Total synthesis of (-)- and (+)-decarbamoyloxysaxitoxin and (+-)-saxitoxin

Article abstract of DOI:10.1002/asia.200800382

Enantioselective total syntheses of (-)- and (-1-)-decarbamoyloxysaxitoxin (doSTX) and (-l-)-saxitoxin (STX) were achieved. The characteristic spiro-fused cyclic guanidine structure of STX was constructed by oxidation at the C4 position with IBX via an α-iminium carbonyl intermediate and acid-promoted cyclization of guanidine at the C5 position. A second-generation methodology was developed for the synthesis of STX, featuring discriminative reduction of the nitro group and N-O bond in nitroisoxazolidine. This approach provides efficient access to the key diamine intermediate for STXs.

Full text of DOI:10.1002/asia.200800382

DOI: 10.1002/asia.200800382
Total Synthesis of (ꢀ)- and (+)-Decarbamoyloxysaxitoxin and (+)-Saxitoxin
Osamu Iwamoto, Ryoko Shinohara, and Kazuo Nagasawa*[a]
Dedicated to Professor Tadashi Nakata on the occasion of his 65th birthday
Abstract: Enantioselective total syn-
theses of (ꢀ)- and (+)-decarbamoylox-
ysaxitoxin (doSTX) and (+)-saxitoxin
(STX) were achieved. The characteris-
tic spiro-fused cyclic guanidine struc-
ture of STX was constructed by oxida-
tion at the C4 position with IBX via an
a-iminium carbonyl intermediate and
acid-promoted cyclization of guanidine
at the C5 position. A second-genera-
tion methodology was developed for
the synthesis of STX, featuring discrim-
inative reduction of the nitro group
ꢀ
and N O bond in nitroisoxazolidine.
This approach provides efficient access
to the key diamine intermediate for
STXs.
Keywords: 1,3-dipolar
cycloaddi-
tion · guanidine · oxidation · natural
products · total synthesis
Introduction
The voltage-dependent sodium channel plays a critical role
in depolarization and conduction in most excitable cells.[1]
Saxitoxin (1; STX; Figure 1), isolated from poisonous shell-
fish,[2] is a neurotoxin that blocks ion influx through the volt-
age-dependent sodium channel by binding to the site-I
region.[3] Many natural STX analogues have been reported,
and each of them shows characteristic biological activity to-
wards the sodium channel.[4] The key structural features of
STXs lie in the bis(guanidine) functional group that con-
structs the compact spiro- and fused-ring system, and a
ketone hydrate stabilized by the neighboring guanidine
group. The complicated structure of STX and its biological
importance as a pharmacological tool for ion channel stud-
ies have greatly attracted synthetic chemists. The first total
synthesis of (+/ꢀ)-1 was reported by Kishi and co-workers
in 1977.[5] Jacobi and co-workers also synthesized (+/ꢀ)-1 by
the intramolecular 1,3-dipolar cycloaddition of azomethine
imine to an imidazolone.[6] Recently, Du Bois and co-work-
Figure 1. Structures of STX (1) and its analogues.
ers succeeded in the synthesis of (+)-STX (1) by using cata-
by Kishi and Du Bois, respectively.[8] We are interested in
the development of isoform-selective sodium channel inhibi-
tors and therefore started a project to synthesize STX deriv-
atives.[9] We recently reported a synthesis of the decarba-
moyloxy analogue of 1, (ꢀ)-decarbamoyloxysaxitoxin
(doSTX, ent-2), a putative unnatural form, by utilizing a-
iminium carbonyl formation with o-iodoxybenzoic acid
(IBX).[10] Herein, we present full details of our synthetic
studies on (ꢀ)-doSTX (2), and synthesis of the natural form
of (+)-2. We also report a synthesis of (+)-STX (1) based
upon our second-generation strategy for STXs.
[7]
ꢀ
lytic C H amination methodology. Other natural STX an-
alogues of (ꢀ)-dcSTX and (+)-GTX3 were also synthesized
[a] O. Iwamoto, R. Shinohara, Prof. Dr. K. Nagasawa
Department of Biotechnology and Life Science
Tokyo University of Agriculture and Technology
2-24-16 Naka-cho, Koganei, Tokyo 184-8588 (Japan)
Fax : (+81)42-388-7295
E-mail: knaga@cc.tuat.ac.jp
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/asia.200800382.
Chem. Asian J. 2009, 4, 277 – 285
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
277

FULL PAPERS
Results and Discussion
Synthesis of (ꢀ)- and (+)-doSTX (2)
Decarbamoyloxysaxitoxin (doSTX; 2) was isolated from the
Australian shellfish Gymnodinium catenatum in 1990.[11] The
structure of doSTX (2) contains the common skeleton of all
the natural STX analogues. Although the absolute stereo-
chemistry of 2 has not yet been reported, it seems likely to
be the same as that of STX (1). Our synthetic plan is illus-
trated in Scheme 1. The stereochemistry at C5 and C6 in 2
Scheme 2. Synthesis of bis-monocyclic guanidine alcohol 14. Reagents
and conditions: a) toluene, 808C (93%); b) LiOH, THF/H2O, 0 8C;
c) (COCl)2, DMF (cat.), toluene; d) NaN3, acetone/H2O, 0 8C; e) 1,4-diox-
ane, 1008C, then 10% HCl, then (Boc)2O, K2CO3, (86% from 6); f) H2,
Pd(OH)2/C, MeOH; g) NCbz=C
h) DEAD, Ph3P, toluene (97% three steps); i) TFA, CH2Cl2; j) NBoc=C-
AHCTUNGTRENGUN(N SMe)NHBoc (16), HgCl2, Et3N, DMF (77% two steps); k) TBAF, THF
ACHTUNGTREN(UNGN SMe)NHCbz (15), HgCl2, Et3N, DMF;
(92%). Boc=tert-butyloxycarbonyl, Cbz=benzyloxycarbonyl, DEAD=
diethyl azodicaboxylate, TBAF=tetrabutylammonium fluoride, TFA=
trifluoroacetic acid.
Scheme 1. Retrosynthetic analysis for (ꢀ)-doSTX (ent-2). PG=protect-
ing group, TIPS=triisopropylsilyl.
room temperature, the corresponding carboxylic acid 9 was
obtained as a mixture with its C5 epimer (1:1). The carbox-
ylic acid 9 was then converted into the N-Boc-protected
amine 10 by Curtius rearrangement reaction: the carboxylic
acid was treated with oxalyl chloride in toluene, and the re-
sulting acid chloride was treated with sodium azide, fol-
lowed by heating at 1008C in dioxane to give the corre-
sponding amine, which was protected with Boc to give N-
Boc-protected amine 10 in 86% yield. Direct protection of
the amino group was attempted with addition of tBuOH to
the isocyanate intermediate, but no reaction occurred. The
are controlled by the chiral center in 7,[12] which corresponds
to the ketone hydrate at C12 in 2, by the intermolecular 1,3-
dipolar cycloaddition reaction of 7 with methyl crotonate
(8). Construction of the characteristic bis-bicyclic guanidine
structure is a challenging issue in the synthesis of STXs. We
planned to cyclize the guanidine at C4 to generate the skele-
ton of 2 via the a-iminium cation 3, which could be formed
by the oxidation of ketone 4.
The bis-monocyclic guanidine 14 was synthesized as fol-
lows (Scheme 2): The 1,3-dipolar cycloaddition reaction of
chiral nitrone 7 with methyl crotonate (8) gave the isoxazoli-
dine 6 stereoselectively in 93% yield.[13] Isomerization of
the pseudoaxially oriented stereochemistry at C5 and subse-
quent hydrolysis of the ester group in 6 proceeded upon
treatment with lithium hydroxide in aqueous THF at 08C to
give the carboxylic acid 9. When this reaction was run at
ꢀ
reductive cleavage of the N O bond of 10 was conducted
with hydrogen in the presence of Pd(OH)2, and a guanidine
group was introduced into the resulting pyrrolidine deriva-
tive with bis(benzyloxycarbonyl)-2-methyl-2-thiopseudourea
(15) and mercury(II) chloride to give guanidine 11.[14] Cycli-
zation of the guanidine group was conducted under the Mit-
sunobu reaction conditions with DEAD and triphenylphos-
phine to give cyclic guanidine 12 in 97% yield from 10.[15]
Deprotection of the Boc group of 12 with TFA, followed by
reaction with bis(tert-butyloxycarbonyl)-2-methyl-2-thio-
pseudourea (16) in the presence of mercury(II) chloride,
and deprotection of the TIPS ether with TBAF, gave the
bis-monocyclic guanidine alcohol 14 in 71% yield from 12.
