Total synthesis of chiriquitoxin, an analogue of tetrodotoxin isolated from the skin of a dart frog

Article abstract of DOI:10.1002/chem.201304110

The first total synthesis of chiriquitoxin, the most structurally complex analogue of tetrodotoxin isolated from a Costa Rican dart frog, has been accomplished from a newly designed intermediate for a variety of tetrodotoxin derivatives. The synthesis includes the third total synthesis of tetrodotoxin in this laboratory, and its intermediate was transformed into chiriquitoxin by a stereocontrolled aldol reaction with a d-camphor-derived lactone for installation of the unique side chain, and a new deprotection of methylthiomethyl (MTM) ether by using a Pummerer rearrangement.

Full text of DOI:10.1002/chem.201304110

DOI: 10.1002/chem.201304110
Communication
&
Natural Products
Total Synthesis of Chiriquitoxin, an Analogue of Tetrodotoxin
Isolated from the Skin of a Dart Frog
Masaatsu Adachi,[a] Takuya Imazu,[a] Ryo Sakakibara,[a] Yoshiki Satake,[a] Minoru Isobe,[a, b] and
Toshio Nishikawa*[a]
the field of neurophysiology as with tetrodotoxin and saxitox-
Abstract: The first total synthesis of chiriquitoxin, the
most structurally complex analogue of tetrodotoxin isolat-
in.[9] However, the scarcity of chiriquitoxin from natural sources
has prevented further detailed investigations of the mecha-
ed from a Costa Rican dart frog, has been accomplished
nisms of the ion-channel inhibitions. Thus, supply of 2 by
from a newly designed intermediate for a variety of tetro-
chemical synthesis has been highly desired.
dotoxin derivatives. The synthesis includes the third total
We have been synthesizing tetrodotoxin and its analogues
synthesis of tetrodotoxin in this laboratory, and its inter-
for a long time.[10–18] In recent years, we have focused on devel-
mediate was transformed into chiriquitoxin by a stereocon-
oping a subtype selective blocker of a voltage-gated sodium
trolled aldol reaction with a d-camphor-derived lactone
for installation of the unique side chain, and a new depro-
channel based on tetrodotoxin and saxitoxin to analyze the
biological functions. Concerning this issue, we have been inter-
tection of methylthiomethyl (MTM) ether by using a Pum-
ested in chiriquitoxin, because it is the sole example that pos-
merer rearrangement.
sesses potassium-channel blocking activity among the tetrodo-
toxin family. Herein, we disclose the first asymmetric total syn-
thesis of chiriquitoxin (2) from a newly designed intermediate
In 1975, Mosher and co-workers reported isolation of chiriqui-
toxin, a potent new neurotoxin, along with tetrodotoxin (TTX,
1) from the skin of a Costa Rican dart frog, Atelopus chiriquen-
sis.[1–3] The molecular weight of this new compound was deter-
mined in 1976 by 252Cf plasma desorption mass spectroscopy
to be 393.[4] The NMR spectra were very similar to those of te-
trodotoxin (1), a toxic principle of puffer fish intoxication;[5]
however, the structure had not been elucidated until 15 years
after the discovery. In 1990, Yasumoto, Yotsu–Yamashita, and
co-workers successfully elucidated the structure of chiriquitox-
in (CHTX, 2) as shown in Figure 1 by extensive analysis of NMR
spectra and derivatization reactions.[6] Chiriquitoxin is an ana-
logue of tetrodotoxin, in which a glycine residue is connected
to the C-11 position of tetrodotoxin.
for the syntheses of a wide variety of tetrodotoxin ana-
logues.[19]
Our synthetic plan for chiriquitoxin is illustrated in
Scheme 1.[20] Since the structural difference between tetrodo-
toxin and chiriquitoxin exists at the side chain at the C-11 posi-
tion, we would synthesize chiriquitoxin by addition of a glycine
Figure 1. Structures of tetrodotoxin (1) and chiriquitoxin (2).
