Revised structure of zamamistatin

Article abstract of DOI:10.1016/j.tetlet.2005.10.157

The structure of zamamistatin, a novel bromotyrosine derivative from the Okinawan sponge Pseudoceratina sp. was revised by comparison of the ¹³C NMR data of zamamistatin with those of synthetic model compounds.

Full text of DOI:10.1016/j.tetlet.2005.10.157

Tetrahedron
Letters
Tetrahedron Letters 47 (2006) 155–158
Revised structure of zamamistatin
Ichiro Hayakawa,a,b Toshiaki Teruyaa,ꢀ and Hideo Kigoshia,b,
*
aDepartment of Chemistry, Graduate School of Pure and Applied Sciences, University of Tsukuba, Tennodai, Tsukuba 305-8571, Japan
bCenter for Tsukuba Advanced Research Alliance, University of Tsukuba, Tennodai, Tsukuba 305-8577, Japan
Received 6 October 2005; revised 24 October 2005; accepted 31 October 2005
Available online 17 November 2005
Abstract—The structure of zamamistatin, a novel bromotyrosine derivative from the Okinawan sponge Pseudoceratina sp. was
revised by comparison of the 13C NMR data of zamamistatin with those of synthetic model compounds.
Ó 2005 Elsevier Ltd. All rights reserved.
Zamamistatin, isolated from Okinawan sponge Pseudo-
ceratina purpurea by Uemura and co-workers,1 exhibits
significant antibacterial activity against the marine bac-
teria Rhodospirillum salecigens SCRC 113 strain. The
structure was elucidated as shown in 1, an exo-type
dimer of the azaoxa-spiro[6.5] unit consisting of a cyclo-
hexadienyl moiety and an isoxazolidine ring. In our
continuing search for new substances from marine
organisms, we investigated the constituents of the
marine sponge Pseudoceratina sp. collected at Okuma,
Okinawa, Japan, and isolated zamamistatin (Fig. 1).
The spectral analysis of zamamistatin let us reconsider
its structure. We report here the revised structure of
zamamistatin, 2, by comparing the 13C NMR data for
zamamistatin with those of synthetic model compounds.
The marine sponge Pseudoceratina sp. (115 g), collected
at Okuma, Okinawa, Japan, in August 2005, was
extracted with MeOH (300 mL) for 7 days. The extract
was filtered, concentrated, and partitioned between
EtOAc and H2O. The EtOAc-soluble material was
further partitioned between 90% aqueous MeOH and
hexane. The material obtained from the aqueous MeOH
portion was subjected to fractionation with column
chromatography (ODS silica gel, MeOH–H2O) and
reversed-phase HPLC (Develosil ODS-HG-5, MeOH–
H2O) to give zamamistatin as a colorless oil (35.0 mg).
HO
O
Br
1
The H and 13C NMR data of isolated zamamistatin in
HN
8
CDCl3 are identical with the reported data.1 Table 1
Br
OMe
6
7
9
MeO
Br
NH
1
O
a
Table 1. NMR data for zamamistatin in CD3OD and acetone-d6
Br
OH
Position
1H
13C
previously proposed structure (1) for zamamistatin
HO
O
Br
500 MHz
CD3OD
270 MHz
acetone-d6
125 MHz 67.8 MHz
Br
HN
8
CD3OD
acetone-d6
OMe
6
7
9
1
2
3
4
5
6
4.10 d (1.1)
4.23 d (7.3)
78.8
114.2
148.9
121.5
78.6
113.8
148.2
120.7
133.6
74.3
MeO
1
O
NH
Br
Br
OH
revised structure 2
6.33 d (1.1)
6.41 d (1.1) 133.5
74.5
Figure 1.
7a
7b
8
2.78 d (16.7) 2.91 br s
2.83 d (16.7)
27.0
26.7
Keywords: Zamamistatin; Revised structure; endo-Type dimer of aza-
oxa-spiro[6.6] unit.
118.4
60.2
117.8
60.0
9
3.70 s
3.70 s
*
Corresponding author. Tel./fax: +81 29 853 4313; e-mail:
kigoshi@chem.tsukuba.ac.jp
–OH
–NH
5.33 d (7.3)
5.33 br s
ꢀ Present address: Research Institute for Biological Functions, Chubu
University, 1200 Matsumoto-cho, Kasugai, Aichi 487-8501, Japan.
a Coupling constants (Hz) are given in parentheses.
0040-4039/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2005.10.157