After the bis-monocyclic guanidine alcohol 14 was ob-
tained, cyclization of the guanidine at C4 through a-iminium
carbonyl formation was examined by oxidation (Scheme 3).
Prior to this conversion, oxidation of the alcohol at C12 was
examined. We tried various oxidants, for example, chromi-
um oxidants, 2,2,6,6-tetramethyl-l-piperidinoxyl (TEMPO),
tetra-n-propylammonium perruthenate (TPAP), and Dess–
Martin periodinane (DMP), but no reaction occurred, and
Abstract in Japanese:
278
www.chemasianj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Asian J. 2009, 4, 277 – 285

(ꢀ)- and (+)-Decarbamoyloxysaxitoxin and (+)-Saxitoxin
This unexpected formation of the bicyclic guanidine
moiety in 22 was interpreted as follows (Scheme 4): After
oxidation of the alcohol at C12 with IBX, the resulting
ketone 21 simultaneously reacted with another molecule of
IBX to form the enol intermediate A. Electron flow in this
intermediate would take place to generate our expected a-
iminium carbonyl cation, but the hydroxy group in IBX
would attack intramolecularly at C4 (A to B), instead of the
amino group in guanidine, to generate the fused-type bicy-
clic guanidine unit in 22.[17,18] X-ray analysis of 22 revealed
ꢀ
that the hydroxy group at C4 and the C N bond at C12 are
in an antiperiplanar orientation. Thus, conversion of 22 into
ent-2 was expected by a pinacol-type rearrangement reac-
tion. Removal of the Boc and Cbz protecting groups by
treatment with TFA and subsequent reduction with hydro-
gen in the presence of Pd(OH)2 afforded fused-type guani-
dine 23, the conformational structure of which was con-
Scheme 3. Oxidation of C12 and C4. Reagents and conditions: a) 5%
TFA/CH2Cl2 (14: 13%, 19: 71%, 20: 16%); b) IBX (4 equiv), DMSO,
708C, (45%). IBX=o-iodoxybenzoic acid.
1
firmed to be similar to that of 22 by H NMR spectroscopy.
alcohol 14 was recovered quantitatively. Swern oxidation af-
forded enone 18 by elimination of the guanidine group at
C5. We assumed that the bulky guanidine group with the
two Boc protecting groups prevented the approach of the
oxidants. Moreover, the character of the Boc group as an
electron-withdrawing group promotes undesired b-elimina-
tion at the C5 guanidine group. Thus, deprotection of one of
the Boc groups in 14 was examined. After many attempts,
mono-Boc 19 was obtained in 71% yield with 5% TFA in
dichloromethane, and 20 was obtained in 16% yield. Even
with the mono-Boc 19, various oxidants were ineffective for
the oxidation at C12.[16] However, IBX was an exception.
Treatment of the alcohol 19 with 4 equivalents of IBX in
DMSO at 708C did not give the desired ketone 21, but an
unexpected fused-type guanidine 22 was obtained in 45%
yield as the sole product, the structure of which was con-
firmed by means of X-ray analysis (Scheme 4).[17]
Exposure of 22 to various acidic conditions (e.g., aq. H2SO4,
aq. 7.5n HCl, 1008C, under vacuum at 1808C,[19] or AcOH/
TFA, 508C[5a]) unfortunately resulted in no reaction.
Although the desired bicyclic guanidine formation from
19 and 23 failed, we were pleased to find that the oxidation
of ketone 21 occurred to generate the a-iminium carbonyl
compound and a hydroxy group could be installed selective-
ly at C4. Therefore, we decided to apply this IBX oxidation
reaction to 24, prior to introduction of the C5 guanidine
group (Scheme 5). Alcohol 24 was obtained by deprotection
Scheme 5. Synthesis of aminal 25. Reagents and conditions: a) TBAF,
THF, 08C (94%); b) IBX (4 equiv), DMSO, 708C (25: 28%, 26: 29%,
24: 30%); c) (COCl)2 (10 equiv), DMSO (12 equiv), Et3N (30 equiv),
CH2Cl2, ꢀ788C; d) IBX (1.1 equiv), DMSO, 508C (25: 64%, 26: 0% two
steps).
of the TIPS ether of 12 with TBAF at 08C in 94% yield.
IBX oxidation (4 equiv) of alcohol 24 in DMSO at 708C
gave the aminal 25 in 28% yield, together with a further
oxidized product 26 (29% yield). Interestingly, ketone 27,
an intermediate for 25, was not observed in this reaction,
despite recovery of the starting material 24 in 30% yield. It
seemed that the reactivity of ketone 27 with IBX is signifi-
cantly higher than that of alcohol 24 or aminal 25. Stepwise
oxidation of 24 was examined to obtain the desired aminal
25 selectively. Thus, oxidation of alcohol 24 was employed
with a large excess (10 equiv) of Swern reagent to give 25.[20]
The resulting ketone was treated with a limited amount of
Scheme 4. Proposed reaction mechanism of IBX oxidation and attempts
at pinacol-type rearrangement of 23. Reagents and conditions: a) TFA,
CH2Cl2; b) H2, Pd(OH)2/C, MeOH; then HCl aq. (60% two steps).
IBA=o-iodosobenzoic acid.
Chem. Asian J. 2009, 4, 277 – 285
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemasianj.org
279

K. Nagasawa et al.
FULL PAPERS
IBX reagent (1.1 equiv) at 508C to give aminal 25 in 64%
yield as a sole product.
The total synthesis of (ꢀ)-doSTX (ent-2) was achieved
from 25 as follows (Scheme 6): Treatment of 25 with sodium
borohydride at 08C diastereoselectively gave the alcohol,
Scheme 7. Synthetic plan for (+)-STX (1) through discriminative diamine
synthesis. [N]=nitrogen-containing functional group.
Scheme 6. Completion of the synthesis of (ꢀ)-doSTX (ent-2). Reagents
and conditions: a) NaBH4, MeOH, 08C (72%); b) TFA, CH2Cl2;
c) NCbz=CACHTUNGTRENNUNG(SMe)NHCbz (16), HgCl2, Et3N, DMF (82% two steps);
d) H2, Pd(OH)2/C, MeOH/EtOAc (2:1); e) TFA, 508C (60% two steps);
f) DMSO, diisopropylcarbodiimide, pyridinium trifluoroacetate (63%).
N7 was planned by the use of the 1,3-dipolar cycloaddition
reaction of nitrone ent-7 with nitroalkene 34.[21] There are
some reports concerning this type of 1,3-dipolar cycloaddi-
tion reaction, although synthetic applications involving con-
ꢀ
the Boc protecting group of which was removed with TFA,
and a Cbz-protected guanidine group was introduced to give
29.[17] Removal of the four Cbz protecting groups with hy-
drogen in the presence of Pd(OH)2 and subsequent treat-
ment with TFA at 508C gave the decarbamoyloxysaxitoxinol
31 in 60% overall yield. Finally, oxidation of the alcohol
with dimethylsulfoxide and diisopropylcarbodiimide afford-
ed (ꢀ)-doSTX (ent-2) in 63% yield.[4b] All the spectroscopic
data of the synthetic material were consistent with the re-
ported data for the natural product. The putative natural
form of (+)-doSTX (2) was also synthesized from (ꢀ)-ni-
trone (ent-7) and 8 similarly to (ꢀ)-doSTX (ent-2). The opti-
cal rotation values of (ꢀ)-doSTX (ent-2) and (+)-doSTX (2)
were ꢀ23.3 (c=0.2, MeOH) and +20.0 (c=0.3, MeOH), re-
spectively.
version of the resulting nitro and N O groups in isoxazoli-
dine into amino groups were not feasible because of the dif-
ficulty of differential reduction of these functional groups.[22]
Therefore, this challenging transformation, that is, discrimi-
native diamine synthesis [see Eq. (1) in Scheme 7], was fur-
ther examined.