Toxicity and inhibitory activity against the voltage-gated
sodium channel of chiriquitoxin resembles those of tetrodotox-
in.[7,8] In 1981, Kao and co-workers reported that chiriquitoxin
is the first tetrodotoxin analogue shown to interfere with the
outwardly rectifying potassium channel.[7b] These characteristic
biological activities of 2 have attracted considerable interest in
[a] Dr. M. Adachi, T. Imazu, R. Sakakibara, Y. Satake, Prof. Dr. M. Isobe,
Prof. Dr. T. Nishikawa
Laboratory of Organic Chemistry
Graduate School of Bioagricultural Sciences,
Nagoya University, Chikusa, Nagoya 464-8601 (Japan)
Fax: (+81)52-789-4111
E-mail: nisikawa@agr.nagoya-u.ac.jp
[b] Prof. Dr. M. Isobe
Department of Chemistry, National Tsing Hua University,
Hsinchu 30013 (Taiwan)
Fax: (+866)3-5736494
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201304110.
Scheme 1. Synthetic plan for chiriquitoxin (2) from a newly designed inter-
mediate 4.
Chem. Eur. J. 2014, 20, 1247 – 1251
1247
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
unit to the aldehyde at the C-11 position of an intermediate.
The required aldehyde A could be obtained by oxidation of
the corresponding hydroxy group at the C-11 position of the
intermediate B, which would also be a promising intermediate
for tetrodotoxin. In order to synthesize the b-hydroxy amino
acid of chiriquitoxin stereoselectively, we selected a d-cam-
phor-derived tricycloiminolactone 3 as a chiral glycine equiva-
lent, because a highly anti-selective aldol reaction with alde-
hyde was reported by Xu and co-workers.[21] The intermediate
B was envisaged to arise from C by oxidative cleavage of the
acetylenic moiety to carboxylic acid and subsequent orthoester
formation. According to the previous synthetic studies in this
laboratory, the acetylene C would be synthesized from vinylep-
oxide D through ozonolysis of the vinyl group followed by ste-
reoselective addition of acetylide as a carboxylic acid equiva-
lent. The oxygen functionalities of D would be installed
through hydroxylation at the C-5 position by allylic oxidation
and subsequent epoxidation from compound 4,[19] a new inter-
mediate possessing two hydroxy groups at the C-8 and C-11
positions for tetrodotoxin and its analogues.
Synthesis began with inversion of the configuration of the
hydroxy group at the C-8 position of the intermediate 4
(Scheme 2). Protection of the primary alcohol with a TBS group
and oxidation with PCC gave enone, which was reduced with
LiAlH4 in ether as a solvent at approximately À110 8C, giving
the desired allylic alcohol 5 in good yield with high stereose-
lectivity (d.r.>20:1).[22] Protection of the resulting alcohol with
a TBS group provided 6. Allylic oxidation at the C-5 position
was next investigated. Upon heating of the bis-TBS ether 6
with SeO2 and pyridine N-oxide in 1,4-dioxane,[10b–c] the desired
allylic alcohol 8 was obtained in 15% yield along with an a,b-
unsaturated aldehyde in 80% yield. Further experiments re-
vealed that the regioselectivity depended on the protective
group of the hydroxy group at the C-11 position.[23] In the
event, the acetate 7 was found to be the best substrate for
the allylic oxidation under the conditions, providing the de-
sired 9 in 52% yield with the unsaturated aldehyde in 22%
yield. Then, 9 was transformed into bis-TBS ether 8 in two
steps because the acetyl group was not compatible for the
subsequent reactions.