156
I. Hayakawa et al. / Tetrahedron Letters 47 (2006) 155–158
1H-NMR
13C-NMR
followed immediately by the addition of NH2OHÆHCl,
yielded an oxime, the hydrogenolysis of which gave
3.15, 3.85
40.2
6.53
132.3
Br
113.8
oxime ester 8. In the oxidative cyclization of 8, we used
160.0
O
O
2,4,4,6-tetrabromo-2,5-cyclohexadienone
in
aceto-
91.5
Br
nitrile.5 This reaction took place readily to yield 9 in a
92% yield. According to Yamamuraꢁs method,6 reduc-
tion of 9 with Zn(BH4)27 gave trans dihydro-1,2-oxazine
methyl ester 10a. On the other hand, 9 was reduced with
NH(CH2)4HN
7
8 155.3
6
N
OMe
60.2
MeO
O
N
148.7
O
O H
Br
Br
OH
5.40
5a
4.18
75.1
NaBH4 to give cis dihydro-1,2-oxazine methyl ester
122.0
10b.8
aerothionin (3)
Figure 2. Selected NMR data of aerothionin (3) in acetone-d6.
Stereochemistry of 10a and 10b was determined by com-
parison of the H NMR spectra of 10a (dH1 4.13, dOH
1
5.30) and 10b (dH1 4.36, dOH 4.89) in acetone-d6 with
1
those of related compounds;6 the H chemical shifts of
summarizes the NMR data in different solvents. The
chemical shifts in acetone-d6 of C1–C5 and C9 in zama-
mistatin closely resembled those of aerothionin
(Fig. 2).2c However, the chemical shifts of C6 (dC6
74.3) and C7 (dC7 26.7; dH7 2.91) in zamamistatin were
apparently different from those of aerothionin (dC6
91.5, dC7 40.2; dH7 3.15, 3.85) (Fig. 2). Therefore, zama-
mistatin was thought to have a different ring system
from an isoxazolidine ring. Considering the molecular
formula of zamamistatin, we proposed structure 2, pos-
sessing a dihydro-1,2-oxazine ring for zamamistatin.
the trans isoxazoline methyl ester 11a, which has a trans
vicinal relationship between a hydroxy group and an
oxime oxygen atom, were dH1 4.22 and dOH 5.38, while
those of cis isoxazoline methyl ester 11b were dH1 4.53
and dOH 4.98 (Fig. 3).6 Thus, the stereochemistry of
compounds 10a and 10b was found to be trans and
cis, respectively.
The 13C NMR data in acetone-d6 of the synthetic 10a
and 10b were similar to those of zamamistatin rather
than aerothionin-related compounds.9 In the synthetic
compound 10a, carbon signal due to C6 appeared at
dC 79.9, while the carbon signal in zamamistatin
appeared at dC 74.3. On the other hand, the carbon
signal due to C6 for 11a appeared at dC 92.4
(Fig. 4).3c These observations indicated that zamamist-
atin has a dihydro-1,2-oxazine ring rather than an isox-
azolidine ring. Based on these results, it was concluded
that the structure of natural zamamistatin, previously
proposed as 1, should be revised to structure 2.
To confirm the proposed structure, we planned to com-
pare the 13C NMR data of zamamistatin with those of
dihydro-1,2-oxazine methyl ester 10a and isoxazoline
methyl ester 11a.3 Although NMR data of isoxazoline
methyl ester 11a were reported by Hoshino,3c dihydro-
1,2-oxazine methyl ester 10a has not been prepared.
We therefore synthesized dihydro-1,2-oxazine methyl
ester 10a.
Our synthetic plan was based on reported procedures3
(Scheme 1). The Wittig-type reaction of 4 and subse-
quent hydrolysis gave aldehyde 5 in a good yield. Alde-
hyde 5 was subjected to Horner–Wadsworth–Emmons
olefination with phosphonate 64 to give silyl enol ether
7. Treatment of silyl enol ether 7 with HFÆpyr in MeOH,
This novel structure of zamamistatin, an endo-type
dimer of the azaoxa-spiro[6.6] unit, can be explained
by a plausible biogenetic pathway shown in Figure 5.
Reductive dimerization of the isoxazoline derivative
12,10 followed by oxidative decarboxylation, yielded an
Br
Br
Br
Br
Br
Br
CO2Me
OTBS
CHO
a
b
c
MeO
CHO
OBn
MeO
MeO
O
P
MeO
MeO
CO2Me
OBn
OBn
4
5
7
OTBS
6
Br
10a
CO2Me
MeO
O
OH
N
N
e
f
trans
Br
Br
CO2Me
NOH
Br
Br
Br
Br
d
MeO
MeO
CO2Me
O
O
N
OH
9
8
10b
cis
MeO
CO2Me
O
Br
OH
Scheme 1. Reagents and conditions: (a) (i) (methoxymethyl)triphenylphosphonium chloride, t-BuOK, THF, rt; (ii) 2 M HCl, THF, reflux, 97% in
two steps; (b) 6, LHMDS, THF, À78 °C, 98%; (c) (i) HFÆpyr., MeOH, rt, then NH2OHÆHCl, rt; (ii) H2, Pd/C, 1,4-dioxane-AcOH, rt, 98% in two
steps; (d) 2,4,4,6-tetrabromo-2,5-cyclohexadienone, MeCN, rt, 92%; (e) Zn(BH4)2, CH2Cl2, rt, 9%; (f) NaBH4, MeOH, rt, 19%.

