Cytotoxic labdane alkaloids from an ascidian Lissoclinum sp.: Isolation, structure elucidation, and structure-activity relationship

Article abstract of DOI:10.1016/j.bmc.2006.06.043

Four labdane alkaloids, haterumaimides N-Q (1-4), were isolated from an ascidian Lissoclinum sp. and their structures were elucidated by chemical and spectral analyses. Investigation of the structure-activity relationships of haterumaimides J-K, N-Q, and 14 related compounds suggested that the presence of hydroxyl groups at C-6, C-7, C-12, and C-18, a chlorine atom at C-2, and an imido NH in ring C should be essential for cytotoxicity against P388 cells.

Full text of DOI:10.1016/j.bmc.2006.06.043

Bioorganic & Medicinal Chemistry 14 (2006) 6954–6961
Cytotoxic labdane alkaloids from an ascidian
Lissoclinum sp.: Isolation, structure elucidation,
and structure–activity relationship
Jasim Uddin,a, Katsuhiro Ueda,a,* Eric R. O. Siwu,b Masaki Kitac
and Daisuke Uemurab,d
aDepartment of Chemistry, Biology and Marine Sciences, University of the Ryukyus, Nishihara-cho, Okinawa 903-0213, Japan
bDepartment of Chemistry, Graduate School of Science, Nagoya University, Furo-cho, Chikusa, Nagoya 464-8602, Japan
cResearch Center for Materials Science, Nagoya University, Furo-cho, Chikusa, Nagoya 464-8602, Japan
dInstitute for Advanced Research, Nagoya University, Furo-cho, Chikusa, Nagoya 464-8602, Japan
Received 9 May 2006; revised 16 June 2006; accepted 17 June 2006
Available online 18 July 2006
Abstract—Four labdane alkaloids, haterumaimides N–Q (1–4), were isolated from an ascidian Lissoclinum sp. and their structures
were elucidated by chemical and spectral analyses. Investigation of the structure–activity relationships of haterumaimides J–K,
N–Q, and 14 related compounds suggested that the presence of hydroxyl groups at C-6, C-7, C-12, and C-18, a chlorine atom at
C-2, and an imido NH in ring C should be essential for cytotoxicity against P388 cells.
ꢀ 2006 Elsevier Ltd. All rights reserved.
1. Introduction
investigated an ascidian Lissoclinum sp. collected off
the coast of Hateruma Island. A lipophilic extract of
the ascidian showed that it could inhibit the division
of fertilized sea urchin eggs. In our recent report, we
described the isolation and structure elucidation of hat-
erumaimides A–I (5–13)10,11 together with dichlorolis-
soclimide (16)12 and chlorolissoclimide (17)13 from the
toxic extract of the ascidian. Further bioassay-guided
fractionation of the same extract led to the isolation
of four new cytotoxic labdane alkaloids, hateruma-
imides N–Q (1–4)14, and haterumaimides J (14) and
K (15).15 Haterumaimides L (21) and M (22), and
3b-hydroxychlorolissoclimide (23) were recently report-
ed from the molluscs Pleurobranchus albiguttatus and
Pleurobranchus forskalii by Schmitz’s group.16 Labdane
alkaloids of this type have attracted considerable
interest because of their potential use as protein
synthesis inhibitors,17 as antitumor drugs,18 and due
to their unusual structural features.19–21 In this report,
we describe the isolation, structure elucidation, and
biological activities of haterumaimides N–Q (1–4), J
(14), and K (15), and the structure–activity relation-
ships (SARs) of haterumaimides A–K (5–15), N–Q
(1–4), dichlorolissoclimide (16), chlorolissoclimide
(17), and three synthetic derivatives of 17 (18–20).
It has been amply demonstrated that ascidians are pro-
lific producers of novel bioactive secondary metabo-
lites.1–3
A significant number of ascidian-derived
compounds have entered into preclinical and clinical
trials as antitumor agents.4 Examples include didemnin
B (went through phase II clinical trials but was with-
drawn by NCI because it proved to be too toxic to
use as a drug),5 aplidine (currently under phase II clin-
ical trials in Europe and Canada),6 and ecteinascidin
743 (passed through phase II trials for the treatment
of sarcoma and is enlisted for phase III in Europe).7
The biomedical potential of the ascidian metabolites
has resulted in the focused interest in these primitive
chordates. As part of our ongoing chemical and biolog-
ical studies on Okinawan marine organisms,8,9 we
Keywords: Diterpene; Labdane alkaloid; Haterumaimide; Cytotoxicity;
Structure–activity relationship.
*
Corresponding author. Tel.: +81 98 895 8894; fax: +81 98 895
8565; e-mail: kueda@sci.u-ryukyu.ac.jp
Present address: Department of Chemistry, Loker Hydrocarbon
Research Institute, University of Southern California, Los Angeles,
CA 90089-1661, USA.
0968-0896/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2006.06.043

J. Uddin et al. / Bioorg. Med. Chem. 14 (2006) 6954–6961
6955
2. Results and discussion
toxic fractions yielded haterumaimides
N
(1,
13.9 · 10ꢀ4 % of wet weight), O (2, 1.1 · 10ꢀ4 %), P
2.1. Isolation, structure elucidation, and relative
stereochemistries
(3, 1.0 · 10ꢀ4 %),
Q
(4, 2.5 · 10ꢀ4 %),
J
(14,
7.5 · 10ꢀ4 %), and K (15, 3.2 · 10ꢀ4 %).