Scheme 2. Synthesis of diacetate 14: a) TBSCl, imidazole, DMF, 08C, 15 min,
96%; b) PCC, 4 ꢁ MS, CH2Cl2, RT, 12 h, 89%; c) LiAlH4, Et2O, ca. À110 8C, 0.5 h,
73%; d) TBSOTf, 2,6-lutidine, CH3CN, À108C, 10 min, 95%; e) HF·Py, pyridine,
THF, À208C, 5 h; f) Ac2O, pyridine, RT, 1 h, 86% in 2 steps; g) SeO2, PNO, 1,4-
dioxane, 1008C, 34 h, 52%; h) K2CO3, MeOH, RT, 30 min; i) TBSOTf, 2,6-luti-
dine, CH2Cl2, À208C, 0.5 h, 99% in 2 steps; j) MCPBA, (CH2Cl)2, RT, 19 h, 88%;
k) PMBOC(=NH)CCl3, TfOH, Et2O, 08C, 0.5 h, 86%; l) O3, CH2Cl2, À788C, 5 h,
then Me2S; m) trimethylsilylacetylene, nBuLi, THF, À208C, 0.5 h, 80% in
2 steps; n) Ac2O, pyridine, RT, 2.5 h; o) TBAF, MeOH, THF, À208C, 15 min,
95% in 2 steps; p) DDQ, CH2Cl2, H2O, RT, 4 h, 93%; q) PCC, MS-4 ꢁ, CH2Cl2,
RT, 12 h, 87%; r) NaBH4, MeOH, À208C, 0.5 h; s) Ac2O, pyridine, DMAP, RT,
11 h, 94% in 2 steps. TBS=tert-butyldimethylsilyl, PCC=pyridinium chloro-
chromate, MS=molecular sieves, Tf=trifluoromethanesulfonyl, Py=pyri-
dine, THF=tetrahydrofuran, Ac=acetyl, PNO=pyridine N-oxide, MCPBA=3-
chloroperbenzoic acid, PMB=p-methoxybenzyl, TBAF=tetrabutylammoni-
um fluoride, DDQ=2,3-dichloro-5,6-dicyano-1,4-benzoquinone, DMAP=N,N-
dimethyl-4-aminopyridine.
Epoxidation of 8 with MCPBA and subsequent protection of
the secondary alcohol as a PMB ether gave 10. The vinyl group
was ozonolyzed to afford an aldehyde, which was subjected to
addition of trimethylsilylacetylene as a carboxylic acid equiva-
lent. The aldehyde reacted with the lithium acetylide in THF at
À208C to give a 10:1 diastereomeric mixture, from which the
desired product 11 was isolated in 80% yield in two steps
from 10.[24] Acetylation followed by selective desilylation of the
TMS group afforded propargyl acetate 12. At this stage, the
configuration of the hydroxy group at the C-5 position of 12
was inverted by an oxidation–reduction sequence; deprotec-
tion of the PMB group with DDQ was followed by oxidation
with PCC to give ketone 13. Reduction with NaBH4 provided
a single alcohol, which was isolated as a diacetate 14 in 94%
yield in two steps.
through epoxide opening was next investigated. Oxidation of
14 was best carried out by RuCl3·nH2O and Oxone[25] to give
a mixture of the desired lactone 15 and a-ketocarboxylic acid
16 (Scheme 3).[26] Without purification, the mixture was treated
with alkaline hydrogen peroxide to give the desired carboxylic
acid, which underwent spontaneous opening of the epoxide
to afford an equilibrium mixture of lactones and orthoester.
