Absolute stereostructures of cell-adhesion inhibitors, macrosphelides C, E-G and I, produced by a Periconia species separated from an Aplysia sea hare

Article abstract of DOI:10.1039/b104337b

Macrosphelides E-I have been isolated, along with known macrosphelides A and C, from a strain of Periconia byssoides originally separated from the sea hare Aplysia kurodai, and the absolute stereostructures of macrosphelides E-G and I have been elucidated on the basis of spectroscopic analyses using 1D and 2D NMR techniques and some chemical transformations. In addition, the absolute configuration of macrosphelide C, previously undetermined, has been established by X-ray analysis and application of the modified Mosher method. Macrosphelides E-H inhibited the adhesion of human-leukaemia HL-60 cells to HUVEC.

Full text of DOI:10.1039/b104337b

View Article Online / Journal Homepage / Table of Contents for this issue
Absolute stereostructures of cell-adhesion inhibitors,
macrosphelides C, E–G and I, produced by a Periconia species
separated from an Aplysia sea hare
1
Takeshi Yamada,a Masashi Iritani,a Mitsunobu Doi,a Katsuhiko Minoura,a Tadayoshi Itob and
Atsushi Numata*a
a Osaka University of Pharmaceutical Sciences, 4-20-1, Nasahara, Takatsuki, Osaka 569-1094,
Japan
b Institute for Fermentation, Osaka, Juso-Honmachi, Yodogawa-ku, Osaka 532-0024, Japan
Received (in Cambridge, UK) 16th May 2001, Accepted 25th September 2001
First published as an Advance Article on the web 23rd October 2001
Macrosphelides E–I have been isolated, along with known macrosphelides A and C, from a strain of Periconia
byssoides originally separated from the sea hare Aplysia kurodai, and the absolute stereostructures of macrosphelides
E–G and I have been elucidated on the basis of spectroscopic analyses using 1D and 2D NMR techniques and some
chemical transformations. In addition, the absolute configuration of macrosphelide C, previously undetermined, has
been established by X-ray analysis and application of the modified Mosher method. Macrosphelides E–H inhibited
the adhesion of human-leukaemia HL-60 cells to HUVEC.
Introduction
Based on the fact that some of the bioactive materials isolated
from marine animals have been produced by bacteria, we have
chosen to seek new antitumour metabolites from microorgan-
isms inhabiting the marine environment.1–3 As part of this pro-
gram, we have made a search for antitumour compounds from
a strain of Periconia byssoides OUPS-N133 which was separ-
ated from the sea hare Aplysia kurodai, and isolated pericosine
A as an antitumour compound by bioassay-directed fraction-
ation (cytotoxicity against P388 lymphocytic leukaemia cells)
from a culture broth of this fungal strain.4 In the course of this
experiment, we have found five new 16-membered macrolides,
macrosphelides E–I 1–5, along with known macrosphelides A 6
and C 7, from non-cytotoxic fractions from column chroma-
tography. Omura and co-workers have found that macrosphel-
ides A 6–D, produced by Microsphaeropsis sp., potently inhibit
cell–cell adhesion,5 and reported the absolute stereostructures
and total synthesis of macrosphelides A 6 and B,6,7 and the
planar structures of macrosphelides C, D, J and K.8,9 Arthritis
and metastasis have been found to be associated with cell-
adhesion molecules.10,11 This has evoked wide interest in inhib-
itors of cell adhesion because of the potential for the treatment
of inflammation and tumour metastasis. Macrosphelides E 1–H
4 isolated in this experiment also inhibited the adhesion of
human-leukaemia HL-60 cells to human-umbilical-vein endo-
thelial cells (HUVEC). We describe herein the absolute stereo-
structures12 of macrosphelides E 1–G 3, I 5 and C 7, and the
planar structure of macrosphelide H 4 in addition to their
inhibition of cell adhesion. The absolute stereostructure of 1
and the planar structures of 2–4 have been briefly reported in a
preliminary form.4
atography and high-performance liquid chromatography
(HPLC) to afford macrosphelides A 6, C 7, and E 1–I 5.
The known macrosphelides A 6 and C 7 were identified by
comparison of spectral data with published values.6,8
Macrosphelide E 1 had the molecular formula C16H22O8
established by the [M ϩ H]ϩ peak of 1 in high-resolution
electron-impact mass spectrometry (HREIMS). Its IR spec-
trum exhibited bands at 3438, 1732, 1720, 1692, 1665 and 1648
cmϪ1, characteristic of hydroxy and carbonyl groups, and a
double bond. A close inspection of the 1H and 13C NMR
spectra of 1 (Table 1) by DEPT and 1H–13C correlation
spectroscopy (COSY) experiments revealed the presence of
three secondary methyl (C-17, C-18 and C-19), one sp3-
hybridized methylene (C-2), five oxygen-bearing sp3-methines
(C-3, C-8, C-9, C-14 and C-15), two 1,2-disubstituted double
bonds (C-6, C-7, C-12 and C-13) and three ester/lactone car-
Results and discussion
The fungal strain was cultured at 27 ЊC for 4 weeks in a medium
containing 1% malt extract, 1% glucose and 0.05% peptone in
artificial seawater adjusted to pH 7.5. The AcOEt extract of
the culture filtrate was purified by fractionation employing a
combination of Sephadex LH-20 and silica gel column chrom-
1
bonyl groups (C-1, C-5 and C-11). The H–1H COSY analysis
of 1 led to three partial structural units as shown by bold-faced
lines in Fig. 1, which were supported by HMBC correlations
(Table 1). The E-geometry of both the ∆6- and ∆12-double
3046
J. Chem. Soc., Perkin Trans. 1, 2001, 3046–3053
This journal is © The Royal Society of Chemistry 2001
DOI: 10.1039/b104337b

View Article Online
Table 1 1H and 13C NMR data of macrosphelide E 1 in CDCl3
a
Position
δH
J/Hz
1H–1H COSY
δC
HMBC (C)
1
2 A
B
3
5
6
7
8
9
11
12
13
14
15
17
18
19
8-OH
14-OH
170.76 (q)b
40.22 (s)
2.52 dd
2.7 dd
5.22 quint d
15.6 (2B), 6.5 (3)
15.6 (2A), 3.5 (3)
6.5 (2A, 19), 3.5 (2B)
3
3
2
1, 3, 19
1, 3, 9
1, 2A, 2B, 5, 19
66.83 (t)
166.48 (q)
122.37 (t)
145.99 (t)
74.18 (t)
75.02 (t)
165.39 (q)
122.77 (t)
145.53 (t)
73.35 (t)
74.38 (t)
17.29 (p)
17.63 (p)
19.32 (p)
6.08 dd
6.97 dd
4.30 br s
5.04 qd
15.7 (7), 1.5 (8)
15.7 (6), 4.2 (8)
6
6, 8
7, 9, 8-OH
8, 18
5, 7, 8
5, 6, 8, 9
6, 7, 9, 18
7, 8, 18, 11
6.8 (18), 2.0 (8)
5.97 dd
6.77 dd
4.09 dd
4.89 qd
1.23 d
1.37 d
1.33 d
15.6 (13), 1.5 (14)
15.6 (12), 6.9 (14)
6.9 (13), 5.3 (15)
6.5 (17), 5.3 (14)
6.4 (15)
6.8 (9)
6.5 (3)
13
12, 14
13, 15, 14-OH
14, 17
15
9
3
8
14
11, 13, 14
11, 12, 14, 15
12, 13, 15, 17
1, 13, 14, 17
14, 15
8, 9
2, 3
3.88 br s
3.72 br s
a 1H Chemical-shift values (δ/ppm from SiMe4) followed by multiplicity and then the coupling constant. Figures in parentheses indicate the proton
coupling with that position. b Letters p, s, t and q, in parentheses, indicate, respectively, primary, secondary, tertiary and quaternary carbons, assigned
by DEPT.