The encrusting gray ascidian Lissoclinum sp. collected
off the coast of Hateruma Island, Okinawa, was
extracted with acetone. The acetone extract was first
partitioned between H2O and EtOAc. The EtOAc
extract was suspended in aqueous MeOH and the sus-
pension was successively partitioned with hexanes,
CHCl3, and 1-BuOH. Only the CHCl3 extract inhibit-
ed the division of fertilized sea urchin eggs. Bioassay-
directed fractionation of CHCl3 extract by a series of
chromatographic processes, including silica gel and
ODS column chromatography, and silica gel and
ODS HPLC, led to the isolation of haterumaimides
A–I (5–13), dichlorolissoclimide (16), and chlorolisso-
climide (17). Further HPLC separation of the minor
Haterumaimide N (1) had a molecular formula of
C22H32ClNO5, as deduced from HRFABMS [m/z
426.2058 (M+H)+, D +1.1 mmu and m/z 428.2062
(M+H +2)+; intensity ratio (3:1)]. Thus, the molecular
formula requires seven degrees of unsaturation. The
IR spectrum of 1 contained characteristic absorption
bands at mmax 3505 (OH), 3400 (NH), 1720 (C@O),
1705 (C@O), 1700 (C@O), and 1605 (C@O), and the
UV spectrum showed one absorption maxima at kmax
(loge) 215 nm (3.6). The detailed analysis of 1H and
13C NMR data (Tables 1 and 2) of 1 indicated the pres-
ence of three methyl groups, an exomethylene, two oxy-
methines, a chloromethine, ester, and imide (or amide)
carbonyls. Therefore, haterumaimide N must be tricyclic

6956
J. Uddin et al. / Bioorg. Med. Chem. 14 (2006) 6954–6961
Table 1. 1H NMRa data for haterumaimides N–Q (1–4)
Position
1b
2c
3c
4c
d (mult, J/Hz)
d (mult, J/Hz)
d (mult, J/Hz)
d (mult, J/Hz)
1b
1a
2
2.20 (ddd, 13.0, 3.5, 1.5)
1.25 (dd, 13.0, 12.5)
4.13 (tt, 12.5, 3.5)
2.20 (ddd, 12.5, 3.5, 1.5)
1.38 (t, 12.5)
4.15 (tt, 12.5, 3.5)
2.18 (ddd, 12.5, 4.0, 1.5)
1.35 (dd, 13.0, 12.5)
4.36 (tt, 12.5, 4.0)
1.60 (ddd, 13.5, 5.5, 3.0)
0.91 (dd, 13.5, 4.0)
1.50 (m)
1.42 (m)
1.36 (ddd, 13.0, 5.5, 3.5)
1.11 (m)
3b
3a
5
2.05 (ddd, 12.5, 3.5, 1.5)
1.23 (t, 12.5)
2.05 (ddd, 12.5, 3.6, 1.5)
1.25 (t, 12.5)
1.95 (ddd, 13.0, 4.0, 1.5)
1.50 (dd, 13.0, 12.5)
1.21 (dd, 12.5, 3.5)
1.88 (ddd, 12.5, 6.0, 3.4)
1.08 (q, 12.0)
1.52 (dd, 12.0, 5.0)
1.98 (dt, 13.0, 5.0)
1.35 (dt, 13.0, 12.0)
5.10 (dd, 12.0, 5.0)
1.74 (dd, 10.0, 6.0)
1.80 (ddd, 14.0, 8.0, 6.0)
1.58 (ddd, 14.0, 10.0, 7.5)
4.28 (dt, 7.5, 1.5)
1.55 (dd, 12.5, 4.0)
1.34 (m)
1.24 (m)
1.12 (dd, 12.0, 3.5)
1.87 (ddd, 12.0, 5.5, 3.5)
1.13 (m)
6b
6a
7
5.15 (dd, 12.5, 5.0)
1.95 (dd, 9.5, 6.5)
2.39 (m)
3.76 (ddd, 12.5, 5.8, 5.0)
1.63 (m)
1.50 (m)
3.75 (ddd, 11.0, 5.5, 4.5)
1.44 (dd, 10.0, 5.5)
1.58 (m)
9
11a
11b
12
2.32 (ddd, 13.0, 9.5, 7.0)
6.65 (tt, 6.5, 2.5)
1.35 (m)
1.50 (m)
1.45 (m)
1.40 (m)
3.99 (dddd, 9.0, 6.5, 5.0, 2.0)
13
2.88 (ddd, 9.0, 5.0, 1.5)
2.82 (dd, 17.5, 5.0)
2.66 (dd, 17.5, 9.0)
5.50 (br s)
2.80 (dddd, 9.0, 8.5, 4.5, 1.5)
2.74 (dd, 17.5, 9.0)
2.36 (dd, 17.5, 4.5)
5.20 (d, 1.5)
2.80 (ddd, 8.5, 5.0, 2.0)
2.52 (dd, 17.5, 5.0)
2.46 (dd, 17.5, 8.5)
5.18 (br s)
14a
14b
17a
17b
18
3.28 (d, 2.5)
3.28 (d, 2.5)
5.10 (br s)
4.55 (br s)
0.98 (s)
4.83 (br s)
0.94 (s)
4.70 (d, 1.5)
0.92 (s)
4.76 (br s)
0.85 (s)
19
20
0.82 (s)
0.73 (s)
0.86 (s)
0.78 (s)
0.81 (s)
0.65 (s)
0.75 (s)
0.57 (s)
22
2.10 (s)
8.45 (s)
2.12 (s)
11.0 (s)
NH
OH-7
OH-12
11.04 (s)
4.92 (d, 5.0)
10.99 (s)
4.91 (d, 4.5)
4.92 (d, 5.0)
a
b
c
Recorded at 500 MHz.
Recorded in CDCl3 (dH 7.24).
Recorded in DMSO-d6 (dH 2.49).
