Turbotoxins A and B, novel diiodotyramine derivatives from the Japanese gastropod Turbo marmorata

Article abstract of DOI:10.1016/S0040-4020(00)00759-6

Bioassay-guided separation of the aqueous ethanol extract of the viscera of the Japanese gastropod Turbo marmorata resulted in the isolation of two toxins, turbotoxins A and B. Their structures were determined by spectral analysis and confirmed by organic synthesis to be diiodotyramine derivatives. Turbotoxins A and B exhibited acute toxicity against ddY mice, with LD₉₉ values of 1.0 and 4.0 mg/kg, respectively. The structure-toxicity relationships of turbotoxins were examined, and it was proved that the iodine atoms and trimethylammonium groups are important for its acute toxicity. Turbotoxin A inhibits acetylcholinesterase with an IC₅₀ of 28 μM. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4020(00)00759-6

TETRAHEDRON
Pergamon
Tetrahedron 56 (2000) 9063±9070
Turbotoxins A and B, Novel Diiodotyramine Derivatives from the
Japanese Gastropod Turbo marmorata
b
b
b,
Hideo Kigoshi,a, Kengo Kanematsu, Kyoko Yokota and Daisuke Uemura
*
*
aResearch Center for Materials Science, Nagoya University, Chikusa, Nagoya 464-8602, Japan
bDepartment of Chemistry, Graduate School of Science, Nagoya University, Chikusa, Nagoya 464-8602, Japan
Dedicated to Professor Paul J. Scheuer on the occasion of his 85th birthday
Received 10 July 2000; accepted 28 July 2000
AbstractÐBioassay-guided separation of the aqueous ethanol extract of the viscera of the Japanese gastropod Turbo marmorata resulted in
the isolation of two toxins, turbotoxins A and B. Their structures were determined by spectral analysis and con®rmed by organic synthesis to
be diiodotyramine derivatives. Turbotoxins A and B exhibited acute toxicity against ddY mice, with LD99 values of 1.0 and 4.0 mg/kg,
respectively. The structure±toxicity relationships of turbotoxins were examined, and it was proved that the iodine atoms and trimethyl-
ammonium groups are important for its acute toxicity. Turbotoxin A inhibits acetylcholinesterase with an IC50 of 28 mM. q 2000 Elsevier
Science Ltd. All rights reserved.
Human intoxication resulting from the ingestion of
shell®sh occurs worldwide, and novel compounds have
been isolated from toxic shell®sh. Toxins from shell®sh,
such as pinnatoxins,1 saxitoxin,2 and neosurugatoxin,3
have attracted interest not only from pharmacologists
but also from biochemists and chemists due to their
novel biological activities and structures (Fig. 1). In
Japan, the gastropod Turbo marmorata is eaten after
the viscera, which cause intoxication, are removed.
Yasumoto and coworkers studied the toxic components of
T. marmorata and obtained several toxic fractions from
this animal. They indicated the occurrence of saxitoxin
and a minor toxin, iodomethyltrimethylammonium salt,
in the water-soluble fraction of T. marmorata.4 Although
they mentioned the presence of other toxins in this
animal,5 not all of them have been identi®ed. We
describe herein the isolation, structures, synthesis, and
structure±toxicity relationships of turbotoxins A and B,
the marine toxins from the Japanese gastropod T. marmorata
(Fig. 2).6
Figure 1. Structures of toxins from shell®sh.
Keywords: toxins; ammonium salts; biologically active compounds; marine metabolites; structure±activity.
* Corresponding authors. Tel.: 181-52-789-2479; fax: 181-52-789-5041; e-mail: kigoshi@chem3.chem.nagoya-u.ac.jp
0040±4020/00/$ - see front matter q 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(00)00759-6

9064
H. Kigoshi et al. / Tetrahedron 56 (2000) 9063±9070
to the scarcity of the sample, the carbon chemical shift was
determined by HMBC (JC±Hˆ6 Hz) experiments.
1
The H NMR data showed the presence of two trimethyl-
ammonium groups. The COSY spectrum allowed ±(CH2)2±
and ±(CH2)3± chains to be constructed. The NMR signals at
dH 7.86 and at dC 92.4, 139.1, 142.7, and 158.8 were
assigned to a 4-substituted-2,6-diiodophenoxy unit by
comparison with those of thyroxine: the unusually high
®eld of the signal at dC 92.4 clearly distinguished the sp2
carbon atoms linked to an iodine atom. The HMBC correla-
tions (H-1/1-NCH3, 1-NCH3/C1, H-2/C3, H-2/C4, H-4/C2,
H-4/C40, H-4/C5, H-4/C6, H-9/9-NCH3, 9-NCH3/C9)
allowed the foregoing partial structures to be connected,
giving two partial structures, Me3N1CH2CH2C6H2I2O±
Figure 2. Structures of turbotoxins A (1) and B (2).
Isolation and Structures
The 75% aqueous ethanol extract of the viscera (4.5 kg, 36
individuals) of T. marmorata collected in Okinawa, Japan,
was partitioned between ethyl acetate and water. The
aqueous layer was chromatographed using bioassay-guided
(intraperitoneal mouse lethality) fractionation, to give two
toxic fractions. The early toxic fraction was puri®ed by
HPLC to give turbotoxin A (1) (2.0 mg; LD99 1.0 mg/kg),
and the late toxic fraction was puri®ed in the same way to
and ±CH2CH2CH2N1Me3. Finally,
a weak HMBC
correlation between H-7 and C6 completed the structure
of turbotoxin A, as shown in Fig. 2.