The mixture was further treated with TIPSOTf in the presence
of 2,6-lutidine to provide the desired orthoester 17 in 68%
overall yield in three steps from 14. To cleave the 1,2-glycol
protected as an acetonide, deprotection of the acetonide was
attempted. Although the selective deprotection in the pres-
ence of acid-sensitive protective groups such as siloxy groups
was difficult, we fortunately found that 17 was treated with
With multigram quantities of 14 in hand, oxidative cleavage
of the acetylenic moiety and subsequent lactone formation
Chem. Eur. J. 2014, 20, 1247 – 1251
www.chemeurj.org
1248
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
Scheme 3. Synthesis of orthoester 19 and 4,9-anhydrotetrodotoxin (21):
a) RuCl3·nH2O, Oxone, NaHCO3, CCl4, CH3CN, H2O, RT, 20 h; b) H2O2 aq.,
NaHCO3, MeOH, RT, 8 h; c) TIPSOTf, 2,6-lutidine, CH2Cl2, 08C, 50 min, 68% in
3 steps; d) SnCl2·2H2O, CH2Cl2, RT, 7 h, 66% (recovered starting material
15%); e) NaIO4, MeOH, H2O, RT, 2.5 h; f) PPTS, CH(OEt)3, EtOH, RT, 12 h, 90%
in 2 steps; g) DIBAL-H, CH2Cl2, À408C, 15 min; h) BocN=C(SMe)NHBoc, HgCl2,
Et3N, DMF, RT, 20 min, 69% in 2 steps; i) HF, H2O, RT, 12 h, 65%. TIPS=triiso-
propylsilyl, PPTS=pyridine p-toluenesulfonate, DIBAL-H=diisobutylalumini-
um hydride, Boc=tert-butoxycarbonyl.
[27]
SnCl2·2H2O in CH2Cl2 to give 1,2-diol 18 in 66% yield along
Scheme 4. Synthesis of an aldehyde 25, a precursor for aldol reaction, and
total synthesis of chiriquitoxin (2): a) BnOH, Na2CO3, DMF, 1508C, 2 h, 85%;
b) 0.05n HCl aq., EtOH, RT, 22 h, 51% (recovered starting material 33%);
c) DMSO, Ac2O, RT, 22 h, 96%; d) iminolactone 3, LDA, LiCl, THF, À208C,
40 min, 69%; e) H2, Pd/C, EtOH, RT, 17 h; f) BocN=C(SMe)NHBoc, HgCl2,
Et3N, DMF, RT, 20 min, 67% in 2 steps; g) MCPBA, CH2Cl2, 08C, 10 min;
h) TFAA, 2,6-lutidine, CH2Cl2, 08C, 1.5 h; i) NH3 aq., EtOH, 08C, 1 h, 37% in
3 steps; j) HF, H2O, RT, 14 h; k) TFA, H2O, 408C, 4 days; l) pyridine, H2O, 358C,
12 h, 41% for CHTX (2), 28% for 4,9-anhydroCHTX (32). Cbz=benzyloxycar-
bonyl, Bn=benzyl, DMSO=dimethylsulfoxide, LDA=lithium diisopropyla-
mide, TFAA=trifluoroacetic anhydride, TFA=trifluoroacetic acid.
with the recovered starting material 17 in 15% yield. Cleavage
of the 1,2-diol of 18 with sodium periodate and subsequent
treatment with PPTS and triethyl orthoformate in ethanol pro-
vided intramolecular acetal 19 as a single diastereomer.[28] This
route is robust and scalable, which allowed us to synthesize
approximately 2 g of 19 to date. Since the product is a fully
functionalized intermediate for tetrodotoxin except for guani-
dine functionality, transformation of 19 into tetrodotoxin was
investigated. The guanidine was installed in two steps; 1) de-
protection of the N-trichloroacetyl group with DIBAL-H[29] and
2) guanidinylation of the resulting amine with N,N’-bis-Boc-S-
methyl-isothiourea in the presence of HgCl2, to give di-Boc
guanidine 20a in 69% overall yield. All the protecting groups
of 20a were deprotected with aqueous HF to afford 65% yield
of 4,9-anhydrotetrodotoxin (21),[30] which could be transformed
into tetrodotoxin (1) by hydrolysis.[31]
developed in this laboratory;[32] heating with benzyl alcohol
and Na2CO3 in DMF at reflux temperature gave benzyl carba-
mate 22 in 85% yield. Selective deprotection of the TBS group
at the C-11 hydroxy group was best carried out by treatment
with aqueous HCl in ethanol to provide primary alcohol 23 in
51% yield along with the recovered starting material 22 in
33% yield. Unexpectedly, oxidation of the primary alcohol to
aldehyde proved to be very difficult. Most of the oxidants
tested gave low yields of the desired product 24 along with
many byproducts including CÀC bond cleavaged products.