Table 2. 13C NMRa data for haterumaimides N–Q (1–4), J (14), and
K (15)
correlations of NH proton signal at dH 8.45 to dC
179.2 (s), 176.8 (s), 46.8 (d), and 29.4 (t). Thus, the H
1
and 13C NMR data suggested that the remaining two
rings are carbocyclic. Further detailed analysis of
DQF-COSY (Fig. 1) and HOHAHA spectral data
allowed us to elucidate three partial structures, C1–C3,
C5–C7, and C9–C14. The connectivity of these partial
structures was established from the HMBC correlations
(Fig. 1) of H3-18/C-3, H3-19/C-3, H3-18/C-4, H3-19/C-4,
H2-3/C-4, H-5/C-4, H3-18/C-5, H3-19/C-5, H2-7/C-8,
H-9/C-8, H2-17/C-7, H2-17/C-8, H2-17/C-9, H-9/C-10,
H2-1/C-10, H3-20/C-10, H3-20/C-1, H3-20/C-9, H3-20/
C-5, H-5/C-10, H-13/C-16, NH/C-16, H2-14/C-15, and
NH/C-15 to give the entire carbon framework of 1, leav-
ing the chlorine atom at C-2, a hydroxyl group (or an
acetoxyl group) at C-12, and an exomethylene double
bond between C-8 and C-17. The downfield proton
chemical shift of H-7 (dH 5.10) compared to that of
H-12 (dH 4.28) indicated that the acetoxyl group residing
at C-7 and the hydroxyl group is at C-12. The position
of acetoxyl group resides at C-7 was further supported
by a strong HMBC correlation between dH 5.10 (s)
and dC 170.2 (s). Therefore, the planar structure of hat-
erumaimide N (1) was elucidated to be a monochlorinat-
ed labdane alkaloid with a succinimide moiety, as shown
in 1. The relative stereochemistry of the decalin part of 1
was determined to be 2S*, 5S*, 7S*, 9R*, and 10S* from
the NOESY correlations (Fig. 2) of H-2/H3-20, H-2/H3-
19, H3-19/H3-20, H-1b/H3-20, H-3a/H-1a, H-3a/H3-18,
H-5/H-3a, H-5/H-1a, H-5/H-7, H-5/H-9, H-5/H-6a,
H-6b/H3-20, H-6b/H3-19, H-7/H-9, and H-9/H-1a,
Compound
1b
2c
3c
4c
14c
15b
d
d
d
d
d
d
1
2
49.1
49.2
48.7
38.0
48.4
49.0
55.0
51.8
35.9
51.9
29.8
74.5
144.6
51.6
41.5
29.0
68.7
46.8
29.4
179.2
176.8
106.1
33.1
22.0
14.8
170.2
21.2
54.8
51.6
56.8
51.7
35.6
50.8
33.0
71.7
149.8
53.0
41.2
28.8
20.2
40.7
35.0
181.6
178.2
104.2
32.9
21.7
14.5
18.9
41.5
33.1
52.2
33.5
72.0
151.1
49.8
38.7
29.5
68.8
45.3
28.9
181.1
178.8
103.6
33.3
21.5
14.2
57.3
45.9
40.3
46.0
22.9
37.0
147.6
51.4
41.3
30.0
68.9
45.5
29.0
181.1
178.8
107.7
69.3
17.9
15.0
54.7
46.2
39.1
48.4
23.6
37.5
147.1
53.5
41.7
29.3
69.2
46.8
29.4
178.8
176.4
109.0
71.7
18.0
15.4
171.0
21.0
3
4
35.9
51.7
5
6
29.6
73.9
7
8
143.9
53.9
41.2
9
10
11
12
13
14
15
16
17
18
19
20
21
22
24.7
138.9
126.7
33.0
173.2
168.9
106.4
33.1
22.0
14.8
169.8
21.1
a
b
c
Recorded at 125 MHz.
Recorded in CDCl3 (dC 77.0).
Recorded in DMSO-d6 (dC 39.5).
to account for the seven sites of unsaturation required
by the molecular formula. The presence of a unique
succinimide moiety in 1 was deduced from the HMBC

J. Uddin et al. / Bioorg. Med. Chem. 14 (2006) 6954–6961
6957
a double bond between C-12 and C-13 in 2. The double
bond was tentatively assigned an E geometry as suggest-
ed by the relatively low field chemical shift (dH 6.65)22 of
H-12 due to deshielding of the carbonyl (C-16) and the
lack of NOEs between H-12 and H2-14. A detailed anal-
ysis of the 1D and 2D NMR data finally disclosed the
planar structure of haterumaimide O. Since the NOEs
observed for the decalin part of 2 resembled those de-
scribed above for 1, both compounds had to possess
an identical stereochemistry for the decalin moiety. We
could not conclude that 2 is a natural product, since it
is possible that 2 may be derived from 1 by dehydration
during isolation.
Figure 1. Planar structure of haterumaimide
DQF-COSY and selected HMBC correlations.
N (1) based on
Haterumaimide P (3) was isolated as a minor constitu-
ent and had a molecular formula of C22H30ClNO3, as
determined by HRFABMS [m/z 368.1982 (M+H)+,
1
D ꢀ1.0 mmu]. The H and 13C NMR data (Tables 1
and 2) were similar to those of chlorolissoclimide (17).
In contrast to 17, compound 3 contained a methylene
instead of an oxymethine. Extensive analysis of the 1D
and 2D NMR data led to the planar structure of 3.
Although the relative stereochemistry of the decalin part
of 3 was determined in the same manner as described
above for 1, the relative stereochemistry at C-13 was
not assigned.
The molecular formula of haterumaimide Q (4) was
deduced to be C22H31NO4 based on HRFABMS [m/z
372.2159 (M+Na)+, D +0.8 mmu] and 13C NMR data.
Despite the absence of a chlorine atom in the molecular
formula, the 1H, 13C NMR, and IR spectra of 4 suggest-
ed that it might belong to the same class of labdane
Figure 2. Selected NOEs of haterumaimide N (1).
1
alkaloids. The H and 13C NMR data resembled those
together with large vicinal coupling constants of H-2 (tt,
J = 12.0, 3.5 Hz), H-5 (dd, J = 12.0, 5.0 Hz), and H-7
(dd, J = 12.0, 5.0 Hz). It was more challenging to assign
the relative configurations at C-12 and C-13 because of
the rotational freedom enjoyed by the succinimide moi-
ety in contrast to the bulk of the molecule. However, a
careful examination of the NOESY correlations of H-
12/H-13, H-12/H-9, H-11a/H-12, H-13/H-14a, and H-
11b/H-14b together with vicinal coupling constants of
1.5 Hz for H-13/H-12, 5.50 Hz for H-13/H-14a, and
9.0 Hz for H-13/H-14b, and comparison with related
compounds isolated from the same organisms suggested
the 12S* and 13R* relative configurations.10–14 Base-cat-
alyzed hydrolysis of 1 furnished chlorolissoclimide (17).
of chlorolissoclimide (17). The only difference between
the two compounds was the presence of a methylene
in 4 instead of the chlorinated methine at C-2 in 17, indi-
cating that 4 is the dechloro analogue of 17. The planar
structure and the relative stereostructure of 4 were deter-
mined in the same manner as described above for 1.
The structures of haterumaimides J (14) and K (15) were
determined in the same way as for 1, as described in the
earlier communication.15 Base treatment (NaOMe/
1
MeOH) of 15 gave 14 with identical H and 13C NMR
data, [a]D, and HPLC retention time.
2.2. Structure–activity relationships
1
The H and 13C NMR data and [a]D of the hydrolysate
of 1 were identical to those of 17, which confirmed the
structure of haterumaimide N (1).
So far we have isolated a total of 17 labdane alkaloids
(1–17) with a unique succinimide moiety, including two
known compounds.10,11,15 Six of these are dichlorinated
(5, 6, 7, 8, 9, and 16), 10 are monochlorinated
(1, 2, 3, 10, 11, 12, 13, 14, 15, and 17), and one is a dechlo-
ro analogue (4). Beyond the basic skeletal structure and
the chlorine substituent(s), this class of alkaloids possess-
es a wide variety of functionalities. Due to their intrinsic
structural variety and impressive biological activities, we
were interested in establishing the structure–activity rela-
tionships of natural haterumaimides A–K and N–Q. We
also prepared three synthetic derivatives of 17 (18, 19,
and 20) to examine the effects of the hydroxyl group at
C-12 and the imide NH in ring C on toxicity. N-Methyl
The molecular formula of haterumaimide O (2) was de-
duced to be C22H30ClNO5 based on HRFABMS [m/z
408.1914 (M+H)+, D ꢀ2.7 mmu]. The molecular formu-
la of 2 suggested an additional degree of unsaturation
1
compared to that of haterumaimide N (1). The H and
13C NMR data (Tables 1 and 2) were similar to those
of 1 except for the two carbon signals resonating at dC
138.9 (d) and 126.7 (s), and a proton chemical shift at
dH 6.65 (tt, J = 7.5, 2.5 Hz). The HMBC correlations
of dH 6.65 (H-12) to dC 126.7 (s, C-13), 24.7 (t, C-11),
and 33.0 (t, C-14) clearly demonstrated the presence of

6958
J. Uddin et al. / Bioorg. Med. Chem. 14 (2006) 6954–6961
compound 18 was obtained by the reaction between 17
and CH2N2. Compounds 17 and 18 were treated with
Ac2O/pyridine to give 19 and 20, respectively. The cyto-
toxicity of compounds 1–20 was evaluated against mouse
lymphocytic leukemia (P388) cells. The toxicity of 1–20
against fertilized sea urchin eggs was also evaluated.
The results are shown in Table 3.
congener 4 shows about a 30-fold decrease in toxicity
compared with the monochloro analogue 17. A similar
trend in the structure–activity relationship for 1–20
was observed in the assay with fertilized sea urchin eggs
(Table 3).
2.3. Conclusion
Compound 14 was the most cytotoxic against P388 cells,
with IC50 value of 0.23 ng/mL, followed by 15 at
0.45 ng/mL. Thus, acetylation of the hydroxyl group at
C-18 of 14 has no significant effect on toxicity. In a com-
parison of 14 and 15 with 17 and its acetyl congener 1,
the cytotoxicity increases three fold when the hydroxyl
group at C-7 in 17 and the acetoxyl group at C-7 in 1
are replaced with those at C-18 in 14 and 15, respective-
ly. Compounds 9 and 10 showed pronounced toxicity
with IC50 values of 4.1 and 5.5 ng/mL. Oxidation of a
C-6 hydroxyl group in 9 or 10 to a keto group in 6 or
11 significantly reduced the toxicity, which suggested
that the relative hydrophilicity of this part of the mole-
cule contributes to the activity. The conversion of an
exomethylene double bond in 6 or 11 to a trisubstituted
double bond in 8 or 12 does not markedly change the
cytotoxic activity. The effect of the hydroxyl group at
C-12 and the imide NH in ring C on the cytotoxicity
is more pronounced, since compounds 2, 3, 7, 13, 18,
19, and 20 show remarkably low cytotoxicity compared
to 17. There is a two fold decrease in cytotoxicity with
the acetyl group at C-7 in 5 or 1, compared to the free
alcohol 16 or 17. Comparison of the cytotoxicity of
dichloro and monochloro analogues clearly showed that
the chlorine atom at C-3 does not significantly contrib-
ute to the cytotoxicity. The chlorine atom at C-2 has a
pronounced effect on the cytotoxicity, since the dechloro
In this report, we have described the isolation, struc-
ture elucidation, and biological activities of haterumai-
mides N–Q (1–4), detailed experimental procedures for
haterumaimides J (14) and K (15), and structure–activ-
ity relationships of haterumaimides A–K (5–15), N–Q
(1–4), dichlorolissoclimide (16), chlorolissoclimide
(17), and three synthetic derivatives of 17 (18, 19,
and 20). Metabolites 1–17 with chlorine and a succini-
mide moiety are rare in nature. Based on the SAR re-
sults, it appears that several structural features of
haterumaimides, such as the presence of hydroxyl
groups at C-6, C-7, C-12, and C-18, a chlorine atom
at C-2, and an imido NH in ring C, are very important
for the cytotoxicity. The hydrophilic OH and (CO)2NH
groups in haterumaimides might increase cell mem-
brane permeability to the molecules and/or enhance
the stereo-electronic interaction between the molecules
and a target molecule. Further chemical and biological
studies on these unique metabolites are in progress in
our laboratory.
3. Experimental
3.1. General experimental procedures
Optical rotations were measured on a JASCO DIP-1000
polarimeter. IR and UV spectra were measured using a
JASCO FT/IR-300 spectrometer and a JASCO UVDEC
610 spectrophotometer, respectively. HRFABMS were
determined on a JEOL JMS-LG2000 mass spectrome-
Table 3. Bioactivities of haterumaimides (A–K), (N–Q) and com-
pounds 16, 17, 18, 19, and 20
1
ter. The H, 13C, and 2D NMR spectra were recorded
Haterumaimides
IC50 (lg/mL)
against P388
Inhibition (%) of the
division of fertilized
sea urchin eggs at 3 ppm
1
on a JEOL a-500 spectrometer, and H and 13C chemi-
cal shifts were referenced to the solvent peaks (dH 2.49
and dC 39.5 in DMSO-d6 and dH 7.24 and dC 77.0 in
CDCl3). Column chromatography was performed on
Kieselgel 60 (70–230 mesh, Merck) and Cosmosil
75C18-OPN (Nacalai Tesque). HPLC was performed
using a COSMOSIL Si60 HPLC (5SL, 10 · 250 mm)
A (5)
B (6)
C (7)
D (8)
E (9)
F (10)
G (11)
H (12)
I (13)
J (14)
K (15)
N (1)
O (2)
P (3)
Q (4)
16
3.5 · 10ꢀ3
2.1
>10
100
100
10
0.9
100
100
100
100
100
50
4.1 · 10ꢀ3
5.5 · 10ꢀ3
>10
and
a
COSMOSIL ODS HPLC column (C18,
10 · 250 mm). Analytical TLC was performed using
Kieselgel 60 F254 DC-fertigplatten (Merck). All solvents
used were of reagent grade.
2.7
>10
2.3 · 10ꢀ4
4.5 · 10ꢀ4
3.4 · 10ꢀ3
1.3
100
100
100
80
3.2. Collection, extraction, and purification
The encrusting gray ascidian was collected by hand from
the coast of Hateruma Island, Okinawa, in June 1999,
and identified as Lissoclinum species. The identified
sponge was kept frozen until use. A voucher specimen
was deposited at the University of the Ryukyus (Speci-
men no. URKU-31). The animal specimen (1.0 kg, wet
weight) was first extracted with acetone and then con-
centrated in vacuo to give an acetone extract (5.7 g).
The acetone extract was partitioned between H2O
1.2
100
100
100
100
10
5.0 · 10ꢀ2
1.0 · 10ꢀ3a
1.7 · 10ꢀ3b
1.4
17
18
19
20
2.4
0.14
0
50
a
Data taken from Ref. 12.
Data taken from Ref. 13.
b

J. Uddin et al. / Bioorg. Med. Chem. 14 (2006) 6954–6961
6959
28
3.2.2. Haterumaimide O (2). Colorless oil; ½aꢁ +16.0 (c
(200 mL) and EtOAc (3 · 200 mL). The EtOAc extract
completely inhibited the first cleavage of fertilized sea
urchin eggs at 10 ppm. The bioactive EtOAc extract
(3.4 g) was suspended in aqueous MeOH (1:1,
500 mL). The suspension was partitioned with hexanes
(300 mL · 2), CHCl3 (300 mL · 2), and n-BuOH
(200 mL · 2), respectively, to give non-polar fatty hex-
ane extract (0.5 g), a lipophilic CHCl3 extract (2.5 g),
and polar n-BuOH extract (0.2 g). All three extracts
were tested for toxicity against fertilized sea urchin eggs.
The lipophilic CHCl3 extract completely inhibited the
first cleavage of fertilized sea urchin eggs at 5 ppm, while
the remaining two extracts were inactive. The active
CHCl3 extract was first subjected to a column chroma-
tography on silica gel (300 g) using hexanes with increas-
ing proportions of EtOAc [hexanes (500 mL)!hexanes/
EtOAc (5:1, 400 mL!1:1, 400 mL!1:3, 400 mL) and
EtOAc with increasing proportions of MeOH [EtOAc
(400 mL)!EtOAc/MeOH (9:1, 400 mL)] as eluents to
give nine fractions. All nine fractions were tested for
toxicity against fertilized sea urchin eggs at 5 ppm. Only
fraction 5 eluted with hexanes/EtOAc (1:1) and fraction
6 eluted with hexanes/EtOAc (1:3) completely inhibited
cell division in fertilized sea urchin eggs. The active fifth
fraction (0.4 g) was further chromatographed on ODS
(150 g) using 35% H2O in MeOH (250 mL) and finally
washed with MeOH (250 mL) to give two fractions.
The active polar fraction (0.3 g) was subjected to further
separation by HPLC on Si60 using hexanes/CH2Cl2/
EtOAc/MeOH (12:4:3:1) to give 12 fractions. The 4th
fraction was purified by reversed-phase HPLC on
ODS using MeOH/H2O/CH3CN (6.5:2.7:0.8) to afford
haterumaimide N (1, 13.9 mg) and haterumaimide K
(15, 3.2 mg). The 10th fraction was subjected to
reversed-phase HPLC on ODS using MeOH/H2O/
CH3CN (11:7:2) to yield haterumaimide O (2, 1.1 mg).
The 6th fraction was purified by repeated reversed-phase
HPLC on ODS using MeOH/H2O/CH3CN (6.5:2.7:0.8)
and MeOH/H2O/CH3CN (6:3:1) as a solvent system to
give haterumaimide P (3, 1.0 mg). The 12th fraction
was further purified by reversed-phase HPLC on ODS
using MeOH/H2O/CH3CN (5.4:3.6:1.0) to furnish hater-
umaimide J (14, 7.5 mg). The active sixth fraction of
aforementioned column chromatography on silica gel
(0.4 g) was chromatographed on ODS (150 g) using
30% H2O in MeOH (250 mL) and MeOH (250 mL) to
give two fractions. Both fractions were tested for toxic-
ity against fertilized sea urchin eggs, and the polar frac-
tion was found to be active at 5 ppm. The active polar
fraction (0.3 g) was subjected to further fractionation
by HPLC on Si60 using hexanes/CH2Cl2/EtOAc/MeOH
(10:3:6:1) to give five factions. The fifth fraction was
purified by normal-phase HPLC on Si60 using
hexanes/CH2Cl2/EtOAc/MeOH (12:3:4:1) to afford hat-
erumaimide Q (4, 2.5 mg).
D
0.08, MeOH); UV (MeOH) kmax (loge) 235 nm (3.9);
FT/IR (film) mmax 3300, 2920, 2840, 1720, 1715, 1710,
1660, 1380, 1280, 1120, and 1040 cmꢀ1 1H and 13C
;
NMR (DMSO-d6) data are listed in Tables 1 and 2;
HRFABMS [m/z (M+H)+ 408.1914 (calcd for
C22H3135ClNO4, 408.1941)].
29
3.2.3. Haterumaimide P (3). Colorless oil; ½aꢁ +68.8 (c
D
0.11, MeOH); UV (MeOH) kmax (loge) 210 nm (3.6);
FT/IR (film) mmax 3400, 2910, 1705, 1700, 1600, 1380,
1200, and 1080 cmꢀ1 1H and 13C NMR (DMSO-d6)
;
data are listed in Tables 1 and 2; HRFABMAS [m/z
(M+H)+ 368.1982 (calcd for C20H3135ClNO3,
368.1992)].
28
D
3.2.4. Haterumaimide Q (4). Colorless oil; ½aꢁ +36.0 (c
0.19, CHCl3); UV (CHCl3) kmax (loge) 204 nm (3.6); FT/
IR (film) mmax 3405, 2930, 1720, 1710, 1600, 1350, and
1165 cmꢀ1; 1H and 13C NMR (DMSO-d6) data are listed
in Tables 1 and 2; HRFABMAS [m/z (M+Na)+
372.2159 (calcd for C22H31NaNO4, 372.2151)].
29
D
3.2.5. Haterumaimide J (14). Colorless oil; ½aꢁ +68.0
(c 0.92, MeOH); UV (MeOH) kmax (loge) 210 nm
(3.6); FT/IR (film) mmax 3400, 2910, 1720, 1710, 1220,
and 1050 cmꢀ1; 1H and 13C NMR (DMSO-d6) data were
described in the earlier paper;11 HRFABMAS [m/z
(M+H)+ 384.1925 (calcd for C20H3135ClNO4,
384.1942)].
29
D
3.2.6. Haterumaimide K (15). Colorless oil; ½aꢁ +59.6
(c 0.19, MeOH); UV (MeOH) kmax (loge) 210 nm
(3.6); FT/IR (film) mmax 3405, 2902, 1720, 1710, 1250,
and 1040 cmꢀ1; 1H and 13C NMR (DMSO-d6) data were
described in the earlier paper;11 HRFABMAS [m/z
(M+H)+ 426.2032 (calcd for C22H3335ClNO5,
426.2047)].
3.3. Methanolysis of 1
To a solution of haterumaimide N (1.1 mg, 2.6 lmol) in
MeOH (0.5 mL) was added sodium methoxide (1.5 mg,
27.7 lmol) and the mixture was stirred at room temper-
ature for 4 h. Next, 0.5 mL of water was added to the
mixture. The organic solvent was evaporated and the
mixture was extracted with ether (3 · 0.5 mL). The com-
bined ether extract was washed with brine, dried
(MgSO4), and evaporated. The residue was purified by
normal-phase HPLC on Si60 using hexanes/EtOAc/
MeOH (6.2:3.3:0.5) to afford the alcohol, chlorolissocli-
29
mide (7, 0.9 mg, 90%) as a colorless oil: ½aꢁ +45.0 (c
D
0.05, MeOH); 1H NMR (DMSO-d6, 500 MHz) dH
11.01 (1H, s), 5.23 (1H, br s), 4.98 (1H, s), 4.93 (1H,
d, J = 4.5 Hz), 4.80 (1H, br s), 4.33 (1H, tt, J = 12.0,
4.0 Hz), 4.00 (1H, m), 3.78 (1H, m), 2.84 (1H, ddd,
J = 9.0, 4.5, 1.5 Hz), 2.55 (1H, dd, J = 17.5, 4.5 Hz),
2.48 (1H, dd, J = 17.5, 9.0 Hz), 2.10 (1H, ddd,
J = 12.0, 4.0, 1.5 Hz), 1.92 (1H, ddd, J = 12.5, 4.0,
1.5 Hz), 1.85 (1H, ddd, J = 12.5, 5.5, 2.0 Hz), 1.64
(1H, ddd, J = 12.5, 5.5, 2.0 Hz), 1.60 (1H, dd, J = 10.5,
9.5 Hz), 1.48 (1H, t, J = 12.5 Hz), 1.40 (1H, ddd,
J = 13.0, 10.5, 6.5 Hz), 1.30 (1H, t, J = 12.0 Hz), 1.23
32
3.2.1. Haterumaimide N (1). Colorless oil; ½aꢁ +59.7
D
(c 0.79, MeOH); UV (MeOH) kmax (loge) 215 nm
(3.6); FT/IR (film) mmax 3505, 3400, 2900, 1720, 1705,
1
1700, 1605, 1370, 1240, 1180, and 1050 cmꢀ1; H and
13C NMR (CDCl3) data are listed in Tables 1 and 2;
HRFABMAS [m/z (M+H)+ 426.2058 (calcd for
C22H3235ClNO5, 426.2047)].

6960
J. Uddin et al. / Bioorg. Med. Chem. 14 (2006) 6954–6961
(1H, dd, J = 13.0, 2.0 Hz), 1.10 (1H, q, J = 12.5 Hz),
0.91 (3H, s), 0.81 (3H, s), 0.61 (3H, s); 13C NMR
(DMSO-d6, 125 MHz) dC 180.8 (s, C-15), 178.5 (s,
C-16), 150.2 (s, C-8), 104.5 (t, C-17), 71.5 (d, C-7),
66.5 (d, C-12), 56.6 (d, C-2), 51.6 (t, C-3), 50.5 (d,
C-5), 49.2 (d, C-9), 48.3 (t, C-1), 45.3 (d, C-13), 41.0 (s,
C-10), 35.5 (s, C-4), 32.8 (t, C-6), 32.8 (q, C-18), 29.9 (t,
C-11), 28.8 (t, C-14), 21.9 (q, C-9), 14.5 (q, C-20).