Turbotoxin B (2) was found to be a demethylated analog of
1, based on its molecular formula, C16H 28I2N2O´(CF3COO)2
(HRFABMS m/z 517.0209 [M2CF3COOH±CF3COO]1, D
1
20.4 mmu). Comparison of the H NMR data of 2 with
give turbotoxin
B
(2) (0.9 mg, LD99 4.0 mg/kg).
those of 1 indicated that a trimethylammonium group at
C9 in 1 was demethylated to a dimethylamino group in 2
(Table 1). Furthermore, correlations suf®cient to establish
the structure of turbotoxin B were found in the HMBC
spectrum of 2. Thus, the structure of turbotoxin B was
determined to be 2.
Compounds 1 and 2 were isolated as tri¯uoroacetate salts
because of the solvent system used for chromatographic
puri®cation.
The molecular formula of turbotoxin A (1) was found to be
C17H30I2N2O´(CF3COO)2 (m/z 645.0286 [M2CF3COO]1,
D 21.2 mmu) by HRFABMS. Resonances in the NMR
spectra were assigned based on the COSY, HSQC, and
HMBC spectra, as shown in Table 1. Although the carbon
signal at C3 was not observed in the 13C NMR spectrum due
Synthesis
To con®rm the structures of turbotoxins A (1) and B (2), and
to subject 1 and 2 to further biological examinations, 1 and 2
were synthesized (Scheme 1). Tyramine was methylated by
a standard procedure to give N,N-dimethyltyramine (3d).
Diiodination of 3d was effected with I2±Hg(OAc)2 to afford
2,6-diiodo-N,N-dimethyltyramine (3a) in 91% yield. This
conversion without mercury salts was achieved with I2±t-
BuNH27 in lower yield. Alkylation of 3a with chloropropyl-
dimethylamine hydrochloride in the presence of Cs2CO3
afforded diamine 4a8 (58%). Diamine 4a was methylated
with 1 molar equivalent of iodomethane to give turbotoxins
A (1, 5%) and B (2, 17%) along with turbotoxin B isomer 5
(22%) and diamine tri¯uoroacetate salt 6 (38%), which were
separated by HPLC. The spectral data and biological
activities of synthetic turbotoxins A (1) and B (2) were
identical to those of natural 1 and 2, respectively, which
con®rmed the structures and toxicity of turbotoxins A and
B. With an excess of iodomethane, 4a was transformed into
1 quantitatively.
Table 1. NMR data of turbotoxins A (1) and B (2) in CD3OD
1
position
1Ha
13Cb
HMBC (13C! H)c
turbotoxin A (1)
1
1-NMe
3.52 (m, 2H)
3.18 (s, 9H)
3.05 (m, 2H)
68.5
54.5
29.1
1-NMe, H-2
1-NMe, H-1
H-1, 4
2
3
139.1d H-1, 2
4, 40
7.86 (s, 2H)
142.7
92.4
158.0
71.1
26.0
66.6
54.5
H-2, 4
H-4
H-4, 7
5, 50
6
7
8
4.10 (t, Jˆ5.5 Hz, 2H)
2.42 (m, 2H)
3.78 (m, 2H)
H-7, 9
9-NMe, H-7, 8
9-NMe, H-9
9
9-NMe
turbotoxin B (2)
1
1-NMe
3.22 (s, 9H)
3.52 (m, 2H)
3.18 (s, 9H)
3.05 (m, 2H)
68.5c 1-NMe, H-2
54.5
29.1
1-NMe, H-1
H-1, 4
2
3
138.0d H-1, 2
4, 40
7.86 (s, 2H)
142.6
92.5
H-2, 4
H-4
To investigate the structure±toxicity relationships of
turbotoxins, ten analogs were synthesized. By using the
same strategy as for turbotoxin A (1), the bromo, chloro,
and hydro analogs, 7, 8, and 9, were prepared from
N,N-dimethyltyramine (3d) as illustrated in Scheme 1.
The benzyl ammonium analogs, 11, 12, and 13, and 4-
phenylbenzyl analogs, 14, 15, and 16, were prepared from
diamine 4a by treatment of 1 molar equivalent of benzyl
bromide or 4-phenylbenzyl bromide followed by an excess
of iodomethane, respectively.
5, 50
6
7
8
158.0d H-4
71.7
4.10 (t, Jˆ5.6 Hz, 2H)
2.33 (m, 2H)
3.55 (m, 2H)
27.1
58.1
44.5
H-7, 9
9-NMe, H-7, 8
9-NMe, H-9
9
9-NMe
2.98 (s, 6H)
a The spectrum was recorded at 600 MHz.
b The spectrum was recorded at 100 MHz.
c Protons correlated to carbon resonances in 13C column.
d Chemical shifts were determined by HSQC and HMBC experiments.

H. Kigoshi et al. / Tetrahedron 56 (2000) 9063±9070
9065
Scheme 1. Synthesis of natural and arti®cial analogs of turbotoxins.
Table 2. Acute toxicity of turbotoxins A (1) and B (2) and their analogs
Dimethyl analog 10 was prepared from 3,5-dimethyl-4-
hydroxybenzaldehyde as follows. 3,5-Dimethyl-4-hydroxy-
benzaldehyde was converted into a cyanohydrin TMS ether,
which was then reduced to the amino alcohol 17 (38%). The
amino alcohol 17 was methylated and deoxygenated with
Et3SiH±CF3COOH to give 3,5,N,N-tetramethyltyramine
(3e) in 70% yield. The compound 3e was transformed
into 10 by the same sequence of reactions as described
above.
Compound
Acute toxicity (LD99, mg/kg)a
Turbotoxin A (1)
Turbotoxin B (2)
5
6
Bromo analog 7
1.0
4.0
8.0
100
4.0
8.0
12
Chloro analog 8
Hydro analog 9
Methyl analog 10
Dibenzyl analog 11
Monobenzyl analog 12
Monobenzyl analog 13
Diphenylbenzyl analog 14
Monophenylbenzyl analog 15
Monophenylbenzyl analog 16
8.0
4.0
2.0
0.5
.32
32
Structure±Toxicity Relationships
The relationships between the structure of turbotoxins and
their acute toxicity (intraperitoneal mouse lethality) were
studied (Table 2).