Eventually, Ac2O/DMSO (Albright–Goldman oxidation) was
found to be the sole condition for yielding aldehyde in an ex-
Towards the total synthesis of chiriquitoxin, an indispensable
intermediate containing aldehyde at the C-11 position was pre-
pared from 19 as shown in Scheme 4. Since deprotection of
the N-trichloroacetyl group would be extremely difficult after
aldol reaction, the N-trichloroacetyl group of 19 was first trans-
formed into benzyl carbamate in one-pot under the conditions
Chem. Eur. J. 2014, 20, 1247 – 1251
www.chemeurj.org
1249
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
cellent yield. Although the tertiary alcohol was protected as
a methylthiomethyl (MTM) ether under the conditions,[33] the
resulting aldehyde 25 was used for the next aldol reaction
with iminolactone 3.
the side chain, and also enables supply of chiriquitoxin and its
derivatives for biochemical investigations. Biochemical investi-
gations using the synthesized chiriquitoxin and its intermedi-
ate are underway.
To our delight, when the aldol reaction of 25 with the lithi-
um enolate of the iminolactone 3 was carried out in the pres-
ence of LiCl in THF at À208C according to the methods report-
ed in the literature,[21] the desired anti-adduct 26 was obtained
as a single product, which was thought to be formed on the
basis of the transition state proposed by Xu and co-work-
ers.[21,34] Interestingly, aldol reaction of the aldehyde 24 under
the same conditions gave a complex mixture of products,
which indicated the importance of the MTM ether.[35]
Acknowledgement
This work was financially supported by Grants-in-Aid for Scien-
tific Research on Innovative Areas “Chemical Biology of Natural
Products”, for Young Scientists (Start-up and B) and for Scientif-
ic Research (B) from MEXT, the Sumitomo Foundation, the
Yamada Science Foundation, the Daiichi Sankyo Foundation of
Life Science, and a SUNBOR GRANT from the Suntory Institute
for Bioorganic Research. We thank Dr. T. Nishikawa (Bruker Bio-
Spin) for assistance with 13C NMR spectroscopic measurement
of chiriquitoxin (2). We are also grateful to Prof. M. Yotsu-Yama-
shita (Tohoku University) for providing the spectra of natural
chiriquitoxin.
Installation of the guanidine functionality was straightfor-
ward; hydrogenolytic deprotection of the Cbz group and sub-
sequent guanidinylation of the resulting amine with N,N’-bis-
Boc-S-methyl-isothiourea in the presence of HgCl2 gave di-Boc
guanidine 27, a synthetic equivalent to chiriquitoxin, in good
overall yield. The remaining issue for the total synthesis was re-
moval of all the protective groups of 27. However, attempted
deprotection of the MTM group under the conventional condi-
tions, such as HgCl2 or AgNO3 in aqueous solvents,[36] was un-
successful. The considerable difficulties encountered at this
stage prompted us to develop a new deprotection method of
the MTM group based on a different deprotection mechanism.
We envisaged that the Pummerer rearrangement of the sulfox-
ide of a MTM ether would provide a thioacetal intermediate,
which collapses to release the corresponding alcohol. In prac-
tice, the MTM group of 27 was oxidized with MCPBA, and the
resulting sulfoxide 28 (2.2:1 diastereomeric mixture) was treat-
ed with trifluoroacetic anhydride and 2,6-lutidine, a condition
of Pummerer rearrangement. Hydrolysis of the product 29
with aqueous ammonia provided the desired diol 30 in 37%
overall yield in three steps from 27.