For the synthesis of nitroisoxazolidine 36, 1,3-dipolar cy-
cloaddition reaction of nitroalkene 34 with nitrone ent-7 was
employed (Scheme 8). The reaction proceeded under sol-
Scheme 8. Synthesis of nitroisoxazolidine 36. Reagents and conditions:
a) No solvent, 408C, 30 min (95%); b) MeOH, MeONa, 08C (35:36=
5:95).
Synthesis of (+)-STX (1)
As described above, the synthesis of doSTX (2) was ach-
ieved by the 1,3-dipolar cycloaddition reaction of optically
active nitrone 7 and 8, and the direct oxidation at C4 of 24
with IBX via an a-iminium carbonyl intermediate. In this
synthesis, the introduction of the nitrogen functional group
at C5 of the 1,3-dipolar cycloaddition adduct 6 through Cur-
tius rearrangement was tedious. Therefore, we developed a
second-generation synthesis of natural (+)-STX (1), focus-
ing on the direct introduction of an amino group at C5. The
synthetic strategy is depicted in Scheme 7. (+)-STX (1) was
expected to be obtained by introducing the guanidine group
into the diamine 32 followed by cyclization of the resulting
bis(guanidine) at C4 and C6, according to our previously de-
veloped route for STXs. In this synthesis, efficient prepara-
tion of the key diamine 32 makes the synthetic route more
practical. For effective access to this key intermediate, direct
and stereoselective introduction of amino groups at N3 and
vent-free conditions at 408C within 30 minutes to give the
endo adduct 35 in 95% yield as a single diastereomer. The
stereochemistry at C5 was epimerized with sodium methox-
ide in methanol at 08C to give 36 in a ratio of 95:5.
Next, selective reduction of the nitro group of 36 was ex-
amined (Table 1).[23] In the case of reaction with NiCl2/
ꢀ
NaBH4, both the nitro group and N O bond in isoxazolidine
were reduced to give diamine 37 (Table 1, entry 1). In con-
trast, reaction of 36 with hydrogen in the presence of Raney
Ni catalyst gave a mixture of diamine 39 and hydroxylamine
38 (Table 1, entry 2). Milder reduction conditions with Lin-
dlar catalyst gave the desired hydroxylamine 37 together
with a mixture of 38 and 39 (Table 1, entry 3). Selective re-
duction of the nitro group in 36 was achieved with zinc
280
www.chemasianj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Asian J. 2009, 4, 277 – 285

(ꢀ)- and (+)-Decarbamoyloxysaxitoxin and (+)-Saxitoxin
Table 1. Reduction of NO2 group of 36.[a]
(Table 2, entry 2). In the case of NiCl2/NaBH4 or CoCl2/
NaBH4, a mixture of 45 and the desired 44 was generated in
very low yield (Table 2, entries 3 and 4). Interestingly, Mill-
erꢁs conditions, namely, TiCl3 and NaOAc in MeOH/aq.
Entry
Conditions
Product
1
2
3
4
5
NiCl2, NaBH4, RT
Raney Ni, H2, RT
Lindlar cat., H2
Zn powder, aq. HCl, 08C
Zn powder, AcOH, 08C
39
38, 39
37, 38, 39
37, 38
37
ꢀ
HCl, were effective to reduce the N O bond in hydroxyla-
mine to give 46 as a sole product (Table 2, entry 5).[25]
Therefore, a combination of the conditions in entries 1 and
5 of Table 2 (i.e., TiCl3/NaOAc and zinc powder in MeOH/
aq. HCl) was employed. We were pleased to find that this
gave diamine 44 exclusively (Table 2, entry 6).
[a] Detected by ESI mass spectrometry.
powder in acetic acid, and hydroxylamine 37 was obtained
(Table 1, entry 5).
Since we had established a discriminative diamine synthe-
sis of nitro isoxazolidine 36, total synthesis of (+)-saxitoxin
(1) was examined (Scheme 10). Guanidination of 44 with
Under these conditions, hydroxylamine 37 could be prac-
tically obtained (on a multigram scale) in 70% yield from
ent-7.[24] These conversions, that is, 1) 1,3-dipolar cycloaddi-
tion reaction, 2) C5 isomerization, and 3) reduction of the
NO2 group, could be performed in one pot by successive ad-
dition of reagents without any work-up (Scheme 9). Prior to
Scheme 9. Synthesis of isoxazolidine 42. Reagents and conditions: a) No
solvent, 408C; then DBU, CH2Cl2, ꢀ608C; then Zn powder, AcOH,
ꢀ608C!RT. (37: 70%, 40: 7%, 41: 9%); b) CbzCl, K2CO3 (3 equiv),
CH2Cl2, 08C (42: 72%, 43: 20%); c) CbzCl, K2CO3 (5 equiv), CH2Cl2,
08C; then MeOH, RT. (42: 87%, 43: 0%). DBU=1,8-diazabicyclo-
Scheme 10. Synthesis of bis(guanidine) 52. Reagents and conditions:
a) TiCl3, Zn powder, NaOAc, aq. HCl, MeOH, CH2Cl2, 08C; b) NBoc=C-
AHCTUNGERTN(GNUN SMe)NHBoc (16), HgCl2, Et3N, DMF (60% two steps); c) ClCH2SO2Cl,
iPr2NEt, CH2Cl2 (98%); d) TBAF, THF, 08C (83%); e) (COCl)2
(10 equiv), DMSO (12 equiv), Et3N (30 equiv), CH2Cl2, ꢀ788C; f) IBX
(1.1 equiv), DMSO, 508C (94%); e) NaBH4, MeOH, 08C (77%); h) H2,
ACHTUNGTRENNUNG[5.4.0]undec-7-ene.
the second-step reduction, the amino group in hydroxyla-
mine 37 was protected with a Cbz group using benzyloxycar-
bonyl chloride in the presence of potassium carbonate. In
this reaction, mixtures of N-Cbz and O-Cbz compounds 42
and 43 were initially generated, but the mixture could be
converted into N-Cbz 42 exclusively simply by treatment
with methanol.
Pd(OH)2/C, MeOH; i) NBoc=CACTHNURTGENUNG(SMe)NHBoc (16), HgCl2, Et3N, DMF
(92% two steps).
bis(tert-butyloxycarbonyl)-2-methyl-2-thiopseudourea (16)
gave guanidine 47 in 60% yield from 42. Subsequent cycli-
zation by treatment with monochloromethanesulfonyl chlo-
ride and diisopropylethylamine gave cyclic guanidine 48 in
98% yield.[26] In this cyclization, Mitsunobu reaction condi-
tions, which successfully gave the cyclized product in the
synthesis of doSTX (2), yielded complex mixtures. Oxida-
tion at C4 was conducted in the same way as for the doSTX
synthesis. Thus, deprotection of the TIPS ether with TBAF
followed by Swern oxidation of the resulting alcohol gave
the corresponding ketone. Oxidation at the C4 position was
performed with IBX (1.1 equiv) to give the aminal 50 in
94% yield from 49. After reduction of the ketone in 50 with
NaBH4, the Cbz group was removed with hydrogen in the
presence of Pd(OH)2, and a Boc-protected guanidine group
was introduced to the resulting amine with 16 and mercury
(II) chloride to give 52 in 92% yield from 51. Prior to the
ꢀ
Next, reduction of two N O bonds in 42 was investigated
to obtain 44 (Table 2).[23] The reaction of 42 with Zn/aq.
ꢀ
HCl gave only 45, in which the N O bond in isoxazolidine
was cleaved (Table 2, entry 1). SmI2 gave similar results
[a]
ꢀ
Table 2. Reduction of two N O bonds in 42.
Entry
Reagent
Solvent
Temp.
Product
1
2
3
4
5
6
Zn
SmI2
NiCl2, NaBH4
CoCl2, NaBH4
TiCl3, NaOAc
TiCl3, Zn, NaOAc
MeOH, aq. HCl
THF
MeOH
MeOH
MeOH, aq. HCl
MeOH, aq. HCl
RT
08C
RT
RT
08C
08C
45
45
44, 45
44, 45
46
44
[a] Detected by ESI mass spectrometry.