8.0
a Upon intraperitoneal injection into ddY mice (n$4).

9066
H. Kigoshi et al. / Tetrahedron 56 (2000) 9063±9070
Compound 6 showed the weakest toxicity among the
compounds 1, 2, 5 and 6, indicating that the quaternary
ammonium group is responsible for the toxicity. The
toxicities of turbotoxin B (2) (4.0 mg/kg) and isomer 5
(8.0 mg/kg) are weaker than that of 1 and stronger than that
of 6. This ®nding shows that the number of the quaternary
ammonium groups is important to the toxicity of turbotoxins.
analogs, however, the benzyl group in 13 might be
stacked against the aromatic gorge to increase its toxicity.
Preliminary neuropharmacological experiments were
effected for turbotoxin A (1), and 1 was proved not to
interact with the peripheral nervous system. Further
studies are required for understanding the mode of action
of turbotoxins.
The comparison of toxicities of turbotoxin A (1), bromo
analog 7, chloro analog 8, and hydro analog 9 (1.0, 4.0,
8.0, and 12 mg/kg) indicates the importance of iodine
atoms in the toxicity of 1. The toxicity of methyl analog 10
(8.0 mg/kg) is more potent than that of hydro analog 9 and the
same as that of chloro analog 8, indicating the importance of
the steric bulkiness of the C5,50 substituent.
In summary, we isolated two marine toxins, turbotoxins A
(1) and B (2), from the Japanese gastropod T. marmorata.
The structures of turbotoxins were determined by spectral
analysis to be diiodotyramine derivatives, and this was
con®rmed by organic synthesis. Turbotoxins A (1) and B
(2) are structurally related to dakaramine (4a)10 and
dibromotyramine derivatives, such as aplysamine-I,11
moloka'iamine,12 and ceratinamine,13 which were all
isolated from marine sponges.
The effect of the substituents of the quaternary ammonium
moieties was examined. While the toxicities of dibenzyl and
monobenzyl analogs, 11 and 12, were weaker than that of 1,
monobenzyl analog 13 exhibited the toxicity stronger than
that of 1. The benzyl substituent at the N-9 ammonium
group increases its toxicity two-fold. The phenylbenzyl
analogs, 14±16, exhibited weaker toxicities than turbotoxin
A (1) and its benzyl analogs, 11±13. These facts indicated
that the bulky 4-phenylbenzyl group prevents the phenyl-
benzyl analogs from binding to target molecules.
The plausible biosynthesis of turbotoxins is illustrated in
Scheme 2. 3,5-Diiodotyrosine, found in corals, sponges,
algae, and other marine organisms, might be important for
the biosynthesis of turbotoxins, considering the food-habits
of T. marmorata. Also, candicine, a quaternary ammonium
salt of tyramine isolated from T. argyrostoma,14 is a
plausible precursor. On the other hand, [3-(dimethyl-
sulfonio)propyl]trimethylammonium salt was isolated
from the same animal,14 which might be a source of the
C7±C9 unit of turbotoxins.
The target biomolecule(s) of turbotoxin A was investigated,
and it was found that turbotoxin A (1) inhibits acetyl-
cholinesterase with an IC50 of 28 mM. X-Ray crystallo-
graphic studies of complexes of acetylcholinesterase with
small molecules, such as decamethonium bromide,
tacrine, and edrophonium bromide, indicated that the
aromatic gorge exists at the bottom of the active site.9
There is as yet no data of relationships between the
toxicity and af®nity to acetylcholinesterase of turbotoxin
Experimental
General
UV spectra were recorded on a JASCO V-550 spectro-
photometer. NMR spectra were recorded on a JEOL
Scheme 2. A plausible biogenesis of turbotoxin A (1).

H. Kigoshi et al. / Tetrahedron 56 (2000) 9063±9070
9067
1
ALPHA600 (600 MHz for H and 150 MHz for 13C), a
Turbotoxin B (2). UV (MeOH) lmax 225 (e 11000), 239
JEOL ALPHA400 (400 MHz for 1H and 100 MHz for
(sh, 5500), 279 nm (sh, 1500); H and 13C NMR data, see
1
13C), or a JEOL EX270 (270 MHz for H). NMR chemical
Table 1; FABMS m/z 517 [M2CF3COOH±CF3COO]1;
HRFABMS calcd for C16H27I2N2O m/z 517.0213
[M2CF3COOH±CF3COO]1, found 517.0209.
1
shifts were referenced to solvent peaks: dH 3.30 (residual
CHD2OD) and dC 49.7 for CD3OD. Mass spectra were
determined on a JEOL JMS LG2000 spectrometer operating
in the FAB mode (glycerin as a matrix). TLC analysis was
conducted on 0.25 mm E. Merck precoated silica gel 60
F254. Fuji Silysia silica gel BW-820 MH was used for
column chromatography unless otherwise noted. Prepara-
tive HPLC and medium-pressure liquid chromatography
(MPLC) were performed using JASCO 880 or Tosoh
CCPM-II pumps. Unless otherwise stated, materials were
obtained from commercial suppliers and used without
further puri®cation.
N,N-Dimethyltyramine (3d). To a solution of tyramine
(546 mg, 3.98 mmol) in MeOH (30 mL) were added 37%
aqueous formalin (1.19 mL, 15.9 mmol) and 10% Pd/C
(421 mg), and the mixture was stirred at room temperature
under hydrogen. The reaction mixture was ®ltered through a
pad of Celite, and the residue was washed with MeOH
(300 mL). The ®ltrate and the washings were combined
and concentrated to give N,N-dimethyltyramine (3d,
611 mg, 93%) as colorless crystals.