Keywords: chiriquitoxin
rearrangement · tetrodotoxin · total synthesis
· natural products · Pummerer
[1] Y. H. Kim, G. B. Brown, H. S. Mosher, F. A. Fuhrman, Science 1975, 189,
151–152.
[2] CHTX was isolated from the eggs of the Costa Rican frog, Atelopus chiri-
quiensis, as a mixture with TTX, see: L. A. Pavelka, Y. H. Kim, H. S.
Mosher, Toxicon 1977, 15, 135–139.
[3] CHTX was also isolated from the skins of Panamanin toads, Atelopus li-
mosus and Atelopus glyphus, see: M. Yotsu-Yamashita, E. Tateki, Toxicon
2010, 55, 153–156.
[4] R. D. Macfarlane, D. F. Torgerson, Science 1976, 191, 920–925.
[5] a) T. Goto, Y. Kishi, S. Takahashi, Y. Hirata, Tetrahedron 1965, 21, 2059–
2088; b) K. Tsuda, S. Ikuma, M. Kawamura, R. Tachibana, K. Sakai, C.
Tamura, O. Amakasu, Chem. Pharm. Bull. 1964, 12, 1357–1374; c) R. B.
Woodward, Pure Appl. Chem. 1964, 9, 49–74.
[6] M. Yotsu, T. Yasumoto, Y. H. Kim, H. Naoki, C. Y. Kao, Tetrahedron Lett.
1990, 31, 3187–3190.
[7] a) C. Y. Kao, C. L. Cohen, F. A. Fuhrman, Fed. Proc. 1976, 35, 846; b) C. Y.
Kao, P. N. Yeoh, M. D. Goldfinger, F. A. Fuhrman, H. S. Mosher, J. Pharma-
col. Exp. Ther. 1981, 217, 416–429.
All the other acid-sensitive protective groups of 30 were re-
moved by treatment with aqueous HF, providing 4,9-anhydro-
chiriquitoxin-13,6-lactone (31) and 4,9-anhydrochiriquitoxin
1
(32) in a 92:8 ratio (from H NMR spectroscopy). The 4,9-anhy-
[8] The toxicity of CHTX reported by Mosher in 1975 is 4000–
6000 MUmgÀ1 (LD50: 8.3–12.5 mgkgÀ1), see ref. [1]. The LD50 values in
mice were reported by Mosher (LD50 in mice, CHTX (2): 13 mgkgÀ1, TTX
dro moiety of both compounds was hydrolyzed with 2% TFA/
H2O at 408C for 4 days, and then the lactone was hydrolyzed
with 0.1% pyridine/H2O at 358C for 12 h to furnish chiriquitox-
in (2) and 4,9-anhydrochiriquitoxin (32) in 41% and 28% yield,
respectively, after purification by HPLC on an ion-exchange
(1): 10 mgkgÀ1) and Yotsu-Yamashita (LD50 in mice, CHTX (2): 14 mgkgÀ1
,
TTX (1): 10 mgkgÀ1) independently, see: F. A. Fuhrman, G. L. Fuhrman,
Y. H. Kim, H. S. Mosher, Proc. West. Pharmacol. Soc. 1976, 19, 381–384
and ref. [6].
1
resin column. The H and 13C NMR spectra of the synthesized
[9] a) T. Narahashi, Physiol. Rev. 1974, 54, 813–889; b) F. Hucho, Angew.
Chem. 1995, 107, 23–36; Angew. Chem. Int. Ed. Engl. 1995, 34, 39–50.
[10] For asymmetric total synthesis of tetrodotoxin in this laboratory, see:
a) N. Ohyabu, T. Nishikawa, M. Isobe, J. Am. Chem. Soc. 2003, 125,
8798–8805; b) T. Nishikawa, D. Urabe, M. Isobe, Angew. Chem. 2004,
116, 4886–4889; Angew. Chem. Int. Ed. 2004, 43, 4782–4785; c) D.