Scheme 1 Reaction conditions: (a) i. K2CO3, ii. HCl, iii. CH2N2; (b) MeOH, H2SO4; (c) p-BrC6H4COCl, pyridine; (d) (CH3)2C(OCH3)2, PPTS.
the proton signals of the two isopropylidene methyl groups
appeared with markedly different shifts (δH 1.38 and 1.52),13
and NOEs were observed from one of the isopropylidene
methyl groups (δH 1.38) to both H-4 and H-5 and from the
other methyl group to H-3. From this evidence compound 12
was deduced to be the anti-acetonide and identified by com-
parison with published data.14 The absolute configuration
Fig. 1 Selected 1H–1H COSY and HMBC correlations in macro-
sphelide E 1.
could not been determined because of a vanishingly small value
of the published specific optical rotation ([α]D Ϫ0.55 × 10Ϫ1 dm3
14
molϪ1 cmϪ1
)
and the small amount of the acetonide. On the
bonds was deduced from the coupling constants (J6,7 15.7 Hz
and J12,13 15.6 Hz) of the olefinic protons. The connection of
these three units and the remaining ester moiety was deter-
mined on the basis of the key HMBC correlations summarized
in Fig. 1, and the planar structure of 1 was elucidated.
Since the stereochemistry of 1 could not be deduced from
NOESY experiments, compound 1 was degraded to the methyl
esters of the constituent carboxylic acids. Alkaline hydrolysis of
1 followed by treatment with diazomethane or acid-catalyzed
methanolysis of 1 in MeOH gave only two products, methyl
esters (respectively 8 and 9) of 3-hydroxybutyric acid and
4,5-dihydroxyhex-2E-enoic acid, which were isolated as p-
bromobenzoates (respectively 10 and 11) because of the
volatility of 8 (Scheme 1). This fact implied that the two dihy-
droxycarboxylic acid moieties of 1 have the same stereo-
chemistry. In the acetonide 12 derived from the resulting ester 9,
other hand, bis-p-bromobenzoate derivative 11 from the methyl
ester 9 exhibited a weak single curve at 240 nm (∆ ϩ1.26) in the
CD spectrum. As reported by Harada, syn-1,2-acyclic diol
dibenzoates exhibit typical exciton split Cotton effects in the
CD spectra, from which their absolute configuration can be
determined,15 whereas the absolute configuration of anti-
dibenzoates is difficult to deduce from the CD spectra because
of a weak single Cotton effect.16 Therefore, the CD data of 11
did not define its absolute configuration, though it was assumed
to be 4R,5S on the basis of the single Cotton effect. The other
p-bromobenzoate 10, derived from 8, was identified as methyl
(3R)-3-(p-bromobenzoyloxy)butyrate by comparison of spec-
tral data including CD spectra with the synthetic compound
prepared from commercial methyl (3R)-3-hydroxybutyrate,
implying that the chirality of the 3-hydroxybutyric acid moiety
in 1 is R.
J. Chem. Soc., Perkin Trans. 1, 2001, 3046–3053
3047

View Article Online
Table 2 1H and 13C NMR data of macrosphelide F 2 in CDCl3
a
Position
δH
J/Hz
1H–1H COSY
δC
HMBC (C)
1
2 A
B
3
5
6
7
8 A
B
9
11
12
13
14
15
17
18
19
14-OH
170.78 (q)b
40.72 (s)
2.59 dd
2.67 dd
5.31 dqd
15.8 (2B), 7.8 (3)
15.8 (2A), 3.3 (3)
7.8 (2A), 6.5 (19), 3.3 (2B)
2B, 3
2A, 3
2
1, 3, 19
1, 3, 19
1, 2A, 2B, 5, 19
66.47 (t)
165.11 (q)
124.55 (t)
143.41 (t)
37.73 (s)
5.80 dt
15.7 (7), 1.3 (8A, 8B)
7
5, 7, 8B
5, 6, 8A, 8B, 9
7, 9, 18
6, 7, 9, 18
7, 8A, 18, 11
6.90 ddd
2.40 dddd
2.72 dddd
5.15 quint d
15.7 (6), 7.5 (8A), 6.8 (8B)
14.6 (8B), 7.5 (7), 6.8 (9), 1.3 (6)
14.6 (8A), 6.8 (7), 4.8 (9), 1.3 (6)
6.8 (8A, 18), 4.8 (8B)
6, 8A, 8B
7, 8B, 9
7, 8A, 9
8A, 8B, 18
68.88 (t)
164.89 (q)
123.02 (t)
144.14 (t)
73.56 (t)
76.04 (t)
17.49 (p)
19.87 (p)
19.81 (p)
6.10 dd
6.86 dd
4.22 br s
4.95 qd
1.32 d
1.39 d
1.36 d
15.6 (13), 1.9 (14)
15.6 (12), 4.2 (14)
13
12, 14
13, 15, 14-OH
14, 17
15
9
3
11, 13, 14
11, 12, 14, 15
12, 13, 15, 17
1, 13, 14, 17
14, 15
8A, 9
2, 3
6.6 (17), 3.9 (14)
6.6 (15)
6.8 (9)
6.5 (3)
3.00 br s
14
a,b As in Table 1.
The third benzoate, 14, was identified as methyl (5S)-5-(p-
bromobenzoyloxy)hex-2E-enoate by comparison of spectral
data including CD, with the enantiomer 15, which was derived
from methyl (5R)-5-hydroxyhex-2E-enoate18,19 prepared from
methyl (3R)-3-hydroxybutyrate by a modification of the
procedure reported previously.19,20 This evidence allowed
assignment of absolute stereostructure 2 to macrosphelide F.
Macrosphelide G 3 had the same molecular formula as 2 as
deduced from HREIMS. Its 1H and 13C NMR signals (Table 3)
1
showed close correspondence with those of 2 (Table 2). H–1H
COSY (8-H/7-H and 9-H, 7-H/6-H, 13-H/12-H, 14A-H and
14B-H, and 15-H/14A-H and 14B-H) and HMBC (7-H/C-5,
3-H/C-5, 9-H/C-11, 13-H/C-11 and 15-H/C-1) correlations in 3
implied that the C-8 methylene and C-14 methine in 2 change
places with respect to each other in 3. Thus, compound 3 is
a positional isomer of 2. In order to determine the absolute
configuration for 3, methanolysis of 3 followed by treatment of
p-bromobenzoyl chloride was carried out using the same pro-
cedure as above with macrosphelide F 2, to give the same three
p-bromobenzoates as those obtained from 2 (Scheme 2). This
evidence allowed assignment of absolute stereostructure 3 to
macrosphelide G.
Fig. 2 1H Chemical-shift differences (∆δ = δS Ϫ δR) between the (R)-
and (S)-MTPA esters 13a and 13b of macrosphelide E 1.
Since the chirality of the 4,5-dihydroxyhex-2E-enoic acid
moiety in 1 could not be determined by CD, the modified
Mosher method17 was applied for determination of the abso-
lute configuration of macrosphelide E 1. The 1H chemical-shift
differences between the ‘(R)- and (S)-2-methoxy-2-phenyl-2-
(trifluoromethyl)acetic acid’ (MTPA) esters 13a and 13b of
compound 1 are shown in Fig. 2. The result suggested 8R and
14R configurations, and hence allowed assignment of absolute
stereostructure 1 with the 3R,8R,9S,14R,15S configuration to
macrosphelide E. From this result, the absolute configuration
of the bis-p-bromobenzoate derivative 11 was confirmed to be
4R,5S. Based on the above evidence, macrosphelide E 1 was
found to be the C-3 stereoisomer of macrosphelide A 6.6,7
Macrosphelide F 2 was assigned a molecular formula which
contained one oxygen atom less than that of 1. The general
features of its UV, IR and NMR spectra closely resembled
those of 1 except that the proton and carbon signals of one
(C-8) of the hydroxymethines in 1 were replaced by those of a
methylene in 2 (Tables 1 and 2). Analysis of 1H–1H COSY (7-H/
8A-H and 8B-H, 9-H/8A-H and 8B-H, and 7-H/6-H) and
HMBC (7-H/C-5, 3-H/C-5 and 9-H/C-11) correlations showed
that the methylene is present at the 8-position and thus led
to the planar structure of 2. Methanolysis of 2 followed by
treatment with p-bromobenzoyl chloride gave three p-bromo-
benzoates (Scheme 2). Two of them were identical with p-
bromobenzoates 10 and 11 obtained from macrosphelide E 1.