N,N-Dimethyldiiodotyramine (3a). (A) To a solution of
N,N-dimethyltyramine (3d, 36.8 mg, 0.223 mmol) in
EtOH (8.9 mL) were added iodine (121 mg, 0.475 mmol)
and Hg(OAc)2 (37 mg, 0.11 mmol). The mixture was stirred
at room temperature for 30 min. The reaction mixture was
diluted with saturated aqueous Na2S2O3 (20 mL) and
extracted with EtOAc (2£10 mL). The combined extracts
were washed with brine, dried, and concentrated. The resi-
due was chromatographed on silica gel (9 g, CHCl3±MeOH
3:1!1:1!1:3) to give N,N-dimethyldiiodotyramine (3a,
84.2 mg, 91%) as a colorless solid.
Isolation of turbotoxins A and B
Internal organs (4.5 kg) of Turbo marmorata (36 indi-
viduals) collected off the coast of the Okinawan Islands,
Japan, in 1998 and 1999 were crushed in a blender in
75% aqueous EtOH and extracted with the same solvent
(9 L). The alcoholic extract was evaporated and partitioned
between EtOAc (12 L) and H2O (4 L). The aqueous layer
was chromatographed on TSK G-3000S polystyrene gel
(74£145 mm, Tosoh Co., Japan) using aqueous ethanol.
The materials from the eluates of 25% aqueous EtOH and
50% aqueous EtOH that were proved to be toxic against
ddY mice (intraperitoneal injection) were chromatographed
using bioassay-guided (intraperitoneal mouse lethality)
fractionation as follows. The oily residue (5.18 g) from
50% aqueous EtOH was dissolved in 30% aqueous
MeOH, and the insoluble materials were ®ltered. The oil
obtained from the ®ltrate was chromatographed on MPLC
(ODS 160 g) with 30% aqueous MeOH containing 0.1%
tri¯uoroacetic acid (TFA). The early fraction (1.19 g) was
chromatographed on MPLC (ODS 160 g) with increasing
amounts (10!60%) of MeOH in H2O containing 0.1%
TFA. The fraction (152 mg) eluted with 25±45% aqueous
MeOH containing 0.1% TFA was further separated with
MPLC (ODS 22 g) with 15% aqueous MeOH containing
0.1% TFA to give two toxic fractions.
(B) To a solution of t-BuNH2 (0.07 mL, 0.635 mmol) in
toluene (0.2 mL) was added a 0.23 M solution of iodine in
toluene (1.4 mL, 0.327 mmol) over 20 min at 2788C. The
mixture was warmed to 08C, and N,N-dimethyltyramine
(3d, 10.5 mg, 63.5 mmol) was added. The mixture was
then warmed to room temperature over 5 h and stirred for
a further 7 h. The reaction was quenched by adding 70%
MeSH in MeOH (0.4 mL), and the mixture was stirred at
room temperature for 10 min and concentrated. The
residual solid was puri®ed by HPLC (Develosil PhA-5,
20£250 mm, 45% aqueous MeOH containing 0.1%
TFA, 5.0 mL/min, UV 254 nm) to give N,N-dimethyl-
diiodotyramine (3a) as the tri¯uoroacetate salt (22.7 mg,
1
67%) as a colorless oil: H NMR (270 MHz, CD3OD) d
7.68 (s, 2H), 3.35±3.25 (m, 2H), 2.94±2.86 (m, 2H), 2.90
(s, 6H); MS (FAB) m/z 418 [M2CF3COO]1; HRMS (FAB)
calcd for C10H14I2NO [M2CF3COO]1 417.9165, found
417.9146.
The early toxic fraction was ®nally puri®ed by HPLC
[Develosil ODS HG-5, 20£250 mm, (1) 20% aqueous
MeOH containing 0.01 M TFA, 5 mL/min; (2) 20%
aqueous MeOH containing 0.01 M TFA, 5 mL/min] to
give turbotoxin A (1) (0.5 mg; LD99 1.0 mg/kg), and the
late toxic fraction was puri®ed by HPLC [Develosil ODS
HG-5, 20£250 mm, (1) 30 to 40% aqueous MeOH contain-
ing 0.01 M TFA, 5 mL/min; (2) 20% aqueous MeOH
containing 0.01 M TFA, 5 mL/min] to give turbotoxin B
(2) (0.9 mg, LD99 4.0 mg/kg).
N,N-Dimethyldibromotyramine (3b). Bromine (0.065 mL,
1.26 mmol) was slowly added to a solution of t-BuNH2
(0.26 mL, 2.5 mmol) in toluene (3 mL) at 2788C. The
resulting mixture was warmed to 58C, and N,N-dimethyl-
tyramine (3d, 83.3 mg, 0.50 mmol) was added. After the
mixture was stirred at 58C for 3 h, the reaction was
quenched by adding H2O (7 mL), and the mixture was
extracted with EtOAc (5£15 mL). The combined extracts
were washed with brine, dried, and concentrated. The
residual solid was chromatographed on silica gel (8.5 g,
EtOAc±MeOH 10:1!5:1!1:1) to give N,N-dimethyl-
dibromotyramine (3b, 159 mg, 98%) as a colorless solid:
1H NMR (270 MHz, CD3OD) d 7.31 (s, 2H), 2.68 (s, 4H),
2.41 (s, 6H); MS (FAB) m/z 322, 324, 326 [M1H]1; HRMS
(FAB) calcd for C10H14Br2NO [M1H]1 321.9442, found
321.9441.
The residue (10.2 g) from the aforementioned eluate of 25%
aqueous EtOH provided turbotoxin A (1.5 mg) by a similar
separation procedure as that described above.
Turbotoxin A (1). UV (MeOH) lmax 226 (e 19000), 239
1
(sh, 9500), 275 nm (sh, 2000); H and 13C NMR data, see
Table 1; FABMS m/z 645 [M2CF3COO]1; HRFABMS
calcd for C19H30F3I2N2O3 m/z 645.0298 [M2CF3COO]1,
found 645.0286.