Urabe, T. Nishikawa, M. Isobe, Chem. Asian J. 2006, 1–2, 125–135.
[11] a) T. Nishikawa, M. Asai, N. Ohyabu, N. Yamamoto, M. Isobe, Angew.
Chem. 1999, 111, 3268–3271; Angew. Chem. Int. Ed. 1999, 38, 3081–
3084; b) M. Asai, T. Nishikawa, N. Ohyabu, N. Yamamoto, M. Isobe, Tetra-
hedron 2001, 57, 4543–4558.
[12] M. Adachi, T. Imazu, M. Isobe, T. Nishikawa, J. Org. Chem. 2013, 78,
1699–1705.
[13] T. Nishikawa, M. Asai, M. Isobe, J. Am. Chem. Soc. 2002, 124, 7847–7852.
[14] a) T. Nishikawa, D. Urabe, K. Yoshida, T. Iwabuchi, M. Asai, M. Isobe, Org.
Lett. 2002, 4, 2679–2682; b) T. Nishikawa, D. Urabe, K. Yoshida, T. Iwabu-
chi, M. Asai, M. Isobe, Chem. Eur. J. 2004, 10, 452–462.
chiriquitoxin were identical to those of the natural chiriquitox-
in.[6] The synthetic [a]D value ([a]3D1 =À14.6 (c=0.185, 0.05 n
AcOH)) was also consistent with that of reported data ([a]2D5 =
À17.3 (c=0.075, 0.05 n AcOH)).[6]
In conclusion, we have successfully achieved the first total
synthesis of chiriquitoxin, the most structurally complex ana-
logue of tetrodotoxin. The success of this total synthesis relied
on the new efficient synthetic route of tetrodotoxin from the
newly designed intermediate 4, stereocontrolled aldol reaction
of a d-camphor-derived iminolactone for introduction of the
glycine unit, and new deprotection procedure of MTM ether
by Pummerer rearrangement. This total synthesis confirmed
the structure of chiriquitoxin including the configurations of
Chem. Eur. J. 2014, 20, 1247 – 1251
www.chemeurj.org
1250
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
[15] T. Nishikawa, M. Isobe, Chem. Rec. 2013, 13, 286–302.
[16] The first total synthesis of racemic tetrodotoxin was reported by Kishi
and co-workers, see: a) Y. Kishi, M. Aratani, T. Fukuyama, F. Nakatsubo, T.
Goto, S. Inoue, H. Tanino, S. Sugiura, H. Kakoi, J. Am. Chem. Soc. 1972,
94, 9217–9219; b) Y. Kishi, T. Fukuyama, N. Aratani, F. Nakatsubo, T.
Goto, S. Inoue, H. Tanino, S. Sugiura, H. Kakoi, J. Am. Chem. Soc. 1972,
94, 9219–9221.
[17] For leading references on the total synthesis of tetrodotoxin and its an-
alogues from other laboratories, see: a) A. Hinman, J. Du Bois, J. Am.
Chem. Soc. 2003, 125, 11510–11511; b) K. Sato, S. Akai, N. Sugita, T.
Ohsawa, T. Kogure, H. Shoji, J. Yoshimura, J. Org. Chem. 2005, 70, 7496–
7504; c) K. Sato, S. Akai, H. Shoji, N. Sugita, S. Yoshida, Y. Nagai, K.
Suzuki, Y. Nakamura, Y. Kajihara, M. Funabashi, J. Yoshimura, J. Org.
Chem. 2008, 73, 1234–1242; d) S. Akai, H. Seki, N. Sugita, T. Kogure, N.
Nishizawa, K. Suzuki, Y. Nakamura, Y. Kajihara, J. Yoshimura, K. Sato, Bull.
Chem. Soc. Jpn. 2010, 83, 279–287.
ed by our laboratory.[10] Transformation of the 4,9-anhydrotetrodotoxin
(21) to tetrodotoxin (1) under acidic conditions was reported.[5a,10].