I. Hayakawa et al. / Tetrahedron Letters 47 (2006) 155–158
157
Br
Br
exo-type dimer of azaoxa-spiro[6.5] unit 1. Hydrolysis
and recyclization of 1 gave the endo-type dimer of aza-
oxa-spiro[6.6] unit 2. This isomerization from 1 to 2
was supported by computational calculation, in which
structure 2 is more stable than structure 1 (52.7 kcal/
mol for 1, 49.1 kcal/mol for 2: MacroModel 6.0,
AMBER*).
Br
MeO
CO2Me
MeO
CO2Me
1
1
O
OH
N
O
OH
N
Br
4.13 (d, J = 8.2 Hz)
4.36 (d, J = 7.8 Hz)
5.30 (d, J = 8.2 Hz)
4.89 (d, J = 7.8 Hz)
10b (cis)
10a (trans)
Br
Br
CO2Me
CO2Me
Acknowledgements
MeO
MeO
1
N
1
N
O
O
We express our gratitude to Professor Daisuke Uemura
(Nagoya University) for his helpful discussion. This
work was supported in part by the 21st COE program
and Grant-in-Aid for Scientific Research on Priority
Areas (No. 16073204, ꢂCreation of Biologically Func-
tional Moleculesꢁ) from the Ministry of Education, Cul-
ture, Sports, Science, and Technology, Japan and
University of Tsukuba Projects. The authors are grate-
ful to the Chemical Analysis Center, University of Tsu-
kuba, for NMR experiments.
Br
4.22 (d, J = 7.5 Hz)
OH
Br
4.53 (d, J = 7.5 Hz)
OH
4.98 (d, J = 7.5 Hz)
5.38 (d, J = 7.5 Hz)
11a (trans)
11b (cis)
Br
MeO
Br
1
O
NH
OH
4.23 (d, J = 7.3 Hz)
5.33 (d, J = 7.3 Hz)
zamamistatin
References and notes
Figure 3. Selected 1H NMR data of zamamistatin and compounds
10a, 10b, 11a, and 11b in acetone-d6.
1. Takada, N.; Watanabe, R.; Suenaga, K.; Yamada, K.;
Ueda, K.; Kita, M.; Uemura, D. Tetrahedron Lett. 2001,
42, 5265–5267.
2. (a) Fattorusso, E.; Minale, L.; Sodano, G.; Moody, K.;
Thompson, R. H. J. Chem. Soc., Chem. Commun. 1970,
752–753; (b) McMillan, J. A.; Paul, I. C.; Goo, Y. M.;
Rinehart, K. L., Jr.; Krueger, W. C.; Pschigoda, L. M.
Tetrahedron Lett. 1981, 22, 39–43; (c) Kernan, M. R.;
Cambie, R. C.; Bergquist, P. R. J. Nat. Prod. 1990, 53,
615–622.
3. (a) Ogamino, T.; Ishikawa, Y.; Nishiyama, S. Heterocycles
2003, 61, 73–78; (b) Boehlow, T. R.; Jonathan, H.;
Spilling, C. D. J. Org. Chem. 2001, 66, 3111–3118; (c)
Murakata, M.; Tamura, M.; Hoshino, O. J. Org. Chem.
1997, 62, 4428–4433.
4. Nakamura, E. Tetrahedron Lett. 1981, 22, 663–666.
5. (a) Forrester, A. R.; Thomson, R. H.; Woo, S.-O. Justus
Liebigs Ann. Chem. 1978, 66–73; (b) Other oxidizing
agents, including phenyliodonium diacetate (PIDA) and
phenyliodonium bis(trifluoroacetate) (PIFA), were not
effective in this case.
79.8
133.6
Br
79.9
131.6
Br
22.8
24.3
18.0
164.2 52.6
CO2Me 60.0
150.5
18.4
164.3 52.6
CO2Me
150.4
O
N
123.3
118.8
MeO
7
7
6
60.1
MeO
6
148.4
148.7
N
O
Br
OH
73.6
113.9
Br
OH
76.4
113.8
10a
10b
92.4
74.3
133.6
Br
120.7
132.0
Br
122.1
60.2MeO
148.7
Br
161.1 52.8
CO2Me
152.4
N
26.7
7
39.8
117.8
7
6
6
60.0MeO
148.2
Br
O
OH
75.1
O
NH
OH
78.6
113.8
113.8
11a
zamamistatin
Figure 4. 13C NMR data of zamamistatin and compounds 10a, 10b,
and 11a in acetone-d6.
HO
O
Br
Br
Br
oxidative
decarboxylation
reductive
CO2R
HN
Br
Br
OMe
RO2C
dimerization
MeO
N
CO2R
O
OH
MeO
NH
O
OH
Br
12
13
HO
O
Br
Br
HO
O
Br
Br
hydrolysis
OMe and recyclization
HN
Br
Br
HN
NH
Br
Br
OMe
MeO
MeO
NH
O
OH
O
OH
1
2
Figure 5. Plausible biogenetic pathway for zamamistatin.