9068
H. Kigoshi et al. / Tetrahedron 56 (2000) 9063±9070
N,N-Dimethyldichlorotyramine (3c). A solution of N,N-
dimethyltyramine (3d, 100 mg, 0.61 mmol) in SO2Cl2
(0.11 mL, 1.3 mmol) was stirred at 658C for 1.5 h. Sulfuryl
chloride (0.02 mL, 0.25 mmol) was added, and the mixture
was stirred at the same temperature for a further 4.5 h. The
reaction mixture was concentrated, dissolved in MeOH to
decompose the excess of SO2Cl2, and concentrated again.
The residual solid was chromatographed on silica gel (8.5 g,
EtOAc±MeOH 10:1!5:1!1:1) to give N,N-dimethyl-
dichlorotyramine (3c, 56.1 mg, 40%) as a colorless solid:
1H NMR (270 MHz, CD3OD) d 7.27 (s, 2H), 2.97±2.90 (m,
4H), 2.91 (s, 6H); MS (FAB) m/z 234, 236, 238 [M1H]1;
HRMS (FAB) calcd for C10H14Cl2NO [M1H]1 234.0453,
found 234.0471.
Turbotoxins A (1) and B (2)
Diamine 4a (6.5 mg, 13 mmol) was dissolved in a 0.1 M
solution of iodomethane in MeOH (0.13 mL, 13 mmol),
and the resulting solution was stirred at room temperature
for 2 h and concentrated. The residue was puri®ed by HPLC
(1. Develosil ODS HG-5, 20£250 mm, 30% aqueous MeOH
containing 0.1% TFA, 5 mL/min, UV 254 nm; 2. Develosil
PhA-5, 20£250 mm, 30% aqueous MeOH containing 0.1%
TFA, 5 mL/min, UV 254 nm) to give turbotoxin A (1,
0.5 mg, 5%), turbotoxin B (2, 1.6 mg, 17%), compound 5
(2.1 mg, 22%), and diamine 4a (3.6 mg, 38%) as the
tri¯uoroacetate salt, respectively.
1
Compound 5 (tri¯uoroacetate salt). H NMR (270 MHz,
Diamine 4a. A mixture of N,N-dimethyldiodotyramine (3a,
16.6 mg, 39.8 mmol), Cs2CO3 (66.1 mg, 0.203 mmol),
ClCH2CH2CH2N(CH3)2´HCl (12.6 mg, 79.6 mmol), and
DMF (0.41 mL) was stirred at 408C for 28 h. The reaction
mixture was diluted with H2O (5 mL) and extracted with
EtOAc (3£5 mL). The combined extracts were washed
with saturated aqueous NaCl, dried, and concentrated. The
residue was chromatographed on silica gel (0.5 g, EtOAc±
MeOH 10:1!5:1!0:1) to give diamine 4a (11.6 mg, 58%)
CD3OD) d 7.86 (s, 2H), 4.11 (t, Jˆ5.6 Hz, 2H), 3.82±3.73
(m, 2H), 3.37±3.28 (m, 2H), 3.22 (s, 9H), 3.00±2.94 (m,
2H), 2.92 (s, 6H), 2.48±2.36 (m, 2H); MS (FAB) m/z 517
[M2CF3COOH±CF3COO]1; HRMS (FAB) calcd for
C16H27I2N2O
found 517.0208.
[M2CF3COOH±CF3COO]1
517.0215,
Diamine tri¯uoroacetate salt 6. 1H NMR (270 MHz,
CD3OD) d 7.83 (s, 2H), 4.10 (t, Jˆ5.6 Hz, 2H), 3.59±
3.50 (m, 2H), 3.36±3.30 (m, 2H), 3.00±2.94 (m, 2H), 2.98
(s, 6H), 2.92 (s, 6H), 2.40±2.28 (m, 2H); MS (FAB) m/z 503
[M2CF3COOH±CF3COO]1; HRMS (FAB) calcd for
C15H25I2N2O [M2CF3COOH±CF3COO]1 503.0059, found
503.0041.
1
as a colorless oil: H NMR (270 MHz, CD3OD) d 7.68 (s,
2H), 3.98 (t, Jˆ6.3 Hz, 2H), 2.72±2.62 (m, 4H), 2.55±2.46
(m, 2H), 2.30 (s, 6H), 2.28 (s, 6H), 2.14±2.01 (m, 2H); MS
(FAB) m/z 503 [M1H]1; HRMS (FAB) calcd for
C15H25I2N2O [M1H]1 503.0059, found 503.0041.
Diamines 4b±4e were also prepared from tyramines 3b±3e,
respectively, by the same experimental procedure as for
diamine 4a.
Turbotoxin analogs 7±10
These compounds were also prepared from the correspond-
ing dihalodimethyltyramines 4b±4e, respectively, by a
similar experimental procedure as that for turbotoxins A
and B except for the use of an excess of iodomethane.
Diamine 4b (72% yield). A colorless oil; 1H NMR
(270 MHz, CD3OD) d 7.45 (s, 2H), 4.01 (t, Jˆ6.1 Hz,
2H), 2.75±2.60 (m, 4H), 2.60±2.50 (m, 2H), 2.29 (s, 6H),
2.28 (s, 6H), 2.06±2.00 (m, 2H); MS (FAB) m/z 407, 409,
411 [M1H]1; HRMS (FAB) calcd for C15H25Br2N2O
[M1H]1 407.0334, found 407.0308.
Bromo analog 7 (tri¯uoroacetate salt, 100% yield). A
colorless oil; H NMR (270 MHz, CD3OD) d 7.63 (s, 2H),
1
4.13 (t, Jˆ5.6 Hz, 2H), 3.78±3.71 (m, 2H), 3.58±3.51 (m,
2H), 3.21 (s, 9H), 3.19 (s, 9H), 3.14±3.05 (m, 2H), 2.44±
2.31 (m, 2H); MS (FAB) m/z 549, 551, 553
[M2CF3COO]1; HRMS (FAB) calcd for C19H30F3Br2N2O3
[M2CF3COO]1 549.0575, found 549.0546.