[32] D. Urabe, K. Sugino, T. Nishikawa, M. Isobe, Tetrahedron Lett. 2004, 45,
9405–9407.
[33] Y. Yamada, K. Kato, H. Nagase, Y. Hirata, Tetrahedron Lett. 1976, 17, 65–
66.
[34] The stereochemical outcome of the aldol reaction was confirmed by
the following experiments; aldol reaction of the trichloroacetyl-protect-
ed aldehyde 33 with lithium enolate of the iminolactone 3 under the
same conditions provided the aldol adduct 34a as a single product. At-
tempted protection of the secondary alcohol at the C-11 position with
PMB gave a methylene acetal 35, instead of the desired PMB ether
34b. Extensive analysis of the NMR spectra including NOESY deter-
mined the configurations of the newly generated asymmetric centers
at the C-11 and C-12 positions in the aldol reaction. Since the coupling
constants between H-11 and H-12 of both the aldol adducts 26 and
34a were the same (J11,12 =8.0 Hz), we concluded that the configura-
tions at the C-11 and C-12 positions of 26 are as shown in 26.
[18] For a review on the chemical synthesis of tetrodotoxin, see: J. Chau,
M. A. Ciufolini, Mar. Drugs 2011, 9, 2046–2074.
[19] Y. Satake, T. Nishikawa, T. Hiramatsu, H. Araki, M. Isobe, Synthesis 2010,
1992–1998.
[20] The numbering used in this paper corresponds to that of chiriquitoxin.
[21] Q. Li, S.-B. Yang, Z. Zhang, L. Li, P.-F. Xu, J. Org. Chem. 2009, 74, 1627–
1631.
[22] When the enone was exposed to conditions of a Luche reduction
(NaBH4, CeCl3 in EtOH/CH2Cl2), the previously optimized conditions for
stereoselective reduction of a similar enone intermediate lacking a hy-
droxy group at the C-11 position,[11] a low conversion with a moderate
stereoselectivity (d.r.=4:1) was observed.
[23] The reason for the regioselectivity of the allylic oxidation has not been
clarified.
[24] For details of determination of the configuration at the C-9 position,
see the Supporting Information.
[25] D. Yang, F. Chen, Z.-M. Dong, D.-W. Zhang, J. Org. Chem. 2004, 69,
2221–2223.
[26] The attempted oxidative cleavage of the acetylenic moiety of 14 with
KMnO4 and NaIO4 in aqueous tert-butanol, which had been employed
in our second total synthesis of tetrodotoxin,[10b–c] gave neither the car-
boxylic acid nor the desired lactone.
[27] W.-B. Yang, S. S. Patil, C.-H. Tsai, C.-H. Lin, J.-M. Fang, Tetrahedron 2002,
58, 253–259.
[28] The C-4 configuration of 19 was established to be S by the coupling
constant between H-4 and H-4a.[13].
[29] T. Oishi, K. Ando, K. Inomoya, H. Sato, M. Iida, N. Chida, Org. Lett. 2002,
4, 151–154.
[35] To maximize the synthetic efficiency, an aldehyde possessing guanidine
from 20 seemed to be one of the ideal substrate for the aldol reaction.
However, the aldehyde prepared from 20b did not undergo the aldol
reaction under the same conditions to give a complex mixture of prod-
ucts.
[36] E. J. Corey, M. G. Bock, Tetrahedron Lett. 1975, 16, 3269–3270.
[30] For NMR spectral data of 4,9-anhydrotetrodotoxin (21), see: M. Naka-
mura, T. Yasumoto, Toxicon 1985, 23, 271–276.
[31] The spectroscopic data of the synthesized 4,9-anhydrotetrodotoxin (21)
were identical to those of the natural product[30] as well as those report-
Received: October 21, 2013
Chem. Eur. J. 2014, 20, 1247 – 1251
www.chemeurj.org
1251
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim