,
75% over two steps; c) NaBH4 (1.1 equiv), MeOH, ꢀ78!ꢀ608C, 3 h, 87%; d) TFA (2.0 equiv),
CH2Cl2, 0!258C, 2 h, 66%; e) (COCl)2 (2.0 equiv), DMSO (4.0 equiv), CH2Cl2, ꢀ608C, 2 h; then
Et3N (8.0 equiv), ꢀ60!ꢀ308C, 1 h, 94%; f) 10 (2.5 equiv), KHMDS (0.5m in toluene,
2.5 equiv), THF, ꢀ788C, 1 h, 92%; g) LiCl (3.0 equiv), [Pd(PPh3)4] (0.2 equiv), nBu3SnH
(3.0 equiv), THF, 258C, 45 min, 95%; h) DIBAL-H (1.0m in toluene, 2.5 equiv), toluene, ꢀ788C,
20 min, 92%; i) (COCl)2 (5.0 equiv), DMSO (11 equiv), CH2Cl2, ꢀ788C, 1 h, ꢀ608C, 1 h; then
Et3N (22 equiv), ꢀ78!ꢀ308C, 1 h, 92%; j) vinylmagnesium bromide (1.0m in THF, 1.6 equiv),
Et2O, 08C, 30 min, 78%; k) Ac2O (5.0 equiv), pyridine (10.0 equiv), DMAP (catalytic), CH2Cl2,
08C, 1 h, 94%; l) LDA (1.5 equiv), TBSCl (1.5 equiv), HMPA (1.5 equiv), THF, ꢀ78!258C, 72 h,
82%; m) MeOH (10.0 equiv), DCC (1.2 equiv), DMAP (0.1 equiv), CH2Cl2, 0!258C, 2 h, 86%;
n) Superhydride (1.0m in THF, 5.0 equiv), THF, ꢀ78!08C, 30 min, 96%; o) PivCl (3.0 equiv),
pyridine (10.0 equiv), DMAP (1.0 equiv), CH2Cl2, 0!258C, 12 h, 95%; p) TBAF (1.0m in THF,
2.0 equiv), THF, 0!258C, 3 h, 93%; p) (COCl)2 (5.0 equiv), DMSO (11 equiv), CH2Cl2, ꢀ788C,
1 h, ꢀ608C, 1 h; then Et3N (22 equiv), ꢀ78!ꢀ308C, 1 h, 89%; r) NaClO2 (4.0 equiv), NaH2PO4
(4.0 equiv), 2-methyl-but-2-ene (5.0 equiv), tBuOH/H2O (5:1), 258C, 2 h, 95%; s) PFPOH,
(1.2 equiv), DCC (1.5 equiv), CH2Cl2, 258C, 2 h, 82%. TBDPS=tert-butyldiphenylsilyl, Piv=piva-
loyl=trimethylacetyl, py=pyridine, DMAP=N,N-dimethyl-4-aminopyridine, NBS=N-bromosuc-
cinimide, 2,6-lut=2,6-lutidine, TFA=trifluoroacetic acid, KHMDS=potassium bis(trimethylsily-
l)amide, DIBAL-H=diisobutylaluminum hydride, DMSO=dimethyl sulfoxide, LDA=lithium dii-
sopropylamide, TBS=tert-butyldimethylsilyl, HMPA=hexamethylphosphoramide, DCC=1,3-
dicyclohexylcarbodiimide, Superhydride=lithium triethylborohydride, TBAF=tetra-n-butylammo-
nium fluoride, PFPOH=pentafluorophenol.
ꢀ
(which generated the desired C8 C9
unsaturation,95% yield) to afford 11.
3644
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2003, 42, 3643 –3648

Angewandte
Chemie
Table 1: Selected data for 3, 16, 30, 33, and 39.
With tetracycle 11 in hand,we then turned our attention to
the installation of the unsaturated C1–C5 carboxylic acid side
3: Rf =0.25 (silica gel, Et2O/hexanes 5:95); [a]D =+76.1 (CHCl3,
[12]
chain. To this end,an Ireland modification
of the Claisen
c=1.4); IR (film): n˜max =2930, 2856, 1632, 1470, 1383, 1085, 1006, 826,
1
775, 650 cmꢀ1; H NMR (500 MHz, CDCl3): d=5.10 (s, 1H), 4.91 (s,
rearrangement was used to ensure the introduction of the
desired trans double bond and the required four-carbon chain.