Diamine 4c (66% yield). A colorless oil; 1H NMR
(270 MHz, CD3OD) d 7.25 (s, 2H), 4.02 (t, Jˆ6.1 Hz,
2H), 2.75±2.70 (m, 2H), 2.66±2.60 (m, 2H), 2.56±2.51
(m, 2H), 2.29 (s, 6H), 2.28 (s, 6H), 2.05±1.95 (m, 2H);
MS (FAB) m/z 319, 321, 323 [M1H]1; HRMS (FAB)
calcd for C15H25Cl2N2O [M1H]1 319.1344, found
319.1318.
Chloro analog 8 (tri¯uoroacetate salt, 100% yield). A
colorless oil; H NMR (270 MHz, CD3OD) d 7.43 (s, 2H),
1
4.13 (t, Jˆ5.6 Hz, 2H), 3.75±3.70 (m, 2H), 3.58±3.52
(m, 2H), 3.21 (s, 9H), 3.20 (s, 9H), 3.13±3.07 (m, 2H),
2.40±2.30 (m, 2H); MS (FAB) m/z 461, 463, 465
[M2CF3COO]1; HRMS (FAB) calcd for C19H30F3Cl2N2O3
[M2CF3COO]1 461.1585, found 461.1565.
Diamine 4d (75% yield). A colorless oil; 1H NMR
(270 MHz, CD3OD) d 7.09 (d, Jˆ8.9 Hz, 2H), 6.82 (d,
Jˆ8.9 Hz, 2H), 3.95 (t, Jˆ6.1 Hz, 2H), 2.73±2.67 (m,
2H), 2.52±2.45 (m, 4H), 2.27 (s, 6H), 2.24 (s, 6H), 1.97±
1.86 (m, 2H); MS (FAB) m/z 251 [M1H]1; HRMS
(FAB) calcd for C15H27N2O [M1H]1 251.2123, found
251.2140.
Hydro analog 9 (tri¯uoroacetate salt, 97% yield). A
colorless oil; 1H NMR (270 MHz, CD3OD) d 7.25 (d,
Jˆ8.9 Hz, 2H), 6.93 (d, Jˆ8.9 Hz, 2H), 4.09 (t, Jˆ5.6 Hz,
2H), 3.62±3.50 (m, 4H), 3.19 (s, 9H), 3.18 (s, 9H),
3.15±3.05 (m, 2H), 2.33±2.23 (m, 2H); MS (FAB) m/z
393 [M2CF3COO]1; HRMS (FAB) calcd for
C19H32F3N2O3 [M2CF3COO]1 393.2365, found 393.2383.
Diamine 4e (47% yield). A colorless oil; 1H NMR
(400 MHz, CD3OD) d 6.83 (s, 2H), 3.77 (t, Jˆ6.1 Hz,
2H), 2.69±2.62 (m, 2H), 2.62±2.55 (m, 2H), 2.54±2.47
(m, 2H), 2.29 (s, 12H), 2.01±1.91 (m, 2H); MS (FAB) m/z
279 [M1H]1; HRMS (FAB) calcd for C17H31N2O [M1H]1
279.2436, found 279.2413.
Methyl analog 10 (tri¯uoroacetate salt, 100% yield). A
colorless oil; H NMR (400 MHz, CD3OD) d 6.98 (s, 2H),
1

H. Kigoshi et al. / Tetrahedron 56 (2000) 9063±9070
9069
3.86 (t, Jˆ5.6 Hz, 2H), 3.72±3.64 (m, 2H), 3.54±3.46 (m,
2H), 3.20 (s, 9H), 3.18 (s, 9H), 3.04±2.96 (m, 2H), 2.34±
2.21 (m, 2H), 2.26 (s, 6H); MS (FAB) m/z 421
[M2CF3COO]1; HRMS (FAB) calcd for C21H36F3N2O3
[M2CF3COO]1 421.2679, found 421.2683.
3.48 (m, 2H), 3.20±3.10 (m, 2H), 3.19 (s, 6H), 3.13 (s, 6H),
2.57±2.45 (m, 2H); MS (FAB) m/z 949 [M2CF3COO]1;
HRMS (FAB) calcd for C43H46F3I2N2O3 [M2CF3COO]1
949.1550, found 949.1556.
1
Analog 15. H NMR (270 MHz, CD3OD) d 7.88 (s, 2H),
7.83±7.77 (m, 2H), 7.70±7.61 (m, 4H), 7.51±7.38 (m, 3H),
4.62 (s, 2H), 4.11 (t, Jˆ5.6 Hz, 2H), 3.83±3.73 (m, 2H),
3.58±3.49 (m, 2H), 3.23 (s, 9H), 3.20±3.14 (m, 2H), 3.13 (s,
6H), 2.50±2.37 (m, 2H); MS (FAB) m/z 797
[M2CF3COO]1; HRMS (FAB) calcd for C31H38F3I2N2O3
[M2CF3COO]1 797.0924, found 797.0902.
Benzyl analogs 11, 12, and 13
Diamine 4a (11.0 mg, 21.9 mmol) was dissolved in a 0.1 M
solution of benzyl bromide in MeOH (0.22 mL, 22 mmol),
and the mixture was stirred at room temperature for 32 h.
Iodomethane (0.14 mL, 2.2 mmol) was added, and the
mixture was stirred at room temperature for 3 h and con-
centrated. The residue was separated by HPLC (Develosil
PhA-5, 20£250 mm, 47% aqueous MeOH containing 0.1%
TFA, 5 mL/min, UV 254 nm) to give analog 11 (6.8 mg,
30%), analog 12 (4.0 mg, 20%), analog 13 (6.0 mg, 30%),
and turbotoxin A (1) (3.7 mg, 20%) as a colorless oil,
respectively.