158
I. Hayakawa et al. / Tetrahedron Letters 47 (2006) 155–158
6. Nishiyama, S.; Yamamura, S. Bull. Chem. Soc. Jpn. 1985,
ddd, J = 5.0, 6.6, 19.6 Hz), 3.71 (3H, s), 3.78 (3H, s), 4.13
(1H, d, J = 8.2 Hz), 5.30 (1H, d, J = 8.2 Hz), 6.32 (1H, d,
J = 1.0 Hz); 13C NMR (125 MHz, acetone-d6) d 18.0, 22.8,
52.6, 60.1, 73.6, 79.9, 113.9, 123.3, 131.6, 148.7, 150.5,
164.2; ESIMS m/z 431.9074, calcd for C12H13Br2NNaO5
58, 3453–3456. Nishiyama and Yamamura reported the
stereostructures of 11a and 11b were established on the
basis of exhaustive comparison of the 1H NMR spectra
with those of trans, trans-aerothionin (3), whose stereo-
structure was established by the X-ray crystallographic
analysis,2b and cis, cis-aerothionin.5a
1
[M+Na]+ 431.9058 10b: H NMR (500 MHz, acetone-d6)
d 2.05 (1H, ddd, J = 6.7, 9.2, 13.8 Hz), 2.15 (1H, ddd,
J = 4.8, 7.3, 13.8 Hz), 2.53 (1H, ddd, J = 7.3, 9.2,
19.5 Hz), 2.65 (1H, ddd, J = 4.8, 6.7, 19.5 Hz), 3.70 (3H,
s), 3.78 (3H, s), 4.36 (1H, d, J = 7.8 Hz), 4.89 (1H, d,
J = 7.8 Hz), 6.45 (1H, d, J = 0.8 Hz); 13C NMR
(125 MHz, acetone-d6) d 18.4, 24.3, 52.6, 60.0, 76.4,
79.8, 113.8, 118.8, 133.6, 148.4, 150.4, 164.3; ESIMS
m/z 431.9031, calcd for C12H13Br2NNaO5 [M+Na]+
431.9058.
7. Nakata, T.; Tani, Y.; Hatozaki, M.; Oishi, T. Chem.
Pharm. Bull. 1984, 32, 1411–1415.
8. Due to the small reaction scale of Zn(BH4)2 and NaBH4
reductions, we could not isolate the minor isomer.
Although these reductions from 9 to 10a or 10b were in
low yield, we could not recover 9 because compounds 9,
10a, and 10b were unstable under these conditions.
9. 10a: 1H NMR (500 MHz, acetone-d6) d1.93 (1H, ddd,
J = 6.6, 8.9, 13.8 Hz), 2.38 (1H, ddd, J = 5.0, 7.0,
13.8 Hz), 2.51 (1H, ddd, J = 7.0, 8.9, 19.6 Hz), 2.64 (1H,
10. Okamoto, K. T.; Clardy, J. Tetrahedron Lett. 1987, 28,
4969–4972.