Macrosphelide H 4 had the molecular formula C18H24O8
established by the [M ϩ H]ϩ peak in HREIMS. The general
1
features of the H and 13C NMR spectra (Table 4) of 4 closely
resembled those of 3 except that the signal for one of the methyl
groups in 3 was replaced by that of an acetonyl group [δH 3.16,
3.19 (19-H2); δC 45.24 (C-19); δC 205.78 (C-20); δH 2.22 (21-H3),
1
δC 30.57 (C-21)] in 4. H–1H COSY (3-H/19-H, 2A-H and 2B-
H) and HMBC (19-H/C-20 and 21-H/C-20) correlations (Table
4) revealed that the acetonyl group connects to the oxymethine
at C-3, and hence the 3-hydroxybutyric acid moiety in 3 is
replaced by a 3-hydroxy-5-oxohexanoic acid moiety in 4. The
position of the methylene groups in 4 was confirmed by 1H–1H
COSY (13-H/12-H, 14A-H and 14B-H, and 15-H/14A-H
and 14B-H) and HMBC (9-H/C-11, 13-H/C-11 and 15-H/C-1)
correlations. Thus, the planar structure of 4 was elucidated.
Macrosphelide I 5 was assigned a molecular formula which
contained two hydrogen atoms more than that of 1 as deduced
from a molecular-ion peak in HREIMS. The general features
of the 1H and 13C NMR spectra (Table 5) of 5 closely resembled
those of 1 except that the signals for one of the disubstituted
double bonds in 1 were replaced by those of an ethylene group
[δH 2.36, 2.63 (H-12), δC 30.41 (C-12); δH 1.46, 1.71 (H-13),
δC 27.45 (C-13)] in 5. The analysis of 1H–1H COSY (13A-H/12-
H, 13B-H and 14-H, and 14-H/15-H) and HMBC (9-H/C-11,
Scheme 2 Reaction conditions: (a) MeOH, H2SO4; (b) p-BrC6H4COCl,
pyridine.
3048
J. Chem. Soc., Perkin Trans. 1, 2001, 3046–3053

View Article Online
Table 3 1H and 13C NMR data of macrosphelide G 3 in CDCl3
a
Position
δH
J/Hz
1H–1H COSY
δC
HMBC (C)
1
2 A
B
3
5
6
7
8
9
11
12
13
14 A
B
15
17
18
19
8-OH
169 (q)b
40 (s)
2.51 dd
2.76 dd
5.2 quint d
15.7 (2B), 6.6 (3)
15.7 (2A), 3.9 (3)
6.6 (2A, 19), 3.9 (2B)
2B, 3
2A, 3
2A, 2B
1, 3, 19
1, 3, 19
1, 2A, 2B, 5, 19
67.2 (t)
165 (q)
122 (t)
145 (t)
75.3 (t)
76.4 (t)
167 (q)
123 (t)
145 (t)
38.3 (s)
6.11 dd
7.05 dd
4.36 br s
5.09 m
15.7 (7), 1.6 (8)
15.7 (6), 3.9 (8)
7
6, 8
7, 8-OH, 9
8, 18
5, 7, 8
5, 6, 8, 9
6, 7, 9, 18
7, 8, 18, 11
5.80 dt
6.80 ddd
2.39 dtd
2.50 m
5.08 m
1.26 d
1.41 d
1.16 d
3.41 br s
15.6 (13), 1.5 (14A, 14B)
15.6 (12), 9.3 (14A), 6.2 (14B)
15.2 (14B), 9.3 (13, 15), 1.5 (12)
13
11, 13, 14A, 14B
11, 12, 14B, 15
12, 13, 15, 17
12, 13, 17
1, 13, 14A, 17
14A, 14B, 15
8, 9
12, 14A, 14B
13, 14B, 15
13, 14A, 15
14A, 14B, 17
15
9
3
70 (t)
6.4 (15)
6.8 (9)
6.6 (3)
20.3 (p)
17.9 (p)
19.3 (p)
2, 3
8
a,b As in Table 1.
Table 4 1H and 13C NMR data of macrosphelide H 4 in CDCl3
a
Position
δH
J/Hz
1H–1H COSY
δC
HMBC (C)
1
2 A
B
3
5
6
7
8
9
169.22 (q)b
36.75 (s)
2.60 dd
2.93 dd
5.52 m
15.4 (2B), 3.5 (3)
15.4 (2A), 4.6 (3)
2B, 3
2A, 3
2A, 19A, 19B
1, 3, 19
1, 3, 19
1, 2A, 2B, 5, 19
66.91 (t)
165.33 (q)
121.26 (t)
145.87 (t)
75.34 (t)
6.10 dd
7.07 dd
4.38 br s
5.12 qd
15.7 (7), 1.6 (8)
15.7 (6), 3.9 (8)
7
6, 8
7, 9, 8-OH
8, 18
5, 7, 8
5, 6, 8, 9
6, 7, 9, 18
7, 8, 18, 11
6.8 (18), 1.6 (8)
77.17 (t)
11
12
13
14 A
B
15
17
18
19 A
B
167.31 (q)
123.28 (t)
145.32 (t)
38.29 (s)
5.83 dt
6.82 ddd
2.42 m
2.47 m
5.05 m
1.29 d
1.43 d
3.16 dd
3.19 dd
15.6 (13), 1.3 (14A, 14B)
15.6 (12), 8.6 (14A), 6.5 (14B)
13
11, 13, 14A,
11, 12, 14B, 15
12, 13, 15, 17
12, 13, 17
1, 13, 14A, 17
14A, 14B, 15
8, 9
12, 14A, 14B
13, 14B, 15
13, 14A, 15
14A, 14B, 17
15
70.03 (t)
20.37 (p)
17.89 (p)
45.24 (s)
6.2 (15)
6.8 (9)
18.2 (19B), 6.6 (3)
18.2 (19A), 7.3 (3)
9
3, 19B
3, 19A
2, 3, 20, 21
20
21
8-OH
205.78 (q)
30.57 (p)
2.22 s
3.56 br s
19A, 19B, 20
8
a,b As in Table 1.
13A-H/C-11 and 15-H/C-1) correlations (Table 5) clarified the
ethylene moiety in 5 to be present at C-12 and C-13, and hence
led to the planar structure of 5. In order to determine the
absolute configuration for 5, it was hydrogenated in the pres-
ence of Pd–C to yield tetrahydro derivative 16, identical with
that derived by hydrogenation from macrosphelide E 1 (Scheme
3). Based on this evidence, the absolute stereostructure for 5
was elucidated.
tical with that of 2, 7 is considered to be a stereoisomer of 2. In
order to determine the relative configuration of 7, an X-ray
crystal-structure analysis was carried out for a single crystal of
7 (obtained by recrystallization from hexane–CH2Cl2). The
result obtained (Fig. 3) allowed assignment of the relative con-
figuration of all the asymmetric centers and the conformation
for 7. The modified Mosher method was applied for determin-
ation of the absolute configuration of 7. The 1H chemical-shift
differences between the (R)- and (S)-MTPA esters 17a and
17b of 7 are shown in Fig. 4, and the result suggested the R
configuration for the asymmetric center at C-14. The above
summarized evidence led to absolute stereostructure 7 for
macrosphelide C.
Compounds 1–5 and derivative 16 were examined together
with 6, 7 and herbimycin A as standard samples in the
adhesion-assay system using HL-60 cells and HUVEC, accord-
ing to a modification of the method reported by Miki and
co-workers.21 As shown in Table 6, compounds 1–4 inhibited
the adhesion of HL-60 cell to HUVEC more potently than
herbimycin A, and compound 4 exhibited the most potent
inhibitory activity. The cytotoxic activities of compounds 1–5
Scheme 3 Reaction condition: (a) H2, Pd–C, MeOH, rt.