Chem. Asian J. 2009, 4, 277 – 285
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemasianj.org
281

K. Nagasawa et al.
FULL PAPERS
1
JASCO DIP 1000 polarimeter using the sodium D line. H and 13C NMR
construction of bicyclic guanidine structure, we planned to
install the carbamoyl group at C13. Then, removal of the
benzyl protecting group was examined. However, this de-
protection was difficult under various conditions. Only re-
duction with hydrogen in the presence of Pd(OH)2 catalyst
gave alcohol 53 in 82% yield, and the reproducibility was
poor.[27]
Therefore, we decided to change the benzyl group to a p-
methoxybenzyl (MPM) group. Thus, the bicyclic guanidine
bearing MPM ether 55 was synthesized from nitrone ent-7
and nitroalkene 54 in a similar way to 55 in 16% yield over
11 steps (Scheme 11). The MPM group was cleanly depro-
spectra were recorded on a JEOL JNM-ECX 400 spectrometer. The
spectra were referenced internally according to residual solvent signals of
CDCl3 (1H NMR; d=7.26 ppm, 13C NMR; d=77.0 ppm), D2O (1H NMR;
d=4.79 ppm), and CD3CO2D (13C NMR; d=179.0 ppm for CD3CO2D).
Mass spectra were recorded on a JEOL JMS-T100X spectrometer in
ESI-MS mode with methanol as solvent.
37, 40, and 41: Nitroolefin 34 (0.96 g, 4.99 mmol, 1 equiv) was added to
the chiral nitrone ent-7 (1.28 g, 4.99 mmol) at room temperature and
stirred for 30 min at 408C without any solvent (1,3-dipolar cycloaddition).
The mixture was then diluted with 50 mL of CH2Cl2 and cooled to
ꢀ608C. DBU (803 mL, 5.70 mmol, 1.2 equiv) was added, and the solution
was stirred for 1 h at that temperature (C5 isomerization). Then, AcOH
(1.36 mL, 23.8 mmol) and freshly activated Zn powder (777 mg,
11.9 mmol) was added to the solution, and the resulting mixture was
warmed to 08C over 5 h. The reaction was then quenched with saturated
aq. NaHCO3. and extracted with EtOAc three times. The combined or-
ganic layer was dried over MgSO4, filtered, and concentrated in vacuo to
give a dark brown oil. The crude mixture was purified by silica gel
column chromatography (hexane/EtOAc; 10:1 to 4:1) to give 5a-hydrox-
ylamine 37 (1.44 g, 70%), 5b-hydroxylamine 40 (140 mg, 7%), and oxime
41 (195 mg, 9%). 37: ½aꢁD22 =+0.4 (c=2.2, CHCl3); 1H NMR (400 MHz,
CDCl3): d=4.56 (d, J=1.4 Hz, 2H), 4.23 (dt, J=2.3, 5.0 Hz, 1H), 4.15 (q,
J=5.5 Hz, 1H), 3.80 (m, 2H), 3.61 (m, 1H), 3.30 (m, 2H), 2.07 (dddd,
J=5.5, 7.3, 9.6, 12.8 Hz, 1H), 1.73 (dddd, J=1.8, 3.7, 5.5, 12.8 Hz, 1H),
1.05 ppm (s, 21H); 13C NMR (100 MHz, CDCl3): d=137.5, 128.4, 127.9,
127.8, 77.9, 77.4, 73.7, 72.0, 68.3, 55.8, 34.4, 17.9, 12.0 ppm; HRMS (ESI,
22
M+Na+) calcd for C23H40N2O4SiNa 459.2655, found 459.2695. 40: ½aꢁD
=
+8.4 (c=3.6, CHCl3); 1H NMR (400 MHz, CDCl3): d=7.32 (m, 5H),
4.70 (dt, J=3.7, 6.9 Hz, 1H), 4.58 (d, J=11.9 Hz, 1H), 4.53 (d, J=
11.9 Hz, 1H), 3.96 (m, 2H), 3.69 (dd, J=3.7, 7.3 Hz, 1H), 3.64 (dd, J=
6.0, 9.6 Hz, 1H), 3.54 (d, J=5.5, 10.0 Hz, 1H), 3.26–3.38 (m, 2H), 2.12
(m, 1H), 1.78 (m, 1H), 1.06 ppm (m, 21H); 13C NMR (100 MHz, CDCl3):
d=137.7, 128.4, 127.8, 127.7, 78.4, 77.2, 73.6, 72.5, 70.5, 70.1, 55.4, 35.0,
18.0, 12.3 ppm; HRMS (ESI, M+H+) calcd for C23H41N2O4Si 437.2836,
found 437.2886. 41: ½aꢁD22 =+44.5 (c=2.4, CHCl3); 1H NMR (400 MHz,
CDCl3): d=8.49 (brd, J=9.2 Hz, 1H), 7.32 (m, 5H), 4.71 (d, J=5.0 Hz,
1H), 4.65 (m, 2H), 4.54 (d, J=12.4 Hz, 1H), 4.36 (s, 1H), 3.71 (dd, J=
8.2, 10.5 Hz, 1H), 3.57 (dd, J=3.2, 10.5 Hz, 1H), 3.49 (dd, J=6.9,
13.3 Hz, 1H), 3.39 (dt, J=5.0, 13.3 Hz, 1H), 2.04 (m, 1H), 1.77 (m, 1H),
1.05 ppm (m, 21H); 13C NMR (100 MHz, CDCl3): d=162.5, 137.9, 128.3,
127.8, 127.6, 76.8, 75.8, 75.7, 73.3, 70.5, 56.0, 34.8, 18.0, 12.2 ppm; HRMS
(ESI, M+Na+) calcd for C23H38N2O4SiNa 457.2499, found 457.2499.
Scheme 11. Formal total synthesis of (+)-STX (1). Reagents and condi-
tions: a) DDQ, CH2Cl2, H2O (90%); b) Cl3CC(O)N=C=O, CH2Cl2, then
K2CO3, MeOH (56: 35%, 57: 20%); c) BACHTUNGTRENUNG(CF3CO2)3, TFA (88%).
DDQ=2,3-dichloro-5,6-dicyano-p-benzoquinone, MPM=p-methoxyben-
zyl.
tected with DDQ to give the alcohol in 90% yield. The re-
sulting alcohol 53 was treated with trichloroacetyl isocya-
nate, followed by hydrolysis with potassium carbonate in
MeOH to give 56 in 35% yield. In this reaction, bis(carba-
moyl) 57 was also obtained in 20% yield. Finally, cyclization
of guanidine and removal of the four Boc protecting groups
of 57 proceeded simultaneously with boron tris(trifluoroace-
tate) in TFA to give an 88% yield of (+)-b-saxitoxinol
(58),[7] which is known to be converted into (+)-saxitoxin
(1) by oxidation at C12.[4b]
42: CbzCl (518 mL, 3.63 mmol) was added to a solution of 5a-hydroxyla-
mine 37 (1.44 g, 3.30 mmol) and K2CO3 (2.28 g, 16.5 mmol) in THF/H2O
(2:1, 30 mL) at 08C. The solution was stirred for 10 min at 08C, then
methanol (30 mL) was added, and resulting mixture was warmed to room
temperature. The reaction mixture was then concentrated in vacuo. The
resulting material was diluted with EtOAc/H2O and extracted with
EtOAc three times. The combined organic layer was dried over MgSO4,
filtered, and concentrated in vacuo to give a yellow brown oil. The crude
mixture was purified by silica gel column chromatography (hexane/
EtOAc; 8:1 to 2:1) to give N-benzyloxycarbonylhydroxylamine 42
(1.61 g, 87%). ½aꢁ2D2 =+18.0 (c=3.6, CHCl3); 1H NMR (400 MHz,
CDCl3): d=7.82 (brs, 1H), 7.22–7.35 (m, 10H), 5.19 (d, J=12.4 Hz, 1H),
5.15 (d, J=11.9 Hz, 1H), 4.92 (m, 1H), 4.48 (s, 2H), 4.33 (m, 1H), 4.14
(dq, J=3.4, 5.5 Hz, 1H), 3.96, (t, J=2.8 Hz, 1H), 3.84 (dd, J=11.0,
5.5 Hz, 1H), 3.68 (dd, J=3.66, 11.0 Hz, 1H), 3.40 (ddd, J=6.4,9.2,
12.8 Hz, 1H), 3.16 (ddd, J=4.1, 7.8,11.9 Hz, 1H), 2.02 (m,1H), 1.75 (m,
1H), 1.02 ppm (m, 21H); 13C NMR (100 MHz, CDCl3): d=156.1, 136.5,
135.9, 128.52, 128.47, 128.26, 128.11, 128.04, 77.4, 77.2, 76.5, 75.3, 74.1,
67.8, 67.1, 55.3, 33.9, 17.9, 17.8, 11.9 ppm; HRMS (ESI, M+Na+) calcd
for C31H46N2O6SiNa 593.3023, found 593.3042.