1
Analog 16. H NMR (270 MHz, CD3OD) d 7.85 (s, 2H),
7.82±7.63 (m, 6H), 7.51±7.26 (m, 3H), 4.68 (s, 2H), 4.11 (t,
Jˆ5.3 Hz, 2H), 3.81±3.71 (m, 2H), 3.56±3.46 (m, 2H), 3.18
(s, 9H), 3.18 (s, 6H), 3.10±3.00 (m, 2H), 2.57±2.45 (m, 2H);
MS (FAB) m/z 797 [M2CF3COO]1; HRMS (FAB) calcd
for C31H38F3I2N2O3 [M2CF3COO]1 797.0924, found
797.0923.
1
Analog 11. H NMR (270 MHz, CD3OD) d 7.85 (s, 2H),
7.66±7.49 (m, 10H), 4.64 (s, 2H), 4.57 (s, 2H), 4.11 (t,
Jˆ5.3 Hz, 2H), 3.77±3.69 (m, 2H), 3.56±3.45 (m, 2H),
3.19±3.10 (m, 2H), 3.14 (s, 6H), 3.09 (s, 6H), 2.56±2.43
(m, 2H); MS (FAB) m/z 797 [M2CF3COO]1; HRMS
(FAB) calcd for C31H38F3I2N2O3 [M2CF3COO]1
797.0924, found 797.0914.
Dimethylamino alcohol 18
A mixture of 3,5-dimethyl-4-hydroxybenzaldehyde (200 mg,
1.33 mmol), ZnI2 (20.6 mg, 0.064 mmol), and TMSCN
(0.2 mL, 1.46 mmol) was stirred at room temperature for
42 h. Trimethylsilyl cyanide (0.2 mL, 1.46 mmol) was
added, and the mixture was stirred at room temperature for
18 h. The mixture was diluted with saturated aqueous
NaHCO3 (5 mL) and extracted with EtOAc (3£5 mL). The
combined extracts were washed with brine, dried, and
concentrated to give a crude cyanohydrin TMS ether
(417 mg).
1
Analog 12. H NMR (270 MHz, CD3OD) d 7.86 (s, 2H),
7.59±7.51 (m, 5H), 4.57 (s, 2H), 4.11 (t, Jˆ5.6 Hz, 2H),
3.83±3.74 (m, 2H), 3.56±3.46 (m, 2H), 3.23 (s, 9H), 3.19±
3.09 (m, 2H), 3.09 (s, 6H), 2.49±2.37 (m, 2H); MS (FAB)
m/z 721 [M2CF3COO]1; HRMS (FAB) calcd for
C25H34F3I2N2O3
721.0623.
[M2CF3COO]1
721.0611,
found
To a solution of the cyanohydrin TMS ether (417 mg) in
THF (1.3 mL) was added a 1.0 M solution of LiAlH4 in THF
(1.33 mL, 1.33 mmol) at 2308C, and the mixture was stir-
red at 08C for 1.5 h. Sodium ¯uoride (280 mg) was added,
and the mixture was stirred vigorously for a while. To the
mixture, H2O±THF (1:9, 10 mL) was carefully added. After
being stirred for 30 min, the mixture was ®ltered through a
pad of Celite. The residue was washed with aqueous THF
thoroughly, and the ®ltrate and washings were combined
and concentrated. The residue was chromatographed on
silica gel (4 g, EtOAc±MeOH 5:1!3:1!1:1) to give
amino alcohol 17 (91.2 mg, 38% for 2 steps) as a colorless
1
Analog 13. H NMR (270 MHz, CD3OD) d 7.85 (s, 2H),
7.66±7.48 (m, 5H), 4.63 (s, 2H), 4.10 (t, Jˆ5.3 Hz, 2H),
3.77±3.69 (m, 2H), 3.56±3.47 (m, 2H), 3.19 (s, 9H), 3.14 (s,
6H), 3.10±3.00 (m, 2H), 2.56±2.47 (m, 2H); MS (FAB) m/z
721 [M2CF3COO]1; HRMS (FAB) calcd for
C25H34F3I2N2O3
721.0601.
[M2CF3COO]1
721.0611,
found
Phenylbenzyl analogs 14, 15, and 16
To a solution of diamine 4a (9.6 mg, 19 mmol) in MeOH
(0.2 mL) was added 4-phenylbenzyl bromide (5.1 mg,
21 mmol), and the mixture was stirred at room temperature
for 20 h. Iodomethane (0.13 mL, 2 mmol) was added, and
the mixture was stirred at room temperature for 3 h and
concentrated. The residue was separated by HPLC (1.
Develosil PhA-5, 20£250 mm, 58% aqueous MeOH
containing 0.1% TFA, 5 mL/min, UV 254 nm; 2. Develosil
PhA-5, 20£250 mm, 40% aqueous MeOH containing 0.1%
TFA, 5 mL/min, UV 254 nm) to give analog 14 (8.0 mg,
32%), analog 15 (2.0 mg, 9%), analog 16 (6.1 mg, 28%),
and turbotoxin A (1) (3.1 mg, 17%) as a colorless oil,
respectively.
1
oil: H NMR (400 MHz, CD3OD) d 6.90 (s, 2H), 4.46 (dd,
Jˆ7.3, 5.1 Hz, 1H), 2.81±2.70 (m, 2H), 2.19 (s, 6H); MS
(FAB) m/z 182 [M1H]1.
A mixture of amino alcohol 17 (91.2 mg, 0.503 mmol), 10%
Pd/C (51 mg), 35% formalin (0.20 mL, 2.5 mmol), and
MeOH (3 mL) was stirred under hydrogen at room tempera-
ture for 14 h. The mixture was ®ltered through a pad of
Celite, and the residue was washed with MeOH (30 mL).
The ®ltrate and washings were combined and concentrated.