The planar structure of macrosphelide C 7 has already been
reported by Omura and co-workers,8 but the stereochemistry
has remained undecided. Since the planar structure of 7 is iden-
J. Chem. Soc., Perkin Trans. 1, 2001, 3046–3053
3049

View Article Online
Table 5 1H and 13C NMR data of macrosphelide I 5 in CDCl3
a
Position
δH
J/Hz
1H–1H COSY
δC
HMBC (C)
1
2 A
B
3
5
6
7
8
9
11
12 A
B
13 A
B
14
15
17
18
19
8-OH
14-OH
168.6 (q)b
41.76 (s)
2.64 dd
2.70 dd
5.65 dqd
16.3 (2B), 10.3 (3)
16.3 (2A), 3.1 (3)
10.3 (2A), 6.4 (19), 3.1 (2B)
2B, 3
2A, 3
2A, 2B
1, 3, 19
1, 3, 19
1, 2B, 5, 19
67.03 (t)
166.9 (q)
121.9 (t)
147.8 (t)
74.84 (t)
78.33 (t)
175.3 (q)
30.41 (s)
6.24 dd
7.21 dd
4.29 br s
4.86 m
15.7 (7), 1.5 (8)
15.7 (6), 4.2 (8)
7
6, 8
7, 9, 8-OH
8, 18
5, 7, 8
5, 6, 8, 9
6, 7, 9, 18
7, 8, 18, 11
2.36 dt
2.64 ddd
1.46 m
1.71 dddd
3.25 dt
4.85 m
1.18 d
1.47 d
1.32 d
3.82 br s
2.53 br s
14.9 (12B), 4.8 (13A, 13B)
14.9 (12A), 11.8 (13B), 4.8 (13A)
12B, 13A, 13B
12A, 13B
12A, 13B, 14
12B, 13A, 14
13A, 13B, 15
14, 17
15
9
3
11, 13A, 14
11, 13A, 13B
11, 12A, 12B, 15
11, 12A, 14, 15
12A, 13B, 15, 17
1, 13B, 14, 17
14, 15
27.45 (s)
16.6 (13A), 11.8 (12B), 4.8 (12A), 2.4 (14)
11.2 (13A), 2.4 (13B, 15)
71.12 (t)
74.38 (t)
18.62 (p)
12.67 (p)
20.04 (p)
6.6 (15)
6.9 (9)
6.4 (3)
8, 9
2, 3
8
a,b As in Table 1.
Table 6 Inhibitory activity of cell adhesion by macrosphelides
Compound
IC50 µ/M
Macrosphelide E 1
19.5
27.2
22.5
8.6
>100
18.7
36.5
F 2
G 3
H 4
I 5
A 6a
C 7a
Derivative 16
Herbimycin Aa
>100
38.0
a Standard samples.
spectrometer, operating at 500 and 125 MHz for H and 13C,
respectively, with TMS as internal reference. CD spectra were
recorded on JASCO J-500A and J-820 polarimeters. The other
procedures are the same as those reported previously.1
1
Fig.
3
X-Ray crystal structure for compound 7. Displacement
ellipsoids are drawn at the 40% probability level.
Culturing and isolation of metabolites
A strain of Periconia byssoides OUPS-N133 was initially iso-
lated from the sea hare Aplysia kurodai, collected in the Osaka
Bay of Japan in April, 1994. The sea hare was wiped with EtOH
and its gastrointestinal tract applied to the surface of nutrient
agar layered in a Petri dish. Serial transfers of one of the result-
ing colonies provided a pure strain of P. byssoides. The fungal
strain was grown in a liquid medium (90 dm3) containing 1%
malt extract, 1% glucose and 0.05% peptone in artificial sea-
water adjusted to pH 7.5 for four weeks at 27 ЊC. The culture
was filtered under suction and the mycelia collected were
extracted thrice with MeOH. The combined extracts were
evaporated in vacuo to give a mixture of crude metabolites
(21.5 g), the CH2Cl2–MeOH (1 : 1)-soluble fraction of which
exhibited cytotoxicity (ED50 >100 µg cmϪ3). The culture filtrate
was extracted thrice with AcOEt. The combined extracts were
evaporated in vacuo to afford a mixture of crude metabolites
(5.7 g), which exhibited cytotoxicity (ED50 4.3 µg cmϪ3). The
AcOEt extract was passed through Sephadex LH-20, using
CH2Cl2–MeOH (1 : 1) as the eluent.
Fig. 4 1H Chemical-shift differences (∆δ = δS Ϫ δR) between the (R)-
and (S)-MTPA esters 17a and 17b of macrophelide C 7.
were also examined against P388 lymphocytic leukaemia cells
and HL-60 cells in vitro. Their ED50-values were greater than
100 µg cmϪ3 against both cells except that the ED50-value of
compound 5 was 20 µg cmϪ3 against P388 cells.
Experimental
The second fraction (3.5 g), in which the activity was concen-
trated, was chromatographed on a silica gel column with a
CH2Cl2–MeOH gradient as the eluent. The MeOH–CH2Cl2
(2 : 98) and (5 : 95) eluates were collected as 2 fractions [Fr. 1
(126.6 mg) and Fr. 2 (54.1 mg)] and 2 fractions [Fr. 3 (315.7
General procedures
Mps were determined on a Yanagimoto micro-melting point
apparatus and are uncorrected. 1D and 2D NMR spectra
were recorded at 27 ЊC on a Varian UNITY INOVA-500
3050
J. Chem. Soc., Perkin Trans. 1, 2001, 3046–3053

View Article Online
mg) and Fr. 4 (254.6 mg)], respectively, all of which exhibited no
cytotoxicity. Fr. 1 was purified by HPLC using MeOH–water
(6 : 4) as the eluent to afford 2 (3.2 mg), 3 (6.7 mg), 4 (5.6 mg),
and 7 (7.2 mg). Fr. 3 was purified by HPLC using MeOH–water
(5 : 5) as the eluent to afford 1 (20.5 mg), 5 (11.1 mg) and 6
(0.8 mg).
J = 15.7, 4.0 Hz, 7-H); CD λ (c 1.66 × 10Ϫ3 mol dmϪ3 in EtOH)/
1
nm 291 (∆ε 0), 247 (Ϫ1.07), 232 (0) and 218 (ϩ3.67). H and
13C NMR data were identical with published values.6
Macrosphelide C 7. Obtained as needles, mp 156–158 ЊC
(from hexane–CH2Cl2); [α]D ϩ46.3 (c 0.54 in EtOH); νmax
(KBr)/cmϪ1 3460 (OH), 1731, 1705 (ester) and 1645 (C᎐C);
᎐
m/z (HREI) Found: [M ϩ H]ϩ, 327.1441. C16H23O7 requires
m/z, 327.1442; CD λ (c 1.66 × 10Ϫ3 mol dmϪ3 in EtOH)/nm 292
(∆ε 0), 245 (Ϫ1.83), 233 (0) and 228 (ϩ1.83). 1H and 13C NMR
data were identical with published values.8
Macrosphelide E 1. Obtained as plates, mp 105–107 ЊC (from
hexane–CH2Cl2); [α]D ϩ56.8 (c 0.46 in EtOH); λmax (EtOH)/nm
213 (log ε/dm3 molϪ1 cmϪ1 4.18); νmax (KBr)/cmϪ1 3438 (OH),
1732, 1720, 1692 (ester), 1665 and 1648 (C᎐C); m/z (EI) 343 ([M
᎐
ϩ H]ϩ, 1.4%), 215 ([C10H15O5]ϩ, 17.7), 111 ([C6H7O2]ϩ, 100), 84
([C5H8O]ϩ, 97.3) and 69 ([C4H5O]ϩ, 47.5) {m/z (HREI) Found:
[M ϩ H]ϩ, 343.1390. C16H23O8 requires m/z, 343.1392}; CD λ (c
6.75 × 10Ϫ4 mol dmϪ3 in EtOH)/nm 289 (∆ε 0), 248 (Ϫ0.59), 238
(0) and 227 (ϩ1.70). 1H and 13C NMR data are listed in Table 1.
Acetonide 12 of degradation product 9 from macrosphelide E 1
A solution (2 cm3) of 5% K2CO3 in water was added to a sol-
ution of macrosphelide E 1 (3.7 mg) in MeOH (0.05 cm3), and
the reaction mixture was left at room temperature overnight.
The mixture was neutralized with 1 M HCl and extracted with
diethyl ether, and then a solution of CH2N2 in diethyl ether
was added to the ether layer. The mixture was left for 2 h at
room temperature, and then the solvent was evaporated off
under reduced pressure to afford a mixture of crude methyl
esters (respectively 8 and 9) of 3-hydroxybutyric acid and 4,5-
dihydroxyhex-2E-enoic acid. To a solution of the mixture in
CH2Cl2 (0.2 cm3) were added 2,2-dimethoxypropane (0.3 cm3)
and pyridinum toluene-p-sulfonate (3.7 mg). After stirring
of the mixture at room temperature for 1 h, the solvent was
evaporated off under reduced pressure. The residue was puri-
fied by HPLC using MeOH–water (9 : 1) as the eluent to afford
acetonide 12 (0.9 mg) as an oil.
Macrosphelide F 2. Obtained as an oil, [α]D ϩ23.3 (c 0.09 in
EtOH); λmax (EtOH)/nm 216 (log ε/dm3 molϪ1 cmϪ1 4.15); νmax
(liquid)/cmϪ1 3445 (OH), 1720 (ester), 1658 and 1651 (C᎐C);
᎐
m/z (EI) 327 ([M ϩ H]ϩ, 0.8%), 111 ([C6H7O2]ϩ, 100), 84
([C5H8O]ϩ, 68.5) and 69 ([C4H5O]ϩ, 50.2) [m/z (HREI) Found:
[M ϩ H]ϩ, 327.1461. C16H23O7 requires m/z, 327.1442]; CD λ
(c 1.66 × 10Ϫ3 mol dmϪ3 in EtOH)/nm 291 (∆ε 0), 240 (Ϫ3.18),
1
225 (0) and 213 (ϩ8.72). H and 13C NMR data are listed in
Table 2.
Macrosphelide G 3. Obtained as an oil, [α]D ϩ66.7 (c 0.48 in
EtOH); λmax (EtOH)/nm 217 (log ε/dm3 molϪ1 cmϪ1 4.17); νmax
(liquid)/cmϪ1 3443 (OH), 1718 (ester), 1658 and 1648 (C᎐C);
᎐
m/z (EI) 326 (Mϩ, 1.8%), 111 ([C6H7O2]ϩ, 100), 84 ([C5H8O]ϩ,
87.3) and 69 ([C4H5O]ϩ, 50.5) [m/z (HREI) Found: Mϩ,
326.1362. C16H23O8 requires M, 326.1364]; CD λ (c 4.65 × 10Ϫ4
mol dmϪ3 in EtOH)/nm 308 (∆ε 0), 235 (Ϫ2.37), 219 (0) and 215
(ϩ0.92). 1H and 13C NMR data are listed in Table 3.
Acetonide 12 [methyl (4R,5S)-4,5-isopropylidenedioxyhex-
2E-enoate]. m/z (EIMS) 200 (Mϩ, 8.5%), 158 (C7H10O4, 100.0)
[m/z (HREI) Found: Mϩ, 200.1046. C10H16O4 requires M,
200.1044]; δH (CDCl3) 1.18 (3H, d, J = 6.4 Hz, 6-H3), 1.38 (3H,
s, acetonide CH3), 1.52 (3H, s, acetonide CH3), 3.76 (3H, s,
COOCH3), 4.43 (1H, quintet, J = 6.4 Hz, 5-H), 4.66 (1H, dd,
J = 6.4, 5.8 Hz, 4-H), 6.13 (1H, d, J = 15.7 Hz, 2-H) and 6.79
Macrosphelide H 4. Obtained as an oil, [α]D ϩ41.7 (c 0.22 in
EtOH); λmax (EtOH)/nm 214 (log ε/dm3 molϪ1 cmϪ1 4.19); νmax
1
(liquid)/cmϪ1 3452 (OH), 1724 (ester), 1661 and 1648 (C᎐C);
(1H, dd, J = 15.7, 5.8 Hz, 3-H). H NMR data were identical
᎐
with published values.14
m/z (EI) 369 ([M ϩ H]ϩ, 1.8%), 111 ([C6H7O2]ϩ, 100), 84
([C5H8O]ϩ, 35.3) and 69 ([C4H5O]ϩ, 10.7) [m/z (HREI) Found:
[M ϩ H]ϩ, 369.1540. C18H25O8 requires m/z, 369.1547]; CD λ
(c 6.75 × 10Ϫ4 mol dmϪ3 in EtOH)/nm 311 (∆ε 0), 239 (Ϫ1.07),
Preparation of methyl (3R)-3-( p-bromobenzoyloxy)butyrate 10
To a solution of methyl (3R)-3-hydroxybutyrate δ (8.2 mg) in
pyridine (2 cm3) was added p-bromobenzoyl chloride (6.5 mg),
and the reaction mixture was left at room temperature over-
night. The mixture was concentrated to dryness under reduced
pressure, and the residue was purified by HPLC using MeOH–
water (9 : 1) as the eluent to afford ester 10 (7.7 mg) as a pale
yellow oil, [α]D Ϫ21.6 (c 0.09 in EtOH); νmax (liquid)/cmϪ1 1725
1
224 (0) and 220 (ϩ0.54). H and 13C NMR data are listed in
Table 4.
Macrosphelide I 5. Obtained as an oil, [α]D ϩ10.3 (c 0.31 in
EtOH); νmax (EtOH)/nm 215 (log ε/dm3 molϪ1 cmϪ1 3.76); νmax
(liquid)/cmϪ1 3436 (OH), 1718 (ester) and 1650 (C᎐C); m/z (EI)
᎐
345 ([M ϩ H]ϩ, 1.1%), 217 ([C10H17O5]ϩ, 66.3), 111 ([C6H7O2]ϩ,
69.6), 84 ([C5H8O]ϩ, 100) and 69 ([C4H5O]ϩ, 62.2) [m/z (HREI)
Found: [M ϩ H]ϩ, 345.1550. C16H25O8 requires m/z, 345.1547];
CD λ (c 5.22 × 10Ϫ4 mol dmϪ3 in EtOH)/nm 295 (∆ε 0), 238
(ester), 1611 and 1593 (C᎐C); m/z (EIMS) 300 (Mϩ, 21.7%), 200
᎐
(p-BrC6H4COOH, 77.6%), 183 (p-BrC6H4CO, 100.0), 100 (Mϩ
Ϫ p-BrC6H4COOH, 12.6) [m/z (HREI) Found: Mϩ, 299.9995.
C12H13BrO4 requires M, 299.9993]; δH (CDCl3) 1.43 (3H, d,
J = 6.8 Hz, 4-H3), 2.63 (1H, dd, J = 13.3, 3.5 Hz, 2-HA), 2.78
(1H, dd, J = 13.3, 4.5 Hz, 2-HB), 3.65 (3H, s, COOCH3), 5.50
(1H, qdd, J = 6.8, 4.5, 3.5 Hz, 4-H), 7.58 (2H, ArH) and 7.88
(2H, ArH); CD λ (c 3.77 × 10Ϫ4 mol dmϪ3 in EtOH)/nm 270
(∆ε 0), 256 (ϩ0.15), 244 (0), 230 (Ϫ0.82) and 210 (0).
1
(Ϫ0.76), 224 (0) and 208 (ϩ2.69). H and 13C NMR data are
listed in Table 5.
Macrosphelide A 6. Obtained as needles, mp 139–141 ЊC
(from hexane–CH2Cl2); [α]D ϩ78.8 (c 0.57 in EtOH); νmax
(KBr)/cmϪ1 3432 (OH), 1718 (ester), 1668 and 1651 (C᎐C); m/z
᎐
(EI) 343 ([M ϩ H]ϩ, 1.8%), 215 ([C10H15O5]ϩ, 10.7), 111
([C6H7O2]ϩ, 100), 84 ([C5H8O]ϩ, 85.7) and 69 ([C4H5O]ϩ, 25.8)
[m/z (HREI) Found: [M ϩ H]ϩ, 343.1390. C16H23O8 requires
m/z, 343.1392]; δH (CDCl3) 1.33 (3H, d, J = 6.4 Hz, 19-H3), 1.37
(3H, d, J = 6.4 Hz, 17-H3), 1.46 (3H, d, J = 6.6 Hz, 18-H3), 2.57
(1H, dd, J = 15.8, 3.7 Hz, 2-HA), 2.62 (1H, dd, J = 15.8, 8.6 Hz,
2-HB), 2.85 (1H, br s, 14-OH), 3.15 (1H, br s, 8-OH), 4.14 (1H,
ddd, J = 6.2, 4.0, 1.5 Hz, 14-H), 4.23 (1H, ddd, J = 4.8, 4.0, 1.7
Hz, 8-H), 4.86 (1H, qd, J = 6.4, 6.2 Hz, 15-H), 4.96 (1H, qd, J =
6.6, 4.8 Hz, 9-H), 5.38 (1H, dqd, J = 8.6, 6.4, 3.7 Hz, 3-H), 6.04
(1H, dd, J = 15.7, 1.7 Hz, 6-H), 6.06 (1H, dd, J = 15.7, 1.5 Hz,
12-H), 6.87 (1H, dd, J = 15.7, 4.0 Hz, 13-H) and 6.88 (1H, dd,
p-Bromobenzoates 10 and 11 of degradation products from
macrosphelide E 1
To a solution of macrosphelide E 1 (7.6 mg) in MeOH (0.2 cm3)
was added conc. H2SO4 (0.01 cm3), and the reaction mixture
was left at room temperature overnight. The mixture was
diluted with water and extracted with diethyl ether, and the
extract was evaporated under reduced pressure to give a mix-
ture of crude 8 and 9. To a solution of the mixture in pyridine
(0.3 cm3) was added p-bromobenzoyl chloride (5.3 mg), and the
reaction mixture was left at room temperature overnight. The
mixture was concentrated to dryness under reduced pressure,
J. Chem. Soc., Perkin Trans. 1, 2001, 3046–3053
3051

View Article Online
and the residue was purified by HPLC using MeOH–water
(9 : 1) as the eluent to afford p-bromobenzoates 10 (1.0 mg) and
11 (4.5 mg) as a pale yellow oil. The spectral data, including CD
spectra, of 10 were identical with those of the synthetic com-
pound methyl (3R)-3-(p-bromobenzoyloxy)butyrate prepared
above.
(C᎐C), 1610 and 1593 (ArC-C); m/z (EIMS) 326 (Mϩ, 15.2%),
᎐
226 (p-BrC6H4COOH, 66.8), 183 (p-BrC6H4CO, 100.0), 126
(Mϩ Ϫ p-BrC6H4COOH, 15.3) [m/z (HREI) Found: Mϩ,
299.9995. C14H15BrO4 requires M, 299.9993]; δH (CDCl3) 1.39
(3H, d, J = 6.8 Hz, 6-H3), 2.60 (2H, dd, J = 4.8, 3.7 Hz, 4-H2),
3.73 (3H, s, COOCH3), 5.28 (1H, qt, J = 6.8, 3.7 Hz, 5-H), 5.91
(1H, d, J = 15.2 Hz, 2-H), 6.95 (1H, dt, J = 15.2, 4.8 Hz, 3-H),
7.59 (2H, ArH) and 7.88 (2H, ArH); CD λ (c 4.58 × 10Ϫ4 mol
dmϪ3 in EtOH)/nm 275 (∆ε 0) and 239 (Ϫ1.79).
p-Bromobenzoate 11 [methyl (4R,5S)-4,5-bis-( p-bromobenzo-
yloxy)hex-2-enoate]. m/z (EIMS) 524 (Mϩ, 4.7%), 324 (Mϩ Ϫ p-
BrC6H4COOH, 5.3%), 183 (p-BrC6H4CO, 100.0) [m/z (HREI)
Found: Mϩ, 324.9462. C21H18Br2O6 requires M, 326.9464]; δH
(CDCl3) 1.50 (3H, d, J = 6.8 Hz, 6-H3), 3.81 (3H, s, COOCH3),
5.54 (1H, qd, J = 6.8, 3.2 Hz, 5-H), 5.96 (1H, ddd, J = 5.3, 3.2,
1.2 Hz, 4-H), 6.20 (1H, dd, J = 15.3, 1.2 Hz, 2-H), 7.09 (1H, dd,
J = 15.3, 5.3 Hz, 3-H), 7.60 (4H, ArH) and 7.88 (4H, ArH); CD
λ (c 4.08 × 10Ϫ4 mol dmϪ3 in EtOH)/nm 275 (∆ε 0) and 240
(ϩ1.26).
p-Bromobenzoates 10, 11 and 14 of degradation products from
macrosphelide F 2
Using the same procedure as above with compound 1, a
solution of macrosphelide F 2 (3.4 mg) in MeOH (0.1 cm3) was
treated with conc. H2SO4 (0.02 cm3). A solution of the resulting
mixture of three crude esters in pyridine (0.2 cm3) was treated
with p-bromobenzoyl chloride (3.2 mg) and purified by HPLC
[MeOH–water (9 : 1)] to afford p-bromobenzoates 10 (0.4 mg),
11 (2.1 mg) and 14 (1.8 mg) as a pale yellow oil.
Formation of the (R)- and (S)-MTPA esters 13a and 13b from
macrosphelide E 1
(R)-MTPA (2.0 mg), dicyclohexylcarbodiimide (DCC) (2.0 mg)
and 4-(dimethylamino)pyridine (DMAP) (1.0 mg) were added
to a CH2Cl2 solution (0.2 cm3) of macrosphelide E 1 (0.6 mg),
and the reaction mixture was left at room temperature for 3 h.
The solvent was evaporated off under reduced pressure, and the
residue was purified by HPLC using MeOH–water (4 : 1) as the
eluent to afford (R)-MTPA ester 13a (0.7 mg). The same reac-
tion with 4 (0.7 mg) using (S)-MTPA (2.2 mg) gave ester 13b
(0.8 mg).
p-Bromobenzoate 14 [methyl (5S)-5-( p-bromobenzoyloxy)-
hex-2E-enoate]. CD λ (c 4.36 × 10Ϫ4 mol dmϪ3 in EtOH)/nm
275 (∆ε 0) and 238 (ϩ1.80). Its spectral data were identical with
those of the synthetic enantiomer 15 except that the curves of
the CD spectrum were the mirror image of those of 15.
p-Bromobenzoates 10, 11 and 14 of degradation products from
macrosphelide G 3
Using the same procedure as above with compound 1, a
solution of macrosphelide G 3 (4.6 mg) in MeOH (0.1 cm3) was
treated with conc. H2SO4 (0.02 cm3). A solution of the resulting
mixture of three esters in pyridine (0.2 cm3) was treated with
p-bromobenzoyl chloride (3.8 mg) and purified by HPLC
[MeOH–water (9 : 1)] to afford p-bromobenzoates 10 (0.7 mg),
11 (2.8 mg) and 14 (2.2 mg).
Ester 13a. Obtained as an amorphous powder; m/z (EI) 774
(Mϩ) [m/z (HREI) Found: Mϩ, 774.2114. C36H36F6O12 requires
M, 774.2112]; δH (CDCl3) 1.20 (3H, d, J = 6.6 Hz, 17-H3), 1.25
(3H, d, J = 6.6 Hz, 19-H3), 1.43 (3H, d, J = 6.6 Hz, 18-H3), 2.46
(1H, dd, J = 14.9, 6.1 Hz, 2-HA), 2.76 (1H, dd, J = 14.9, 3.0 Hz,
2-HB), 3.45 (3H, s, OMe), 3.50 (3H, s, OMe), 5.16 (1H, qdd, J =
6.6, 6.1, 3.0 Hz, 3-H), 5.23 (1H, qd, J = 6.6, 3.2 Hz, 9-H), 5.30
(1H, qd, J = 6.6, 4.3 Hz, 15-H), 5.45 (1H, ddd, J = 7.8, 4.3, 1.2
Hz, 14-H), 5.62 (1H, ddd, J = 4.8, 3.2, 1.8 Hz, 8-H), 5.95 (1H,
dd, J = 15.9, 1.2 Hz, 12-H), 5.97 (1H, dd, J = 15.9, 1.8 Hz, 6-H),
6.74 (1H, dd, J = 15.9, 7.8 Hz, 13-H), 6.91 (1H, dd, J = 15.9, 4.8
Hz, 7-H), 7.37 (6H, m, ArH) and 7.45 (4H, m, ArH).
Hydrogenation of macrosphelide E 1
To a solution of macrosphelide E 1 (2.2 mg) in MeOH (0.5 cm3)
was added 10% Pd/C (5.7 mg), and the reaction mixture was
stirred under hydrogen atmosphere (1 atm) at room temper-
ature for 2 h. The catalyst was filtered off and the solvent was
evaporated off under reduced pressure. The residue was puri-
fied by HPLC using MeOH–water (1 : 1) as the eluent to afford
compound 16 (2.0 mg) as an oil, [α]D ϩ15.0 (c 0.20, EtOH); λmax
(EtoH)/nm 215 (log ε/dm3 molϪ1 cmϪ1 3.18); νmax (liquid)/cmϪ1
3448 (OH) and 1721 (ester); m/z (EI) 346 (Mϩ, 0.1%), 215
([C10H15O5]ϩ, 45.5), 199 ([C10H15O4]ϩ, 58.4), 113 ([C6H9O2]ϩ,
100) and 85 ([C5H9O]ϩ, 42.7) [m/z (HREI) Found: Mϩ,
456.1616. C16H26O8 requires M, 456.1626]; δH (CDCl3) 1.21
(3H, d, J = 6.6 Hz, 17-H3), 1.24 (3H, d, J = 6.4 Hz, 18-H3), 1.30
(3H, d, J = 6.6 Hz, 19-H3), 1.60 (1H, m, 13-HA), 1.75 (1H, m,
13-HB), 1.77 (1H, m, 7-HA), 2.12 (1H, m, 7-HB), 2.34 (1H, m,
12-HA), 2.36 (2H, m, 6-H2), 2.52 (1H, dd, J = 14.9, 10.1 Hz,
2-HA), 2.59 (1H, dd, J = 14.9, 2.5 Hz, 2-HB), 2.68 (1H, m,
12-HB), 3.68 (1H, ddd, J = 12.9, 10.6, 4.1 Hz, 8-H), 3.76 (1H,
ddd, J = 10.9, 4.1, 1.9 Hz, 14-H), 4.82 (1H, qd, J = 6.63, 1.9 Hz,
15-H), 5.10 (1H, dq, J = 12.9, 6.4 Hz, 9-H) and 5.64 (1H, dqd,
10.1, 6.6, 2.5 Hz, 3-H); δC (CDCl3) 12.31 (C-17), 16.33 (C-18),
20.48 (C-19), 25.29 (C-7), 28.47 (C-13), 30.11 (C-6), 31.44
(C-12), 42.43 (C-2), 67.09 (C-3), 69.91 (C-14), 71.87 (C-9),
73.05 (C-8), 74.59 (C-15), 169.09 (C-1), 174.34 (C-11) and
174.94 (C-5); CD λ (c 7.23 × 10Ϫ4 mol dmϪ3 in EtOH)/nm 238
(∆ε 0) and 223 (ϩ0.90).
Ester 13b. Obtained as an amorphous powder; m/z (EI) 774
(Mϩ) [m/z (HREI) Found: Mϩ, 774.2115. C36H36F6O12 requires
M, 774.2112]; δH (CDCl3) 1.14 (3H, d, J = 6.6 Hz, 17-H3), 1.24
(3H, d, J = 6.6 Hz, 19-H3), 1.49 (3H, d, J = 6.6 Hz, 18-H3), 2.41
(1H, dd, J = 14.7, 6.1 Hz, 2-HA), 2.77 (1H, dd, J = 14.7, 3.0 Hz,
2-HB), 3.46 (3H, s, OMe), 3.49 (3H, s, OMe), 5.16 (1H, qdd, J =
6.6, 6.1, 3.0 Hz, 3-H), 5.18 (1H, qd, J = 6.6, 4.0 Hz, 15-H), 5.19
(1H, qd, J = 6.6, 3.3 Hz, 9-H), 5.40 (1H, ddd, J = 7.8, 4.0, 0.9
Hz, 14-H), 5.62 (1H, ddd, J = 5.0, 3.3, 1.6 Hz, 8-H), 5.98 (1H,
dd, J = 15.9, 0.9 Hz, 12-H), 6.13 (1H, dd, J = 15.9, 1.6 Hz, 6-H),
6.78 (1H, dd, J = 15.9, 4.0 Hz, 13-H), 6.97 (1H, dd, J = 15.9, 3.3
Hz, 7-H), 7.37 (6H, m, ArH) and 7.45 (4H, m, ArH).
Preparation of methyl (5R)-5-p-bromobenzoyloxy)hex-2E-
enoate 15
Methyl (5R)-5-(p-bromobenzoyloxy)hex-2E-enoate, [α]D Ϫ16
(c 0.18 in EtOH),18,19 was prepared from methyl (3R)-3-
hydroxyhex-2E-enoate by
a modification of the method
reported previously.19,20 To a solution of the ester (7.1 mg) in
pyridine (2 cm3) was added p-bromobenzoyl chloride (6 mg),
and the reaction mixture was left at room temperature over-
night. The mixture was concentrated to dryness under reduced
pressure, and the residue was purified by HPLC using MeOH–
water (9 : 1) as the eluent to afford methyl (5R)-5-(p-bromo-
benzoyloxyhex-2E-enoate 15 (5.7 mg) as a pale yellow oil, [α]D
Ϫ38.7 (c 0.15 in EtOH); νmax (liquid)/cmϪ1 1715 (ester), 1644
Hydrogenation of macrosphelide I 5
Using the same procedure as above with compound 1, a
solution of macrosphelide I 5 (1.7 mg) in MeOH (0.5 cm3) was
hydrogenated in the presence of 10% Pd/C (4.7 mg), and the
3052
J. Chem. Soc., Perkin Trans. 1, 2001, 3046–3053

View Article Online
resulting product was purified by HPLC [MeOH–water (1 : 1)]
96-well plate in medium 199 (Invitrogen Corp., USA) contain-
to afford 16 (1.5 mg), identical with the compound derived from
1.
ing 10% fetal calf serum (FCS, Invitrogen Corp.) and washed
with phosphate-buffered saline (PBS, DIA-IATRON Co., Ltd.)
containing 20% FCS. The HUVEC were stimulated with a
solution of lipopolysaccharides (LPS, Sigma) in RPMI 1640
medium (Invitrogen Corp.) containing 10% FCS for 4 h in the
presence of various concentrations of macrosphelides, and then
MTT-labelled HL-60 cells were added and incubated for 40 min
at 37 ЊC in 5% CO2. Unbound cells were gently washed out with
PBS containing 10% FCS, and DMSO was added to lyse the
adherent HL-60 cells. Absorbance at 540 nm was measured
using a microplate reader (Model 450, Bio-Rad).
Formation of the (R)- and (S)-MTPA esters 17a and 17b from
macrosphelide C 7
Using the same procedure as above with macrosphelide E 1,
macrosphelide C 7 (0.7 and 0.8 mg) was treated with (R)-MTPA
(2.0 mg) and (S)-MTPA (2.0 mg) to afford esters 17a (0.9 mg)
and 17b (0.7 mg), respectively.
Ester 17a. Obtained as an amorphous powder; m/z (EI) 542
(Mϩ) [m/z (HREI) Found: Mϩ, 542.1761. C26H29F3O9 requires
M, 542.1763]; δH (CDCl3) 1.30 (3H, d, J = 6.4 Hz, 17-H3), 1.32
(3H, d, J = 6.6 Hz, 19-H3), 1.37 (3H, d, J = 6.4 Hz, 18-H3), 2.35
(1H, dddd, J = 13.3, 4.7, 3.4, 1.1 Hz, 8-HA), 2.49 (1H, dd, J =
15.2, 8.1 Hz, 2-HA), 2.51 (1H, dtd, J = 13.3, 10.1, 0.9 Hz, 8-HB),
2.61 (1H, dd, J = 15.2, 3.0 Hz, 2-HB), 3.50 (3H, s, OMe), 5.11
(1H, dqd, J = 10.1, 6.4, 4.7 Hz, 9-H), 5.14 (1H, dq, J = 7.6, 6.4
Hz, 15-H), 5.27 (1H, dqd, J = 8.1, 6.4, 3.0 Hz, 3-H), 5.43 (1H,
ddd, J = 7.6, 6.3, 1.1 Hz, 14-H), 5.76 (1H, ddd, J = 15.9, 1.1, 0.9
Hz, 6-H), 5.86 (1H, dd, J = 15.9, 1.1 Hz, 12-H), 6.70 (1H, dd,
J = 15.9, 6.3 Hz, 13-H), 6.82 (1H, ddd, J = 15.9, 10.1, 3.4 Hz,
7-H), 7.42 (3H, m, ArH) and 7.47 (2H, m, ArH).
Acknowledgements
We thank Professor M. Kobayashi (Osaka University) and
Dr K. Kawai (Fuso Pharmaceutical Industries, Ltd.) for their
helpful advice on NMR spectral chemical shifts of acetonides
and cell-adhesion assay, respectively. This work was supported
in part by a Grant-in-Aid for Scientific Research from the
Ministry of Education, Science, Sports and Culture, Japan.
References
1 T. Amagata, M. Doi, T. Ohta, K. Minoura and A. Numata, J. Chem.
Soc., Perkin Trans. 1, 1998, 3585 and references cited therein.
2 T. Amagata, M. Doi, M. Tohgo, K. Minoura and A. Numata,
Chem. Commun., 1999, 1321.
3 C. Iwamoto, K. Minoura, T. Oka, T. Ohta, S. Hagishita and
A. Numata, Tetrahedron, 1999, 55, 14353.
4 A. Numata, M. Iritani, T. Yamada, K. Minoura, E. Matsumura,
T. Yamori and T. Tsuruo, Tetrahedron Lett., 1997, 38, 8215.
5 M. Hayashi, Y.-P. Kim, H. Hiraoka, M. Natori, S. Takamatsu, T.
Kawakubo, R. Masuma, K. Komiyama and S. Omura, J. Antibiot.,
1995, 48, 1435.
6 S. Takamatsu, Y.-P. Kim, M. Hayashi, H. Hiraoka, M. Natori,
K. Komiyama and S. Omura, J. Antibiot., 1996, 49, 95.
7 T. Sunazuka, T. Hirose, Y. Harigaya, S. Takamatsu, M. Hayashi,
K. Komiyama, S. Omura, P. A. Sprengler and A. B. Smith, J. Am.
Chem. Soc., 1997, 119, 10247.
8 S. Takamatsu, H. Hiraoka, Y.-P. Kim, M. Hayashi, M. Natori,
K. Komiyama and S. Omura, J. Antibiot., 1997, 50, 878.
9 A. Fukami, Y. Taniguchi, T. Nakamura, M.-C. Rho, K. Kawaguchi,
M. Hayashi, K. Komiyama and S. Omura, J. Antibiot., 1999, 52,
501.
10 Y. Iigo, T. Takashi, T. Tamatani, M. Miyasaka, T. Higashida,
H. Yagita, K. Okumura and W. Tsukuda, J. Immunol., 1991, 147,
4167.
11 D. Lauri, L. Needham, I. Martin-Padura and E. Dejana, J. Natl.
Cancer Inst., 1991, 83, 1321.
12 T. Yamada, M. Doi and A. Numata, The 50th Annual Meeting of the
Kinki Branch of the Pharmaceutical Society of Japan, Osaka,
October, 2000, Abstract p. 137.
13 H. Oikawa, I. Matsuda, T. Kagawa, A. Ichihara and K. Kohmoto,
Tetrahedron, 1994, 50, 13347.
14 S. Nagumo, I. Umezawa, J. Akiyama and H. Akita, Chem. Pharm.
Bull., 1995, 43, 171.
15 N. Harada, A. Saito, H. Ono, S. Murai, H.-Y. Li, J. Gawronski,
K. Gawronska, T. Sugioka and H. Uda, Enantiomer, 1996, 1, 119.
16 N. Harada, A. Saito, H. Ono, H. Uda, J. Gawronski and
K. Gawronska, The 33rd Symposium on the Chemistry of Natural
Products, Osaka, Japan, October 1991, Symposium Papers, p. 464.
17 I. Ohtani, T. Kusumi, Y. Kashman and H. Kakisawa, J. Am. Chem.
Soc., 1991, 113, 4092.
18 E. Hungerbühler, D. Seebach and D. Wasmuth, Helv. Chim. Acta,
1981, 64, 1467.
19 H. Tsutui and O. Mitsunobu, Tetrahedron Lett., 1984, 25, 2159.
20 P. Schnurrenberger, E. Hungerbühler and D. Seebach, Tetrahedron
Lett., 1984, 25, 2209.
21 I. Miki, N. Ishihara, M. Otoshi and H. Kase, J. Immunol. Methods,
1993, 164, 155.
22 G. M. Sheldrick, SHELXS-97, Program for the Solution of Crystal
Structures, University of Gottingen, Gottingen, 1997.
23 G. M. Sheldrick, SHELXL-97, Program for the Refinement of
Crystal Structures from Diffraction Data, University of Gottingen,
Gottingen, 1997.
Ester 17b. Obtained as an amorphous powder; m/z (EI) 542
(Mϩ) [m/z (HREI) Found: Mϩ, 542.1762. C26H29F3O9 requires
M, 542.1763]; δH (CDCl3) 1.16 (3H, d, J = 6.2 Hz, 17-H3), 1.30
(3H, d, J = 6.6 Hz, 19-H3), 1.37 (3H, d, J = 6.4 Hz, 18-H3), 2.33
(1H, dddd, J = 13.3, 4.7, 3.1, 1.2 Hz, 8-HA), 2.47 (1H, dd, J =
15.0, 8.2 Hz, 2-HA), 2.51 (1H, dtd, J = 13.3, 10.0, 1.0 Hz, 8-HB),
2.52 (1H, dd, J = 15.0, 3.0 Hz, 2-HB), 3.52 (3H, s, OMe), 5.09
(1H, dq, J = 7.6, 6.2 Hz, 15-H), 5.14 (1H, dqd, J = 10.0, 6.4, 4.7
Hz, 9-H), 5.26 (1H, dqd, J = 8.2, 6.4, 3.0 Hz, 3-H), 5.41 (1H,
ddd, J = 7.6, 6.3, 1.0 Hz, 14-H), 5.76 (1H, ddd, J = 15.9, 1.2, 1.0
Hz, 6-H), 6.0 (1H, dd, J = 15.9, 1.0 Hz, 12-H), 6.75 (1H, dd, J =
15.9, 6.3 Hz, 13-H), 6.81 (1H, ddd, J = 15.9, 10.0, 3.1 Hz, 7-H),
7.42 (3H, m, ArH) and 7.47 (2H, m, ArH).
X-Ray crystallography of macrosphelide C 7†
Macrosphelide C 7 was crystallized from hexane–CH2Cl2
solution by the vapor diffusion method. Crystal data: C16H22O7,
M = 326.34, orthorhombic, P212121, a = 10.514(2), b =
28.460(2), c = 5.860(2) Å, V = 1753.4(7) Å3, Z = 4, Dx = 1.236
Mg mϪ3, F(000) = 696, µ(Cu-Kα) = 0.817 mmϪ1. Data collec-
tion was performed by a Rigaku AFC5R using graphite-
monochromated radiation (λ = 1.5418 Å). A total of 2473
reflections were collected until θ = 67.63 Å, in which 2185 reflec-
tions were observed [I > 2σ(I )]. The crystal structure was solved
by the direct method using SHELXS-97.22 The structure was
refined by the full matrix least-squares method on F 2 using
SHELXL-97.23 In the structure refinements, non-hydrogen
atoms were refined with anisotropic temperature factors.
Hydrogen atoms were calculated on the geometrically ideal
positions by the ‘ride on’ method, and were included in the
calculation of structure factors with isotropic temperature fac-
tors. In the final stage, R = 0.0437, Rw = 0.1055 and S = 1.060
were obtained.
Cell-adhesion assay
This assay was carried out according to a modification of
Miki’s method using 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-
2H-tetrazolium bromide (MTT)-labeled cells.20 HUVEC (DIA-
IATRON Co., Ltd., Japan) were cultured until confluent in a
† CCDC reference number 169025. See http://www.rsc.org/suppdata/
p1/b1/b104337b/ for crystallographic files in .cif or other electronic
format.
J. Chem. Soc., Perkin Trans. 1, 2001, 3046–3053
3053