Conclusions
In conclusion, total syntheses of (ꢀ)- and (+)-doSTX (2)
and (+)-STX (1) have been achieved based upon oxidation
at C4 with IBX. In the synthesis of (+)-STX (1), discrimina-
ꢀ
tive reduction of the nitro group and N O bond in nitroi-
soxazolidine was newly developed to provide efficient
access to the key diamine intermediate for STXs.
Experimental Section
47: TiCl3/HCl (7.30 mL, 5.64 mmol) was added to a solution of N-benzy-
loxycarbonylhydroxylamine (1.61 g, 2.82 mmol), NaOAc (2.31 g,
28.2 mmol), and freshly activated Zn powder (1.84 g, 28.2 mmol) in
CH2Cl2/MeOH at 08C under an N2 atmosphere. The reaction mixture
was stirred for 5 h, quenched with saturated aq. NaHCO3, and extracted
General
Flash chromatography was performed on silica gel 60 (spherical, particle
size 0.040–0.100 mm; Kanto). Optical rotations were measured on
a
282
www.chemasianj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Asian J. 2009, 4, 277 – 285

(ꢀ)- and (+)-Decarbamoyloxysaxitoxin and (+)-Saxitoxin
with EtOAc three times. The combined organic layer was dried over
MgSO4, filtered, and concentrated in vacuo to give diamine 44 (1.49 g).
HgCl2 (766 mg, 2.82 mmol) was added to a solution of crude diamine 44
was quenched with 10% aq. Na2S2O3 and saturated aq. NaHCO3, and the
solution was diluted with EtOAc. The organic layer was separated and
washed with saturated aq. NaHCO3 and brine. The organic layer was
dried over MgSO4, filtered, and concentrated in vacuo. The residue was
purified by silica gel column chromatography (hexane/EtOAc; 4:1 to 1:1)
to give aminal 50 (60 mg, 94%) as an inseparable mixture of its C12 hy-
drate. NaBH4 (4 mg, 0.094 mmol) was added to a solution of aminal 50
(60 mg, 0.096 mmol) in methanol (2.0 mL) at 08C. After stirring for
30 min, the reaction mixture was quenched with H2O, and the solution
was extracted with EtOAc three times. The combined organic layer was
dried over MgSO4, filtered, and concentrated in vacuo. The residue was
purified by silica gel column chromatography (hexane/EtOAc; 4:1 to 1:1)
to give alcohol 51 (46 mg, 77%). ½aꢁ2D2 =+75.9 (c=2.4, CHCl3); 1H NMR
(400 MHz, CDCl3): d=7.30 (m, 10H), 5.55 (brd, J=8.7 Hz, 1H), 5.10 (d,
J=11.9 Hz, 1H), 5.03 (d, J=11.9 Hz,1H), 4.57 (d, J=11.9 Hz, 1H), 4.53
(dd, J=6.0, 9.6 Hz, 1H), 4.47 (d, J=11.6 Hz, 1H), 4.18 (dt, J=3.2,
6.4 Hz, 1H), 4.05 (d, J=2.75, 1H), 3.87 (dd, J=3.7, 9.2 Hz, 1H), 3.82 (t,
J=9.2 Hz, 1H), 3.66 (dd, J=6.9, 9.2 Hz, 1H), 3.51 (dt, J=6.9, 11.0 Hz,
1H), 2.14 (m, 1H), 1.87 (dd, J=5.5, 13.3 Hz, 1H), 1.44 (s, 9H), 1.43 ppm
(s, 9H); 13C NMR (100 MHz, CDCl3): d=159.0, 157.2, 151.4, 148.8, 138.0,
035.6, 128.6, 128.4, 128.2, 128.1, 127.5, 127.4, 91.0, 83.2, 79.0, 76.9, 73.3,
71.7, 67.7, 57.0, 53.0, 46.3, 29.0, 28.2, 28.1 ppm; HRMS (ESI, M+Na+)
calcd for C33H44N4O9Na 663.3006, found 663.3004.
(1.49 g),
bis(tert-butyloxycarbonyl)-2-methyl-2-thiopseudourea
(16)
(819 mg, 2.82 mmol) and Et3N (1.18 mL, 8.46 mmol) in DMF (14 mL) at
room temperature under an N2 atmosphere. The reaction mixture was
stirred for 1 h, diluted with EtOAc, and filtered through a pad of celite.
The filtrates were washed with H2O and brine twice. The organic layer
was dried over MgSO4, filtered, and concentrated in vacuo to give pyrro-
lidine as a yellow brown oil. The crude mixture was purified by silica gel
column chromatography (hexane/EtOAc; 15:1 to 5:1) to give guanidine
47 (1.35 mg, 60% steps). ½aꢁ2D2 =+5.6 (c=2.8, CHCl3); 1H NMR
(400 MHz, CDCl3): d=10.32 (brs, 1H), 7.32 (m, 10H), 5.70 (brs, 1H),
5.44 (d, J=9.6 Hz, 1H), 5.09 (d, J=12.4 Hz, 1H), 5.04 (d, J=12.4 Hz,
1H), 4.49 (m, 1H), 4.47 (d, J=11.9 Hz, 1H), 4.42 (d, J=11.9 Hz, 1H)
4.38 (m, 1H), 3.92 (m, 2H), 3.58 (dq, J=7.33, 11.0 Hz, 1H), 3.50 (m,
2H), 3.36 (dd, J=6.9, 9.6 Hz, 1H), 2.36 (m, 1H), 1.49 (s, 9H), 1.44 (s,
9H), 0.97 ppm (s, 21H); 13C NMR (100 MHz, CDCl3): d=161.3, 157.6,
156.5, 150.2, 138.1, 136.4, 128.4, 128.3, 128.0, 127.9, 127.5, 82.3, 79.5, 73.3,
72.5, 70.8, 70.1, 66.8, 53.0, 46.8, 32.1, 28.1, 28.0, 17.8, 12.0 ppm; HRMS
(ESI, M+H+) calcd for C42H67N4O9Si 799.4677, found 799.4637.
48: Chloromethanesulfonyl chloride (219 mL, 2.21 mmol) was added to a
solution of guanidine 47 (352 mg, 0.441 mmol) and iPr2NEt (911 mL,
5.29 mmol) in CH2Cl2 (5.0 mL) at 08C under an N2 atmosphere. The re-
action mixture was stirred for 12 h, quenched with saturated aq.
NaHCO3, and extracted with CH2Cl2 twice. The combined organic layer
was dried over MgSO4, filtered, and concentrated in vacuo to give a
yellow brown oil. The crude mixture was purified by silica gel column
chromatography (hexane/EtOAc; 10:1 to 4:1) to give cyclic guanidine 48
(337 mg, 97%). ½aꢁ2D2 =+96.0 (c=2.0, CHCl3); 1H NMR (400 MHz,
CDCl3): d=7.32 (m, 10H), 5.08 (m, 2H), 5.00 (d, J=11.9 Hz, 1H), 4.52
(d, J=11.0 Hz,1H), 4.42 (d, J=2.3 Hz, 1H), 4.35 (d, J=10.5 Hz, 1H),
4.25 (m, 1H), 4.00 (dt, J=4.6, 10.1 Hz, 1H), 3.84 (dd, J=2.8, 9.2 Hz,1H),
3.74 (dd, J=4.6, 9.2 Hz, 1H), 3.59 (dt, J=6.7, 11.0 Hz, 1H), 3.42 (m,
1H), 3.19 (d, J=10.5 Hz, 1H), 1.76 (m, 2H), 1.48 (s, 18H), 0.98 ppm (s,
21H); 13C NMR (100 MHz, CDCl3): d=159.6, 155.6, 151.8, 151.6, 138.1,
135.9, 128.5, 128.3, 128.2, 128.1, 127.9, 127.5, 82.6, 78.1, 74.1, 73.3, 71.6,
68.5, 67.2, 59.4, 54.4, 46.0, 32.5, 28.5, 28.1, 17.9, 12.0 ppm; HRMS (ESI,
M+H+) calcd for C42H65N4O8Si 781.4572, found 781.4606.
52: 20% Pd(OH)2/C (30 mg) was added under an N2 atmosphere to a so-
lution of alcohol 51 (99 g, 0.155 mmol) in methanol (5.0 mL), and the sus-
pension was vigorously stirred under H2 atmosphere (balloon) at room
temperature. After 8 h, the reaction mixture was filtered through a pad
of celite. The filtrates were concentrated in vacuo to give crude amine
(84 mg). HgCl2 (51 mg, 0.186 mmol) was added to a solution of amine
(84 mg), Et3N (65 mL, 0.465 mmol), and bis(tert-butyloxycarbonyl)-2-
methyl-2-thiopseudourea (16) (54 mg, 0.186 mmol) in DMF (1.5 mL) at
room temperature under an N2 atmosphere. The reaction mixture was
stirred for 1 h, diluted with EtOAc, and filtered through a pad of celite.
The filtrates were washed with H2O and brine twice. The organic layer
was dried over MgSO4, filtered, and concentrated in vacuo to give a
yellow brown oil. The crude mixture was purified by chromatorex NH
(Fuji Silysia) gel column chromatography (hexane/EtOAc; 4:1 to 1:1) to
give bis(guanidine) 52 (106 mg, 92%). ½aꢁ2D2 =+50.5 (c=2.6, CHCl3);
1H NMR (400 MHz, CDCl3): d=11.32 (brs, 1H), 8.72 (brd, J=8.2 Hz,
1H), 7.25 (m, 5H), 6.87 (brs, 1H), 4.81 (dd, J=5.5, 8.7 Hz, 1H), 4.55 (d,
J=11.5 Hz, 1H), 4.49 (d, J=11.5 Hz, 1H), 4.33 (m, 1H), 4.01 (m, 1H),
3.92 (dd, J=4.6, 9.2 Hz, 1H), 3.87 (m, 1H), 3.61 (m, 2H), 2.95 (brs, 1H),
2.20 (m, 1H), 1.98 (dd, J=6.41, 13.3 Hz, 1H), 1.48 (s, 9H), 1.47 (s, 9H),
1.46 (s, 9H), 1.44 ppm (s, 9H); 13C NMR (100 MHz, CDCl3): d=162.2,
159.0, 156.0, 152.6, 151.1, 148.37, 138.0, 128.3, 127.4, 127.3, 91.5, 83.8,
83.2, 79.9, 79.0, 76.3, 73.1, 71.7, 56.8, 53.4, 46.4, 18.9, 28.2, 28.1, 28.0,
27.9 ppm; HRMS (ESI, M+Na+) calcd for C36H56N6O11Na 771.3905,
found 771.3898.
55: ½aꢁ2D2 =+65.1 (c=1.9, CHCl3) 1H NMR (400 MHz, CDCl3): d=11.33
(brs, 1H), 8.72 (brd, J=8.7 Hz, 1H), 7.19 (d, J=8.7 Hz, 2H), 6.88 (brs,
1H), 6.81 (d, J=8.7 Hz, 2H), 4.80 (dd, J=5.5, 8.7 Hz, 1H), 4.49 (d, J=
11.5 Hz, 1H), 4.41 (d, J=11.5 Hz,1H), 4.32 (m, 1H), 4.01 (d, J=3.8 Hz,
1H), 3.89 (m, 2H), 3.78 (s, 3H), 3.61 (m, 2H), 2.15 (m, 1H), 2.00 (dd, J=
6.4, 13.3 Hz, 1H), 1.49 (s, 9H), 1.48 (s, 9H), 1.47 (s, 9H), 1.45 ppm (s,
9H); 13C NMR (100 MHz, CDCl3): d=162.2, 159.0, 158.9, 156.0, 152.6,
151.2, 148.3, 130.2, 128.9, 113.7, 91.4, 83.8, 83.3, 80.0, 79.1, 76.3, 72.9, 71.5,
56.9, 55.2, 53.3, 46.5, 28.9, 28.2, 28.1, 28.0, 27.9 ppm; HRMS (ESI,
M+Na+) calcd for C37H58N6O12Na 801.4010, found 801.4010.
49: TBAF (286 mg, 1.09 mmol) was added to a solution of cyclic guani-
dine 48 (569 mg, 0.729 mmol) in a solution of THF (14 mL) at 08C. The
reaction mixture was stirred for 1 h at that temperature, quenched with
saturated aq. NH4Cl, and extracted with EtOAc three times. The com-
bined organic layer was dried over MgSO4, filtered, and concentrated in
vacuo. The residue was purified by silica gel column chromatography
22
(hexane/EtOAc; 3:1 to 1:3) to give alcohol 49 (377 mg, 83%). ½aꢁD
=
+141.5 (c=2.7, CHCl3); 1H NMR (400 MHz, CDCl3): d=7.32 (m, 10H),
6.05 (brs, 1H), 5.11 (d, J=12.4 Hz, 1H), 5.07 (d, J=12.4 Hz, 1H), 4.48
(d, J=10.1 Hz, 1H), 4.34 (d, J=10.1 Hz, 1H), 4.21 (m, 1H), 4.11 (m,
1H), 4.01 (dt, J=2.8, 8.7 Hz, 1H), 3.85 (dd, J=2.3, 9.2 Hz, 1H), 3.74 (dd,
J=3.2, 8.7 Hz, 1H), 3.54 (m, 1H), 3.36 (d, J=10.1 Hz, 1H), 3.02 (m,
1H), 1.71 (m, 2H), 1.45 (s, 9H), 1.41 ppm (s, 9H); 13C NMR (100 MHz,
CDCl3): d=159.1, 156.2 ,152.0, 151.8, 137.9, 136.1, 128.5, 128.2, 128.1,
127.6, 83.4, 78.2, 73.7, 73.4, 71.9, 67.0, 65.5, 61.4, 54.8, 46.2, 31.4 ppm;
HRMS (ESI, M+H+) calcd for C33H44N4O8 625.3237, found 625.3217.
51: (COCl)2 (85 mL, 1.00 mmol) was slowly added to a solution of DMSO
(85 mL, 1.2 mmol) in CH2Cl2 (2.0 mL) at ꢀ788C under an N2 atmosphere.
The reaction mixture was stirred for 10 min, and then alcohol 49 (63 mg,
0.100 mmol) was added dropwise as a solution in CH2Cl2 (1.0 mL) over
15 min under an N2 atmosphere. The reaction mixture was stirred for 1 h,
then Et3N (418 mL, 3.00 mmol) was added, and the mixture was stirred
for 5 min, rapidly quenched with H2O, and warmed to room temperature.
The solution was extracted with CH2Cl2 three times. The combined or-
ganic layer was dried over MgSO4, filtered, and concentrated in vacuo to
give crude ketone (79 mg). IBX (31 mg, 0.110 mmol) was added to a so-
lution of crude ketone (79 mg) in DMSO (1.0 mL), and the resulting sus-
pension was stirred for 20 min at room temperature. Then, the reaction
mixture was warmed to 508C and stirred for a further 1 h. The reaction
53: DDQ (86 mg, 0.381 mmol) was added to a solution of MPM ether 55
(64 mg, 0.095 mmol) in CH2Cl2/H2O (2:1, 3.0 mL) at room temperature.
The reaction mixture was stirred for 1 h, saturated aq. NaHCO3 was
added, and the resulting solution was extracted with EtOAc three times.
The combined organic layer was dried over MgSO4, filtered, and concen-
trated in vacuo. The residue was purified by Chromatorex NH (Fuji Sily-
sia) gel column chromatography (hexane/EtOAc; 2:1 to pure EtOAc) to
give alcohol 53 (47 mg, 90%). ½aꢁ2D2 =+52.1 (c=2.2, CHCl3); 1H NMR
(400 MHz, CDCl3): d=11.42 (s, 1H), 8.69 (d, J=9.16 Hz, 1H), 6.68 (brs,
1H), 5.26 (brs, 1H), 4.60 (dd, J=7.3, 8.2 Hz, 1H), 4.12 (dt, J=3.2,
Chem. Asian J. 2009, 4, 277 – 285
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemasianj.org
283

K. Nagasawa et al.
FULL PAPERS
6.4 Hz, 1H), 3.98 (m, 2H), 3.82 (dd, J=8.2, 11.0 Hz, 1H), 3.47 (m, 1H),
2.2 (m, 1H), 2.00 (dd, J=6.0, 13.3 Hz, 1H), 1.50 (s, 9H), 1.49 (s, 9H),
1.47 (s, 9H), 1.45 ppm (s, 9H); 13C NMR (100 MHz, CDCl3): d=162.0,
161.7, 156.1, 152.8, 150.7, 150.3, 92.1, 84.2, 83.4, 80.3, 80.2, 77.2, 76.2, 61.3,
59.4, 51.0, 46.6, 29.1, 28.1, 28.0, 27.9 ppm; HRMS (ESI, M+Na+) calcd
for C29H50N6O11Na 681.3435, found 681.3423.
Chem. Soc. 1957, 79, 5230–5235; b) W. Schuett, H. Rapoport, J.
Am. Chem. Soc. 1962, 84, 2266–2267.
[3] L. E. Llewellyn, Nat. Prod. Rep. 2006, 23, 200–222.
[4] a) E. J. Schantz, J. D. Mold, D. W. Stanger, J. Shavel, F. J. Riel, J. P.
Bowden, J. M. Lynch, R. S. Wyler, B. Riegel, H. Sommer, J. Am.
Chem. Soc. 1957, 79, 5230–5235; b) F. E. Koehn, V. E. Ghazarossi-
an, E. J. Schantz, H. K. Schnoes, F. M. Strong, Bioorg. Chem. 1981,
10, 412–428; c) A. A. Genenah, Y. Shimizu, J. Agric. Food Chem.
1981, 29, 1289–1291; d) C. F. Wichmann, W. P. Niemczura, H. K.
Schnoes, S. Hall, P. B. Reichardt, S. D. Darling, J. Am. Chem. Soc.
1981, 103, 6977–6978; e) E. Moczydlowski, S. Hall, S. S. Garber,
G. S. Strichartz, C. J. Miller, J. Gen. Physiol. 1984, 84, 687–704;
f) J. J. Sullivan, M. M. Wekell, L. L. Kentala, J. Food Sci. 1985, 50,
26–29; g) Y. Oshima, Manual on Harmful Marine Microalgae (Eds.:
G. Hallegraeff, D. Anderson, A. Cembella), UNESCO, Paris, 1995,
pp. 81–94; h) H. Onodera, M. Satake, Y. Oshima, T. Yasumoto,
W. W. Carmichael, Nat. Toxins 1997, 5, 146–151; i) G. Usup, C.-P.
Leaw, M.-Y. Cheah, A. Ahmad, B.-K. Ng, Toxicon 2004, 44, 37–43;
j) L. Llewellyn, A. Negri, M. Quilliam, Toxicon 2004, 43, 101–104;
k) M. Yotsu-Yamashita, Y.-H. Kim, S. C. Dudley Jr. , G. Choudhary,
A. Pfahnl, Y. Oshima, J. W. Daly, Proc. Natl. Acad. Sci. USA 2004,
101, 4346–4351.
[5] a) H. Tanino, T. Nakata, Y. Kaneko, Y. Kishi, J. Am. Chem. Soc.
1977, 99, 2818–2819; b) Y. Kishi, Heterocycles 1980, 14, 1477–1495.
[6] a) P. A. Jacobi, M. J. Martinelli, S. Polanc, J. Am. Chem. Soc. 1984,
106, 5594–5598; b) M. J. Martinelli, A. D. Brownstein, P. A. Jacobi,
S. Polanc, Cort. Chem. Acta 1986, 59, 267–295; c) P. A. Jacobi, Strat-
egies and Tactics in Organic Synthesis, Vol. 2 (Ed.: T. Lindberg),
Academic Press, New York, 1989, pp. 191–219.
[7] a) J. J. Fleming , J. Du Bois, J. Am. Chem. Soc. 2006, 128, 3926–
3927; b) J. J. Fleming, M. D. McReynolds, J. Du Bois, J. Am. Chem.
Soc. 2007, 129, 9964–9975.
56 and 57: Trichloroacetyl isocyanate (3.0 mL, 0.025 mmol) was added to
a solution of alcohol 53 (15 mg, 0.023 mmol) in CH2Cl2 at room tempera-
ture under an N2 atmosphere. The reaction was stirred for 1 h, K2CO3
and methanol were then added at 08C, and the solution was stirred for
another 24 h at 08C. The reaction mixture was quenched with H2O and
extracted with EtOAc three times. The combined organic layer was dried
over MgSO4, filtered, and concentrated in vacuo. The residue was puri-
fied by preparative TLC (hexane/EtOAc; 1:2, twice) to give monocarba-
22
mate 56 (6 mg, 35%) and bis(carbamate) 57 (3 mg, 20%). 56: ½aꢁD
=
+39.4 (c=2.2, CHCl3); 1H NMR (400 MHz, CDCl3): d=11.37 (s, 1H),
8.74 (d, J=8.7 Hz, 1H), 6.64 (brs, 1H), 4.90 (brs, 1H), 4.62 (dd, J=6.9,
8.7 Hz, 1H), 4.38 (m, 1H), 4.26 (m, 2H), 4.00 (d, J=3.4 Hz, 1H), 3.66
(dt, J=6.0, 11.0 Hz, 1H), 2.18 (m, 1H), 1.97 (dd, J=6.0, 13.3 Hz, 1H),
15.0 (s, 9H), 1.49 (s, 9H), 1.46 (s, 9H), 1.45 ppm (s, 9H); 13C NMR
(100 MHz, CDCl3): d=162.0, 158.7, 156.5, 156.0, 152.7, 152.7, 151.1,
147.8, 91.2, 84.2, 83.6, 80.3, 79.2, 76.2, 65.1, 56.2, 51.9, 46.6, 28.8, 28.2,
28.1, 28.03, 28.00 ppm; HRMS (ESI, M+Na+) calcd for C30H51N7O12Na
724.3493, found 724.3534. 57: ½aꢁ2D2 =+16.5 (c=2.4, CHCl3); 1H NMR
(400 MHz, CDCl3): d=11.35 (s, 1H), 9.07 (d, J=7.9 Hz, 1H), 5.1 (brs,
1H), 5.01 (d, J=3.2 Hz, 1H), 4.71 (m, 1H), 4.48 (dd, J=6.0, 6.9 Hz, 1H),
4.38 (dd, J=7.8, 11.0 Hz, 1H), 4.25 (dd, J=4.1, 11.0 Hz, 1H), 3.91 (dd,
J=9.6, 10.1 Hz, 1H), 3.46 (dt, J=6.9, 11.5 Hz, 1H), 2.39 (m, 1H), 2.07
(dd, J=6.9, 13.7 Hz, 1H), 1.49 (s, 9H), 1.47 (s, 18H), 1.44 ppm (s, 9H);
13C NMR (100 MHz, CDCl3): d=162.3, 158.9, 157.6, 155.6, 155.3, 152.6,
150.9, 148.0, 89.2, 83.7, 83.2, 79.9, 79.5, 79.1, 64.1, 56.2, 52.3, 44.8, 29.7,
28.1, 28.0, 27.8 ppm; HRMS (ESI, M+Na+) calcd for C31H52N8O13Na
767.3552, found 767.3569.
[8] a) C. Y. Hong, Y. Kishi, J. Am. Chem. Soc. 1992, 114, 7001–7006;
b) J. V. Mulcahy, J. Du Bois, J. Am. Chem. Soc. 2008, 130, 12630–
12631.
58: A 0.5m solution of BACHTUNRGTNEUNG(OCOCF3)3 (1.0 mL) in TFA was added to a so-
lution of carbamate 56 (21 mg, 0.029 mmol) in TFA (0.5 mL) at room
temperature. The reaction mixture was warmed to 708C, stirred for 6 h,
then cooled to room temperature, quenched with methanol, and concen-
trated in vacuo. The resulting residue was then dissolved with 5 mL of
H2O and desalinated with Amberlyst A 26(OH) (strongly basic resin).
The resin was removed by filtration, and the filtrates were neutralized
with Amberlite IRC-50 (weekly acidic resin) to trap the basic STXol.
The resin was washed with H2O (20 mL) and then 10% aq. TFA
[9] M. S. Gold, Exp. Neurol. 2008, 210, 1–6.
[10] O. Iwamoto, H. Koshino, D. Hashizume, K. Nagasawa, Angew.
Chem. 2007, 119, 8779–8782; Angew. Chem. Int. Ed. 2007, 46, 8625–
8628.
[11] Y. Oshima, K. Sugino, H. Itakura, M. Hirota, T. Yasumoto, Toxic
Marine Phytoplankton (Eds.: R. Graneli, B. Sundstrom, L. Edler,
D. M. Anderson), Elsevier Science, Amsterdam, 1990, pp. 391–396.
[12] a) O. Iwamoto, M. Sekine, H. Koshino, K. Nagasawa, Heterocycles
2006, 70, 107–112; b) J. Shimokawa, K. Shirai, A. Tanatani, Y. Ha-
shimoto, K. Nagasawa, Angew. Chem. 2004, 116, 1585–1588;
Angew. Chem. Int. Ed. 2004, 43, 1559–1562; c) J. Shimokawa, T. Ish-
iwata, K. Shirai, H. Koshino, A. Tanatani, T. Nakata, Y. Hashimoto,
K. Nagasawa, Chem. Eur. J. 2005, 11, 6878–6888.
[13] A. Goti, M. Cacciarini, F. Cardona, A. Brandi, Tetrahedron Lett.
1999, 40, 2853–2856.
[14] K. S. Kim, L. Qian, Tetrahedron Lett. 1993, 34, 7677–7680.
[15] D. S. Dodd, A. P. Kozikowski, Tetrahedron Lett. 1994, 35, 977–980.
[16] Under the Swern oxidation conditions, C12 O-methyl thiomethyl
(MTM) ether 59 was obtained in 79% yield.
(30 mL). The aqueous TFA fraction was collected and concentrated in
22
vacuo to give b-saxitoxinol (58) as the 2TFA salt (12 mg, 88%). ½aꢁD
=
+90.2 (c=1.2, MeOH); 1H NMR (400 MHz, D2O): d=4.76 (d, J=
1.4 Hz, 1H), 4.31 (d, J=4.1 Hz, 1H), 4.24 (dd, J=9.2, 11.5 Hz, 1H), 4.00
(dd, J=5.5, 11.5 Hz, 1H), 3.80 (ddd, J=1.4, 5.5, 9.2 Hz, 1H), 3.74 (dt, J=
2.3, 10.1 Hz, 1H), 3.64 (m, 1H), 2.38 (m, 1H), 2.22 ppm (ddd, J=1.8, 8.2,
14.7 Hz, 1H); 13C NMR (100 MHz, 4% CD3CO2D in D2O): d=160.7,
159.3, 157.4, 85.1, 76.1, 64.8, 59.4, 54.5, 45.4, 30.4 ppm; HRMS (ESI,
M+H+) calcd for C10H18N7O3 284.1471, found 284.1481.
Acknowledgements
This work was financially supported in part by a Grant-in-Aid for Scien-
tific Research (B) from JSPS (No. 20310130) and by a Grant-in-Aid for
Scientific Research on Priority Areas from MEXT (No. 17035025). O.I. is
grateful for financial support from an education program “Human Re-
source Development Program for Scientific Powerhouse” provided
through the Tokyo University of Agriculture & Technology.
[17] See the Supporting Information.
[18] Nicolaou et al. reported that IBX can oxidize alcohols, ketones, and
aldehydes to the corresponding a,b-unsaturated systems through
enol formation of a carbonyl group and subsequent single-electron-
transfer (SET) based oxidation reaction. See: a) K. C. Nicolaou, Y.-
L. Zhong, P. S. Baran, J. Am. Chem. Soc. 2000, 122, 7596–7597;
[1] W. A. Catterall, A. L. Goldin, S. G. Waxman, Pharmacol. Rev. 2005,
57, 397–409.
[2] a) E. J. Schantz, J. D. Mold, D. W. Stanger, J. Shavel, F. J. Riel, J. P.
Bowden, J. M. Lynch, R. S. Wyler, B. Riegel, H. Sommer, J. Am.
284
www.chemasianj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Asian J. 2009, 4, 277 – 285

(ꢀ)- and (+)-Decarbamoyloxysaxitoxin and (+)-Saxitoxin
b) K. C. Nicolaou, T. Montagnon, P. S. Baran, Y.-L. Zhong, J. Am.
Chem. Soc. 2002, 124, 2245–2258.
Ramasubbu, S. Muthusubramanian, S. Sivasubramanian, Synth.
Commun. 1999, 29, 21–26; j) P. Borrachero, F. Cabrera, M. J.
Dianez, M. D. Estrada, M. Gomez-Guillen, A. Lopez-Castro, J. M.
Moreno, J. L. de Paz, S. Perez-Garrido, Tetrahedron: Asymmetry
1999, 10, 77–98; k) A. Voituriez, J. Moulinas, C. Kouklovsky, Y.
Langlois, Synthesis 2003, 1419–1426; l) V. Sridharan, S. Muthusubra-
manian, S. Sivasubramanian, K. Polborn, Tetrahedron 2004, 60,
8881–8892; m) A. Banerji, M. Gupta, P. K. Biswas, T. Prange, A.
Neuman, J. Heterocycl. Chem. 2007, 5, 1045–1049.
[19] J. A. McCauley, K. Nagasawa, P. A. Lander, S. A. Mischke, M. A.
Semones, Y. Kishi, J. Am. Chem. Soc. 1998, 120, 7647–7648.
[20] Oxidation of alcohol 24 was difficult even though it has a less bulky
group, NHBoc, at C5 than bis(guanidine) compounds 14 and 19.
[21] D. Lucet, S. Sabelle, O. Kostelitz, T. Le Gall, C. Mioskowski, Eur. J.
Org. Chem. 1999, 2583–2591.
[22] a) A. Padwa, K. F. Koehler, A. Rodriguez, J. Am. Chem. Soc. 1981,
103, 4974–4975; b) A. Padwa, L. Fisera, K. F. Koehler, A. Rodri-
guez, G. S. K. Wong, J. Org. Chem. 1984, 49, 276–281; c) A. M. N.
El-Din, Bull. Chem. Soc. Jpn. 1986, 59, 1239–1243; d) M. Burdisso,
A. Gamba, R. Gandolfi, P. Pevarelloa, Tetrahedron 1987, 43, 1835–
1846; e) M. Burdisso, R. Gandolfi, P. Grunanger, Tetrahedron 1989,
45, 5579–5594; f) T. Yakura, M. Nakazawa, T. Takino, M. Ikeda,
Chem. Pharm. Bull. 1992, 40, 2014–2018; g) K. Tanaka, T. Mori, K.
Mitsuhashi, Bull. Chem. Soc. Jpn. 1993, 66, 263–268; h) S. Moriya-
ma, Y. Vallee, Synthesis 1998, 393–404; i) V. Ramamoorthy, A.
[23] All these reactions were monitored by ESI mass spectrometry.
[24] Small amounts of C5 epimer 40 and oxime 41 were generated.
[25] P. G. Mattingly, M. J. Miller, J. Org. Chem. 1980, 45, 410–415.
[26] T. Shimizu, T. Ohzeki, K. Hiramoto, N. Hori, T. Nakata, Synthesis
1999, 1373–1385.
[27] M. B.-U. Surfraz, M. Akhtar, R. K. Allemann, Tetrahedron Lett.
2004, 45, 1223–1226.
Received: October 7, 2008
Published online: November 28, 2008
Chem. Asian J. 2009, 4, 277 – 285
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemasianj.org
285