IR (CHCl3) mmax 3400, 3027, 2964, 1726, 1372 cmꢀ1
;
1H NMR (CDCl3, 500 MHz) dH 7.89 (1H, br s, NH),
5.43 (1H, m, H-12), 5.20 (1H, br s, Ha-17), 5.18 (1H,
br s, Hb-17), 5.05 (1H, dd, J = 9.0, 4.0 Hz, H-7), 4.13
(1H, m, H-2), 3.01 (1H, m, H-13), 2.80 (2H, m, Ha,b
-
14), 2.19 (1H, br d, J = 10.5 Hz, Ha-1), 2.12 (3H, s,
COCH3), 2.06 (1H, m, Ha-11), 2.02 (1H, m, Ha-6),
2.01 (1H, m, Ha-3), 1.98 (3H, s, COCH3), 1.61 (1H, m,
H-9), 1.59 (1H, m, Hb-11), 1.50 (1H, t, J = 13.0 Hz,
Hb-3), 1.32 (1H, q, J = 11.5 Hz, Hb-6), 1.20 (1H, m,
Hb-1), 1.19 (1H, m, H-5), 0.94 (3H, s, CH3-18), 0.84
(3H, s, CH3-19), 0.74 (3H, s, CH3-20); 13C NMR
(CDCl3, 125 MHz) dC 176.5 (s, C-15), 175.2 (s, C-16),
169.9 (s, COCH3), 169.7 (s, COCH3), 142.5 (s, C-8),
107.3 (t, C-17), 74.0 (d, C-7), 70.6 (d, C-12), 54.7 (d,
C-2), 52.1 (d, C-5), 51.8 (t, C-3), 51.0 (d, C-9), 49.2 (t,
C-1), 43.8 (d, C-13), 41.4 (s, C-10), 35.9 (s, C-4), 33.1
(q, C-18), 29.9 (t, C-14), 29.7 (t, C-6), 26.4 (t, C-11),
22.0 (q, C-19), 21.1 (q, COCH3), 20.8 (q, COCH3),
14.7 (q, C-20); HRESIMS [m/z (M+Na)+ 490.1972
(calcd for C24H3435ClNNaO6, 490.1964)].
3.4. Methanolysis of 15
To a solution of haterumaimide K (0.5 mg, 1.2 lmol) in
MeOH (0.5 mL) was added sodium methoxide (1.0 mg,
18.5 lmol) and the mixture was stirred at room temper-
ature for 4 h. Next, 0.5 mL of water was added to the
mixture. The organic solvent was evaporated and the
mixture was extracted with ether (3 · 0.5 mL). The com-
bined ether extract was washed with brine, dried
(MgSO4), and evaporated. The residue was purified by
reversed-phase HPLC on ODS using MeOH/H2O/
CH3CN (5.4:3.6:1.0) to afford haterumaimide J (14,
1
0.4 mg, 90%) as a colorless oil. The H and 13C NMR
data and [a]D of the hydrolysate of 15 were described
3.7. Methylation of 17 followed by acetylation
previously.15
Compound 17 (4.0 mg, 10.41 lmol) dissolved in MeOH
(1.0 mL) and a solution of diazomethane in ether were
added until all of 17 had been consumed as determined
by TLC. The reaction mixture was concentrated and
dried in vacuo. The crude product was dissolved in pyr-
idine (0.25 mL) and Ac2O (0.10 mL), and the solution
was allowed to stand for 8 h. Evaporation of the solvent
in vacuo followed by HPLC separation [COSMOSIL,
3.5. Methylation of 17
Compound 17 (4.0 mg, 10.41 lmol) was dissolved in
MeOH (1.0 mL), and a solution of diazomethane in
ether was added until all of 17 had been consumed as
determined by TLC. The reaction mixture was concen-
trated and the crude product was purified by HPLC
[COSMOSIL, 5SL, hexanes/EtOAc (1:2.5)] to give com-
5SL, hexanes/EtOAc(1:1)] yielded 2.9 mg of 20: colorless
16
D
16
pound 18 (2.3 mg, 55%): colorless oil; ½aꢁ +70 (c 1.0
oil; ½aꢁ +38 (c 1.0, MeOH); IR (CHCl3) mmax 3019,
D
1
MeOH); IR (CHCl3) mmax 3612, 3019, 2969, 1699,
2965, 1736, 1705, 1223, 1208 cmꢀ1; H NMR (CDCl3,
500 MHz) dH 5.45 (1H, m, H-12), 5.20 (1H, br s, Ha-
17), 5.18 (1H, br s, Hb-17), 5.04 (1H, dd, J = 11.5,
5.0 Hz, H-7), 4.13 (1H, m, H-2), 2.97 (3H, s, NCH3),
2.94 (1H, m, H-13), 2.73 (2H, m, Ha,b-14), 2.19 (1H,
br d, J = 10.5 Hz, Ha-1), 2.12 (3H, s, COCH3), 2.06
(1H, m, Ha-11), 2.02 (1H, m, Ha-6), 2.01 (1H, m,
Ha-3), 1.93 (3H, s, COCH3), 1.61 (1H, m, H-9), 1.57
(1H, m, Hb-11), 1.50 (1H, t, J = 11.5 Hz, Hb-3), 1.32
(1H, q, J = 11.5 Hz, Hb-6), 1.20 (1H, m, Hb-1), 1.19
(1H, m, H-5), 0.94 (3H, s, CH3-18), 0.84 (3H, s, CH3-
19), 0.74, (3H, s, CH3-20). 13C NMR (CDCl3,
125 MHz) dC 176.9 (s, C-15), 175.9 (s, C-16), 169.9 (s,
COCH3), 169.8 (s, COCH3), 142.6 (s, C-8), 107.3 (t,
C-17), 74.4 (d, C-7), 70.7 (d, C-12), 54.8 (d, C-2), 52.1
(d, C-5), 51.8 (t, C-3), 51.1 (d, C-9), 49.3 (t, C-1), 42.6 (d,
C-13), 41.4 (s, C-10), 36.0 (s, C-4), 33.1 (q, C-18), 29.8 (t,
C-6), 28.9 (t, C-14), 26.1 (t, C-11), 25.0 (q, NMe), 22.0
(q, C-19), 21.1 (q, COCH3), 20.9 (q, COCH3), 14.7 (q,
C-20); HRESIMS [m/z (M+Na)+ 504.2129 (calcd for
C25H3635ClNNaO6, 504.2120)].