The residue was chromatographed on silica gel (4 g,
EtOAc±MeOH 5:1!3:1!1:1) to give dimethylamino
alcohol 18 (74 mg, 70%) as a colorless oil: 1H NMR
(400 MHz, CD3OD) d 6.91 (s, 2H), 4.65 (dd, Jˆ9.3,
3.7 Hz, 1H), 2.67 (dd, Jˆ12.7, 9.3 Hz, 1H), 2.45 (dd,
Jˆ12.7, 3.7 Hz, 1H), 2.37 (s, 6H), 2.19 (s, 6H); MS
1
Analog 14. H NMR (270 MHz, CD3OD) d 7.86 (s, 2H),
7.82±7.61 (m, 12H), 7.51±7.34 (m, 6H), 4.69 (s, 2H), 4.62
(s, 2H), 4.11 (t, Jˆ5.3 Hz, 2H), 3.81±3.71 (m, 2H), 3.57±

9070
H. Kigoshi et al. / Tetrahedron 56 (2000) 9063±9070
(FAB) m/z 210 [M1H]1; HRMS (FAB) calcd for
C12H20NO2 [M1H]1 210.1494, found 210.1477.
1156; Chou, T.; Kamo, O.; Uemura, D. Tetrahedron Lett. 1996,
37, 4023±4026. Chou, T.; Haino, T.; Kuramoto, M.; Uemura, D.
Tetrahedron Lett. 1996, 37, 4027±4030.
2. Schantz, E. J.; Mold, J. D.; Stanger, D. W.; Shavel, J.; Riel, F. J.;
Bowden, J. P.; Lynch, J. M.; Wyler, R. S.; Riegel, B.; Sommer, H.
J. Am. Chem. Soc. 1957, 79, 5230±5235.
Dimethyltyramine 3e. A mixture of dimethylamino
alcohol 18 (16.8 mg, 80.3 mmol), Et3SiH (0.016 mL,
96 mmol), and TFA (0.4 mL) was stirred at 08C for
40 min. The mixture was diluted with saturated aqueous
NaHCO3 (8 mL) and extracted with EtOAc (3£10 mL).
The combined extracts were washed with brine, dried, and
concentrated. The residue was chromatographed on silica
gel (3 g, EtOAc±MeOH 5:1!3:1) to give dimethyltyramine
3e (20.1 mg, 100%) as a colorless oil: 1H NMR (400 MHz,
CD3OD) d 6.79 (s, 2H), 2.97±2.89 (m, 2H), 2.81±2.71 (m,
2H), 2.62 (s, 6H), 2.17 (s, 6H); MS (FAB) m/z 194 [M1H]1;
HRMS (FAB) calcd for C12H20NO [M1H]1 194.1545,
found 194.1548.
3. Kosuge, T.; Tsuji, K.; Hirai, K.; Yamaguchi, K.; Okamoto, T.;
Iitaka, Y. Tetrahedron Lett. 1981, 22, 3417±3420.
4. Kanno, K.; Kotaki, Y.; Yasumoto, T. Bull. Jpn. Soc. Sci. Fisher.
1976, 42, 1395±1398; Yasumoto, T.; Kotaki, Y. Bull. Jpn. Soc. Sci.
Fisher. 1977, 43, 207±211; Kotaki, Y.; Oshima, Y.; Yasumoto, T.
Bull. Jpn. Soc. Sci. Fisher. 1981, 47, 943±946.
5. Kotaki, Y.; Yasumoto, T. Bull. Jpn. Soc. Sci. Fisher. 1977, 43,
1467.
6. Preliminary communication: Kigoshi, H.; Kanematsu, K.;
Uemura, D. Tetrahedron Lett. 1999, 40, 5745±5748.
7. Reagents for aromatic bromination, t-BuNH2±Br2, were
reported: Pearson, J. H.; Wysong, J. W.; Breder, C. V. J. Org.
Chem. 1967, 32, 2358±2360.
Acknowledgements
8. Diamine 4, or dakaramine, was isolated as a metabolite of the
Senegalese sponge Ptilocaulis sp.: Diop, M.; Samb, A.;
Costantino, V.; Fattorusso, E.; Mangoni, A. J. Nat. Prod. 1996,
59, 271±272.
This work was supported in part by the Yamada Science
Foundation and by Grants-in-Aid for Scienti®c Research
and for COE Research from the Ministry of Education,
Science, Sports, and Culture, Japan. We thank Dr Hiroshi
Yamauchi (Eisai Co.) for evaluating the inhibitory activity
of turbotoxin A against acetylcholinesterase and Professors
Hideaki Karaki and Hiroshi Ozaki (The University of
Tokyo) for testing the neuropharmacological activities of
turbotoxin A.
9. Harel, M.; Schalk, I.; Ehret-Sabatier, L.; Bouet, F.; Goeldner,
M.; Hirth, C.; Axelsen, P. H.; Silman, I.; Sussman, J. L. Proc. Natl.
Acad. Sci. USA 1993, 90, 9031±9035.
10. Xynas, R.; Capon, R. J. Aust. J. Chem. 1989, 42, 1427±1433.
11. Hamann, M. T.; Scheuer, P. J.; Kelly-Borges, M. J. Org.
Chem. 1993, 58, 6565±6569.
12. Tsukamoto, S.; Kato, H.; Hirota, H.; Fusetani, N. J. Org.
Chem. 1996, 61, 2936±2937.
References
13. Yasumoto, T.; Endo, M. Bull. Jpn. Soc. Sci. Fisher. 1974, 40,
217±221; Yasumoto, T.; Endo, M. Bull. Jpn. Soc. Sci. Fisher.
1974, 40, 841±845.
1. Uemura, D.; Chou, T.; Haino, T.; Nagatsu, A.; Fukuzawa, S.;
Zheng, S.-z.; Chen, H.-s. J. Am. Chem. Soc. 1995, 117, 1155±