Thus,the pivaloate group was cleaved from 11 by exposure to
DIBAL-H (92% yield),and the resulting alcohol was treated
under Swern conditions to afford aldehyde 12 in 92% yield.
The latter compound, 12,was immediately treated with
vinylmagnesium bromide,and the resulting allylic alcohol
(78% yield) was converted into its acetate 13 (mixture of
diastereoisomers,92% yield). Acetate 13 was exposed to
LDA at ꢀ788C in the presence of TBSCl and HMPA,and the
mixture was then warmed to room temperature to provide the
corresponding g,d-unsaturated acid,which was subsequently
esterified (MeOH,DCC,DMAP),furnishing the targeted
methyl ester 14 in 71% overall yield for the two steps.
With the C1–C5 side chain in place,attention was then
focused on the further elaboration of the C20 terminus in
1H), 4.53 (d, J=11.4 Hz, 1H), 4.17 (d, J=15.0 Hz, 2H), 4.16 (d,
J=10.6 Hz, 1H), 2.97–2.91 (m, 1H), 2.86–2.80 (m, 3H), 2.18 (ddd,
J=12.2, 8.4, 4.0 Hz, 1H), 2.12–2.05 (m, 2H), 1.86–1.77 (m, 2H), 1.13
(d, J=7.0 Hz, 3H), 1.04 (s, 9H), 0.98(s, 9H), 0.84 ppm (d, J=7.0 Hz,
3H); 13C NMR (126 MHz, CDCl3): d=146.0, 114.0, 81.5, 67.4, 54.5, 37.0,
36.5, 36.3, 31.4, 30.6, 27.3, 26.9, 36.4, 21.4, 20.9, 18.2, 16.3 ppm; MS
(ESI): calcd for C21H40NO2S2SiClꢀ [M+Clꢀ]: 451, found: 451
16: Rf =0.24 (silica gel, CH2Cl2/MeOH 9:1); [a]D =ꢀ31.5 (CHCl3,
c=2.0); IR (film): n˜max =3436, 2960, 1727, 1456, 1396, 1320, 1285, 1231,
1
1160, 1090, 1026, 976, 878, 802, 732, 671 cmꢀ1; H NMR (600 MHz,
CDCl3): d=5.99 (ddd, J=9.9, 5.7, 2.2 Hz, 1H), 5.69 (d, J=5.7 Hz, 1H),
5.69–5.65 (m, 1H), 5.50 (dd, J=15.4, 6.1 Hz, 1H), 4.85 (dd, J=8.5,
7.9 Hz, 1H), 4.41 (ddd, J=11.8, 5.1, 4.4 Hz, 1H), 4.21 (m, 1H), 4.03
(dd, J=6.6, 6.4 Hz, 2H), 4.01 (m, 1H), 2.54 (dd, J=13.6, 7.4 Hz, 1H),
2.24–1.95 (m, 11H), 1.72–1.68(m, 2H), 1.50–1.46 (m, 1H), 1.19 (s,
9H), 0.91 ppm (d, J=6.6 Hz, 3H); 13C NMR (150 MHz, CDCl3):
d=178.6, 176.0, 131.2, 130.8, 128.8, 128.7, 111.4, 104.3, 77.6, 76.0, 75.4,
68.9, 63.6, 38.5, 35.6, 33.2, 30.7, 30.0, 28.7, 28.0, 27.2, 27.2, 23.5,
15.5 ppm; HRMS (MALDI): calcd for C26H38O8Na+ [M+Na+]: 501.2459,
found: 501.2467
ꢀ
preparation for the anticipated C20 C21 bond-forming
reaction. The first task was the reduction of methyl ester 14
(Superhydride,96% yield) and protection of the resulting
primary alcohol as a pivaloate ester (95% yield),a precau-
tionary maneuver undertaken to enable the pending dithiane
coupling step. The obtained intermediate 15 was then trans-
formed into carboxylic acid 16 (Table 1) by desilylation
(TBAF,93% yield),followed by a two-step oxidation
involving Swern conditions (89% yield) and NaClO2 (95%
yield). Finally,coupling of carboxylic acid 16 with pentafluor-
ophenol in the presence of DCC led to pentafluoroester 2 in
82% yield,a coupling partner which proved,as we shall see,
superior to its aldehyde progenitor.
30: Rf =0.45 (silica gel, EtOAc/hexanes 1:4); [a]D =ꢀ26.5 (CHCl3,
c=0.85); IR (film): n˜max =2955, 2914, 1745, 1697, 1460, 1392, 1350,
1253, 1170, 1067, 841, 741 cmꢀ1; 1H NMR (500 MHz, CDCl3): d=4.57
(m, 2H), 4.20 (dd, J=7.0, 4,4 Hz, 1H), 4.18–4.08 (m, 2H), 3.69 (ddd,
J=13.2, 4.0, 1.5 Hz, 1H), 3.03 (dd, J=13.0, 11.9 Hz, 1H), 2.79 (dd,
J=14.3, 7.3 Hz, 1H), 2.55 (m, 2H), 2.28–2.15 (m, 2H), 2.05 (ddd,
J=16.5, 10.3, 6.2 Hz, 1H), 1.89 (m, 1H), 1.68 (dt, J=13.9, 4.6 Hz, 1H),
1.58–1.42 (m, 2H), 1.29 (q, J=12.5 Hz, 1H), 1.09 (d, J=6.6 Hz, 3H),
1.02–0.95 (m, 2H), 0.94 (t, J=7.9 Hz, 9H), 0.80 (d, J=6.6 Hz, 3H), 0.79
(d, J=7.7 Hz, 3H), 0.59 (q, J=7.9 Hz, 6H), 0.04 ppm (s, 9H); 13C NMR
(126 MHz, CDCl3): d=172.3, 156.2, 96.6, 82.8, 76.3, 71.8, 63.1, 49.2,
43.6, 37.9, 37.8, 37.5, 31.1, 30.9, 24.0, 21.0, 18.6, 17.7, 16.7, 6.8, 4.6,
ꢀ1.5 ppm; HRMS (MALDI): calcd for C28H53NO6Si2Na+ [M+Na+]:
578.3303, found: 578.3303
The construction of the dithiane C21–C27 fragment 3
commenced from the previously synthesized lactone 17[13]
(Scheme 3). Reduction of 17 with DIBAL-H afforded the
corresponding lactol in equilibrium with its open-chain
hydroxy–aldehyde,which was trapped with 13,-propanedi-
thiol in the presence of BF3·OEt2 to furnish hydroxy dithiane
18 in 99% overall yield. Swern oxidation then led to aldehyde
19 (94% yield),which was coupled with vinyl iodide 21
(obtained from propargyl alcohol in two steps as shown in
Scheme 3) under the Nozaki–Hiyama–Kishi conditions[14]
(CrCl2,NiCl 2),leading to an epimeric mixture of allylic
alcohols 22 in 95% yield. The desired stereochemistry at C25
was then set by an oxidation (IBX,90% yield)–reduction
(Red-Al,80% yield) protocol,yielding the expected alcohol
23 (stereochemistry confirmed by NMR spectroscopic anal-
ysis),which was then desilylated by the action of TBAF to
give diol 24 (99% yield). Protection of the 1,3-diol system
within 24 as its silylene acetal (tBu2Si(OTf)2,26,-lut,75%
yield) then completed the synthesis of fragment 3 (Table 1).
The remaining coupling partner,C28–C40 fragment 4,
was constructed as outlined in Scheme 4. Thus,following our
previous studies,[9] azide 25 was reduced (10% Pd/C,H 2) and
the resulting primary amine was protected as its 2-(trime-
thylsilyl)ethoxy carbamate (Teoc) with reagent 26[15] in the
presence of Et3N to afford derivative 27 (80% overall yield),
setting the stage for the required spiroaminal formation. This
crucial cyclization was found to proceed in good and
33: Rf =0.40 (silica gel, EtOAc/hexanes 1:6); [a]D =+12.1 (CHCl3,
c=7.0); IR (film): n˜max =2955, 1695, 1459, 1394, 1351, 1254, 1218, 1171,
1
1132, 1065, 840, 757 cmꢀ1; H NMR (600 MHz, CDCl3): d=5.76–5.69
(m, 1H), 5.24 (d, J=17.1 Hz, 1H), 5.13 (d, J=17.1 Hz, 1H), 4.74 (m,
1H), 4.24 (d, J=5.3 Hz, 1H), 4.10–4.06 (m, 3H), 3.86 (d, J=6.1 Hz,
1H), 3.74–3.71 (m, 2H), 3.18(t, J=12.9 Hz, 1H), 2.98(dd, J=13.8,
9.0 Hz, 1H), 2.49 (dd, J=14.0, 5.7 Hz, 1H), 2.26 (dd, J=14.5, 6.1 Hz,
1H), 2.03–1.97 (m, 1H), 1.87 (dt, J=14.0, 5.7 Hz, 1H), 1.57–1.50 (m,
2H), 1.34–1.23 (m, 3H), 0.96 (d, J=5.7 Hz, 3H), 0.97–0.93 (m, 2H),
0.79 (d, J=6.1 Hz, 3H), 0.79 (d, J=6.1 Hz, 3H), 0.03 ppm (s, 9H);
13C NMR (150 MHz, CDCl3): d=157.1, 132.0, 120.5, 98.1, 97.9, 79.6,
75.0, 72.9, 63.8, 50.0, 49.8, 47.4, 41.0, 39.7, 37.5, 32.2, 31.8, 29.2, 25.7,
19.6, 18.8, 17.3, ꢀ0.5 ppm; HRMS (MALDI): calcd for C25H42INO5SiNa+
[M+Na+]: 614.1769, found: 614.1785
39: Rf =0.30 (silica gel, EtOAc/hexanes 1:1); [a]D =ꢀ9.8(CHCl 3,
c=0.2); IR (film): n˜max =3436, 2930, 1725, 1725, 1460, 1402, 1350, 1155,
1025, 979, 879, 585 cmꢀ1; 1H NMR (600 MHz, CDCl3): d=5.98(ddd,
J=9.8, 5.6, 2.1 Hz, 1H), 5.70 (ddd, J=9.8, 2.3, 1.0 Hz, 1H), 5.42 (s,
1H), 5.22 (s, 1H), 4.66 (dt, J=10.1, 6.1 Hz, 1H), 4.31 (d, J=6.1 Hz,
1H), 4.24–4.20 (m, 1H), 4.15 (m, 1H), 4.14–4.09 (m, 3H), 4.08–4.05 (m,
2H), 3.89 (m, 1H), 2.23–2.20 (m, 2H), 2.16–2.13 (m, 1H), 2.11–2.07 (m,
2H), 2.04–1.94 (m, 6H), 1.87–1.77 (m, 2H), 1.46–1.37 (m, 2H), 1.19 (s,
9H), 1.02 (d, J=6.6 Hz, 3H), 0.96 (d, J=6.6 Hz, 3H), 0.86 ppm (d,
J=6.6 Hz, 3H); 13C NMR (150 MHz, CDCl3): d=178.9, 146.4, 129.6,
128.1, 118.1, 112.0, 104.5, 100.4, 98.6, 81.5, 78.2, 75.9, 75.6, 73.7, 67.0,
66.9, 63.4, 38.7, 36.7, 36.1, 34.3, 33.7, 31.1, 27.6, 27.0, 24.6, 24.2, 23.8,
18.0, 16.5, 16.3 ppm; HRMS (MALDI): calcd for C33H47ClO10Na+
[M+Na+]: 661.2750, found: 661.2749
Angew. Chem. Int. Ed. 2003, 42, 3643 –3648
www.angewandte.org
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3645

Communications
azaspiracid-1 skeleton,and therefore an alternative tactic was
sought to achieve this objective. It was in the search of such a
protocol that the model substrate 32 was synthesized from 31
by reaction with allyltri-n-butyltin in the presence of
[21]
[Pd2dba3],TFP,
and LiCl (95% yield). Upon considerable
experimentation,it was found that selective removal of the
TES group from 32 under the mild conditions provided by
HF·py buffered with additional pyridine led to the generation
of the desired hydroxy enol ether precursor,which upon
Scheme 3. Construction of 3: a) DIBAL-H (1.0m in CH2Cl2, 1.1 equiv),
CH2Cl2, ꢀ788C, 1.5 h; b) 1,3-propanedithiol (1.1 equiv), BF3·OEt2 (1.5 equiv),
CH2Cl2, 08C, 1 h, 99% over two steps; c) (COCl)2 (1.2 equiv), DMSO
(2.4 equiv), CH2Cl2, ꢀ788C, 30 min; then Et3N (5.0 equiv), ꢀ78!ꢀ208C,
94%; d) TMSCl (1.2 equiv), NaI (1.2 equiv), H2O (0.6 equiv), CH3CN, 0!
258C, 1.5 h, 51%; e) TBSCl (1.2 equiv), imidazole (2.5 equiv), DMF, 258C,
36 h, 96%; f) NiCl2 (0.02 equiv), CrCl2 (4.0 equiv), DMF, 08C; then 19
(1.0 equiv), 21 (2.5 equiv), 0!258C, 15 h, 95%; g) IBX (2.0 equiv), DMSO/
THF (4:1), 258C, 2 h, 90%; h) Red-Al (2.5 equiv), toluene, ꢀ788C, 1 h, 80%;
i) TBAF (2.2 equiv), THF, 258C, 1 h, 99%; j) tBu2Si(OTf)2 (1.6 equiv), 2,6-luti-
dine (4.0 equiv), CH2Cl2, ꢀ308C, 30 min, 75%. TMS=trimethylsilyl, imid=
imidazole, DMF=N,N-dimethylformamide, IBX=o-iodoxybenzoic acid, Red-
Al=sodium bis(2-methoxyethoxy)aluminum hydride, Tf=trifluoromethane-
sulfonyl.
reproducible yields in the presence of Yb(OTf)3,[16] in
[17]
contrast to the previously employed BF3·OEt2 procedure,
which proved to be unreliable. An oxidation–reduction
sequence was then carried out to invert the stereochemistry
at C34. Thus,exposure of benzoate lactone 28 to DIBAL-H
resulted in the formation of the corresponding hydroxylactol,
which was oxidized stepwise back to the hydroxylactone
(NIS,TBAI, [18] 88% overall yield) and then to ketolactone 29
(DMP,95% yield). Treatment of ketolactone 29 with L-
Selectride at ꢀ788C furnished,exclusively,the a-hydroxy
compound (85% yield),which was silylated (TESOTf,26, -
lutidine,86% yield) to afford the desired intermediate 30
(Table 1). In preparation for the projected Stille coupling,
lactone 30 was efficiently converted into vinyl stannane 4
through the intermediacy of enol triflate 31 (KMHDS, 10,[10]
Scheme 4. Construction of 4 and model FGHI ring system 34: a) Pd/C
(10% w/w, 0.1 equiv), EtOAc, H2, 258C, 5 h; then filtration through celite;
then Et3N (5.0 equiv), 26 (4.0 equiv), 258C, 12 h, 80%; b) Yb(OTf)3
(0.1 equiv), CH3CN, 258C, 3 min, 71%; c) DIBAL-H (1.0m in THF,
4.0 equiv), toluene, ꢀ788C, 25 min; d) NIS (10.0 equiv), TBAI (2.0 equiv),
CH2Cl2, 258C, 40 min, 88% over two steps; e) DMP (1.5 equiv), pyridine
(12.5 equiv), CH2Cl2, 0!258C, 30 min, 95%; f) L-Selectride (1.0m in THF,
1.8equiv), THF, ꢀ788C, 20 min, 85%; g) TESOTf (1.1 equiv), 2,6-lutidine
(1.5 equiv), ꢀ788C, 10 min, 86%; h) 10 (5.0 equiv), KHMDS (5.0 equiv),
THF, ꢀ788C, 40 min, 91%; i) (Me3Sn)2 (10.0 equiv), TFP (0.5 equiv), LiCl
(3.0 equiv), [Pd2dba3] (0.1 equiv), THF, 258C, 1 h, 98%; j) allyltri-n-butyltin
(10.0 equiv), TFP (0.5 equiv), LiCl (3.0 equiv), [Pd2dba3] (0.1 equiv), THF,
258C, 1 h, 95%; k) HF·py (5.0 equiv), THF/pyridine (1:1), 0!258C, 2 h;
l) NIS (10.0 equiv), NaHCO3 (30.0 equiv), THF, 08C, 12 h, 75% over two
steps; m) Et3B (1.0m in hexanes, catalytic), nBu3SnH/toluene (1:2), 08C,
5 min, 94%. Bz=benzoyl, NIS=N-iodosuccinimide, TBAI=tetra-n-buty-
lammonium iodide, DMP=Dess–Martin periodinane, L-Selectride=lithium
tri-sec-butylborohydride, TFP=tri-2-furylphosphane, dba=dibenzylideneace-
tone.
91% yield; (Me3Sn)2,[Pd dba3],[19] 98% yield). Although
2
triflate 31 itself can potentially serve as a coupling partner
with an allyl stannane,model studies suggested the reverse
partnership as the most efficient variant for the azaspiracid-1
framework—and thus the choice of 4 as the FHI coupling
partner. At this juncture,we needed further improvements
of our approach to the final steps for the casting of the FGHI
ring system. Specifically,our previously developed protocol
[20]
for the formation of ring G employing Hg(OAc)2 proved
problematic in the presence of the entire,highly oxygenated
3646
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2003, 42, 3643 –3648

Angewandte
Chemie
treatment with N-iodosuccinimide[22] in the presence of
NaHCO3 furnished the FGHI iodide 33 (Table 1) in a
pleasing yield of 75% overall for the two steps. The formation
of 33 as a single,crystalline product gave us the opportunity to
confirm its stereochemistry by X-ray crystallographic analysis
[23]
(see ORTEP drawing of ent-33,Figure 1)
and to test the
next step,that of removing the facilitating,but now extra-
neous,iodine residue,without compromising the rest of the
molecule. Indeed,this crucial operation ( 33!34) was suc-
cessfully executed by the reductive rupture of the carbon–
iodine bond by the mild action of nBu3SnH and Et3B[24] (94%
yield).
Having chartered an appropriate route to access the
FGHI domain of azaspiracid-1,we then turned our attention
to scouting the ground for a possible means to construct the
Figure 1. ORTEP drawing of compound ent-33.
ꢀ
demanding C20 C21 bond and completing the requisite
ABCDE domain. Thus,we targeted advanced model system
39 (Scheme 5),which contains all the rings and stereocenters
embedded in the C5–C27 region of structure 1. It was after the
recognition of the reluctance of the aldehyde functionality at
C20 of various ABCD-domain substrates to participate as an
electrophilic partner in coupling reactions with organometal-
Scheme 5. Model dithiane coupling and structural confirmation of all stereocenters in ABCDE ring system (39, C5–C27 domain): a) TBAF (1.0m
in THF, 5.0 equiv), THF, 258C, 2 h, 88%; b) (COCl)2 (5.0 equiv), DMSO (11 equiv), CH2Cl2, ꢀ788C, 30 min; then Et3N (22 equiv), ꢀ78!08C,
94%; c) NaClO2 (6.0 equiv), NaH2PO4 (6.0 equiv), 2-methyl-but-2-ene (excess), tBuOH/H2O (4:1), 258C, 1.5 h; d) PFPOH (1.5 equiv), DCC
(2.0 equiv), 258C, 2.5 h, 56% over two steps; e) 4 (9.0 equiv), nBuLi–nBu2Mg (1.1m in hexanes, 6.0 equiv), THF, 0!258C, 1.5 h; then 35, ꢀ908C,
15 min, 63%; f) DIBAL-H (1.0m in CH2Cl2, 10.0 equiv), CH2Cl2, ꢀ908C, 1.5 h, 55%; g) PivCl (3.0 equiv), pyridine (10.0 equiv), 0!258C, 12 h,
75%; h) PhI(OCOCF3)2 (2.2 equiv), MeCN/pH7 buffer (4:1), 08C, 78%; i) TBAF (1.0 m in THF, 5.0 equiv), THF, 258C, 16 h, 85%; j) triphosgene
(2.0 equiv), pyridine (15 equiv), CH2Cl2, ꢀ78!258C, 1 h, 54%.
Angew. Chem. Int. Ed. 2003, 42, 3643 –3648
www.angewandte.org
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3647

Communications
[3] E. Ito,M. Satake,K. Ofuji,M. Higashi,K. Harigaya,T.
McMahon,T. Yasumoto, Toxicon 2002, 40,193 – 203.
[4] For synthetic studies towards 1,see: a) R. G. Carter,D. J.
Weldon, Org. Lett. 2000, 2,3913 – 3916; b) R. G. Carter,D. E.
Graves, Tetrahedron Lett. 2001, 42,6035 – 6039; c) R. G. Carter,
T. C. Bourland,D. E. Graves, Org. Lett. 2002, 4,2177 – 2179;
d) C. J. Forsyth,J. Hao,J. Aiguade, Angew. Chem. 2001, 113,
3775 – 3779; Angew. Chem. Int. Ed. 2001, 40,3663 – 3667;
e) A. B. Dounay,C. J. Forsyth, Org. Lett. 2001, 3,975 – 978;
f) J. Auguade,J. Hao,C. J. Forsyth, Tetrahedron Lett. 2001, 42,
817 – 820; g) J. Hao,J. Aiguade,C. J. Forsyth, Tetrahedron Lett.
2001, 42,821 – 824.
[5] See the following Communication in this issue: K. C. Nicolaou,
D. Y.-K. Chen,Y. Li,W. Qian,T. Ling,S. Vyskocil,T. V. Koftis,
M. Govindasamy,N. Uesaka, Angew. Chem. 2003, 115,3777–
3781; Angew. Chem. Int. Ed. 2003, 42,3649–3653.
[6] E. J. Corey,D. Seebach, Angew. Chem. 1965, 77,1134 – 1135;
Angew. Chem. Int. Ed. Engl. 1965, 4,1075 – 1077.
[7] L. Del Valle,J. K. Stille,L. S. Hegedus, J. Org. Chem. 1990, 55,
3019 – 3023.
[8] K. C. Nicolaou,W. Qian,F. Bernal,N. Uesaka,P. M. Pihko,J.
Hinrichs, Angew. Chem. 2001, 113,4192 – 4195; Angew. Chem.
Int. Ed. 2001, 40,4068 – 4071.
[9] K. C. Nicolaou,P. M. Pihko,N. Diedrichs,N. Zou,F. Bernal,
Angew. Chem. 2001, 113,1302 – 1305; Angew. Chem. Int. Ed.
2001, 40,1262 – 1265.
[10] K. Tsushima,K. Araki,A. Murai, Chem. Lett. 1989,1313 – 1316.
[11] J. K. Stille, Angew. Chem. 1986, 98,504 – 519; Angew. Chem. Int.
Ed. Engl. 1986, 25,508 – 524.
lic species (such as the lithiated derivative of dithiane 4) that
we focused on the corresponding pentafluorophenyl ester as a
potential alternative for this pivotal coupling. Thus,silyl
derivative 11 (obtained as described in Scheme 2) was
desilylated by treatment with TBAF (88% yield),and the
resulting alcohol was oxidized stepwise,first under Swern
conditions (94% yield) and then with NaClO2,to afford
carboxylic acid, 34,as shown in Scheme 5. Subsequent
coupling of the latter compound 34 with pentafluorophenol
was accomplished in the presence of DCC in 56% overall
yield from the corresponding aldehyde. The rather labile
pentafluorophenyl ester 35 was carefully purified by flash
column chromatography and immediately employed in the
coupling reaction with the lithiated form of dithiane 4
(nBuLi–nBu2Mg)[25] at ꢀ908C to afford ketone 36 in 63%
yield. The required stereoselective reduction of the C20
carbonyl group was then accomplished through the use of
DIBAL-H. However,the pivaloate group at C1 was cleaved
concomitantly,leading to the corresponding diol in 55%
yield. This diol was reprotected selectively (PivCl,py,75%
yield) at the primary position,leading to advanced inter-
mediate 37. In a carefully orchestrated sequence,the dithiane
moiety was cleaved from 37 through the action of PhI(O-
COCF3)2[26] to liberate the corresponding ketone (78% yield),
from which the silicon tether was removed by exposure to
TBAF,furnishing the azaspiracid-1-like lactol 38 through
spontaneous ring closure as a single isomer (85% yield). With
the aim of imparting more structural rigidity to the ABCDE
truncate of azaspiracid-1 for detailed NMR studies,the
hydroxy groups at C20 and C21 were conformationally
constrained within a cyclic carbonate upon exposure to
triphosgene and pyridine to provide cyclic carbonate 39
(54% yield) in which the terminal position was now occupied
by a chlorine residue. Despite this unexpected occurrence,
however,chloride 39 proved quite useful in providing us with
the required structural information (Table 1). Thus,extensive
NOE studies of compound 39 (see Scheme 5) revealed the
key spatial correlations for the assigned stereochemistry,
particularly around the C20 stereocenter.
[12] R. E. Ireland,R. H. Mueller,A. K. Willard, J. Am. Chem. Soc.
1976, 98,2868 – 2877.
[13] D. B. Collum,J. H. McDonald,W. C. Still, J. Am. Chem. Soc.
1980, 102,2118 – 2120.
[14] a) K. Takai,M. Tagashira,T. Kuroda,K. Oshima,K. Utimoto,H.
Nozaki, J. Am. Chem. Soc. 1986, 108,6648 – 6050; b) H. Jin,J.
Uenishi,W. J. Christ,Y. Kishi, J. Am. Chem. Soc. 1986, 108,
5644 – 5646.
[15] A. Rosowsky,J. E. Wright, J. Org. Chem. 1983, 48,1539 – 1541.
[16] S. Kobayashi,M. Sugiura,M. Kitagawa,W. Lam,
Chem. Rev.
2002, 102,2227 – 2302.
[17] A. M. P. Koskinen,L. A. Otsomaa, Tetrahedron 1997, 53,6473 –
6484.
[18] S. Hanessian,D. H. Wong,M. Therien, Synthesis 1981,394 – 396.
[19] C. Barker,K. Jarowicki,P. Kocienski, Synlett 1991,197 – 198.
[20] a) H. C. Brown,P. J. Geoghegan, J. Org. Chem. 1970, 35,1844 –
1850; b) S. J. Danishefsky,W. H. Pearson, J. Org. Chem. 1983, 48,
3865 – 3866.
[21] V. Farina,B. Krishman, J. Am. Chem. Soc. 1991, 113,9585 – 9595.
[22] a) M. Adinolfi,M. Parrilli,G. Barone,G. Laonigro,L. Mangoni,
Tetrahedron Lett. 1976,3661 – 3662; b) G. Haaima,R. T. Wea-
vers, Tetrahedron Lett. 1988, 29,1085 – 1088.
[23] CCDC-210122 (ent-33) contains the supplementary crystallo-
graphic data for this paper. These data can be obtained free of
charge via www.ccdc.cam.ac.uk/conts/retrieving.html (or from
the Cambridge Crystallographic Data Centre,12 Union Road,
Cambridge CB21EZ,UK; fax: ( + 44)1223-336-033; or deposit@
ccdc.cam.ac.uk).
The described chemistry provided ample quantities of
building blocks 2, 3,and 4,and powerful synthetic technol-
ogies pertaining to their coupling as a prelude to further
forays along the designed path to the proposed structure 1 for
azaspiracid-1. In the following Communication,[5] we report
the final drive to two diastereoisomers of this structure and
their comparison with the natural substance.
Received: May 7,2003 [Z51825]
Keywords: asymmetric synthesis · azaspiracid-1 · natural
.
[24] K. Miura,Y. Ichinose,K. Nozaki,K. Fugami,K. Oshima,K.
Utimoto, Bull. Chem. Soc. Jpn. 1989, 62,143 – 147.
[25] M. Ide,M. Nakata, Bull. Chem. Soc. Jpn. 1999, 72,2491 – 2499.
[26] G. Stork,K. Zhao, Tetrahedron Lett. 1989, 30,287 – 2902.
products · neurotoxins · structural revision · total synthesis
[1] T. MacMahon,J. Silke, Harmful Algae News 1996, 14,2.
[2] M. Satake,K. Ofuji,H. Naoki,K. J. James,A. Fruey,T.
McMahon,J. Silke,T. Yasumoto, J. Am. Chem. Soc. 1998, 120,
9967 – 9968; see also: Y. Romµn,A. Alfonso,M. C. Louzao,
L. A. de la Rosa,F. Leira,J. M. Vieties,M. R. Vieytes,K. Ofuji,
M. Satake,T. Yasumoto,L. M. Botana, Cellular Signalling 2002,
14,703 – 716.
3648
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2003, 42, 3643 –3648