1
1387, 1222 cmꢀ1; H NMR (CDCl3, 500 MHz) dH 5.33
(1H, br s, Ha-17), 4.90 (1H, br s, Hb-17), 4. 34 (1H, m,
H-12), 4.14 (1H, m, H-2), 3.98 (1H, dd, J = 11.0,
5.5 Hz, H-7), 2.97 (3H, s, NCH3), 2.81 (1H, m, H-13),
2.79 (1H, m, Ha-14), 2.63 (1H, dd, J = 18.0, 10.5 Hz,
Hb-14), 2.20 (1H, br d, J = 10.5 Hz, Ha-1), 2.08 (1H,
m, Ha-6), 2.00 (1H, m, Ha-3), 1.81 (1H, m, Ha-11),
1.60 (1H, m, Hb-11), 1.52 (1H, t, J = 12.5 Hz, Hb-3),
1.24 (1H, m, Hb-6), 1.22 (1H, m, Hb-1), 1.16 (1H, dd,
J = 12.5, 1.5 Hz, H-5), 0.96 (3H, s, CH3-18), 0.84 (3H,
s, CH3-19), 0.72 (3H, s, CH3-20): 13C NMR (CDCl3,
125 MHz) dC 178.9 (C-15), 176.7 (s, C-16), 169.9 (s,
COCH3), 149.5 (s, C-8), 105.4 (t, C-17), 73.3 (d, C-7),
68.9 (d, C-12), 55.0 (d, C-2), 52.3 (d, C-5), 51.9 (t,
C-3), 51.6 (d, C-9), 49.3 (t, C-1), 45.4 (d, C-13), 41.7
(s, C-10), 35.9 (s, C-4), 33.2 (q, C-18), 33.0 (t, C-6),
29.1 (t, C-11). 28.2 (t, C-14), 24.9 (q, NMe). 22.0 (q,
C-19), 14.9 (q, C-20); HRESIMS [m/z (M+Na)+
420.1917 (calcd for C21H32335ClNNaO4, 420.1910)].
3.6. Acetylation of 17
Compound 17 (5.0 mg, 13.02 lmol) was dissolved
in pyridine (0.250 mL) and Ac2O (0.100 mL) and the
solution was allowed to stand for 30 min. Evaporation
of the solvent in vacuo followed by HPLC separation
Acknowledgments
We thank Professor Tatsuo Higa, University of the
Ryukyus, Okinawa, for his invaluable suggestions
throughout this study, and Dr. Yuichi Hirose of the
same University for identifying the ascidian. This work
[COSMOSIL, 5SL, hexanes/EtOAc (1:1)] yielded
16
D
1.3 mg of 19: colorless oil; ½aꢁ +47 (c 0.58 MeOH);

J. Uddin et al. / Bioorg. Med. Chem. 14 (2006) 6954–6961
6961
11. Uddin, M. J.; Kokubo, S.; Suenaga, K.; Ueda, K.;
Uemura, D. J. Nat. Prod. 2001, 64, 1169.
12. Malochet-Grivois, C.; Cotelle, P.; Biard, J. F.; Henichart,
J. P.; Debitus, C.; Roussakis, C.; Verbist, J. F. Tetrahe-
dron Lett. 1991, 32, 6701.
was supported in part by Wako Pure Chemical Indus-
tries Ltd. and by Grants-in-Aid (Nos. 11558079 and
12045235) from the Ministry of Education, Science,
Sports and Culture of Japan.
13. Biard, J. F.; Malochet-Grivois, C.; Roussakis, C.; Cotelle,
P.; Henichart, J. P.; Debitus, C.; Verbist, J. F. Nat. Prod.
Lett. 1994, 4, 43.
References and notes
14. Uddin, M. J.; Kokubo, S.; Suenaga, K.; Ueda, K.;
Uemura, D. The World Chemistry Congress, Brisbane,
Australia, 2001, abstract, PFD25, p 562.
15. Uddin, M. J.; Kokubo, S.; Suenaga, K.; Ueda, K.;
Uemura, D. Chem. Lett. 2002, 10, 1028.
1. Davidson, B. S. Chem. Rev. 1993, 1771.
2. Blunt, J. W.; Copp, B. R.; Munro, M. H. G.; Northcote,
P. T.; Prinsep, M. R. Nat. Prod. Rep. 2006, 23, 26, and
previous reports in this series.
3. Rinehart, K. L. Med. Res. Rev. 2000, 20, 1, and references
cited therein.
4. Newman, D. J.; Cragg, G. M. J. Nat. Prod. 2004, 67, 1216.
5. Haefner, B. Drug Discovery Today 2003, 8, 536.
6. Jimeno, J.; Aracil, M.; Tercero, J. C. J. Transl. Med. 2006,
4, 3.
7. Cesne, A. L.; Blay, J. Y.; Van Oosterom, J. A.; Verweij, J.;
Radford, J.; Lorigan, P.; Rodenhuis, S.; Ray-Coquard, I.;
Bonvalot, S.; Collin, F.; Jimeno, J.; Paola, E. D.; Van
Glabbeke, M.; Nielsen, O. S. J. Clin. Oncol. 2005, 23, 576.
8. Ueda, K.; Hu, Y. Tetrahedron Lett. 1999, 40, 6305.
9. Takada, N.; Sato, H.; Suenaga, K.; Arimoto, H.; Yamada,
K.; Ueda, K.; Uemura, D. Tetrahedron Lett. 1999, 40,
6309.
16. Fu, X.; Palomar, A. J.; Hong, E. P.; Schmitz, F. J.;
Valeriote, F. A. J. Nat. Prod. 2004, 67, 1415.
17. Robert, F.; Gao, H. Q.; Donia, M.; Merrick, W. C.;
Hamann, M. T.; Pettetier, J. RNA 2006, 12, 717.
18. Malochet-Grivois, C.; Roussakis, C.; Robillard, N.; Biard,
J. F.; Riou, D.; Debitus, C.; Verbist, J. F. Anti-Cancer
Drug Des. 1992, 7, 493.
19. Jung, M. E.; Gomez, A. Tetrahedron Lett. 1993, 34, 2891.
20. Toupet, L.; Biard, J. F.; Verbist, J. F. J. Nat. Prod. 1996,
59, 1203.
21. Gribble, G. W. Acc. Chem. Res. 1998, 31, 141.
22. Evaluation of the chemical shift (dH-12 = 6.56 for an E
isomer, dH-12 = 6.08 for a Z isomer) of the proton at
C-12 by ChemNMR (CambridgeSoft Corporation, MA,
USA) also supported the E configuration of the double
bond.
10. Uddin, M. J.; Kokubo, S.; Suenaga, K.; Ueda, K.;
Uemura, D. Heterocycles 2001, 54, 1039.